2 Tank with 3 pump running in parallel at 13 GPM

[bmw318be]

Hi,

I have a system that have the following:

Atmospheric Tank is 10 meter height
3 reciprocating pumps running in parallel at 13 GPM
4 Inch common header from the tank and then
3 pumps are tees from this 4 inch header with 2 Inch pipe diameter at 13 GPM.
Pumping is Chlorine

Question,

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar ?

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?

3. What is the average velocity of the 4" common header when 2 tank are opened ?

4. Is there anyway to calculate the velocity in the pipe header ?

5. Would the pump P8 starved as it is the furthest distance ?

 

[LittleInch]

Really needs a diagram and a bit more explanation but here's a first go.

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar ?
What "2 tank" ?? But No, static head is static head, you can't increase it by adding a second identical tank

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?
Almost certainly, but depends on the Liquid level in the tanks and a section drawing

3. What is the average velocity of the 4" common header when 2 tank are opened ?
Depends on the difference in liquid level, length of lines between the two, any other flow restriction.

4. Is there anyway to calculate the velocity in the pipe header ?
Velocity = volumetric flow divided by area, e.g m^3/second divided by inside pipe area in m^2 = m/sec

5. Would the pump fighting and starved the furthest distance ?
Probably, but impossible to tell with the (lack of) data provided

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

 

[bimr]

Needs a diagram, and type of pump, and fluid properties to make a reasonable answer.

 

[mk3223]

Agreed. a system diagram with detail of the components and options could be understood better than the words described. With it, I believe some of the questions could be answered automatically.

 

[bmw318be]

Hi,

Thanks, i have edited my question and attached the sketch

 

[pennpiper]

bmw318be,
My thoughts:
Atmospheric Tank is 10 meter height. These Tanks should be supported at the same elevation. if they are not at the same elevation you can run the risk of over flowing the lower tank.
3 reciprocating pumps running in parallel at 13 GPM Here you say 3 pumps. Your sketch shows 10 pumps and all 10 connected to the Tanks. What gives?
4 Inch common header from the tank and then. Here you say 4" Header then next you say 3" Header. Which is correct?
3 pumps are tees from this 3 inch header with 2 Inch pipe diameter at 13 GPM.
Pumping is Chlorine

Question,

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar?
What do you mean "If the 2 tank is open"? If your two tanks are not supported at the same elevation and both Tank Isolation Valves are open then you only have one liquid level and you have only one static head.

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?
If your two tanks are not supported at the same elevation and both Tank Isolation Valves are open then you only have one liquid level and you have only one static head.

3. What is the average velocity of the 4" common header when 2 tank are opened ?
It depends on the liquid level.

4. Is there anyway to calculate the velocity in the pipe header ?
Yes.

5. Would the pump fighting and starved the furthest distance ?
Sorry, this question does not make sense.

Sometimes its possible to do all the right things and still get bad results

 

[KevinNZ]

Re question 3

The flow in the suction header will be the same as the flow through the pumps. Positive displacement pump flow does not vary much with suction pressure, so level in the tank/s will make little difference. If you know the pumped flow then you can work the velocity in the header. You can also work out pressure drops to the pumps for question 5.

 

[bmw318be]

Hi kevin,

My undersanding, if the pressure is igher than thr discharge pressure, we would have uncontrolled flow . Let say the back pressure is only 1 barg. If the tank is 10 m height with 1 atm so total would be 2 barg.

Do you think the static height would create a greater static flow along the pipe ? .

And my question if the tank 1 and 2 opened, what would be my suction pressure ?

Q4. The velocity at common header is it based on the total of 3 pump flow in parallel ? Q =13 x 3 = 39 GPM . My doubt, there would be natural flow due to high static height or pulsating flow of the reciprocating pumo , should i factor in by 3.14?

 

[bmw318be]

bmw318be,
My thoughts:
Atmospheric Tank is 10 meter height.

These Tanks should be supported at the same elevation. if they are not at the same elevation you can run the risk of over flowing the lower tank.

Could you elaborate, what support at the same elevation ? . The tank 1 and 2 is the same elevation.
3 reciprocating pumps running in parallel at 13 GPM Here you say 3 pumps. Your sketch shows 10 pumps and all 10 connected to the Tanks. What gives?

There is various dosing point with the automatic valve teed of from suction header. 10 Pumps might not be run at the same time in parallel. 3 would be run at the same time. At current scenario is P8-P9

4 Inch common header from the tank and then. Here you say 4" Header then next you say 3" Header. Which is correct?

I corrected, it is 4" header.

3 pumps are tees from this 3 inch header with 2 Inch pipe diameter at 13 GPM.
Pumping is Chlorine

Question,

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar?
What do you mean "If the 2 tank is open"? If your two tanks are not supported at the same elevation and both Tank Isolation Valves are open then you only have one liquid level and you have only one static head.

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?
If your two tanks are not supported at the same elevation and both Tank Isolation Valves are open then you only have one liquid level and you have only one static head.

