A post from a few years ago, Separate Versus Integral Final Element Accessories in Safety Applications, continues to generate some good questions. I thought I’d make these questions and associated answers more visible in today’s post.
Here’s the first question:
I understood the whole description, But while calculating PFDavg, We will have to add the Failure rate of valve assembly (Valve+Actuator) which brings down the total PFDavg (Sensor+Logic solver+Final control element). I don`t understand that, How providing redundancy at the accessories will bring down the PFDavg. As I understand the LAMBDA DU value for SOV and Pneumatic positioner is much lower than that of the Valve and Higher LAMBDA DU of the valve is responsible for low PFDavg. On contrary, I think by adding more accessories like SOV would increase spurious trip.
I ran this question by Emerson’s Riyaz Ali, who was featured in the original post. Here’s the reply from Riyaz that I posted:
PFDavg will be equal to ג du x T2 /3 (Note: ג du is dangerous undetected failure rate of Equipment under control (EUC) and T is test interval). 1oo2 provides improved PFDavg over 1oo1 single box approach, where PFDavg will be equal to ג du x T /2, using ISA TR84.00.02-2002 part 2 simplified equation).If pneumatic path has two devices to exhaust in case of Demand, either or (DVC or SOV) device can take the valve to safe state, improving reliability of “Final Element”.
Since use of reverse B relay, any electrical failure of DVC will not cause the valve to safe state. MTTFs (mean time to fail spuriously) is just as good as of single SOV, classified as Type A device.
The next question came in yesterday:
The article mentions that many process manufacturers still opt to use solenoid valves (SOVs) pneumatically mounted in series with the digital valve controllers, which is a safe and conservative thing to, however is there enough field data yet to support the use of smart positions on their own? The reliability figures for SOV and for DVC6200 SIS are those comparable? Will it be possible to achieve the same SIL level, when SOV replaced by the positioner? Or by positioner + quick exhaust valve [QEV], for fast closure requirements? Would much appreciate if you can share your experience in this area. Thank you.
Smart positioners are certified standalone as SIL suitable for use in SIF loop. However, process industry still prefers to use SOV (considered as Type A device per IEC61508, part 2, Table 2) pneumatically in series with Smart positioners with Reverse Relay to provide spurious trip protection.
Applications where Process Safety Time is critical from stroking speed time consideration than use of Volume Booster (Larger orifice meeting stroking speed requirement) is much preferred when using smart positioner standalone instead of QEV. It is highly recommended not to use QEV while using smart positioners for partial stroke time as QEV gets activated with very small DeltaP, causing uncontrolled travel during partial stroke test.
A great place to continue the questions and interact with other safety experts is in the Safety Instrumented Systems group in the Emerson Exchange 365 community.