Selecting a pump for HC tanks heal transfer (high Vapor Pressure, very low NPSHA)

[ZahidM]

Hello Dears,
Please advise what type of pump/pumping system is best for the following duty. Kindly share your experience as well.
Vendors please if you can offer a solution.

We need to often decommission large size hydrocarbon product tanks for inspection by transferring liquid from one tank to another tank (heel/slop transfer).
The tank products are like gas condensate, naphtha, kerosene, diesel, gas oil etc.
We are looking for an appropriate portable pump package which can do the job safely and reliably.

The pump is required for the following operating conditions.
Products: gas condensate, naphtha, kerosene, diesel, gas oil etc.
Pressure: 0.93 BarA
SG: 0.66 - 1
NPSHA: 1 m or slightly above
Flow: 100 m3/h or close
Pressure: 4 barg
Presence of some solid particles

Thanks in advance.

 

[Artisi]

Why are saying NPSHa is 1 m +?


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.)

 

[LittleInch]

You'll struggle to do much better than an AODD.

Relatively small, air powered so no issues with electric cables or engines

Flow at 100 m3/hr might be a bit big, but see what you can find.

NPSHA at 1m though is very low - there is very little that will do that, but the AODDs don't mind a bit of two phase flow.

If you're basically sucking va=pour then just slow the pump rate down and let it flow into the pump.

At 0.93bara vapour pressure you're into very volatile material.

If you are really pumping rubbish then an option is a progressive cavity pump but they tend to be very long (3-4m) but they accept any old material. If it can get through the pump inlet flange then the pump will move it along and out the outlet flange.

what do you use now?

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

 

[ZahidM]

Thanks for the reply LittleInch. I think you got a good idea of what our issue is. We use AODD which give low flow rate around 20 m3/h so it takes many days to empty the tanks remaining volume. The AODD pumps at this duty has reliability issues as well which also adds to the time.
Progressive cavity pumps we thing would not be able to completely empty the tanks so leaving a considerable quantity of liquid in the tank when the pump will start loosing suction. As a portable system, it will be complex as well.
Side channel pumps can work at low NPSHA but we have some solid particles as wee at tank bottom products.
I wonder how people in industry do empty their large HC product tanks like ours.

Dear Artisi, NPSHA of 1m is based on, the tanks bottom height which is about one meter above ground level, some head loss in suction piping and then subtracting head due to vapor pressure of 0.93 barA which is the main negative one.

Thanks for helping.

 

[Artisi]

Have you tried a vacuum assisted self-priming contractor pump.
Check net using above description.

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.)

 

[ZahidM]

Artisi, thanks. A vacuum assisted self-priming pump I think is not safe and suitable for our application bcz while initial priming, lot of HC vapors plus air mixture will be ejected which is an explosive mixture. Secondly, while pumping it continuously require higher NPSHa than what is actually available.

 

[georgeverghese]

The largest I can find in the metal diaphragm version for Wilden is the 4inch T20, which has a max capacity of approx 60m3/hr if there is adequate NPSHr. In your case, if we apply a notional 50% derate for inadequate NPSHr, flow will be approx 30m3/hr, which is far from the 100m3/hr you want. So is there space to drop 3 suction hoses down from the tank top in to the heel of these tanks to run 3 pumps?

 

[LittleInch]

At that level of NPSH / volatility (0.92 bara) there isn't much you can do other than pull it out in buckets (seriously).

Your only other option if this is a reparative task is to excavate or drill a hole down and insert a pipe and then drop in a Can pump and attach that to the pipe.

I am very surprised your products are that high in vapour pressure other than the gas condensate.

Naptha, kerosene, diesel etc are normally quite low vapour pressure in a storage tank.

Only other thing which works is to lower in a submersible into the sump.

If you run it slowly enough it shouldn't cavitate too much.

A diagram of your tank with elevations and lengths / diameters of pipe etc would help.

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

 

[ZahidM]

georgeveghese, thanks for taking time and suggesting Wilden T20 AODD. Our tanks dia are in the range of 60-80 meter and the liquid level to be removed is about 1.5 m, making it 7500 m3 volume. So with an average flow of 30 m3/h would take around 10.5 days to empty the tank. That too if the pump keeps on operating continuously and does not stop or breakdown, which is not the case.
Can we operated two or three of these AODDs in parallel taking suction from a manifold and discharging to another manifold?

LittleInch, you are spot on once again. I have confirmed that the NPSHA is even less than 1m because:
NPSHa = (head due to atm press) + (net height of the pump centerline from tank bottom) – (friction head of piping @ say 100 m3/h flow) – (vapor pressure head)
NPSHa = (1*10/0.66) + 0.5 – XX – (0.933*10/0.66)
NPSHa = 15.15 + 0.5 – XX – 14.13
NPSHa = 1.514 – XX where XX is head loss due to piping friction. Drain line is 3 inch dia with an isolation valve, to be connected to the pump through hose, length say up to 10 m.

