Dosing Pump Double Diapraghm

[bmw318be]

Hi,

I am using the dosing double diapraghm, I try to troubleshoot and eliminate possible cause that caused the reduced flow by 50%


The application is a low pressure about 1 bar as I am testing it on the short discharge pipe less than 2 meters, when running I am using the following

1. calibration column as a tank to check on the capacity.

2. After discharge there is an external relief valve

3. Pulsation dampener

4. Pressure gauge reading

Seek an opinion if metering pump was flooded the suction with calibration column is sufficent to create postive suction ?


I feek the reduced flow is puzzled me when we have tested in the workshop, the pump under flooded suction and about 9 bar, we could meet the duty under high discharge prrssure condition.

Current site condition is
6 LPH
1 bar

I am getting around 1 bar,did anyone has ever encountered this synthoms of 30 to 50% drop in flow under low discharge pressure ?

What are possible caused.

We have hydro test the system to 8 bars for 30 mins.

Is there any method to identify the causes of this low flow.

I understand that air trapped in diapraghm could cause this issue ?

Would the 2 check ball valves in suction and discharge does not seat it properly due to hydo test


What is the correct procedures of hydro test the system ,is it wrong to from the suction pressurize through the discharge line , would this caused the seat valves to not seating in due to the hydro test

 

[bimr]

The check valves are most likely defective or worn out.

 

[sshep]

leaking internal check valves or internal (or external) relief valve is the most likely cause. An undersized suction is possible, but your flow is so low that seems unlikely. If you have veriable stoke or speed then you must test the capacity running at 100% conditions. Good luck, sshep

 

[itdepends]

Try increasing the backpressure before you pull the pump apart. Under low backpressure conditions the suction and discharge check valves may not seat quickly or well.

In particular you'll want a reasonable static discharge- if you have a short pipe that you're using a valve to provide the backpressure then during the transition from full discharge to suction stroke the discharge pressure (which is what seats your valves) will drop to almost zero.

If that doesn't work- it's likely to be the check valves inside the pump.

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"

 

[bmw318be]

If we pump in water to the suction line and blanked off the discharge line subject to 8 bar, as this part of the FAT test, understand metering has 2 check ball valves in suction and discharge, would the water pressure from the suction damage the check valves since we do it for 8 bar ?.

I seek your advise if this hydro test with metering pump off is common in the system, i feel that we should test run the pump before hydro test so we could actually get the actual factory test conditions.

Would low backpressure caused the lower flow for metering, understand that for positive displacement pump, lower back pressure would normally increase slightly the flow under operating conditions.

However when measuring the repeatability, we see the flow at 25, 50, 75 and 100% are good less than 3 %

 

[LittleInch]

From what you say I gather you are going to pressurise the pump from the suction side to 8 bar? So long as the pump is rated for this there shouldn't be an issue.

In terms of your low flow issues, it depends has it I think, i.e. these pumps work much better when you have a decent back pressure on the discharge, 1 bar or less is not a good working environment. You might need to include a pressure relief valve set at say 4 bar in the discharge line to create this back pressure. Try it and see and see what the vendor recommends.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[moltenmetal]

Agree that you need some backpressure on the discharge to force the discharge check valves to seat, or else you will have no luck using an air operated diaphragm pump for dosing/metering. As a transfer pump they don't care much, but if you want constant volumetric efficiency you need some backpressure.

 

[bmw318be]

Molten metal, I think you made a good logic, the ball valve has to reseat in order to create a consistent flow, however what I am experiencing now not the actual condition that the pump is designed, you are right that factory might have tested at high 8 bar during testing, in FAT, we normally only have short piping connection that can stimulate only 1 bar or less.

I believe I can make an elevation to create a higher discharge pressure, it is recommended to restrict the flow to create backpressure for metering pump ?. I normally did this on internal gear pump, throttle the discharge valve to create backpressure. I need an advise on to safely create the backpressure


Little inch,

The pump is rated at discharge beyond 8 bar, however if we pressurized from suction, correct me if i am wrong, understand metering pump is not designed for high suction pressure, for discharge, understand that it can go very high 10 above depending on the pump.

