Running Vertical Turbine Well Pumps In Parallel

[plcguru]

Our plant uses 2 American Turbine Well pumps to provide water for 2 paper mills. There is no elevated tank so the pumps must provide flow and pressure for proper mill operation. On very rare occasions are we allowed to use city water and only for a very short time. The well company convinced management to install VFDs on the pump motors. They can not "tune" the drives to provide a stable pressure inside the plant so I have been asked to look at the control system to see if it can be resolved. I think there may be some design problems.

The first pump is an American Turbine 12-M-70 (1760 RPM) designed for 700 GPM at 490 FT TDH. It is a 7 stage bowl. The pump is 360 feet deep. The pump is driven with a 150HP motor with a VFD. There is 675 feet of 8 inch pipe with 11-90 degree and 5-45 degree elbows before it ties into the common header inside the plant. There is an 8 inch swing check valve just before it comes into the header.

The second pump is a 12-RKBM (1760 RPM) designed for 800 GPM at 490 FT TDH. It is a 7 stage bowl. The pump is 300 feet deep. The pump is driven with a 125 HP motor with VFD. There is 170 feet of 8 inch pipe with 4-90 degree elbows before it ties into the common header close to the other pumps tie point. There is no swing check valve in this line at the header.

Both pumps have swing check valves in their respective pump houses to prevent any backflow.

The well company is trying to set up the control system to use Proportional, Integral & Derivative (PID) control to maintain a stable pressure inside the plant. The plant pressure transducer is about 15 feet from the common header assembly. When the pressure exceeds 110 PSI (in the Plant)it blows off to the city sewer system

The question is this: Should either pump 1 or pump 2 be run at a constant speed and use the other pump to "trim" to maintain a stable plant pressure, or, should both pumps be controlled with PID to maintain a stable pressure?

The plant flow requirements will change with the demands of the mill. As an example; the desired pressure in the plant should be 70 PSI. If it falls below 65 PSI for 2 minutes we want to correct (speed up pump 1 or pump 2 or both). If it goes above 85 PSI for 20 minutes we want to shut one or the other pump down.

The well company is not achieving the results that we expected so any comments or guidance will be greatly appreciated.

As always they bring in the electrical guy at the 11th hour!

 

PLCGURU:

It sounds like that well company is overstepping their role and is providing engineering consultation to you which is all to often the reason people like you run into trouble. This is a complex engineering problem that needs an engineering solution. You didn't provide enough information for me to help, but I would say that planning for both pumps to operate is a good idea since its a mill and flows vary so much and you must have reliability of pumping, meaning that both pumps may at any time have to take the role of lead or lag. As for your control, its a classic problem. To turn your pumps on you need a pressure signal, to turn one or both off, you need a flow signal. This is where the engineering comes in and where your well company missed it. I would prefer a PLC to monitor the system in liu of a PID.

If you need help, call me...

Bob Yurick, PE
Greeley and Hansen Engineers
215 563-3460

 

[stressriser]

I have just finished a study on pressure control for municipal water well systems. Your problem is in the system design. We have had many well systems try to install off the shelf VFD control programs, only to discover that they are not adaptable and can not be "Dialed in".

A PLC is indeed the right choice, they have the range you require. Two pumps are easily controlled by one PLC. Each pump will have it's own motor controller.

Pressure transducers should be used, NEVER use pressure switches. Each pump has two transducers, one upstream and one downstream.


The motors should be oversized, this way they run more efficiently. Instead of pressure switches hammering the system on, the motors are brought to speed gradually. They can be programmed to run together, seperatly or a combination of both.

Our water well distributors usually aid the engineers with requirements and system components. They are professionals at these systems and guarantee they will work. Every system is different.

Incidentally, although a PID is not directly used, it is a PID program mimicked in the PLC. Off the shelf components that the pump manufactures sell are usually for residential systems. Every time they try to adopt one to a municipality, they screw it up.

I just got back from Boggs Municipal Services in Wise, Virginia. They can help you out. 276 328 5144.

We manufacture pitless water wells. We see a lot of things. Make sure you get help from people in the well business, no two are the same.

 

[BobPE]

Again, I can't stress enough the need to seek the services of a registered professional engineer. By law, only a registered professional engineer can dispense advice, perform designs and calculations, and specify equipment to meet the needs of the public. This becomes fuzzy with industry because of something called the industy exemption that allows engineers working for an comapny to only consult for that company without professional registration status. They can not legally consult outside that company without registration.

Service companies often cross this line and end up consulting to industry because they know this is a fuzzy area. Often times this is not questioned since industry is just as confused about the laws as most others are. Now there are service companies out there that do have engineers on staff and this is fine as long as they are registered in the state where they are doing the design work.

Sorry to dwell on this, but I have had to work too many hours fixing things for industrial clients that were designed by outside the company non-engineers and it just seems to be a repatitive cycle.

 

[stressriser]

Nevertheless,
Even though you are right, the answer to all his questions can be answerd by Boggs.

Since he is in a private industry, anyone can help him.

This was his question:
The question is this: Should either pump 1 or pump 2 be run at a constant speed and use the other pump to "trim" to maintain a stable plant pressure, or, should both pumps be controlled with PID to maintain a stable pressure?


"The plant flow requirements will change with the demands of the mill. As an example; the desired pressure in the plant should be 70 PSI. If it falls below 65 PSI for 2 minutes we want to correct (speed up pump 1 or pump 2 or both). If it goes above 85 PSI for 20 minutes we want to shut one or the other pump down."

Here is the answer:
Boggs system will keep the pressure AT 70psi +/- 1 psi, response time is in milliseconds. If it falls below 70 psi for less than a second, the main pump will speed up. It will never go over 71 psi, so you don't have to worry about 85 psi. The pumps will alternate so as not to wear one more than another. If there is no water demand, the pumps will not run. As soon as a valve opens, and the pressure begins to drop, the main pump will gradually come to speed. The valves are controlled so that the pump starts BEFORE the valve fully opens, thus eliminating water hammer.

Again, call the people whom make the system. I hope your questions get answered. But Bob is right, consult a good registered firm if the well is municiple (for approval.)

 

[BobPE]

Stressriser:

The law is as I stated. No one can consult anyone outside of their own company without being a registered professional engineer. Even if they are private industry. You can look at the law at

http://WWW.NSPE.ORG

and even ask questions there.


PLCGURU's question is as you stated, but he also stated his problem :"They can not "tune" the drives to provide a stable pressure inside the plant so I have been asked to look at the control system to see if it can be resolved. I think there may be some design problems."

Your answer will not work unfortunately because he implied that he would like both pumps to run in his statement as a possible solution. Again, I don't mean to dwell on this, but PLCGURU's problem is a complex one that has no simple answer.

Bob

 

[stressriser]

I can't argue with you, and I do not want to.

I think that Boggs Municipal can help, maybe they can't, but it does not hurt to call. They make constant pressure booster station controls for water wells. All approved by a licensed PE. I have seen this problem before, and seen it fixed.

Regardless, you are right, the problem is much too complex for this forum, I'll leave it at that. I'm gone from this thread forever.

 

[butelja]

Be very careful applying VFD's to vertical turbine pumps. They very likely operate above the 1st critical speed. If you slow them down, you can run into resonance, high vibration, and rapid failure. This can be overcome by locking out certain critical ranges of RPM's around the resonant frequencies. Work with the pump OEM to obtain critical speed estimates. Then double check vibration levels during commisioning.