The data acquisition and control interface 94 receives indications e. In a slightly overbalanced managed pressure drilling operation, the setpoint pressure would likely increase, due to the reduced equivalent circulating density, in which case flow resistance through the choke 34 would be increased in response. However, in some operations such as, underbalanced drilling operations in which gas or another light weight fluid is added to the drilling fluid 18 to decrease bottom hole pressure , the setpoint pressure could decrease e.
In step , the restriction to flow of the fluid 18 through the choke 34 is changed, due to the changed desired annulus pressure in step As discussed above, the controller 96 controls operation of the choke 34 , in this case changing the restriction to flow through the choke to obtain the changed setpoint pressure.
- Managed Pressure Drilling (MPD) Services | Schlumberger.
- PEH:Well Control: Procedures and Principles;
- Exercises in Pressure Control During Drilling;
- Educating Israel: Educational Entrepreneurship in Israel’s Multicultural Society!
- Pressure Control.
- The Lighthouse (Inspector Adam Dalgliesh Book 13).
Also as discussed above, the setpoint pressure could increase or decrease. Steps , and are depicted in the FIG. In step , the bypass flow control device 74 gradually opens. This diverts a gradually increasing proportion of the fluid 18 to flow through the bypass line 72 , instead of through the standpipe line In step , the setpoint pressure changes due to the reduced flow of the fluid 18 through the drill string 16 e.
Flow through the drill string 16 is substantially reduced when the bypass flow control device 74 is opened, since the bypass line 72 becomes the path of least resistance to flow and, therefore, fluid 18 flows through bypass line In a slightly overbalanced managed pressure drilling operation, the setpoint pressure would likely increase, due to the reduced equivalent circulating density, in which case flow restriction through the choke 34 would be increased in response.
However, these steps could be performed non-concurrently in other examples. In step , the pressures in the standpipe line 26 and the annulus 20 at or near the surface indicated by sensors 36 , 38 , 40 , 44 equalize. At this point, the bypass flow control device 74 should be fully open, and substantially all of the fluid 18 is flowing through the bypass line 72 , 75 and not through the standpipe line 26 since the bypass line represents the path of least resistance.
Static pressure in the standpipe line 26 should substantially equalize with pressure in the lines 30 , 73 , 75 upstream of the choke manifold In step , the standpipe flow control device 81 is closed. The separate standpipe bypass flow control device 78 should already be closed, in which case only the valve 76 would be closed in step In step , a standpipe bleed valve 82 see FIG. At this point, the standpipe line 26 is vented to atmosphere. In step , the kelley or top drive is disconnected from the drill string 16 , another stand of drill pipe is connected to the drill string, and the kelley or top drive is connected to the top of the drill string.
Pressure Control During Oil Well Drilling
This step is performed in accordance with conventional drilling practice, with at least one exception, in that it is conventional drilling practice to turn the rig pumps off while making a connection. In the method , however, the rig pumps 68 preferably remain on, but the standpipe valve 76 is closed and all flow is diverted to the choke manifold 32 for annulus pressure control.
Non-return valve 21 prevents flow upward through the drill string 16 while making a connection with the rig pumps 68 on. In step , the standpipe bleed valve 82 is closed. The standpipe line 26 is, thus, isolated again from atmosphere, but the standpipe line and the newly added stand of drill pipe are substantially empty i.
In step , the standpipe bypass flow control device 78 opens in the case of the valve and flow restrictor configuration of FIG. In this manner, the fluid 18 is allowed to fill the standpipe line 26 and the newly added stand of drill pipe, as indicated in step Eventually, the pressure in the standpipe line 26 will equalize with the pressure in the annulus 20 at or near the surface, as indicated in step However, substantially all of the fluid 18 will still flow through the bypass line 72 at this point. In step , the standpipe flow control device 76 is opened in preparation for diverting flow of the fluid 18 to the standpipe line 26 and thence through the drill string The standpipe bypass flow control device 78 is then closed.
Note that, by previously filling the standpipe line 26 and drill string 16 , and equalizing pressures between the standpipe line and the annulus 20 , the step of opening the standpipe flow control device 76 does not cause any significant undesirable pressure transients in the annulus or mud return lines 30 , Substantially all of the fluid 18 still flows through the bypass line 72 , instead of through the standpipe line 26 , even though the standpipe flow control device 76 is opened. Considering the separate standpipe flow control devices 76 , 78 as a single standpipe flow control device 81 , then the flow control device 81 is gradually opened to slowly fill the standpipe line 26 and drill string 16 , and then fully opened when pressures in the standpipe line and annulus 20 are substantially equalized.
In step , the bypass flow control device 74 is gradually closed, thereby diverting an increasingly greater proportion of the fluid 18 to flow through the standpipe line 26 and drill string 16 , instead of through the bypass line During this step, circulation of the fluid 18 begins through the drill string 16 and wellbore In step , the setpoint pressure changes due to the flow of the fluid 18 through the drill string 16 and annulus 20 e. The desired annulus pressure may either increase or decrease, as discussed above for steps and In step , the flow rate output from the pump 68 may be increased in preparation for resuming drilling of the wellbore This increased flow rate maintains the choke 34 in its optimum operating range, but this step as with step discussed above may not be used if the choke is otherwise maintained in its optimum operating range.
