Hydrocarbon Processing - March 2021 - 65

Valves, Pumps and
Turbomachinery
L. STEWART and S. STATHAM, exida,
Sellersville, Pennsylvania

Optimizing failure data analysis
A critical objective of engineers who design an automatic
protection system is to lower the false trip rate and to lower a
key metric called average probability of failure on demand (PFDavg). The PFDavg metric represents the chance that the automatic protection will not work when needed. The false trip rate
and PFDavg are based on several variables, including the failure
rate of all devices used for each safety instrumented function
(SIF). These devices are classified into three groups:
1. The sensor assembly, which detects a dangerous condition
2. The logic solver, which determines when to initiate
the protection
3. The final element, which does the protection work,
and is often a remote actuated valve that opens or closes.
Statistical analysis of more than 80,000 SIF designs shows
that nearly 70% of the PFDavg comes from the final element
(FIG. 1). Many engineers are looking for help from manufacturers
to reduce this percentage and make their plants safer. Some manufacturers have had their products IEC 61508 certified for functional safety applications. While this helps, plant operators and
engineers recognize that utilizing IEC 61508-certified products
does not guarantee the reliability of the final element assembly.
Many field-failure reports show root causes due to inadequate
torque matching, insufficient assembly testing, manufacturing
errors or application mismatching. More needs to be done to apply the lifecycle engineering rigor required by IEC 61508 to the
entire final element supply chain.
Pre-engineered final element assemblies. Recently, orga-

nizations (such as WIB-The Process Automation Users' Association) have undertaken efforts to address this challenge by
developing a new recommended practice (S 2812-X-1), which
is targeted specifically at quarter-turn emergency isolation valve
assemblies. This new recommended practice-titled " Actuated
Valve Assembly " -significantly increases the rigor associated
with the engineering, design, testing and documentation of complete final element assemblies. To satisfy the S 2812-X-1 recommended practice, those companies that supply final element assemblies must have strong engineering processes that allow the
complete assembly to be engineered, designed and tested as a preengineered product. This added rigor has the potential to reduce
the failure rates that contribute to false trips and the PFDavg.
To provide verification and structure, IEC 61508 certification
can be applied to the assembly. To address this need, the authors'
company has developed a certification schemea that assesses the

complete final element assembly. Several standards have been
leveraged in the development of the certification scheme from
organizations such as WIB, the International Organization for
Standardization (ISO), the Instrument Society of America
(ISA), European Norms (EN), the American Petroleum Institute (API) and the American Society of Mechanical Engineers
(ASME). The resulting requirements from these relevant standards define a certification scheme that will reduce engineering
errors and omissions, characterize lifetime torque, match actuator output, provide complete design testing and verify manufacturing quality. The program also requires a user document called
a " safety manual " where safety design data-including application limitations, predicted failure rates, maintenance procedures

Failure data analysis
68.64%-Final elements
23.27%-Sensor(s)
7.64%-Logic solver
Projects
SIFs

5,029
84,165

FIG. 1. Average PFD contribution results from more than 80,000 SIF
designs. Source: exida exSILentia usage study, November 2019.
0.6
Typical stress
Industry typical

0.5
0.4
0.3
0.2
0.1
0

FIG. 2. A stress vs. strength distribution model.
Hydrocarbon Processing | MARCH 2021

Hydrocarbon Processing - March 2021

Table of Contents for the Digital Edition of Hydrocarbon Processing - March 2021