11 replies to this topic

[Neelrahs2006]

Hi All,

Just wondering does anyone know what is the maximum pump suction line length? I have read from somewhere that it is around 10m? Is that really the limit? The pump i am currently dealing with is a rotary pump. The reason i have asked, is because currently, i have a existing plant that requires me to route a pump suction line for about 100m from tank to pump with some elevation (crossing roads).

I have doubt whether this will work or not. Hydraulics can be used to calculate it but, in practically wise, i am doubting it.

Can someone share a thought about this with me?

Regards,
Shar

[Art Montemayor]

Shar:

From the manner in which you form your questions, I have to assume you are a student and not a professional engineer. What you ask is very, very basic fluid mechanics.

You cannot address the length of a suction line to a centrifugal pump (and I have to assume that since you also fail to state the type of pump) without taking into consideration the fluid’s vapor pressure, the flow rate, and the static head plus the source tank´s vapor pressure. All these parameters form part of the usual NPSH equation and should be taken into consideration when discussing the system’s NPSH availability.

Additionally, if you are going to have high spots within the line, you will have to always consider that you MUST establish AND MAINTAIN a 100% liquid-filled suction line 100% of the time. Depending on the complexity and the specifics of your application, you may not be able to make this proposed system work.

You certainly will require detailed and accurate calculations before you attempt to do this.

[sheiko]

The function of suction piping is to supply an evenly distributed flow of liquid to the pump suction (typically by providing the suction side with a straight run of pipe, in a length equivalent to 5 to 10 times the diameter of that pipe, between the suction reducer and the first obstruction in the line), with sufficient pressure to the pump to avoid cavitation ," and related damage ," in the pump impeller (by ensuring that the Net Positive Inlet Pressure available, or NPIPa, is always higher than the NPIP required, or NPIPr).

Note: NPIP is usually used for rotary pumps (your case) while NPSH (Net Positive Suction Head) is used for centrifugal type.

NPIPa = minimum source pressure + source elevation pressure (at the lowest source elevation) - suction pressure losses (sum of the equipment, instrument, piping specialty, straight pipe, fitting and valve pressure losses between the source and the pump suction) - vapor pressure (at the highest pumping temperature).

To sum up, the maximum pump suction line length is the one that will minimize the NPIPa. For most pump, it is safe to consider that the minimum NPIPa is equal to 1.3 times the NPIPr of your pump (or use the safety factor recommended by your company. The higher is the safer).

Good luck!

Edited by sheiko, 25 October 2009 - 12:26 PM.

[Neelrahs2006]

Thanks Sheiko. Hi Art, I am not a student. But you could say that i am considered a junior engineer. The reason why i ask for only the maximum suction length is because i have considered all hydraulics calculation, but in logical and practicality, i wanted to know how long it could be.

Our company's practice is to have a 1m margin between the NPIPa and NPIPr. Sheiko, could i get a reference to the minimum NPIPa equal 1.3 times the NPIPr of the pump?

But even if i can fulfill the NPIPa through calculation, the practicality of it?

To "MAINTAIN a 100% liquid-filled suction line 100% of the time" how is this ensured? Can i say the line is always filled as the outline of the tank is always below the LLL.

Regards,
Sharleen

[sheiko]

Sharleen,

Note that NPIP is in pressure unit not in head unit (you said that your pump is rotary), so the NPIPr margin should be in pressure unit.

As for the reference, please read: Budris A.R., Designing pump piping, Chem. Processing, Aug. 2002 (http://www.chemicalp...s/2002/263.html) which deals with centrifugal pumps. I have asssumed the margin still applies to rotary pump but check it with your supervisor.

Rgds.

Edited by sheiko, 26 October 2009 - 02:15 PM.

[shan]

No, you can not assume a 100% liquid-filled suction line simply because your tank outlet nozzle is always submerged into the liquid. The pressure drop due to friction or elevation change may make the pressure at some points are lower than vapor pressure of the fluid in the suction line. Therefore, vapor will be flashed at the points where the pressure is lower than the fluid vapor pressure.

