EP2090742A1 - Verfahren und Systeme zur Schätzung von Bohrlochereignissen - Google Patents

Verfahren und Systeme zur Schätzung von Bohrlochereignissen Download PDF

Info

Publication number
EP2090742A1
EP2090742A1 EP08002721A EP08002721A EP2090742A1 EP 2090742 A1 EP2090742 A1 EP 2090742A1 EP 08002721 A EP08002721 A EP 08002721A EP 08002721 A EP08002721 A EP 08002721A EP 2090742 A1 EP2090742 A1 EP 2090742A1
Authority
EP
European Patent Office
Prior art keywords
well
probability
measurements
network
event
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.)
Ceased
Application number
EP08002721A
Other languages
English (en)
French (fr)
Inventor
designation of the inventor has not yet been filed The
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.)
ExxonMobil Upstream Research Co
Original Assignee
ExxonMobil Upstream Research Co
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
Application filed by ExxonMobil Upstream Research Co filed Critical ExxonMobil Upstream Research Co
Priority to EP08002721A priority Critical patent/EP2090742A1/de
Publication of EP2090742A1 publication Critical patent/EP2090742A1/de
Ceased legal-status Critical Current

Links

Images

Classifications

    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B44/00—Automatic 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
    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B47/00—Survey of boreholes or wells
    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B43/00—Methods or apparatus for obtaining oil, gas, water, soluble or meltable materials or a slurry of minerals from wells

