Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations GregLocock on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Throttling pump to get it back on the curve

Status
Not open for further replies.

waseem19

Civil/Environmental
Nov 23, 2002
82
Hi all,

I have a situation where the pumps are pumping outside of its curve, max flow.

the client has no problem with throttling the outlet to move the duty point back on the pump curve, he knows all about the energy losses and don't care. I was concerned about the pressure drop required to reach the required duty point, I'll need approximately 3.4bar drop.

We will get this by semi closing a butterfly valve. Will this pressure drop cause cavitations? Even if it does and the only thing that will get damaged is the valve and a piece of pipe then no problem. Does this valve need to be away from the pump “few meters” to prevent cavities from affecting the pump it self?

Is there something else I should worry about?

Regards.
 
Replies continue below

Recommended for you

If you are not concerned over energy losses then your best solution is to install orifice plates in the downstream pipeline. Throttling a butterfly valve is not a good solution.

If cavitation is a problem then you can install several plates in series.

Whether or not cavitation is a problem will depend on the pressure at the pump delivery.
 
Throttling is nothing to do with cavitation. In your case, the unthrottled flow may be more influential in pump cavitation. Just check for the safe minimum flowrate of the pump. You would get better responses if you provide, initial and final flowrates and discharge pressures.

 
I assume he was questioning cavitation of the valve. Cavitation at the valve will affect the valve and perhaps the downstream pipe. It will not affect the pump. But its occurrence depends whether the velocity through the valve is sufficient to drop the pressure to vapour pressure. This is unlikely on a pump discharge. You can do the calculations or throttle the valve and listen if it is cavitating you will hear it.
 
I just noted that you said a 3.4bar drop. That needs a velocity of some 25m/sec. Orifice plates are the best solution. Do you really want to waste so much energy - what is the flow rate?

 
Waseem, you have received good advice here so my only comment is that you should look at this as a developmental process rather than as an ultimate solution.

During commissioning it is not uncommon to find this situation where you have an over-sized pump. Typically you would overcome this for the first few hours by throttling the valve, then for the next few days by inserting a properly sized restriction orifice. Once the system has settled down and you have the final flowrates and pressure drops you would implement a permanent and economic solution - either install a VSD as recommended by Keith, or get the impeller trimmed to the correct diameter.

BRIS - how did you get from 3.4 bar to 25 m/sec?

Harvey

Katmar Software
Engineering & Risk Analysis Software
 
Katmer - agreed trimming the impeller, or other economic solution, would generally be recommended. It depend son flow, how often the pump operates and why you want to reduce the flow and increase the pressure to get the pump back on curve (excessive amperage, inadequate NPSH). Not much point in improving efficiency if you are throwing away the energy at the orifice plate.

The 25m/sec is simply the velocity to be destroyed to create a head loss of 34m. orifice loss = (Vvc-Vds)^2/2g (for water) Vvc = velocity at vena contracta, Vds = velocity in downstream pipe.


 
I would use a control valve, but not a butterfly, use a globe valve. See Fischer Control Valves or the like. 3 or 4 bar drop is no problem for a globe. If you have 50 bar drops, you can get special anticavitational trims for the valve. Noise will be you're biggest concern.

No point using a VFD if you're not too concerned with energy loss, they'll add more maintenance headaches.

Orifice plates usually do not have enough variability if the flow changes later, but could be a quick solution, esp if the flow doesn't vary much.

Throttling is likely to control cavitation if cavitation is an existing problem. It won't add to that problem.
 
Cavitation (NPSHr) will be reduced at the pump but may be a problem at the valve/orifice plate depending on your system. If you want to keep costs to a minimum you can place the orifice plate downstream of your existing butterfly valve. This will provide additional back pressure on the valve and reduce the potential for cavitation. You may then use the butterfly valve for control. You need to size the orifice for the maximum flow you require.
 
