DEVON COUNTY COUNCIL

NORTH DEVON TRANSPORT MODEL

LOCAL MODEL VALIDATION REPORT

OCTOBER 2015

CONTENTS

Page

1. INTRODUCTION 1

1.1 Introduction 1 1.2 Model Specification 1

2. NETWORKS 2

2.1 Highway Network 2 2.2 SATURN Coding 3

3. CALIBRATION & VALIDATION DATA 8

3.1 Traffic Growth 8 3.2 Traffic Counts 9 3.3 Journey Time Data 11

4. TRIP MATRICES 13

4.1 Highway Matrix Building Procedure 13 4.2 Car Trip Ends 13 4.3 Car Distribution 14 4.4 Matrix Building 18 4.5 Matrix Estimation 21

5. TRAFFIC MODEL CALIBRATION 27

5.1 Procedure 27 5.2 Highway Network 27 5.3 Traffic Assignment 30 5.4 Generalised Costs 30 5.5 Model Convergence 31

6. TRAFFIC MODEL VALIDATION 32

6.1 Validation Criteria & Acceptability Guidelines 32 6.2 Network Validation 33 6.3 Journey Time Validation 34 6.4 Validation at A361 Portmore RSI Site 36 6.5 Traffic Flow Validation 38

7. SUMMARY & CONCLUSIONS 39

TABLES

Table 1: SATURN Junction Simulation Parameters for Extended Network ...... 3 Table 2: Traffic Growth ...... 8 Table 3: Details of Traffic Counts ...... 9 Table 4: Average Journey Times ...... 11 Table 5: Model Time Period Factors ...... 14 Table 6: Calibration of Gravity Models ...... 16 Table 7: Time Period Proportions ...... 18 Table 8: Composition of Initial Trip Matrices ...... 20 Table 9: Matrix Adjustment to Cordon & Screenline Counts ...... 21 Table 10: Initial Calibration of Trip Matrices ...... 21 Table 11: SATME2 Results ...... 22 Table 12: Trip Matrix Totals After Matrix Estimation ...... 22 Table 13: WebTAG Matrix Estimation Performance ...... 23 Table 14: 2014 Sector Trip Matrices After Matrix Estimation ...... 25 Table 15: 2013 PPM and PPK values ...... 31 Table 16: Convergence Results – Base Year ...... 31 Table 17: Screenline Flow Validation Criteria ...... 32 Table 18: Link Flow and Turning Movement Validation Criteria ...... 32 Table 19: Journey Time Validation Criteria ...... 32 Table 20: AM Peak Journey Time Validation ...... 34 Table 21: Inter Peak Journey Time Validation ...... 34 Table 22: PM Peak Journey Time Validation ...... 35 Table 23: Screenline Traffic Flow Validation ...... 38 Table 24: Summary of Validation of Link Flows ...... 38 Table 25: Summary of Model Validation ...... 39

FIGURES

Figure 1: Barnstaple & Fremington Zones ...... 4 Figure 2: & Northam Zones ...... 5 Figure 3: External Zones ...... 6 Figure 4: Base Year Highway Network ...... 7 Figure 5: Traffic Count Locations ...... 10 Figure 6: Journey Time Routes ...... 12 Figure 7: Calibration of Infill Gravity Models ...... 17 Figure 8: Sector System ...... 19 Figure 9: Trip Length Distributions ...... 24 Figure 10: Selected Modelled Routes ...... 28 Figure 11: Sector Movements at RSI Sites 1 & 2 ...... 37

APPENDIX 1 – Time / Distance Diagrams

APPENDIX 2 – Traffic Flow Validation

LOCAL MODEL TRANSPORT MODEL VALIDATION REPORT

1. INTRODUCTION

1.1 Introduction

1.1.1 Devon County Council have updated the SATURN based transport model for Barnstaple and Bideford to inform their ongoing response and contribution to the joint Local Plan in order to identify transport schemes and measures to support the Local Plan. The aim of the transport study for Barnstaple and Bideford is to assess the impact of future development and identify appropriate transport infrastructure requirements.

1.1.2 The previous SATURN highway assignment model was developed for the working weekday AM and PM peak hours and an average inter peak hour, and has been updated to the new 2014 base year. The network and zoning was reviewed and extended to reflect the level of present trip generation and attraction and the likely future development areas. Validation was aimed to achieve near DMRB/Webtag critieria.

1.1.3 This report describes the development and validation of the updated base year transport model. The data presented shows that the model meets the required WebTAG validation criteria of networks, traffic flows, journey time, model convergence and trip distribution.

1.2 Model Specification

1.2.1 The detailed SATURN simulation area covers Barnstaple, Bishops Tawton, Landkey, Fremington and Yelland in North Devon District and Bideford and Northam in . Other parts of Devon and the rest of the UK are included as external zones.

1.2.2 The updated SATURN based highway model has been developed for the following time periods in the 2014 base year:

. AM peak hour (08:00 to 09:00 hrs); . Inter peak (IP) hour (average hour 10:00 to 16:00 hrs); . PM peak hour (17:00 to 18:00 hrs); 1.2.3 The model performs multiple user class assignment with the following user classes:

. UC1 – Car - Commuting (COM); . UC2 – Car - Employer’s business (EMP); . UC3 – Car - Education (EDU); . UC4 – Car – Other purposes (OTH); . UC5 – Light goods vehicles (LGV); . UC6 – Other goods vehicles (HGV).

NDevonNewModelLMVRV1 Page 1 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

2. NETWORKS

2.1 Highway Network

2.1.1 The previous SATURN highway network was extended to in the Bideford area with a more detailed network and a finer 166 zone system. Internal zones have been increased from 121 to 138 to include development zones in Barnstaple and Bideford, as shown on Figure 1 and Figure 2. The 28 external zones, Figure 3, have been retained but were linked to an external road network to replace traffic loading point zones on the external modelling cordon.

2.1.2 Link information was obtained for sections of similar characteristics, including:

. Width . Number of lanes . One or two-way . Central reserve width (if any) . Extent of kerbside parking (if any) . Extent of bus lane, times of operation, vehicle types . Speed limit . Location of bus stops . Location of controlled and uncontrolled pedestrian crossings . Lane and road markings

2.1.3 Junction information was collected for all junctions, including:

. Type of junction . For each approach no. of lanes and widths and lane length for short approach lanes . Lane and road markings indicating turning movements . Stop lines and give way lines . Prohibited turns . For traffic signals - cycle times, no. and length of stages, intergreen times, pedestrian phases, permitted movements for each stage etc.

2.1.4 Highway link and junction details were obtained from:

. Devon County Council’s highway information GIS data; . Traffic signal data from Devon County Council; . Site visit and Google aerial mapping to check for completeness and for recent changes.

NDevonNewModelLMVRV1 Page 2 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

2.2 SATURN Coding

2.2.1 The highway network was coded into SATURN format with the main trafficked roads and significant junctions included, see Figure 1. The coded base year network consists of 1,346 simulation road links (single direction) and 515 modelled nodes with 485 priority, 8 roundabouts and 22 traffic signal junctions. Most roundabouts were coded as a series of priority junctions.

2.2.2 The highway network was extended outside the external modelling cordon to provide improved modelling of internal-external and external-external trips and to provide full costs for these movements required for economic assessment.

2.2.3 The main bus routes serving Barnstaple and Bideford were coded into the SATURN highway network

2.2.4 The whole model area was included in the simulation area and all junctions, apart from minor and access roads, were simulated in SATURN. Speed/flow relationships were included only for the long links on the A39 and A361. Junction coding from the previous model was reviewed and generally retained. For junctions in the extended network a simplified coding method was adopted with standardised saturation capacities and gap sizes.

Table 1: SATURN Junction Simulation Parameters for Extended Network Saturation Flow Minimum Gap Parameter (pcu/hr) (sec) 1. Priority: Minor Arm Left Turn Right Turn Good Visibility 1200 875 Good vis.: 5.0 Poor Visibility 1100 800 Poor vis.: 5.5 Bad Visibility 1000 740 Bad vis.: 6.0 Major Arm 2-lane 4-lane 2000 3000 (in each direction) 2. Roundabout Approach: Dual 2 lane 2200 Standard: 1.5 Flared 1650 Mini: 2.5 1 lane 1100 Circulation: 2400 3. Traffic Signals Left turn 1500 Opposed right Straight 1800 turn: 3.5 Right turn 1500

NDevonNewModelLMVRV1 Page 3 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 1: Barnstaple & Fremington Zones

NDevonNewModelLMVRV1 October 2015 Page 4

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 2: Bideford & Northam Zones

NDevonNewModelLMVRV1 October 2015 Page 5

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 3: External Zones

NDevonNewModelLMVRV1 October 2015 Page 6

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 4: Base Year Highway Network

BARNSTAPLE

A361

NORTHAM

A39

BIDEFORD

NDevonNewModelLMVRV1 October 2015 Page 7

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

3. CALIBRATION & VALIDATION DATA

3.1 Traffic Growth

3.1.1 ATC data in Barnstaple and Bideford was analysed to identify the level of traffic growth from the 2008 base year of the previous traffic model as shown in Table 2 . This showed that over the 6 year period 2008 to 2014 traffic flows fluctuated with low levels of growth and an overall increase of 4%. The traffic count data used to calibrate and validate the new traffic model was largely for 2012 and 2014, see Section 3.2 below, and traffic growth over this period was 1.2%.

3.1.2 As the levels of traffic growth was insignificantly low and variable over the period when calibration and validation traffic count data was collected it was decided not to apply growth to normalise the data. For the same reasons it was considered that the data was also consistent with base year of 2014 for the new model.

Table 2: Traffic Growth

Location 2008 2009 2010 2011 2012 2013 2014 1. A39 Abbotsham Cross 6567 7690 7507 7848 7365 7369 7670 2. A386 South of Bideford 8248 8493 8551 8827

3. A377 Bishops Tawton 6830 6859 6667 6774 6696 6616 6898 4. A361 Ashford 18648 18776 18498 18444 18150 18331 18161 5. A39 North Rd, Barnstaple 11875 12172 12167 12212 12084 11832 12345 6. Barnstaple Bypass South 15024 16118 16310 17369

7. Braunton Road Barnstaple 18130 17532 17424

8. Barnstaple Bypass North 21197 22197 22243 23395 23873 24499 24537 9. B3233 Bickington 11964 11808 11629 11658

10. Barnstaple Longbridge 16778 16712 14744 17125 17035 16757 17407 11. Eastern Ave, Barnstaple 14848 14807 14586 13895 13559 13407 13627 12. A39 Roundswell 23756 24241 24199 24207 23879 24597 24186 Total 1,3,4,5,8,10,11,12 120499 123454 120611 123900 122641 123408 124831 Growth From 2008 100% 102% 100% 103% 102% 102% 104%

NDevonNewModelLMVRV1 Page 8 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

3.2 Traffic Counts

3.2.1 Traffic count data was collected from Devon’s manual classified and automatic count database and from Transport Assessments submitted in support of development planning applications. Details of the traffic count data is shown in Table 3