Definitions

  • This invention relates to the production of hydrocarbons from a subsurface reservoir. More specifically, the present invention provides for methods and systems for building a probability calculator and using the probability calculator to estimate or predict the status of a producing well.
  • Oil and gas fields are often metered with a wide array of measuring devices, and engineers routinely monitor production volume flow rates and other data in real time or with a short delay. Engineers monitor these measurements in order to assist in making decisions about field operations. Adverse events may severely impact the production performance of the oil and gas wells, and the accurate and timely identification of these events are the engineer's responsibility. If such events are identified in an accurate and timely manner, it is often the case that remedial action can be taken which will lower production cost, reduce downtime, improve hydrocarbon recovery, and/or increase well productivity. Avoiding adverse well production events is an important part of meeting the world's energy needs.
  • the data are viewed using automated charting tools drawing on a production database.
  • Engineers use their experience to discern wellbore related events from these time-series measurements, and then use their professional judgment to decide what action to take. For example, an engineer might see rapid variation in production volumes in combination with temperature variations on the time scale of hours. Should this pattern repeat several times, an experienced engineer suspects that the well may have liquid loading issues, and might pick up the phone to call a field engineer, asking them to invest gate and potentially change the flow control setting (e.g. choke) to alleviate the variations.
  • the flow control setting e.g. choke
  • Van Zandvoord, et al. (SPE #100342) have developed an event detection and alarming system for wells produced with electric submersible pumps as well as wells that are produced using gas lift.
  • Poulisse, et al. (SPE# 99963) describe detecting water coning and water breakthrough, two phenomena afflicting subterranean wells.
  • Lentini, et. al. (SPE# 102139) used tubing head pressure, flowline pressure, and pump amperage measurements as the basis for detecting gas-lock, low production, and slugging, events that affect electric submersible pumps.
  • the detection scheme used basic trend determination and threshold counters.
  • Jervis, et al. U.S. Pat. No. 5,952,569
  • Jervis, et al. have described a method and apparatus for the identification of adverse drilling events that include well kick, formation fluid influx, stuck drill pipe, pipe washouts, and other drilling events.
  • the method relies upon measurements from flowmeter paddles, electrochemical transducers, measurement while drilling parameter sensors, and mud tank volume sensors.
  • Bayesian networks have been used in non-real-time applications and have relied largely upon expert knowledge and not statistical learning in their construction. For example, Woronow, et al. (WO 2006/112864 ) have used Bayesian networks to predict sand quality in geological formations. Only in drilling applications have others seen utility in real-time applications of Bayesian networks so far.
  • a method for estimating the status of a producing well comprises obtaining at least one real-time well production measurement; (b) calculating at least one mathematical derivative of the at least one real-time well production measurement; (c) developing a probability calculator based on the at least one mathematical derivative; (d) calculating a status of the producing well utilizing the probability calculator; and (e) displaying the status of the producing well for a user.
  • the method may further include obtaining and using facilities data in the probability calculator and may also include updating the mathematical derivatives using updated well production data and calculating and displaying an updated status of the producing well for the user using the updated probability calculator.
  • a method for well event detection includes developing a probability network; measuring at least one physical property of at least one well to generate well measurements; analyzing the well measurements, wherein analyzing the well measurements includes at least calculating mathematical derivatives of the well measurements; transferring at least the derivatives of the well measurements to the probability network; calculating the probability of an event utilizing at least the probability network and the derivatives of the well measurements; and displaying the probability of the event for a user.
  • the method may also incorporate facilities data, and may run in a predictive or recognition mode. Further, the method may include utilizing updated data to provide an updated probability of the event for the user.
  • a system for estimating the status of a producing well comprises a database configured to store at least one well production measurement; a processor configured to determine a mathematical derivative of the at least one well production measurement; a probability network in communication with the processor and the database, wherein the probability network is configured to calculate the status of the producing well utilizing at least the mathematical derivative of the at least one well production measurement; and a display configured to display the status of the producing well calculated by the probability network.
  • Figure 1 is an exemplary process diagram of one method of estimating the status of a producing well
  • FIG. 2 is an illustrative diagram of one approach to developing a probability calculator as shown in FIG. 1 ;
  • FIG. 3 is an illustration of a schematic of an exemplary event detection system utilizing the probability network of FIG. 2 ;
  • FIG. 4 is an illustration of an exemplary sample Bayesian network of FIGs. 2 and 3 developed to report on the status of wells in a gas field, including possible events;
  • FIG. 5 is an illustration of an exemplary sample display of a Bayesian network-determined most probable well event for a group of wells.
  • the term "facilities data” refers generally to values that do not change continuously over time. For example, the state of the choke on a flowline remains fixed in one position and does not change until it is fixed in a different position by an operator. These kinds of data are discrete. In this embodiment, these "facilities data" are generally related to the fluid gathering system, such as the diameter of the tubing string, choke state, and other system operational limits or setpoints.
  • the terms "derivative” or “mathematical derivative” are used interchangeably with “characteristic mathematical quantities” and "preferred mathematical measures.” These quantities are computed from time-dependent variables and change with time. It is typical that such measures would be computed over a window of time. For example, a seven day running average is an example of a mathematical derivative of a time series variable. The term is not limited to its definition from the mathematical field of calculus.