 

[moltenmetal]

Backpressure can be provided by means of backpressure regulators or spring-loaded check valves etc. Throttling will give a backpressure which varies with flow and which will affect your testing results.

A true PD pump gives a constant flow irrespective of backpressure. However, an air operated diaphragm pump is definitely not a true PD pump given that it has a compressible gas as a motive fluid. And even a mechanically or motor-driven diaphragm pump without this issue is going to have check valves on both the suction and discharge side which leak at least a little- their efficiencies will not be 100% and will vary somewhat with backpressure. For metering service, you want the fixed (controlled) backpressure to be more significant than the variable (process) backpressure so the flow you get at a given operating rate is as close to constant as practical.

 

[bmw318be]

Hi Moltenmetal,

The pump is electric dosing pump with double diapraghm, the principle is

All dosing pump ( hydraulic oil) is having bottom suction and top discharge:
1.when piston retract position, check ball valve drop and seat in, suction ball valve goes up and draw the liquid to the diapraghm half

2. When the piston pumped hydraulic oilto the diapraghm, the product being pumped up and the suction ball valve move upward

I believe as what "litteInch",sufficient backpressure may be needed for the ball valve must be able to drop or seat in. As the ball is very small about 5 mm and light, if not enough backpressure and if the suction pressurr push in, it may not have delay to seat in .


Is it recommended to use a hose to increase the elevation discharge to stimulate the higher discharge line ?

 

[itdepends]

It's a pretty small pump- 6 LPH (I missed that during the first review). A double diaphragm pump to me is an air operated unit with a diaphragm on each side of an air inlet and exhaust.

Sounds like the one you're describing is a single diaphragm. I know the type- they can be a pain in the proverbial to get going. Hydrotest was not likely an issue assuming the pump was rated for it.

Make sure you've primed it really well- get all the air out.
A good suction head will help with this- or better yet connecting a hose to a t-piece on the suction and pushing water through the pump.

Check to make sure it doesn't have variable stroke length as well as variable speed- many of the dosing pumps do. If it does have variable stroke length- be sure to check the manufacturers instructions about how to adjust it (e.g. adjust while pump is running or pump stopped).

For FAT you should be testing it with the backpressure you expect it to operate at- up to you how you generate that- something that provides a bit of buffering would be best - e.g. a short section of larger diameter pipe/hose that is only partially flooded and has your restricting valve on the low point.

Heck- just set up a water hose with a t-piece. Test your injection pump by having it "dose" into the t-piece and control the dosing pressure by adjusting a valve on the outlet of your "main" water hose- the hose flow will be >>> than the dosing flow and will provide a constant back pressure for your test.

Finally- given the size of the pump - a small piece of debri would be sufficient to cause a check valve to not seat. The pump should be pretty easy to pull apart and check them- before you do- try applying pressure (again via hose connection) on the discharge- if both suction and discharge valves pass you'll soon find out.

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"

 

[moltenmetal]

Sorry, other posters had concluded it was an air operated double diaphragm.

We use hydraulic double diaphragms of the size you're talking about by the hundred, made by many different vendors. If you don't have significant springs to seat the discharge checks, which is sometimes required by the service, in our experience you will need substantial back-pressure to get them to pump at all, and even more to keep them pumping with constant volumetric efficiency.

 

[bmw318be]

Hi Sorry did not elaborate the details of the types of pump specifically, this is actually API 675, Dosing pump, Double diaphragm using 2 balls (NOT a SPRING), using electrical motor where accuracy and repeatability shall be +- 3 %, We did achieving the repeatability but not maximum rated flow that shall be designed and tested.


I managed to do a sketch on the layout of the pump, I have also tried to increase the back pressure using back pressure valves to around 4 bar, however no change in flow from 1 bar discharge pressure

Link of the sketch of the P & ID belowLink

http://expirebox.com/download/aedc0beecb78b609a215cc416e508689.html

Could it be the suction pressure used in calibration column is not enough to prime the pump?

I observed, no vibration,bi abnormality, as the capacity is achieving the repeatability less than 1 %. Even when the stroke length adjusted to 25, 50, 100 %, however the pump much lower than the rated by 60 %, I am not sure if this pump is self priming and able to work against low suction pressure.

