This is my first post and hoping this topic will raise some discusion.
We have several rotary screw air compressors located throughout the US, operating temperature is around 180 and the ones in the southern states have issues with water contamination 2000-8300ppm based on Karl Fischer tests from our oil analysis company.
We are using Sullube 32 oil which is a Polyglycol synthetic and this oil also gets tested by our Compressor Services to validate the oil has not lost lubrication properties, and they test only for Acid number, viscosity, Methanol Isolubles. This insures we get their so called warranty. We know now that they only care about their lubricate and not our compressors, which is why we use 2 oil analysis compaines.
Our questions are:
What should the limit be set for water content? We've been told that a new drum of polyglycol probably has about 600-700ppm of water and our threshold should be set 500ppm above this (1100-1200ppm).
Second question how do we keep the water out? Desiccant breathers are an option but expensive. Another option is to raise the operating temp to 200-210 degrees to vaporize the water out and this probably would require a different oil since Sullube should operate between 180-200 degrees.
We are sure others have the same problem in these climate zones and any help is greatly appreciated.

Your limits for moisture seem too low. PAG lubricants made from EO/PO (ethylene oxide / propylene oxide)normally saturate with moisture at somewhere around 10,000 ppm. PAGs made from EO are water miscible. Exposure to atmospheric humidity will result in the PAG approaching saturation in 3 or 4 days.

Water does not readily react with the ether bonds in the PAG, so moisture levels less than saturation are not a problem for the lubricant. If the rest of the system is compatible with less than saturation levels of moisture, then the moisture levels you indicate in the in-service PAG may be acceptable.

Also in addition to refig guy's valid comments,
if the compressors are wet screws where the lubricant and air are intermixed then separated during the compresion cycle the water cannot be avoided as it will be atmospheric water and the difference between the two sites will be relative humidity in each area,
we watch 10+ large screws via WDA and chemical lab, the water is always there and as long hard particles are kept out normally causes no difficulty,

regards Rob S

Sullube, UltraCoolant and other Polyalkylene Glycol / Polyol Ester blend compressor oils will pick up water. These have about 25 to 40% POE depending on the viscosity grade and manufacturer.

The PAG used is polyproplyleneglycol (not the PO/EO or EO mentioned by others). At temperatures above 50F, the amount of water can normally be from1 to as much as 3% wiht this type of PAG. As such the Karl Fisher method is not used normally for air compressor applications with this oil.

PAGs have a certain degree of inverse solublility with water. So you might want to ask the compressor manufacturer if you can adjust the temperature up a few degrees if the water is an issue (accumulating). Be careful - ever 18 deg F shortens the life of the oil by about up to half.

If you are not experiencing corosion or other water realted problems, then there is no problem. PAGs hydrogen bond with water - and there is no issue with lubricity problems unless you see two phases in the lubricant feed to the compressor - meaning water to a certain degree is to be expected in these fluids.

As a background - th use of PAG with POE was first developed back in the late 70s by Dow. The idea was to combine the extreme long life of a POE (oxidatively) and to overcome the hydorlytic stability issues of the POE by adding the PAG which ideally draws the water away from the POE. Water being natural to air compresor applications (you are always at fully saturated conditions).

Today is is possible to purchase POEs that are much more hydrolytically stable than back then. However they are at a very high cost. They are made with branched acid POEs (avoiding low molecular weight acids in the production) branching also adds to oxidative stability. Degredation of POEs in air and moisture leads to their breakdown - creating organic acids (just like any organic oil - except POEs have a greater tendency). Properly made, the POEs resist that breakdown and can operate several multiples of hours over the other synthetics. branching protects the ester structure. the downside is a lower Viscosity index. But careful blending of branched acids and correct selection in making the polymer can overcome that for the most part. As I said - it adds to the cost (I think worth it as the result is a very long life oil). I have tested these to more than 24,000 hours in some cases - even at elevated temperatures.

Otherwise, the PAOs offer much better water separation if obtaine from a company that knows how to formulate them.

(35 years experience with synthetic oil formulations)