Compressed Air Purity Checks in Pharmaceutical/Medical Facilities

Specifications
The Pharmaceutical industry generally adopts CGA Grade D as the specification for compressed instrument air, but it is poorly written. Your air can be sampled, tested and found to be compliant with the specification, and yet still result in major problems. The following discussion is intended to help you discover the true condition of your compressed air.

Sampling
There are two rules in sampling compressed air:
(1) Do it in a way that can discover solid debris (desiccant dust, charcoal dust, and metallic debris) as well as any oil that may have accumulated on the interior of the piping. This always means obtaining an air sample at a high flow rate (10- 20 SCFM or faster). Not all sampling kits can sample the air at a high flow rate. Do not confuse pressure with flow rate; pressure is only something to control in order to develop a high flow rate. It is the fast moving air that helps you discover oil and debris in your air system. Our kits sample at reduced pressure but high flow.

(2) Always capture enough air to have a good statistical sample, and to provide your laboratory with enough air to perform several analyses. If you think you have a problem, but you do not have enough sample left, the laboratory will not be able to solve or even verify the problem. Generally, sampling kits that use small vials can never do more than one analysis of a simple specification because their vials typically hold only 30 cc of non-pressurized sample .

Analysis
Although there are at least 4 ways to monitor compressed air purity, Pharmaceutical and Medical facilities require data that is generally best obtained with sensitive instrumentation normally found in laboratories. The laboratory techniques commonly used are Gas Chromatography (including GCMS) and Infrared Absorption Spectrometry (IR). Although the IR technique has some unique advantages, the technique is not universal and is not as specific as GC techniques.

In the Pharmaceutical industry, three contaminants of major significance are moisture, particulate and non-volatile hydrocarbons. Moisture is best determined with a dew point meter (also called a hygrometer). Meters that can measure moisture down to a dew point of -78 ºC are readily available. The meters have sensors that do not last forever, and when they fail, they do so in the “dry” state (meter reads dry even if the air is wed). Therefore, it is necessary to occasionally check their reliability. We use a color indicator tube (Dräger Tube) to accomplish this.

Non-volatile hydrocarbons (compressor oil) are removed from the compressed air stream when the air passes through several physical filters (molecular sieves and charcoal). These filters do not last forever, and in the event of a premature failure, the oil mist would not be discovered until an oil-related problem is detected.

Quantification of oil in compressed air can be accomplished two ways: (1) a color indicator tube that is sensitive to oil, or (2) examination of a membrane filter through which a large amount of compressor air has passed. The tube works by developing a color or stain within the tube when the sulfuric acid in the tube contacts the entrapped hydrocarbons, but not all compressor oils will react with sulfuric acid to form a stain.

A technique that we developed is to extract membrane filters (through which about 4 cubic meters of air has passed) and analyze the extract. This is the technique we use to detect very low levels of oil in your air. It also can detect the oil in the presence of particulate.

The ACI Universal Air Sampling Kit: Model RD2-137
With all of the above considerations in mind, we developed an air sampling kit that more than meets the above criteria with our patented Sampler block. The kit can accomplish many tasks involving a wide variety of situations. Some of these are as follows:

(1) The kit can sample your air at nearly any flow rate (2 to greater than 20 SCFM), and it does so safely. The equipment is safe because it has 2 failsafe components that are active up to 3400 psi.

(2) The sampling block has a built-in holder for the Dräger Moisture tube to determine dryness down to 4 ppmv moisture (dew point -70º C), and it does it in 10 minutes.

(3) The membrane depth filter collects oil mist and particulate down to 0.7 microns. By extracting and concentrating the filter extract, we can measure oil levels equivalent to 0.01 mg/cubic meter based on the volume of air (usually 4 cubic meters) that went through the membrane filter. (See attached data.)

(4) An optional filter (0.45 micron porosity Millipore type) can be used with the kit for applications requiring a filter that can be used for particle counting by microscopy.

ANALYTICAL CHEMISTS, INC.
7551 Convoy Court, San Diego, CA 92111 Voice & FAX (800) 466-8898

ATTN :
FAX :
TEL : LABORATORY # :
DATE REPORTED :
DATE RECEIVED :
DATE SAMPLED :
SAMPLED BY :
AIR SAMPLE VOLUME : cubic ft sampled at SCFM
IDENTIFICATION :

COMPRESSED PROCESSING AIR*
TESTS PERFORMED ALLOWABLE LIMITS PRESENT IN THIS SAMPLE CONCLUSION
TOTAL NON-VOLATILE RESIDUE (NVR) 1 ppmv
TOTAL NON-VOLATILE OIL HYDROCARBONS 2 mg/m3
TOTAL VOLATILE HYDROCARBONS3 ppmv
MOISTURE4 Dew Pt ºF – 50 ºF
MOISTURE4 mg/L
ODOR5 Description None
PARTICULATE6 Description
FILTER APPEARANCE7 Description N.A. Clean & Dry
*Compressed Processing Air is very clean air used in manufacturing parts. It is not a test for compressed breathing air.

1. This is the same as the graimetric test for Oil Mist + Particulate; it is essentially an NVR test.
2. This is a high sensitivity test for only Oily Hydrocarbons (down to 0.005 mg/m3). Chemlube 751 is the standard.
3. Total Volatile Hydrocarbons (TVHC) measures solvents and fuels. The standard is methane.
4. When air becomes wet it indicates that the primary filter is spent. Wet air may suggests a water dump valve problem.
5. Odor is the first significant clue of oil; you can smell it on the filter even before you see the yellow spot.
6. The approximate number and size of particles on the membrane filter can hint at the condition of the particle filter.
7. The type of particles indicates the source of the particles: charcoal, mole sieves, brass pipe steel or stainless steel.
8. Archiving the filters can reveal normal gradual piston wear or sudden excessive wear.

COMMENTS BASED ON OUR OWN TESTS

Data : Air Log xxx
Invoice : 27xxx-A
Code : ast
Review :
David H. Elgas
Laboratory Director