  • the present invention provides for a method and system for building a probability calculator and using the probability calculator to estimate or predict the status of a producing well. More specifically, the estimation method includes taking time-based measurements and system state measurements (these are generally static or non-time based), then inputting the measurements into a probability calculator (which may be a Bayesian network) to determine a likely status for the producing well, which may be displayed as a probability or likelihood of an event occurring.
  • a probability calculator which may be a Bayesian network
  • the analysis of downhole and surface measurements during production is automated (e.g. pressure, rate, temperature measurements, and others) in order to rapidly identify adverse production system events.
  • Such analysis may be automated by statistically characterizing expert judgment of the events, and by linking through hardware and software interfaces directly to the measurements.
  • FIG. 1 is an exemplary process diagram of one method of estimating the status of a producing well.
  • the method 100 begins at 102 and includes obtaining a well production measurement (e.g. a real-time well production measurement or a historic well production measurement) 103 , calculating a mathematical derivative based on the well production measurement 104 , and developing a probability calculator (e.g. probability network) 106 based on the derivative.
  • the process 100 includes calculating the status of the well 108, and displaying the well status 110.
  • the process 100 may continue by gathering updated well measurement data and calculating updated derivatives 112, then repeating steps 108 and 110 to show an updated list of probabilities in real-time or approximately real-time.
  • the well status may be calculated using the probability calculator developed in step 106 .
  • the probability calculator may be developed 106 using well production measurements 103 such as pressure, temperature, and production rate in the well. These well production measurements 103 may also be referred to as "surveillance variables.” Some exemplary equipment related measurements include pump rotational speed and amperage. Pressure measurements may be recorded by dedicated downhole devices, associated pump intake and outtake pressure measurement devices, or surface flowline pressure measurements. Temperature measurements may be recorded, for example, by dedicated downhole or surface thermometers. Rate measurements may be recorded from single fluid phase or multiphase fluid wellhead or downstream flow meters.
  • the probability calculator may alternatively be developed 106 by selecting a field for which there exists abundant pressure and rate data of high quality, compiling a list of all of the possible states ("events") that the wells in the selected field may experience, gathering the historical data for the wells of interest and reviewing the historical data with the engineers ordinarily responsible for operation of those wells. The engineer(s) then assigns their best judgment of the status of a well for each hour in the time period in review. It is preferable that this historical set of data is representative of the full range of behaviors exhibited by the wells.
  • the process 100 may further include computing mathematical derivatives 104 of the well production measurements 103 that are likely to be indicative of the events of interest. Using the mathematical theorem of Bayes or any other mathematically related algorithm that computes joint probabilities to mathematically relate the mathematical derivatives of the measured historical signals to the expert identified events. Commercially available software that does the mathematics is available.
  • the probability calculator developed in step 106 should be integrated into the method of estimating the status of a producing well 100 in such a manner that the same measurement types 103 that were used to create the probability calculator are transmitted to the mathematical derivative calculator in a real-time fashion.
  • the mathematical derivative calculator may transmit its output to the probability calculator in a real-time fashion.
  • An alarm system may be designed that can take advantage of the probabilistic nature of the probability network and is compatible with how engineers respond to field problems.
  • the probability calculator may be updated and the alarm system adjusted as engineers gain experience with the system.
  • the production method also controls which well production measurements are of interest.
  • the most important measurements for a particular well or field may be referred to as the "key performance indicators.”
  • injection rates and water production rates may be the key performance indicators.
  • the key performance indicators for a cold flow heavy oil production process may include cavity pump speed, and other variables.
  • Drilling measurements may include flow into and out of the well rather than production through a production string. Drilling measurements may also include a stand-pipe pressure transducer (SPT) for monitoring mud tank volumes. Mud tank volumes, for example, are irrelevant in production operations.
  • SPT stand-pipe pressure transducer
  • the probability calculator developed at 106 may utilize Bayes' theorem or another method of calculating probabilities such as the Hidden Markov Model or other such probability calculation method.
  • the well measurement derivative at 104 may be time derivatives, sums, percent change, averages, day over day change, day over week change, day over month change, week over month change, standard deviation, and counting the frequency of repeated values, and other similar mathematical derivations or combinations of the measured signals.
  • FIG. 2 is an illustrative diagram of one approach to developing a probability calculator as shown in FIG. 1 .
  • the probability calculator diagram 200 includes historical measurements from similar wells 202 and facilities data (e.g. discreet or system state data) 204, which is used to develop an engineering judgment of the data 206.
  • the historical measurements 202 are also used in a computation of characteristic mathematical quantities 208 .
  • the facilities data 204, and engineering judgments 206 are combined to form a static or relative data set 210. This static data set 210 is then combined with the computational data set 208 to form the probability calculator 212 .
