Variable Speed Motor Control Under Pressure Versus Flow

[Sawsan311]

Dear All,

I am investigation control schemes for parallel injection pumps application, typically the injection flow rates are set at the individual wellhead chokes and system is required to achieve a target wellhead pressure.

I would like to get your technical views on the robustness of the following option for a variable speed motor (VFD) control:

A- setting the primary controlled variable as the flow rate with a required design contingency to ensure that in case one pump trips, the controller will not force the pumps to speed up beyond their run out conditions.
B- Pressure control at the common discharge header to maintain the network pressure at a target corresponding to the wellhead pressure. In case of high backpressure, the pumps will tend to slow down to maintain the set point.
From energy optimization perspectives, the throttling losses across the individual wellhead choke valves are not wasted as in the case of flow control. Additionally, the same injection flow rate can be managed under various target operating pressure and it will not be required for operator to continuously adjust the injection rates for the pump control.

On the other hand, if a fixed speed design is adapted with individual discharge throttling, in case of flat pump curve, operating the discharge valve as PCV will jeopardize the performance of the pump with respect to the risk of operating towards the left end of the curve. While the same concern is not envisaged in the case of a discharge FCV.

Do you agree that the pump curve shape is a determining factor for a throttling control scheme since the action of the valve modifies the system curve. While the pump curve shape will not be endangered in case of a variable speed control under pressure?? since system curve in this case is not being modified by a discharge throttling action?

Appreciate your views.
Thanks

Regards,

 

[georgeverghese]

Would prefer B with an additional feature if it is possible, to save on power:
If each of the wellhead chokes is remotely operated on flow control or manual, get the valve open position readout and compare with all other wellhead chokes. Take the largest open readout, go through a high signal select and feed a "stroke controller" with a gap setpoint of say 85(low)-90(high)%. Output of this controller is used to reset the pressure controller's current setpoint. So the pressure controller acts as a slave to the master stroke controller. Injection pressure is then minimised. Pump pressure and speed goes up when ANY ONE choke position exceeds say 92% for more than 60sec, and pump pressure and speed goes down (with another low signal select) when ALL choke positions are less than say 83% for than 60sec.

 

[Sawsan311]

Thanks Mr.George, in light of the fact that usually more than 100 manual chokes would be present in the system, it will be highly complex and challenging to employ an override under high flow rate choke opening.

The effect of opening more chokes would correspond to lower backpressure from the network, mandating the pump's speed to increase to maintain the set point, only if the flow is excessively high due to a leak , an override on high flow can be introduced to slow down the pump and hence mitigating pump run out.

 

[Sawsan311]

Hi again Mr.George, I would like to ask about a specific point which I have been discussing:

- For a variable speed driven motor pumps running in parallel, for a liquid injection application, in case one pump trips out of a group of running pumps, then the following will happen:

- The VFD controller will be requesting the remaining pumps to increase their speed to reach the preset set point which corresponded to the total requirement without changing the set point after the pump trips.

- When the pumps speed up, there is also an increase in discharge pressure i.e. the point will be moving along the original system curve which tends to slowly drift to the right by the increase in suction pressure (see below). Eventually, the system will restabalize to a new operating point where the pressure generated by the pump will be killed across the downstream the injection wellhead chokes.

- Upon the one pump tripping, the upstream suction pressure will increase, forcing the pumps to push more flow along with the requirement for the pump's VFD to speed up,

Do you agree with the above interpretation of the system dynamics and how VFD control will maintain almost constant backpressure, ofcourse pressure control will be more efficient with regards to energy efficiency?

regards,