3 phase separator fire relief

[ChemEngSquirrel]

I’m doing a fire case relief calculation on a three phase (oil / water / gas) separator using the API 521 method:

Is it fair to treat the liquid phase (oil / water) as a mixed solution for calculating overall latent heat of vapourisation?

Given that the two phases are immiscible, I’m not sure if this is a fair assumption.

Thanks.

 

[TD2K]

If there are significant volumes of oil and water, then I do the calculations assuming both phases are at the same temperature and both exert their individual vapor pressures.

I have the total Q and knowing the vapor composition (from the vapor pressures), I can estimate the amounts of water and oil being vaporized to develop the physical properties necessary to size the relief valve.

If I have a vessel with say a water boot then I just use the HC inventory as I would argue the water would be quickly vaporized due to the limited inventory. Likewise, for something with a aqueous phase with a thin HC layer I would likely ignore the HC layer. Alternatively, you do both sets of calculations and go with the larger area.

 

[don1980]

TD2K's advice is good.

If there could be a significant volume of either fluid, then I'll size for the one that requires the largest orifice. In almost every case, that's the HC rather than the water (assuming the HC isn't a heavy oil with a relieving temp >> vessel design temp).

 

[MortenA]

TD2K, Im not quite sure i follow you,

Say that the boiling point of the HC liquid is below that of water at the relief pressure. So water wont boil. What i would do would then be to calculate the HC vaporisation rate based on the latent heat of the HC fluid (initial composition). What you are saying then is that you would determine the vapour pressure of water at the relief conditions and then "add" a proportional amount of water? Would you accont for the additional latent heat - or is it small compared to the HC?

Best regards

Morten

 

[123engr]

I do prefer to size the PSV using the latent heat of vapourization of the HC liquid instead of water.

My past shall not limit my future....

 

[TD2K]

Sorry Morten, missed your question to me.

Depending on the boiling points of the HC and water, it may not make a lot of difference.

The example that brought this to mind was a caustic extractor. The PSV was set at 400 psig (fire was the only relieving case) and there was a significant inventory of caustic in the base of the vessel. At relieving pressure, we calculated that about 10% of the vapor stream was water and the remainder hydrocarbon. We knew the heat input into the vessel and could calculate how much of each phase was being vaporized to give the vapor stream and then used the bulk properties to confirm the PSV capacity. The other check we did was to calculate how long it would take the vessel's contents to get hot enough to begin to vaporize at 400 psig which was another layer of credit the client was willing to take.

Taking credit for the higher latent heat of water is definitely something to discuss with the client.