dmrb volume 15 - sias.comsias.com/2013/ts/the nesa manual - october 2015.pdf · urban roads...

October 2015 The NESA Manual

DESIGN MANUAL FOR ROADS AND BRIDGES

VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND

SECTION 1 THE NESA MANUAL

CONTENTS

Contents

Chapter

1. List of Contents

2. List of Tables

3. List of Figures

4. Program Availability

5. Enquiries

PART 1

Volume 15 Section 1Part 1 Contents

The NESA Manual October 2015

Volume 15 Section 1 Chapter 1Part 1 Contents List of Contents

October 2015 The NESA Manual 1-1-1

1 LIST OF CONTENTS

PART1 CONTENTS

Chapter

1. List of Contents2. List of Tables3. List of Figures4. Program Availability5. Enquiries

PART2 INTRODUCTION TO NESA

Chapter

1. Introduction2. NESA Model Structure3. Applicability of NESA4. Structure of the Manual5. Acknowledgements

PART3 ECONOMIC CONCEPTS IN NESA

Chapter

1. Cost-Benefit Analysis2. The Do-Minimum and the Do-Something Options3. Discounting, the Price Basis and Economic Indicators4. Evaluation Over 60 Years5. The Fixed Trip Matrix6. The Treatment of Taxation

PART4 THE APPLICATION OF NESA

Chapter

1. Economic Decision Criteria: Viability of Schemes2. Appraisal Under Uncertainty3. Incremental Analysis: The Choice of Route and Standards4. The Appraisal of Competing and Complementary Schemes5. Effects of Delaying Construction6. The Appraisal of Schemes with Different Opening Years

Chapter 1 Volume 15 Section 1List of Contents Part 1 Contents

1-1-2 The NESA Manual October 2015

PART5 TRAFFIC MODELLING IN NESA

Chapter

1. Principles of Traffic Modelling2. Definition of Terms3. Road Network4. Trip Matrices5. Route Choice6. Traffic Forecasts

PART6 VALUATION OF COSTS AND BENEFITS

Chapter

1. Introduction2. The Valuation of Time Savings3. The Valuation of Vehicle Operating Costs4. The Valuation of Accidents5. The Valuation of Accidents on Links6. The Valuation of Accidents at Junctions7. Carbon Emissions8. Construction Costs9. The Preparation of Scheme Cost Data for Use in NESA10. An Example of Scheme Cost Inputs11. Road Maintenance12. Delays During Construction

PART7 SPEEDS ON LINKS

Chapter

1. The NESA Speed/Flow Types2. Rural Single Carriageways (Speed/Flow Type 1)3. Rural All-Purpose Dual Carriageways and Motorways (Speed/Flow Types 2-6)4. Urban Roads (Speed/Flow Types 7 and 8)5. Suburban Roads (Speed/Flow Types 9 and 10)6. Small Town Roads (Speed/Flow Type 11)7. Single Track Roads (Speed/Flow Type 12)8. Treatment of Overcapacity on Links9. Representative Diagrams of Speed/Flow Relationships

Volume 15 Section 1 Chapter 1Part 1 Contents List of Contents


PART8 JUNCTIONS IN NESA

Chapter

1. When to Model Junctions2. Junction Choice3. Junction Types Modelled4. The Concept of Maximum Delay5. Geometric Delay6. Queuing Delay7. Formulae for Junction Capacity8. Geometric Parameters

PART9 VALIDATING A NESA ASSESSMENT

Chapter

1. NESA Traffic Model Calibration and Validation2. Goodness of Fit Statistics3. Economic Assessment Validation4. A Summary of the Items of Cost and Benefit in NESA5. The Timing and Documentation for NESA Validations

PART10 HOW TO USE THE NESA PROGRAM

Chapter

1. Introduction2. Using NESA3. Building the Base Network4. Amending the Base Network5. Building the Trip Matrices6. Building the Base Network Trees7. Assignment8. Evaluating the Base Network Model9. Modelling Junctions for Delay10. The Modelling of Accidents11. Assessment of Alternative Design Schemes12. Building the Design Network13. Building the Design Network Trees14. Reassignment of Traffic15. Evaluating the Design Network16. NESA Print Command Options17. Amalgamation of Stages in NESA18. Accident only assessments19. Example Input Files and Output Tables20. NESA File Formats21. NESA Error Messages

Chapter 1 Volume 15 Section 1List of Contents Part 1 Contents


PART11 INDEX AND ABBREVIATIONS

Chapter

1. Index2. Abbreviations

Volume 15 Section 1 Chapter 2Part 1 Contents List of Tables


2 LIST OF TABLES


Table 4/3/1: Example of Scheme Option RankingTable 4/3/2: Example of Incremental AnalysisTable 4/4/1: Strategy CombinationsTable 4/4/2: Appraisal of Local Authority and Developer Schemes


Table 5/2/1: Network Classification DefinitionsTable 5/2/2: Seasonality Indices by Network ClassificationTable 5/2/3: Default E-Factors by Network ClassificationTable 5/2/4: Default a, b Coefficients and Typical M Factors by Month of CountTable 5/2/5: Trip Matrix User ClassesTable 5/2/6: Default User Class Proportions by Network ClassificationTable 5/2/7: Default Vehicle Categories by Network ClassificationTable 5/2/8: Default Hours in each Flow GroupTable 5/3/1: NESA Road Categories, Link Speeds and Link CapacitiesTable 5/3/2: [Contd] NESA Road Categories, Link Speeds and Link CapacitiesTable 5/6/1: National Road Traffic Forecasts (NRTF 1997) - Annual Percentage Growth Rates


Table 6/2/1: Values of Time per Person and per Vehicle in NESA (2010 prices and values)Table 6/2/2: Compound Annual Rates of Change in Car Occupancies (%)Table 6/2/3: Assumed Compound Annual Rates of Growth in the Real Value of Time (%)Table 6/3/1: VOC Formulae Parameter Values (2010 prices and values)Table 6/3/2: Fuel Resource Costs, Fuel Duty and VAT Rates (2010 prices and values)Table 6/3/3: Proportion of Cars and LGV’s Using Petrol or diesel by Vehicle Kms (%)Table 6/3/4: Change in Vehicle Efficiency (%PA)Table 6/4/1: Components of Accident Costs (2010 prices and values)Table 6/4/2: Proportions of Fatal, Serious and Slight Accidents on Links (Average 1999- 2001)Table 6/4/3: Proportions of Fatal, Serious and Slight Accidents at Junctions (Average 1999 - 2001)Table 6/4/4: Assumed Compound Annual Rates of Growth in Accident Values (%)Table 6/5/1: Default Link Only Accident Rates (2000 Base)Table 6/5/2: Default Link and Junction Combined Accident Rates (2000 Base)Table 6/5/3: Average Number of Casualties per Accident (2000 Base)Table 6/5/4: Casualties per Accident Change Coefficients bTable 6/6/1: Junction Accident Parameters (2000 Base)Table 6/6/2: Observed Ranges of Flow in NESA Junction Accident Model CalibrationTable 6/6/3: Accident and Casualties per Accident Change Coefficient b for JunctionsTable 6/7/1: Carbon Emissions per Litre of Fuel Burnt (gCarbon/l)Table 6/7/2: Non Traded Values of Carbon £ per Tonne of Carbon (2002 prices and values)

Chapter 2 Volume 15 Section 1List of Tables Part 1 Contents


Table 6/8/1: The Treatment of Land and Property CostsTable 6/8/2: Optimum Bias Adjustment Factors Generally Applied for Different Stages of Assessment

of Trunk Road SchemesTable 6/9/1: CPI Values from 2010 to 2014Table 6/9/2: Default Profile for Construction Costs (Excluding Land Costs)Table 6/10/1: Example of Scheme Cost Profile (£ thousands in 2010 prices)Table 6/11/1: Non-Traffic Related Maintenance Costs (2010 values and prices)Table 6/11/2: Estimated Maintenance Works Cost Profiles (average 2010 prices)


Table 7/1/1: NESA Speed/Flow TypesTable 7/1/2: NESA Road Categories and Speed/Flow TypesTable 7/1/3: [Contd] NESA Road Categories, Link Speeds and Link CapacitiesTable 7/1/4: Link Geometric Properties Used in Speed/Flow RelationshipsTable 7/1/5: Default Link Geometric Variables used in Speed/Flow relationships by Road CategoryTable 7/2/1: Definition of Variables Used in Speed Prediction Formulae for Rural Single CarriagewaysTable 7/3/1: Definition of Variables Used in Speed Prediction Formulae for Rural All-Purpose Dual

Carriageways and MotorwaysTable 7/4/1: Definition of Variables Used in Speed Prediction Formulae for Urban RoadsTable 7/5/1: Definition of Variables Used in Speed Prediction Formulae for Suburban RoadsTable 7/6/1: Definition of Variables Used in Speed Prediction Formulae for Small Town RoadsTable 7/7/1: Definition of Variables Used in Speed Prediction Formulae for Single Track RoadsTable 7/9/1: Values of Variables Used in Representative Speed/Flow Relationships - Rural RoadsTable 7/9/2: Values of Variables Used in Representative Speed/Flow Relationships - Urban RoadsTable 7/9/3: Values of Variables Used in Representative Speed/Flow Relationships - Suburban RoadsTable 7/9/4: Values of Variables Used in Representative Speed/Flow Relationships - Small Town

Roads


Table 8/5/1: Geometric Delays at Major/Minor Priority Junctions

PART9 VALIDATING A NESA ASSESSMENT

Table 9/2/1: Assignment Calibration/Validation: Acceptability GuidelinesTable 9/4/1: Conversion of Travel Costs to Market Prices by Vehicle Category (Table 14 of the NESA

output)Table 9/4/2 : The Economic Efficiency of the Road System in Market Prices (Table 15A of the NESA

output)Table 9/4/3 : Public Accounts in Market Prices (Table 15B of the NESA output)Table 9/4/4 : Analysis of Monetised Costs and Benefits in Market Prices (Table 15C of the NESA

output)


Volume 15 Section 1 Chapter 2Part 1 Contents List of Tables


Table 10/3/1: Mandatory Link VariablesTable 10/3/2: Optional Link VariablesTable 10/3/3: [Contd] Optional Link VariablesTable 10/5/1: Base Network MATRIX_OPTIONS ParametersTable 10/15/1: Scheme Cost Data InputTable 10/15/2: Alternative Scheme Cost Data InputTable 10/21/1: NESA Error MessagesTable 10/21/2: [Contd]: NESA Error MessagesTable 10/21/3: [Contd]: NESA Error MessagesTable 10/21/4: [Contd]: NESA Error MessagesTable 10/21/5: [Contd]: NESA Error MessagesTable 10/21/6: [Contd]: NESA Error Messages

Chapter 2 Volume 15 Section 1List of Tables Part 1 Contents


Volume 15 Section 1 Chapter 3Part 1 Contents List of Figures


3 LIST OF FIGURES

PART2 INTRODUCTION TO NESA

Figure 2/1/1: The Components of a Road Scheme AppraisalFigure 2/2/1: NESA Outline StructureFigure 2/2/2: NESA Accident Only Outline Structure

PART3 ECONOMIC CONCEPTS IN NESA

Figure 3/1/1: The NESA Economic Appraisal MethodFigure 3/1/2: Process for Calculating Network User Costs


Figure 4/4/1: Testing the StrategyFigure 4/4/2: Exclusion Analysis for Scheme AFigure 4/4/3: Exclusion Analysis for Scheme CFigure 4/4/4: Exclusion Analysis for Scheme DFigure 4/4/5: Isolation Analysis for Scheme AFigure 4/4/6: Isolation Analysis for Scheme CFigure 4/4/7: Isolation Analysis for Scheme DFigure 4/6/1: Selection of a Common Terminal Year


Figure 5/1/1: Outline of the NESA Traffic Modelling ProcessFigure 5/2/1: Defining the Seasonality IndexFigure 5/2/2: The Variation of M-Factor with Seasonality IndexFigure 5/2/3: NESA Vehicle CategoriesFigure 5/2/4: The Relationship between Vehicle Categories and User ClassesFigure 5/2/5: Flow Groups Representing Annual FlowFigure 5/4/1: Day Type 1 & 3 Daily Profiles for Three User ClassesFigure 5/4/2: Annual Profiles for Three User ClassesFigure 5/4/3: An Example of an Adjusted ProfileFigure 5/4/4: Example Total Weekly Traffic Flow Profile (January)Figure 5/5/1: Node Expansion


Figure 6/1/1: Calculation of User Costs by Flow Group and Forecast Year from Assigned Link FlowsFigure 6/6/1: Rural 4-Arm Major/Minor Junction

Chapter 3 Volume 15 Section 1List of Figures Part 1 Contents



Figure 7/1/1: Measurement of Road Geometry on Rural RoadsFigure 7/9/1: Typical Rural Speed/Flow Relationships - Vehicles/Hour/LaneFigure 7/9/2: Typical Rural Speed/Flow Relationships - Vehicles/Hour/DirectionFigure 7/9/3: Typical Urban Speed/Flow RelationshipsFigure 7/9/4: Typical Suburban Speed/Flow RelationshipsFigure 7/9/5: Typical Small Town Speed/Flow RelationshipsFigure 7/9/6: Single Track Roads Speed/Flow Relationships


Figure 8/6/1: High and Low Definition Peak ModelsFigure 8/6/2: Queuing Delay CurvesFigure 8/8/1: Major Road WidthFigure 8/8/2: Lane Width for Non-Priority StreamsFigure 8/8/3: Measurement of Visibility Distances VL and VRFigure 8/8/4: Geometric Parameters of RoundaboutsFigure 8/8/5: Entry Angle on Straight Weaving SectionsFigure 8/8/6: Entry Angle on Curved Weaving SectionsFigure 8/8/7: Entry Angle on Short Weaving AreasFigure 8/8/8: Flare Lengths for RoundaboutsFigure 8/8/9: Geometric Parameters for Signalised Roundabouts


Figure 10/1/1: The NESA Procedure SummarisedFigure 10/1/2: Example Base NetworkFigure 10/1/3: Example Design NetworkFigure 10/9/1: Priority JunctionFigure 10/9/2: Carriageway Lane DefinitionFigure 10/9/3: Signalised IntersectionFigure 10/9/4: Signalised Intersection DimensionsFigure 10/9/5: Signalised T - JunctionFigure 10/9/6: Motorway Merge

Volume 15 Section 1 Chapter 4Part 1 Contents Program Availability


4 PROGRAM AVAILABILITY4.1 A copy of the NESA program can be obtained on CD-ROM from Transport Scotland:

Technical Analysis BranchMajor Transport Infrastructure Projects DirectorateTransport ScotlandBuchanan House58 Port Dundas RoadGLASGOWG4 0HF

Email: [email protected]

4.2 Please note that the NESA Manual forms Volume 15, Section 1 of the Design Manual for Roads andBridges. The latest version of the NESA Manual is available online in pdf form from TransportScotland:

http://www.transportscotland.gov.uk/about-scot-tag

4.3 NESA will handle network sizes of up to 2000 links, 800 nodes and 120 zones and is supplied for PCsystems only. There is currently no charge for NESA.

