Marlborough District Council Roading Assets - Activity Management Plan 2015 - 2018 SECTION 1 EXECUTIVE SUMMARY The Treatment Selection Algorithm (TSA) available in RAMM was originally used by Council to determine the immediate (or short term) and annual maintenance works programme for the road network. The RAMM TSA did not have a predictive modelling capability and did not consider the life cycle analysis for optimisation purposes. The predictive modelling analysis process is completed in three stages: Analysis of the RAMM database: Define network status and issues; Performance Based Analysis: determine the appropriate level of service; and Extended Performance Analysis: Prepare optimised multi-year work programme. Analysis of the RAMM database is undertaken to understand the current network conditions and issues requiring attention. Performance based analysis is used to define appropriate/optimum long-term maintenance strategies for the road network. Extended Performance Analysis has been undertaken to generate optimised multi-year works programmes based on the economic optimisation of the network condition to satisfy the selected long-term maintenance standards. Different funding scenarios are considered and the most practical one selected in preparing the Optimised Work Programme (OWP). Four scenarios are compared in the following table, namely the trigger based model and three optimised models. Three optimised budgets are used to demonstrate that the proposed funding request is at about the right level of investment to provide an acceptable level of service while ensuring that consumption of the asset is not occurring. The model funding scenarios are: Name Annual Budget Variance Trigger based Unlimited $ Optimised Normal 2.5 M $2,500,000* $2.5mil = summary of resurfacing, AWT and rehabilitation Optimised Very High 3.2 M $3,125,000 + 25% of Normal budget Optimised Very Low 1.9 M $1,875,000 - 25% of Normal budget Model Funding Scenarios Table 6.1.6.2a * The normal funding scenario is the proposed $ FWP for resurfacing and renewals The trigger based model is run first to validate the model inputs. The following graph shows the proposed quantity of treatments predicted by the pavement performance models. The Optimised ‘normal’ model is predicting quantities that align well with the current FWP (71km of resurfacing and 7km of AWT). Treatment Quantities - 20 year average 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 Resurfacing Pavement Optimsed Very Low 1.9 M 61,577 5,421 Optimised Normal 2.5 M 72,802 6,808 Optimised Very High 3.2 M 80,909 7,692 Trigger based 81,572 10,800 Treatment Quantities Figure 6.1.6.2a 1.6.1.7 Pavement Renewals 30 September 2014 Page 40 of Section 1