Critical valves are tested at regular intervals to ensure their functionality and maintain their Safety Integrity Level (SIL). One method of proving the degree of valve reliability is to fully stroke the valve to the closed (or open) position, providing immediate confirmation of the operability and performance. However, when performing a full-stroke test while the plant is operating, the process flow would come to a halt. To minimize impact on the process, operators usuallychoose to perform an offline full stroke test during a
scheduled shutdown.

A downside of a full stroke testing during shutdown is that these tests are performed in a ‘cold’ and therefore abnormal process condition. There are examples where ESD valves and actuators were tested only during shutdown and then worked properly, but when tested in operation they failed. Clearly, the valves reacted differently in ‘hot’ and ‘cold’ conditions.

Partial stroke testing prevents failure of emergency valves, and helps determine the likelihood that an emergency shutdown valve will operate on demand. During a partial stroke test, the system is activated but the valve is only partially closed or opened to a pre-set degree. Thereby the performance of the entire actuator and valve package is tested at real speed and force – and under actual operation conditions. The process flow is maintained and the impact on the production is minimal. Whilst a partial stroke test never replaces the need for full stroke testing, by doing partial stroke tests between full stroke ones, the facility may lower the Probability of Failure on Demand (PFD) of their safety system, for example by lowering the risk of an increasing break-loose torque, and/or extend the time interval between full stroke tests.

Standard electronic partial stroke test devices use an electronic module that connects between the supply from the ESD system and the solenoid valve. To perform a test, the timer de-energizes the solenoid valve to simulate a shutdown and re-energizes the solenoid when the required degree of partial stroke is reached. These systems are fundamentally a miniature PLC dedicated to the testing of the valve.

A mechanical partial stroke test device allows a true system test. All the actual components, controls and elements used in an ESD valve will be activated. The user has real information about the exact controls that will be relied upon, to protect plant and personnel.

Most non-mechanical partial stroke test devices assume a relatively smooth movement of the valve actuator. They also reckon that the automated valve will act in a consistent manner, independent of environmental conditions, such as temperature and humidity or the length of time the valve rests between test cycles. This is rarely the case, however. One of the advantages of mechanical partial stroke test devices is that they stop valve movement mechanically at a specified percentage or degree of valve closure. This metal-to-metal blocking prevents overshooting of the valve past the set point due to spurious trips or other extraneous conditions.

When using controls-driven partial stroke test systems it is assumed that the test device or system will indeed prevent the actuator from driving the valve past the set point or to the fully closed position. However, actuators don’t always act smoothly and ‘stiction’ can prevent a spring return actuator from stroking immediately upon release of air pressure from the actuator cylinder. The result is the possibility of stored energy driving the valve past the set point and allowing the valve to nearly or fully close, potentially forcing a process shut down.

• Mechanical partial stroke testing has some distinctive advantages over electronical testing.
• It is a ‘true system test’, all the actual components, controls and elements used in an ESD valve package will be activated under reality conditions.
• The actuator’s full force is tested, and the entire system will stroke in its ‘real world’ time sequence and speed of operation.
• The metal-to-metal blocking system prevents stroking past the specified point, eliminating the danger of overshooting.
• No additional power or wiring required, resulting in cost savings and reduction of system complexity.
• No need for calibration of the mechanical test device after installation.
• The ‘human-machine’ Interface assures that valve movement can be ‘eyeball’ verified.

In a typical mechanical partial stroke test scenario, the key to initiate the test mode is kept in a supervised location, such as a key cabinet in the control room. When it is time to test the valve, the operator inserts the key into the device, engages the device and then informs the control room that the valve is now ready to test. When the partial stroke test is initiated, the person witnesses the valve movement and reports the event, usually by radio, to the control room (this can also be verified by limit switches). After completing the test, the key is removed, disengaging the device, and the key is returned to supervised control.

By design, the key cannot be removed from the mechanical testing device while it is in the test (or ‘engaged’) position. So, if plant operations personnel know the key is in their control, then they also know the device cannot be engaged.