4.4 In addition to supplying the program, the Technical Analysis Branch (TAB) at Transport Scotland alsooffer a user support service. Problems in using the program should be reported to Transport Scotland.NESA support is available by emailing [email protected].

Chapter 4 Volume 15 Section 1Program Availability Part 1 Contents


Volume 15 Section 1 Chapter 5Part 1 Contents Enquiries


5 ENQUIRIESApproval of this document for publication is given by the undersigned:

All technical enquiries or comments about this publication should be sent in writing to the above.

Chief EngineerTransport ScotlandMajor Transport Infrastructure ProjectsBuchanan House58 Port Dundas StreetGLASGOWG4 0HF A McLaughlin

Chief Engineer

Chapter 5 Volume 15 Section 1Enquiries Part 1 Contents




VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


INTRODUCTION TO NESA

Contents

Chapter

1. Introduction

2. NESA Model Structure

3. Applicability of NESA

4. Structure of the Manual

5. Acknowledgements

PART 2

Volume 15 Section 1Part 2 Introduction to NESA


Volume 15 Section 1 Chapter 1Part 2 Introduction to NESA Introduction


1 INTRODUCTION1.1 NESA (Network Evaluation from Surveys and Assignment) is a computer program developed and

maintained by Transport Scotland, Major Transport Infrastructure Projects (MTRIPS) for the trafficand economic appraisal of road schemes. NESA is fully Windows compatible and does not depend onany third party libraries or applications such as DBOS. Please consult with the overseeing authority(i.e. Transport Scotland) before commencing any traffic and economic assessment using NESA.

1.2 Figure 2/1/1 illustrates the main components of a road scheme appraisal. The appraisal hinges on theoutcome of operational, economic and environmental assessments. An important input to each of theseassessments are the changes in traffic conditions that result from the road improvement and the growthof travel demand through time. Integrating the economic assessment with a traffic model, NESAprovides an easily used and powerful aid to scheme appraisal.

1.3 An economic assessment determines a road improvement’s value for money, through an objectivecomparison of the quantifiable costs of the improved and base situations. NESA not only performs thistest, but also outputs economic indicators that can be used to rank alternative schemes (for example,whether a bypass to town X gives greater benefit than a bypass to town Y), or to indicate which ofseveral mutually exclusive scheme options (for example, alternative routes for a bypass) is to bepreferred in quantifiable economic terms.

1.4 The program, however, is not simply an aid to economic assessment. It also includes the facility tomodel traffic movements on an existing road network and predict future flows on alternative networks.These flows provide useful intermediate outputs that can assist in the operational and environmentalassessment of proposed schemes. The program can be run as a single process leading directly toeconomic evaluation, or in steps permitting the user to examine the output at each stage. This flexibilityallows specific aspects of the networks, for example junctions, to be examined and modified prior tothe economic evaluation stage being undertaken.

NESA15

1.5 NESA15 was released in October 2015 and supersedes NESA14. NESA15 differs from the previousrelease, NESA14, in the following ways:

• Changes to parameters in TAG databook version 1.3b (November 2014) for:

- Value of Time growth percentages

Figure 2/1/1: The Components of a Road Scheme Appraisal

Chapter 1 Volume 15 Section 1Introduction Part 2 Introduction to NESA


- Fuel prices

- Fuel consumption

- Forecast fuel efficiency

- Emission factors

- Carbon values

- Accident/casualty costs

- Annual rates of growth of accident values

- Accident proportions

• Correction to beta factors for junction casualties

• Inclusion of additional options to enable user input of accident data to be defined for a specific year

• Current year default now 2015

NESA14

1.6 NESA14 was released in June 2014 and supersedes NESA11. NESA14 differs from the previousrelease, NESA11, in the following ways:

• Changes to Values of Time and Vehicle Operating Costs in line with the TAG data book (previously TAG Unit 3.5.6), January 2014

• Changes to Vehicle Emissions Costs in line with the TAG data book (previously TAG Unit 3.3.5), January 2014

• Inclusion of Electric Cars in the calculation of Vehicle Operating Costs and Vehicle Emissions Costs

• Changes to Accident Costs in line with the TAG data book (previously TAG Unit 3.4.1), January 2014

• Inclusion of additional option to print link flows by user class

• Current year assumed to be 2014 but can be user defined

• New error check on the length of input file path (max=100 characters)

NESA11

1.7 NESA11 was released in March 2013 and superseded NESA10. NESA11 differed from the previousrelease, NESA10, in the following ways:

• NESA11 includes the calculation and valuation of carbon emissions based on TAG Unit 3.3.5, April 2011

• Values of Time have been updated to accord with TAG Unit 3.5.6, April 2011

• Vehicle Operating Costs have been updated to accord with TAG Unit 3.5.6, April 2011

• Fuel Vehicle Operating Costs are now derived by fuel usage and fuel efficiency improvements (petrol and diesel)

• Non-fuel Vehicle Operating Costs have been disaggregated by work and non-work vehicles

• Evaluation calculations are undertaken every year instead of every 5 years

Volume 15 Section 1 Chapter 1Part 2 Introduction to NESA Introduction


• Future year matrices can be input for any year instead of every 5 years

• Print statements involving YEAR= can be for any year rather than multiples of 5 years

• Non-work outputs have been disaggregated by commuting and other

• The Current Year is assumed to be 2011 - previously 2010.

• The Current Year can be user defined.

NESA10

1.8 NESA10 was released in March 2011 and superseded NESA05. NESA10 differed from the previousrelease, NESA05, in the following ways:

• NESA10 had the ability to run accident only assessments

• Users could decide whether a junction being assessed / modelled for accidents was a 'Major' or 'Minor' junction - see paragraph 6.11 and Table 6/6/5 on page 6-6-7 for further details.

• The Current Year is assumed to be 2010 - previously 2005.

1.9 The main economic parameter values within NESA10 are consistent with NESA05.

NESA05

1.10 NESA05 was released in July 2005 and superseded NESA03. NESA05 differed from the previousrelease, NESA03, in the following ways:

• NESA05 was fully compatible with TAG Unit 3.5.6, Values of Time and Vehicle Operating Costs (December 2004)

• The accident rates and costs in NESA05 were updated to a 2000 Base and presented in 2002 prices and values, based on the Highways Economics Note No. 1 (2002)

• NESA05 calculated net scheme benefits for each year of operation of an improvement over a 60 year appraisal period (previously 30 years)

• Table 15, Summary of Expenditure and Benefits in Market Prices, was split into three separate tables, Tables 15A, B and C

1.11 NESA05 continued to present all Transport Economic Efficiency (TEE) results in market prices. Thetraffic growth forecasts in NESA05 remained consistent with the 1997 National Road Traffic Forecasts(NRTF(1997)).

NESA03

1.12 NESA03 was released in January 2003 and superseded NESA02. The only significant differencebetween NESA03 and NESA02 was that NESA03 incorporated a 3.5% discount rate whereas NESA02incorporated a 6% discount rate. Both NESA03 and NESA02 were fully compatible with the TransportEconomics Note (TEN, March 2001) and presented all Transport Economic Efficiency (TEE) results inmarket prices. The traffic growth forecasts in NESA03 were consistent with the 1997 National RoadTraffic Forecast (NRTF(1997)).

NESA02

1.13 NESA02 was released in June 2002 and superseded NESA98. NESA02 differed from NESA98 in thatit was fully compatible with the Transport Economics Note (TEN, March 2001). Therefore the

Chapter 1 Volume 15 Section 1Introduction Part 2 Introduction to NESA


appraisal process within NESA02 (onwards) moved from a method based on overall social costs andbenefits to one based on willingness to pay. At the same time, the unit of account changed from factorcost prices (prices excluding all indirect taxation) to market prices. The willingness to pay approach isintended to show how different groups are affected by a proposal, where previously the aggregation ofresource costs and benefits masked transfer payments between different groups. The willingness to payapproach is intended to show how different groups are affected by a proposal, where previously theaggregation of resource costs and benefits masked transfer payments between different groups. Thesechanges do not affect the principles of appraisal contained within the NESA program, merely thepresentation of the results. Indeed, NESA02 onwards continues to operate in the factor cost unit ofaccount but reflects the new methodology by presenting the results in market prices and in terms ofhow a scheme will impact on different groups i.e. users, operators and government. NESA02incorporated a 6% discount rate an a present value year of 1998. The traffic growth forecasts inNESA02 were consistent with the 1997 National Road Traffic Forecasts (NRTF(1997)).

NESA98

1.14 NESA98 was released in February 1998 and superseded NESA97. NESA98 differed from NESA97 inthat it incorporated traffic and economic growth forecasts consistent with the 1997 National RoadTraffic Forecasts (NRTF(1997)).

NESA97

1.15 NESA97, released in May 1997, superseded NESA90. The traffic modelling component of NESA97differed subtly from its predecessor, NESA90, whilst the economic evaluation component differedsignificantly from its predecessor. By embracing a flow group methodology to represent the variationsin travel demand throughout a year, NESA97 had many similarities with the Department forTransport’s economic evaluation program, COBA. NESA97 also included an update of the price basefrom 1988 to 1994, a change in the discount rate from 8% to 6%, as well as relatively major revisionsto the junction delay methodology and the link speed/flow curves. In addition there were minorchanges in the accident cost calculations and the vehicle operating cost formula.

The NESA Manual

1.16 With each significant update to the NESA program updates are made to the NESA manual (DMRBVolume. 15). These updates are normally supplied to users on the release CD and paper copies areissued with the normal rounds of DMRB updates.

Future DMRB Volume 15 Updates

1.17 Whilst the NESA manual will remain part of the DMRB series as DMRB Volume 15, given the leadtimes required to issue paper copies of any updates, electronic updates (as PDFs) will be available fordownload at:


1.18 Users are advised always to ensure they have the latest documentation by checking for updates at:


Volume 15 Section 1 Chapter 2Part 2 Introduction to NESA NESA Model Structure


2 NESA MODEL STRUCTUREOverview

2.1 Figure 2/2/1 illustrates the structure of a standard NESA run. In common with earlier versions, the newversion of NESA consists of two sub-models run sequentially; a traffic assignment model whichallocates vehicle trips to a network, and an economic assessment model which calculates the user costs,benefits and economic return of a road improvement scheme.

2.2 Figure 2/2/2 illustrates the structure of an accident only NESA run. There is no traffic assignmentelement as part of an accident only assessment and the economic assessment only calculates theaccident costs and benefits of a road user scheme. More details are provided in Part 6, Chapter 4 andPart 10, Chapter 18.

Economic Assessment

2.3 NESA’s primary use is in the economic assessment of a road scheme. The assessment takes the form ofa cost-benefit analysis, in which the sum of the benefits generated is compared to the sum of theadditional costs imposed. If the benefits exceed the costs, then implementing the road improvementwould be advantageous to society overall; conversely, if the benefits amount to less than the costs,society would remain better off by not undertaking it. The economic assessment is viewed as providinga value for money test for the road improvement.

2.4 A NESA economic assessment assumes that the same pattern of travel demand occurs in both the baseand the improved situations, that is the trip matrix is fixed between the two scenarios (Fixed TripMatrix). This assumption implies that the only significant response to a road improvement is there-routeing of traffic. All other responses such as re-timing, re-distribution and modal switching areassumed to have negligible effects on the economic assessment. (Further advice regarding theapplicability of the fixed trip matrix assumption is given in Part 3 Chapter 5.)

2.5 NESA converts all costs and benefits to monetary units so that their summation and comparison cantake place. This monetary valuation process is based on the willingness to pay principle and placespractical limitations on the types of cost and benefit that can be incorporated within the economicassessment. NESA currently considers the following cost types:

• capital costs of the road improvement

• maintenance costs (non-traffic and traffic related1)

• travel time costs (link and junction delays)

• vehicle operating costs

• accident costs

• indirect tax revenues

2.6 Since July 2005 Scottish trunk road schemes are appraised over a 60 year period following theiropening. To allow for the fact that the capital cost of a road improvement is usually incurred at thebeginning of this period, whilst benefits accrue steadily throughout, a discounting technique isemployed. This technique prevents excessive weight being placed on benefits received towards the end

1.Traffic related maintenance costs (e.g. cost of works and the delays imposed by roadworks) wouldnormally be estimated using the Highways Agency computer program, QUADRO and manuallyincorporated into the NESA results. Further information on QUADRO can be found in DMRB Vol. 14.

Chapter 2 Volume 15 Section 1NESA Model Structure Part 2 Introduction to NESA


of the evaluation period. For economic assessments, benefits are estimated as they accrue to society asa whole. This implies that as far as possible all measurements should be in terms of resource costs, withtransfer payments such as taxes and subsidies eliminated (see Part 3).The costs detailed in Paragraph2.4 are clearly not the only costs or benefits imposed by or accruing to a road improvement. Othersinclude those broadly defined as environmental (e.g. noise, severance and pollution). The exclusion ofthese costs implies that economic assessment, as embodied within NESA, is only a partial technique.Consequently, NESA (and economic assessment in general) must be seen as only one element in theappraisal process, to be used along with assessments of environmental and other considerations. (Usersare directed to the Scottish Transport Appraisal Guidance (STAG) for further information regarding allthe factors which should be taken into account within the assessment of a trunk road scheme. Part 3Chapter 1 of this manual gives a brief introduction to STAG.)

2.7 Cost-benefit analysis techniques provide a methodology to produce economic indicators such as Firstand Single Year Rates of Return (FYRR and SYRR) and a Net Present Value (NPV) for the roadimprovement (see Part 3). These techniques calculate user costs and benefits by flow group from traveldemands, link speeds and junction delays (see Part 6).

2.8 Flow groups are used to represent average travel conditions for sub-groups of the 8760 hours in a year(e.g. the peak 500 hours). The flow groups are derived using link flows output from the assignmentmodel and user defined parameters that describe how travel demand varies throughout a year (see Part5).

2.9 Link speeds and junction delays are calculated for travel conditions represented by each flow groupusing speed/flow relationships (see Part 7) and junction delay models (see Part 8) respectively.

Traffic Modelling

2.10 Incorporated within NESA is a traffic modelling component. Amongst its primary objectives are to:

(i) identify current and future stress points, such as problem junctions or heavilytrafficked links, and assess the extent to which potential solutions alleviate theseproblems

(ii) provide output that can be used in the operational and environmental assessments

(iii) provide input to the economic assessment

By integrating the economic assessment with a traffic model NESA ensures consistency between thetwo processes and minimises any error that may arise through incompatibility.

