RU2591870C1 - Method for adaptive automatic control of gas condensate wells - Google Patents

Abstract

FIELD: oil and gas industry.

SUBSTANCE: invention can be used to control operating modes of gas producing well. Control of operating modes of gas producing well is formed on the basis of an adaptive pulse regulator acting on a time quantizer, in which there happens locking of control signal magnitude u_imi(t) during a given period (unit) of time with further action on the actuator mechanism controlling the regulating valve changing the amount of gas fed into the collector, ensuring quantized signal u_kv. At that, the adaptive regulator has control law

where k_I, k_D are coefficients of integral and proportional components of PID regulation law, k_P(t) -common transmission coefficient, which varies according to equations: k_P(t)=k_H+γy(t), T_ay′(t)+y(t)=u_PIDa,kv(t), where filtration time T_a is calculated in standard units, and k_H>0 and γ>0 are adjustable parameters of the adaptive regulator.

EFFECT: increasing of resource of technical means of automatic elements.

1 cl, 11 dwg

Description

The invention relates to the field of gas production, and in particular to methods for controlling the operating modes of a gas producing well.

A known method of controlling the operating modes of a gas well in which gas production is carried out with the restriction of the selection of gas coming from the annulus by throttling while maintaining a constant pressure drop using an automatic control system in which the received data is analyzed and a command is sent to the automatic gas flow control valve optimizing the total debit of the well in accordance with a certain dependence taking into account the filtration resistance of the well (see, for example, atent RU №345266, A1 01.01.1972).

The disadvantage of this method of operating gas wells is that it does not provide an optimal level of control of operating modes and has a low resource of technical means of automation elements.

As a prototype, a method for controlling flow in a well is described as described in patent RU No. 2513942 C2, 01/27/2014.

In the known method for controlling the operating modes of a gas well with the restriction of gas extraction from the annular space by throttling while maintaining a constant pressure drop using an automatic control system, the obtained data are analyzed and commands are sent to the automatic gas flow control valve, optimizing the total debit of the well in according to a certain dependence and taking into account the filtration resistance of the well.

The advantage of the prototype is a higher accuracy of maintaining the constancy of the output streams.

The disadvantage of the prototype, as well as the analogue, is the low resource of technical means of automation elements. In addition, during the operation of a gas well, the physical characteristics of the gas produced from the well change, the gas pressure and flow rate (debit) of the well change. This requires reconfiguration of the gas production control system. In the analogue and prototype, this is not provided.

The objective of the invention is to increase the resource of technical means of automation elements.

The technical result of the invention is the optimization of an adaptive system for automatic control of a gas production well while maintaining the required high accuracy of the pressure of the outlet gas under the conditions of a changing gas flow rate and the changing parameters of the gas production itself and ensuring a high resource of technical means of automation elements.

The technical result is achieved due to the fact that in the method of controlling the operating modes of a gas well with the restriction of gas extraction coming from the annulus by throttling while maintaining a constant pressure drop using an automatic control complex, in which the received data is analyzed and commands are sent to the automatic control valve gas flow rate, optimizing the total debit of the well in accordance with a certain dependence, taking into account the filtration resistance of the well, according to As per the invention, the control of gas production well operating modes is formed on the basis of an adaptive pulse controller acting on a time quantizer, in which the value of the control signal u _ui (t) is fixed for a predetermined period (quantum) of time with subsequent action on the actuator controlling the control valve, changing the amount of gas entering the collector, following the quantized signal u _kv , while the adaptive controller has a control law

,

,

where k _AND , k _D are the coefficients of the integral and proportional components of the PID regulation law, respectively, k _P (t) is the total transmission coefficient, which varies in accordance with the equations:

k _P (t) = k _n + γy (t),

T _a y ′ (t) + y (t) = u _{PID, sq} (t),

where the filtering time T _{a is} calculated in arbitrary units, and k _n > 0 and γ> 0 are the tunable parameters of the adaptive controller.

,Control of the gas well’s operating modes based on an adaptive pulse controller acting on a time quantizer, in which the value of the control signal _uii (t) is fixed for a specified period (quantum) of time with subsequent action on the actuator controlling a control valve that changes the amount of gas entering the collector, following the quantized signal and the _square at which the adaptive controller has a control law

,

where k _AND , k _D are the coefficients of the integral and proportional components of the PID regulation law, respectively, k _P (t) is the total transmission coefficient, which varies in accordance with the equations:

k _P (t) = k _n + γy (t),

T _a y ′ (t) + y (t) = u _{PID, sq} (t),

where the filtration time T _{a is} calculated in arbitrary units, and k _n > 0 and γ> 0 are the adaptive regulator's adjustable parameters, which allows maintaining the required high accuracy of the outlet gas pressure under the conditions of the changing gas flow rate and the changing parameters of the gas producing well and provides a high resource of technical means elements of automation.

