This post examines the attributes of state based control and the value delivered to manufacturing from the initial design through the operating life of the facility by improving the effectiveness of operators. This is part 5 of a 5 part post and discusses safety and asset utilization.
Coordinated Safety Response - Beyond a Safety Function, a Safe State
Beyond the safety instrumented systems ability to keep the plant safe by interfacing with the instrumentation identified in the hazard analysis, state based control takes the entire unit operation to a safe state. Other units respond by moving to their appropriate states in response. For instance, you may have a Safety Instrumented Function (SIF) that cuts the steam to a reactor on a high pressure. With state based control, you could define a safe state for the reactor that would perform the actions that the operator would have to take in this scenario. The transition to this state could come from the high pressure or other causes. The safe state, once created, can be used for any number of scenarios the user thinks are appropriate. In this state, the control system would take actions that the operator might need to take, such as cut the steam, cut off the feed, put on full cooling and open the vent to the flare heater automatically and immediately. In addition to the safety function, alarms are dynamically managed so that only alarms that are applicable are enabled. Therefore, there is no alarm flood to confuse the situation. The reactor trip is an identified degradation scenario. A degradation scenario is a case where something has happened in the plant which the other units need to respond to in order to keep the facility in the maximum possible readiness to return to normal operation. Beyond handling the issues to safely manage the reactor, the reactor unit will communicate to the other up and down stream units that it has tripped. The other units can respond by going to the appropriate state to minimize the time required to come back on line. For instance, the back end of the plant can put itself on recycle. When the reactor is restarted and seen to be in a normal run state, the back end of the plant can leave the recycle state and continue with normal operation minimizing any impact to production. A crude unit, for instance, can have a similar response for a crude tower heater trip and automatically recycle the back end.
Figure 4: State Based Control Process Coordination
Avoiding a Safety Response at All
Better than handling a safety response well is not having one. Some analysis has been done linking unplanned events to alarm rates and the level of automation of a facility, with the facilities divided into four quadrants. There was a strong relationship between unplanned events and the quadrant a plant was in. Quadrant 1 was the best case with low alarm loading and a high level of automation, and Quadrant 4 was at the other end of the spectrum with high alarm loading and a low level of automation. Not surprisingly, Quadrant 1 plants had the lowest level of unplanned events, and Quadrant 4 plants had the highest level of unplanned events over the period studied.
Interestingly, Quadrant 2 plants that had high alarm loading but also a high level of automation, had less unplanned events than Quadrant 3 plants with low alarm rates and a low level of automation.
Moving from a quadrant number to one lower reduced the unplanned events by about 2 – 3 times per thousand I/O over the period studied. So, lowering the alarm rate and moving a plant from Quadrant 4 to Quadrant 3 would reduce the unplanned events by 2 to 3 times over a similar time period, but raising the level of automation and moving from Quadrant 4 to Quadrant 2 would have twice the effect on unplanned events. As we know, unplanned events are costly and dangerous.
To put it in perspective, the take away here is that lowering alarm rates is a powerful tool to reduce unplanned events. Raising the level of automation can have twice the impact. The surface has twice the slope on the automation axis as the alarm loading axis. Doing both is of course the best and moving from Quadrant 4 to Quadrant 1.
Figure 5: Relative Unplanned Events by Quadrant
State based control has also been seen to increase asset utilization. Some key areas of lost asset utilization such as equipment failures, operating discipline issues, and reprocessing can be significantly reduced through automation.
State based control is a viable option to improve the design process through the use of standard reusable architecture and instrument and alarm justification. It reduces time and costs and improves quality. State based control maximizes the investment in the DCS by capturing knowledge in the form of operating discipline that can be leveraged with greatly reduced training costs in a dynamic workforce. Safety and operability are enhanced through the uses of safe states in units and the communication between units to optimize the response to degradation scenarios.
Operators are in a position to manage the process through state changes while having their heads up to see the big picture - avoiding problems and optimizing performance.
State based control allows project teams to meet and exceed the rising expectations they are challenged with.