2.11 Traffic modelling is traditionally split into two stages, that of assessing the travel demand and that ofallocating the demand to the road network, commonly known as assignment (see DMRB Volume 12).It is the latter (traffic assignment) modelling stage that is incorporated in NESA. As can be seen fromFigure 2/2/1 the two main inputs to NESA’s traffic assignment model are a road network descriptionand a trip matrix or trip matrices (by trip purpose (user class) and/or flow group).

2.12 The assignment technique employed is based on Burrell’s multi-routeing method. Average vehicle linkspeeds can be randomly varied to simulate how different drivers perceive a route. Because each linkspeed is fixed, that is it does not vary with traffic volume, in certain situations this assignment methodcan have some limitations (e.g. forecasting future year scenarios in a heavily congested urban area).However, in the majority of trunk road appraisals, multi-route assignments are perfectly suitableproducing a good spread of traffic throughout the network. The traffic modelling component of NESAis described in greater detail in Part 5 of this manual.

Volume 15 Section 1 Chapter 2Part 2 Introduction to NESA NESA Model Structure


Figure 2/2/1: NESA Outline Structure

Chapter 2 Volume 15 Section 1NESA Model Structure Part 2 Introduction to NESA


Figure 2/2/2: NESA Accident Only Outline Structure

2.13 The integrated nature of NESA means that modelling a design (improved) situation requires fewerinput parameters than modelling the base situation. This occurs firstly, because NESA allows the userto specify a design network by updating the base and secondly because the trip matrix specified withthe base network is also used for the traffic assignment to the design network. The latter propertyensures that the Fixed Trip Matrix assumption in the economic assessment is not violated.

2.14 Given that traffic predictions rely upon assumptions about future conditions and about the behaviour ofpeople, the output from traffic models can never be precise and should never be presented as such.Moreover, it must be appreciated that traffic flows alone cannot justify an investment. Schemes mustgenerally be justified against all five of the Scottish Government’s objectives, namely: economy;environment; integration; safety and accessibility (ref. Scottish Transport Appraisal Guidance).

Volume 15 Section 1 Chapter 3Part 2 Introduction to NESA Applicability of NESA


3 APPLICABILITY OF NESAEconomic Assessment

3.1 It is mandatory to undertake a formal economic assessment for all trunk road schemes costing over £1million (see DMRB 5.1.4 SH1/97). A formal economic assessment may also be required for someschemes in Scotland costing under £1 million. Although NESA is the standard Transport Scotland toolfor the economic assessment of road schemes, alternative programs maybe required in certaincircumstances. When other assessment techniques are used, it is essential that they should beequivalent to NESA in the sense that they conform to the same underlying economic principles and usethe same economic parameters.

3.2 Apart from indicating whether a scheme is socially desirable NESA also has several other applications,for which its use is recommended:

• the assessment of the need for a corridor improvement

• the selection of the preferred scheme out of several mutually-exclusive options

• the selection of optimal link design standards

• the selection of optimal junction types

• the optimal timing of the scheme

• the ranking of schemes in the Roads Programme

• the assessment of accidents only

3.3 There are four situations in which an alternative economic assessment is either allowed orrecommended. These are:

(a) schemes costing less than £1m where a less sophisticated approach may beappropriate

(b) where the network is operating at or is expected to operate close to capacity (e.g. in acongested urban area)

(c) where traveller responsiveness to changes in travel times or costs is high, as mayoccur where trips are suppressed by congestion and then released when the network isimproved (e.g. in a congested urban area)

(d) where the implementation of a scheme causes large changes in travel costs (e.g. a newestuarial crossing)

3.4 If any one of conditions (b), (c) or (d) applies then a Variable Trip Matrix (VTM) evaluation should beconsidered (see DMRB Volumes 12 and 13) This type of assessment differs from the Fixed Trip Matrixevaluation (see Paragraph 2.3) because the pattern of travel demand is assumed to differ between thebase and the improved situations.

3.5 If one or more of conditions (b), (c) or (d) are partially fulfilled then the sensitivity of a Fixed TripMatrix evaluation should be tested against a Variable Trip Matrix scenario in line with the proceduresset out in DMRB Volume 12. These practices are consistent with the recommendations made by theStanding Advisory Committee on Trunk Road Assessment (SACTRA) in 1994. While the terms ofreference of SACTRA do not extend to Scotland, Transport Scotland considers its findings to begenerally applicable.

Chapter 3 Volume 15 Section 1Applicability of NESA Part 2 Introduction to NESA


3.6 Conditions (b), (c) and (d) are most likely to be found in congested urban areas or with large scaleschemes such as estuarial crossings and greenfield motorways.

3.7 It should be noted that NESA cannot undertake a Variable Trip Matrix assessment. If a VTMassessment is required the user should use the Department for Transport’s TUBA program.Alternatively, a fully compatible economic evaluation program may be used where this has beenapproved and agreed by Transport Scotland.

3.8 To ensure consistency between appraisals throughout Scotland the undertaking of and the reporting ofthe economic assessment are subject to a number of design standards:

• Traffic and Economic Evaluation of Trunk Road Schemes, DMRB 5.1.4 SH1/97

• Scheme Assessment Reporting, DMRB 5.1.2 TD 37/93

• Choice between Options for Trunk Road Schemes, DMRB 5.1.2 TA 30/82

Traffic Modelling

3.9 The traffic assignment component of NESA is suitable for the majority of situations that necessitate amodel (see DMRB 5.1.4 SH1/97). However, the fact that the NESA assignment technique uses fixedlink speeds, means that care should be taken when interpreting the output in situations where trafficconditions are forecast to be substantially congested (e.g. future year scenarios in an urban area). Insuch a situation, NESA may, for example, assign traffic flows to links in excess of their capacities. Anassignment methodology incorporating some form of capacity restraint may be more suitable in suchan area.All the traffic modelling standards and advice notes from the Department for Transport areapplicable in Scotland. The user is referred to three manuals in particular:

• DMRB Volume 12 - Traffic Appraisal of Road Schemes

• DMRB Volume 5 - Assessment and Preparation of Road Schemes

• DMRB Volume 6 - Road Geometry

3.10 Volume 12 concentrates on the technical issues of traffic modelling and contains the Traffic AppraisalManual, Traffic Appraisal in Urban Areas and Induced Traffic Appraisal; whilst Volumes 5 and 6describe the traffic appraisal procedures and geometric design standards, respectively.

3.11 In addition, users are referred to the Scottish Transport Appraisal Guidance (STAG) for further adviceand information on traffic modelling.

Volume 15 Section 1 Chapter 4Part 2 Introduction to NESA Structure of the Manual


4 STRUCTURE OF THE MANUAL4.1 The manual is divided into ten parts, each part being further subdivided into chapters. It has been

written in such a manner that the most technical aspects are contained in the later parts. For example,NESA’s operating instructions are contained in the final part.

4.2 Parts 3 and 4 introduce the user to the economic principles underlying a NESA economic assessmentand detail the different ways in which a NESA assessment can be used for a scheme appraisal.

4.3 Parts 5, 6, 7, 8 and 9 describe the individual components of the methodology that allow NESA to takeas input a trip matrix and a road network and calculate a road scheme’s economic return. Part 5describes the traffic model, Part 6 the calculation of each of the user costs and Parts 7 and 8 thecalculation of link travel times and junction delays respectively.

4.4 Part 10 details the precise network coding instructions required to run the program.

Chapter 4 Volume 15 Section 1Structure of the Manual Part 2 Introduction to NESA


Volume 15 Section 1 Chapter 5Part 2 Introduction to NESA Acknowledgements


5 ACKNOWLEDGEMENTS5.1 The preparation of this manual has included the collation of material from a number of sources. The

primary source is the COBA manual (DMRB Volume 13) prepared and maintained by the HighwaysAgency. Transport Scotland’s Chief Road Engineer gratefully acknowledges these sources.

Chapter 5 Volume 15 Section 1Acknowledgements Part 2 Introduction to NESA




VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


ECONOMIC CONCEPTS IN NESA

Contents

Chapter

1. Cost-Benefit Analysis

2. The Do-Minimum and the Do-SomethingOptions

3. Discounting, the Price Basis and EconomicIndicators

4. Evaluation Over 60 Years

5. The Fixed Trip Matrix

6. The Treatment of Taxation

PART 3

Volume 15 Section 1Part 3 Economic Concepts in NESA


Volume 15 Section 1 Chapter 1Part 3 Economic Concepts in NESA Cost-Benefit Analysis


1 COST-BENEFIT ANALYSIS1.1 At the heart of a NESA economic assessment is a technique called cost-benefit analysis. Cost-benefit

analysis is designed to measure the net social benefit of a scheme by comparing the sum of the benefitsgenerated to the sum of the costs incurred. If benefits exceed costs, implementing the roadimprovement would be advantageous for society overall; conversely, if the benefits amount to less thanthe costs, society would remain better off by not undertaking the improvement.

1.2 The justification for this is that, if total benefits exceed total costs, the beneficiaries (the gainers) of theproject would still feel better off even after fully compensating those (the losers) who suffer the costs.Thus no-one would feel worse off than prior to the project, and some would feel better off. Cost-benefitanalysis, however, does not require that this compensation actually be carried out, so the gainers ingeneral need not pay anything for the benefits they enjoy, and perhaps more importantly, the losers ingeneral are not compensated for the costs imposed upon them.

1.3 Cost-benefit analysis does not fully distinguish between benefits, disbenefits and money costsexperienced by different segments of the population, for example road users, local residents and centralgovernment. Consequently, for a scheme in which the benefits outweigh the costs, there may be someindividuals, or groups, who receive the majority of the benefits, and others who bear the majority of thedisbenefits or money costs. The distribution of disbenefits on the local population may be such that, nomatter what the outcome of the economic evaluation, the scheme is unacceptable on social grounds.For example, a road improvement benefiting high income car owners commuting from a city’s suburbsmay prove unacceptable on social grounds, because the scheme imposes disbenefits on low income,low car ownership residents neighbouring the scheme.

The Need for Cost-Benefit Analysis

1.4 The use of cost-benefit analysis in the transport sector stems primarily from the fact that allgovernments are concerned to secure value for money from their investment expenditure. This occursbecause resources, particularly public sector investment resources, are scarce.

1.5 Cost-benefit analysis techniques were developed precisely for markets with no marketable output (e.g.the roads sector). The lack of a marketable output in the roads sector occurs because though charges arelevied for road use, through taxation and tolls, it is only through some tolling regimes that charges areallowed to vary by road type or quality of service. In such a situation, it is impractical to use the moretraditional financial profitability measures.

1.6 Cost-benefit analysis provides a methodology that is transparent to the public, thus helping to fosterconfidence in the appraisal process. It can also be used to direct scarce public funds towards highyielding schemes, achieving value for money from any public investment. In addition to this,cost-benefit analysis provides a consistent methodology that allows Transport Scotland to objectivelymeasure the value for money of an individual scheme and compare its value for money to otherschemes within the roads programme.

The Scottish Transport Appraisal Guidance (STAG)

1.7 As well as trying to ensure that an improvement scheme provides value for money there are otherfactors which need to be taken into account. The Scottish Transport Appraisal Guidance (STAG)outlines the Scottish Government’s five objectives which new proposals, including road improvementschemes, need to be tested against:

• environment

• safety

Chapter 1 Volume 15 Section 1Cost-Benefit Analysis Part 3 Economic Concepts in NESA


• economy

• integration

• accessibility

1.8 Cost-benefit analysis is tested against the Economy objective, but is only one part of the overallappraisal which needs to be carried out. Users are recommended to familiarise themselves with theoverall integrated appraisal methodology outlined in STAG. (The Scottish Transport AppraisalGuidance document is available via http://www.transportscotland.gov.uk/analysis/scottish-transport-analysis-guide/STAG.)

The NESA Economic Appraisal

1.9 The NESA program was originally developed to compare the costs and benefits of inter-urban roadimprovements. It has been extensively developed since its introduction and is applicable to a variety ofroad investments. In order to appraise road projects consistently with other modes, the appraisalprocess within NESA (NESA02 onwards) moved from a method based on overall social costs andbenefits to one based on willingness to pay. At the same time, the unit of account changed from factorcost prices (prices excluding all indirect taxation) to market prices. The willingness to pay approach isintended to show how different groups are affected by a proposal, where previously the aggregation ofresource costs and benefits masked transfer payments between different groups. Users are directed toTAG Unit 3.5.4, Cost Benefit Analysis (January 2014), for a full discussion on the methodology.

1.10 The above changes do not affect the principles of appraisal contained within the NESA program,merely the presentation of the results. Indeed, NESA still operates in the factor cost unit of account butreflects the new methodology by presenting the results in market prices and in terms of how a schemewill impact on different groups i.e. users, operators and government.

1.11 NESA is principally concerned with estimating the effect of a road network improvement on the usersof the road system, namely changes in time, accident and vehicle operating costs. These are balancedagainst the construction and maintenance costs which are borne by central or local government, or theprivate sector. However, the impact of a network improvement may be felt more widely in thecommunity than NESA suggests. For example, accident costs will also be distributed over pedestrians,the emergency services and friends and relatives of accident victims. Environmental (noise, airpollution etc.) and other impacts (accessibility, integration etc.), although not included within theNESA framework, must be taken into account and assessed separately (see STAG).

TRAFFIC FLOWS

1.12 Traffic flows with and without the road scheme under evaluation are generally obtained from NESA’straffic model, however, it is possible to interface NESA’s economic component with other trafficmodels. The essence of a NESA economic appraisal is that the travel cost for each component (link andjunction) of the network is calculated separately according to the flows and turning movementsassigned to it (a link based evaluation). These individual link and junction costs (that is, time, vehicleoperating costs and accidents) are summed to yield the total costs over the network.

THE FIXED TRIP MATRIX

1.13 In a NESA assignment, the matrix of trips is assumed to be fixed, (see Part 3 Chapter 5) and acomparison is made between total costs before and after the improvement in question. The differencein costs is taken as a measure of the scheme benefits as they arise in each year of the evaluation period.This estimate rests on the assumption that the improvement in question does not affect the number oftrips made nor their origins and destinations. Benefits are calculated for all traffic on the whole roadnetwork affected, including the traffic using the new road.



USER COSTS

1.14 Figures 3/1/1 and 3/1/2 illustrate the process used by NESA to arrive at estimates of network user costsand hence scheme benefits.

1.15 User costs are measured in units of time costs, vehicle operating costs, and numbers of accidents.NESA applies monetary values to each of these units of measurements. Thus, time costs are measuredin £/hour. Money values for a given time saving are based on what people would, on average, bewilling to pay for this saving (see Part 6).

1.16 Vehicle operating costs may be either perceived or unperceived by the user. Perceived costs are takeninto account by the user in his/her decision whether, and on which route, to undertake a journey.Unperceived costs are not taken into account in this decision.

