Process Modelling improves plant safety

Process Modelling improves plant safety

Dynamic simulation can provide a way to document and understand the complex risks of process plant, while paying for itself in improvements.

The first step of risk assessment is by now very familiar to all of us: Identify the hazards. This is true whether we’re assessing an office or a chemical plant. Once we know what the hazard is, we then move on to understanding the impacts it can have, and how we may prevent it. However, there is a significant difference between assessing slips, trips, and falls, and the hazards of industrial chemical processes. Complexity.

Complex processes present complex hazards. In fact, even seemingly straightforward processes present complex hazards. Runaway reactions, overfilling, utility failures, etc. can happen at any site. While larger chemical processes tend to involve significant interactions between plant, assessing process hazards properly is never simple. This is why we rely so heavily on well-run hazard studies, with a multidisciplinary team to carry them out.

Over the last half a century, hazard studies have developed from a set of internal processes within ICI, to a standardised, methodical approach used around the world. In parallel, quantitative risk assessment has grown, responding to the needs of process safety professionals to answer the big question: “What if?”

Meanwhile, dynamic process simulation has emerged from the technological revolution. Process simulation is the modelling of plants and equipment using computer software. Dynamic simulation involves running these models in a time-based way, to see the effects of changes or failures. Physics, chemical, engineering, and thermodynamic equations combine to calculate every aspect of a process, every fraction of a second. Originally limited by the available hardware, complete plants can now be built, running real-world processes in a virtual setting, faster than real-time on today’s powerful PCs. While many are familiar with these models in Operator Training, or occasionally for one-off safety studies, their use in hazard studies is a newer phenomenon.

Imagine carrying out a hazard study by going out onto the built plant, and testing the failures. “What if this pump fails?” Turn off the pump. “What if this controller fails?” Override the controller. Obviously impossible, both for design (when the plant doesn’t yet exist) and for existing plant (where intentionally introducing failures is manifestly unacceptable). This is what dynamic simulation offers the hazard study team; the opportunity to break the plant, and see and record what happens.

By integrating models into the hazard study, the risk assessment process can be greatly enhanced. Once a model is built, systematically going through the failure modes for equipment is trivial. This means that answers can be generated at the HAZOP table. Complex cases, such as site wide power or instrument air failure, can be run with a single click. Identifying process hazards has never been more straightforward.

In the same act of identifying the hazards, the modeller is instantly presented with the effects. Vessels overpressure, relief valves open, pipework floods. These quantifiable results can then feed directly into the risk assessment. No need for gut feel, or rules of thumb.

Moving further through the steps, the mitigation measures can be studied. The effectiveness of controls and trips can be validated. Timings, peak pressures and temperatures, or any other process data can be instantly used as part of the Layer of Protection Analysis (LOPA). Set points can be tuned, and unexpected effects identified. The design can be demonstrated to be effective before a penny is spent.

Throughout this process, a library of hazardous scenarios is compiled, fully accessible and repeatable. This can then be integrated into the site’s Process Safety Management system. When new safeguards are designed, their testing in the models can be validated and documented. Management of Change can integrate model testing into the process, so that the impacts of the change can be instantly assessed against the scenarios associated with the equipment.

By creating an accurate model of the plant, efficiency and debottlenecking opportunities are practically a by-product. In our experience, a model will ultimately pay for itself in terms of the plant improvements it makes available. The model provides a meeting place for engineers of all disciplines. Planners can predict the effects of new feedstocks. Control engineers can test their designs against a simulation which responds realistically. Operators can be trained in how to handle the hazards they will hopefully never see. All of this may seem to put dynamic simulation into the “too good to be true” category, but it’s not. It’s a platform which improves safety while increasing profits, available to everyone. It’s the future.

For more information visit
www.flexprocess.co.uk
or contact enquiries@flexprocess.co.uk