If you introduce water in known amounts in the argon gas holder, it will evaporate by itsedlf, until the saturation is reached at the temperature at which the holder is kept.

So simply introduce known amounts of water in the gas holder, starting with small amounts. The water will evaporate by itself. Then cool down the gas holder wall until dew appears on it.

Dear Claude-Alain Roulet,

I'd like to thank you for your kind reply. How can I introduce water in the argon gas holder? do I need a special kind of equipment?

Claude-Alain Roulet

You need of course some entry in the gas holder. One way is to use a syringe with a needle though a small rubber cap.

Another is to have a sas with two valves. The steps are as follows:

Close the inner valve and open the outer one. Make vacuum in the sas and then fill it with argon at the pressure of the gas holder and with the required amount of water.

Close the outer valve and open the inner valve.

I have a question: The boiling or dew point of water depends only on the pressure, and not on the other gases of the gas mix when they don't chemically react with water (see https://en.wikipedia.org/wiki/Vapour_pressure_of_water). Why measure the dew point of water in argon, which is chemically inert? The dew point in argon will be the same as in air with the same content in vapour pressure, and will depend only on the partial pressure of vapour in argon-vapour mix.

Dear Claude-Alain Roulet,

Thanks again. I really do not want to measure the dew point in the argon environment. I name it as an example. I want to measure it in a hydrocarbon mixture.

Reza,

I would think the balance in this context must be methane rather than argon. Of course beacause of explosion dangerous, one can use N2 instead of Ar.

Cloude-Alian,

As you pointed out the gas temperature, gad pressure and water content in the gas are the main factors which can change the WDP in the gas mixture.

Theoretically gas flow must not change the WDP, however, practicaly the gas flow rate in the close system such as pipe or sampling system of the DP Analyzer may cause errors in the measurement of WDP.

I would say in this context the flow rate must be carefully considered the measurements.

We need to measure 1 to 40 ppm water content in the natural gas (say 90% CH4, up to 7% other hydrocarbons C2 to C5 and 2-3% Co2 and N2).

Would you please give us an advice in terms of experimental set-up to measure the WDP in this context.

Thank you very much for your time.

If I understand you well, the problem is to measure the water content of natural gas in a duct in which the gas flows at a given flowrate, and the velocity of the gas perturbs the dew point measurement. With 1 to 40 ppm water vapour concentration, the dew point should be between -75 and -50 °C at atmospheric pressure or between - 65 and -36 °C at 5 bar.

The direct way to measure the dew point is to cool down a polished surface exposed to the gas and detect (e.g with a laser beam which is diffused by the dew) the temperature at which the dew appears on the surface. The polished surface can be in good thermla contact with the duct wall, and the laser source and detector can be inside the duct. I don't expect a problem with the gas velocity in that method.

A problem may appear with butane and pentane that boil respectively at -1°C and 36 °C if their concentrations are high enough to condense above the dew point of water.

The main issue is to distinguish the WDP and HDP of natural gas in one integrated system say chilled mirror or polished surface coupled to laser and detector as you mentioned.

Imagine in the gas pressure between 1 to 50 bar mainly hydrocarbons are condensed earlier than water vapour and one would need to distinguish them.

Do you think the intensity of light received from detector must be different due to water and hydrocarbon(s) condensation on the same surface? Probably because of their different surface tension or viscosity?

Do you think this indicator (different light intensity) makes sense to measure them unambiguously?

Since the surface tension of water and hydrocarbons are not the same, the shape of the droplets will certainly differ, and this will have an influence on the light scaterring patterns. The lens-shape of water droplets (on a weetting surface) will focus the laser beam, which will then be diffused, while perfectly spherical droplets (such as those from non-wetting hydrocarbons) will reflect it back. Therefore, the diffused light intensity measured perpendicularly to the incident light beam should mainly result from the diffusion from water droplets.

However, I have no experience on this issue and therefore cannot give a straight answer. I propose to look at the scaterring patterns of pure water and pure hydrocarbons on a polished metallic or glas surface to get this information.