  • data set 206 may include asking surveillance engineers to identify events of interest in historical sets of measured time-series variables 202 using simple charts, their personal knowledge of the well and static (non time-series) facility data 204 (choke state, flowline sizes, prevailing compressor conditions) and no mathematical aides. Separately, simple mathematical functions such as standard deviation, day over day percent change, and other measures are calculated for several variables over the time range of interest 208 . In this manner, the probability network 212 probabilistically associates with each event a particular characteristic mathematical signature. The structure of the probability network 212 is defined by a set of variables and the probabilistic relationships between them. Because the particular set of measured variables and events of interest may differ from field-to-field, the probability calculator building process 200 must be completed separately for each field (though not necessarily for each wellbore).
  • the probability network 212 functions according to Bayes' theorem (in which case the network 212 may be called a Bayesian network 212 or BN 212 ).
  • Bayes' theorem is an efficient mathematical framework for calculating probabilities according to Bayesian probability. The theorem is stated as:
  • the theorem may be stated in English as the conditional probability of event A, given event B, is equal to the conditional probability of event B, given event A, times the prior probability of event A, normalized by the prior probability of event B.
  • the Bayesian network 212 may be incorporated into engineering work procedures by integrating it into surveillance software that is coupled to databases where real-time measurements are stored.
  • FIG. 3 is an illustration of a schematic of an exemplary eve it detection system utilizing the probability network of FIG. 2 .
  • the event detection system 300 (which may run in a real time mode) includes time series measurements 302, which are transferred to a database 304 , such as a Supervisory Control And Database (SCADA) frequently enough to be useful to the surveillance engineer (e.g. daily frequency or higher).
  • SCADA Supervisory Control And Database
  • a processor or computer computes characteristic mathematical quantities 306 , which are used in combination with system state (e.g. facilities) data 308 in a probability calculator 310 (which may be the probability calculator 212 ) to produce estimated event probabilities 312 .
  • the event probabilities 312 and stored user settings 314 are fed into an alarm and prioritization calculator 316.
  • the results from the alarm and prioritization calculator 316 are then sent to a display 318, such as a graphic user interface (GUI), and storage 320 for archiving and later reference.
  • GUI graphic user interface
  • the time series measurements 302 may include at least pressure, temperature, rate, and other equipment-related measurements.
  • potential equipment related measurements include pump rotational speed and amperage.
  • Pressure measurements may be recorded by dedicated downhole devices, associated pump intake and outtake pressure measurement devices, or surface flowline pressure measurements.
  • Temperature measurements may be recorded by dedicated downhole or surface thermometers, for example.
  • Rate measurements may be recorded from single fluid phase or multiphase fluid wellhead or downstream flow meters.
  • the production method also controls which variables are of interest. For example if the field is producing using water injection, injection rates and water production rates will be key for understanding events in injection wells. All of these measurements are relevant to production surveillance. Measurements that are useful for the diagnosis of drilling events are largely different than those useful for the diagnosis of production events.
  • the database 304 may be a SCADA and may be saved on a centralized server as part of a network, may be remotely accessible through wired or wireless networks, may be password protected or encrypted data, redundantly available, or may be local or located remote from the field. The recorded measurements are then transferred to a centralized server and analyzed. In one embodiment, the database 304 may be accessible worldwide, located in Houston, Texas, and populated by data transfer from field location compilers or similar devices.
  • the computed characteristic mathematical quantities 306 may include mathematical derivatives of the recorded measurements from the database 304 . In one exemplary embodiment, these mathematical derivatives are looked at before absolute values are assessed because adverse events typically may be best identified by looking for changes in process parameters. Some exemplary calculations 306 include day over day change, day over week change, day over a month change, week over month change, and standard deviation of all process parameters computed on an absolute or percentage basis. Additional calculations 306 may include counting the frequency of repeated values. Some prior art methods for predicting drilling events calculate residuals left after subtraction of the signal data from a stored database of representative model event data, which is more suited :o looking for small changes and searching high frequency data sets. In production applications, such as the one described in the present invention, small changes may not be as significant as in drilling applications.
  • facilities data 308 may include any and all data which is not time-series data, and is related to the system or facility state. Such data may include, for example, pump depth, choke type, flowline diameter, and well coordinates.
  • the probability calculator 310 is a Bayesian network that computes, for all of the possible events that are known to it, the probability that the analyzed time series measurements 306 indicate that the event occurred during that time window 312. Whereas a threshold-based method provides a simple pronouncement that the event did or did not occur, the probability calculator 310, taking into account all available information and prior knowledge, provides the rank order of the most probable events, and the associated probabilities for each one. This helps engineers prioritize their decisions. A well with a high probability of a particularly adverse event, such as the shutdown of a well, will receive attention more quickly than a well with a lower probability of well shutdown, which might in turn receive attention more quickly than a well with a high probability of a minor event. The probabilities themselves can serve as the alarm function 316. This is a recognition that not all alarms are of equal importance.
  • the stored user settings 314 may be set based at least in part on the historic performance of related wells 202 , facilities data 308 , and engineering judgment 206 . It could also be a default setting or group of settings. These settings could be manually or automatically updated by an engineer locally or remotely.
  • the alarm and prioritization calculator 316 may list the events according to their calculated probability 312 and the thresholds and other information from the user settings 314 .
  • the display of the event prioritized probabilities 318 may be on any type of display including, for example, a computer screen, a GUI, a personal digital assistant or phone or other handheld device and may display a single alarm in certain circumstances, a list of probabilities, or some combination depending on settings 314 .
