PT Notes
Part 2 - Are you Addressing Functional Resonance in Process Safety?
PT Notes is a series of topical technical notes on process safety provided periodically by Primatech for your benefit. Please feel free to provide feedback.
Part 1 of this PT Note introduced the concept of functional resonance and provided examples of functional resonances in classic process safety accidents. Part 2 describes a framework for applying the concept of functional resonance in the process industries.
The concept of functional resonance recognizes that unexpected outcomes can emerge in a process as a result of normal variations in different process functions interacting in unanticipated ways leading to nonlinear consequences. Functional resonance modeling focuses on understanding the variability of system functions and how they can resonate under certain conditions, that is, where variabilities may coincide, and produce unexpected and possibly hazardous outcomes.
As used in functional resonance, the term “resonance” does not refer to the phenomenon of synchronized vibrations in objects. Rather it denotes that variabilities can combine or "resonate" in unexpected ways, leading to unforeseen outcomes. The concept does draw on the analogy from physics, where minor inputs to a system at the correct frequency can produce substantial vibrations in the system. A key feature of functional resonance modeling is that resonances can lead to outcomes that are not directly proportional to the inputs or disturbances but can be much larger.
Many process safety accidents have occurred because of functional resonances where multiple variabilities or factors aligned, leading to significant disasters. The concept of functional resonance provides a means to understand such complex interactions and their cascading effects in complex systems, such as modern process plants. Functional resonance modeling provides the means to identify such resonances so that preventive measures can be taken to avoid process safety accidents.
A framework for applying the concept of functional resonance to processes is provided by its premises:
Complex Systems: Functional resonance modeling recognizes that real-world processes are inherently complex, and their behaviors, especially under adverse conditions, cannot always be predicted from simply studying their individual parts.
Sociotechnical Systems. Processes consist of both social (human and organizational) and technical (equipment and technology) components. Variabilities can arise from both social and/or technical functions. Consequently, processes must be understood as sociotechnical systems, which enables the interactions of people (operators, decision makers, etc.) with process equipment and technology to be addressed.
Safety as an Emergent Property: Emergent behaviors arise within complex systems. Their defining characteristic is that they cannot be predicted or deduced merely by examining the individual components of the system. Instead, emergent behaviors manifest as a result of interactions between the components of a system.
Traditional safety models view safety as a direct result of the barriers or safeguards that have been put in place. In contrast, functional resonance modeling interprets safety (or its converse, the lack of safety) as an emergent property of the process, that is, it arises from the complex interactions and relationships between individual components of a process, its environment, and human operators, but is not a property of any single component taken in isolation. Recognizing safety as an emergent property emphasizes the importance of holistic approaches to hazard analysis and risk assessment.
Human Factors Integration: Traditional safety models treat human operators and other people who work in process plants as sources of error or unreliability. In contrast, functional resonance modeling integrates human factors holistically, recognizing that human variability can both help and hinder system performance.
Adaptive Nature: Complex systems, especially those that are sociotechnical systems, can adapt based on experience or changing conditions. This adaptability can reshape the interactions between functions and lead to evolving emergent behaviors over time.
No Root Causes: Traditional safety thinking aims to trace back adverse events to a singular "root cause." Functional resonance modeling operates on the principle that in complex systems such as process plants, it is more meaningful to understand the confluence of variabilities across functions that lead to an outcome, rather than searching for a singular linear path.
Variability is Inevitable: Functional resonance modeling recognizes that there is always variability in how process functions are executed. For example, work as done in a process plant may well be different from how work is prescribed. Such variability in performance is often necessary to keep processes operating. For example, an operating procedure might dictate that operators follow a certain sequence of operations to perform a task but, in practice, they follow a different sequence because it is quicker or more efficient. Variabilities in functions can resonate with each other and cause process safety accidents. Rather than seeing variabilities as deviations or errors, functional resonance modeling views them as a natural and normal part of system performance.
Dynamic Adjustments: Functional resonance modeling recognizes that functions and their performance adjust dynamically based on process conditions and interactions with other functions. These adjustments can lead to non-linear shifts in system behavior based on changing contexts. People and systems often make dynamic adjustments to cope with system variability. These adjustments, while generally beneficial, might sometimes be sources of variability themselves.
Functions, not Failures: Functional resonance modeling is centered on understanding system functions and their variability. Unlike traditional safety approaches, it does not focus on failures or errors but seeks to understand how normal variability in functions can lead to different outcomes.
Interactions Over Components: While traditional safety models focus on the failure of specific system components, functional resonance modeling focuses on the interactions and relationships between functions. This represents a shift from a cause-effect model to understanding emergent properties of a process.
Complex Interdependencies: In complex systems, functions or components do not operate in isolation. Their operation is dependent on and influenced by multiple other functions, often in unexpected ways. This interdependence and the nature of their interactions make the overall system behavior difficult to predict.
Feedback Loops: Complex systems often have feedback mechanisms, both positive and negative. These feedback loops can amplify or mitigate effects, leading to behaviors that are not directly deducible from understanding individual functions.
Non-linearity: Outcomes in complex systems are not always linearly related to the inputs or causes. Small changes or disturbances can propagate, interact, and amplify leading to significant and non-proportional outcomes. Traditional safety models focus on sequences or chains of events (cause-effect relationships). In contrast, functional resonance modeling is centered around the interactions between functions in a system. These interactions are not merely linear chains but can be complex networks where feedback, couplings, and parallel processes exist.
Example of the Application of Functional Resonance to a Distillation Column
First, functions provided within the column are identified:
- Feed Pre-heating
- Reflux Ratio Control
- Reboiler Heat Supply
Next, possible variabilities in these functions are considered:
Feed Pre-heating: The feed entering the distillation column is occasionally hotter than expected due to minor fluctuations in another part of the plant.
Reflux Ratio Control: Operators in the control room adjust the reflux ratio in the column based on product quality metrics. Due to intermittent instrument errors, the reflux ratio sometimes varies slightly.
Reboiler Heat Supply: The reboiler experiences slight fluctuations in its heat output due to wear and tear in its heating element.
Lastly, possible resonances are identified.
Individually, each of these variations might be within acceptable limits and not pose any concern. However, on a particular day, hotter feed may coincide with a momentary lower reflux ratio, and simultaneously the reboiler may supply more heat than usual. These small variations resonate with each other, leading to a rapid and unexpected increase in the column's internal temperature and pressure, which may cause the column to exceed safety limits.
This functional resonance is an example of how the unforeseen interaction of normal variations in different functions can cumulatively lead to significant consequences in a process plant. It emphasizes the need for holistic monitoring and adaptive control strategies to anticipate and mitigate these kinds of risks.
Functional resonance modeling is a paradigm shift from traditional safety models. It emphasizes the interconnectedness and variability inherent in complex systems. It looks at processes not just from a "what could go wrong" perspective but from a broader "how things usually go right and how they sometimes do not" viewpoint. Its ability to provide a holistic understanding of processes and their variabilities makes it a powerful tool for enhancing process safety.
If you would like further information, please click here.
To comment on this PT Note, click here.
You may be interested in:
