Vapor recovery unit (VRU). NPSH pump of rich absorbent

[Platonicus]

Hello!
I design a vapor recovery unit for a gasoline tank farm and a railroad loading and unloading ramp. The customer decided to use a vapor recovery unit (also for environmental reasons). In order to reduce costs, it was decided to use an absorption type VRU, which is an absorption column with a liquid tank and 2 pumps for lean and rich absorbent. Diesel fuel is used as an absorbent.

In many analogue projects developed by manufacturers of complete supplies of VRUs, the absorber has a low support (about 0.5-1.0 m or 1.64-3.28 ft). For this reason, the question arises as to how the pumping of the rich absorbent from the column is ensured without cavitation. Because the liquid absorbs hydrocarbon vapors, then, accordingly, it is in a state of saturation. Considering that conventional centrifugal pumps have a NPSH in the region of 2-3 m or 6.56-9.84 ft, it is suspected that there may be cavitation.

Therefore, I ask for help with understanding this aspect. Why don't suppliers make a higher support for the absorption tower? Did I correctly decide that the liquid is in a state of saturation and the NPSH, therefore, depends only on the geometric height of the liquid minus hydraulic losses? Or do they use some kind of special pumps that can work with a very low NPSH of the system?

 

[1503-44]

NPSHA for these things is mostly controlled by vapor pressure. NPSHA = Height x liquid density - pipe friction loss - vapor pressure. If you cavitate before reaching the pump, not all the height counts.

You will need every mm of height you can get. These liquids are very high vapor pressure and you will probably cavitate inside the suction piping even before reaching the pump intake. Raising the height of the unit just increases pipe length and resulting pressure drop, so it doesn't help much if any.

Speciality pumps for high VP liquids

https://www.corken.com/products_corken

For higher flows of hot gasoline you may need a can pump.

https://www.gouldspumps.com/en-US/Products/VIC/

 

[Platonicus]

That was the question. It seems to be a liquid in a state of saturation with gas. Accordingly, the difference between operating pressure and saturated vapor pressure is 0.

 

[1503-44]

If its a cold day, it might work. If its a hot day, you'll cavitate.

 

[LittleInch]

Even a casual look for absorber units shows that these are quite complex units, often using chilled absorber.

I think the key is that the sentence "Because the liquid absorbs hydrocarbon vapors, then, accordingly, it is in a state of saturation" is not correct.

The rich absorbent will have a higher VP than the lean absorbent, but maybe not very much and is not high enough to cause an issue.

Without knowing the precise make and model of the absorber and how it functions, this sounds like a question for the vendor, but I suspect my assessment is correct - i.e. the rich absorbent has a low enough Vapour pressure that you don't need the extra height.

Liquid coming from a 3 phase separator is a completely different thing.... Now that IS fully saturated.

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

 

[shvet]

See document attached - NPSH margine for dissolved gases.

https://files.engineering.com/getfi...es_for_Pumping_Liquids_with_Dissolved_Gas.pdf

 

[Platonicus]

The true vapor pressure of gasoline in the winter season can be equal to 100 kPa(a). Since gasoline will be poured into railway tank cars, light hydrocarbon vapors will go along with the air and saturate the absorbent to about the same value. VRU operates at a pressure close to atmospheric pressure.

 

[LittleInch]

Vapour pressure without temperature is meaningless.

It depends on the ratio better absorbent and vapour whether there is saturation or not.

You offer no more information so can't really go any further.

Why do you think this is a problem if the vendors of the package don't

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