The invention is illustrated by 11 figures.

In FIG. 1 shows a block diagram of a gas production well.

FIG. 2 shows a schematic structural diagram of control links.

In FIG. 3 presents: (schedule a) the change in u _them - the control signal of the actuator of the control valve when exposed to a conventional adaptive PID controller; transient pressure change P _k (graph b) with a sampling of 0.01 s in real time t, s.

In FIG. 4 there is a graph similar to FIG. 3, a transient process using a conventional adaptive PID controller with 5 s sampling in real time t, s.

In FIG. 5 shows the transient process with changed parameters: u _them - the control signal of the actuator (graph a); conventional adaptive PID controller and the transition process of pressure change P _k (graph b) with a sampling of 0.01 s in real time t, s.

In FIG. 6 there is a transition process of changing the pressure p _ZPA to a predetermined pressure value p _kz = 10 (graph b) with the initial u _them = 0 graph (a). The settings of the PID controller are the same as in FIG. 5. The time quantizer worked with a sampling time of 5 s.

In FIG. 7 - transients in real time: (a) for u _them during pressure testing (b) r _ZPA with modified parameters of a conventional PID controller.

In FIG. 8 there is a block diagram of an adaptive control device with a PID controller, with the control law proposed in this application.

In FIG. Figure 9 gives an idea of the transient process of changing the pressure in the well collector to a predetermined value in a system in which the usual adaptive PID controller is replaced by the proposed adaptive controller: a) u _it is the control signal for the actuator of the control valve; b) R _PD - bottom-hole pressure u _them - the control signal of the actuator of the control valve; c) the overall transmission coefficient k _p . The process is shown in real time t, s.

In FIG. 10 shows a transient in real time t, c, similar to FIG. 9, when working out the pressure p _k (b) in the reservoir with other values u _them (a), k _p (s).

In FIG. 11 shows transients of pressure change p _k (b) in a manifold similar to FIG. 9 and 10 with modified initial values of u _them (a), k _p (s).

The optimal adaptive automatic control system for gas and gas condensate wells is arranged as follows. The gas flow 1 (Fig. 1) from the well passes through a control valve 2, which is connected to the actuator 3. The latter is under the influence of a time delay relay (RVZV) 4 of the adaptive controller 5 and a smoothing filter 6. In turn, the smoothing filter 6 is connected to a comparison unit 7, which responds to a signal from a pressure sensor 8 located in the gas stream in the (manifold) 9 after the control valve and pressure regulator 10. After the control valve 2, gas enters the block of switching equipment 11. Regulating The valve 2 may change the flow area and thereby change the amount of gas flowing through the collector.

The adaptive controller 5, in turn, contains an adaptation block 12 (Fig. 2), a PID calculation block of the control law 13 and a filter 14. The first input of the adaptation block 12 is the signal of the adaptive controller 5. The output of the adaptation block 12 is connected to the input of the PID calculation block - control law 13. The output of the PID calculation block — control law 13 is the output of the adaptive controller 5 and is connected to the input of the filter 14. The output of the filter 14 is connected to the second input of the adaptation unit 12. Transients were modeled for various parameters of a conventional system p -regulation with various parameters and control system when changing the time discretization discredited (FIGS. 4, 5, 6, 7).

Adaptive controller system control with regulation law

based on the block diagram of FIG. 8. The overall transmission coefficient k _P (t) varies in accordance with the equations:

k _P (t) = k _n + γy (t),

T _a y ′ (t) + y (t) = u _{PID, sq} (t),

where T _a > 50 conventional units,

k _n > 0 and γ> 0 are tunable parameters of the adaptive controller, ay (t) is the assessment of the position of the actuator.

Figures 9, 10, 11 give a clear idea that the transition processes go more smoothly, there are no fluctuations, and the number of inclusions becomes several orders of magnitude less when using the proposed optimal adaptive PID controller.