1.17 The total costs borne by the road user are the sum of perceived and unperceived costs. The net effect ofa road improvement on that user is the aggregate of perceived and unperceived costs and benefits.

1.18 The NESA method measures user benefits from journey time savings and perceived vehicle operatingcost savings in terms of changes in consumer surplus. This is the difference between what the userwould be prepared to pay to make a journey, and what he/she actually pays. NESA measures userbenefits from unperceived vehicle operating cost savings in terms of the amount the user saves.

1.19 Time and vehicle operating cost elements, together with changes in accident costs represent the fullimplications of a road improvement as measured by NESA. These costs are then set alongside thecapital and maintenance costs. The output tables in NESA are structured as follows;

• Table 15A - The Economic Efficiency of the Road System in Market Prices,

• Table 15B - Public Accounts, and

• Table 15C - Analysis of Monetised Costs and Benefits in Market Prices.

The economic results in Tables 15A, 15B and 15C are presented in market prices. Table 14, Conversionof Travel Costs to Market Prices by Vehicle Category, outlines how the tax elements are added to theresource costs of the user benefits. Tables 15A, 15B & 15C summarise the expenditure and benefits (inmarket prices) in willingness-to-pay terms in a format compatible with the Transport EconomicEfficiency analysis (see STAG Chapter 8 and also TAG Unit 2.7.1). Part 9 Chapter 4 contains fulldetails of output Tables 14, 15A, 15B & 15C as well as the conversion process.

SUNK COSTS

1.20 As in any Cost Benefit Analysis, the economic assessment of road schemes is based on the assessmentof future costs and future benefits. One of the features of the progressive analysis of schemes is that theeconomic assessment is undertaken at each stage based on the return on future investments. This meansthat scheme costs incurred prior to the current economic assessment which are already spent andcannot be recovered (whether or not the scheme goes ahead) should be excluded from the overallscheme costs input to the economic assessment. Such costs are referred to as Sunk Costs.



Figure 3/1/1: The NESA Economic Appraisal Method



Figure 3/1/2: Process for Calculating Network User Costs

Volume 15 Section 1 Chapter 2Part 3 Economic Concepts in NESA The Do-Minimum and the Do-Something Options


2 THE DO-MINIMUM AND THE DO-SOMETHING OPTIONS

2.1 All NESA economic appraisals require a comparison between the base and the improved situations.The first stage in a NESA appraisal is therefore to define these alternatives. Generally the base situationis referred to as the Do-Minimum, whilst the improved situation is referred to as the Do-Something.Although there will only be one Do-Minimum situation there may well be several Do-Somethingoptions under consideration. While these terms may appear self explanatory, in fact they raise someissues of fundamental importance.

The Do-Minimum

2.2 In some cases, the definition of the Do-Minimum is straightforward; it is simply the existing networkwithout modification, that is, a Do-Nothing scenario. This corresponds to the general maxim that allexpenditure which is not sunk (firmly committed, but not spent) or past requires an economicjustification.

2.3 It should be noted that the Calibration Base network (see Part 9 Chapter 1) used in building the trafficmodel may differ from both the Do-Nothing and Do-Minimum scenarios. Also it is important to notethat even a literal Do-Nothing case is not a no change one. With traffic growth in the future,Do-Nothing user costs will increase over time reflecting increased congestion.

2.4 The Do-Minimum network can in any one of the following cases differ from the Do-Nothing:

(i) the case where works will be carried out regardless of whether or not theDo-Something scheme is built. An example is where a local authority is firmlycommitted to improving a junction in the existing network, regardless of the roadimprovement proposal. In this case, the improved (not the existing) junction should becoded in the NESA Do-Minimum and Do-Something networks. It should not becoded into the traffic model’s Calibration Base. The cost of the junction improvementmay be regarded as committed and is irrelevant for the NESA appraisal of the roadimprovement proposal; it should not therefore be included in the scheme costs;

(ii) the case where the existing network may be improved to form a Do-Minimumscheme which can be tested as an alternative to carrying out majorDo-Something improvements. An example is where an existing junction or link isforecast to become heavily overcapacity in the near future and where relatively minorimprovements can be undertaken to increase capacity, for example, by adding a laneat a traffic signal junction. This class of Do-Minimum improvement is particularlyimportant where the existing network is congested and where a literal Do-Nothingscheme would represent an unrealistically poor baseline for comparison. The NESAprintout identifies junctions and links which exceed their capacity; one useful rule ofthumb is that Do-Minimum network improvements should always be consideredwhere junctions or links are forecast to go overcapacity outside peak hours, that is,outside flow groups 4/5 (see Part 5). When such improvements are included, theDo-Minimum should include both the cost of the improvement and the changes in thenetwork compared to the Do-Nothing. One would generally expect that the benefits(or difference in user costs) comparing the Do-Minimum and Do-Nothing wouldgreatly outweigh the construction costs involved in the Do-Minimum. It isrecommended that where this type of Do-Minimum expenditure is significant, that is,20% or more of the cost of the cheapest Do-Something scheme, the Do-Minimumitself should be appraised against a Do-Nothing; provided the expenditure is

Chapter 2 Volume 15 Section 1The Do-Minimum and the Do-Something Options Part 3 Economic Concepts in NESA


justifiable in this way the improvements should be included in the Do-Minimum forcomparison with the major Do-Something improvements;

(iii) the case where traffic conditions can be improved without significant capitalexpenditure. An example is where traffic management measures can be undertakento reduce existing traffic delays. In such a situation the Do-Minimum network shouldinclude these measures. In a congested urban area it may, however, be appropriate tooptimise traffic operations in the Do-Minimum using microsimulation models orSATURN. The traffic flows output from these models can be input to NESA. Publictransport/private traffic restraint options may also be relevant, for example in someurban areas. In such cases, changes in modal split will be inconsistent with the fixedmatrix assumption, and a standard NESA assessment will not be applicable;

(iv) the case where the area covered by the NESA network includes trunk road ornon-road improvement proposals other than the one under immediateconsideration. Given the uncertainty associated with any scheme during itspreparation stage it is recommended that the sensitivity of the scheme underimmediate consideration is tested to the inclusion or exclusion of any other networkimprovements. Related to this is the question of how to evaluate a number ofinter-related road schemes, which may be either complementary or competing. This isdiscussed in Part 4 Chapter 4.

The Do-Something Scheme

2.5 The Do-Something scheme is the road improvement proposal under consideration. Usually there willbe more than one feasible Do-Something option. The number and nature of the Do-Something optionswill change as the planning of the trunk road scheme proceeds. At early stages in scheme planning, awide range of different options may be considered. At later stages, the range will be narrower butDo-Something options may be refined to highlight more detailed differences such as junction design orlink standards.

2.6 Where public transport is considered to be one of the feasible options for solving a particular roadproblem, it is important that all options, road and public transport alike, are assessed on a consistentbasis. The evaluation techniques used should match the particular circumstances. A cost benefitapproach will be most appropriate in those cases where the public transport option is intended, mainlyor exclusively, to meet particular social objectives.

Volume 15 Section 1 Chapter 3Part 3 Economic Concepts in NESA Discounting, the Price Basis and Economic Indicators


3 DISCOUNTING, THE PRICE BASIS AND ECONOMIC INDICATORS

3.1 Throughout the life cycle of a road improvement scheme, benefits and dis-benefits, some of whichhave monetary values applied to them, will accrue to road users and the local population, whilstconstruction and/or maintenance costs will be incurred by the body responsible for the road (e.g. thelocal authority, central government or the private sector). This stream of costs and stream of benefitsmust be presented in a consistent way to arrive at a view of the overall worth of the scheme. However,the total value of costs and/or benefits cannot be calculated by simply adding up their relevant streams.This is because by doing so it would assume that any capital invested or received now would haveexactly the same value (net of inflation) at any point in the future. In general, investments are expectedto bring positive rates of return, thus both consumers and the private sector value capital today morehighly than capital tomorrow. Public sector capital investments are no different and the Governmentsets the minimum acceptable return using the discount rate (see Paragraph 3.9).

Present Value (PV)

3.2 Costs and benefits arising in different years are therefore expressed in terms of their value from thestandpoint of a given year, known as the present value year. In principle any year can be taken aspresent value year but in NESA15, 2010 is used.

3.3 In NESA, prices are also expressed in a common year, which is called the price base year. This is also2010. Conversions to the price base year are normally carried out using the Consumer Price Index. Byexpressing all prices throughout the lifetime of a project at 2010 levels, the distorting effects thatinflation will have on any comparisons can be overcome.

3.4 Costs and benefits arising in different years are transformed into their present values by the process ofdiscounting. This can be understood by considering the principle of compound interest. If £1 isinvested at a real (net of inflation) interest rate of r, at the end of one year it would be worth £(1 + r) andin two years £(1 + r)2 and so on. Conversely £1 received in n years is worth £1/(1 + r)n now. Note thatthis illustration ignores the effect of inflation and assumes that £1 has the same real value in each year.

3.5 Since discounting involves the notion of charging interest against a project, rather than paying interestto an investor, r is known as the discount rate (as opposed to interest rate). Any monetary sum maytherefore be reduced to its present value (PV) by means of this formula:

3.6 For example, £100 to be received in one year from now is worth £96.62 in today’s money. This iscalculated using a discount rate of 0.035 or 3.5% (see 3.9 below) to derive a discount factor of 0.9662.

Equation 3/3/1

where PV is the present valueS is the sumr is the discount rate, expressed as a fractionn is the year in which the sum is receivedn=0 is the present value year

PV S/ 1 r+ n=

Chapter 3 Volume 15 Section 1Discounting, the Price Basis and Economic Indicators Part 3 Economic Concepts in NESA


3.7 The default present value year in NESA is 2010. The present value of a stream of benefits (PVB) iscalculated according to this formula:

3.8 The present value of the stream of costs (PVC) is calculated by a similar formula.

Discount Rate (r)

3.9 The discount rate used in NESA is 3.5% for the appraisal period 0 to 30 years and then reduces to 3.0%for years 31 to 75 (ref. Table 6.1, The Green Book). See Chapter 4 for further information regarding theappraisal period. Therefore for example, in year 32 any benefits are discounted at 3.0% for 2 years andat 3.5% for 30 years.'

Traffic Growth

3.10 Users should refer to the Overseeing Organisation’s guidance on economic assessment when selectingthe appropriate Traffic Growth. Local growth, or growth based on national forecasts, is generally used.In Scotland for trunk road schemes, Transport Scotland’s advice is to consider using TMfS as theprinciple source of growth forecasts; or alternatively Scottish Trip End Program (STEP); or NationalRoad Traffic Forecasts (NRTF97). Where NRTF97 forecasts are used, users should note that theseassume zero traffic growth post 2031.

3.11 In NESA15 (as in previous versions of NESA), the same economic growth rates are used underCentral, Low or High traffic growth scenarios.

Equation 3/3/2

where B(n) is the benefit occurring in year n

VB B 2010 B 2011 1 r+

----------------- ...B n

1 r+ n 2010–--------------------------------+ + +=

Volume 15 Section 1 Chapter 3Part 3 Economic Concepts in NESA Discounting, the Price Basis and Economic Indicators


Economic Indicators

3.12 The following four economic indicators are output by NESA and are contained in Tables 11, 12 and15A-15C of the NESA design output (see Part 10 Chapter 19). Users should note that the values givenin Tables 11 and 12 are in resource costs and are presented for information only, and that only Tables15A-15C, which summarise the expenditure and benefits, present the economic indicators in marketprices - see Part 9 Chapter 4 for further details. The application of these indicators for schemeevaluation is discussed further in Part 5.

NET PRESENT VALUE (NPV)

3.13 The Net Present Value (NPV) of a scheme in 2010 prices and values is the Present Value of Benefits(PVB) minus the Present Value of Costs (PVC).

Note: Any costs to Government entered in Table 15B, Public Accounts, are entered as positive values.

BENEFIT COST RATIO (BCR)

FIRST YEAR RATE OF RETURN (FYRR)

SINGLE YEAR RATE OF RETURN IN YEAR n (SYRRn)

Current Year Basis

3.14 The 2010 based NPV calculated by NESA can be converted manually to a current year basisrelatively simply by using the appropriate price indices. Expressing the NPV in current prices anddiscount year may make the significance of the sums involved more readily understood. For example,if the economic submission was made in August 2013, the NESA NPV could be updated to July 2013prices and values as follows:

(i) NESA NPV = £1m in average 2010 prices and values

(ii) at July 2013 prices and 2010 values

Equation 3/3/3

Equation 3/3/4

Equation 3/3/5

Equation 3/3/6

NPV PVB PVC–=

BCR PVBPVC------------=

FYRR Undiscounted Benefits in Opening YearCapital Costs Compounded to Opening Year-----------------------------------------------------------------------------------------------------------=

SYRRnUndiscounted Benefits in Year n

Capital Costs Compounded to Opening Year-----------------------------------------------------------------------------------------------------------=

NPV £1m x Consumer Price Index at July 2013 (125.8)Average Retail Price Index for 2010 (114.5)---------------------------------------------------------------------------------------------------------- £1.099m= =

Chapter 3 Volume 15 Section 1Discounting, the Price Basis and Economic Indicators Part 3 Economic Concepts in NESA


(iii) at July 2013 prices and 2013 values at a 3.5% discount rate

3.15 The Consumer Price Index (CPI) is prepared by the Office for National Statistics and is published intheir statistical periodicals, including the Monthly Digest of Statistics. CPIs can also be obtained byvisiting the Office for National Statistics website at www.ons.gov.uk.

NPV £1.099m x 1.035 2013 2010– £1.218m= =

Volume 15 Section 1 Chapter 4Part 3 Economic Concepts in NESA Evaluation Over 60 Years


4 EVALUATION OVER 60 YEARS4.1 The current version of NESA calculates net scheme benefits for each year of operation of the

improvement over a period of 30 or 60 years from the scheme opening year, whereas versions prior toNESA05 evaluated schemes over a 30 year assessment period only.

4.2 From 1 July 2005 the appraisal period used for the economic assessment of trunk road projects isnormally taken as 60 years (previously 30 years).

4.3 The 60 year appraisal has been prompted by the reduction to the discount rate in the latest version ofthe HM Treasury Green Book and the declining long term discount rate (i.e. 3.5% for years 0 to 30,reducing to 3% for years 31 to 75 etc. - ref. Table 6.1, The Green Book) and aims to cover the period ofuseful life of the asset and capture the streams of costs and benefits accrued during the life of a project.

4.4 The move to a 60 year appraisal period removes the need for the external calculation/estimation of aproject’s Residual Value (previously based on the scheme’s estimated total scheme cost). While it ispossible that, with full maintenance throughout the lifespan of a transport project, the period ofusefulness of the project may extend well beyond 60 years, and technically such projects may be seento have an infinite life, a 60 year appraisal period is considered pragmatic, given the risk anduncertainty of estimating costs and benefits any further into the future.