  • the storage 320 may be local or remote, network connected to a central server or database or redundantly available on multiple locations.
  • the system 300 can be programmed by the user or the creator to interpret the probabilities 312 in an automated fashion.
  • the system may be designed to send alarms, with their associated event probabilities, to system users over a display or other device 318 . If users elect not to receive alarms in real time, the probabilistic analysis may be stored, and retrieved later for further use.
  • rate measurements 302 and their mathematical derivatives 306 may be the primary inputs, with pressure and temperature measurements playing a secondary role.
  • an electronic flowmeter of high precision and highfrequency installed at the wellhead may transmit the recorded fluid flow rates of all flowing phases to the SCADA 304. Downhole and wellhead pressures may also be transmitted, and temperature since it is usually available when pressure measurements are available. If downhole equipment is installed, such as an electric submersible pump or a progressing cavity pump, the preferred implementation should include the transmission of amperage and rotational speed.
  • the preferred mathematical measures 306 depend on the time scale of the events of interest. If daily operational events are of interest then the day-over-day change in the oil or gas flowrate will be significant in detecting events. Many physical phenomena cause instabilities in measurements. For example, the process of the liquid load up of gas wells includes a phase in which the gas flowrate of the well is unstable (sometimes called metastable). The standard deviation of a set of measurements recorded over a particular time scale closely corresponding to the time scale over which the instability occurs may be an advantageous mathematical derivative 306 for enabling a robust detection method of the instability.
  • probability calculator 212 is the preferred probability calculator 310 in the event detection scheme 300, other implementations are also possible. Not all of the variables used in probability calculator 310 are equally important for the interpretation of the events. It is natural that some variables contribute strongly and others less so to the quality of the interpretation.
  • One advantage of the invention over prior inventions is that probability calculator 310 will allow the engineer to precisely calculate which variables are most important to the well or group of wells.
  • One exemplary implementation of the invention is to choose the subset of variables that contribute most strongly and build a decision tree based on those variables. The decision tree can be used as a simplified event detector instead of the full-fledged probability calculator 310 .
  • the advantage of this less robust alternative implementation is that it is often easier to incorporate this method into existing software systems than to incorporate a true probability calculator 310 .
  • the alarming system 316 is tunable by the user.
  • the user might specify that they want to see all alarms that indicate a 90% or higher chance of an unexplained well shutdown.
  • the calculated event probabilities 312 are stored to disk 320 for all measurement history. This enables an engineer to examine the alarm h.story of a well. A rising probability 312 for an event over time indicates greater and greater confidence that the interpretation is true.
  • One desirable method to filter the alarm data may be to only display alarms for less detrimental events if the probability of the event has been high for a sustained period of time.
  • an operator may have access to contextual data that is not captured in the event detection system 300 , despite the inclusion of as much facilities data 308 as possible when the probability calculator 310 is constructed.
  • the invention provides the probabilities of all possible events in a user-friendly format, and is configured such that the user may access the basis for calculating the probabilities 312. This allows the user to interpret how to respond based on any special contextual information that they may possess. This interface allows for the user to provide feedback so that the event detection system 300 can be improved, and its estimated probabilities 312 made more accurate.
  • FIG. 4 is an illustration of an exemplary sample Bayesian network of FIGs. 2 and 3 developed to report on the status of wells in a gas field, including possible events. As such, FIG. 4 may be best understood with reference to FIGs. 2 and 3 .
  • the Bayesian network (BN) 400 may be the probability calculator or network 12 or 310.
  • Five mathematical measures 404, 406, 410, 412 , and 414 are used to interpret the status of the well during the last 24 hours of production.
  • All five exemplary characteristic mathematical quantities, SDLW 404 , ZRLW 406, SD24 414 , ZR24 410, and AV24 412 may be characteristic mathematical quantities 208 derived from historical measurements 202 and used to build the BN 400 by the process 200 . The same quantities may also utilize the BN 400 in the process 300.
  • the BN 400 gives its most accurate probability estimate 402 (which may be the estimated event probabilities 312 as shown on display 318 ) of well status during the last 24 hours.
  • the network is operating in a "recognition mode," estimating the well status during the previous 24 hours.
  • the BN 400 can make a prediction about the next (future) 24-hour period. In this manner the BN 400 operates in a "predictive mode.”
  • a recognition mode is more accurate because it utilizes more data.
  • the possible events 402 are the three different phases of liquid loading -- metastable, loading, and unloading -- as well as unexplained well shutdowns, field shutdowns, fixed choke changes, and adjustable choke changes.
  • the event summary 408 simply groups certain events together to give the engineer an additional tool for monitoring and evaluating well events.
  • FIG. 5 is an illustration of an exemplary sample display of a Bayesian network-determined most probable well event for a group of wells.
  • the Bayesian network may be probability calculator 212 or 310 and the display may be display 318 .
  • FIG. 5 may be best understood with reference to FIGs. 2 and 3 .
  • the display 500 includes a column 502 labeled "Most Probable Well Status" (actual probability not shown) for each of the listed wells in column 503, which may all be associated gas wells in a single field.
  • the engineer would look at charts of the flow rate over time for all wells with production drops in an effort to determine the cause(s).
  • the surveillance engineer can use the display 500 to prioritize which wells 503 to examine in more detail, skipping the wells for which the BN-determined event is known not to be detrimental (e.g. where the value in column 508 is 24 and the statement in 502 is null).