Table 3: Details of Traffic Counts Ref. Date Location 1 2009 A39 Heywood Rd Rbt 2 2012 A39 North Rd / Westaway Plain 3 2012 A3125 Cedars Rbt 4 2012 A3125 Old Bideford Rd Rbt 5 2012 A39 Roundswell Rbt 6 2012 A39 Lake Rbt 7 2012 A361 Rumsam Rbt 8 2012 A3125 Old Torrington Rd Mini Rbt 9 2013 A39 Pilton Causeway / Alexandra Rd 10 2013 A39 Alexandra Rd / Belle Meadow Rd 11 2013 A39 Eastern Ave / Hollowtree Rd 12 2012 A361 Taw Bridge Approach / Riverside Rd 13 2012 A361 Taw Bridge / Braunton Rd 14 2012 Manteo Way 15 2012 A386 New Road / Torridge Hill 16 2012 Old Town / Meddon St 17 2012 Abbotsham Rd / Belvoir Rd 18 2012 A39 Buckleigh Rd 19 2012 Longbridge / The Quay 20 2012 A39 Big Sheep 21 2012 A39 Abbotsham Cross 22 2012 Clovelly Rd / Atlantic Village 23 2012 A361 Squareabout 24 2012 A361 Portmore Rbt 25 2009 A39 Westleigh 26 2012 Hollowtree Rd / Newport Rd 27 2012 A39 Eastern Ave / Tesco Rbt 1 2015 A361 NDL E of Portmore 2 2015 A377 Bishops Tawton 3 2012 A39 Rumsam Bridge 4 2014 Barnstaple Longbridge 5 2014 A361 Taw Bridge Approach 6 2014 A361 Braunton Rd

NDevonNewModelLMVRV1 Page 9 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

7 2014 A39 North Rd 8 2014 A39 E of Westleigh 9 2014 A39 W of Abbotsham Cross 10 2014 A386 S of Bideford

Figure 5: Traffic Count Locations

BARNSTAPLE

KEY: MCC & TA ATC

BIDEFORD

NDevonNewModelLMVRV1 Page 10 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

3.3 Journey Time Data

3.3.1 Trafficmaster journey time data on roads in Barnstaple and Bideford was obtained from the Strat-e-gis congestion monitoring database system for Local Authorities. Trafficmaster is a service that provides vehicle fleet management and GPS tracking and also provides historic journey time and speed data from over 100,000 vehicles across the UK road network. The DfT provide access to the data linked to the road network of the Ordnance Survey’s Integrated Transport Network (ITN).

3.3.2 Journey time data was extracted for working weekdays in the period 1 April to 30 June 2014 for A and B roads and a selection of C roads in the model area. It became clear that the roundabout works at Roundswell caused major diversions and delays during the January to September 2014 construction period and so it was decided instead to use data for the 1 April to 30 June 2013 period for the affected roads.

3.3.3 Average journey times for all vehicle types were obtained for each ITN link in both directions, and were summed to give journey time estimates given in Table 4 for the routes shown on Figure 6.

Table 4: Average Journey Times

Inbound Outbound Route AM IP PM AM IP PM 1. Chaddiford Lane - The Square 06:36 06:43 07:54 07:03 07:08 08:52 (2.8km) 2. N Devon Hospital – The Square 05:44 06:10 07:26 07:44 07:04 08:35 (2.9km) 3. Landkey – The Square via 06:55 06:21 08:36 05:42 06:10 05:57 Portmore (4.2km) 4 - Landkey – The Square via 07:26 06:33 05:23 07:42 06:22 06:37 Newport (4.4km) 5. Roundswell – The Square via 05:17 05:12 04:31 04:39 04:46 06:42 Lake (3.8km) 6. Roundswell – The Square via 08:01 06:14 06:00 06:57 06:04 07:19 Cedars (3.3km) 7. Abbotsham Cross – Westleigh 04:52 04:54 05:47 04:49 04:46 04:37 via A39 (5.3km) 8. Abbotsham Cross – Westleigh 09:29 09:37 09:33 09:44 09:58 10:06 via Bideford (6.3km) 9. Westleigh – The Square via A39 17:07 12:19 11:29 11:33 11:39 13:53 Roundswell (12.9km) 10. Westleigh – The Square via 21:08 18:08 17:52 16:14 17:58 18:23 B3233 Cedars (13.0km)

NDevonNewModelLMVRV1 Page 11 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Figure 6: Journey Time Routes

Route 1: Chaddiford Lane – Route 2: North Devon The Square Hospital – The Square

Route 6: Roundswell – The Square via Cedars Route 3: Landkey – The Square via Portmore

Route 10: Westleigh – The Square via Cedars

Route 4: Landkey – The Square via Newport Route 5: Roundswell – The Square via Lake

Route 7: Abbotsham Cross – Westleigh via A39 Route 9: Westleigh – The Square via Roundswell

Route 8: Abbotsham Cross – Westleigh via Bideford

NDevonNewModelLMVRV1 Page 12 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

4. TRIP MATRICES

4.1 Highway Matrix Building Procedure

4.1.1 The previous model trip matrices were built using roadside interview data from surveys carried out at 6 sites in 2008. It was considered that the origin- destination data from these surveys was out of date and as no subsequent RSI’s had been carried out it was decided to use the National Trip End Model (NTEM) to develop car trip ends for the new 2014 base year.

4.1.2 Car trip ends were synthesised according to NTEM methods as set out in the WebTAG Supplementary Guidance using the NTEM databases and trip end program (CTripEnd50.exe). Population and employment input data was obtained from the 2011 Census data.

4.1.3 The car trip ends were then distributed using gravity models developed for the previous model from the 2008 RSI data.

4.1.4 Education trips in the AM peak hour were directly obtained from school student data which allowed the origin-destination matrix to be directly obtained from information on student home location and mode of travel for each primary and secondary school. Data was available from the 2011 database held by Devon County Council and for 2014 from the Ecocheck organisation. Traffic zones were identified for home and school locations for car trips, the data expanded to total student numbers and the two sources averaged to obtain an origin-destination matrix and it was assumed that these trips would be made in the AM peak hour. The data did not provide information on trips made in the other model hours and so the NTEM method was used for the IP and PM model periods.

4.1.5 National Road Traffic Forecast growth was applied to the 2008 LGV and HGV trip matrices to produce the new 2014 base year trip matrices.

4.2 Car Trip Ends

4.2.1 The NTEM 2011 and 2016 databases were interpolated to give trip end data control totals for the 2014 base year for the 8 NTEM zones covering the model area. The land use, population and mode of travel data required by the NTEM trip end calculations was obtained from the 2011 Census as follows.

4.2.2 Population data for the 88 categories by age, employment status, household size and car availability was obtained from the following 2011 Census datasets:

. S001-S008 Children 0-15 years – LC4109EW Car or van availability by age and sex; . S009-S088 Adults – LC4609EW Car or van availability by economic activity.

NDevonNewModelLMVRV1 Page 13 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

4.2.3 Mode split data for the 12 categories consisting of 6 each for workplaces and home (walk, cycle, car driver, car passenger, bus and rail) was obtained from the following 2011 Census datasets:

. K01-K08 Workplace – WP703EW Method of travel to work (workplace zones); . K07-K12 Residence – QS701EW Method of travel to work (output areas). 4.2.4 Employment data by workplace zone for the 15 categories was obtained from 2011 Census dataset WP605EW - Industry (Workplace population) where the industry classification is based on the UK Standard Industrial Classification of Economic Activities (UKSIC).

4.2.5 The derived population data by output zone and employment data by workplace zone was then converted to study zones.

4.2.6 The trip end program was then used to output 2014 trip ends for each zone by purpose, mode and time period and the 15 output purposes collated to the 4 car purposes used in the model (commuting, employer’s business, education and other purposes).

4.2.7 NTEM trip ends are output for 5 working weekdays and the factors shown in were applied to convert to a single weekday and then from the NTEM time periods (AM 07-10hrs, IP 10-16hrs and PM 16-19hrs) to the model hours (AM 08-09hrs, IP average 10-16hrs and PM 17-18hrs).

Table 5: Model Time Period Factors

Week to Time Period to Average Period Model Hour Weekday AM 0.2 0.4075 IP 0.2 0.1667 PM 0.2 0.3696

4.3 Car Distribution

4.3.1 The base year car trip ends derived using the NTEM method were distributed using gravity trip distribution models calibrated using 2008 RSI data for the previous model. The gravity models were calibrated for private vehicle purposes in the AM peak hour, inter-peak hour and PM peak hour and the calibration results, see Table 6 and Figure 7, show close correspondence in each time period between observed and modelled data with high values of the correlation coefficient between the two sets of data. It was found that a β value of -0.005 produced close calibration for all user classes in all time periods.

NDevonNewModelLMVRV1 Page 14 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

4.3.2 The form of the gravity used was:

Tij = Ai Bj Oi Dj dij exp( -βdij )

Where:

Tij = trips from zone i to zone j

Oi, Dj = origin and destination trip totals

Ai, Bj = balancing factors

dij = distance of travel from zone I to zone j

β = calibration coefficient

NDevonNewModelLMVRV1 Page 15 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Table 6: Calibration of Gravity Models

AM Peak Hour Inter-Peak Hour PM Peak Hour Calibration Parameters UC3 UC3 UC3 & Indicators by Time UC1 UC2 EDU & UC1 UC2 EDU & UC1 UC2 EDU & Period & User Class COM EMP UC4 COM EMP UC4 COM EMP UC4 OTH OTH OTH Coefficient β -0.005 -0.005 -0.005 -0.005 -0.005 -0.005 -0.005 -0.005 -0.005 Total Trips: - Observed 533 146 549 93 169 1025 275 153 1207 - Modelled 532 146 549 93 169 1025 275 153 1207 Av. Trip Distance (km): - Observed 8.4 8.8 5.7 6.9 7.4 6.9 8.6 8.0 5.8 - Modelled 8.2 8.5 5.8 7.1 7.4 6.8 8.4 7.8 5.8

NDevonNewModelLMVRV1 Page 16 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 7: Calibration of Infill Gravity Models

TRIP LENGTH DISTRIBUTION - AM PEAK - UC1 HBW TRIP LENGTH DISTRIBUTION - AM PEAK - UC2 EMB TRIP LENGTH DISTRIBUTION - AM PEAK - UC3 OTH

18.0 20.0 25.0 Observed Observed 16.0 18.0 Observed Modelled Modelled Modelled 14.0 16.0 20.0 14.0 12.0 12.0 15.0 10.0 10.0 8.0

8.0 10.0

% OF TRIPS OF % TRIPS OF % 6.0 TRIPS OF % 6.0 4.0 4.0 5.0 2.0 2.0

0.0 0.0 0.0

TRIP LENGTH (KM) TRIP LENGTH (KM) TRIP LENGTH (KM)

TRIP LENGTH DISTRIBUTION - IP - UC1 HBW TRIP LENGTH DISTRIBUTION - IP - UC2 EMB TRIP LENGTH DISTRIBUTION - IP - UC3 OTH

20.0 20.0 25.0

18.0 Observed 18.0 Observed Observed Modelled Modelled Modelled 16.0 16.0 20.0

14.0 14.0

12.0 12.0 15.0

10.0 10.0

8.0 8.0 10.0

% OF TRIPS OF % TRIPS OF % % OF OF TRIPS % 6.0 6.0

4.0 4.0 5.0

2.0 2.0

0.0 0.0 0.0

TRIP LENGTH (KM) TRIP LENGTH (KM) TRIP LENGTH (KM)

TRIP LENGTH DISTRIBUTION - PM PEAK - UC1 HBW TRIP LENGTH DISTRIBUTION - PM PEAK - UC2 EMB TRIP LENGTH DISTRIBUTION - PM PEAK - UC3 OTH

18.0 20.0 25.0 Observed Observed Observed 16.0 18.0 Modelled Modelled Modelled 14.0 16.0 20.0 14.0 12.0 12.0 15.0 10.0 10.0 8.0

8.0 10.0

% OF TRIPS OF % TRIPS OF % 6.0 TRIPS OF % 6.0 4.0 4.0 5.0

2.0 2.0

0.0 0.0 0.0

TRIP LENGTH (KM) TRIP LENGTH (KM) TRIP LENGTH (KM)

NDevonNewModelLMVRV1 Page 17 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

4.4 Matrix Building

4.4.1 The gravity distribution models were calibrated for trips within the external cordon of the previous model and were therefore only applied to internal zones in the updated model.

