How will the VFD act?

[rutherford703]

I got a pump system like attached drawing. My questions are:
1) Pump D is a booster pump. How do you expect the VFD will response the suction pressure, high pressure to increase the pump speed to pump out more liquid, or low pressure to increase the pump speed to boost the pump discharge pressure?
2) Considering that the flow inside suction piping of pump D essentially does not change with the speed change of pump D, will the VFD cause any operating problem?

 

[TD2K]

As drawn, the VFD on pump D is just maintaining suction pressure. I would expect that as the suction pressure goes up the pump speed increases. As the suction pressure goes down, the pump speed would slow down. The discharge pressure isn't a control point as it's not seen by the control loop as drawn.

Maintaining the level in the tank by adjusting the speed of pumps A/B/C basically is adjusting the discharge pressure from pump C to create the necessary pressure difference between the discharge of pump C and the suction of pump D to get the required flow through the pipeline.

How much pressure drop do you take over the 1.5 km pipeline?

However, it's impossible to be certain just from the sketch what the control action of the PC on the pump's speed (though the system wouldn't operate properly in the field if the control action was reversed from what I gave). The action is usually described in a control narrative.

 

[BigInch]

VSD to level control. Unless the tank has multiple products of differing SG, the VSD would be better controlled from it's suction pressure.

If installed as shown, but on min suction pressure control instead of Liquid Level, usually, pump speed would slow as suction pressure decreased to wind up stopping the pump just as the minimum pressure, just greater than needed for NPSH of the pump, is reached. Similar to what I think pump D appears to be doing now.

There is no apparent need for 2 VSDs in such a short pipeline like this system. Why feed a short pipeline more than what you're taking out.

I would arrange this as an on/off liquid level control running the feed pumps, fully pressurizing the line. Put a flow, or pressure, control valve at the end of the pipeline, eliminating the last pump if possible, or leaving it there, if a pressure boost is still needed when the line is flowing at the required delivery rate and flow friction losses are excessive. In that case, leave your pump D as is, running with min suction pressure control.

There usually isn't much need for pipeline feed pumps to run on VSD/VFD control. You've made a nice pipeline, so why force it to run at less than capacity. You will have flow or pressure control at the downstream end and your upstream pumps will be on when needed and off when not, little better efficiency is possible, except in unusual cases.

What would you be doing, if you knew that you could not fail?

 

[rutherford703]

Thanks for your responses

After posted the thread I found the answer in my files. Your both answers are great. Possibly I should give each of you a star.

The reason we use two VSDs is the discharge locations which can be 4 km difference. sometimes the pump system has too much engergy than required. Sometimes the pump system is just OK. The location for the booste pump is based on the pump house and gland water available. The pumping liquid is slurry with specific gravity of about 1.5. Putting control valve in such a system is very difficult because of the erosion and corrosion of the system.

 

[bimr]

You need to define the main control objective in the simple Piping & Instrumentation Diagram in order to understand the control target and prepare the System Control Narrative

 

[BigInch]

OK, It now makes sense why a control valve isn't good.

I'd set the D pump vfd for maintaining a minimum suction pressure to avoid cavitation possibilities. Increasing suction pressure would speed up giving you the maximum possible flow at D, without cavitation, at any given time.

Was that the answer you found in your files?

What would you be doing, if you knew that you could not fail?

 

[bimr]

Biginch,

The missed it, the answer was shown on the second coctail napkin.

 

[BigInch]

Ya I didn't see the second attachment.
"speed control on booster inlet to control suction pressure."

I agree that an alarm on some low speed limit would be a good addition. I forgot about the possibility of plugging as the slurry slows down. I had the same problem with a hot heavy crude pipeline. If the pump station goes down and the pipeline cools off, the viscosity goes from 80 cP to 1300 resulting in a 130 mile long plug in the pipeline. 3 pumps, 2 x 100% spares, minimize that possibility. Hot molten sulfur lines are worse, but usually a whole lot shorter and heat traced to remelt it if it freezes up.

What would you be doing, if you knew that you could not fail?

 

[BigInch]

Oh. I didn't forget. you didn't mention the slury until your second post.

What would you be doing, if you knew that you could not fail?