engineers kill me

So to start the story... I have a new instillation of a water pump. They had me on startup limit the max freq. to 42hz on VFD. I did so without even asking why.

After running the pump it does about 1700 to 1800 gpm depending on the head pressure. At 42hz and 29amps (rated 26amps) 30.5 trip. But they want more flow.

They have a restrictor plate on the line...and some discussion started about how to increase the flow.

Engineer says.. just remove the restrictor plate. problems solved.. well needless to say I tried to explain that "work" is work. If you remove the plate the current will go down..agreed.. but the flow will decrease. He says.. then increase the max frequency... well to my dismay he could not comprehend that this would increase flow..and increase current to the same (or close to) the overload.

oh and yes the pump is spinning the right direction..lol

Why is the restrictor plate installed in the first place?

If you remove it then flow should increase and head should decrease just like a discharge throttle valve. Think I would start pump at about 10 Hz and step up in 5 Hz intervals and take data to draw an as installed system curve.

Dan Bentler

The head I refer to is the head pressure from the container they/we are filling... it's open to air..and flow decreases as the container fills.

I don't belive the pump is following the pump curve they provided...and they scratch their heads.

Kinda reminds me of the enginner that said..."these two water towers are the exact same elevation"....... 3hrs later the 2nd one overflows.

The head I refer to is the head pressure from the container they/we are filling... it's open to air..and flow decreases as the container fills.

Ahh good to know VITAL info -- flow decreases as head increases -did not some guy named Bernuli dream that up?

I don't belive the pump is following the pump curve they provided...and they scratch their heads.

So pull out the restrictor plate and build your own pump curve. Can you pump to tank with max flow and let tank overflow with no bad effects?
IF so now you can draw a "at max head" speed vs flow vs power curve for the pump(s ??)

Uhhh these are centrifugal pumps right??

Dan Bentler

I am unable..to pull the restrictor but might "make" the contractor do it. But I have run the pumps at max head within a few inches...since it runs out an overflow.... at at max head 1800 gpm @ 30amps Thats when I told them it was wrong.

I'll have to look at the pumps in the am to see the type I don't recall tonight.

robo77 said:

I am unable..to pull the restrictor but might "make" the contractor do it. But I have run the pumps at max head within a few inches...since it runs out an overflow.... at at max head 1800 gpm @ 30amps Thats when I told them it was wrong.

I'll have to look at the pumps in the am to see the type I don't recall tonight.

Click to expand...

Gotta love design by trial and error.

robo77 said:

Engineer says.. just remove the restrictor plate. problems solved.. well needless to say I tried to explain that "work" is work. If you remove the plate the current will go down..agreed.. but the flow will decrease.

Click to expand...

Well, there certainly is a lot of misconception rolling around the jobsite, but not all of it is by the engineers.

First off, based on your description I'm going to assume that this is a centrifugal pump.

Second, while it is possible that the pump isn't performing along the curve, that is very unlikely. Most pump suppliers are pretty good about that. It is more likely that either some instrumentation is scaled wrong or the pump suction is being restricted. Most likely is that noone is reading the pump curve properly!

Look at the attached for an example pump and system curve.

Now, for a couple of basics: Pumps follow affinity laws similar to fans. If you change impleller diameter or speed a given flow point on the curve drops in direct proportion to the ratio of the change. The pressure the pump produces at the original curve point drops as the square of the ratio. Power associated with the original curve point drops as the cube of the ratio. I've attached a typical example. (Note: Bernoulli's Law isn't the same as pump affinity laws, but Bernoulli's Law is important in analyzing pump systems. It has to do with the relationship of total energy, static head, friction losses, and velocity head.)

Another important point is that the pump curve defines pump capability. The system curve defines the system resistance to flow. Most often, as in the attached, the system curve has some static lift (40 ft in the example) and some friction losses (System curve 1 has a friction loss of 40 ft. at 2,000 gpm). At a given flow the sum of the two is what the pump must overcome at a given flow. The intesection of the system curve with the pump curve defines the actual operating point for a given pumping system.

Removeing a restrictor plate will increase flow, all other things being equal. In the example system curve 1 would be with the restrictor, system curve 2 without. If the pump is operating at 52 Hz removing the plate would increase flow from 1,700 gpm to 2,000 gpm. The motor power and current would also increase.

You are correct that increasing speed will increase flow. If the plate is left in the system's restriction will follow curve 1. Increasing the speed from 52 Hz to 60 Hz will increase flow from 1,700 gpm to 2,300 gpm, the discharge pressure guage should read about 94 ft H2O (41 psig) and the power will increase to about 68 hp.

This looks like another example of Jenkins Law. Nobody really understands pumps, so the pump must be bad!

(Why they want a restrictor plate I don't understand unless it was a bad pump selection and the plate is needed to keep flow and power down. That works, but isn't the best way to do it.)

(Why they want a restrictor plate I don't understand unless it was a bad pump selection and the plate is needed to keep flow and power down. That works, but isn't the best way to do it.)

Click to expand...

I don't know if it's applicable in this case, but one reason to use a restriction on the pump discharge is if the NPSH requirement is not met. Because the pressure is higher at the discharge, it can also be higher at the suction, so I have worked on systems where throttling the discharge did actually increase the flow.

AutomaticLeigh said:

I don't know if it's applicable in this case, but one reason to use a restriction on the pump discharge is if the NPSH requirement is not met. Because the pressure is higher at the discharge, it can also be higher at the suction, so I have worked on systems where throttling the discharge did actually increase the flow.

Click to expand...

I'm skeptical. I'm sure you are reporting what you saw, but I'm willing to bet there are more pieces to the puzzle to explain your observation.

I don't see how adding restriction to the discharge would increase pressure at the suction side of the pump. Pressure at the suction is determined by elevation difference between water surface & pump inlet and the friction loss through the suction piping. The higher the flow rate, the lower the suction pressure because piping losses increase as the square of the flow. That reduces NPSH (Net Positive Suction Head). To make things more interesting, NPSHR (NPSH Required) increases as the flow increases for a given pump.

I suspect that you are describing a system where they changed to throttling the discharge instead of throttling the suction. That would be consistent with your observations. If that is the case then indeed you could get more flow out of the pump. However, it wasn't restricting the discharge that created the improvement, it was reducing the restriction on the inlet.

I should also add that it is considered bad practice to throttle a pump on the suction side. This can lead to cavitation problems. If you have to throttle to reduce flow, do it on the pump discharge.

Sock it to em', Tom! I thought Robo's perceptions sounded flaky from the first post, all twisted around.

Hmh, yeah robo should clarify where the restrict plate is installed into.

I ASSUMED the
1 pump was centrifigul
2. the restrictor plate was on the discharge

I really need to stop making these foolish assumptions

Well, Leitmotif, I don't see anything wrong with making assumptions as long as you say that's what they are! Sometimes there isn't an alternative.

I'm skeptical. I'm sure you are reporting what you saw, but I'm willing to bet there are more pieces to the puzzle to explain your observation.

Click to expand...

I guess the answer is in your post. When the pump has something to get hold of, the flow is high, which pulls down the suction pressure due to losses on the suction side. Cavitation then occurs, which stops the pump from pumping effectively. If the flow is restricted on the discharge, this situation doesn't occur. OK, it's a stretch to say throttling the discharge increases the flow, but that's certainly what it looks like if you're trying to control the system.

In one case, this was pumping hot water out of vacuum chamber. Normally I'd say dig a hole and chuck your pump down it, but if you're standing on the bottom plates of a ship, that dimension is a bit limited.