4.5 The standard appraisal period for trunk road projects shall therefore be taken as 60 years and alleconomic results should be presented based on a 60 year appraisal.

4.6 NESA15 also contains a set of assumptions about economic parameters, namely the values of time,vehicle operating costs and accident costs (see Part 6, Chapters 2, 3, 4, 5 and 6). The values of time andvehicle operating costs are based on the values contained within the TAG data book (previously TAGUnit 3.5.6 Values of Time and Vehicle Operating Costs), November 2014. Carbon costs in NESA15 arebased on the values contained within the TAG data book (previously TAG Unit 3.3.5 The GreenhouseGases Sub-Objective), November 2014. Accident costs in NESA15 are derived from the TAG databook (previously TAG Unit 3.4.1 The Accidents Sub-Objective), November 2014. A single set ofeconomic growth factors are applied in NESA, which are independent of the traffic growth beingapplied.

4.7 The value of each item of benefit is expressed in terms of prices in the NESA present value year, 2010.This means that the change in money value of a particular cost or benefit from one year to the next isignored to the extent that the change is attributable solely to the effects of general inflation. BecauseNESA uses constant prices, construction costs must be adjusted to the price level of the base year. Theprocedure for doing this is described in Part 6 Chapter 7.

4.8 The NESA facility for estimating a stream of benefits arising over the life of a road scheme allows amore sound basis for evaluation than is afforded by single year measures. Such measures can beparticularly deceptive since two scheme options may yield similar returns for a given year but performdifferently as traffic flows change over time.

4.9 The default traffic forecasts used within NESA are based on National Road Traffic ForecastsNRTF(97) which are not projected beyond 2031. If the default forecasts are being used NESA willassume zero traffic growth post 2031.

Chapter 4 Volume 15 Section 1Evaluation Over 60 Years Part 3 Economic Concepts in NESA


Volume 15 Section 1 Chapter 5Part 3 Economic Concepts in NESA The Fixed Trip Matrix


5 THE FIXED TRIP MATRIX 5.1 When a road improvement takes place, several changes in travel demand patterns are possible:

• reassignment:traffic travelling from A to B may use a new route

• redistribution:traffic may change its origin or destination, e.g. go to C instead of B

• generation:trips may be made when previously travel did not take place

• modal transfer:trips to the same destination maybe made by a different mode of transport. (Note: This includes changes in vehicle occupancy)

• choice of departure time:trips may be made at a different time of day.

5.2 In addition to these user responses land use changes may also occur, changing the location of orintroducing new activities and therefore also altering travel demand patterns.

5.3 A standard NESA economic assessment operates on the assumption that only the first response,reassignment, takes place. This assumption (the Fixed Trip Matrix) is, however, inappropriate in thefollowing circumstances:

• where the implementation of a scheme causes large changes in travel costs

• where traveller responsiveness to changes in travel times or costs is high (i.e. where the elasticity of demand is high)

• where the network affected by the scheme is or will be close to capacity

The above three conditions are most likely to be found in congested urban areas or with large scaleschemes such as estuarial crossings and greenfield motorways1.

5.4 The computation of user cost savings in the context of a fixed trip matrix is the central principle of astandard NESA appraisal. Taking a fixed trip matrix for the Do-Minimum and Do-Something in aninitial year, it can estimate the effects of daily and seasonal flow variations on total user costs (on a linkby link basis) through the application formulae modelling link speed (see Part 7), junction delay (seePart 8) and accident and vehicle operating costs (see Part 6). Using the appropriate traffic forecastsNESA can repeat these calculations for every year from the base over the 60 year evaluation period.This allows ready estimation of a stream of benefits which can be discounted to give a base year NetPresent Value.

5.5 The advantage of adopting a fixed trip matrix is the absence of a direct calculation of the effects oftraffic redistribution and generation, and of changes in the timing of trips (for example, from peak tooff peak) or in modal split. It should also be noted that NESA assumes that traffic growth in eachvehicle category will be similar on all links of the network, however, if non-uniform growth is expectedto have a significant effect on traffic flows it can be modelled in NESA.

1.Advice regarding the applicability of the fixed matrix assumption for any particular scheme is contained inTransport Scotland's Technical Memorandum SH1/97 (see DMRB 5.1.4). It should be noted that NESA canundertake only a Fixed Trip Matrix appraisal. Users wishing to undertake a Variable Trip Matrix (VTM)appraisal should seek advice from Transport Scotland regarding the most appropriate modelling software touse.

Chapter 5 Volume 15 Section 1The Fixed Trip Matrix Part 3 Economic Concepts in NESA


Volume 15 Section 1 Chapter 6Part 3 Economic Concepts in NESA The Treatment of Taxation


6 THE TREATMENT OF TAXATION6.1 As outlined previously in paragraph 1.9, the method of accounting used in the appraisal of transport

schemes changed in NESA02. In order to appraise projects across the modes, it was been necessary tomove from a calculus based on social costs to one based on willingness to pay. In addition, the unit ofaccount changed from factor cost prices to market prices. These changes merely affect the presentationof results and not the fundamental principles of appraisal. In particular, the results using a willingnessto pay calculus show clearly how different groups are affected by the project, whilst previouslyresource costs and benefits were aggregated and consequently masked transfer payments. A fullerdiscussion of the conversion between the factor cost and market price units of account is given in TAGUnit 3.5.4, Cost Benefit Analysis, January 2014.

6.2 The existence of indirect taxation in the economy, which involves transfers between members ofsociety (e.g. consumers to Government) rather than a cost to society at large, means that for thepurposes of running NESA an adjustment must be made to the valuation of benefits to avoidoverstatement. Although the accounting approach has changed, NESA continues to work in the factorcost unit of account and it is therefore still necessary to make corrections to ensure the values used inNESA are in factor costs. NESA02 onwards takes the outputs in factor costs and also presents them inmarket prices.

6.3 The largest proportional adjustment to be made from market to factor costs is in the valuation of thechange in fuel consumption in making a trip. In NESA, for each litre of diesel, the non-business userpays 86.95 pence (the average pump price in 2009), of which 45.51 pence (52.3%) is excise duty, and11.34 pence (13%) is VAT. The factor cost price is net of these tax transfers. In this case the factor costprice is 30.1 pence per litre (that is, 34.6% of the pump price). Business users who can reclaim VATpay 75.61 pence (i.e. 86.95 - 11.34 pence) but the factor cost price remains the same. A similaradjustment is made in respect of indirect taxation on other items of vehicle operating cost namely oil,tyres, depreciation and maintenance. (These items of course bear only VAT and not excise duty).

6.4 An adjustment is also made to the valuation of non-working time. This valuation is based onobservations of behaviour which provide an average value per hour of time saved. This means thatpeople can be considered as prepared to sacrifice that amount of expenditure on other goods andservices in order to save one hour of non-working time. But these alternative goods and services bearprices that include the indirect tax element (mainly VAT). They are, therefore, in the market price unitof account. The factor cost of these other goods and services is less by the average rate of indirecttaxation (net of subsidies) in consumers’ expenditure. The average rate of indirect taxation inconsumers’ expenditure is calculated as:

(indirect taxes - subsidies) on consumers’ expenditureconsumers expenditure at factor cost

6.5 The factor cost value of non-working time may be regarded as the value to the economy (in terms offactor inputs) of the alternative goods and services that would yield the individual the same benefit.

6.6 In calculating the value of work time, gross wage rates including National Insurance and pensioncontributions, are used rather than rates net of taxation. This is because, in competitive equilibrium, itis the employee’s gross remuneration that represent how much he is costing his employer and thusreflects his worth to his employer, and by analogy, the value to the economy of his output. These valuesare already in the factor cost unit of account and thus do not need any further adjustment.

Chapter 6 Volume 15 Section 1The Treatment of Taxation Part 3 Economic Concepts in NESA




VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


THE APPLICATION OF NESA

Contents

Chapter

1. Economic Decision Criteria: Viability ofSchemes

2. Appraisal Under Uncertainty

3. Incremental Analysis: The Choice of Routeand Standards

4. The Appraisal of Competing andComplementary Schemes

5. Effects of Delaying Construction

6. The Appraisal of Schemes with DifferentOpening Years

PART 4

Volume 15 Section 1Part 4 The Application of NESA


Volume 15 Section 1 Chapter 1Part 4 The Application of NESA Economic Decision Criteria: Viability of Schemes


1 ECONOMIC DECISION CRITERIA: VIABILITY OF SCHEMES

1.1 As discussed in Parts 2 and 3 NESA’s evaluation model uses cost-benefit analysis techniques to deriveeconomic indicators of a scheme’s value for money. These indicators (defined in Part 3 Chapter 3)allow Transport Scotland to determine:

(i) whether a scheme is economically justified

A scheme is justified in purely economic terms if it displays a Net Present Valuegreater than zero, that is the sum of its discounted benefits exceeds the sum of itsdiscounted costs. If the NPV is negative, this indicates that its discounted benefits areless than its discounted costs; therefore the reference case (e.g. the Do-Minimum) ispreferred in economic terms.

(ii) which of several mutually exclusive options is the preferred scheme

When there are a number of competing Do-Something options, the preferred option ineconomic terms is selected using a technique described in Part 4 Chapter 3 calledincremental analysis.

(iii) the optimal combination of interrelated competing or complementary schemes

Most trunk road schemes are free-standing. However, where the planning ofindividual road schemes becomes interrelated there arises a question of how toappraise a package of schemes which may be complementary or offering competingsolutions to a traffic, operational or environmental problem. This is often a strategicevaluation problem and is described in Part 4 Chapter 4.

(iv) the optimal timing of a scheme

NESA can be used to determine the optimum opening year for a proposed roadscheme from an economic point of view (see Part 4 Chapter 6).

(v) which of several schemes with different opening years is the preferred scheme

It is sometimes necessary to appraise and compare route options with differentopening years. Chapter 6 describes the methodology that should be used to do this.

(vi) the ranking of schemes in the roads programme

The trunk road programme budget is cash limited so it is not always possible to buildevery economically justifiable scheme. Therefore, the best projects must be selectedfrom the range of worthwhile schemes available, in other words, the economicobjective is to maximise the benefits from the available funds. In such circumstancesthe ratio of PVB to PVC, the benefit cost ratio (BCR), can be used to rank schemes inorder of declining BCR, selecting the highest ones first until the budget is exhausted.

Chapter 1 Volume 15 Section 1Economic Decision Criteria: Viability of Schemes Part 4 The Application of NESA


Volume 15 Section 1 Chapter 2Part 4 The Application of NESA Appraisal Under Uncertainty


2 APPRAISAL UNDER UNCERTAINTY2.1 It is important for decision makers to have some idea about how robust the results may be in order to

know what weight to attach to them. It is therefore necessary to consider a range of possible outcomes.

2.2 Any investment decision, whether public or private sector, is bound to be subject to uncertainty.Decisions regarding long lived investments with distant forecast horizons, such as road improvementproposals, are subject to a high degree of uncertainty. It is constructive to set bounds on this uncertaintyby carrying out tests on key variables to identify those variables to which the NESA results areparticularly sensitive. These are the variables on which decision makers’ judgment should focus.

2.3 Errors that occur in a NESA appraisal can be classified as one of three types. Firstly, there may bespecification errors introduced by the incorrect formulation of the model, for example, in theassumptions underlying the value of work time in NESA. Secondly, there are measurement errorsresulting primarily from the use of sampling techniques, for example the confidence placed on thecalibration of modelled journey times depends on the size of the sample of measured timed runs. Thisincludes measurement errors in the traffic appraisal process which leads to the traffic flow input toNESA. Thirdly, there may be prediction errors introduced by forecast inputs, such as the future growthof GDP which cannot be forecast with certainty. DMRB Volume 12 describes how uncertainty shouldbe handled in the context of traffic modelling procedures.

2.4 Effort is made to minimise specification and measurement errors in NESA by carrying out external andinternal reviews of NESA incorporating the latest empirical evidence wherever possible. However, itwill often be useful for the user to carry out sensitivity tests on variables which are both uncertain in thelocal context and likely to affect the NESA result significantly. Sensitivity testing using local data isdescribed in Part 5 Chapter 2.

2.5 It may well be the case that a sensitivity test highlights variation in the NPV, but that the sign of theNPV and ranking of options remains unchanged. This would clearly increase the weight which can beput on the economic results. However, sensitivity tests are not costless to carry out and need to beconsidered carefully.

2.6 Errors introduced by forecasting inputs are known to have a significant impact on NPV results, inparticular the impact of GDP and fuel price assumptions on traffic forecasts and the associated valuesof time, accidents and VOC used in NESA. It may be therefore necessary to present NESA resultsbased on both high and low traffic/economic growth. The high and low assumptions used are the sameas those presented in the National Road Traffic Forecasts and WebTAG. While this range does notexhaust the possibilities it should capture the more plausible outcomes. Decisions should necessarilytake account of both forecasts, with the cost of being wrong a key consideration, both in the sense ofthe cost of unnecessary investment in the light of low growth and the cost of future congestion andpossible remedial action in the light of high growth.

2.7 The introduction of high and low growth NPVs complicates the decision criteria described in Part 4Chapters 1 and 3. Investment choices should be considered in the context of risk analysis; an exampleis where the choice is between D2 and D3 standards for a motorway. On high growth, the NPV of theD3 standard scheme may be higher; on low growth the D2 NPV may be higher. This clearly presents adilemma.

Chapter 2 Volume 15 Section 1Appraisal Under Uncertainty Part 4 The Application of NESA


2.8 It is not possible to attach reliable probabilities to the range of low to high growth NPVs. Users arerecommended to present NESA results as follows:

(i) list all the possible options, including staged construction where feasible (forexample, adding lanes in future years) and determine the NPVs for both high and lowgrowth scenarios

(ii) rank options by NPV for both high and low growth. If the same option is ranked firston both high and low growth, this is the overall preferred option from an economicpoint of view. This will be a common situation where uncertainty does not greatlycomplicate the incremental analysis. The preferred Do-Something option has also tobe justified economically against the Do-Minimum. If the preferred Do-Somethingoption has a positive NPV under both high and low growth, the economic justificationagainst the Do-Minimum is unambiguous. If it has a positive NPV on high growth butnegative on low growth, its economic justification with respect to the Do-Minimum isdependent on whether high or low growth materialises. In this instance a scheme mustbe considered to be marginal from an economic point of view, dependent in part onthe weight attached to the high or low growth scenario result

(iii) if different options are ranked first on high and low growth, it is prudent to considerthe balance of advantage of each option. For example, consider whether the schemeranking is marginal on high growth but very clear cut on low growth. Consider alsothe possible implications of choosing the wrong option, for example, the optionpreferred on low growth in the light of high growth actually materialising, and viceversa, in terms of user costs and unnecessary capital costs. The non-economicadvantages and disadvantages of each option should also be considered

2.9 Where (iii) arises, there is no simple decision rule which can be applied to arrive at a clearly preferredoption. Intuitively one wishes to select a portfolio of road schemes in the trunk road programme whichis balanced in the sense that if low growth materialises there is not too much overcapacity in the trunkroad system and if high growth materialises there is not too much congestion and the possible need forlater widening. The balance between on one hand possible over provision and over-expenditureinitially and, on the other hand, possible under capacity in later years has to be presented and judged foreach scheme explicitly.

