5 replies to this topic

[aushulz]

I have a stream of (approximatively) this composition:

• 60.3% mol water
• 32.7% mol oxygen
• 5.4% mol HCl
• 1.1% mol VCM
• 0.3% mol nitrogen

• Flowrate: 257 kmol/h
• Temperature: 500 K
• Pressure: 3 bar

So, I want to separate essentially oxygen from water.
What is the better method to do this?

I tried on SimSci Pro/II with a flash drum, but didn't work (all the product remain on vapor phase).

Some suggestion?

[MrShorty]

Not too surprising that they all stayed in the vapor. You need to calculate or look up bubble and dew point curves for this. At 500 K, the vapor pressure of pure water is 25-30 bar: at 3 bar it would all be a vapor even if it was pure water. In order to separate them by some kind of vapor-liquid separation operation, you're going to have to locate the two phase region so you can operate at those conditions.

hope that helps.

[aushulz]

Not too surprising that they all stayed in the vapor. You need to calculate or look up bubble and dew point curves for this. At 500 K, the vapor pressure of pure water is 25-30 bar: at 3 bar it would all be a vapor even if it was pure water. In order to separate them by some kind of vapor-liquid separation operation, you're going to have to locate the two phase region so you can operate at those conditions.

hope that helps.

I tried to change stream conditions at 298 K and 1 bar, but nothing changes.
I tried to draw on Pro/II the T-xy diagram at 1 bar: at 400 K it indicates a mole fraction of water (dew point) of 250000!!
If I try to draw a P-xy diagram at 500 K, it appears this warning: "Bubble point flash did not solve. Final temperature(k)= 500, pressure(kpa)= 578490, lfrac= 1, function= 0, tolerance= 1e-5. (a dew or bubble flash in the supercritical region is an arbitrary phase change that is expected to fail.)."

[aushulz]

Not too surprising that they all stayed in the vapor. You need to calculate or look up bubble and dew point curves for this. At 500 K, the vapor pressure of pure water is 25-30 bar: at 3 bar it would all be a vapor even if it was pure water. In order to separate them by some kind of vapor-liquid separation operation, you're going to have to locate the two phase region so you can operate at those conditions.

hope that helps.

I tried to change stream conditions at 298 K and 1 bar, but nothing changes.
I tried to draw on Pro/II the T-xy diagram at 1 bar: at 400 K it indicates a mole fraction of water (dew point) of 250000!!
If I try to draw a P-xy diagram at 500 K, it appears this warning: "Bubble point flash did not solve. Final temperature(k)= 500, pressure(kpa)= 578490, lfrac= 1, function= 0, tolerance= 1e-5. (a dew or bubble flash in the supercritical region is an arbitrary phase change that is expected to fail.)."

Using NRTL or UNIQUAC correlations instead of Redlich-Kwong, the liquid-phase diagram is fine.
Do you now the reason of this?

[milenkom]

Not too surprising that they all stayed in the vapor. You need to calculate or look up bubble and dew point curves for this. At 500 K, the vapor pressure of pure water is 25-30 bar: at 3 bar it would all be a vapor even if it was pure water. In order to separate them by some kind of vapor-liquid separation operation, you're going to have to locate the two phase region so you can operate at those conditions.

hope that helps.

I tried to change stream conditions at 298 K and 1 bar, but nothing changes.
I tried to draw on Pro/II the T-xy diagram at 1 bar: at 400 K it indicates a mole fraction of water (dew point) of 250000!!
If I try to draw a P-xy diagram at 500 K, it appears this warning: "Bubble point flash did not solve. Final temperature(k)= 500, pressure(kpa)= 578490, lfrac= 1, function= 0, tolerance= 1e-5. (a dew or bubble flash in the supercritical region is an arbitrary phase change that is expected to fail.)."

Using NRTL or UNIQUAC correlations instead of Redlich-Kwong, the liquid-phase diagram is fine.
Do you now the reason of this?

It shouldn't be that difficult to separate air from water. It's quite easy, all you need to do is to cool down the product below dew point of water. Use some kind of knock out drum. Water will condense and fall to the bottom of drum. Also please note that HCL in normal state is gas, which is highly soluble in water. So when water condenses HCL will dissolve in water (ionization). So in order to simulate this properly you will have to use electrolyte package. Neither NRTL nor UNIQUAC won't do the trick here as all HCL will be vented off as gas.

Edited by milenkom, 17 February 2010 - 12:58 PM.

[aushulz]

Not too surprising that they all stayed in the vapor. You need to calculate or look up bubble and dew point curves for this. At 500 K, the vapor pressure of pure water is 25-30 bar: at 3 bar it would all be a vapor even if it was pure water. In order to separate them by some kind of vapor-liquid separation operation, you're going to have to locate the two phase region so you can operate at those conditions.

hope that helps.

I tried to change stream conditions at 298 K and 1 bar, but nothing changes.
I tried to draw on Pro/II the T-xy diagram at 1 bar: at 400 K it indicates a mole fraction of water (dew point) of 250000!!
If I try to draw a P-xy diagram at 500 K, it appears this warning: "Bubble point flash did not solve. Final temperature(k)= 500, pressure(kpa)= 578490, lfrac= 1, function= 0, tolerance= 1e-5. (a dew or bubble flash in the supercritical region is an arbitrary phase change that is expected to fail.)."

Using NRTL or UNIQUAC correlations instead of Redlich-Kwong, the liquid-phase diagram is fine.
Do you now the reason of this?

It shouldn't be that difficult to separate air from water. It's quite easy, all you need to do is to cool down the product below dew point of water. Use some kind of knock out drum. Water will condense and fall to the bottom of drum. Also please note that HCL in normal state is gas, which is highly soluble in water. So when water condenses HCL will solve in water (ionization). So in order to simulate this properly you will have to use electrolyte package. NRTL nor UNIQUAC won't do the trick here as all HCL will be vent off as gas.

If HCl is solved on water how can I eliminated it as vent gas, which unit operation I need?

In Pro/II there is a table for thermodynamic systems such this (on square parenthesis there are my assumed model):

• K value (VLE) - [UNIQUAC]
• K value (LLE) - [None]
• K value (SLE) - [None]
• Liq enthalpy - [Library]
• Vap enthalpy - [Library]
• Liq density - [Library]
• Vap density - [Ideal]
• Vap fugacity - [Ideal]
• Liq entropy - [SRK]
• Vap entropy - [SRK]

Which element of the list I need to change?

If I choose "electrolyte" correlation, a lot of messages like this appears: "Component 2, 'oxygen', is not valid with electrolyte thermodynamic model hcl (set = hcl01). If this component enters a unit operation that uses this model, an error will result."