T he subject of valve emissions appears to be generating a fair amount of comment in some circles. As discussed last December, ISO 15848 is now the subject of a five year review.
While this may appear to be a good opportunity to generate some consensus this does not appear to be happening. There are probably four major testing standards that are of concern to packing and valve manufacturers: ISO 15848, Shell standard MESC SPE 77/312, the German standard VDI 2440 and the relatively new API 622.
Just to recap, the ISO and Shell tests are reasonably similar and are a test on a complete valve. The VDI and API tests are carried out in a test fixture on the packing. The other major difference is that the API test has a qualification level in ppm; the others measure a leakage rate.
There does appear to be a genuine will to try and unify the various test criteria and qualification requirements of the ISO, Shell and VDI standards, which would be an enormous help to the supplier industry. However, there is also a vociferous lobby supporting the API standard. As mentioned above there is a major technical difference between this and the other standards as it requires a measurement of gas concentration in ppm. While this is extremely relevant, particularly out on a plant, the fact remains that a concentration is not a measurement of leakage. While an approximate correlation can be established this is only valid in still air. If one requires a qualification of a given valve design with specific stem seal and body sealing then surely an actual leakage rate is required? And to help this along it would be very useful to agree some units of leakage that the average user can understand. Measurements in mbar.l/s are of less use than cubic football pitches/ millennium.
Trying to measure leakage rates out on a plant is extremely difficult and time consuming -I know from having tried it 30 years ago. Measurements with a sniffer of concentration are quite reasonable to establish the extent of a leak, but for initial qualification surely a leak rate is required?
The first step appears to be for the major users to get together and agree what qualification method they would jointly accept. It happened for mechanical seals on pumps; why not for valves?