In the description of the invention, the following notation:

R _PD - bottom-hole pressure;

Ρ _s - gas pressure in front of the control valve;

P _to - gas pressure after the control valve;

Ρ _KZ - task for gas pressure;

P _ZPA - gas pressure in the intake manifold ZPA;

q _c - gas flow from the well

q _to - gas flow after the control valve;

u _them - the control signal of the actuator of the control valve.

The adaptive automatic control method of gas and gas condensate wells is as follows. The pressure adjuster 10 (Fig. 1, 2) generates the required value of the gas pressure in the manifold. In the comparison unit 7, the set pressure value is compared with the current value generated in the sensor 8, the pressure in the manifold, which comes from the pressure sensor 8, which measures the pressure in the manifold 9. The error signal (errors) received in the comparison unit 7 is fed to the smoothing filter 6, in which the measurement noise of pressure sensor 8 is smoothed out. The filtered error signal is fed to the input of adaptive controller 5, which generates a control signal. The control signal from the output of the adaptive controller 5 is fed to a time quantizer 4, in which the magnitude of the control signal is fixed for a given period (quantum) of time. The quantized control signal from the output of the temporary quantizer 4 is supplied to the actuator 3, which, acting on the control valve 2, changes the amount of gas entering the collector 9.

Evaluation of the effectiveness of the optimal adaptive automatic control system for gas and gas condensate wells was carried out on the basis of comparison with the well-known automatic control complex of the prototype. So in FIG. 3 shows a transition process at working pressure p in the reservoir _{to the} wellbore from the _PAD value p corresponding to _them u = 0, to a predetermined pressure value p = 20 _kOe. A PID controller with a transfer function of the form was used as a controller

, $$w_{Reg}(R)=k_P\left(\mathrm{one}+\frac{k_{\mathrm{AND}}}{p}+k_{D}R\right)$$

,

where k _P = 0.019, and k _I = 0.1, k _D = 2.

The time quantizer worked with a discretization of 0.01 s. At the same time, the number of changes u by _{them was} 110, the integral error was 3615, the time for establishing pressure was 78 s, and the overshoot coefficient was 1.47.

FIG. 4 shows how the transient will change when working out the pressure p _k in the reservoir of the well from the value of p _ZPA at u _im = 0 to the preset pressure value p _kz = 20. A PID controller with settings k _П = 0,016, and k _И = 0,09, k _Д = 2 was also used as a controller. The time quantizer worked with 5 s sampling. In this case, the number of changes u _was equal to 13, the integral error was 4463, the time for establishing pressure was 88 s, and the overshoot coefficient was 1.27.

From the transient comparison of FIG. 3 and FIG. 4 it can be seen that the introduction of a temporary quantizer into the control loop allows, by substantially reducing the number of inclusions of the executive body, to practically not worsen the quality performance of PID control systems.

The operation of a control system with a PID controller with constant settings at various values of technological parameters is simulated.

In FIG. 5 - transient at working pressure p in the reservoir _to the well by the values of p when u _{them PAD} = 0 to a predetermined pressure value p = 30 _kOe. The controller was a PID controller with settings k _P = 0.018, and k _И = 0.08, k _Д = 2.4. The time quantizer worked with 5 s sampling.

In FIG. 6 - transient at working pressure p in the reservoir _to the well by the values of p when u _{them PAD} = 0 to a predetermined pressure value p = 10 _kOe. The settings of the PID controller are the same as in the processes of FIG. 5, i.e. k _P = 0.018, and k _I = 0.08, k _D = 2.4. The time quantizer worked with 5 s sampling. From FIG. 6 it follows that when the controller settings are saved in the transition process, fluctuations occur and the quality of the control system becomes worse.

In FIG. 7 transient mining pressure p in the reservoir _{to the} wellbore at p = 10 _kOe. In this case, in order to ensure stability in the control loop, the controller tuning parameters were taken equal to k _P = 0.005, and k _И = 0.09, k _D = 2.2. Thus, to maintain the quality of regulation, you need to change the overall gain in the regulator.

From the graphs of FIG. 3-7, two conclusions can be made:

1) the introduction of time quantization did not practically worsen the quality indicators of the considered control system in comparison with the quality indicators of the operation of a conventional PID controller with a high frequency of control signal change;

2) when the gas pressure in the collector decreases, the parameters of the technological control object change significantly, which, with constant adjustments of the regulator, lead to a loss of stability of the control system.

From the first conclusion it follows the expediency of introducing into the composition of the control system a temporary quantizer, the presence of which, due to a significant decrease in the number of actuators, will significantly increase its resource. The second conclusion implies the need to use an adaptive regulator in the control loop, which could evaluate the changing parameters during the process operation and, based on the results of the evaluations, optimally adjust the regulator settings so that it is guaranteed to ensure stable and high-quality regulation in all possible well operation modes.