4.4.2 The proportions of the trip ends with internal origins or destinations were estimated using 2001 Census origin-destination journey to work data. This was available at the output area level and was converted to model zones. Origin-destination data from the 2011 Census was only available at the middle super output area level which are generally much larger than model zones. It was decided to use the 2001 data as the distributions of external work trips would not have changed significantly since 2001 and the splitting of 2011 middle super output area level data to model zones would have introduced other inaccuracies.

4.4.3 The derived 2001 journey to work trip matrix was converted to model time periods using proportions derived from TEMPRO NTEM data for North Devon and Torridge, see Table 7. This enabled the production-attraction daily trip matrix to be converted to origin-destination trip matrices by time period. It was assumed that the derived time period proportions were applicable to the modelled hours.

Table 7: Time Period Proportions

AM IP PM

To Work 67% 15% 6% From Work 5% 21% 58%

4.4.4 The gravity models were applied to the proportioned trip ends to produce trip matrices of internal movements for each user class. The same internal trip proportions were assumed for all user classes. The internal proportions were applied at a sector level with the sectors defined below. Sectors 1 to 5 represent the internal area and are shown on Figure 2:

. Barnstaple, Bishops Tawton, Landkey . Roundswell, Bickington, . Fremington, Yelland, . Bideford East the Water . Bideford, Northam, Appledore . South Molton . Braunton, Ilfracombe, Lynton . A377, Crediton, Exeter Area . A386, Torrington, Plymouth . Cornwall, Clovelly . , Rest of UK

NDevonNewModelLMVRV1 Page 18 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Figure 8: Sector System

7 – Braunton, Ilfracombe, Lynton 11 – Somerset, Rest of UK

1 – Barnstaple, 2 – Roundswell, Bishops Tawton, Bickington Landkey

6 – South Molton

3 – Fremington, 5 – Bideford, Yelland, Instow Northam, Appledore 4 – Bideford East the Water

8 – A377, Crediton, 10 – Clovelly, 9 – A386, Exeter Area Cornwall Torrington, Plymouth

NDevonNewModelLMVRV1 Page 19 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

4.4.5 The distribution of trips between internal and external zones was estimated from the journey to work trip matrices derived for the model time periods. The journey to work trip matrices were factored to external trip end totals for each user class. In this way the sum of internal and external trips equalled the total NTEM trip ends estimated for each zone in the model time periods.

4.4.6 A number of external to external movements that were judged to pass through the model area were also included. These comprised movements to and from Braunton, Ilfracombe and Lynton (sector 7).

Table 8: Composition of Initial Trip Matrices

AM IP PM

Internal-Internal 10,936 4,415 5,608 Internal-External 1,881 1,952 2,889 External-Internal 2,404 1,846 2,313 External-External 1,039 290 1,099 Total 16,260 8,503 11,909 Note: Car user classes only Vehicles/hour

4.4.7 The derived base year trip matrices were then calibrated to the totals of counts on a number of cordons and screenlines in the model area, represented by the sector boundaries shown on Figure 8. The 2008 base year LGV and HGV user class trip matrices were factored to the new 2014 base year using 2015 NTEM forecast growth and included with the car matrices. The all vehicle trip matrices were then assigned to the updated model networks and select link assignments carried out for links corresponding to each of the count sites. The select link trip matrices were then summed to give matrices for each of the cordons and screenlines.

4.4.8 The corresponding origin-destination movements in the full matrices were then factored so that total trips in the select link matrix portions were the same as the count totals. This was repeated for each cordon and screenline and 10 iterations of the process carried out so that the matrix totals converged to the count totals.

4.4.9 The differences between the initial and calibrated matrices were quite small, as shown in Table 1, which validated the matrix building method using NTEM trip ends, the calibrated gravity model and the 2001 journey to work distribution.

NDevonNewModelLMVRV1 Page 20 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Table 9: Matrix Adjustment to Cordon & Screenline Counts AM Peak Hour Inter Peak Hour PM Peak Hour Cordon/ Screenline Matrix Matrix Matrix Count Count Count Before After Before After Before After Barnstaple Inbound 5788 5994 5769 3993 4062 3984 4840 4862 4811 Barnstaple Outbound 4614 5593 4612 4111 4211 4104 5794 5753 5804 EW Screenline Eastbound 1257 2348 1273 1138 1942 1147 1075 1908 1097 EW Screenline Westbound 875 2011 882 1164 2033 1171 1806 2156 1806 Bideford Inbound 739 941 739 673 620 673 755 737 758 Bideford Outbound 572 734 573 642 657 642 881 866 881

4.4.10 Because the initial trip matrices had been fully synthesised it was necessary to derive sector to sector calibration factors to improve the representation of traffic movements. Matrix estimation was used iteratively to derive the sector to sector factors which were then applied at the zonal level. This method adjusted sector to sector movements to traffic count data without altering the traffic distributions within sectors. The change in trip matrix totals is shown in Table 10.

Table 10: Initial Calibration of Trip Matrices

AM IP PM

Initial Matrix Build 18,115 9,900 13,148 With Cordon / Screenline Adjustments 16,045 9,054 12,477 Difference -2,070 -845 -671 With Calibration Factors 16,931 11,207 14,549 Difference +886 +2,153 +2,072 Note: All user classes (Car, LGV & HGV) Vehicles/hour

4.5 Matrix Estimation

4.5.1 Matrix estimation was used to calibrate the initial trip matrices to traffic count data using available ATC and MCC data. Six iterations of SATURN assignment and the SATME2 matrix estimation program were carried out using traffic count data on 167 entries, exits and turns at 25 junctions and in both directions at 10 permanent ATC sites. In total 187 traffic flow values in the peak hours and 151 values in the inter peak were included in SATME2.

It was found that the matrices before SATME2 produced a reasonable correspondence with observed data but after matrix estimation the correspondence was close, as shown on Table 11. Changes to the matrices caused by SATME2 were quite small overall, a high proportion of cells remained unchanged and very few matrix cells changing by more than 20 vehicles. GEH values for the comparison of observed and modelled counts were less than 6.0 for 94% of links in the AM, 93% in the PM peak hours peak hour and 95% in the inter-peak hour.

NDevonNewModelLMVRV1 Page 21 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Table 11: SATME2 Results

AM Peak Hour Inter Peak Hour PM Peak Hour Measure Before After Before After Before After Observed/Modelled Flows -3% 0% -8% +1% -7% -1% Total Average Difference – -13 -1 -27 -3 -34 -4 Absolute Regression - R-squared 0.81 0.98 0.77 0.94 0.75 0.93 Average GEH 6.38 1.19 5.70 1.13 6.93 1.48 GEH – 0 to 6 56% 95% 58% 95% 52% 94% GEH – 6 to 10 23% 3% 29% 4% 24% 4% GEH – above 10 21% 2% 13% 1% 24% 2% 17,509 12,392 15,763 Matrix Total 16,931 11,207 14,549 +3% +11% +8% Cells – -1 to +1 - 94% - 95% - 93% Cells – +/-1 to +/- 20 - 6% - 5% - 7% Cells - > +/- 20 - 0% - 0% - 0%

Table 12: Trip Matrix Totals After Matrix Estimation User Class AM IP PM UC1 – Car etc - Home Based Work 5,061 1,355 4,437 UC2 – Car etc - Employers Business 962 515 1,016 UC3 – Car etc - Education 5,677 689 579 UC4 – Car etc - Other Purposes 3,693 7,861 7,685 UC5 – LGV 1,809 1,549 1,806 UC6 – HGV 306 424 239 Total (pcu/hr) 17,509 12,392 15,762

Matrix estimation is intended to be used to refine estimated matrices and the WebTAG matrix estimation criteria are very strict and if followed rigorously effectively eliminate any significant matrix changes. It is therefore not surprising that the WebTAG matrix estimation criteria were only partially met as shown in Table 13. However, the differences from the criteria are not generally excessive and the results shown in Table 12 show that matrix estimation improves the accuracy of link flow modelling without causing excessive changes to matrix cells.

NDevonNewModelLMVRV1 Page 22 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Table 13: WebTAG Matrix Estimation Performance

Significance Criteria AM IP PM Matrix Zonal Cell Values:

Slope within 0.98 and 1.02 0.96 1.029 0.97 Intercept near zero (absolute & % 0.05 0.03 0.06 of mean) (3.7%) (3.5%) (5.4%) R2 in excess of 0.95 0.71 0.68 0.566 Matrix Zonal Trip Ends

Slope within 0.98 and 1.02 0.96 1.03 0.96 Intercept near zero (absolute & % 7 5 11 of mean) (6.8%) (7.0%) (11.4%) R2 in excess of 0.95 0.91 0.82 0.82 Trip Length Distributions:

Means within 5% -6.2% diff. -16.2% diff. -12.7% diff. Standard deviations within 5% -3.9% -11.8% -10.6% Sector to sector level matrices:

75% of cells 72% of cells 65% of cells Differences within 5% (GEH 88%) (GEH 89%) (GEH 88%) Note: WebTAG Unit M3-1 Highway assignment modelling Table 5 Significance of matrix estimation changes

NDevonNewModelLMVRV1 Page 23 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Figure 9: Trip Length Distributions

Trip Length Distribution - AM

50% 45% 40% 35% 30% 25%

Trips (%) Trips 20% 15% 10% 5%

0%

5

-

15 20 25 30 70 75 80 85 90 95 35 40 45 50 55 60 65

10

------

-

0

105 110 150 155 115 120 125 130 135 140 145

100

------

-

5

10 15 20 25 65 70 75 80 85 90 30 35 40 45 50 55 60

95

100 105 145 150 110 115 120 125 130 135 140 Distance (km)

Prior Post

Trip Length Distribution - IP

30%

25%

20%

15%

Trips (%) Trips 10%

5%

0%

Distance (km)

Prior Post

Trip Length Distribution - PM

25%

20%

15%

Trips (%) Trips 10%

5%

0%

5

-

15 20 25 30 70 75 80 85 90 95 35 40 45 50 55 60 65

10

------

-

0

105 110 150 155 115 120 125 130 135 140 145

100

------

-

5

10 15 20 25 65 70 75 80 85 90 30 35 40 45 50 55 60

95

100 105 145 150 110 115 120 125 130 135 140 Distance (km)

