A common problem that occurs when releasing chlorine from cylinders during industrial chlorination processes is that of liquefaction of the gas and ice deposition in and around the cylinders which slow the rate of withdrawal of chlorine from the cylinders. Thermodynamically, its a legit process when chlorine will undergo expansion, with the help of Joule Thompson coefficient, it will cool down.

My issue with the reasoning is that since the expansion is occurring at the valve itself and adiabatically, there shouldn't be cooling inside the cylinder yet I do have pictures of a cylinder with ice deposited all over its surface due to expansion via cooling. How do we explain the fact that the cooling effect is travelling back through the expansion nozzle and more importantly, how do we calculate the cooling rate and total energy removed from the cylinder via expansion through nozzle?


2 Answers 2


You have it backwards. The major part of the cooling occurs within the cylinder itself, with very little cooling (if any) occurring in the valve due to Joule Thomson.

The gas remaining within the cylinder at any time has expanded adiabatically and reversibly to do work in forcing the gas exiting the cylinder ahead of it into the valve. On the other hand, the gas passing through the valve is experiencing both expansion cooling and viscous frictional heating, so that the net effect is much smaller.

To calculate the cooling effect within the cylinder, one merely needs to apply the open system version of the 1st law of thermodynamics to this adiabatic expansion into the valve. For an ideal gas, this would read $$nC_vdT=RTdn$$ where n is the number of moles of gas within the cylinder at any time.

  • $\begingroup$ But Cl2(l) evaporation must play role as well. It's boiling point is just about 6 deg C below b.p. of propane. The pressure cannot be very high, so gas expansion thermal effects would be much weaker than e.g. for air, N2 or O2. $\endgroup$
    – Poutnik
    Aug 18 at 20:29
  • $\begingroup$ Yes, that would have to be taken into account that too. I just gave the ideal gas as a simple example. $\endgroup$ Aug 18 at 20:35
  • $\begingroup$ I see... Ideal gas examples never hurts. $\endgroup$
    – Poutnik
    Aug 18 at 20:37
  • $\begingroup$ Correction, remembering chlorine and propane b.p. wrong. 8 deg C above propane. $\endgroup$
    – Poutnik
    Aug 19 at 7:18

Gases, that at room temperature condense before reaching sufficient storage pressure, are provided in the liquified form with gaseous headspace, what allows much better storage capacity.

Liquid chlorine has the vapor pressure $\pu{5830 mm Hg}$ at $\pu{25^{\circ}C}$, what is about $\pu{7.7 atm}$. Having cylinders with just few atmospheres of gaseous chlorine would not make much sense.

Adiabatic or Joule-Thompson expansion cooling has minor impact here. Additionally, considering what the output chlorine pressure may be, there may not be much expansion at all, depending on application.

The major effect is evaporation cooling of liquid chlorine, based on its evaporation enthalpy. The similar effect is observed on the outdoor camping LPG burners. When cooking, you can feel how cold the thin wall LPG cartriges become.

Therefore chlorine is even more easily liquified than propane. With an intense output flow, the boiling chlorine may reach deep frost temperature.

If the pressure regulator has equal or lower temperature than the liquid chlorine - e.g. for the minor expansion effect, then the gaseous chlorine will condense there. Similarly as water vapor condenses on glasses, if you enter a vapor chamber in a spa, or if you enter a warm room from cold outside during winter. Mild heating of regulators may help.

Dew or frost deposition on cylinders indicates another issue, as low cylinder temperature significantly decreases the gas pressure. (LPG cartridges are often kept in warm sleeping bags overnight, as butane boiling point is near $\pu{0^{\circ}C}$.) It would help if there was applicable some technical way to keep cylinders at room temperature.

  • $\begingroup$ So you are saying that, in the case of chlorine, under typical starting conditions, most of the chlorine will be liquid? $\endgroup$ Aug 19 at 11:39
  • $\begingroup$ Poutnik is right on this one so I downvoted your answer Chet. I calculated the total cooling by plotting taking adiabatic work done for the duration and the cooling effect from adiabatic expansion is nowhere near the cooling effect I observed but I started talking to the guy who does HVAC on site and he also hinted at the liquefaction of gas and cited the LPG example. I checked the pressure of the Cl2 and it was comfortably above 10 kg. So, the gas was in liquid state inside the cylinder. I calculated the evaporative cooling and the answer was legit. $\endgroup$ Aug 19 at 16:28
  • $\begingroup$ @Poutnik Now you've made me look stupid in front of the whole plant. I overdesigned my chlorine handling system badly. I was more comfortable with other compressed gases like N2 CO2 etc and they have higher pressures like 80 bar or above so I got a 35 bar Pressure relief valve. I apparently have installed a show piece in the line. I will plug it tomorrow. First the overdesign with cooling tower pumps and now this $\endgroup$ Aug 19 at 16:35
  • $\begingroup$ I am sorry for that. :-) $\endgroup$
    – Poutnik
    Aug 19 at 16:42
  • $\begingroup$ Hypothetically, putting aside chlorine reactivity and chemical weapon quality, it could be used in fridges or AC, similarly as halocarbons or ammonia. $\endgroup$
    – Poutnik
    Aug 20 at 7:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.