3. What is the average velocity of the 4" common header when 2 tank are opened ?
It depends on the liquid level.

4. Is there anyway to calculate the velocity in the pipe header ?
Yes.

5. Would the pump fighting and starved the furthest distance ?
Sorry, this question does not make sense.

I mean would the pump starving when it run in parallel ?

 

[LittleInch]

bmw,

Looking through the items above I think I can summarise as below:

1) Opening the second tank will not make much difference and the head will not change and certainly not double.
NPSH is in absolute pressure, so atmospheric pressure at sea level plus 10m max liquid is 2 bara or 1 barg MAX.

2) In general with a header system you want to make your header as big as possible and to get equal spread of pressure make the header equal to the area or even size of all the offtakes added to gether. Hence your 4" header with multiple 2" oftakes is too small and will give you varying pressures at the pump inlet.

3) You seem confused about Static flow ( can't have both) and "natural flow"

4) Do you have an actual problem you're trying to solve? If so please tell us what it is.
As said, I think your inlet pipes are too small for the number of pumps you have and with PD pmps you might be getting soe sort of pressure pulsing or fluctuations.

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

 

[bimr]

Pumping is Chlorine

Are you pumping a 10% hypochlorite solution?

 

[bmw318be]

Hi Littleinch,

I appreciate your good points, I am actually verifying the design that is already designed by the consultant. I am the contractor that is going to built the system which has to follow the design and verify it. The challenges of oversizing the common header is the space given .

Let me add the response below

Looking through the items above I think I can summarise as below:

1) Opening the second tank will not make much difference and the head will not change and certainly not double.
NPSH is in absolute pressure, so atmospheric pressure at sea level plus 10m max liquid is 2 bara or 1 barg MAX.

Thanks you, so it is still better to use 1 pump in this case, anyway correct me if i am wrong, would the rate of static drop from the tank would be slower so the static head would be maintained at high level which is better to the pump, if we are opening up the 2 tank. ?

2) In general with a header system you want to make your header as big as possible and to get equal spread of pressure make the header equal to the area or even size of all the offtakes added to gether. Hence your 4" header with multiple 2" oftakes is too small and will give you varying pressures at the pump inlet.

Sorry, the offtakes is it the one after the common header to the pump inlet ? . Should I make sure the volume of liquid from 4" common header is larger than total 10 suction lines . May I know what is the rules of thumb larger by any specific % ?

3) You seem confused about Static flow ( can't have both) and "natural flow"
[highlight #4E9A06]I am trying to differentiate this term and try to conceptualize what is static flow, how much would it flow along the pipe when the valve is opened with 2 tank or 1 tank, would the 2 bara suction pressure create high acceleration along the suction header which might damage the pipe when undersized ?[/highlight]

4) Do you have an actual problem you're trying to solve? If so please tell us what it is.
As said, I think your inlet pipes are too small for the number of pumps you have and with PD pmps you might be getting soe sort of pressure pulsing or fluctuations.

[COLOR=]Yes, pulsating flow and pressure is the concern here. Is there anyway that i can show in the calculation for the amount of pulsating pressure along the common header or offtakes lines
[/color]

 

[bimr]

Question,

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar ?

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?

3. What is the average velocity of the 4" common header when 2 tank are opened ?

4. Is there anyway to calculate the velocity in the pipe header ?

5. Would the pump P8 starved as it is the furthest distance ?

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar ?

If Tank 2 is full of water (10 meters), then the static head will be approximately 1 bar.

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?

If one tank had a liquid level that was 5 meters less than the other tank, and both tanks were connected with a header, then the tank levels would eventually equalize.

3. What is the average velocity of the 4" common header when 2 tank are opened?

With a 4-Inch manifold between tanks, your flow would start out at an estimated 300 gpm if you have one tank full and the other empty. The flow would drop off quickly as the tank levels equalized. I would estimate an initial 10 feet per second velocity in the 4-Inch manifold pipe interconnection. Don't understand why you are concerned about this. You can slow the flow by closing back the valves.

4. Is there anyway to calculate the velocity in the pipe header ?

Calculate the head loss of a 4-Inch manifold pressure pipe flowing full and add the head loss through the valves and fittings. The head loss should equal the static head. As the static head drops, the flow will also drop.

5. Would the pump fighting and starved the furthest distance ?

No. The equalization between the tanks will not have an effect on the flow to the pumps. The flow to the pumps is only 39 gpm and requires only a 2-Inch pipe (4 feet per sec). Pump suction pipes should be 6-8 feet per second. If these are positive displacement pumps, I would recommend a 3-Inch pipe instead of the 2-Inch Pipe.

 

[bmw318be]

Dear Bimr,

Thanks, may I ask

1. If the 2 tank is open, would the static head is added to be 20 meter or 0.2 bar ?

If Tank 2 is full of water (10 meters), then the static head will be approximately 1 bar.

So with atm it shall be 20 m despite 2 tank, we would still based on 1 tank that is highest ?

2. If one scenario, tank 2 is lower 5 meter, would the tank 1 of 10 meter flowback to the tank and balance up ?

If one tank had a liquid level that was 5 meters less than the other tank, and both tanks were connected with a header, then the tank levels would eventually equalize.