Prior Post

NDevonNewModelLMVRV1 Page 24 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Table 14: 2014 Sector Trip Matrices After Matrix Estimation

From/To Time Period 1 2 3 4 5 6 7 8 9 10 11 Total Sector AM Peak Hour - 08.00-09.00hrs (all vehicles/hour): Barnstaple, Landkey 1 3304 591 234 88 293 170 687 167 39 51 76 5701 Roundswell, Bickington, 2 908 111 216 36 54 88 117 22 12 7 76 1649 Fremington, Yelland, Instow 3 487 141 207 83 117 28 49 12 24 55 26 1229 Bideford East the Water 4 68 48 60 167 416 2 4 3 13 16 5 803 Bideford, Northam 5 487 155 129 259 2294 29 15 36 99 265 31 3800 South Molton 6 232 117 24 4 7 0 77 0 1 0 0 462 Braunton, Ilfracombe, Lynton 7 1322 44 28 5 13 161 28 159 30 20 229 2038 A377, Crediton, Exeter Area 8 315 97 15 3 30 0 131 0 2 0 0 594 A386, Torrington, Plymouth 9 62 74 25 9 104 1 25 1 0 36 1 336 Cornwall, Clovelly 10 61 31 10 35 357 0 3 2 34 6 5 544 Somerset, Rest of UK 11 58 138 13 2 9 0 132 0 0 0 0 353 Total 7304 1547 960 690 3697 480 1269 403 253 457 449 17509 Inter Peak Hour – Average Hour 10.00-16.00hrs (all vehicles/hour): Barnstaple, Landkey 1 1685 553 349 71 262 152 842 228 39 54 57 4293 Roundswell, Bickington, 2 590 13 37 43 208 57 105 23 42 33 29 1180 Fremington, Yelland, Instow 3 341 34 20 40 139 22 45 12 24 26 14 718 Bideford East the Water 4 48 43 18 6 317 0 2 2 8 42 1 487 Bideford, Northam 5 306 238 278 219 1058 4 10 25 75 570 15 2800 South Molton 6 186 70 31 2 8 0 44 0 0 1 0 342 Braunton, Ilfracombe, Lynton 7 836 103 44 2 8 40 16 59 7 4 61 1181 A377, Crediton, Exeter Area 8 268 21 15 2 37 0 68 0 1 2 0 413 A386, Torrington, Plymouth 9 40 35 19 2 134 0 9 1 0 5 2 245 Cornwall, Clovelly 10 40 25 21 25 367 1 6 2 1 0 2 490 Somerset, Rest of UK 11 92 40 22 0 21 0 65 0 1 2 0 244

NDevonNewModelLMVRV1 Page 25 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Total 4434 1175 854 411 2559 277 1213 351 198 739 181 12392 PM Peak Hour - 17.00-18.00hrs (all vehicles/hour): Barnstaple, Landkey 1 1898 773 564 88 435 132 1370 462 53 94 44 5912 Roundswell, Bickington, 2 761 58 104 71 350 64 88 45 40 31 18 1630 Fremington, Yelland, Instow 3 309 42 30 39 290 16 42 12 37 49 11 878 Bideford East the Water 4 33 46 38 9 401 0 1 2 18 36 2 587 Bideford, Northam 5 338 200 412 183 1050 14 13 24 113 677 13 3036 South Molton 6 209 100 35 5 16 0 144 0 0 0 0 510 Braunton, Ilfracombe, Lynton 7 870 97 26 2 4 95 6 132 11 1 140 1384 A377, Crediton, Exeter Area 8 208 39 13 2 51 0 200 0 0 2 0 515 A386, Torrington, Plymouth 9 30 18 29 13 147 0 26 0 0 9 0 274 Cornwall, Clovelly 10 29 18 83 38 325 0 13 1 6 0 3 515 Somerset, Rest of UK 11 118 97 36 5 25 0 227 0 1 11 0 520 Total 4804 1486 1371 455 3096 322 2130 678 279 909 233 15762

NDevonNewModelLMVRV1 Page 26 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

5. TRAFFIC MODEL CALIBRATION

5.1 Procedure

5.1.1 The traffic model calibration was an iterative process consisting of the estimation of essential parameters defining the SATURN traffic model, and related to the following main items:

 Highway network;  2008 base year origin-destination matrices;  Traffic assignment techniques and assumptions. 5.1.2 The accuracy of the developing model was assessed by comparing observed and assigned traffic movements on the survey cordons and screenlines by analysing the volume and distribution of traffic at each RSI site.

5.2 Highway Network

5.2.1 The SATURN network was checked to ensure important roads were included to provide the necessary level of detail within the study area, and that it was consistent with the traffic zoning system. All coded link lengths were checked to ensure that they were correct as well as the saturation flows and signal timings within the simulated network. Similarly all banned movements were reviewed and checked and that all possible movements at junctions were included. The loading points of zones onto the network were also reassessed and if need be altered to allow for sensible and accurate traffic flows.

5.2.2 SATURN plots of ‘trees’, the routes used by traffic between origins and destinations, were checked to ensure that the selected routes were logical and valid. ‘Trees’ from various locations were checked and a number of origins and destinations were selected as being representative of a large cross section of the movements occurring through and within the study area. The modelled routes in the AM and PM peak hours between the selected origins and destinations are described as follows and illustrated on Figure 10:

 Barnstaple Town Centre – Bideford Quay : Torridge Bridge, A39 and Barnstaple Western Bypass;  A361 North Devon Link – Braunton Road : A39 and Barnstaple Western Bypass;  A39 Bude – North Devon Hospital : Torridge Bridge, A39, Barnstaple Western Bypass, Fair View and North Road;  Gorwell – Roundswell : Derby Road, Alexandra Road, Longbridge, Sticklepath Hill and B3232. 5.2.3 These and other paths are consistent with known routes area are considered to be reasonable.

NDevonNewModelLMVRV1 Page 27 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Figure 10: Selected Modelled Routes

BIDEFORD QUAY TO BARNSTAPLE TOWN CENTRE – AM PEAK HOUR

A361 NORTH DEVON LINK TO BRAUNTON ROAD – AM PEAK HOUR

A39 BUDE TO NORTH DEVON HOSPITAL – AM PEAK HOUR

GORWELL TO ROUNDSWELL – AM PEAK HOUR

NDevonNewModelLMVRV1 Page 28 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

BARNSTAPLE TOWN CENTRE TO BIDEFORD QUAY – PM PEAK HOUR

BRAUNTON ROAD TO A361 NORTH DEVON LINK – PM PEAK HOUR

NORTH DEVON HOSPITAL TO A39 BUDE – PM PEAK HOUR

ROUNDSWELL TO GORWELL – PM PEAK HOUR

NDevonNewModelLMVRV1 Page 29 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

5.3 Traffic Assignment

5.3.1 Calibration of traffic assignment in the model firstly concentrated on identifying the most appropriate type of traffic assignment to use. Available options within SATURN were the Wardrop equilibrium and the stochastic user (SUE) equilibrium assignments, with also the possibility of using all or nothing and pure stochastic assignments.

5.3.2 It was considered appropriate to use an equilibrium assignment to take account of and optimise capacity restraint effects. The Wardrop equilibrium assignment seemed to be most appropriate because there is already significant congestion in the peak hours and there are limited route options within Barnstaple and Bideford.

5.3.3 The appropriateness of using the different assignment methods was assessed using the procedures recommended in the SATURN user manual. In particular, the level of congestion on the network was investigated in assessing the merits of using either the Wardrop or stochastic user methods. It was found that the ‘epsilon-2’ values of 1% in the AM peak hour and 2% in the PM peak hour were much lower than the recommended maximum of 25% that indicates the upper limit of the use of stochastic assignment. However the values were distorted by roads between Barnstaple and Bideford that are largely uncongested during the peaks and it was considered that congestion in the towns is widespread indicating the use of the Wardrop equilibrium method was.

5.3.4 The values of significant assignment parameters used were:

 MASL = 50 Maximum number of assignment/simulation loops.  NITA = 30 Maximum number of iterations within each assignment.  NITS = 15 Maximum number of iterations within each simulation. 5.3.5 Minimum time assignments were carried out to be consistent with previous modelling. It was found that the use of the traffic signal green split optimisation parameter (SIGOPT=T) produced improved correspondence and were used for all model time periods.

5.3.6 Termination of the simulation-assignment loops was set to occur when the number of loops exceeds MASL or %FLOWS exceeded ISTOP / RSTOP (1%) and %GAP was less than STPGAP (1.0) on NISTOP (4) successive loops (KONSTP=5).

5.4 Generalised Costs

5.4.1 The generalised cost parameters were calculated according to DfT guidance in WebTAG Unit A1.3 and associated Databook issued in November 2014. The 2013 values of time (pence per minute – PPM) and distance (pence per kilometre – PPK) output from the calculation spreadsheet and used in the 2013 model update are provided in Table 15 below.

NDevonNewModelLMVRV1 Page 30 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Table 15: 2013 PPM and PPK values

Time (PPM) Distance (PPK) User Class AM Inter- PM AM Inter- PM Peak peak Peak Peak peak Peak Car –Commuting 13.46 13.36 13.17 11.17 11.06 11.14 Car - Employers 54.26 53.05 52.23 15.32 15.21 15.28 Business Car - Education 16.99 17.65 18.14 11.17 11.06 11.14 Car - Other 16.99 17.65 18.14 11.17 11.06 11.14 LGV 24.20 24.20 24.20 16.56 16.54 16.55 OGV 24.87 24.87 24.87 41.14 40.58 40.98

5.5 Model Convergence

5.5.1 The convergence criteria are set out in WebTAG unit M3.1 and are defined by two measures; proximity indicators, reflecting how close the current flow and cost pattern is to the assignment objective, and stability for ensuring model convergence. The standards for each of the criteria are set out below: . Proximity measures: Delta (δ) < 0.1%; and %GAP < 0.1%.

. Stability measures: %Links with Flows changing by less than 1% > 98% (“P1”); and %Links with Costs changing by less than 1% > 98% (“P2”). 5.5.2 These criteria are appropriate for a model using an equilibrium based assignment and the stability measures are reported for all user classes.

5.5.3 The convergence statistics of the final 2014 base models, Table 16, shows that the models converged in a relatively small number of loops given their complexity in terms of geographical extent, numbers of zones and links. All the convergence criteria were met in all time periods. Delta and %GAP values were considerably less than the required 1% and there were more than four final iterations when over 98% of flows on links changed by less than 1%.

Table 16: Convergence Results – Base Year

Criteria AM IP PM

Number of loops 39 33 42 Final 99.9 / 100 100 / 100 99.4 / 99.9 % of Flows / 99.9 / 100 99.7 / 100 99.4 / 99.9 Delays < 1% Previous 99.9 / 100 99.7 / 100 99.1 / 99.9 (>98) 3 loops 98.4 / 99.9 99.5 / 100 98.9 / 99.9 % Gap (<0.1) 0.016 0.0000 0.00069 % Delta (<0.1) 0.0021 0.0000 0.0007

NDevonNewModelLMVRV1 Page 31 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

6. TRAFFIC MODEL VALIDATION

6.1 Validation Criteria & Acceptability Guidelines

6.1.1 The validation criteria for highway assignment models are set out in WebTAG unit M3.1. There are three main validation criteria:

. Assigned flows and counts totalled for each screenline or cordon, as a check on the quality of the trip matrices (see Table 17); . Asigned flows and counts on individual links and turning movements at junctions as a check on the quality of the assignment (see Table 18); and . Modelled and observed journey times along routes, as a check on the quality of the network and the assignment (see Table 19).

