H2S equilibria

[sevy]

Hello,

I am working on modelling a neutralization tank where a strong acid is dosed. This will of course lower pH and shift the sulfide equilibrium. I want to specifically know how much H2S (g) will be produced in the headspace of my tank.

I have started by modelling the reaction by titrating my caustic solution and letting it interract with a GAS phase with a fixed volume. This is true in practice but what happens is the pressure increases in the headspace, which in reality does not happen because in my case there is an offgas vent to prevent pressure build-up beyond 1,5 bars. I noticed that pressure build-up and constant volume affects my H2S gas-liquid equilibrium. Does anyone have any suggestions or comments to help me generate reliable results?

Any response would be much appreciated!

PS: I tried to equilibriate with a constant pressure GAS phase but the method won't converge. Besides PHREEQC defined this option for gas bubble which form and stay dissolved so I do not think it is the right way to go for me anyways...

[dlparkhurst]

Seems like the constant-volume gas phase is the way to go until a pressure of 1.5 atm is reached.

I would switch to a constant-pressure gas phase after that addition of acid (probably in a separate calculation that starts with that addition of acid). The volume will increase above your tank volume, but the solution composition should be correct. In detail, you would remove enough of the gas phase to reduce the gas volume to the fixed volume of the tank (at a pressure of 1.5), but in fact, this should have no effect on the solution composition.

There is a GAS_PHASE_MIX keyword that is not documented; however, this problem revealed a bug that I need to fix. In the next version of PHREEQC, you should be able to do something like the following to decrease the volume of gas phase 1 to 20 percent of its current volume. However, you don't know the total volume of the gas phase until after the calculation, so you either have to iteratively emend the input file to reduce the gas volume, or do the calculation in a script using IPhreeqc. In any event, the solution composition should be the same, so it is probably not worth doing.

GAS_PHASE_MIX 1
1 0.2
END

[dlparkhurst]

On second thought, my answer is not right if there are gas components other than H2S. Certain components could be removed from the gas phase at different times during the sequential addition of acid, which could, in fact, affect the evolving solution composition.

[sevy]

Thank you for you response dlparkhurst.

Indeed, I forgot to mention that the headspace of my tank has an initial blanket of fuel gas to keep a minimum pressure of 1 atm. So my initial 1 atm - constant volume (headspace) gas phase is mostly composed of methane (as other hydrocarbon gases are not in the database). Reaction temperature is 55°C.

After getting my desired pH, the pressure builds up to 5.73 atm. This is much lower than any of the saturating vapor pressures of H2S, water vapor or CH4. What I do not understand (and this might well be a theoretical notion I am not seeing) is that if I double my headspace volume, then more H2S (g) moles will be present in the gas phase. This is disturbing because at the end of the line, I want to know how much hydrogen sulfide gas will be produced due to the neutralization.

[dlparkhurst]

t is not exactly right, but I think it may be close enough, just to use a fixed pressure gas phase.
Suppose had a balloon that expanded to 5 liters and was in equilibrium with the solutions. Then you
released enough gas to bring the volume down to the head space of the tank. The solution composition would
not change as you released that gas.

Where the calculation may be wrong, is that gas incrementally escapes from the head space. The first gas
that escapes probably is mostly methane and a little bit of H2S. Later, the fraction of H2S probably
increases and methane decreases. So the gases will fractionate as you loose more and more gas. To model this fractionating process, maybe you want to sequentially add a little acid, create a gas/water equilibrium, and release
enough gas to reduce the pressure to 1.5 atm. You can do this with a sequence of REACTION and GAS_PHASE_MIX (requires version 3.1.4 or later) calculations:

SOLUTION
END
GAS_PHASE 1
-fixed_pressure
-pressure 1.5
-volume ?
CH4(g) 1.0
END
REACTION 1
  HCl 1
  xxx moles
END
RUN_CELLS
  -cell 1
END
# repeating section
GAS_PHASE_MIX
  1   ?
END
RUN_CELLS
  -cell 1
END
repeat ...

It will require a run to determine the mixing fraction that is needed in the GAS_PHASE_MIX definition at each step. After the run (up to GAS_PHASE_MIX), the mixing fraction for GAS_PHASE_MIX can be determined; it will be the calculated head space volume divided by the calculated gas volume. It will vary depending on the amount of acid added in the reaction step. You can compare how the solution composition changes as the reaction steps are decreased.

I'm not sure that what happens in the real-world tank is well defined. It depends on the rate of gas evolution and the rate of gas release through the valve. I think the 1-step versus infinite steps should bracket reality.

Alternatively, you can use a MIX calculation instead of REACTION to simulate the addition of acid.