Would the flow from higher level tank which is of 10 m equalize by flowing to the common header then dinding ways to the lowest pressure of any of the pump ? .

would the lower level tank which is of 5 meter remain it level until the 10 m tank reached 5 meter ?.




3. What is the average velocity of the 4" common header when 2 tank are opened?

With a 4-Inch manifold between tanks, your flow would start out at an estimated 300 gpm if you have one tank full and the other empty. The flow would drop off quickly as the tank levels equalized. I would estimate an initial 10 feet per second velocity in the 4-Inch manifold pipe interconnection. Don't understand why you are concerned about this. You can slow the flow by closing back the valves.

Could you please elaborate how you got 300 GPM with velocity of 10 feet per seconds in 4 inch .

4. Is there anyway to calculate the velocity in the pipe header ?

Calculate the head loss of a 4-Inch manifold pressure pipe flowing full and add the head loss through the valves and fittings. The head loss should equal the static head. As the static head drops, the flow will also drop.

What flow you are using to calculate the Pressure drop ?

5. Would the pump fighting and starved the furthest distance ?

No. The equalization between the tanks will not have an effect on the flow to the pumps. The flow to the pumps is only 39 gpm and requires only a 2-Inch pipe (4 feet per sec). Pump suction pipes should be 6-8 feet per second. If these are positive displacement pumps, I would recommend a 3-Inch pipe instead of the 2-Inch Pipe.

This pump is plunger which has acceleration head,

should the total flow combinations in parallel be considered alon section 1 and 2 for vrlocity calculation ?

As you refer to pipe section 1 and section 2 before header was only 2", would the velocity is high for the 2" pipe to handle?

 

[bimr]

So with atm it shall be 20 m despite 2 tank, we would still based on 1 tank that is highest ?

When the tank with the higher fluid level is opened, the pressure in the manifold will immediately increase. The pressure level that will be seen will depend on the static pressure in the tank, less the head losses in the tank outlet and manifold.

Don't understand why you are fixated on this as the tank levels will equalize within a short period of time.

If one tank had a liquid level that was 5 meters less than the other tank, and both tanks were connected with a header, then the tank levels would eventually equalize.

Would the flow from higher level tank which is of 10 m equalize by flowing to the common header then dinding ways to the lowest pressure of any of the pump ? .

would the lower level tank which is of 5 meter remain it level until the 10 m tank reached 5 meter ?.

When the difference in tank levels is high, the fluid will be flowing faster between tanks. When the fluid level between the tanks is slight, the effects of pump suction may be noticed.

What flow you are using to calculate the Pressure drop ?

The instantaneous flow may be as high as 300 gpm initially, but will rapidly diminishes as the difference in tank levels diminishes. You have a 4-Inch interconnection that is capable of passing 00 gpm.


This pump is plunger which has acceleration head,

should the total flow combinations in parallel be considered alon section 1 and 2 for vrlocity calculation ?

As you refer to pipe section 1 and section 2 before header was only 2", would the velocity is high for the 2" pipe to handle?

I understand that you will only have 3 pumps operating with a total flow of 39 gpm. A rule of thumb is to size pulsing positive displacement pump piping at 3X the average flow. The rule of thumb compensates for the velocity head. Suction pump piping should generally be sized at 6-8 feet per second. Conservatively sizing at 3X 39 gpm would give a velocity of 5 feet per second in 3-Inch pipe.

 

[LittleInch]

bmw,

You say that you are verifying the design. To verify a design you need to know more or at least as much as the person undertaking the design.

Unfortunately the rather basic questions you're asking and the apparent inability to understand the issues leads me to advise strongly that you employ a consultant experienced in design and operation of parallel positive displacement pumps to review and verify it.

There are several strange things going on here:

1) why are there 10 pumps in parallel if you state somewhere only three are in operation at one time?
2) Is the operation fixed and described in a philosophy or operating document?
3) How are these tanks filled and maintained full of fluid?

You seem to be concerned about the inlet side which is fair enough, but don't neglect the discharge side.

In general with piston plunge pumps, there are issues with acceleration head and pulsation which ar enot easy to investigate without some transient studies and specific analysis. In general however, the larger the header and inlet pipework, the fewer issues you will get.

SO to get back to your points
1) If you have two tanks in parallel then the level will not fall as fast as with one tank, but you need to look at the design conditions and make sure the system will operate at low level with only one tank open. If it can't then you need to either make sure both tanks are open or that you have a higher low level alarm
2) "offtakes" means the pipes coming off the header to the pumps. The guide is simply to give you an idea of how big the header should be ideally. space is an issue I can accept that, but can have an impact on operation.
3) "static" means not moving and "flow " implies movement, hence you can't have "static flow". You won't get high enough velocities to worry about damage to the pipe, at least on the inlet side.
4) There are many guides for inlet on pulsating flow and bimrs advice is good also. Generally the larger the better...

Like I said at the beginning, I can't see how you can verify a design when you're asking questions of this sort, sorry.


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