What I seek advise to isolate the issue one by one:

1. Back pressure increasing (Done)
2. increase the elevation of the Calibration column higher to ensure that, it has NPSHa sufficient

3. Decoupled the instruments ? Pulsation dampener, Back pressure valve, External Relief valve, Stainer


I was given a data form manufacturer than the NPSHr is 0,3 bar


QUOTED from Itdepends

Make sure you've primed it really well- get all the air out.
A good suction head will help with this- or better yet connecting a hose to a t-piece on the suction and pushing water through the pump.

My response: Yes, you are right, since this is hydraulic diapraghm pump, we did primed the diapraghm, understand if you look at my P & ID, my suction could be small, I am not really be able to estimate, do you think, it pump air, so it drop by 50 to 60 % ?

Check to make sure it doesn't have variable stroke length as well as variable speed- many of the dosing pumps do. If it does have variable stroke length- be sure to check the manufacturers instructions about how to adjust it (e.g. adjust while pump is running or pump stopped).

My response: Yes, The pump could be adjusted the stroke length when the pump is running and stop conditions, Stroke length is 11 mm only when at 100 %

For FAT you should be testing it with the backpressure you expect it to operate at- up to you how you generate that- something that provides a bit of buffering would be best - e.g. a short section of larger diameter pipe/hose that is only partially flooded and has your restricting valve on the low point.

My response:Sorry can you elaborate this, short section of larger diameter, the diameter is 1/2 inch pipe

Heck- just set up a water hose with a t-piece. Test your injection pump by having it "dose" into the t-piece and control the dosing pressure by adjusting a valve on the outlet of your "main" water hose- the hose flow will be >>> than the dosing flow and will provide a constant back pressure for your test.

My response:Can you elaborate or help me get a better idea what is the T pieve, and adjusting valve, where shall i install this and shall i restrict the flow to increase pressure

Finally- given the size of the pump - a small piece of debri would be sufficient to cause a check valve to not seat. The pump should be pretty easy to pull apart and check them- before you do- try applying pressure (again via hose connection) on the discharge- if both suction and discharge valves pass you'll soon find out.

My response: Would it be damage if I pressure from Discharge instead of suction, as the ball valve shall be restrict the air to come in once you pressure, Would the diapraghm get rupture as i have rupture diapraghm detection which consist of 2 diapraghm balanced

Thanks you

 

[itdepends]

Couldn't get the P+ID because the timer on the file life had expired.

Regarding your test to date with increasing backpressure- it sounds like you have a pulsation dampener on the discharge? Does it provide enough volume and buffer to keep the backpressure at 4bar when you test at higher backpressures? Did the pressure drop significantly during the pump stoke (e.g. if "set" for 4 bar, does the discharge pressure drop lower than 3bar at any stage during the pump stroke).

Ignore the comments RE discharge into something with backpressure and a buffering- just use a hose as described below.

To provide backpressure for testing.

Take a water hose- put a t-piece on the end.
Install a restricting valve and pressure gauge on one branch- this will restrict the flow of water out the end of the hose. The pressure gauge should be upstream of the restricting valve (this will require a second t-piece).
Connect your dosing pump to the other branch of the first t-piece
Turn the water supply on a small amount and adjust the restricting valve to get the backpressure you want to test against
Start the dosing pump

Regarding overpressure
Which direction you backpressure from isn't the issue- you need to check the pressure rating of the pump against your hydrotest. You shouldn't be relying on the internal check valves to protect the pump during a hydro- if it's not rated for the required pressure it shouldn't be connected during hydro. To test for a passing check valve you don't need a lot of pressure- a couple of bar will do it. But it will only show flow through to your calibration column if both valves are passing- if only one passes you won't see it- and you'll still have a problem.

Regarding pump priming- connect a hose to the suction and push flow through the pump while it is operating at 100% stroke length. Speed isn't as important. Please make sure the discharge of the pump is vertical to allow air to be expelled. After pushing all air out you can then test against your calibration column. These pumps will not self prime- you need to get all the air out first.

As a chem eng/metallurgist the first part of any answer I give starts with "It Depends"