Embedding Whole Life Cost Modelling: Development of the London Underground Track Model

摘要

Over the last five years, London Underground has developed a comprehensive asset model for its track assets, to enable it to optimise its renewals plans based on Whole Life Cost and performance on an asset by asset basis. This model has been fully embedded within London Underground's decision making, planning and forecasting processes. It is supported by comprehensive validation and forecast assessments to ensure that rules are continuously improved and refined to increase long term accuracy. This paper summarises the development of this model, its functionality, how it operates, how it is deployed within the business, and lessons learned in the process which may benefit the development of other asset models. A Whole Life Cost (or Life Cycle Value) model for track assets is at the same time both simple and complex. The general rules governing asset life expectation under different environments and loading conditions are reasonably straightforward and understood, as are many of the relationships between asset condition and maintenance effort or asset availability. However, at a detailed level a section of track will be made up of various assets which may have different ages, expected lifespans, maintenance requirements and performance risks. Options for intervention will range from wholesale replacement of all assets in a track section (which may destroy some residual asset value) to localised patch and repair of specific components. When trying to define a renewals workbank within constraints on budget and resources (which is balanced across the entire network to deliver the best achievable value for money), these calculations become complex. Hence a detailed and robust modelling approach is required. Unlike other Track Models which have been produced over a number of years, the LU track model calculates at a tactical level, producing specific cost, performance and condition forecasts for every section of track on the network which can be independently validated to ensure accuracy. These forecasts are then rolled up to a strategic level to allow decisions to be fully optimised within known constraints.