Emerging trends in intelligent safety-valve control

Juha Yli-Petäys, EHC Business Manager, Metso’s Automation business line

The reliability of safety valves is crucial to process control. Their design has changed little in recent years and partial stroke testing (PST) has been the accepted method of monitoring them. Newer methods linked to the introduction of on-line intelligent safety valve controllers are offering new levels of performance monitoring. Analysis results of automatic PST implementation, especially when using Fieldbus technology, offer possibilities for enhanced diagnostic and maintenance programming.

Reliability is critical
Safety valves protect processes, personnel and the environment against process disruption.  However, they are operated only in trip situations, but it is essential that they perform, when required. Because these shutdown valves are rarely cycled, there is always concern over whether they will operate on demand.  In practice, unless they are periodically stroked, it is virtually certain that they will not work when required. 

It is important to realise that safety valves are moving mechanical devices that normally operate in very difficult environmental conditions. Furthermore, they are the most important components in the safety loop (sensor, safety logic and final element), because most of the problems that occur are related to the functionality of the final element [1].  The need for regular valve testing and for testing while the process is running is therefore absolutely essential.

Fig. 1 Typical safety valve designs

 

Safety standards and PFD measurement
To reduce the risk of toxic emissions, uncontrollable fires or explosions, or any combination of the above, oil, chemical and petroleum plants need to pay ever-increasing attention to plant safety. The principal influence in such projects has been the IEC61511 safety standard. .This requires companies to evaluate every area of their plant and assign a targeted Safety Integrity Level (SIL) to it. It applies when equipment meets the IEC61508 standard and is integrated into an overall system. Thus IEC61508 is the standard applicable to device manufacturers and IEC61511 the standard for end-users.

Probability of Failure in Demand (PFD) is the statistical measure of availability in an emergency. Although there are many ways to calculate PFD for individual components, it can be measured using the following well-known equation. [2]

Where:
DC = Diagnostics Coverage factor
ld   = Dangerous Failure Rate = 1/MTBF (Mean Time Between dangerous Failures
TIA  = Testing interval for automatic tests (on-line)
TIM  = Testing interval for manual tests (off-line)

These are two-part calculations: on-line testing and off-line testing. For safety valves, the on-line diagnostics part relates to Partial Stroke Testing (PST) and the off-line part to periodic maintenance. With frequent on-line testing, better diagnostic coverage, shorter mean times for repair and good communication methods, lower PFD figures can be achieved.

Traditional on-line testing
Partial stroke testing is the method most commonly used to test valve functionality while the process is running. Conventionally, this involves using a mechanical stroke-limiting device (or a pneumatic switch for each safety valve) to restrict the movement of the safety valve, to permit partial stroking without disturbing the process. Once movement is restricted, technicians send a signal from the control room to determine whether or not the valve will respond.

The manual work involved is expensive and unreliable for many reasons including the lack of real-time data and the absence of trend data. Another significant drawback is that the valve under test is not operable during testing - a hazard, should a real safety issue be encountered.

Advent of the intelligent controller
In the early 2000s, PST implementation was considerably improved by the introduction of Neles ValvGuard, the first intelligent safety valve controller, which was taken into use almost immediately by BP, who reported significant improvements in both plant safety and operating costs [3].

With partial stroke testing, the emergency isolation valve is moved so slightly that it does not disturb the protected process, yet the degree of movement is sufficient to detect the most critical problems with the emergency valve.

However, when PST is carried out using an intelligent emergency valve controller, the test results can be analysed automatically after test implementation.

Problem solving
Partial stroke testing allows more frequent valve stroking without disturbing the process. The required SIL level for Safety Instrumented Functionality (SIF) can therefore be maintained for longer periods. Even though the valve stroke does not cover the whole valve travel range, when an intelligent emergency valve controller is used to implement PST, the build-up of most of the problems typically encountered with emergency valves can be detected.

Figure 2. PST testing example: Actuator pressure measurement v. valve position.

 

PST detects random hardware failures related to the final element. Detection is based typically on the change in valve dynamics, which can be seen when the latest PST results are compared to historical data. For example, Neles ValvGuard measures the breakaway pressure and load factor from PST and Figure 3 shows one method of doing a historical comparison using a graph. 

Breakaway pressure indicates the pressure measurement level at which the valve starts to move during a valve test. Breakaway-pressure trend information can be used to analyze valve load changes. The load factor is calculated from PST data, which indicate the friction changes of the valve. A high load-factor value means increased friction due to an undersized actuator. Load-factor historical values are shown in the same way as a breakaway pressure trend.

Typical valve problems which can be detected with partial stroke testing are: Sticking, stem seal leakage, valve seat leakage, mechanical valve or actuator failure and failure in the controller itself.

Diagnostics of partial valve stroke test makes maintenance planning easy

Technology developments
In recent decades, developments in emergency valve design have remained rather static when compared to developments in emergency valve controllers. New technologies have been continuously adopted to meet the required safety specifications, while simultaneously ensuring the cost efficiency of installation and operational costs. In the past few years these technologies has been used increasingly as a part of SIF implementation.

Looking ahead
One clear future trend seems to be the use of fieldbuses. FOUNDATION fieldbus and Profibus organisations are the most active in this area and are enlarging their specifications to cover Safety Instrumented Functionality. The reasoning is clear: Fieldbuses are used more often and have proved the technology.

New upstream processes are making greater use of Fieldbus technology, largely because it reduces hardware, wiring and engineering costs and improves the asset management integration of final elements into the host system. Traditionally, on installation all control valves have fully supported Fieldbus communication technology, but emergency valve controllers are integrated into the asset management systems through HART communication using separate communication hardware.

Foundation-based emergency valve controllers would give much faster response times for the status information and easier integration of higher-tier diagnostic information into the host system, when compared to the parallel HART network. This would improve cost efficiency when additional communication hardware is not required. At the same time, operation of the device is much easier for the end-user.

The use of intelligent functions with safety valve controllers has already created considerable interest. Nowadays, automatic PST implementation is seen more often as a serious possibility, rather than as a possible source of spurious trips. Furthermore, the analysis results of PST implementation are being widely used as a part of the maintenance program. This trend raises the need for easier field-device integration up to control system level and expectations now lie in the wider use of fieldbus technology.

References:
[1] Sintef; Offshore Reliability Data Handbook; OREDA; 2002.
[2] Laaksonen J.; Reducing the Window of Potential Failures for Emergency Shutdown and Venting Valves by Automatic Partial Stroke Testing Devices; 2005; TÜV Symposium Cleveland, OH, USA; 12p.
[3] ARC White Paper; Neles ValvGuard Allows BP to Increase Safety While Reducing Costs; 2001; 12p.