In Part I of this three-part blog series we discussed Structuring your Bridge Database in dTIMS. In this blog post (Part II) we will explore Life Cycle Cost Modelling of Bridges.
Modelling
Mathematical or scientific models allow us to predict the future. We don’t have a crystal ball to show us the future, so we use models based on observable data in a certain environment to predict performance of assets. For example, the performance of a bridge.
Why Model the Performance of a Bridge?
Performance prediction is the basis for the assessment of the life cycle of a bridge, and enables us to compare different maintenance treatment strategies from the technical and economic point of view:
- Life-cycle cost analysis: comparison of different maintenance treatment strategies using benefit and cost indicators
- Life-cycle risk analysis: comparison of different maintenance treatment strategies using risk and cost indicators, where the maintenance risk is defined by the probability of failure (PoF) and the consequences of failure (CoF) of doing or not doing maintenance actions
Looking into the future is essential because we don’t want to be in a position where our bridges are unsafe and we have to close them. We need to be aware of the status of the bridge in the short, medium, and long-terms.
Predicting Performance
Performance prediction depends on selecting the correct life cycle cost model. This is strongly dependent on the level of analysis and level of data granularity. Models can by Mechanistic or Empirical. Examples of empiral models include Deterministic and Probabilistic.
- Mechanistic – uses a theory to predict what will happen in the real world. The alternative approach, empirical modeling, studies real-world events to develop a theory. Deterministic and Probabilistic modeling are examples.
- Deterministic – forecast asset deterioration in the future by using a mathematical correlation between condition and parameters
- Probabilistic – forecast asset deterioration in the future by using a probability distribution
The following graphic presents an overview of each level managed and which type of model would potentially be used. We will discuss deterministic and probabilistic models as they are the most used in bridge modelling.
Deterministic Models
Deterministic models can be created on different levels so it's possible to use these models on the object level, as well as a component and element level. Depending on the level, the specific element might have different dependencies that can be integrated in the model.
The example model below is based on the component level. Shown are the main dependencies based on lifespan and on the material for the different components. Different types of deterioration curves include progressive or linear curves. The main input parameters are the lifespan and some shaping parameters which allow us to shape the deterioration curve individually. The starting value in these examples is the age of the component, which could either be the real age of the component or a back dated age calculated from the last condition service.
Probabilistic Models
Markov Chain modeling is a probabilistic approach that allows us to predict a condition distribution as opposed to one definite value. This is based on transition probability matrices which gives us that probability to which the condition will change within one year. For example, what is the probability that the condition is changing from a Class 1 into a Class 2 within one year? What is the probability that the condition will change from Class 2 to Class 3?
The number of asset classes gives us the size of that condition transition matrix. This transition probability matrix will then be multiplied with our starting vector and will result in the condition factor in the next year.
Final Word
This is a short look at bridge models that can be used in dTIMS. Flexibility is key when choosing an asset management software that can accommodate models on different levels. Watch for Part III in this blog series which explores the treatment catalogue for bridges and maintenance strategies, and bridge analysis results.