Table 17: Screenline Flow Validation Criteria Criteria Acceptability Guidelines Differences between modelled flows and counts All or nearly all screenlines should be less than 5% of the counts

Table 18: Link Flow and Turning Movement Validation Criteria

Criteria Description of Criteria Acceptability Guidelines 1 Individual flows within 100 veh/h of > 85% of cases counts for flows less than 700 veh/h Individual flows within 15% of counts > 85% of cases for flows from 700 to 2,700 veh/h Individual flows within 400 veh/h of > 85% of cases counts for flows more than 2,700 veh/h 2 GEH* < 5 for individual flows > 85% of cases

(푀−퐶)2 Where GEH = √ 0.5(퐶+푀) M = modelled flows; and C = observed flows.

Table 19: Journey Time Validation Criteria

Criteria Acceptability Guidelines Modelled times along routes should be within 15% > 85% of route of surveyed times (or 1 minute, if higher than 15%)

6.1.2 The aims of model validation are to:

 Compare model estimates with information not used in model calibration;  Directly measure the accuracy of the model estimates.

NDevonNewModelLMVRV1 Page 32 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

6.1.3 The following elements have been validated in detail:

1) Network : link details, link length, routes and journey times; 2) Assignment : sector movements, screenline and link counts. 6.1.4 Assignment routes for zone to zone movements have been checked to ensure that logical and known routes are used that take account of signposting, major roads, rat-runs etc.. The aim was to ensure that the assignment model gives an adequate representation of routes chosen by traffic in the North Devon area.

6.1.5 Modelled times for the 2014 base year were compared with those observed on the same routes from Trafficmaster journey time data.

6.1.6 Select link assignments have been used to identify the distribution of movements assigned through the RSI stations and these have been compared with the origin-destination distributions found from the roadside surveys and presented as sector to sector comparisons.

6.1.7 Assigned flows were compared with survey volumes at each of the RSI sites. The GEH statistic has been used to compare observed and assigned flow. GEH is based on the chi-squared parameter and takes account of both the absolute and relative difference and provides a practical measure of important differences at both low and high traffic levels. For transport models it is recommended that GEH values of less than 5 indicate a good fit between assigned and observed flows while GEH values greater than 10 require closer attention.

6.2 Network Validation

6.2.1 Following calibration of the traffic model, the finalised link characteristics of the highway network were checked to ensure valid classification for each link. The speed/flow relationships allocated to each link were confirmed as appropriate. It was confirmed that no exceptions to the standard values for speed/flow relationships were used. Junction coding was checked to confirm the correct interpretation of priorities, lane allocation, saturation flow etc..

6.2.2 Link lengths were checked by comparing coded lengths with crow-fly distances derived from the co-ordinates coded for network plotting. Co- ordinates were based on OS grid references modified for short links to assist visibility for plotting. Those links where the coded link length was greater than 1.3 times the crow-fly distance, and where the difference was greater than 100 metres, were screened and checked. Only a few such exceptional links were identified and these were located in the extremities of the modelled area and were all bendy rural roads with fairly long link lengths.

NDevonNewModelLMVRV1 Page 33 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

6.3 Journey Time Validation

6.3.1 The Trafficmaster journey times described in Section 3.3 were compared with modelled times in Table 20 to Table 22 and time / distance comparison diagrams are included in APPENDIX 1. The average modelled journey times were compared with the average observed time and 95th percentile range which is also given in the Trafficmaster data. The validation criteria of journey time modelling accuracy within 1 minute, 15% or the 95th percentile range for at least 85% of the routes was met for the AM and PM peak hours, with 90% and 100% respectively, but only 75% was achieved for the inter peak period.

Table 20: AM Peak Journey Time Validation

Within 1 Mean Route Observed Modelled %age min, 15% Road Direction 95%ile Mean (s) Time (s) Diff. or 95%ile Range (s) Range JT1 Chaddiford Lane – The Inbound 396 339-453 384 -3% Yes JT2 Square Outbound 423 359-486 394 -7% Yes JT3 ND Hospital – The Inbound 344 294-394 361 5% Yes JT4 Square Outbound 464 386-543 387 -17% Yes JT5 Landkey– Square via Inbound 415 386-472 439 6% Yes JT6 Portmore Outbound 342 316-393 342 0% Yes JT7 Landkey– Square via Inbound 446 207-982 357 -20% Yes JT8 Newport Outbound 462 274-930 359 -22% Yes JT9 Roundswell – Square via Northbound 317 296-337 325 3% Yes JT10 Lake Southbound 279 254-304 280 0% Yes JT11 Roundswell – Square via Northbound 481 423-539 415 -14% Yes JT12 Cedars Southbound 417 354-481 375 -10% Yes JT13 Abbotsham Cross – Northbound 281 266-295 264 -6% Yes JT14 Westleigh via A39 Southbound 278 255-301 250 -10% Yes JT15 Abbotsham Cross – Northbound 569 574-663 591 4% Yes JT16 Westleigh via Bideford Southbound 584 585-716 583 0% Yes JT17 Westleigh – The Square Northbound 1027 883-1170 783 -24% No JT18 via A39 Southbound 693 630-755 671 -3% Yes JT19 Westleigh – The Square Northbound 1268 1043-1493 1042 -18% No JT20 via B3233 Southbound 974 890-1058 1024 5% Yes %age of Journey Times Meeting Validation Criteria 90% .

Table 21: Inter Peak Journey Time Validation

Within 1 Mean Route Observed Modelled %age min, 15% Road Direction 95%ile Mean (s) Time (s) Diff. or 95%ile Range (s) Range JT1 Chaddiford Lane – The Inbound 403 345-462 306 -24% No JT2 Square Outbound 455 393-517 328 -28% No JT3 ND Hospital – The Inbound 370 314-426 278 -25% No JT4 Square Outbound 467 400-534 279 -40% No JT5 Landkey– Square via Inbound 381 355-433 393 3% Yes JT6 Portmore Outbound 371 348-420 297 -20% No JT7 Landkey– Square via Inbound 393 307-480 351 -11% Yes JT8 Newport Outbound 382 295-468 352 -8% Yes JT9 Roundswell – Square via Northbound 312 280-345 301 -4% Yes JT10 Lake Southbound 286 266-305 269 -6% Yes JT11 Roundswell – Square via Northbound 374 329-419 348 -7% Yes

NDevonNewModelLMVRV1 Page 34 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

JT12 Cedars Southbound 364 324-405 358 -2% Yes JT13 Abbotsham Cross – Northbound 306 251-362 272 -11% Yes JT14 Westleigh via A39 Southbound 304 250-359 251 -18% Yes JT15 Abbotsham Cross – Northbound 577 551-664 590 2% Yes JT16 Westleigh via Bideford Southbound 598 577-724 582 -3% Yes JT17 Westleigh – The Square Northbound 739 671-806 725 -2% Yes JT18 via A39 Southbound 699 661-737 660 -6% Yes JT19 Westleigh – The Square Northbound 1088 918-1259 974 -10% Yes JT20 via B3233 Southbound 1078 888-1269 1009 -6% Yes %age of Journey Times Meeting Validation Criteria 75%

Table 22: PM Peak Journey Time Validation

Within 1 Mean Route Observed Modelled %age min, 15% Road Direction 95%ile Mean (s) Time (s) Diff. or 95%ile Range (s) Range JT1 Chaddiford Lane – The Inbound 474 381-567 386 -19% Yes JT2 Square Outbound 532 391-673 391 -26% Yes JT3 ND Hospital – The Inbound 446 362-531 368 -18% Yes JT4 Square Outbound 515 416-614 359 -30% Yes JT5 Landkey– Square via Inbound 516 426-637 440 -15% Yes JT6 Portmore Outbound 357 324-414 343 -4% Yes JT7 Landkey– Square via Inbound 323 285-689 357 10% Yes JT8 Newport Outbound 397 314-600 359 -10% Yes JT9 Roundswell – Square via Northbound 271 251-291 301 11% Yes JT10 Lake Southbound 402 362-441 295 -27% Yes JT11 Roundswell – Square via Northbound 360 297-423 358 -1% Yes JT12 Cedars Southbound 439 380-499 421 -4% Yes JT13 Abbotsham Cross – Northbound 348 259-438 294 -16% Yes JT14 Westleigh via A39 Southbound 293 257-329 265 -9% Yes JT15 Abbotsham Cross – Northbound 573 538-627 591 3% Yes JT16 Westleigh via Bideford Southbound 606 583-725 587 -3% Yes JT17 Westleigh – The Square Northbound 689 643-735 739 7% Yes JT18 via A39 Southbound 833 771-895 696 -16% Yes JT19 Westleigh – The Square Northbound 1072 826-1319 983 -8% Yes JT20 via B3233 Southbound 1103 965-1241 1062 -4% Yes %age of Journey Times Meeting Validation Criteria 100%

NDevonNewModelLMVRV1 Page 35 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

6.4 Validation at A361 Portmore RSI Site

6.4.1 As matrix building did not include RSI origin-destination data validation of the process was carried out by compaing the distribution of traffic movements assigned through the A361 Portmore RSI site link corresponded to the origin- destination distribution found from the roadside survey. The A361 RSI was carried out in July 2015 in the eastbound direction between Portmore roundabout and Landkey.

6.4.2 The comparison was made using the 11 sector system described earlier, see Figure 8. Expanded survey data at each RSI station was compressed to sectors and select link analysis used to establish the modelled movements.

6.4.3 The resulting origin and destination totals for each roadside station, illustrated on Figure 11, show reasonable correspondence between survey and assignment for most movements with only a few that show significant differences.

6.4.4 The correspondence between observed and modelled volumes at the roadside interview stations for each sector movement was also analysed by the distribution GEH values calculated for each non-zero sector to sector movement. 80% of the non zero sector to sector movements at the RSI site had GEH values less than 5 in the AM peak hour, 83% in the inter peak and 81% in the PM peak hour. These results showed very good correspondence between observed and modelled sector to sector volumes at the roadside interview site.

NDevonNewModelLMVRV1 Page 36 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

Figure 11: Sector Movements at RSI Sites 1 & 2

RSI A361 PORTMORE - AM RSI A361 PORTMORE - AM

OBSERVED MODELLED OBSERVED MODELLED 350 400 300 350 250 300 250 200 200 150 150

100 100

VOLUME (PCU/HR) VOLUME (PCU/HR) 50 50 0 0 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 ORIGINS BY SECTOR DESTINATIONS BY SECTOR

RSI A361 PORTMORE - IP RSI A361 PORTMORE - IP

OBSERVED MODELLED OBSERVED MODELLED 350 300

300 250 250 200 200 150 150 100

100 VOLUME (PCU/HR) VOLUME (PCU/HR) 50 50 0 0 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 ORIGINS BY SECTOR DESTINATIONS BY SECTOR

RSI A361 PORTMORE - PM RSI A361 PORTMORE - PM

OBSERVED MODELLED OBSERVED MODELLED 450 300 400 250 350 300 200 250 150 200 150 100

100 VOLUME (PCU/HR) VOLUME (PCU/HR) 50 50 0 0 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 ORIGINS BY SECTOR DESTINATIONS BY SECTOR

NDevonNewModelLMVRV1 Page 37 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

6.5 Traffic Flow Validation

6.5.1 Total observed and modelled traffic flows were compared for a number of Barnstaple screenlines:

 Barnstaple Cordon;  East West Screenline;  Bideford Screenline.. 6.5.2 The results, Table 23 show a significant proportion of screenlines had differences between observed and modelled traffic flows greater than the 5% WebTAG criteria. However, the differences were acceptable according to the criteria when all the links making up the screenlines were considered.

