Pump data sheet specified "pressure" versus "head" and confusion ensued 2

[SNORGY]

I just want to make sure I am not missing something.

A pump data sheet for a PD pump specified three operating conditions: low flow, normal flow and rated flow, with suction and discharge PRESSURES expected for each. The PD pump was long lead, so a very well known vendor quoted a centrifugal pump with suction, discharge and *differential* conditions reported in the same PRESSURE units, based on pumping water. The centrifugal pump was purchased.

The actual fluid pumped is light hydrocarbon condensate, SG = 0.51.

The vendor is of the opinion that the pump will be able to achieve the same delivery PRESSURE as long as the suction PRESSURE is the same. I do not share this opinion. I think it will deliver the same HEAD at about half the PRESSURE. In which case, it might spin but other than that, might not work at all.

Have I got this all wrong? If not, I am a little disappointed with a set of pump curves beautifully and elegantly spat out by vendor's software with TDH reported in units of "barg", referenced to water.

No, I was not involved in the design, although as a favor I am being asked to review it.

 

[1gibson]

Never ever trust head in units of pressure for a centrifugal pump (and don't trust a vendor that gives you a curve like that.) You are correct that the differential pressure will be about half. But if you decide to pump water instead, you'll be just fine...

As a side note, that means that the NPSH margin might not be as expected, and vendor probably does not know what the pump thrust will be. Could be an issue depending on the pump type, and motor thrust bearing may be oversized, again depends on pump type/size. Conditions for the mechanical seal will be off if .51 SG fluid was specified, with "water" pressures.

Stop the bleeding, escalate to vendor's engineering manager and get someone who knows that they are doing to look at it. Yikes.

 

[SNORGY]

1gibson,

Already done. Equally disturbing was the suction line sizing, fluid at vapor pressure moving at 20 ft/s through 300 feet of pipe before arriving at the pump. I suggested going up about 3 line sizes there.

Pump seal system is Plan 53; so yes, now that you mention it, I can see that their predicted seal chamber pressures require review as well. Thanks for the confirmation and the help.

 

[LittleInch]

Oh dear.

It is difficult to comment on whether it will actually work or not as if a PD pump was originally specified, this implies that flow rate was the key factor to be achieved with a fixed or variable pressure requirement, presumably by variable speed drive or some sort of wasteful bypass line. However you have now got essentially a constant pressure unit.

It is difficult to understand how an experienced vendor didn't pick up the SG from the pump data sheet and as you have given us no data we can't check this, but if we assume that you are not mistaken, then if the pressure differential was quoted in differential head based on water then I believe you are correct, i.e. your inlet head will be substantially higher (x2), but the differential head is fixed for both fluids, then discharge pressure in water will always be higher than that of the condensate. Exactly how much will depend on the ratio between inlet head and differential head.

Using round numbers, if inlet is 10 bar, then water = 100m and your condy 200m. Your pump adds 100m head for both - total 300 for condy, 200 for water translates to 15 bar condy, 20 bar water. As you increase differential head it gets worse in terms of difference. The differential head is the same regardless of contents, but the pressure rise is a fraction of water based on SG.

If you need a certain minimum pressure before flow occurs - we have no idea about your system - then yes, you might get no flow.

I think you're talking to the sales guy not the technical sales / design person....

Can you post the data sheet and pump curves and a bit more detail?

Sounds like a new pump to me.... or a bigger impellor

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[SNORGY]

LittleInch,

Thanks. It is a real present day problem, so that plus the fact I am somewhat well known in these parts (locally), I'll post more detail if I can find a way to do so without betraying my client's or the vendor's rights to confidentiality. The system is for LPG truck loading. In truth, the pump only needs to do enough to get LPG to the truck at a pressure that allows vapor back to the LPG tank at a compatible rate. The problems that might show up will be with the calibration and settings on the in-line devices (control valves, transmitters, switches) on the LPG loading skid, into which the pump is being free-issued by the pump vendor. It might be a rescuable system but there is obviously a miscommunication between "requirements" and "offering". The latest revision to the correspondence shows the pump pressure and motor speed both doubling and the horsepower tripling even though the fluid density is halved. So, things are going in an entirely wrong direction; hence my review. My OP was just a gut check to make sure there wasnt something fundamental about the behavior of centrifugal pumps or my interpretation of the curves that I was missing.

 

[LittleInch]

Hmmm, Sounds like they're trying to speed it up and possibly with a bigger impellor to get the pressure up to what you want.

Not sure that's the best way to do it, but I can't see anything wrong with your assumptions about this - looks like a screw up by the pump vendor.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Artisi]

Does this help

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[Artisi]

Second help.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[TenPenny]

If the fluid SG was specified on the datasheet, the vendor has no ground to stand on.

Any centrifugal pump guy will produce info based on head, and sometimes will put pressure units in parantheses for information.

 

[BigInch]

The difficult part of these problems is that LPG pumps often draw from a rather low pressure tank with little head and vapor pressure reduces your NPSHa to well below nothing even with a very short run to the pump.

If pressure was specified, the engineer has no ground to stand upon.

you must get smarter than the software you're using.

 

[SNORGY]

Artisi,

Yes, that's how my Cameron book reads as well. I just wanted to see if something was amiss in my interpretation because, when I was asked to review this, I couldn't believe what I was seeing, and I thought, "OK, what am I missing here?", like, maybe their software automatically corrected for head versus pressure or maybe their "barg" meant something other than what I thought it meant.

As for the data sheet, the original data sheet was written for an API-676 pump such that suction and discharge PRESSURES were specified coincident with flow rates and fluid SG for the three cases given. The problem came when people read the 20+ week delivery and asked the vendor to quote something lower cost and shorter lead. The vendor quoted a centrifugal pump and rated it for water based on the original data. All I was supposed to do was review the NPSHA - it took me about 5 minutes to advise that there were larger causes for concern than NPSH.

I have a few things on the go today but I will provide a summary of this situation so that other posters are not left in the dark - I apologize for that. This "head vs pressure" issue comes up a lot in my work and it continues to surprise me how often people - otherwise intelligent people - get tripped up by it.

 

[GHartmann]

I was blessed to be taught to do all my pump requirements in feet of liquid pumped. This has kept me out of trouble for over 30 years. Unfortunately, even the API-610 data sheet mixes pressures and head. These type of conversions will often lead to error.

I had to create my own Excel spreadsheet to keep myself straight between the units.

The LPG applications are also tough because of the low NPSH requirements.

I have seen Corken utilized often in this application (on tank trucks, etc). Not sure of their delivery or flow range covered.

They utilize a sliding vane design. The Corken website has all types of good info and LPG transfer, etc.

Corken

GOOD LUCK!

We look forward to hearing how this turns out.

G.Hartmann
BS ChE NC State University