WorkBench can be used for building models, customizing them and reasoning with them
If you already have access to WorkBench, and know how to open a model, as outlined in Get access to WorkBench and models then you are ready to create your first model.
Models in WorkBench are built based on the Multilevel Flow Modelling theory that must adhere to a syntax, that is a set of rules that define how model elements can be combined to build a model.
When opening a model, the user is presented with the canvas (1) where models can be built on, and three menus.
The model elements menu (2) can be used for dragging elements to the canvas, to combine them into a model. The model elements must be combined based on the model syntax and are advised to follow a set of modelling principles. The model element details menu (3) can be used for searching for model elements on the canvas, for adding or editing information to a selected model element on the canvas and for viewing errors from syntax verification. The navigation menu (4) can be used for seeing the current model name, for exiting a model to go to My Documents, for changing between model building, model customization and model reasoning, or to see the current WorkBench version or sign out. The modelling support menu (5) can be used for importing, saving and exporting the model, for model syntax verification, zooming, undoing and redoing.
The structures, flow functions, objectives, controllers and causal relations in the model elements menu (2) can be dragged to the canvas and be combined. For each element, a label and a description can be added from the model element details menu (3) to provide an explanation of the model. Dependent on the element, different types of details are available. When a model has been built, the syntax verification in the modelling support menu (5) can be used to check for errors. If any errors exist, the errors are shown in the model element details menu (3) and must be corrected before the model can be used for reasoning.
WorkBench can be used for three purposes: model building (A), labelling and model customization of generic models (B) and reasoning (C).
When modelling in build mode, the purpose is to use generic models. As an example, a model of heat exchanger could be the same independent of where it is placed in an asset. However, the information that must be shown in other products such as the Control Room Assistant or the HAZOP Assistant should differ dependent on where and what it is used for. Therefore, the Label Editor is used for making generic component or system models asset specific.
When labelling in the Label Editor (B), asset specific information, such as tags, system names and standardized counter-actions for verifying, correcting and preventing causes and consequences can be overlaid on generic models to produce asset specific models.
The model can be for simulating in Reasoning (C), to observe the behaviour of the system based on an input. The inputs are deviations to the system, and the user can choose to view either the potential causes or the potential consequences for the deviation. A path is shown on the model, to visualize the entire path through the model (and system) between the input deviation and the output. Both the input and the output are qualitative and discrete states such as low, normal, high, breach, true or false.