It's hard to believe in today's environment, with today's electric costs and bottom-line requirements, that he's able to get a manager to say wasted energy is "not important" - long term fuel costs are much greater than the (apparetly lower) initial equipment cost.

The noise, lack of variablity, and excess backpressure from a control valve + orifice plate(s) required by a too-large pump are incredibly short-sighted solution by management. VFD is a much better solution in the long run....
 
Why is the pump running outside of it's max curve?- is it pumping more than required (and being stop/started) or is the downstream system demanding the higher flow?

Throttling the pump will use less power than the pump is currently using- but as per above- what effect does the reduction in flow have on your process?
 
Like a few of the other postings, I would be interested in the amount of power you are going to consume destroying 3.4 bar of discharge head.
What is the flow and unthrottled head and what is the required head of this pump.

Naresuan University
Phitsanulok
Thailand
 
All,

these are existing pumps, throttling will be for a short period of time. its pumping 300 l/s against 72m head through 600mm GRP pipeline.

control valve and VSD are more expensive in capital cost and will take long time to be done. the client wants something low in capital (high in O&M is acceptable due to budget allocation) as long as it solves the problem and can be done fast.

The orifice plate seems to be the best solution, I've never done any, I'll Google it and see what comes up, in the mean time can anybody point me in the right direction on how to size these , installation , cost ? is not possible to get these on a rotating disk ? ( so you can increase or decrease the area )

these pumps are not pumping outside of their curve now, but will be ,because the client is looking to add a new connection to the existing main. The idea is to throttle the pump some where (maybe at the connection point) to pump exactly the same flow it is pumping now (not more) by splitting the flow into the two mains.

trimming the impeller is a good idea but cannot be done cause the operator might wish to turn off the new connection and pump through the old one only.

thanks all for your post
 
We commonly use orifice plates to throttle back pumps in groundwater abstraction where we are mining the ground water and the pump is over sized for the initial conditions.

In your case I would recommend two orifice plates - (to reduce noise) construction is simply a plate with a hole placed between flanges on the pipeline. Head loss across the orifice is H = (Vcc - Vds)^2/(2g) (Vcc = velocity at the vena contracta. You can take the area at the vena contracta to be about 0.65 x the area of the hole in the orifice. Vds = the velocity in the downstream pipe. (note most calculations are based on using orifice plates as a flow measuring device and are not relevant to your application).

I have never considered a variable orifice. It is a simple procedure to swap plates when you want to change system characteristics. If you oversize slightly you can throttle back (fine tune) with your butterfly valve.
 
I am sure that with Google you will quickly find a source to enable you to calculate the orifice bore required. But there is another problem you need to address.

In a pipe this diameter you will need to check the mechanical strength of the orifice plate. I have seen 1/8 inch plate used for orifices that later looked liked they had been deliberately dished. A worst case would be for the orifice to actually tear and have a chunk of steel going into some downstream piece of equipment. Get a competent mechanical engineer to check the plate thickness required.

Katmar Software
Engineering & Risk Analysis Software
 
You seem to have answered your own question - from the description you have given - all you are doing is maintaining the head on the pumps to what currently exists.
Using orifice/s plates to restrict flow is common usage for controlling power input to a pump or to control discharge and seems a logical way for you to go.

Naresuan University
Phitsanulok
Thailand
 
If you have a variable head use a valve.

If the short duration still is months/year yu should consider the trimming of the impellar (same relation in flow/head/power) vs impellar size as for VFD. But then you must order a new impellar to replace the one thats been altered.

Remember that 3.4 bar at a flow rate of .3 m3/sec corersponds to a power consumption of approx 100 kW (with an assumed 100% efficiency!).

Best regards

Morten
 
I must agree with Morten and others regarding the high costs of restricting the flow - although you haven't specified a period of time for operating under this condition. If it's a tempory condition "maybe" it can be justified but anything else is money down the drain - Burning-off something like 120kW is a small fortune every day - take the time to calculate the number of kWh's you will use every day.

Naresuan University
Phitsanulok
Thailand
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor