Mr. Pellinghelli, do you think that SVRP (SIS valves reliability prediction) is properly understood by people who are working within flow control?
I believe that in recent years all the issues related to security and the ways to implement it and therefore the relevant knowledge have increased. All the companies who want to have an important role in flow control have incorporated the rules on the reliability of components and they are qualified to perform probability of failure calculations according to standards like IEC 61508. So there have been some important steps forward but, considering the importance of this issue, I would say it is important for this knowledge to be increasingly implemented.
If that knowledge is not properly implemented, what might the consequences be?
Obviously it depends on the application but it can be very serious. Just think of the failure of an emergency valve or a High Integrity Pressure Protection valve (HIPPS). These systems are often among the last lines of defence against potential injury or fatalities to workers and environmental or community disasters.
Have there been any such instances of failures to date as regards safety instrumented systems?
Speaking about pneumatic and hydraulic actuator applications I am aware of several end users who have had some failures on their plants. In a previous role I was asked to look at the failure of some actuators fitted to emergency valves on an offshore platform. Fortunately there were no serious consequences as the facility was provided with redundant safety valves. I am also aware of an investigation into degraded pneumatic actuators on platforms following an incident where a spring can failed due to corrosion.
Can you identify some of the main causes of SVRP issues?
Within my own field, namely pneumatic and hydraulic actuators, I can mention several as follows:
The technical specifications issued by engineering departments are sometimes not sufficiently clear and defined which can lead to different interpretations that often do not give sufficient consideration to safety issues.
- Reference standards for actuator qualification have been issued in the past 10-15 years but they need to be implemented in order to ensure a high level of product qualification. And actuators with an emergency function should really have dedicated reference standards because of their importance.
- The final tests used to test the complete system (e.g. actuator + valve + control system) are often not performed under actual operating conditions and are not sufficient to detect eventual problems.
- The aging factor and corrosion of components that occurs under actual operating conditions is not always taken into consideration nor properly assessed. An emergency actuator, for example, can remain stationary in the compressed spring position and submitted to severe ambient conditions for years before being called on to operate, which it is a very heavy duty service. The qualification methods are often based on historical data yet such figures may not be reliable or may not cover the actual operating conditions and do not guarantee that the system is field proven, This is definitely one of the possible causes of failures.
What about your customers? Do they ever ask about actuator reliability?
Our direct customers are normally the valve maker or the engineering contractor, and so not the actual users of the product. The applications may be very different from each other. In most cases it is required that the actuator should be in compliance with the product standards and of course be in the customer’s vendor list. In other cases we are subjected to real additional qualification tests defined by the engineering company or by the final customers.
For emergency applications additional requirements are normally applied, such as third party certification of the product reliability level. There tends to be an assumption that the actuators will work as expected if they are supplied with the right “paperwork”. Some background aspects such as correct selection and configuration tend to be taken for granted, even if it is clear that the best Probability of Failure on Demand (PFD) rating for example is not much use if the actuator is wrongly sized with respect to the valve.
It is important to understand that the product certification is not sufficient in itself to guarantee reliability. The actuator needs to be correctly selected based on the specific functions and operating conditions, correctly sized to the valve, correctly configured in terms of control equipment, and finally properly built, installed and maintained. For this reason it is important to refer to suppliers that can guarantee the quality of their operations across all functions.
Which standards are available that cover reliability?
The most widely used standard for actuator reliability measurement is IEC 61508 . Actuator qualification is based on this standard and witnessed by a third party authority. Other standards such as EN 15714 also prescribe endurance testing to guarantee the longevity of the actuators. Individual actuators should then be tested for torque output, cylinder tightness, spring thrust and general functionality.
What is Limitorque’s policy about actuator testing?
All our hydraulic and pneumatic actuators are tested. As a starting point we follow a standard procedure which we always inform the clients about in advance. If they have any specific requirements they can of course be implemented.
How are you making your actuators more reliable?
Thanks to a major investment by Flowserve in R & D in Italy we have been able to redesign our pneumatic and hydraulic actuators to make them more suitable to meet the current requirements of applicable actuator standards as well as of the market. We have looked to improve actuator reliability through superior design solutions. Further, we have improved corrosion resistance, worked on the efficiency of the torque transmission parts and we are now implementing new spring status monitoring and PST solutions.
Are there any actuator designs which you would advise users not to specify in critical applications?
Bearing in mind that you cannot make general rules and that various cases should be examined individually, we can say that in many cases, where emergency valves are concerned, single-acting actuators have a greater reliability compared to double-acting actuators and that scotch yoke actuators are more suitable than rack and pinion actuators. I can also add that the actuators directly coupled to the valve (without an interface adaptor) are preferable.
Finally, what advice would you give to a customer who wants to have absolute confidence in the reliability of his safety instrumented systems?
I’d recommend that buyers should always ask for quotations from various qualified suppliers and evaluate the offers by placing an appropriate balance evaluation between the technical characteristics of the product and the price. Unfortunately on many occasions it is the price that is the most important variable and that does not favour safety. Further, it is important to write very clear and unambiguous technical specifications in order to simplify the evaluation of multiple quotations.
Before starting with technical and production activities it is also necessary to plan meetings with the valve and actuator manufacturers in order to be sure that all the points (in particular the most critical) are clarified. Buyers should also be clear about their requirements for the final testing of the actuator and valve assembly.
And finally, they should plan periodical visits to the major suppliers to verify their efficiency and quality assurance in order to maintain valid and up-to-date supplier qualifications.
Mr. Ferruccio Pellinghelli
Engineering Senior Supervisor at
Limitorque Fluid Power Systems in Italy