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Geology (AREA)
  • Mining & Mineral Resources (AREA)
  • Physics & Mathematics (AREA)
  • Environmental & Geological Engineering (AREA)
  • Fluid Mechanics (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Geochemistry & Mineralogy (AREA)
  • Geophysics (AREA)
  • Testing And Monitoring For Control Systems (AREA)
EP08002721A 2008-02-14 2008-02-14 Verfahren und Systeme zur Schätzung von Bohrlochereignissen Ceased EP2090742A1 (de)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP08002721A EP2090742A1 (de) 2008-02-14 2008-02-14 Verfahren und Systeme zur Schätzung von Bohrlochereignissen

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
EP08002721A EP2090742A1 (de) 2008-02-14 2008-02-14 Verfahren und Systeme zur Schätzung von Bohrlochereignissen

Publications (1)

Publication Number Publication Date
EP2090742A1 true EP2090742A1 (de) 2009-08-19

Family

ID=39485090

Family Applications (1)

Application Number Title Priority Date Filing Date
EP08002721A Ceased EP2090742A1 (de) 2008-02-14 2008-02-14 Verfahren und Systeme zur Schätzung von Bohrlochereignissen

Country Status (1)

Country Link
EP (1) EP2090742A1 (de)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140124265A1 (en) * 2012-11-02 2014-05-08 Saudi Arabian Oil Company Systems and methods for expert systems for underbalanced drilling operations using bayesian decision networks
WO2014070962A3 (en) * 2012-11-02 2014-09-04 Saudi Arabian Oil Company Systems and methods for expert systems for well completion using bayesian decision networks
WO2014070823A3 (en) * 2012-11-02 2014-11-06 Saudi Arabian Oil Company Systems and methods for an expert system for well control using bayesian intelligence
WO2014070967A3 (en) * 2012-11-02 2014-11-13 Saudi Arabian Oil Company Systems and methods for drilling fluids expert systems using bayesian decision networks
WO2021029858A1 (en) * 2019-08-09 2021-02-18 Landmark Graphics Corporation Model based preference learning and optimization systems and methods
WO2022071980A1 (en) * 2020-09-30 2022-04-07 Cleaver-Brooks, Inc. System and method for predicting shutdown alarms in boiler using machine learning
US11396804B2 (en) 2018-08-30 2022-07-26 Landmark Graphics Corporation Automated rate of penetration optimization for drilling
US20220235628A1 (en) * 2021-01-28 2022-07-28 Saudi Arabian Oil Company Controlling fluid flow through a wellbore tubular