In step , the setpoint pressure changes due to the increased flow of the fluid 18 e. In a slightly overbalanced managed pressure drilling operation, the setpoint pressure would likely decrease, due to the increased equivalent circulating density, in which case flow restriction through the choke 34 would be decreased in response.
In step , drilling of the wellbore 12 resumes. When another connection is needed in the drill string 16 , the steps - can be repeated. Steps and are included in the FIG. That is, the data acquisition and control interface 94 continues to receive data from the sensors 36 , 38 , 40 , 44 , 46 , 54 , 56 , 58 , 62 , 64 , 66 , 67 and supplies appropriate data to the hydraulics model The hydraulics model 92 continues to determine the desired annulus pressure corresponding to the desired downhole pressure.
The controller 96 continues to use the desired annulus pressure as a setpoint pressure for controlling operation of the choke It will be appreciated that all or most of the steps described above may be conveniently automated using the control system For example, the controller 96 may be used to control operation of any or all of the flow control devices 34 , 74 , 76 , 78 , 81 automatically in response to input from the data acquisition and control interface Human intervention would preferably be used to indicate to the control system 90 when it is desired to begin the connection process step , and then to indicate when a drill pipe connection has been made step , but substantially all of the other steps could be automated i.
However, it is envisioned that all of the steps - can be automated, for example, if a suitable top drive drilling rig or any other drilling rig which enables drill pipe connections to be made without human intervention is used. The control system 90 of FIG.
The predictive device preferably comprises one or more neural network models for predicting various well parameters. These parameters could include outputs of any of the sensors 36 , 38 , 40 , 44 , 46 , 54 , 56 , 58 , 60 , 62 , 64 , 66 , 67 , the annulus pressure setpoint output from the hydraulic model 92 , positions of flow control devices 34 , 74 , 76 , 78 , drilling fluid 18 density, etc. Any well parameter, and any combination of well parameters, may be predicted by the predictive device The predictive device may be trained by inputting to the predictive device data obtained while drilling at least one prior wellbore.
The training may include inputting to the predictive device data indicative of past errors in predictions produced by the predictive device. The predictive device may be trained by inputting data generated by a computer simulation of the well drilling system 10 including the drilling rig, the well, equipment utilized, etc. The predicted parameter values can be supplied to the data validator for use in its data validation processes. The predictive device does not necessarily comprise one or more neural network models.
The predictive device may perform a regression analysis, perform regression on a nonlinear function and may utilize granular computing. The predictive device receives the actual parameter values from the data validator , which can include one or more digital programmable processors, memory, etc. The data validator uses various pre-programmed algorithms to determine whether sensor measurements, flow control device positions, etc. For example, if a received actual parameter value is outside of an acceptable range, unavailable e.
The desired annulus pressure setpoint is communicated from the hydraulics model 92 to the predictive device for use in predicting future annulus pressure setpoints. However, the predictive device could receive the desired annulus pressure setpoint along with the other actual parameter values from the data validator in other examples. The predictive device is trained in real time, and is capable of predicting current values of one or more sensor measurements based on the outputs of at least some of the other sensors.
Thus, if a sensor output becomes unavailable, the predictive device can supply the missing sensor measurement values to the data validator , at least temporarily, until the sensor output again becomes available. If, for example, during the drill string connection process described above, one of the flowmeters 62 , 64 , 66 malfunctions, or its output is otherwise unavailable or invalid, then the data validator can substitute the predicted flowmeter output for the actual or nonexistent flowmeter output.
It is contemplated that, in actual practice, only one or two of the flowmeters 62 , 64 , 66 may be used. Thus, if the data validator ceases to receive valid output from one of those flowmeters, determination of the proportions of fluid 18 flowing through the standpipe line 26 and bypass line 72 could not be readily accomplished, if not for the predicted parameter values output by the predictive device Validated parameter values are communicated from the data validator to the hydraulics model 92 and to the controller The hydraulics model 92 utilizes the validated parameter values, and possibly other data streams, to compute the pressure currently present downhole at the point of interest e.
Although the predictive device may stop training one or more neural network models when a sensor fails, it can continue to generate predictions for output of the faulty sensor or sensors based on other, still functioning sensor inputs to the predictive device. Upon identification of a faulty sensor, the data validator can substitute the predicted sensor parameter values from the predictive device to the controller 96 and the hydraulics model The predictive device is preferably also able to train a neural network model representing the output of the hydraulics model A predicted value for the desired annulus pressure setpoint is communicated to the data validator If the hydraulics model 92 has difficulties in generating proper values or is unavailable, the data validator can substitute the predicted desired annulus pressure setpoint to the controller The parameters a, b, c,.
Differences between the actual and predicted values for the parameter y can be useful in training the neural network model e. During training, weights are assigned to the various input parameters and those weights are automatically adjusted such that the differences between the actual and predicted parameter values are minimized.
- Advancing the Technology of Bileaflet Mechanical Heart Valves.
