VFD vs. recirc line

[mjpetrag]

I have a low flow pump application with a centrifugal pump discharging 5 GPM at 100' head. Currently, the flow downstream of the pump is metered by a control valve that adjusts its position controlled by the level inside the tank feeding the pump. The suction head to the pump is slightly above atmospheric.

The problem we are currently having is that the control valve cycles between completely closed and part to full open, causing the pump casing to overpressure and blow seals out. The pump is constantly run.

One of the ways to remedy this is to install a VFD to control the discharge head and flow rate, and leave the control valve full open in manual. The easier way is to install a recirc line with a smaller diameter than the discharge piping on the pump discharge to the suction line of the pump. The control valve will still operate the way it does now.

What are the drawbacks to a recirc line?

Also, is there a possibility of creating a feedback loop and blowing out the pump with a recirc line? For example, the control valve is closed and the pressure is all going to the recirc line, therefore the pump suction pressure is the same as the discharge pressure and the pressure keeps building as a feedback loop until it breaks.

-Mike

 

[BigInch]

First of all, install a relief valve in the pump.

Is it a positive displacement pump? Why/When do the pump seals blow now, when the CV is closed?

The recirculation line method may use more energy when the pump is at lower flow rates. If running that 5 HP motor costs you a lot, and you run low flowrates a lot of the time, you might consider that VFD solution. The VFD will tend to save energy at lower flows, but it will not give you your 100 ft head at those lower flowrates. At 50% flow, it will only give you 100 ft / 1^2 * 0.5^2 = 25 ft of head. At 25% flow, it will be down to 6 ft. So a VFD might not work at all.

A recirculation line might be used down to about 20% flow and still give 100 ft head. Below 20% flow expect to begin running into overheating trouble.



**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[mjpetrag]

This is a centrifugal pump. The seal failure mode analysis done by the seal supplier points to excessive pressure from throttling the pump (from the control valve). The pump is at its shutoff head frequently.

The relief valve you are talking about...do you mean a relief valve that would dump back to the tank?

-Mike

 

[BigInch]

Dump to anywhere in order to protect the pump seals.

Sounds like this is not a well designed pump, if it can't even handle its own shutoff head pressure.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[BigInch]

Or the supply tank pressure is way too high.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[zdas04]

I've had good luck with a constant pressure valve on a recirc (with positive displacement pumps). The valve looks at delivery pressure and increases or decreases recirc volume to keep it constant. Something like a Kimray backpressure valve would work at this pressure. In an application where you have to have constant delivery pressure (in my case I was driving a downhole jet pump) and can tolerate wasted energy it works well.

I don't see why the same scheme would not work for a centrifugal pump that requires 100 ft to deliver liquid into the next process. As BigInch says, a VFD will lower the available head with the pump speed. That is often not a problem, but it can be in some applications.

David

 

[BigInch]

For a 5 HP, where energy use is pretty insignificant, and you need a relief valve anyway, install a relief valve off the pump discharge, but between pump and the control valve, and pipe it back to suction piping or back to the tank, whatever's closer. Seals: Happy.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[mjpetrag]

The pump is almost 40 years old, and the seals are the only part that is failing. I was just out looking at it and it cycled between 40 and 60 (shut off head) psi about 3 times in 10 minutes. The seals really aren't robust enough to handle this.

The supply tank pressure is 15 psig, so I don't believe that is much of an issue as the throttling is.

Just talked to the planner and it looks like the VFD is out of the question for now (with all the red tape, cost approvals, etc.) to get it installed. I'll take a look at that Kimray valve, maybe something similar will help.

Also, in my OP I wrote "Also, is there a possibility of creating a feedback loop and blowing out the pump with a recirc line? For example, the control valve is closed and the pressure is all going to the recirc line, therefore the pump suction pressure is the same as the discharge pressure and the pressure keeps building as a feedback loop until it breaks." I always wondered about this. Can this happen?

-Mike

 

[BigInch]

Yes, that's why I suggested a relief valve recirculation line combination. Set the relief valve low enough to prevent the seals from blowing and when flow drops enough, the pump tends to increase output pressure and when it gets to the relief set pressure, that activates before the seals blow.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[CH5OH]

normally the seals shut last pumping with dead head pressure
since the flow regulating valve cycles, does it operate correctly?what type of level sensor you are using to control the flow valve?Is it a pneumatic proportional
system or just open/close?
the seals blowing out is caused by chock wave, rather than cut off pressure.
I think you should take a close look to the flow valve controlling mechanism and at least dampen its respond time.

 

[dcasto]

on a small pump like this, I'd put a 1/2" line running from the discharge line upstream of the level valve back to the suction of the pump. The 1/2" line would have a (about) 1/8" orifice in the union. This would gaurentee that you always have a minimum flow through the pump.

You can either size the orifice by calculation of by experimenting.

 

[mjpetrag]

It seems a little uncertain to size the orifice because I only know the pressure of the line and the pressure of the tank. I'm thinking I would have to do an iterative solution and probably wouldn't be too accurate.

I'm thinking about putting a 1" line off the pump discharge with a check valve and a backpressure regulator set a few psi below the shutoff head back into the tank.



-Mike

 

[MJCronin]

Perhaps there are alternatives....

For this small amount of flow,(5 gpm) and the need for delivery over a range of conditions, I believe that the replacement pump should be a PD type.

It is possible to get a DC motor driven pump in this small size and get the full range of floew conditions.

Alternately, if you can afford the cost of compressed air, a larger Air driven Diaphragm pump can meet the 5 gpm at 100 TDH requirements.

My opinion only

-MJC