Table 23: Screenline Traffic Flow Validation %age of Screenlines Criteria AM IP PM Differences between modelled flows and counts 67% 83% 50% should be less than 5% of the counts GEH < 5.0 or Acceptable Flow Differences 100% 83% 83% %age of Modelled Flows at all Criteria Screenline Sites AM IP PM GEH < 5.0 or Acceptable Flow Differences 91% 86% 91%

Comparison of traffic counts and assigned flows from the ATC and MCC data used for matrix estimation, 187 traffic flow values in the peak hours and 151 values in the inter peak, were used for link flow and turning movement validation, Table 24. It can be seen that the WebTAG validation requirements of at least 85% of values meeting the criteria was achieved in all of the modelled time periods. Full details are provided in APPENDIX 2.

Table 24: Summary of Validation of Link Flows

% of Links Meeting Criteria Criteria Description of Criteria AM IP PM 1 Individual flows within 100 veh/h of 94% 95% 91% counts for flows less than 700 veh/h Individual flows within 15% of counts 100% 100% 81% for flows from 700 to 2,700 veh/h Individual flows within 400 veh/h of - - - counts for flows more than 2,700 veh/h 2 GEH < 5 for individual flows 90% 93% 85% 1 & 2 Either Criteria 95% 96% 90%

NDevonNewModelLMVRV1 Page 38 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

7. SUMMARY & CONCLUSIONS

7.1.1 The purpose of this model update is to provide Devon County Council with a robust model for the assessment of future development and transport schemes in Barnstaple and Bideford. This report provides details of the model update process and data inputs using in the calibration and validation of the model.

7.1.2 The highway network and zone structure was refined to provide more detail in the modelling of areas with future development.

7.1.3 Traffic demand matrices for the new 2014 base year were developed using the NTEM trip end model, school trip data and previously development trip distribution models. Matrix estimation was then used to match modelled flows to traffic counts and screenlines across the model area.

7.1.4 Model performance against WebTAG / DMRB criteria is summarised in Table 25 and shows that the models are within the acceptable guidelines. The 2014 base year models are therefore considered to be suitable as a base to undertake future year option testing.

Table 25: Summary of Model Validation

% Within Criteria Criteria Description of Criteria AM IP PM Calibration / Individual flows within 100 veh/h of counts 94% 95% 91% Validation - for flows less than 700 veh/h Links Individual flows within 15% of counts for 100% 100% 81% flows from 700 to 2,700 veh/h Individual flows within 400 veh/h of counts - - - for flows more than 2,700 veh/h GEH < 5 for individual flows 90% 93% 85% GEH < 5 or Acceptable Flow Differences 95% 96% 90% Calibration / Differences between modelled flows and Validation - counts should be less than 5% of the 91% 97% 91% Screenlines counts All or nearly all screenlines GEH < 5.0 or Acceptable Flow Differences 91% 86% 91% at screenline links, >85% Journey Times Modelled times along routes should be within 15% of surveyed times (or 1 minute, 90% 75% 100% if higher than 15%) Convergence Number of loops, % of Flows / Delays < 1% (>98%) 100% 100% 100% % Gap & % Delta (<0.1%)

NDevonNewModelLMVRV1 Page 39 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – AM Peak Hour (1)

1 - Braunton Road Inbound 6 - A361 West Outbound

500 450 450 400 2099 2623 400 350 350 2192 2242 300 2584 300 2232 2572 2211 250 250 2500

2049 200 Time (s) Time Time (s) Time 200 2029 2352 1940 150 2247 150 2197 100 1873 100 501840 502099 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

2 - Braunton Road Outbound 7 - Landkey Inbound

600 600

500 500 1840 400 1873 400 2126 2283 300 1940 300 2026

Time (s) Time 2444 Time (s) Time 2042 200 2198 200 2525 2232 21972244 100 100 2099 2621 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

3 - A39 North Inbound 8 - Landkey Outbound

450 600 400 2099 500 350 2192 2621 300 2242 400 2232 250 2211 300 200 2049 2525 Time (s) Time 2444 150 2066 (s) Time 200 100 100 2283 2126 502254 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 500 1000 1500 2000 2500 3000 3500 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

4 - A39 North Outbound 9 - Roundswell via Lake Inbound

600 400 350 2099 500 2047 300 2254 1926 400 250

300 2066 200 Time (s) Time Time (s) Time 2043 150 200 2198 2079 2232 21972244 100 100 50 2099 1794 0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

5 - A361 West Inbound 10 - Roundswell via Lake Outbound

600 400 350 500 2099 1796 300 400 22422192 2346 250 2073 300 2493 200 2563 Time (s) Time 2582 150 200 (s) Time 1929 100 100 2047 50 2623 2099 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -50 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 40 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – AM Peak Hour (2)

11 -Roundswell via Cedars Inbound 16 - Abbotsham Cross - Longbridge - Westleigh WB

600 800 700 500 2099 1014 2047 600 400 1926 500 1067 300 1867 400 1212 Time (s) Time 300 Time (s) Time 200 1350 1762 1752 200 100 100 1794 1446 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 0 1000 2000 3000 4000 5000 6000 7000 8000 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

12 - Roundswell via Cedars Outbound 17 - Westleigh - Barnstaple via Lake EB

600 1400

500 1200 1796 1000 400 20472099 1764 800 1926 1766 300 2079

Time (s) Time 600 1867 1794

Time (s) Time 200 1923 400 100 2047 200 2099 1435 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 0 2000 4000 6000 8000 10000 12000 14000 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

13 - Abbotsham Cross - A39 - Westleigh EB 18 - Westleigh - Barnstaple via Lake WB

400 900 350 800 1435 700 300 1446 600 250 1233 500 200 400

150 1796 Time (s) Time Time (s) Time 1094 300 2073 100 1029 200 1929 50 2047 1010 1002099 0 0 0 1000 2000 3000 4000 5000 6000 0 2000 4000 6000 8000 10000 12000 14000 -50 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

14 - Abbotsham Cross -A39 - Westleigh WB 19 - Westleigh - Barnstaple via Cedars EB

400 1600 350 1400

300 1200 20472099 1013 1926 250 1000 1867 1029 1762

200 800 Time (s) Time Time (s) Time 1094 150 600 1589 1538 1205 100 400 1484 50 200 1446 1435 0 0 0 1000 2000 3000 4000 5000 6000 0 2000 4000 6000 8000 10000 12000 14000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

15 - Abbotsham Cross - Longbridge - Westleigh EB 20 - Westleigh - Barnstaple via Cedars WB

700 1200 1446 1435 600 1000 1484 500 800 1350 1538 400 1589 1212 600

Time (s) Time 300

Time (s) Time 400 1757 200 1067 1867 200 1923 100 20992047 1017 0 0 0 2000 4000 6000 8000 10000 12000 14000 0 1000 2000 3000 4000 5000 6000 7000 8000 -200 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 41 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – Inter Peak Hour (1)

1 - Braunton Road Inbound 6 - A361 West Outbound

500 500 450 400 400 2099 350 2192 2623 300 2242 2232 300 2211 250 2584 2572 200 2049 2029 (s) Time 200

Time (s) Time 1940 2500 150 2352 100 1873 100 2247 1840 2197 50 0 2099 0 500 1000 1500 2000 2500 3000 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

2 - Braunton Road Outbound 7 - Landkey Inbound

600 500 450 500 400 2126 350 400 2283 1840 300 1873 300 250 2444

1940 Time (s) Time Time (s) Time 200 2525 20422026 200 150 2198 2232 100 100 21972244 2099 502621 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

3 - A39 North Inbound 8 - Landkey Outbound

500 500 450 450 400 400 2621 350 350 2099 300 300 22422192 250 2232 250 2525

Time (s) Time 2211 200 200 2444 2049 (s) Time 150 150 2066 100 100 2283 502254 502126 0 0 0 500 1000 1500 2000 2500 3000 3500 -50 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

4 - A39 North Outbound 9 - Roundswell via Lake Inbound

600 400 350 500 2099 2047 300 400 1926 250 2254

300 200 Time (s) Time Time (s) Time 2066 2043 150 200 2079 2198 2232 100 100 2244 2197 50 2099 1794 0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

5 - A361 West Inbound 10 - Roundswell via Lake Outbound

500 400 450 2099 350 400 2192 2242 300 1796 350 2346 300 2493 250 2073 250 2563 200 2582

Time (s) Time 200 150 Time (s) Time 150 1929 100 100 50 2047 502623 2099 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -50 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 42 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – Inter Peak Hour (2)

11 -Roundswell via Cedars Inbound 16 - Abbotsham Cross - Longbridge - Westleigh WB

500 800 450 700 1014 400 2099 2047 600 350 1926 500 1067 300 400 250 1867 1212 200 (s) Time 300 Time (s) Time 1350 150 1762 200 100 1752 100 501794 1446 0 0 0 1000 2000 3000 4000 5000 6000 7000 8000 -50 0 500 1000 1500 2000 2500 3000 3500 4000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

12 - Roundswell via Cedars Outbound 17 - Westleigh - Barnstaple via Lake EB

450 900 400 1796 800 20472099 1926 350 700 1764 300 600 2079 1766 250 500 1794 200 400 1867 (s) Time

Time (s) Time 300 150 1923 100 200 2047 100 502099 1435 0 0 0 2000 4000 6000 8000 10000 12000 14000 -50 0 500 1000 1500 2000 2500 3000 3500 4000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

13 - Abbotsham Cross - A39 - Westleigh EB 18 - Westleigh - Barnstaple via Lake WB

400 900 350 800 1435 300 1446 700 600 250 1233 500 200 400

150 1796 Time (s) Time Time (s) Time 1094 300 2073 100 1029 200 1929 50 1010 10020992047 0 0 0 1000 2000 3000 4000 5000 6000 0 2000 4000 6000 8000 10000 12000 14000 -50 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

14 - Abbotsham Cross -A39 - Westleigh WB 19 - Westleigh - Barnstaple via Cedars EB

400 1400

350 1200 20472099 300 1000 1926 1013 1867 250 1762 1029 800 200

Time (s) Time 600 1589 Time (s) Time 1094 150 1538 400 100 1205 1484 50 200 1446 1435 0 0 0 1000 2000 3000 4000 5000 6000 0 2000 4000 6000 8000 10000 12000 14000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

15 - Abbotsham Cross - Longbridge - Westleigh EB 20 - Westleigh - Barnstaple via Cedars WB

700 1400 1446 600 1200 1435 500 1000 1350 1484 800 400 1538 1212 600 1589

Time (s) Time 300 Time (s) Time 200 1067 400 1757 1867 200 1923 100 2047 1017 2099 0 0 0 1000 2000 3000 4000 5000 6000 7000 8000 0 2000 4000 6000 8000 10000 12000 14000 -200 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 43 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – PM Peak Hour (1)

