In this article:
By Dr.Neill Renton, ORS Country Manager, UK.
Introduction
Complex process plants can fail in many ways. Valve failures can mean a loss of liquid level in a vessel and high-pressure gas being routed into a lower design pressure system (gas blowby), or low temperatures developing as a result of an unplanned pressure let-down (Joule-Thomson cooling), or even excessive flowrates leading to large amounts of vibration in the pipework. These are just some examples of component failures that can lead to a loss of containment with a resulting fire and explosion.
To try to prevent large-scale damage to equipment following a component failure, process plants contain safety instrumented systems or ‘functions’ (SIFs). These uses transmitters to ‘trip’ the plant in the event of high/low pressure, level, and temperature and isolate the main unit operations from the relevant source of pressure using emergency shutdown valves or other end-elements.
What is the ‘Process Safety Time’ and the SIF Response Time?
One question that often comes up when designing or operating these SIFs is ‘how quickly do they need to work to keep the plant safely inside its design envelop?’. To answer this question, let’s define two different time periods of interest:
How quickly does the failure event develop? This is taken from the moment a valve or controller fails to the hazardous event occurring, in this case, loss of containment. This time is a property of the process and depends on the equipment volumes, flow rates, compositions, and valve sizes. It is known as the Process Safety Time (PST) in IEC 61511-1.
How much time does the SIF have to respond? This is shorter than the process safety time and is known as the SIF Response Time (SRT). It is a performance target that the SIF needs to meet and is again a property of the process. It represents the length of time available between the SIF set-point being reached and the design condition of the equipment being breached following failure.
This is best understood using a diagram:
The graph shows a plot of pressure changing with time in a vessel following a vapour outlet valve failing closed. With vapour trapped in the vessel, pressure begins to build. Without action from the operators, the SIF, or finally the mechanical protection in the form of the pressure safety valve (PSV), the vessel will be over-pressurised and a loss of containment will occur.
The graph shows a number of time intervals of interest. These are not all defined in 61511-1 [1] or the CCPS Guidance [2], but are discussed in [3]:
: This is the ‘process safety time’ as defined in 61511-1 and represents the total time from the initial valve failure first occurring to the ‘hazardous event’ when the vessel is over-pressurised.
: Early Response Time. This is the time interval in which the process variable is changing after the failure occurs – no alarm has yet been activated.
: Operator Reaction Time. This is the time interval from the alarm set-point being reached to the SIF set-point being reached in which the operator could take manual action to shut-down.
: The required SIF Response Time[1] (SRT). This is the interval between the SIF set-point being reached, and the hazardous event occurring. The SIF must detect and then operate its primary end elements, usually shutdown valves (SDVs) within this interval and do this quickly enough to keep the process variable below the design value.
The graph also shows the relationship between these different intervals such that:
Which one should you use?
During the design phase, the PST as defined in [1] is the most useful measure as it allows the determination of a suitable set-point of the SIF such that the SIF can react and prevent the hazardous event. Once a SIF is installed, and has been in operation for some time, perhaps with set-point changes, the SIF Response time
is a more appropriate performance measure.
How to calculate the PST & SRT.
Assessing the Process Safety Time and the SIF Response time can be done in different ways:
Simply: By making some conservative assumptions about the process conditions at the point of failure and the vessel, valve, and pipework sizes. This gives a conservative lower bound on the PST & SRT which is a robust approach for monitoring SIF performance. Works well if the SRT is greater than about 15-20 seconds and the unit operation is simple e.g. a vessel with one inlet and one/two outlets.
More Complex: using Dynamic Simulation in a commercial process simulation software package. This gives a more accurate picture of the process behavior during a failure and is good for shorter PSTs / SRTs less than 15 seconds or more complex unit operations with multiple inlets and outlets, rotating equipment, and temperature problems.
Verification of SIF Performance
Having calculated the PST and SRT, the final step is to assess whether the SIF (transmitter, logic solver and valve/end-element) can meet the performance target and keep everyone safe in the event of failure. This in turn can be used to confirm required ESD valve closure times and SIF test times that can be checked throughout the lifetime of the plant.
Conclusion
In conclusion, the Process Safety Time and the SIF Response Time are properties of the process and describe how process variables change during failure events up to the point when the design envelope is breached. They can be used as performance measures for SIFs that can be turned into testing time requirements for instruments, logic solvers, and valve closure times.
References
[1]
"BSi, Functional Safety - Safety instrumented systems for the process industry sector – Part 1: Framework, definitions, system, hardware and application programming requirements BS EN ISO 61511-1+A1:2017". https://webstore.iec.ch/publication/24241
[2]
"CCPS, Guidelines for Safe & Reliable Instrumented Protective systems, American Institute of Chemical Engineers, 2007.". https://www.wiley.com/en-gb/Guidelines+for+Safe+and+Reliable+Instrumented+Protective+Systems-p-9781118209691
[3]
"G. Bernard & W. Creel, Impacts of Process Safety Time on Layer of Protection Analysis, Process Safety Progress, Vol.34, No 4, pp383-388, American Institute of Chemical Engineers, 2015.". https://aiche.onlinelibrary.wiley.com/doi/10.1002/prs.11759
