Genesis is spot on here with his analysis.
One must remember the visual humidity indicator is measuring relative humidity at pressure hence you are concentrating the water vapour in what would be very dry air, in the ball park of 0.07 percent relative humidity (RH) at 1 atm at 20 C, to the 10 percent RH at the pressure and temperature one decides to read the humidity disc at. Ten percent RH at 207 bar is a much lower RH at 1 bar if at the same temperature.
Quarterly testing for moisture levels in breathing gas is a reasonable minimum level of moisture monitoring, but real-time moisture monitoring is better. The same goes for any of the other breathing air contaminants. Using the drinking water quality analogy even with the best purification system one could monitor every three months for fecal coliforms and most of the time find the water quality would be fine. There will come a time though when for various unforeseen reasons fecal coliforms will be found in the drinking water and people will get sick. We now monitor drinking water supplies supplies at least daily and in many large cities every six hours.
The problem is much can happen in the 3 month interval between quarterly air quality tests to leave your desiccant excessively moisture-saturated. Once the desiccant bed is moisture-saturated it is no longer able to protect the activated charcoal bed which relies on dry air to effectively remove the volatile hydrocarbons from the gas stream. In fact an on-going private research project using Drager tubes to monitor moisture levels at fill stations in Canada, Mexico, and the Caribbean has shown that at any one time about 50 percent of fill stations have breathing gas with moisture levels above the 35 mg/m3 allowed in EN 12021. Most of these stations are not checking moisture at all, but many are using quarterly testing which has proven insufficient, particularly in areas with high ambient temperature. People are not making the required temperature corrections to their filtration processing capacities, and hence they are exceeding the end-of-service life hourly ratings in much less time and not aware of this. The gas stream once the desiccant is moisture-saturated is no longer protected from contaminants entrained into the compressor either from the ambient air, or from those produced within the compressor typically from off-gassing of the lube oils. These are gaseous hydrocarbon off-gasing products, and not liquid oil mist which is detected on the Drager oil mist tube. There is no Drager tube for gaseous hydrocarbons on the market which is broad-spectrum or sensitive enough to be used to monitor breathing gas quality.
One way around this tedious temperature/hour logging is to install a real-time moisture monitor. If you look at the low pressure compressor systems used for industrial breathing air and for some of the hook-up umbilical systems the filter housing can be made of clear plastic. The manufacturers put a humidity strip in the filter which one can read through the filter housing’s thick clear plastic wall. When the humidity indicator changes from blue to pink the operator knows in real-time it is time to change out the filter. This works well on the low pressure compressor filtration systems.
Unfortunately with our high pressure breathing air compressors the filter appliances are made of aluminum or steel and not plastic, and so one must physically depressurize the system and remove the filter cartridge in order to look at a humidity strip in the filter cartridge. This is the system Lawrence Factor (LF) uses on many of their filters. There are two problems with this. First having to depressurize your filtration system, and pull the filter to see what the strip is doing is like having your auto’s gas gauge in the boot (trunk) of your car. At some point you are going to forget to check the gauge and run out of petrol (gas). At some point something will happen such that the strip turns pink well before you expect it to rendering the gas stream unprotected from potential contaminants. Personally I have seen this happen on two occasions despite temperature/hour logging. On one compressor the o-rings went on the auto-drains, and on another someone accidentally put a tank on the condensate drain line. Unexpected things do happen.
Secondly the LF strips are designed to change at 40 percent RH which is too moist if one wants to meet the moisture requirements for EN 12021 or CGA Grade E. If one speaks with Mike Casey at LF he will say the cartridge humidity strips were only meant to be used in multi-tower systems such that when the first tower hit 40 percent RH you would change out all the downstream towers at the same time knowing they were somewhere less than 40 percent RH and hopefully in the 20 percent RH territory. Unfortunately many operators are using these LF filters in solo tower filter systems which is not how these cartridges with humidity strips were designed to be used.
To make life much easier and not have to pull filter cartridges in order to determine your desiccant’s moisture status, the external visual humidity indicator came along which allows one to monitor moisture in real time. It simply has removed the humidity indicator from inside the filter cartridge, and placed it in an external position on the purifier’s discharge line. It is an inexpensive, simple, and safe way to monitor moisture and gives better information than just a quarterly Drager test. One now knows on a day to day basis what is happening to the desiccant bed within the purifier. Depending on one’s budget there are many ways to monitor moisture levels in real-time from this inexpensive visual humidity device to the Bauer Securus capacitance hygrometer to very expensive ($US1600) high pressure dew point meters from
Nyad Inc. which require regular calibration of its sensor. The only caveat when using the visual humidity indicator is it must be read at pressure and so one must have a pressure maintaining valve on the system set at a high enough pressure to concentrate the remaining moisture after the final cartridge. A pressue gauge is required between the final separator and the PM valve to monitor the moisture indicator's pressure.
Here are some charts which will allow you to convert from RH percent to mg/m3 as used in Europe or ppm as used on this side of the pond.
Hope that helps a bit to explain what is a very poorly understood aspect of breathing gas quality monitoring. The document posted by Brent from the HSE is in fact for low pressure compressor systems and does not apply to the high pressure breathing air compressor systems which follow much tighter moisture limits.
http://www.afcintl.com/pdf/Dew.pdf