2.10 Two examples of the application of high and low growth illustrate the approach described above.(These examples are limited to the economic evaluation. Environmental factors would also need to beconsidered in practice and these may be subject to uncertainty themselves.) NPV results are expressedin £ million 2010 price and discount base and include delays during construction and changes to futuremaintenance costs.

EXAMPLE 1

High growth Low growth£m £m

Option 1 20 10Option 2 10 8

Incremental NPV of Option 1 over Option 2: High growth = £10mLow growth = £ 2m

Volume 15 Section 1 Chapter 2Part 4 The Application of NESA Appraisal Under Uncertainty


Conclusion:

Option 1 is unambiguously preferred overall and is also unambiguously positive. The decision criteriaset out in Part 4 Chapter 1 apply unchanged.

EXAMPLE 2

Conclusion:

Both options have unambiguously positive NPVs, but option 1 is clearly preferred if high growthmaterialises. If low growth materialises, option 2 is marginally preferred. The lowest NPV return ofoption 1 within the high/low range is £4m while the lowest NPV return of option 2 is £6m. If onewished to minimise the chance of getting a relatively low economic return one would choose option 2.However, if one took a weighted average (for example 60/40) of the low and high growth NPVs onewould choose option 1 (this weighting gives an average NPV of £10.4m for option 1 and £7.6m foroption 2). The conclusion is that there is no simple economic criterion for choosing the optimumoption. In this example where the NPV range appears to be non-linear in option 1, one might examinethe consequences of traffic forecasts within the high and low range. The analysis of uncertainty can beused creatively to suggest new options for testing, for example, in the above example 2 a third optiondefined as staged addition of capacity if high growth materialises during the next 10 or so years may beoptimal.

High growth Low growth£m £m

Option 1 20 4Option 2 10 6

Incremental NPV of Option 1 over Option 2: High growth = £10mLow growth = £ -2m

Chapter 2 Volume 15 Section 1Appraisal Under Uncertainty Part 4 The Application of NESA


Volume 15 Section 1 Chapter 3Part 4 The Application of NESA Incremental Analysis: The Choice of Route and Standards


3 INCREMENTAL ANALYSIS: THE CHOICE OF ROUTE AND STANDARDS

3.1 When there is more than one Do-Something option, an analysis of incremental BCRs is required.Given the funds available for the trunk road programme it may not be appropriate to accept a smallscale high BCR option if this precludes a larger scale option with a lower, but still acceptable, BCR. Aprocedure called incremental analysis should be carried out to ensure that while not only is the schemeeconomically justified, it also helps to maximise the benefits of the overall programme. Incrementalanalysis increases the scope for the application of NESA. Two areas are discussed below, namely, thechoice of route and the choice of standards.

3.2 Incremental analysis uses BCRs to choose the option which is best given the funding constraints. Theprocedure is as follows:

(i) arrange the Do-Something options in rising order of cost;

(ii) beginning with the lowest cost option, consider the next higher cost option andcalculate the ratio of the present value of incremental benefits to the present value ofincremental costs, where incremental benefits and costs are those in excess of thebenefits and costs of the lower costs option;

(iii) an incremental ratio greater than the incremental cut off BCR implies that theincremental benefits are achieved at a sufficiently high rate per pound spent tocompete with other independent schemes. Thus if the incremental ratio is:

- greater than or equal to the incremental cut off, reject the lower costs option and make the higher cost option the basis for comparison with the next higher cost option; or,

- less than the incremental cut off, reject the higher cost option and use the lower cost option as the basis for the next increment;

(iv) repeat this procedure until all the options have been analysed; and,

(v) choose the option with the highest capital cost which has an incremental BCR equal toor greater than the incremental cut off value.

3.3 To illustrate how incremental analysis should be applied consider the five mutually exclusive schemeoptions ranked in order of increasing cost in Table 4/3/1.

Table 4/3/1: Example of Scheme Option Ranking

OPTION NET PRESENT BENEFIT

£m

NET PRESENT COST

£m

NET PRESENT VALUE

£m

BENEFITCOST RATIO

A 6.5 1.5 5.0 4.3B 8.0 2.5 5.5 3.2C 15.5 4.5 11.0 3.4D 20.0 6.5 13.5 3.1E 22.0 10.0 12.0 2.2

Chapter 3 Volume 15 Section 1Incremental Analysis: The Choice of Route and Standards Part 4 The Application of NESA


3.4 Assume that the budget constraint is such that the incremental cut off BCR is set at 2.0. Theincremental BCR of each higher cost option is calculated, rejecting those below the incremental cut offvalue. The results are shown in Table 4/3/2.

Hence option D is chosen.

Note that under a tighter budget constraint which had an incremental cut off BCR of say 2.5, the lowercost option C would be selected.

3.5 For most trunk road schemes there will be a number of possible routes. A common situation might bewhere there is an on-line alternative and two off-line routes either side of the town. In this case, thereare three Do-Something options and incremental analysis should be used to select the preferred option,other things being equal. The choice of route can involve radically different options which may to someextent serve different traffic movements and objectives. At the other extreme, the choice of route caninvolve very fine alternatives for example in vertical and horizontal alignment, and may not appearsignificantly different in NESA terms.

3.6 NESA should also be used to contribute to the evaluation of the choice of link and junction standardsincorporating future maintenance costs and associated traffic delays where appropriate (see Part 6Chapter 10 and 11). Local site specific characteristics, such as local construction cost estimates, can beinput to NESA and can be used to demonstrate that the preferred standard is economically optimal, orthe extent to which it is sub-optimal if non-economic factors dominate. Examples are where there is achoice between building the Do-Something option to D2AP standard or wide single 10 metre; or wherethere is a choice between designing a Do-Something junction as a roundabout or a grade separatedjunction. The two Do-Something options can be compared using NESA and incremental BCR analysisto select the preferred design. For example, a single carriageway solution can be treated as theDo-Minimum and a dual carriageway solution as the Do-Something for incremental analysis, providedthat the fixed trip matrix assumption is maintained. The scheme itself, incorporating the preferredstandards, should of course be justified against the Do-Minimum.

3.7 The following considerations should always be addressed when NESA is used to appraise highwaystandards:

(i) the economically preferred standard in terms of incremental BCR analysis may notnecessarily be preferred when non-economic factors are taken into account. Tradeoffs between economic and non-economic factors should be set out explicitly, usingthe implicit valuation approach described in Part 9 Chapter 4 where appropriate;

(ii) the attention paid to differences in BCR between options should be based on anunderstanding of the speed/flow relationships, robustness of capital cost estimates,etc. which can give rise to these differences. Small differences in BCR may not beconsidered significant;

Table 4/3/2: Example of Incremental Analysis

BASIS FOR COMPARISON

NEXT HIGHERCOST OPTION

INCREMENTALBCR

PASS/FAILCUT OFF = 2.0

A B 1.5 failB C 3.0 passC D 2.3 passD E 0.6 fail

Volume 15 Section 1 Chapter 3Part 4 The Application of NESA Incremental Analysis: The Choice of Route and Standards


(iii) where NESA analysis favours standards which appear to conflict with other criteriaissued, Transport Scotland should be consulted. For example when extra capacitymay be required on short links involving high weaving proportions, the safety andcongestion implications that are not fully modelled in NESA should be set out againstan economic appraisal;

(iv) the appraisal of standards should incorporate the range of low and high traffic andeconomic growth forecasts. Their use is described in Part 5 Chapter 6.

3.8 One area of incremental analysis worthy of more extensive description is the analysis of the provisionof additional link capacity, either as a road widening scheme or as a choice of standards in a newscheme. For example, consider a proposed motorway scheme where forecast traffic flows indicate thateither D2 or D3 may be the appropriate standard. The first step is to appraise the two Do-Somethingoptions (D2 and D3) using NESA. The D3 option will have higher construction costs but can alsoproduce extra user benefits (primarily time savings through the effect of lower flows per lane). Thesecond step is to allow for predicted future lane or carriageway closures, due to future majormaintenance works, using QUADRO. The predicted maintenance works should be checked to ensurethat they are optimal in terms of minimising combined works and delay costs in present value terms.

3.9 The analysis of major maintenance works is described in detail in Part 6 Chapter 10. It is necessary toconsider both the profile of future maintenance works and the associated traffic delays which are likelyto occur. These will be higher if there are no suitable diversion routes. If D2 is chosen, the works costsof maintenance will be lower because there is one lane less in each direction; but this is likely to bemore than offset by the higher traffic delay costs if one less lane is available during maintenance works.Finally, non-economic Framework considerations need to be set out if significant. The economicanalysis may reveal further options to be tested, for example a phased construction providing D2 nowand widening to D3 at some future date; or providing a pavement design with alternative maintenanceprofiles. The analysis becomes slightly more complicated when high and low traffic growth areallowed for (see Part 5 Chapter 6).

Chapter 3 Volume 15 Section 1Incremental Analysis: The Choice of Route and Standards Part 4 The Application of NESA


Volume 15 Section 1 Chapter 4Part 4 The Application of NESA The Appraisal of Competing and Complementary Schemes


4 THE APPRAISAL OF COMPETING AND COMPLEMENTARY SCHEMES

4.1 Most trunk road schemes are free-standing. However, where the planning of individual road schemesbecomes interrelated there arises a question of how to appraise a package of schemes which may becomplementary or offering competing solutions to a traffic, operational or environmental problem.This is a strategic evaluation problem often referred to as the Scheme in Route issue which involvesthe economic evaluation of a complete scheme strategy and the contribution made by its componentsections.

4.2 The scheme in route problem requires that four economic issues be addressed. These are:

(i) the determination of the optimal extent of the strategy and its overall economic worth

(ii) the determination of priority ranking of start dates for sections of the strategy

(iii) the appraisal of the component sections within the strategy

(iv) the determination of precise design standards and alignments for each component

4.3 In practice these four aspects are interrelated problems, in particular (i) and (ii) are closely linkedwhilst (iii) and (iv) are often part and parcel of the same assessment. For ease of exposition eachaspect will, however, be considered in turn. While each of these issues needs to be fully considered, theextent of the analysis should depend on what is relevant to the decisions under consideration; forexample, options which are clearly impracticable should not be appraised. Throughout the analysis,economic considerations need to be set in the context of non-economic factors.

4.4 In order to simplify exposition, the following sections assume that budget constraints do not affect theselection procedure and therefore that maximising NPV is the relevant criterion. In situations whereconstraints do restrict the choice of schemes for which funds are available, BCRs should be used to aidthe selection process. Thus in drawing up the overall strategy it is important to ensure that schemeshave a sufficiently high return to justify their inclusion. Once the strategy has been determined theprogression of schemes within the strategy should reflect the BCR of each constituent part with higherpriority being attached to those schemes with higher BCRs. Such a procedure will enable those partswith the best returns to be completed first. It therefore follows that later sections, with lower BCRs,will inevitably reduce the overall BCR of the strategy. This is acceptable provided that the latersections provide sufficient incremental returns to justify their construction.

Stage 1: The Determination of the Optimal Extent of the Strategy and its Overall Economic Worth

4.5 Consider the example on Figure 4/4/1 of a strategic improvement to an existing road or base networkinvolving the bypasses of a number of towns.

4.6 The notional strategy comprises four component sections; A, B, C, and D which may be competing,complementary or neutral with respect to each other. The evaluation of the strategy will assess thepreferred combination of individual schemes and therefore the optimal extent of the strategy. On thesimplifying assumption that all the schemes open in the same year, there are 15 possible combinationsto assess. (See Table 4/4/1). Appraisal will involve the building from observed data a model of presentand future traffic flows that may use each of the combinations. A NESA economic appraisal may thenbe undertaken to assess the potential benefits of each strategy combination.

4.7 In principle, the preferred combination of schemes will be that which yields the highest aggregate NetPresent Value (NPV). For example, if A + C + D displays the highest NPV over all other combinations

Chapter 4 Volume 15 Section 1The Appraisal of Competing and Complementary Schemes Part 4 The Application of NESA


it will form the economically preferred strategy, even over the notional full strategy of A + B + C + D.However, there may be a non-economic justification to retain scheme B within the strategy.

Stage 2: Priority Ranking of Scheme Options

4.8 The second problem is to decide which section in the preferred strategy of A + C + D should be builtfirst, although it should be noted that the best phasing may be to complete the strategy all at once. Thatcomponent which offers the highest NPV should be undertaken first. It is therefore necessary to testeach scheme independently against the base Do-Minimum network that excludes all the other elementsof the strategy to establish that section which delivers the highest NPV.

4.9 If it is established that scheme A should be given priority then clearly it will form part of the networkbefore other schemes in the strategy become operational. A is, therefore, added to the Do-Minimumnetwork against which C and D are tested respectively in order to establish which one of these schemestakes precedent over the other.

Stage 3: The Appraisal of Component Sections

4.10 Having established the order in which to progress the individual schemes within a strategy, the nextaspect is to assess the worth of each component to the strategy. This is carried out by Exclusion andIsolation analysis. This incremental analysis will provide a check against possible over provision andensure the best value for money is obtained. It establishes the economic worth of a scheme, bearing inmind the other schemes in the surrounding area. Clearly, whilst a strategy may be economically viable,

Figure 4/4/1: Testing the Strategy

Table 4/4/1: Strategy Combinations

STRATEGY COMBINATIONSA A + B + D B + C + DA + B A + C + D CA + C B C + DA + D B + C DA + B + C B + D A + B + C + D



an individual section may prove to be neutral to the strategy or even in competition with othercomponents and would leave the strategy NPV largely unchanged or increased if it was not built.

4.11 Exclusion analysis regards each component as a missing link in the preferred strategy. Each link isappraised against a Do-Minimum base network which changes through time. For example, let usassume that the opening dates for the preferred strategy of A + C + D have been established as year 0,year 5 and year 10 respectively. The economic appraisal of section A will incorporate a base Do-Minimum network which includes scheme C from year 5 and D from year 10 onwards over theappraisal period. The test for scheme C will have regard to a Do-Minimum network that will include Afrom year 0 and D from year 10. The exclusion test of D appraises the component section against aDo-Minimum comprising A from year 0 and C from year 5. These test networks are showndiagrammatically on Figures 4/4/2 to 4/4/4.