- Well drilling forum;
- Under New Management: Universities, Administrative Labor, and the Professional Turn.
If the underlying structure of the neural network model and the input parameters are properly chosen, training should result in very little difference between the actual parameter values and the predicted parameter values after a suitable and preferably short training time. It can be useful for a single neural network model to output predicted parameter values for only a single parameter.
Multiple neural network models can be used to predict values for respective multiple parameters. In this manner, if one of the neural network models fails, the others are not affected. However, efficient utilization of resources might dictate that a single neural network model be used to predict multiple parameter values. Such a configuration is representatively illustrated in FIG. If multiple neural networks are used, it is not necessary for all of the neural networks to share the same inputs.
In an example representatively illustrated in FIG.
Oil well control
The neural network models , share some of the same input parameters, but the model has some parameter input values which the model does not share, and the model has parameter input values which are not input to the model If a neural network model outputs predicted values for only a single parameter associated with a particular sensor or other source for an actual parameter value , then if that sensor or other actual parameter value source fails, the neural network model which predicts its output can be used to supply the parameter values while operations continue uninterrupted.
Since the neural network model in this situation is used only for predicting values for a single parameter, training of the neural network model can be conveniently stopped as soon as the failure of the sensor or other actual parameter value source occurs, without affecting any of the other neural network models being used to predict other parameter values. The configuration of FIG. However, in the FIG. The choke manifold 32 , pressure sensor 46 and flowmeter 58 may also be provided as a separate unit. Note that use of the flowmeters 66 , 67 is optional.
For example, the flow through the standpipe line 26 can be inferred from the outputs of the flowmeters 62 , 64 , and the flow through the mud return line 73 can be inferred from the outputs of the flowmeters 58 , In this configuration, the flow control device 76 is connected upstream of the rig's standpipe manifold This arrangement has certain benefits, such as, no modifications are needed to the rig's standpipe manifold 70 or the line between the manifold and the kelley, the rig's standpipe bleed valve 82 can be used to vent the standpipe 26 as in normal drilling operations no need to change procedure by the rig's crew, no need for a separate venting line from the flow diversion unit , etc.
Well drilling forum
The flow control device 76 can be interconnected between the rig pump 68 and a flow control device 77 in the standpipe manifold 70 using, for example, quick connectors 84 such as, hammer unions, etc. A specially adapted fully automated flow control device 76 e. The entire flow control device 81 can be customized for use as described herein e.
grupoavigase.com/includes/398/4682-encuentros-sexuales.php It may now be fully appreciated that the above disclosure provides substantial improvements to the art of pressure and flow control in drilling operations. The above disclosure provides a well drilling system 10 for use with a pump 68 which pumps drilling fluid 18 through a drill string 16 while drilling a wellbore A flow control device 81 regulates flow from the pump 68 to an interior of the drill string 16 , with the flow control device 81 being interconnected between the pump 68 and a rig standpipe manifold Another flow control device 74 regulates flow from the pump 68 to a line 75 in communication with an annulus 20 formed between the drill string 16 and the wellbore Flow is simultaneously permitted through the flow control devices 74 , The flow control device 81 may be operable independently from operation of the flow control device The pump 68 may be a rig mud pump in communication via the flow control device 81 with a standpipe line 26 for supplying the drilling fluid 18 to the interior of the drill string The system 10 is preferably free of any other pump which applies pressure to the annulus The system 10 can also include another flow control device 34 which variably restricts flow from the annulus An automated control system 90 may control operation of the flow control devices 34 , 74 to maintain a desired annulus pressure while a connection is made in the drill string The control system 90 may also control operation of the flow control device 81 to maintain the desired annulus pressure while the connection is made in the drill string The above disclosure also describes a method of maintaining a desired bottom hole pressure during a well drilling operation.
The method includes the steps of: dividing flow of drilling fluid 18 between a line 26 in communication with an interior of a drill string 16 and a line 75 in communication with an annulus 20 formed between the drill string 16 and a wellbore 12 ; the flow dividing step including permitting flow through a standpipe flow control device 81 interconnected between a pump 68 and a rig standpipe manifold 70 , the standpipe manifold 70 being interconnected between the standpipe flow control device 81 and the drill string The flow dividing step may also include permitting flow through a bypass flow control device 74 interconnected between the pump 68 and the annulus 20 , while flow is permitted through the standpipe flow control device The method may also include the step of closing the standpipe flow control device 81 after pressures in the line 26 in communication with the interior of the drill string 16 and the line 75 in communication with the annulus 20 equalize.
The method may include the steps of: making a connection in the drill string 16 after the step of closing the standpipe flow control device 81 ; then permitting flow through the standpipe flow control device 81 while permitting flow through the bypass flow control device 74 ; and then closing the bypass flow control device 74 after pressures again equalize in the line 26 in communication with the interior of the drill string 16 and in the line 75 in communication with the annulus The method may also include the step of permitting flow through another flow control device e.
The method may also include the step of determining the desired annulus pressure in response to input of sensor measurements to a hydraulics model 92 during the drilling operation. The step of maintaining the desired annulus pressure may include automatically varying flow through the flow control device e.