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5072387A (en) * 1989-12-20 1991-12-10 Chevron Research And Technology Company Method for determining a transit time for a radioactive tracer
GB2318437A (en) * 1996-10-21 1998-04-22 Anadrill Int Sa Alarm system for wellbore site
US6732052B2 (en) 2000-09-29 2004-05-04 Baker Hughes Incorporated Method and apparatus for prediction control in drilling dynamics using neural networks
US6820702B2 (en) 2002-08-27 2004-11-23 Noble Drilling Services Inc. Automated method and system for recognizing well control events
WO2006112864A2 (en) 2004-07-07 2006-10-26 Exxonmobil Upstream Research Company Bayesian network applications to geology and geophysics
US7128167B2 (en) 2002-12-27 2006-10-31 Schlumberger Technology Corporation System and method for rig state detection
US20070175633A1 (en) 2006-01-30 2007-08-02 Schlumberger Technology Corporation System and Method for Remote Real-Time Surveillance and Control of Pumped Wells
US20070276603A1 (en) * 2003-11-05 2007-11-29 Abb As Detection of Water Breakthrough

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5072387A (en) * 1989-12-20 1991-12-10 Chevron Research And Technology Company Method for determining a transit time for a radioactive tracer
GB2318437A (en) * 1996-10-21 1998-04-22 Anadrill Int Sa Alarm system for wellbore site
US5952569A (en) 1996-10-21 1999-09-14 Schlumberger Technology Corporation Alarm system for wellbore site
US6732052B2 (en) 2000-09-29 2004-05-04 Baker Hughes Incorporated Method and apparatus for prediction control in drilling dynamics using neural networks
US6820702B2 (en) 2002-08-27 2004-11-23 Noble Drilling Services Inc. Automated method and system for recognizing well control events
US7128167B2 (en) 2002-12-27 2006-10-31 Schlumberger Technology Corporation System and method for rig state detection
US20070276603A1 (en) * 2003-11-05 2007-11-29 Abb As Detection of Water Breakthrough
WO2006112864A2 (en) 2004-07-07 2006-10-26 Exxonmobil Upstream Research Company Bayesian network applications to geology and geophysics
US20070175633A1 (en) 2006-01-30 2007-08-02 Schlumberger Technology Corporation System and Method for Remote Real-Time Surveillance and Control of Pumped Wells

