Pipeline sizing for fuel flow

[raisie]

Hi guys.

I have a situation I would appreciate some help on.

There are two gasoline storage tanks (Tank 1 and Tank 2), two pumps (Pump 1 and Pump 2). Currently, Tank 1 and Tank 2 are connected to Pump 1 through a 4 inch pipeline. For future operation purposes, an additional connection will be made from Tank 1 and Tank 2 to Pump 2. Thus, in the future operation mode, fuel will be supplied to from either Tank 1 or Tank 2 to both pumps simultaneously. This means that, in order to deliver the nominal flow to both pumps, the flow through the 4 inch pipeline will double. My colleagues seem to think that the 4 inch pipeline will have to be replaced with a 6 inch pipeline in order to optimally deliver fuel to both pumps and are trying to justify this thought with calculations. With “optimally” I mean that both pumps receive a flow rate equal to their nominal flow rate continuously. The question is, which calculations do you perform to check this??? (Please see the attached sketch)

I think that the size of the pipeline has nothing to do with whether or not both pumps can be provided with their nominal flow, because once the pipelines are filled with fuel and the tank has enough fuel, the pumps will just keep receiving the needed flow through suction. However, I think that due to the doubled amount of required fuel flowing through the same size pipeline, the speed of the flow will be larger and THAT is why the size of the pipeline must be increased. So basically, now I’m looking for a code/standard which specifies recommended speeds of fuels through pipelines.

Am I right on this so far??

 

[georgeverghese]

NPSH adequacy calculations should be carried out whenever there are changes in suction piping flows to pumps, and line diameters should be adjusted to suit.
The NPSHa at the pump suction flange should be higher than that required by the pump (NPSHr)at the corrsponding flow by a margin that is usually specified by the plant operator - if there isnt one, suggest 1m margin as a minimum. This NPSH calc should be done with tank level at low low trip level.
In addition, believe there is a max permissible liquid velocity requirement for some fuels - believe this is applicable to jet fuel.
Suction lines should also slope continously down towards the pump with no pockets in between. In some high capacity pumps, the pump vendor may request for a minimum straight length upstream of the pump.

 

[LittleInch]

raise,

The size of the pipeline has an effect. The issue is one of vapour pressure. To move the fuel along the pipe to the pump you need a difference in pressure. The driving force for the flow is atmospheric pressure (approx. 1 bara at sea level), which reduces with altitude (this can be important). The lower pressure is formed by the pump flowing and creating pressure lower than the static pressure to create flow.
As George mentions above, the calculation is termed NPSH (Net Positive suction Head) and the NPSHA (available) must be at least 1m higher than the the NPSHR (required) which is a figure obtained from the pump curve. If NPSHA< NPSHR, your pump will cavitate and not work properly as it will have a mixture of liquid and vapour coming into it.

The factors involved in the calculation are:
(A) Static height of the lowest liquid level above (or below) the pump inlet ( this is usually a positive number (elevation of liquid - elevation of pump inlet nozzle)
(B) Vapour pressure of the liquid at the temperature you are flowing at. This is a number in bara. For gasoline this can be quite high and is the (sub atmospheric) pressure at which the liquid boils. This is a chart or figures available from your supplier or on the internet (it can change with season so be careful)
(C) Friction in the pipe. Whilst the other two are fixed, this varies by diameter and flow rate. Hence a larger pipe will have less friction

NPSH is basically A -B -C

If your system is as shown and the length of the line from the tank to the junction quite small compared to the rest, then you might get away with it as only the section from tank to branch will see double flow compared to your current situation.

There is no code or standard for this, only the calcualtion I explained above. As a guideline, velocity in lines like this are often 2-3m/sec and rarely more than 5m/sec, but if you have long lines from the tank, you might need to drop this 1/sec or less or lower your pump into a pit.

Hope you understand this, but if not come back to us. Some idea of the lengths on your schematic and the flowrates would be very useful this time and next time you post..

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

 

[georgeverghese]

Actual vapor pressure of some of these finished petroleum products may be well in excess of published values, and must be used with caution, especially with products that are stated to be of low published vapor pressure at max ambient temp.
Deviations in actual sat vap pressure in some of these less volatile streams may be due to contamination with lighter fluids in heat recovery schemes as the product runs down to storage (ie tube leaks in heat exchangers), dumping of lighter components into some products because there is no where else to dump it, adding lights to some heavy products to increase flowing characteristics etc.
Instrumented blending controls for products leading to tankage are often prone to breakdown due the nature of field analysers.

 

[LittleInch]

Surprised no one noticed my error, but in the NPSH calc it should be

NPSH = A - B -C + D, where D is the atmospheric pressure (converted to metres head) acing on the top of the liquid, i.e. corrected for altitude.

All figures should be expressed in metres head of the liquid in question.

So raisie, does this answer your question??

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

 

[raisie]

Hi guys,

Thank you all for your responses.
I am already doing the NSPH calculations.

However, my main concern here is that when both pumps are on and "sucking" from one tank, is there a chance that one of the pumps will "starve"? Please keep in mind that one of the pumps is much further from the source (tank) than the other, which means that one of them has a larger suction head.
Or should I just assume that this will not be a problem, as long as the NSPH conditions are fulfilled for both pumps?

Thanks in advamce everyone.

 

[LittleInch]

Pumps don't suck.

So long as there is enough NPSH available then it shouldn't matter. If pumps are "identical" then the one with the longest run might have a slightly lower discharge pressure, but probably not noticeable.

If the common section of line is quite short compared to the single flow and it works now then you should be ok.

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