Adaptive control was based on the model of FIG. 8, in which the conventional PID controller is replaced by an adaptive controller with a control law

the total transfer coefficient k _P (t), which varies in accordance with the equations:

T _a y ′ (t) + y (t) = u _PID , _sq (t),

where T _a > 50 conventional units,

where k _n > 0 and γ> 0 are the adjustable parameters of the adaptive controller, ay (t) is the assessment of the position of the actuator.

In FIG. 8 is a block diagram of an adaptive PID controller, which is included as a controller in the block diagram of FIG. one.

Checking the use of the adaptation unit was carried out on the basis of modeling its operation at various values of the technological parameters.

In FIG. 9 depicts transient at working pressure p in the reservoir _to the well by the values of p when u _{them PAD} = 0 to a predetermined pressure value p = 10 _kOe. The initial value k _n = 0.0024, the final value found in the adaptation mode, k _P (t) = k _n + γu _PID (t) = 0.005125. In the upper graph of FIG. 9 shows the control signal u _them , on average - pressure p _to , at the bottom - the change in the transition process of the total coefficient of the controller k _P (t). From the graphs of FIG. 9 it follows that the process, which at the settings corresponding to FIG. 6, was unstable, due to adaptation it became not only stable, but also quite high quality.

The transient time of FIG. 9 is approximately 170 conventional units, while the optimally tuned process of FIG. 7 lasted approximately 200 conventional units.

In FIG. 10 - transition process at working pressure p in the reservoir wells _{to the} bush from the _PAD value p when u = 0 _{to them} to a predetermined value of pressure p = 20 _kOe. The initial value of k _n = 0.0024, the final value found in the adaptation mode, k _P (t) = k _n + γu _a (t) = 0.01115. In the upper graph of FIG. 10 - control signal u _him , on average - pressure p _to , at the bottom - a change in the transition process of the total coefficient of the controller k _P (t). The adaptive control processes of FIG. 10, in this case, are somewhat inferior in speed to the processes of FIG. 4 since in the case of FIG. 4, the total gain k _P at the initial moment of the transition process was selected from the calculation of the high-speed control system and amounted to 0.016, while in the case of adaptive control the initial value k _P (0) = 0.0024, which then increased during operation by approximately 5 times.

In FIG. 11 is a transient process when working out the pressure p _k in the reservoir of the wellbore from the p value of the _ZPA at u _im = 0 to the preset pressure value p _kz = 30. The initial value of k _n = 0.0024, the final value found in the adaptation mode, k _P (t) = k _n + γu _a (t) = 0.0185. In the upper graph of FIG. 11 - control signal u _him , on average - pressure p _to , at the bottom - a change in the transition process of the total coefficient of the controller k _P (t). Adaptive control processes, as in the previous case, are somewhat inferior in speed to similar processes of FIG. 5 for the same reasons discussed for the case of FIG. 10.

Thus, the introduction of an adaptive regulator into the gas well control system made it possible to ensure high efficiency of this system in all pressure stabilization modes in the well reservoir. There is an optimization of the adaptive system for automatic control of a gas well while maintaining the required high accuracy of the pressure of the outlet gas under the conditions of a changing gas flow rate and the changing parameters of the gas well itself and ensuring a high resource of technical means of automation elements.

Claims (1)

A method for controlling gas well operating modes with restriction of gas extraction from the annulus by throttling while maintaining a constant pressure drop using an automatic control complex in which the received data is analyzed and commands are sent to the automatic gas flow control valve, optimizing the total well flow rate in accordance with a certain dependence, taking into account the filtration resistance of the well, characterized in that the control of gas production The borehole is formed on the basis of an adaptive pulse controller acting on a time quantizer, in which the control signal u _ui (t) is fixed for a given period (quantum) of time with subsequent action on the actuator, controlling a control valve that changes the amount of gas entering to the collector, following the quantized signal u _kv , while the adaptive controller has a control law

where k _AND , k _D are the coefficients of the integral and proportional components of the PID regulation law, respectively, k _P (t) is the total transmission coefficient, which varies in accordance with the equations:
k _P (t) = k _n + γy (t),

where the filtering time T _{a is} calculated in arbitrary units, and k _n > 0 and γ> 0 are the tunable parameters of the adaptive controller.

Images