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9366129B2 (en) 2012-11-02 2016-06-14 Saudi Arabian Oil Company Expert systems for well completion using bayesian decision networks with a multilateral junction design, and a junction classification decision node
US9202175B2 (en) 2012-11-02 2015-12-01 The Texas A&M University System Systems and methods for an expert system for well control using Bayesian intelligence
US20140124265A1 (en) * 2012-11-02 2014-05-08 Saudi Arabian Oil Company Systems and methods for expert systems for underbalanced drilling operations using bayesian decision networks
WO2014070823A3 (en) * 2012-11-02 2014-11-06 Saudi Arabian Oil Company Systems and methods for an expert system for well control using bayesian intelligence
WO2014070967A3 (en) * 2012-11-02 2014-11-13 Saudi Arabian Oil Company Systems and methods for drilling fluids expert systems using bayesian decision networks
US9140112B2 (en) 2012-11-02 2015-09-22 Saudi Arabian Oil Company Systems and methods for expert systems for well completion using Bayesian decision models (BDNs), drilling fluids types, and well types
US9202169B2 (en) 2012-11-02 2015-12-01 Saudi Arabian Oil Company Systems and methods for drilling fluids expert systems using bayesian decision networks
US9366130B2 (en) 2012-11-02 2016-06-14 Saudi Arabian Oil Company Expert system for well completion using bayesian probabilities and a consequences node dependent on the zonal isolation types, reliability level, cost level, productivity level, the completion type, and the junction classification decision nodes
WO2014070960A3 (en) * 2012-11-02 2014-09-18 Saudi Arabian Oil Company Systems and methods for expert systems for underbalanced drilling operations using bayesian decision networks
WO2014070962A3 (en) * 2012-11-02 2014-09-04 Saudi Arabian Oil Company Systems and methods for expert systems for well completion using bayesian decision networks
US9376905B2 (en) 2012-11-02 2016-06-28 Saudi Arabian Oil Company Expert systems for well completion using Bayesian probabilities, open hole gravel pack types, gravel pack design details, open hole gravel packs, and completion types
US9371726B2 (en) 2012-11-02 2016-06-21 Saudi Arabian Oil Company Expert systems for well completion using Bayesian probabilities, and a packers consequence node dependent on wellbore fluids, hydrocarbon types, completion fluids, packers decision, and treatment fluids nodes
US9382791B2 (en) 2012-11-02 2016-07-05 Saudi Arabian Oil Company Expert systems for well completion using bayesian decision networks including a fluids damage and temperature effects uncertainty node, a perforation considerations uncertainty node, a perforation analysis uncertainty node, and a perforation type decision node
US11396804B2 (en) 2018-08-30 2022-07-26 Landmark Graphics Corporation Automated rate of penetration optimization for drilling
GB2599835B (en) * 2019-08-09 2023-05-03 Landmark Graphics Corp Model based preference learning and optimization systems and methods
GB2599835A (en) * 2019-08-09 2022-04-13 Landmark Graphics Corp Model based preference learning and optimization systems and methods
WO2021029858A1 (en) * 2019-08-09 2021-02-18 Landmark Graphics Corporation Model based preference learning and optimization systems and methods
WO2022071980A1 (en) * 2020-09-30 2022-04-07 Cleaver-Brooks, Inc. System and method for predicting shutdown alarms in boiler using machine learning
US20220235628A1 (en) * 2021-01-28 2022-07-28 Saudi Arabian Oil Company Controlling fluid flow through a wellbore tubular

Similar Documents

Publication Publication Date Title
CA2703857C (en) Methods and systems to estimate wellbore events
EP2090742A1 (de) Verfahren und Systeme zur Schätzung von Bohrlochereignissen
AU2013296744B2 (en) Monitoring, diagnosing and optimizing electric submersible pump operations
US5952569A (en) Alarm system for wellbore site
CA2492608C (en) System and method for obtaining and analyzing well data
RU2723805C9 (ru) Способ и компьютерная система управления бурением скважин
EP2519843B1 (de) Verwendung allgemeiner bayesscher netzwerke in ölfeldoperationen
CA2994479C (en) Method and system for detecting at least one of an influx event and a loss event during well drilling
Gupta et al. Applying big data analytics to detect, diagnose, and prevent impending failures in electric submersible pumps
US20140278302A1 (en) Computer-implemented method, a device, and a computer-readable medium for data-driven modeling of oil, gas, and water
US20190093468A1 (en) Real time measurement of mud properties for optimization of drilling parameters
AU2013296908B2 (en) Monitoring and diagnosing water flooded reservoirs
CN105408914A (zh) 流体回流预测
CA3115224A1 (en) Oilfield system
Wang et al. Time Series Data Analysis with Recurrent Neural Network for Early Kick Detection
US20220195861A1 (en) Identifying operation anomalies of subterranean drilling equipment
Lafond et al. Detecting Pressure Anomalies While Drilling Using a Machine Learning Hybrid Approach
US20230383633A1 (en) Machine-learned based real-time virtual gas metering
Zhang et al. Intelligent kick detection using a parameter adaptive neural network
Mahmoud et al. Application of Artificial Neural Networks in Predicting Discharge Pressures of Electrical Submersible Pumps for Performance Optimization and Failure Prevention
Shi et al. Novel Kick Detection Tool Based on Knowledge-Data-Dual-Driven Approaches
Lentini et al. A New Approach for Field Surveillance: Enhanced Monitoring Associating Smart Alarms to Produce More Oil

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MT NL NO PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA MK RS

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION HAS BEEN REFUSED

18R Application refused

Effective date: 20090911