1 - Braunton Road Inbound 6 - A361 West Outbound

600 450 400 500 2623 350 2099 400 300 2584 22422192 2572 2232 250 300 2500 2211 200 20292049 (s) Time 2352 Time (s) Time 200 1940 150 2247 2197 100 100 1873 1840 502099 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

2 - Braunton Road Outbound 7 - Landkey Inbound

700 450 400 600 2126 350 2283 500 1840 300 1873 400 250 2444

1940 200 Time (s) Time

Time (s) Time 300 2525 20422026 2198 150 200 2232 21972244 100 100 2099 502621 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

3 - A39 North Inbound 8 - Landkey Outbound

600 500

500 400 2621 2099 400 300 2192 2242 2525 300 2232 200 2444

Time (s) Time 2211

200 2049 (s) Time 2066 100 2283 100 2126 2254 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 500 1000 1500 2000 2500 3000 3500 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

4 - A39 North Outbound 9 - Roundswell via Lake Inbound

700 350 2099 2047 600 300 1926 500 250

400 2254 200

Time (s) Time 150

Time (s) Time 300 2066 2043 2079 2198 200 2232 100 21972244 100 50 2099 1794 0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

5 - A361 West Inbound 10 - Roundswell via Lake Outbound

700 500

600 400 500 2099 1796 2192 300 400 2242 2346 2073 2493 200

Time (s) Time 300 2563 2582 (s) Time 1929 200 100 2047 100 2099 2623 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 44 October 2015

LOCAL MODEL NORTH DEVON TRANSPORT MODEL VALIDATION REPORT

APPENDIX 1 – Time / Distance Diagrams – PM Peak Hour (2)

11 -Roundswell via Cedars Inbound 16 - Abbotsham Cross - Longbridge - Westleigh WB 450 800 400 2099 700 2047 1014 350 1926 600 300 500 1067 250 1867 400 1212

200 (s) Time 300

Time (s) Time 1350 150 1762 200 100 1752 100 501794 1446 0 0 0 1000 2000 3000 4000 5000 6000 7000 8000 -50 0 500 1000 1500 2000 2500 3000 3500 4000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

12 - Roundswell via Cedars Outbound 17 - Westleigh - Barnstaple via Lake EB

600 900 800 20472099 500 1926 1796 700 400 1764 600 2079 1766 500 1794 300

1867 400 Time (s) Time

Time (s) Time 200 300 1923 200 100 2047 2099 1001435 0 0 0 500 1000 1500 2000 2500 3000 3500 4000 0 2000 4000 6000 8000 10000 12000 14000 -100 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

13 - Abbotsham Cross - A39 - Westleigh EB 18 - Westleigh - Barnstaple via Lake WB

450 1200 400 1000 350 1446 300 800 1435 250 1233 600 200

Time (s) Time 400 1796 Time (s) Time 150 1094 2073 100 1029 200 1929 20992047 50 1010 0 0 0 2000 4000 6000 8000 10000 12000 14000 -50 0 1000 2000 3000 4000 5000 6000 -200 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

14 - Abbotsham Cross -A39 - Westleigh WB 19 - Westleigh - Barnstaple via Cedars EB

400 1400 350 1200 2099 19262047 300 1013 1000 1867 250 1762 1029 800 200 1094

Time (s) Time 600 1589 Time (s) Time 150 1538 1205 400 100 1484 50 200 1446 1435 0 0 0 1000 2000 3000 4000 5000 6000 0 2000 4000 6000 8000 10000 12000 14000 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

15 - Abbotsham Cross - Longbridge - Westleigh EB 20 - Westleigh - Barnstaple via Cedars WB

700 1400 1446 600 1200 1435 1000 500 1484 1350 400 800 1538 1212 1589 600

Time (s) Time 300 Time (s) Time 400 1757 200 1067 1867 200 1923 100 2047 1017 2099 0 0 0 1000 2000 3000 4000 5000 6000 7000 8000 0 2000 4000 6000 8000 10000 12000 14000 -200 Distance (m) Distance (m)

Modelled Observed DMRB Range Modelled Observed DMRB Range

NDevonNewModelLMVRV1 Page 45 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

APPENDIX 2 - Traffic Flow Validation – Traffic Count Sites

Location / Count Model GEH Count Location Direction AM IP PM AM IP PM AM IP PM A361 E of Portmore WB 853 605 806 812 515 685 1.4 3.8 4.4 A361 E of Portmore EB 789 627 774 744 583 690 1.6 1.8 3.1 A377 Bish. Tawton NB 448 252 275 496 292 281 2.2 2.4 0.4 A377 Bishp Tawton SB 256 248 447 288 264 488 1.9 1.0 1.9 Rumsam Bridge EB 1108 543 1086 1133 771 1063 0.7 8.9 0.7 Rumsam Bridge WB 1191 611 1200 1218 766 1181 0.8 5.9 0.5 Barn. Longbridge EB 847 645 648 795 662 526 1.8 0.7 5.0 Barn. Longbridge WB 538 649 731 505 667 611 1.5 0.7 4.6 Barn. W Bypass N NB 1173 984 1084 1166 987 1016 0.2 0.1 2.1 Barn. W Bypass N SB 999 1003 1203 963 1013 1239 1.2 0.3 1.0 Ashford A361 EB 878 671 651 871 670 624 0.2 0.0 1.1 Ashford A361 WB 588 691 985 587 692 950 0.0 0.0 1.1 A39 Shirwell Cross SB 481 293 290 410 235 281 3.4 3.6 0.6 A39 Shirwell Cross NB 253 282 454 208 241 394 3.0 2.5 2.9 B3233 Bickington EB 419 452 395 542 458 388 5.6 0.3 0.3 B3233 Bickington WB 303 471 710 302 474 730 0.1 0.1 0.7 A39 Westleigh EB 838 686 680 890 682 626 1.8 0.2 2.1 A39 Westleigh WB 572 693 1096 564 676 1078 0.3 0.7 0.5 A39 Abbotsham X EB 331 320 310 332 325 301 0.0 0.3 0.5 A39 Abbotsham X WB 216 317 414 304 537 596 5.5 10.6 8.1 Bideford A386 S NB 408 353 445 406 361 416 0.1 0.4 1.4 Bideford A386 S SB 356 325 467 362 326 450 0.3 0.0 0.8 Northam Rd Rbt SB 822 - 730 838 900 733 0.6 - 0.1 Northam Rd Rbt NB 526 - 1104 518 961 1061 0.3 - 1.3 A39 E – Rbt WB 559 - 1225 605 701 1206 1.9 - 0.6

NDevonNewModelLMVRV1 Page 46 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

A39 E – Rbt EB 727 - 764 779 806 692 1.9 - 2.7 Heywood Rd – Rbt NB 615 - 915 591 903 851 1.0 - 2.1 Heywood Rd – Rbt SB 772 - 873 790 660 858 0.6 - 0.5 A39 W – Rbt EB 456 - 552 497 563 556 1.9 - 0.2 A39 W - Rbt WB 427 - 681 445 640 659 0.8 - 0.9 Westaway - A39 N SB 485 251 304 557 353 401 3.2 5.8 5.2 Westaway - A39 N NB 253 252 452 318 332 537 3.9 4.7 3.8 Westaway – Hosp. WB 84 236 391 84 235 384 0.0 0.1 0.4 Westaway – Hosp. EB 502 206 116 508 204 116 0.3 0.1 0.0 Westaway - A39 S NB 610 433 555 368 449 539 10.9 0.8 0.7 Westaway - A39 S SB 549 462 619 589 492 617 1.7 1.4 0.1 Westaway Plain EB 193 85 105 457 85 114 14.6 0.0 0.9 Westaway Plain WB 68 85 168 52 93 167 2.1 0.9 0.1 Cedars - Barnstaple WB 408 698 990 400 695 907 0.4 0.1 2.7 Cedars - Barnstaple EB 897 670 660 885 694 622 0.4 0.9 1.5 Cedars - Roundwell NB 671 612 765 675 645 768 0.2 1.3 0.1 Cedars - Roundwell SB 399 642 798 431 631 712 1.6 0.5 3.1 Cedars - Bickington EB 535 466 462 542 458 388 0.3 0.4 3.6 Cedars - Bickington WB 304 462 748 302 474 730 0.1 0.6 0.7 Sainsburys - Cedars SB 435 641 788 431 631 712 0.2 0.4 2.8 Sainsburys - Cedars NB 680 619 789 675 645 768 0.2 1.1 0.7 Old Bideford Rd E WB 100 45 87 100 47 56 0.0 0.3 3.7 Old Bideford Rd E EB 29 46 109 29 47 80 0.0 0.2 2.9 Sains - Roundswell NB 433 616 904 430 638 890 0.1 0.9 0.5 Sains- Roundswell SB 694 622 680 690 605 608 0.2 0.7 2.8 Old Bideford Rd W EB 532 168 176 529 166 173 0.1 0.2 0.2 Old Bideford Rd W WB 97 183 377 96 183 374 0.1 0.0 0.2 Rounds – A3125 SB 583 617 786 596 598 819 0.5 0.8 1.2 Rounds – A3125 NB 672 545 565 622 584 590 2.0 1.6 1.0 Roundswell - A39 E WB 837 799 1244 956 836 1204 4.0 1.3 1.1

NDevonNewModelLMVRV1 Page 47 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Roundswell - A39 E EB 1377 795 810 1449 854 919 1.9 2.0 3.7 Roundswell – Torr. NB 370 184 207 374 191 199 0.2 0.5 0.6 Roundswell – Torr. SB 186 184 285 182 193 261 0.3 0.7 1.5 Roundswell - A39W EB 921 565 579 890 682 626 1.0 4.7 1.9 Roundswell - A39W WB 476 641 1156 564 676 1078 3.8 1.4 2.3 Lake - Squareabout SB 731 590 886 714 593 898 0.6 0.1 0.4 Lake - Squareabout NB 1312 605 717 1291 606 731 0.6 0.0 0.5 Lake – Rumsam WB 1203 798 1187 1218 766 1181 0.4 1.1 0.2 Lake – Rumsam EB 1066 823 1086 1133 771 1063 2.0 1.8 0.7 Lake – Roundswell EB 1403 849 917 1449 854 919 1.2 0.2 0.1 Lake – Roundswell WB 959 809 1187 956 836 1204 0.1 0.9 0.5 Rumsam - Newport SB 536 334 520 570 349 503 1.4 0.8 0.7 Rumsam - Newport NB 596 352 341 603 367 359 0.3 0.8 1.0 Rumsam - Portmore WB 723 521 789 651 468 775 2.8 2.4 0.5 Rumsam - Portmore EB 736 508 585 741 483 595 0.2 1.1 0.4 Rumsam- B Tawton NB 478 248 281 496 292 281 0.8 2.7 0.0 Rumsam- B Tawton SB 287 239 513 288 264 488 0.1 1.6 1.1 Rumsam – Lake EB 1099 788 1035 1133 771 1063 1.0 0.6 0.9 Rumsam – Lake WB 1217 792 1186 1218 766 1181 0.0 0.9 0.1 Sticklepath Hill Old Torr. Rd 458 135 181 451 135 152 0.3 0.0 2.3 Bickington Road 407 645 945 393 644 828 0.7 0.0 3.9 Old Torrington Rd Bickington Road 80 95 98 80 98 97 0.0 0.3 0.1 Sticklepath Hill 243 156 170 245 158 153 0.1 0.1 1.3 Bickington Road Sticklepath Hill 827 625 601 736 644 554 3.3 0.7 2.0 Old Torr. Rd 154 86 94 157 88 92 0.2 0.2 0.2 Pilton Causeway SB 588 441 532 598 456 525 0.4 0.7 0.3 Pilton Causeway NB 564 406 468 563 393 434 0.1 0.7 1.6 Alexandra Rd WB 711 730 646 722 717 617 0.4 0.5 1.2 Alexandra Rd EB 732 681 838 741 665 725 0.3 0.6 4.0 North Walk NB 436 436 550 438 411 433 0.1 1.2 5.3