You are correct in saying that we need to dig a pit and install a pump there. This is how the main transfer pumps are (5 meter below G.L).
As you hinted, we probably need a different arrangement/pump for condensates and another arrangement/pump type for low vapor pressure liquids like kerosene and diesel etc.

Progressive cavity pumps though should work for other than gas condensate service BUT they are very long for required duty, I think must require a PSV at discharge, a dry run protection and a tandem seal with API seal plan 52 as a minimum for our case. All these with a portable configuration makes this option also complex. What are your views on it?

Do you have any experience with Side Channel Pumps? Can we use them for condensates? Websites though mention that they are not good for liquids with solid particles because of close clearances.

Thanks dears.

 

[georgeverghese]

If you were to run 3 Wilden pumps in parallel, guestimate on suction manifold would be some 8inch or 10inch.

 

[Artisi]

How about a sketch for the heights, positions of pipe / access points etc. you have given us nothing other than a bunch of numbers.

What is the pump centre height from the tank bottom?

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.)

 

[LittleInch]

Well 80m diameter is about 500,000 bbls storage so pretty big.

And then the designers give you a 3" line from the drain to empty it??

You really need to find some bigger drain lines somewhere or gradually retrofit at least a 6" or 8" connection to the drain point.

I'm assuming these are floating roof tanks at that sort of vapour pressure, but it's higher than most people would use floating roofs for.

Never seen a side channel pump before, but if you've got some dirty, gritty fluid it doesn't look good and they don't look very high capacity.

No I think you need to dig a big hole somewhere close to the tank and either drop in a drain vessel and pump it out from there or drop in a can pump. And find some bigger pipes. Or drop in a submersible if you can get a manhole entry and can handle the fumes.



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

 

[ZahidM]

LittleInch,
Thanks again, I agree with you. Other than AODD, next option for condensates is as what you mentioned. How about kerosene type products where we have about 10 m of NPSHA, can we use a centrifugal pump? Discharge side head head will be varying depending on the distance and level of the discharge side tank. So the pump will be giving a varying flow, which can be fine though. Also, do we have to provide a minimum flow line in that case? Centrifugal pump in portable configuration will be simpler I think.

Artisi, sorry that I gave only general idea but close to actual situation. If you have noticed, I even left the piping & flow hydraulic losses as uncalculated XX which have to be subtracted from 1.5m to know the NPSHA. Anyway I am attaching now a sketch.

I am so grateful for all of you trying to support.

 

[Artisi]

Having NPSHa 10 metres for the kerosene / diesel is now a different story - and completely changes what has already been posted by others.

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]

ZahidM: yes, you can use a centrifugal pump for the kerosene and diesel, but trust you have plenty of time available if pumping from a 3", line.

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.)

 

[LittleInch]

Your real problem for anything other than your condensate is that 3" nozzle / pipe.

It's far too small for the flow and even if you get to 100m3/hr, you're looking at 3-5 days to empty tanks of that size with 1.5m of heel left in it.

I've seen modifications which used low level nozzles having dip pipes installed to approx. 100mm off the floor or they wait until the level goes below a manhole then open the manhole and either drop in a submersible or a hose / pipe to pump to pump more out faster. Usually needs a tank outage to fit them but at least next time they empty faster.

BTW with a EFR tank I'm surprised you can store product with a vapour pressure as high as 0.93 bara. That's close to the liquid boiling creating a vapour block and sinking the roof or having high emissions.

Normally a limit of 0.75 bara (11 psia) is used as the highest vapour pressure for a EFR tank.

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

 

[ZahidM]

Artisi, thanks. So what do you think, how to manage minimum flow line if we use CF portable pump on Kerosene like products?

littleInch, thanks again. I took Vapor pressure numbers from MSDS and data sheets of main transfer pumps. The tanks have floating roofs, I checked the drawings.
I have given up on Condensates, only Largest capacity AODDs to be used no matter how much time is taken. We will try to run more than one if possible and reduce down time.

What are your views for kerosene type products, what is the best option in portable configuration? I wish to use a centrifugal pump but not sure how to manage the minimum flow line and for mechanical seal I think tandem seal with API plan 52. Will fixed speed pump be an issue too?

Regards

 

[LittleInch]

Where you have such wide discharge pressures I would use some sort of PD pump myself - maybe a screw or piston with a built in re-circ line.

Centrifugals are good but can be limited in terms of max to min operating pressures.
Why are you worried about re-circulation / min flow? This will just take even longer.
Seals I leave to pump suppliers.

Fixed speed is OK.

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

 

[ZahidM]

Then the only PD pump I can think of is Progressive Cavity pump for this application bcz of some solid particles present in the slop. So no matter how complex it will be in portable configuration, we will have to go for it. Screw does not like solid particles as I know.

For a centrifugal pump, my understanding is that at least minimum flow needs to be ensured somehow to avoid overheating of liquid inside the casing which would be an unsafe condition. I do not recall seeing a CF pump on critical application without having a min flow line. So for my application, I am not getting an idea how to deal with this requirement.