4.12 The opening year of the first scheme, in this case section A, is taken as the starting point for theappraisal. Where component sections have differing opening dates, the terminal year of the appraisalperiod should be 60 years after the opening of the last scheme, in this case section D. Each test revealsthe NPV of the preferred strategy without the contribution of one of its component sections, that is thestrategy NPV that would be foregone. If each section is highly complementary to other schemes thenone may expect the difference in NPV values of the strategy and the exclusion of sections to besignificant. Further information on the common terminal date approach is given in Part 4 Chapter 6.

4.13 Isolation assessment appraises each component against a Do-Minimum network that does not changethrough time, incorporating only those schemes that have an earlier opening date scheduled. Thus, A iscompared in isolation to the more familiar Do-Minimum network in year 0. Scheme C is measuredagainst a base network including A from year 0, while D is compared to a base network incorporatingA from year 0 and C from year 5. Again, the common terminal date approach should be adopted.

4.14 The isolation assessment of each section reveals the economic worth of a scheme if later sections in thestrategy are not built. If sections are largely neutral with respect to each other NPV values should not

Figure 4/4/2: Exclusion Analysis for Scheme A



Figure 4/4/3: Exclusion Analysis for Scheme C

Figure 4/4/4: Exclusion Analysis for Scheme D



change dramatically for previous sections between each additive test. However, NPV values for say Aand C will improve significantly if D complements their services. This assessment technique, therefore,displays the sensitivity of each schemes justification to the assumption that the strategy is completed.

4.15 It should be noted that NPV values of individual sections appraised according to the above methodscannot usually be added together to determine the overall strategy NPV. In the case of exclusionanalysis there may be double counting of benefits whereas inclusion assessment may overlookcomplementary benefits. The results of scheme in route analysis should therefore normally bepresented showing the NPV results for the component sections in exclusion and isolation and also forthe strategy as a whole, where the entire strategy has been tested against a Do Minimum base case (seePart 4 Chapter 2). Where component sections have different opening years, the appraisal period for thestrategy should run to 60 years after the opening date of the last section.

Figure 4/4/5: Isolation Analysis for Scheme A

Figure 4/4/6: Isolation Analysis for Scheme C



4.16 A scheme which has a positive NPV in isolation is justified economically, provided that the otherfuture schemes in the strategy are not actually competing with it. However, on the assumption that therest of the strategy is in fact completed, a scheme is justified even if it is negative in isolation, providedit is positive in exclusion and that the strategy itself is positive. In such cases, the negative isolationresult shows the sensitivity of the scheme’s economic justification to the assumption that the strategy iscompleted.

Stage 4: Design Standards and Alignments

4.17 By adopting the same techniques employed in exclusion and inclusion analysis, incrementalassessment can be performed to examine the impact of varying design standards and alignments oneach section and upon the strategy to arrive at an optimal strategy not only in terms of whichcomponent sections are to be included but also their precise design (see Part 4 Chapter 3).

Practical Problems

4.18 The scheme in route methodology set out above is somewhat schematic. Complete conformity to thisapproach is neither possible nor desirable. Each strategy appraisal tends to have its own particularproblems; some of the possible problems are discussed below. It is important, however, that theframework for traffic modelling and economic appraisal should be set out clearly at an early stage inthe assessment of strategies and component schemes, in the light of the principles set out above andtheir relevance to the decisions to be made. The capital and user costs of wrong decisions can faroutweigh the costs incurred in further analysis. Some of the practical problems that may arise with theanalysis of competing and complementary schemes are:

(i) where there are several sections or schemes within a strategy the potential number ofNESA runs to be undertaken can be very large. Simplification should be sought

Figure 4/4/7: Isolation Analysis for Scheme D



wherever possible. However, there remains a trade-off between computational easeand the robustness of end results

(ii) the choice of strategy and choice of timing of component sections are interrelated.There is a simultaneity problem because it is not strictly possible to identify theoptimal strategy without making assumptions about start dates for sections. At theinitial choice of strategy stage, judgements have to be made about the likely optimaltiming of sections, informed where possible by the initial economic analysis along thelines of Paragraphs 4.8 and 4.9. In reality, however, start dates are often dictated bystatutory procedures and resource availability rather than selected on grounds ofoptimal timing (see Part 4 Chapter 6)

(iii) if the start dates of individual sections are spread widely through time an evaluationperiod of 60 years pegged to the opening of the first scheme may ignore the benefitsassociated with the strategy which may only be fully realised once the last section isin place. On the other hand, to attempt a 60 year appraisal from the opening of the lastsection may extend the analysis into a future which is very uncertain, significantlyreducing the confidence which may be attached to the results. However, it is clearlynecessary to ensure that all the potential benefits of a strategy are measured and it isnormally advised to adopt the common terminal year approach for the appraisal, 60years from the opening year of the last section (see Part 4 Chapter 6). In most casesthis should not be greatly removed from the start dates of any earlier components.This advice may extend the analysis further into an uncertain future but the 60 yearevaluation is already viewed as an approximation to infinity. Forecasts employed in aNESA analysis do not lend themselves easily to tests of statistical confidence

(iv) priority ranking of component sections suggested by NPVs may be at odds with thereality of programming the construction which will ultimately dictate what can bebuilt and when

(v) variable trip matrix models may be more appropriate in the consideration of missinglinks in a strategy than the fixed trip model assumption employed by NESA (see Part3 Chapter 5). Adding component sections together implies changes in traffic flowswhich may be significant enough to alter trip assignment thereby breaking the NESAfixed trip matrix assumption

(vi) isolation and exclusion analysis may result in a conflict. Exclusion analysis maydisplay that a component section is beneficial to the whole strategy. Evaluation of asection in isolation may, on the other hand, reveal that as a stand alone scheme it is noteconomically worthwhile and would not be built. As a stand alone scheme that meritsconsideration it may be to a different alignment, terminal points and cost than as apart of a strategy. In this case the final decision on whether or not to build thatcomponent should be delayed until it becomes clear that other sections of the strategywill be built. Greater emphasis in the decision can then be placed on the results ofexclusion analysis to determine the complementary worth of the particular section tothe strategy

Local Authority and Developer Funded Schemes

4.19 Part of road improvement strategy may involve schemes commissioned and funded by local highwayauthorities or private developers. For the purpose of appraisal these schemes can be classified asfollows:



(i) schemes for which there is a commitment to build independent of TransportScotland’s scheme(s)

(ii) schemes dependent on the outcome of Transport Scotland’s choice of route(s) andstandards

(iii) uncommitted schemes that may be either:

(a) independent of Transport Scotland’s scheme(s); or

(b) dependent on the outcome of Transport Scotland’s decision making

4.20 Do-Minimum and Do-Something networks used in the appraisal of the Scottish Executive’s routechoices and standards will be affected in a number of ways depending on the classification given to thelocal authority or private developer scheme(s).

4.21 Where local authority or private developer schemes enter the Do-Something network for appraisal theircosts of construction, land, preparation and supervision should also be included along with the capitalcosts associated with Transport Scotland’s scheme(s). All costs input to NESA should be expressed inundiscounted 2010 values (see Part 6 Chapter 7).

Table 4/4/2: Appraisal of Local Authority and Developer Schemes

LOCAL AUTHORITY/PRIVATE DEVELOPERSCHEME CLASSIFICATION NESA APPRAISAL

i (See paragraph 4.19) Scheme enters Do-Minimumii. (See paragraph 4.19) Scheme enters Do-Somethingiii. (See paragraph 4.19) Sensitivity tests only:If iii (a) Scheme enters Do-MinimumIf iii (b) Scheme enters Do-Something

Volume 15 Section 1 Chapter 5Part 4 The Application of NESA Effects of Delaying Construction


5 EFFECTS OF DELAYING CONSTRUCTION5.1 NESA can be used to determine the optimum opening year for a proposed road scheme from an

economic point of view. This can be calculated using Table 12 of the NESA printout. If a scheme isdelayed for a given number of years (‘n’), there are three effects:

(i) the construction costs and delays during construction of the scheme are delayed ‘n’years. This will reduce the PVC of the scheme because costs are discounted by ‘n’more years.

For example, where ‘n’ = 1 year and using a discount rate of 3.5% real per annum, thePVC is reduced by:

It will not usually be possible to forecast real changes in the price of constructioncosts during the period of delay. Constant real prices should therefore be assumed;

(ii) ‘n’ years of discounted user benefits will be lost if the scheme is delayed for ‘n’ years.This is given in Table 12 of the NESA printout;

(iii) on the other hand if a 60 year evaluation period is maintained this will extend by afurther ‘n’ years into the future if the scheme is delayed. This can be extrapolatedfrom Table 12 of the NESA printout.

5.2 The net discounted benefit of delaying a scheme by ‘n’ years is (iii) - (ii) + (i). If this sum is positive itmay be beneficial to delay scheme opening. Alternatively, a number of NESA runs can be undertakeneach assuming a different scheme opening year. NESA results can then be compared to appraise theincremental impact of an ’n’ year delay to opening.

5.3 If a scheme has a negative NPV (a BCR of less than 1.0), it may be worthwhile conducting sensitivitytests to see if the NPV is improved by delaying the scheme one or more years. This of course will alsodelay potential environmental benefits.

5.4 A positive NPV (a BCR greater than 1.0) indicates that a scheme is economically justifiable. However,consideration of the profile of benefits of the scheme may still be worthwhile when determining thepriority to be given to the scheme. For example, a scheme may have a positive NPV but most of thebenefits may accrue in later years. It may make economic sense to delay scheme opening in order notto displace other schemes in the trunk road programme.

5.5 In addition to NPV and BCR, the principle of the First Year Rate of Return (FYRR) can be used to aiddecision making concerning the selection and timing of scheme options. For example, wherecomparisons have to be made between traffic management and other measures, which may be effectivein the short run, and road schemes where the bulk of total benefits will be realised over a longer timescale as the alternative Do-Minimum network becomes more congested. If the calculated first year rateof return is less than the discount rate, this is an indication that it would be economically beneficial todelay construction of the scheme until a time when the two measures are equated.

5.6 NESA calculates the FYRR. This is shown in Tables 11 and 12 of the NESA print out. It is calculatedas the ratio of the user benefits in the opening year of the scheme to the sum of the capital costs, up toand including that year (see Part 3 Chapter 3).

1 1/1.035n – 1 0.966– 0.034= =

Chapter 5 Volume 15 Section 1Effects of Delaying Construction Part 4 The Application of NESA


Volume 15 Section 1 Chapter 6Part 4 The Application of NESA The Appraisal of Schemes with Different Opening Years


6 THE APPRAISAL OF SCHEMES WITH DIFFERENT OPENING YEARS

6.1 It is sometimes necessary to appraise and compare route options with different opening years. Whenthis occurs it is imperative that a common terminal date for assessments is adopted.

6.2 Although it is possible that with full maintenance throughout the lifespan of a transport project, theperiod of usefulness of the project may theoretically be infinite, NESA uses a 60 year appraisal periodas default. The 60 year appraisal period is considered pragmatic given the risks and uncertainty ofestimating costs and benefits any further into the future. However, rigid application of a 60 yearappraisal period to a number of options programmed to open at different times would ignore benefitsaccruing beyond 60 years to schemes completed the earliest.

6.3 In a comparison of options with different opening dates the adoption of a common terminal datedisregards the same amount of time for each option. For ease, 60 years after the opening of the optionwith the most distant first year of operation can form a convenient terminal point in time. Clearly thiswill extend the appraisal period beyond 60 years for those schemes programmed to open prior to thelater option (see Figure 4/6/1). The benefits for the additional years can either be extrapolated from theNESA printout or calculated by running a separate NESA for the extra period.

6.4 On a practical point, it should be remembered that the option that opens earlier is likely to requireadditional maintenance and this should be included in the assessment. Conversely, schemes with lateropenings dates may need additional network maintenance before the scheme opens.

Figure 4/6/1: Selection of a Common Terminal Year

Chapter 6 Volume 15 Section 1The Appraisal of Schemes with Different Opening Years Part 4 The Application of NESA




VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


TRAFFIC MODELLING IN NESA

Contents

Chapter

1. Principles of Traffic Modelling

2. Definition of Terms

3. Road Network

4. Trip Matrices

5. Route Choice

6. Traffic Forecasts

PART 5

Volume 15 Section 1Part 5 Traffic Modelling in NESA


Volume 15 Section 1 Chapter 1Part 5 Traffic Modelling in NESA Principles of Traffic Modelling


1 PRINCIPLES OF TRAFFIC MODELLING1.1 NESA’s traffic modelling process is illustrated in Figure 5/1/1.

1.2 The two main inputs are the network, specifying the physical structure of the roads, and the tripmatrix specifying the number of zone to zone trips. These can be respectively thought of as the supplyand demand components of the model.

1.3 Both the network and the matrix are input to the route choice component, which allocates trips toroutes through the network according to user defined instructions. The resulting flows and travel timecosts are output for analysis.

1.4 In order to achieve a good representation of the traffic situation in the model’s base year (see DMRBVolume 12), a detailed calibration and validation process must be undertaken (see Part 9). Thevalidated base year model is often referred to as the Calibration Base.

1.5 Once the traffic model has been calibrated and validated, the Do-Minimum and Do-Somethingnetworks can be defined (see Part 3 Chapter 2) and traffic forecasts produced (see Part 5 Chapter 6).

1.6 In determining the traffic forecasts, the same trip matrices are assigned to both the Do-Minimum andDo-Something networks. This methodology is consistent with the Fixed Trip Matrix approach (see Part3 Chapter 5).

Figure 5/1/1: Outline of the NESA Traffic Modelling Process

Chapter 1 Volume 15 Section 1Principles of Traffic Modelling Part 5 Traffic Modelling in NESA


1.7 The traffic forecasts from the Do-Minimum and Do-Something situations form the basis of the inputinto NESA’s economic assessment module.

Volume 15 Section 1 Chapter 2Part 5 Traffic Modelling in NESA Definition of Terms


2 DEFINITION OF TERMS2.1 Traffic input to NESA is used for both the traffic model and the economic evaluation. Due to its

economic assessment component, NESA’s traffic input requirements are more detailed than those thatwould be required for a stand-alone traffic model.

2.2 This chapter therefore aims to define the key terms used in describing the traffic data input to NESA(e.g. network classification, seasonality index, user classes, vehicle category proportions, flow groupsand flow profiles).

Network Classification

2.3 The purpose of classifying a network is to activate default economic evaluation parameters (e.g.E-Factors or the Seasonality Index) where appropriate local data is unavailable.

2.4 NESA contains eight network classifications that reflect the wide range of seasonal variationsencountered on trunk and principal roads in Scotland.