NDevonNewModelLMVRV1 Page 48 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

North Walk SB 439 520 422 454 525 416 0.7 0.2 0.3 Alexandra Rd SB 923 900 1061 912 862 933 0.4 1.3 4.0 Alexandra Rd NB 790 769 705 788 740 658 0.1 1.1 1.8 Belle Meadow Rd E WB 881 805 763 892 801 625 0.4 0.2 5.2 Belle Meadow Rd E EB 759 832 1018 749 810 829 0.4 0.8 6.2 Belle Meadow Rd W EB 478 612 627 457 625 526 1.0 0.5 4.2 Belle Meadow Rd W WB 733 716 728 725 738 597 0.3 0.8 5.1 Eastern Ave. E WB 632 570 675 691 546 548 2.3 1.0 5.1 Eastern Ave. E EB 663 637 741 641 625 596 0.9 0.5 5.6 Hollowtree Rd NB 435 509 434 510 503 377 3.5 0.3 2.8 Hollowtree Rd SB 499 552 585 622 538 521 5.2 0.6 2.7 Eastern Ave. W EB 894 873 1003 929 862 846 1.2 0.4 5.2 Eastern Ave. W WB 798 764 792 868 748 655 2.4 0.6 5.1 Riverside Road Braunton Road 29 40 48 25 22 23 0.7 3.2 4.2 Western Bypass 141 144 321 104 188 308 3.3 3.4 0.8 Braunton Road Western Bypass 875 731 982 858 825 931 0.6 3.4 1.6 Western Bypass Riverside Road 261 138 89 220 234 108 2.7 7.0 2.0 Braunton Road 905 723 922 947 754 907 1.4 1.1 0.5 Manteo Way NB 169 68 222 168 69 242 0.1 0.1 1.3 SB 174 75 256 176 76 107 0.2 0.1 11.0 Torridge Hill New Rd N 162 178 204 159 120 251 0.2 4.8 3.1 New Rd S 130 109 173 128 102 133 0.2 0.7 3.2 New Rd N New Rd S 279 269 450 279 261 378 0.0 0.5 3.5 Torridge Hill 121 119 106 90 91 99 3.0 2.7 0.7 New Rd S Torridge Hill 166 101 122 148 100 107 1.5 0.1 1.4 New Rd N 339 308 343 336 312 326 0.2 0.2 0.9 Old Town Meddon St 46 41 53 60 40 87 1.9 0.1 4.1 Clovelly Rd 130 166 192 131 168 129 0.1 0.2 5.0 Meddon St Clovelly Rd 131 201 210 124 147 156 0.7 4.1 4.0 Old Town 51 48 54 114 44 50 6.9 0.6 0.6

NDevonNewModelLMVRV1 Page 49 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Clovelly Rd Old Town 193 156 163 193 132 46 0.0 2.0 11.5 Meddon St 230 204 256 227 182 297 0.2 1.6 2.5 Belvoir Rd Abbotsham Rd E 13 10 8 3 5 7 3.4 1.8 0.3 Abbotsham Rd W 14 10 10 14 5 10 0.0 1.8 0.0 Abbotsham Rd E Abbotsham Rd W 242 157 216 246 161 176 0.2 0.3 2.8 Belvoir Rd 12 6 13 7 5 5 1.4 0.4 2.8 Abbotsham Rd W Belvoir Rd 11 13 16 8 13 16 0.8 0.1 0.0 Abbotsham Rd E 181 170 195 183 171 192 0.2 0.1 0.2 Buckleigh Rd A39 N 62 60 40 39 61 41 3.2 0.1 0.1 A39 S 36 50 41 58 28 42 3.2 3.6 0.1 A39 N A39 S 396 515 567 432 574 570 1.8 2.5 0.1 Buckleigh Rd 43 65 90 13 65 89 5.7 0.1 0.1 A39 S Buckleigh Rd 35 57 98 38 56 107 0.4 0.2 0.9 A39 N 474 481 549 458 502 515 0.7 1.0 1.5 The Quay Longbridge 166 167 245 166 168 287 0.0 0.1 2.6 New Rd 196 194 262 210 209 255 1.0 1.1 0.4 Longbridge New Rd 167 169 198 146 121 182 1.7 4.0 1.2 The Quay 354 291 354 325 295 332 1.6 0.2 1.2 New Rd The Quay 289 281 262 274 286 325 0.9 0.3 3.7 Longbridge 229 161 268 221 146 252 0.5 1.2 1.0 Bridge St The Quay 65 138 107 66 93 99 0.2 4.2 0.8 Longbridge 77 58 83 77 60 82 0.0 0.2 0.1 New Rd 5 15 12 13 23 39 2.6 1.8 5.4 Heywood Rd Rbt Abbotsham X Rbt 352 423 457 354 476 457 0.1 2.5 0.0 The Big Sheep 131 68 100 136 126 155 0.4 5.9 4.8 Abbotsham X Rbt The Big Sheep 51 32 24 53 31 25 0.3 0.1 0.1 Heywood Rd Rbt 414 400 493 351 422 496 3.2 1.1 0.1 The Big Sheep Heywood Rd Rbt 144 60 99 145 135 126 0.1 7.6 2.6 Abbotsham X Rbt 13 11 16 15 33 34 0.6 4.7 3.5 Abbotsham EB 45 68 84 48 68 80 0.5 0.0 0.5

NDevonNewModelLMVRV1 Page 50 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

A39 Bideford SB 360 477 486 369 509 491 0.5 1.5 0.2 Clovelly Road WB 302 360 461 316 373 446 0.8 0.7 0.7 A39 Bude NB 315 322 277 332 325 301 0.9 0.2 1.4 Abbotsham WB 61 58 69 61 9 13 0.0 8.4 8.8 A39 Bideford NB 398 471 539 404 454 520 0.3 0.8 0.8 A39 Bude SB 190 322 414 304 537 596 7.3 10.4 8.1 Clovelly Road EB 345 357 191 296 276 188 2.7 4.5 0.2 Town Centre Atlantic Village 13 74 18 12 61 12 0.2 1.5 1.6 Abbotsham Cross 233 205 309 148 193 297 6.2 0.9 0.7 Atlantic Village Abbotsham Cross 6 103 80 168 181 149 17.4 6.5 6.4 Town Centre 4 72 48 3 53 37 0.3 2.4 1.7 Abbotsham Cross Town Centre 326 227 165 182 131 117 9.0 7.2 4.0 Atlantic Village 19 130 26 114 145 71 11.7 1.3 6.5 Squareabout Sticklepath EB 1124 - 860 1101 854 796 0.7 - 2.2 Squareabout Sticklepath WB 899 - 1268 886 871 1138 0.4 - 3.7 Squareabout Taw Bridge SB 982 - 1435 963 1013 1239 0.6 - 5.3 Squareabout Taw Bridge NB 1193 - 1027 1166 987 1016 0.8 - 0.4 Squareabout - Town WB 509 - 1030 484 609 913 1.1 - 3.8 Squareabout - Town EB 1121 - 746 1073 631 628 1.5 - 4.5 Squareabout A361 Lake NB 1331 - 736 1291 606 731 1.1 - 0.2 Squareabout A361 Lake SB 733 - 1020 714 593 898 0.7 - 4.0 Portmore - A361 E WB 771 597 706 812 515 685 1.5 3.5 0.8 Portmore - A361 E EB 781 617 822 744 583 690 1.3 1.4 4.8 Portmore - A361 W NB 693 536 605 741 483 595 1.8 2.3 0.4 Portmore - A361 W SB 604 590 711 651 468 775 1.9 5.3 2.4 Portmore – East. Av EB 589 597 730 558 563 725 1.3 1.4 0.2 Portmore – East. Av WB 668 523 508 716 509 540 1.8 0.6 1.4 Westleigh - Bideford A386 158 258 275 83 272 239 6.8 0.9 2.2 Westleigh - Bideford Barnstaple 707 492 472 696 535 452 0.4 1.9 0.9 Westleigh - A386 Barnstaple 201 136 160 194 147 174 0.5 0.9 1.1

NDevonNewModelLMVRV1 Page 51 October 2015

NORTH DEVON TRANSPORT MODEL LOCAL MODEL VALIDATION REPORT

Westleigh - A386 Bideford 134 107 119 167 128 325 2.7 1.9 13.8 Westleigh – Barn. Bideford 410 522 928 438 573 880 1.4 2.2 1.6 Westleigh – Barn. A386 133 116 213 125 103 198 0.7 1.3 1.1 Hollowtree Road Landkey Road 137 - 165 140 154 136 0.2 - 2.4 Hollowtree Road South Street 123 - 264 249 277 251 9.3 - 0.8 Hollowtree Road Newport Road 33 - 75 1 0 0 7.9 - 12.2 Landkey Road South Street 89 - 109 0 0 0 13.3 - 14.8 Landkey Road Newport Road 146 - 129 62 7 40 8.3 - 9.7 Landkey Road Hollowtree Road 145 - 106 107 186 81 3.4 - 2.6 South Street Newport Road 99 - 104 137 26 99 3.5 - 0.5 South Street Hollowtree Road 207 - 152 340 219 154 8.0 - 0.2 South Street Landkey Road 76 - 49 0 0 0 12.3 - 9.9 Newport Road Hollowtree Road 31 - 30 0 0 0 7.7 - 7.7 Newport Road Landkey Road 77 - 96 77 35 72 0.0 - 2.6 Newport Road South Street 90 - 98 86 26 85 0.4 - 1.4 Tesco Whiddon Dr. SB 527 - 488 523 428 476 0.2 - 0.5 Tesco Whiddon Dr. NB 505 - 448 501 443 405 0.2 - 2.1 Tesco Eastern Av. E WB 674 - 580 716 509 540 1.6 - 1.7 Tesco Eastern Av. E EB 576 - 743 558 563 725 0.8 - 0.7 Tesco - Rose Lane NB 233 - 447 186 120 305 3.3 - 7.3 Tesco - Rose Lane SB 321 - 420 317 129 239 0.2 - 10.0 Tesco Eastern Av. W EB 665 - 806 641 625 596 0.9 - 7.9 Tesco Eastern Av W WB 697 - 710 691 546 548 0.2 - 6.5 All Sites GEH<5 90% 93% 85%

5

GEH>10 3% 1% 3%

Acceptability of flow differences 95% 96% 88%

NDevonNewModelLMVRV1 Page 52 October 2015