[Art Montemayor]

Shar:

Shan is entirely correct and shows foresight in warning about the effects of a high vapor pressure liquid - especially if it saturated. I would also reiterate my warning about having any high spots within the line. Any high spot will have a tendency to accumulate vapor and/or absorbed non-condensables in the liquid. Should this happen, it will destroy the ability to maintain the suction line flooded and 100% liquid - something that is detrimental to any pump´s operation.

[sheiko]

Shar:

Shan is entirely correct and shows foresight in warning about the effects of a high vapor pressure liquid - especially if it saturated. I would also reiterate my warning about having any high spots within the line. Any high spot will have a tendency to accumulate vapor and/or absorbed non-condensables in the liquid. Should this happen, it will destroy the ability to maintain the suction line flooded and 100% liquid - something that is detrimental to any pump´s operation.

Just for info. there is an interesting article related to this discussion: Chen C.C., Cope with dissolved gases in pump calculations, Chem. Eng, Oct 1993

Edited by sheiko, 26 October 2009 - 02:13 PM.

[Shahnawaz Makroo]

Dear friend.

Maximum suction lenth for different cases will be different there is no such generaliztion but you cal calculate how much length at pump suction is permisible for particular case.

steps to calculate:

1.Obtain value of NPSHr from pump vendor or rotating department of your company.
2.Add margin over it by 1.0 m or more to fix NPSHa at pump suction.
3.From basic formula of NPSHa
NPSH a = Psource (m)+ suction liquid head (m) -Vapor pressure (m)@pumping temp. - line loss (m)

line loss (m) =Psource (m)+ suction liquid head (m) -Vapor pressure (m)@pumping temp - NPSHa

= equivalent pipe length from pressure drop calculation formula(delp = f(L/D)(v2/2g)).



I hope answer of ur question.


Regards
SAM

[kkala]

To the valuable contributions already made, following may be useful.
1a. First condition to be realized is a generous NPSHa-NPSHr (or NPIPa-NPIPr) margin. In refineries there may be suction lines over 1000 m long, e.g. from light fuel tanks to truck loading pumps.
1b. However risks like: locked air/gases in spots, suction pipe not full, semi-clogged pipe if slurry is handled,
increase as suction length increases. All suction line, from tank nozzle (with antivortex inside?) to pump inlet, should be below tank low - low liquid level and it had better be straight, of constant (downward) slope. This in addition to NPSH requirement 1a.
1c. In case of atmospheric suction tank, deviation from 1b "suggestions" may still result in generous NPSH margin, but the pump may not prime, especially for 100 m suction length.
2. I would worry if the suction line had to cross roads, and look into two matters:
2a. Does local regulations permit this passage, and with what protective measures? It depends on the kind of liquid handled and the kind of road (public, private, in refinery, etc).
2b. Is any "suggestion" of para 1 above deviated? Most probably yes, as understood from the description, but there is also possibility of underground routing.
3. Knowing what liquid is handled would probably help.

[patron]

Dear all,

I'm a rookie and interested in this discussion,

In refineries there may be suction lines over 1000 m long, e.g. from light fuel tanks to truck loading pumps.

Is it a common practice in refinery design if I place a smaller tank right before the suction point just to make sure the pump suction always filled with liquid and prevent the cavitation and water hammer effect?

[kkala]

Dear all,

I'm a rookie and interested in this discussion,

In refineries there may be suction lines over 1000 m long, e.g. from light fuel tanks to truck loading pumps.

Is it a common practice in refinery design if I place a smaller tank right before the suction point just to make sure the pump suction always filled with liquid and prevent the cavitation and water hammer effect?

No it is not a common practice, as indicated from local (Greek) refineries.
My understanding is that this is also a safety issue. A tank (even small) near the pumphouse contains a stock of flammable liquid that would affect pumps and vicinity piping in case of fire.
For this reason liquid fuel tanks are in dikes, while pumps are placed out of these dikes, usually not far from them. Suction line (with isolation valve on tank) can be (say) up to 250 m long in such case.
Truck loading pumps can have much longer suction lines because they are placed at the station where trucks are loaded with fuel; operator should supervise both pump and truck in this way.
So far remote operated truck loading pumps (i.e. far from the station) have not been heard of.