Water in plant air line

[dnflsmsgksk]

I am looking at a plant air line with water problem.

The plant air originates from compressors with aftercooler that uses water as a cooling medium.

The plant air line has a pressure of around 93psig and temperature around 70degF.

I have checked drain location right next to the compressors and get no water.

However, at two ends of multiple branches, I get significant water around

1600 cubic inches and 1000 cubic inches of water from 2 minute drain test.

I was wondering if anyone else has looked into a similar problem

and would like to lead me in the right direction to the potential cause of this water. Thank you.

 

[Bribyk]

What's cooling the intercooler water?

 

[dnflsmsgksk]

I believe its just room temperature water. Since the outlet temperature is around 70degF.

 

[MintJulep]

If your cooler water exit temperature is 70F then your air temperature leaving the cooler is probably around 85 - 90F.

By the time it has reached the ends of the branch lines the air has cooled down to whatever the ambient temperature is, and the water has condensed.

Get a refrigerated or a desiccant air drier.

 

[zdas04]

What he said.

The amount of water vapor that is present in moderately humid air (I'm not talking Houston here, much less humidity than that) is staggering. At elevated pressures, the amount of water vapor possible decreases and dryers become economical.

Now you have to decide how dry you need the air. If you are just trying to prevent liquid accumulation in the line, then you need to lower the dew point to below your lowest expected pipe temperature. If all the piping is inside or buried, then the lowest expected temp is around 60F--reaching that dew point with chilled water is generally pretty economic. If the number is around 40F, then delequescent dryers (by themselves or in conjunction with a chilled water system) can work very well. If the number is 30F, then refrigeration can work and if it does the economics are ok.

Anything lower than about 30F at your pressure requires some sort of Temp Swing Adsorption (TSA) unit that generally have pretty poor operating economics. The problem with these units is that they do a really good job of capturing water vapor until they get saturated (time dependent on flow rate and water content, but is more often hours than days). Once they're saturated you have to apply heat to cook the water out of the bed. It is very energy intensive to apply that much heat with only gases for heat transfer media. It is doable, but really expensive.

David

 

[rmw]

Search this site for other threads that have dealt extensively with the topic. In addition to the excellent suggestions given, there is the phenomenon of ambient moisture diffusing into the pipeline at air leak points. It was discussed in depth a couple of years ago or so. I will leave the searching to you but do use 'diffusion' as one of your search terms.

rmw

 

[hacksaw]

if you are dealing with plant air as opposed to instrumentation air, water is always a problem, severa lgallons a day

driers would have been prohibitive

we had a resevoir in the system and simply put a timer operated electrical actuated valve on the tank drain, worked fine

you have to sort out th epiping from your description, the air coolers (water coolant) need to be arranged so that you do not get drainage back to the compressors

 

[chief]

You can fit an auto drain trap at the lowest point in your supply line ( see Spirax Sarco).
You can also install a Hydrocyclone type separator at the outlet from your cooler ( see Wright-Austin) (with an auto drain trap). This will remove approx. 95% of the moisture.
If you require the air for instrumentation purposes, you will need further drying ,either using a refrigerant type dryer ( Hankinson) or an absorption type using a desiccant with regenerative heating/

Offshore Engineering&Design

 

[zdas04]

Chief,
I think you need to be more precise in your terminology. Mechanical separation can remove a high percentage of liquid water, and zero percent of water vapor, so depending on where it is it might remove a lot of the "moisture" or very little indeed.

A desiccant bed is an adsorption unit not an absorption unit. This makes a pretty big difference in the method used to regenerate it (absorption media like sponges can be "wrung out", adsorption media for water vapor must be regenerated through elevated temperature.

I don't mean to bust your chops, but I see a lot of people buy high dollar hydrocyclone units and dessicant beds, put them in the wrong place and end up very unhappy with the results. You need to understand process constraints to decide what technology has the best chance of meeting goals.

David

 

[hacksaw]

Back to the cause,

We did a water balence based on psycometric readings of local ambient air. The amount of water obtained from buck tests was entirely due to ambient moisture, not leakage from the air cooler.

In compression of moist air, the moisture drops out other than the small amount needed to saturate the air at the new pressure and temperature.

Good luck

 

[FredRosse]

When humid air is compressed, much of the H2O in the air cannot remain in the mixture, and falls out as condensation, based on the "dew point" of the mixture. An ordinary Psychometric chart cannot be used here, as it is based on one atmosphere absolute pressure, and the compressed air is at several atmospheres. This is explained in any mechanical / thermo handbook.

Removing this water is often accomplished with a separator just downstream of the aftercooler, which removes most, but not all condensate. If there is no further drying of the gas flow, then residual condensate will accumulate downstream, and will also require draining at low points.

An air dryer installed downstream of the separator may be used to get further reduction in the "dew point" temperature. There are several types available, the selection depends on the size of the system. I usually recommend that "heatless" dryers not be purchased, they use way too much energy.