Autonomous Control

Autonomous Control

Current artificial lift systems have certain elements of control. For example, the Subsurface Compressor System (SCS) and the Electric Submersible Pump (ESP) use a variable speed drive (VSD) on the surface to control their motors downhole. There are also other elements in the well system that provide either information or actuation to control the well. For instance, operators install downhole sensors to measure the temperature and pressure of a well at different depths. However, these elements are seldomly linked together to provide better control of the overall system. Even if the information from different elements is used to improve the performance of the well, it is often used in an ad-hoc and non-real-time fashion by engineers located remotely after the fact. The adverse situations must be handled in real time in-situ by a system level control scheme to avoid damages to the downhole equipment and to ensure continuous operations of the artificial lift system. There is a need to link all of these elements together to achieve optimal performance of the system.

Block diagram of the SCS system
Block diagram of the SCS system

The SCS has several plants (e.g. motor, bearings and compressor), actuators (e.g. PM motors to generate rotation, bearing actuators to levitate rotors, bearing damping actuators to dampen the vibrations, and compressors to increase the gas pressure), sensors (e.g. motor back EMF voltages, bearing sensors and bearing damping sensors), and controllers (e.g. VSD and magnetic bearing controller (MBC)) throughout the SCS system. The VSD was originally designed to control SCS and ESP motors. The MBC can control the magnetic bearings in the SCS.

Without an autonomous system control scheme, as the environment changes, the motors will only respond to the VSD’s control commands, while the bearings will only respond to the MBC’s control commands. For example, if an environmental change affects the bearing performance that can be mitigated by varying the motor speed, then in the current configuration, this action can only be done by an engineer. To avoid damages to the downhole equipment, the engineer notified by the MBC must have the knowledge to analyze the information and make the right decision about what actions to take on the VSD, must have access to change the setting of VSD, and must act in a timely fashion. Therefore, for faster response and to eliminate human error, there is a need to have an autonomous control at the system level to protect and optimize artificial lift systems.

How Autonomous Control Works

The autonomous control of the SCS
The autonomous control of the SCS

As shown above, energy is transmitted into the SCS from the surface (as indicated by “Electricity”). The motor will convert the energy to torque and transmit torque to the compressor via a thrust bearing unit. The compressor will convert the torque to the pressure ratio of the gas flow across the intake and discharge of the SCS. With the pressure ratio by the SCS, there will be a lower downhole flowing pressure (as indicated by “Drawdown”) generated in the wellbore to induce more gas production. The thrust loads from the compressor will be taken by the thrust bearing unit.

An autonomous system controller links all the elements to control the whole system without human interference. Each plant (e.g. motor, bearing, etc.) could have its own controller. The information (as indicated by “Info”) from the sensors of a plant will feed into both the plant controller (e.g. motor controller, bearing controller, etc.) and the system controller (as indicated by all the red lines shown). For example, the position sensor in a magnetic bearing will send the information about the rotor position to both the bearing controller and the system controller. With the information from some of the sensors (sensors of the plants or other external sensors), the system controller, based on a predetermined autonomous control scheme (algorithm with logics), will decide what action to take by sending out control commands (as indicated by all the green lines) to the actuators in the plants to optimize the operation of the system according to an initial setting (as indicated by the “Setting”). In the meantime, the system controller will also send the information from the sensors to the surface (as indicated by “Monitoring”) to set off alarms, display in a monitor for the engineer to review in real time, or store in a database for later use.

To summarize, a system controller uses information from the SCS elements and other external sensors to take appropriate actions autonomously. This will help optimize the performance and reliability of the SCS system in-situ without the delay of decision-making and manual control by personnel located remotely.

Four types of autonomous control schemes
Four types of autonomous control schemes

There are four types of autonomous control schemes as shown above. When a system (either SCS or ESP) is surrounded by its environment, both the system and its environment can be either active or reactive.

When the environment is actively/constantly changing, the autonomous system controller will receive the information about this change from the sensors. The system control will take appropriate actions to ensure that the system will reactively keep up with the environment in order to maintain the operation of the system. Under these circumstances, when the environment makes the first move and the system catches up autonomously, the autonomous control is in the mode of optimization of system performance under external disturbances (as shown in the lower right corner of the table).

The system can also actively change the state of operation to induce changes to the environment (as shown in the upper left corner of the table), particularly inducing changes to producing reservoirs. The purpose to do so is to understand the reservoir responses to the changes of the SCS. In this case, the SCS is performing a programmed well testing, wherein the pressure measurement records form the basis for transient well-test analysis and are primarily used for determining reservoir-rock properties and producing-formation limits.

When both the system and its environment are not active (as shown in the lower left corner of the table), there is information from the sensors in the meantime, indicating that some of the SCS parameters are deviating from its set values, and there could be deteriorations of the SCS components causing it is to move away from its operating point without active changes from either the SCS or its environment. In this situation, the system controller can set off alarms on the surface to indicate the need for maintenance on the SCS. Essentially, the system controller is performing diagnostics (or even prognostics) of the SCS.

Instead of responding to the changes of the environment, the system controller can predict the changes of its environment and the SCS itself, and can proactively change the state of the SCS operation to ensure that it always stays in the optimal operating conditions. To ensure that the SCS can always adapt to its environment, the knowledge of how the reservoir and SCS will evolve with time is required upon the installation of the SCS in the well. In this case, both the SCS and its environment are active (as shown in the upper right corner of the table).