Aug 31, 2012

When a valve is tight enough?

The tightness discussion has been going around in the automation industry since the first valves were introduced.

Pipes by Neles

A lot of debate is involved with both control and isolation valve tightness; zero leakage, bubble tight, no leakage… well those are the dear wishes of many instrumentation people working daily to maximize process efficiency, reliability, safety and of course meeting environmental specifications. But what is realistic and economical to wish for? Continuous development is no doubt needed, as that is one of the main task of engineering; finding solutions with better performance and lower price.

Neles trunnion mounted ball valve

One very interesting article about control valve tightness is this, published in Hydrocarbon Processing August 2011 by Sanders, D. from GE Energy (access article here).

What about the isolation valve tightness in certain applications?

 Is it worthwhile in requesting tightness Class VI on applications were soft seated valves are out because of such demanding process conditions with high temperatures, erosion, and high-pressure differentials? One good example of such an application is molecular sieve valves, used in many industries from natural gas dryers in LNG/NGL gas plants to isomerization dryers in refineries and cracked gas dryers in ethylene crackers. The Class VI tightness requirement pops up every now and then to Neles as well. But isn’t it more the long-lasting tightness that this kind of application calls for? What is the true performance of the valve in molecular sieve applications where the valve is facing continuous shifts of temperature from 40 °C to 310 °C between processing and regeneration cycles, at a shutoff pressure of 20 to 50 barg and molecular sieve dust coming from the beds, that requires the use of hard, erosion-resistant materials that are porosive in nature, like chromium carbides (CrC)?

Neles has a long history with many installations at molecular sieve applications with trunnion mounted and seat supported ball valves with long-lasting performance, where usual tightness requirement is Class V. Testing Class V tightness according to standards, such as API 598 and FCI 70-2 matches rather well with the true operating conditions as the test is carried out with water at the rated pressure.

Testing Class VI tightness gives only limited information about the true performance at operating conditions because the specified test procedure requires the use of air or nitrogen gas at a temperature of 10°C to 50°C and at a pressure differential of 3,5 bar. How long can the tightness last with a metal seated valve at the operating conditions described above, rated as bubble-tight Class VI? Tested and passed, but for how long?

Quoting Mr. Sanders:

“Clearly, there are a multitude of factors that influence seat leakage measurement and performance in control valves. While there are various industry standards that can be utilized in the specification process, ANSI/FCI 70-2 is the most widely used despite its shortcomings. Given that a shop-based seat leakage test provides limited prediction of field performance, the end user is wise to understand the more durable attributes that lead to satisfactory seat leakage performance over the installed life of a control valve.”


Written by Sari Aronen. For additional information on the topic, please contact

This blog post has been up-dated in July 2020, due to company name change to Neles.