Modelling and Aprraisal

Modelling and Appraisal Note Note01Version: 1

1Introduction

1.1In February 2013, Sheffield City Council submitted a bid to the Local Pinch Point. The bid covers a scheme that addresses traffic congestion at 4 pinch-point junctions along a short section of the A61 Penistone Road, roughly 2km north of the city centre. Penistone Road is one of the most heavily trafficked radial routes into the city centre.

1.2The scheme will help to unlock the development potential of a key site in the Upper Don Valley, an area identified by the Local Enterprise Partnership as a key economic zone.

1.3Amongst the evidence required by the Department for Transport is a pro-forma setting out traffic flows and journey time savings. This information had been extracted from a transport model of the area and this note explains how the forecasts for the model are robust and fit for the purpose of appraising the proposed scheme.

2Source of the Forecasts

2.1The forecasts in the pro-forma come from the Sheffield and Rotherham Transport Model version 3 (SRTM3). SRTM3 is a multi-modal variable demand model prepared specifically to support major scheme bids for two schemes:

Penistone Road Smart Route; and

South Yorkshire Bus Rapid Transit Southern Route

2.2The first of these is a park and ride scheme whose route runs through the junctions included in the local pinch point funding proposal.

2.3The second is a scheme that the DfT has approved for funding, following extensive scrutiny of the SRTM3 model.

2.4The forecasts used in the pro-forma were taken from the models used to support the major scheme business case for the Penistone Road Smart Route. They therefore include a small amount of mode-shift away from car that would not apply in the absence of the park and ride service.

3Model Development

Overview

3.1The SRTM3 model used in the appraisal is a multi-modal model of Sheffield. It was specifically tailored to support a major scheme business case for the Penistone Road scheme. The model network and zoning system was refined in the area of influence of the scheme, new data were gathered for calibration and validation, and the trip matrices were rebuilt from first principles using the new zone system.

3.2The model was built with a base year of 2009 and forecast years of 2015 and 2029. It was calibrated and validated for the following time periods:

morning peak 0800 to 0900;

inter-peak average 1000 to 1600; and

evening peak 1700 to 1800.

3.3The model was used to support the Bus Rapid Transit scheme through the Best and Final Financial Bid process. It has undergone detailed scrutiny from the Department for Transport.

Network

3.4The network represents every motorway, A Road, B Road and C Road. A selection of unclassified roads are also represented so that every road on which buses travel is included.

Matrices

3.5The matrices for the model were built from a series of roadside interview surveys, which form 9 cordons around areas of the city. Two of the survey sites are on Penistone Road, one at the Northern end of the treated area and one at the southern end. Their location can be seen in Figure 3.1.

3.6The matrices were infilled using population estimates and trip-rates to produce trip-ends and gravity models used to distribute the trips. The gravity models were calibrated to match the trip-length distributions observed in the roadside interview surveys. Separate models were created for 12 journey purposes.

Figure 3.1 Roadside Interview Survey Sites and Cordons

Calibration and Validation

Traffic Counts

3.7The model has been calibrated to match classified counts undertaken at all the key intersections along the route.

3.8The fit between modelled flows and counts is very good in the Penistone Road corridor, comfortably exceeding the WebTAG guidelines of 85% in the morning peak and inter-peak, and coming very close in the evening peak.

Table 3.1 Fit to Counts in the Penistone Road Corridor

3.9The model has also been validated against a set of 88 independent counts, a set kept entirely separate from the calibration counts. Against this set of counts, the overall demand in the model validates very well in the morning and evening peak - to within +/- 3%. In the inter-peak the total modelled flows matches less well – to within 13%.

3.10At the level of individual counts, the inter-peak achieves a better fit than the peaks. On the DMRB measure, 51% of links pass in the morning peak, 46% in the evening peak, and 68% in the inter-peak. Unsurprisingly, these values are well below the levels achieved for the calibration count set. However, the good fit of the modelled totals indicates that overall the model is sound. This conclusion is supported by the very high R squared values, which show that in every time period the model explains more than 90% of the variation between counts (it has been calculated for the line Y=X rather than Y=aX +b).

Table 3.2 Validation Against Independent Counts

Journey Times

3.10.1The model has been validated against a set of journey times on 24 routes, two of which traverse the Penistone Road Corridor.

3.10.2As Table 3.3 shows, the journey time validation is generally good in the Penistone Road Corridor (OK in 8 out of 12 cases) and there is no systematic bias, the model is just as likely to be fast (2 cases out of 12) as slow (2 cases out of 12).

3.10.3During the peaks, Route 17 validates well in both directions and Route 9 validates well inbound but outbound it is slow in the morning and fast in the evening. In the inter-peak inbound, Route 9 runs slow and Route 17 runs fast.

Table 3.3 Journey Time Validation

4Forecasting

Without Scheme

4.1The model forecasts have been taken from historic work used for the major scheme business case for the Penistone Road Smart Route. When the work was undertaken, the latest version of the TEMPRO software was 6.0 but the latest available data set was version 5.4. Version 5.4 had slightly higher car growth compared to Version 6.2, which is currently the latest version.

4.2The forecasting approach used an uncertainty log and included in its core scenario only those developments that had already received planning consent or had submitted a planning application that was expected to be granted consent.

4.3The growth forecasts were constrained to TEMPRO Growth and then the multi-modal model was run to take account of changes in the levels of congestion, and growth in the value of time.

With Scheme

4.4The scheme affords journey time savings along Penistone Road. The levels of traffic flow inbound on the treated sections of Penistone Road are shown in Table 4.1.

Table 4.1 Morning Peak Traffic Flow (pcu/hr) Towards Sheffield City Centre

5Estimated Present Value of Benefits

5.1A rough and ready estimate of benefits has been calculated from the value of time and the reduction in journey times in the pro-forma. This estimate suggests an overall benefit of around £20m, split roughly equally between business and consumers.

6Completing the Pro Forma

6.1The traffic flows and time savings in the pro-forma have been taken from the 2014 model, which is the closest to the opening year.

6.2The hour to period factors have been derived from local counts:

Morning Peak 2.61

Interpeak 6.00

Evening Peak 2.80

6.3The journey purpose splits have been taken from the transport model.

Modelling and Appraisal
1