2.5 The Network Classification default is All Roads highlighted in bold above.

2.6 A network classification is recommended for all economic evaluations.

Seasonality Index (SI)

2.7 The Seasonality Index is a measure of the variation that occurs in daily traffic flows throughout theyear. It is defined as the ratio of the 24hr Average Weekday Traffic (AWDT) (all vehicles) for the peakholiday period (16th July to 15th August) to the 24hr Annual Average Weekday Traffic (AAWDT) (allvehicles).

2.8 Figure 5/2/1 shows the relationship between the annual all vehicle profile, the 24hr AWDT flow bymonth, the 24hr AAWDT flow and the Seasonality Index.

2.9 Default Seasonality Indices are available by Network Classification (see Table 5/2/2), however, wherepossible it is recommended that a local SI is used. Ideally, a local SI should be derived using long term

Table 5/2/1: Network Classification Definitions

Network Classification Definition Example1. Urban (URB) Within a large town’s boundary Glasgow, Dumfries

2. Inter-Urban Local (IUL) Roads which have a large proportion of short distance trips (e.g. commuting, shopping)

M8 Edinburgh to WhitburnM80 Stepps to Haggs

3. Inter-Urban General (IUG) Main inter-urban road network outside IUL category

M90 Kelty to Bridge of EarnM80/M9 Haggs to Dunblane

4. Inter-Urban Tourist (IUT) Mix of tourist and IUG traffic A9 Perth to Inverness

5. Rural Local (RL) As IUL, but not an inter-urban strategic route A84 Stirling to CallanderA915 Kirkcaldy to Leven

6. Rural General (RG) As IUG, but not an inter-urban strategic route A91 Dollar to St. AndrewsA71 Larkhall to Darvel

7. Rural Tourist (RT) As IUT, but not an inter-urban strategic route A84/A85 Callander to CrianlarichA93 Blairgowrie to Ballater

8. All Roads Default -

Chapter 2 Volume 15 Section 1Definition of Terms Part 5 Traffic Modelling in NESA


Automatic Traffic Count (ATC) data and input to NESA. Alternatively, a good estimate of the local SIcan be obtained by comparing two or three weeks continuous weekday counts from the peak holidayperiod to those from a neutral month (e.g. March, April, May, October, November).

2.10 A local seasonality index is recommended for all economic evaluations.

12 Hour Traffic Flow Input (E-Factor)

2.11 The E-Factor is used to convert a 12 hour matrix into a 16 hour matrix. The user can either specify alocally derived E-Factor or use the program default. The latter should only be used where local data isnot available.

2.12 The default E-Factor converts 12 hour (0700-1900) AAWDT matrices to 16 hour (0600-2200)AAWDT matrices. E-Factors vary according to the user class proportions in a network, hence differentNetwork Classifications have different E-Factors. Table 5/2/3 contains the NESA default E-Factors byNetwork Classification.

2.13 Ideally a locally derived E-Factor will be based on long term traffic counts details of which should beincluded in the economic evaluation report.

Figure 5/2/1: Defining the Seasonality Index

Table 5/2/2: Seasonality Indices by Network Classification

Network Classification Range of SI Mean/Default Values1. Urban 0.87 to 1.07 0.972. Inter-urban local 0.93 to 1.09 1.023. Inter-urban general 1.00 to 1.26 1.094. Inter-urban tourist 1.30 to 1.42 1.355. Rural local 0.94 to 1.29 1.056. Rural general 1.02 to 1.34 1.167. Rural tourist 1.10 to 1.99 1.448. All roads 0.87 to 1.99 1.07



2.14 The use of a local rather than default E-Factor is recommended.

Flow Factor (F-Factor)

2.15 For most scheme assessments, the input matrix will either contain 12 or 16 hour flows. However, incertain situations local circumstances may dictate otherwise, and a matrix may be input containing, say8 or 14 hour flows. In these situations, a locally derived F-Factor needs to be input to NESA (andapplied to all user classes) to convert the matrix to 12 hour AAWDT.

2.16 Ideally a locally derived F-Factor will be based on long term traffic counts details of which should beincluded in the economic evaluation report.

2.17 A local F-Factor is optional in the economic evaluation.

16 Hour Traffic Flow Input (M-Factors)

2.18 The M-Factor is used to convert a 16 hour matrix to an Annual Flow matrix. Program defaults shouldonly be used where local data is not available. The derivation of the M-Factor should be from long termautomatic traffic counts, details of which should be included in the economic evaluation report.

2.19 If a local M-Factor is not input, NESA will automatically calculate one, according to the followingrelationship:

Note: If a local M-Factor is input, the program default calculated through the seasonality index will beoverridden.

2.20 The NESA default M-Factors convert a 16 hour (0600-2200) AAWDT matrix to an Annual Flowmatrix. The same factors are applied to all user classes and are only dependent upon the month of thetraffic count and the Seasonality Index (see Equation 5/2/1).

Table 5/2/3: Default E-Factors by Network Classification

Network Classification E-Factors1. Urban 1.1812. Inter-urban local 1.1563. Inter-urban general 1.1554. Inter-urban tourist 1.1425. Rural local 1.1566. Rural general 1.1487. Rural tourist 1.1398. All roads 1.158

Equation 5/2/1

where a,b = coefficients (vary with the month that the trip matrix represents) SI = Seasonality Index (a default SI, from the network classification is

only used when an observed SI is NOT input)

M a bSI+=



2.21 The NESA values for coefficients a and b and typical M Factors are given in Table 5/2/4. The straightline relationships these coefficients represent are shown graphically in Figure 5/2/2.

2.22 The use of a local rather than default M-Factor is recommended, particularly if a local seasonalityindex has not been input.

Table 5/2/4: Default a, b Coefficients and Typical M Factors by Month of Count

Parameters Typical M FactorsAll Months a b SI=1.00 SI=1.25 SI=1.50January -181 603 422 573 724February 41 331 372 455 538March 99 277 376 445 514April 263 102 365 390 416May 333 27 360 367 374June 442 -68 374 357 340July 485 -117 368 339 310Holiday 561 -200 361 311 261August 483 -136 347 313 279September 394 -42 352 342 331October 239 115 354 383 412November 38 329 367 449 532December -80 439 359 469 578

Figure 5/2/2: The Variation of M-Factor with Seasonality Index



User Classes

2.23 Traffic data can be input to NESA in up to 15 trip matrix user classes (see Table 5/2/5). The user classesare consistent with standard roadside interview survey procedures (see DMRB Volume 12).

Note: Other - Short distance includes trips up to 50kmOther - Long distance includes trips over 50km

2.24 NESA requires the disaggregation of traffic by vehicle category and journey purpose, so thatappropriate growth rates, vehicle operating costs and values of time can be applied in the forecastingand economic evaluation procedures.

2.25 If the user has not specified the input matrices by user class, then NESA applies default user classproportions by Network Classification (see Table 5/2/6).

Table 5/2/5: Trip Matrix User Classes

User Vehicle Journey Purpose MatrixClass Category Code

1 Car To work/education from home (HBTW)2 Car From work/education to home (HBFW)3 Car To employer’s business from home (HBTEB)4 Car From employer’s business to home (HBFEB)5 Car To other from home - short distance (HBTOS)6 Car From other to home - short distance (HBFOS)7 Car To other from home - long distance (HBTOL)8 Car From other to home - long distance (HBFOL)9 Car Non-home based employer’s business (NHBEB)

10 Car Non-home based other - short distance (NHBOS)11 Car Non-home based other - long distance (NHBOL)12 LGV All journey purposes (LGV)13 OGV1 All journey purposes (OGV1)14 OGV2 All journey purposes (OGV2)15 Coach All journey purposes (COACH)

Table 5/2/6: Default User Class Proportions by Network Classification

User Classes

Network Classification UC 1 UC 2 UC 3 UC 4 UC 5 UC 6 UC 7 UC 8 UC 9 UC 10 UC 11 UC 12 UC 13 UC 14 UC 15

(Car) (Car) (Car) (Car) (Car) (Car) (Car) (Car) (Car) (Car) (Car) (LGV)(OGV1)(OGV2)(Coach)1. Urban 0.16 0.15 0.04 0.04 0.13 0.11 0.02 0.02 0.13 0.08 0.02 0.06 0.03 0.02 0.012. Inter-urban local 0.12 0.12 0.08 0.09 0.06 0.05 0.05 0.04 0.15 0.04 0.04 0.07 0.04 0.05 0.013. Inter-urban general 0.07 0.06 0.10 0.11 0.03 0.03 0.10 0.08 0.14 0.02 0.07 0.07 0.05 0.07 0.014. Inter-urban tourist 0.01 0.01 0.06 0.08 0.03 0.03 0.17 0.13 0.08 0.03 0.17 0.08 0.05 0.07 0.015. Rural local 0.17 0.16 0.03 0.04 0.12 0.10 0.03 0.02 0.11 0.07 0.02 0.06 0.04 0.03 0.016. Rural general 0.00 0.10 0.06 0.06 0.08 0.07 0.07 0.06 0.10 0.05 0.05 0.07 0.06 0.07 0.017. Rural tourist 0.04 0.04 0.05 0.06 0.06 0.06 0.14 0.12 0.09 0.05 0.12 0.08 0.05 0.03 0.018. All Roads 0.12 0.11 0.06 0.07 0.08 0.07 0.06 0.05 0.13 0.05 0.04 0.07 0.04 0.05 0.01



Vehicle Categories

2.26 The five vehicle categories associated with the trip matrix user classes are defined below and illustratedin Figure 5/2/3.

CARS

(CAR) including taxis, estate cars, three wheeled cars, motor invalid carriages, cars towing trailers andall light vans with side windows to the rear of the drivers seat (e.g. minibuses and camper vans);

LIGHT GOODS VEHICLES

(LGV) consisting of all car type delivery vans and those of a larger carrying capacity range butexcluding any vehicle with twin rear tyres. Also including three wheeled goods vehicles;

OTHER GOODS VEHICLES

(OGV1) consisting of all goods vehicles with

(i) two axles with twin rear tyres

(ii) three axles, rigid

(iii) miscellaneous vehicles such as tractors, tractor engines, ambulances, road rollers, etc;

(OGV2) consisting of all goods vehicles with

(i) three axles articulated

(i) with 4 or more axles whether rigid or articulated

BUSES AND COACHES

(COACH) including work buses and mini-buses over six passengers.



Figure 5/2/3: NESA Vehicle Categories



2.27 If the user has specified input trip matrices by vehicle category, NESA will accept a network wide Carin Work Time (CWK) proportion.

2.28 If the user has specified an all vehicle matrix, then the facility exists for the user to define network widevehicle category proportions. While default vehicle category proportions (based on Table 5/2/6) areavailable by network classification (see Table 5/2/7), it is recommended that vehicle categoryproportions are derived from local data.

2.29 There is also a facility within NESA to specify local vehicle category proportions by flow group (seeParagraph 2.31). These proportions need to be supported by long term traffic count data, which has tobe included in the economic evaluation report. In the absence of local vehicle category proportions byflow group, NESA calculates defaults within the routine that defines Hourly Flow Group Multipliers(see Part 5 Chapter 4).

2.30 The relationship between vehicle categories and user classes together with the subdivision of car tripsby working and non-working time is illustrated in Figure 5/2/4.

Table 5/2/7: Default Vehicle Categories by Network Classification

Vehicle CategoriesNetwork Classification Car LGV OGV1 OGV2 Coach

(UC 1-11) (UC 12) (UC 13) (UC 14) (UC 15)1. Urban 0.879 0.055 0.031 0.023 0.0122. Inter-urban local 0.832 0.068 0.042 0.052 0.0063. Inter-urban general 0.807 0.070 0.050 0.066 0.0074. Inter-urban tourist 0.790 0.082 0.049 0.068 0.0115. Rural local 0.858 0.064 0.041 0.029 0.0086. Rural general 0.801 0.071 0.055 0.066 0.0077. Rural tourist 0.841 0.076 0.045 0.028 0.010

8. All roads 0.837 0.067 0.043 0.045 0.008

Figure 5/2/4: The Relationship between Vehicle Categories and User Classes



Flow Groups

2.31 At any location on the road network, the total number of vehicles varies from:

• Hour to hour (e.g. AM and PM peak flow versus night time flow)

• Day to day (e.g. weekday versus weekend flow)

• Season to season (e.g. holiday season versus off-season flow)

Not only does total traffic flow vary in this way, but also the proportion of the different types of tripwhich it contains can vary as well. For example, the relative proportion of cars to goods vehicles islikely to be higher during peak hours than in other hours of the day.

2.32 If an economic evaluation does not account for hourly, daily and seasonal variations in demand, thiscould result in inaccurate results. For example, the variation in hourly traffic demands throughout theday can lead to large junction delays in the AM and PM peaks when a junction is close to capacity andminimal delays at other times. Calculating delays using 12hr AAWDT flows may underestimate totaldelay by suggesting minimal delays at all times, thereby leading to an underestimate of schemebenefits. The same effect could occur with a rural road improvement if the route is situated in a touristarea subject to high seasonal flows.

2.33 To take account of the variation in the level of traffic flow and its vehicle composition, the 8760 hoursof the year are sub-divided into different portions (number of hours) called flow groups. Each flowgroup represents a different level of flow: one flow group represents peak conditions and containssignificantly less hours than that of a flow group representing night time conditions. However, the levelof flow is much higher in the peak period than during the night. To compensate for this large variationin flow, between four and six flow groups can be used. Figure 5/2/4 illustrates the typical distributionof the hourly flow levels over a year, and the manner in which four default flow groups capture thatvariation.

Figure 5/2/5: Flow Groups Representing Annual Flow



2.34 By default there are four flow groups in NESA. Flow group four contains 500 hrs and represents thepeak period, flow group three has 760 hrs and is the shoulder to the peak period, flow group two is thebetween peak period, and has 2500 hrs and flow group one the off - peak (night time), has 5,000hrs (seeFigure 5/2/5).

2.35 In addition to the default four hourly flow groups NESA also allows the use of up to six user definedflow groups, thus enabling the characteristics of particular time periods (e.g. AM and PM weekdaypeak periods) to be modelled. This facility is required if time period matrices other than all daymatrices are input but can also be invoked when all day matrices are specified as input. Table 5/2/8indicates the default number of hours in each flow group.

2.36 In NESA the user can define two types of flow group. The first type relates to specific time periods(e.g. weekday AM, PM and between peak periods) and is defined in the terms of the hours beginning(0-23), the day types (1-4)(see Part 5 Chapter 4) and the months (1-13, see Figure 5/2/1). For example:

(i) AM peak FG: hours beginning 7 and 8; day types 1 and 2; and months 1 to 13

(ii) PM peak FG: hours beginning 16 and 17; day types 1 and 2; and months 1 to 13

(iii) Between peak FG: hours beginning 9 to 15; day types 1 and 2; and months 1 to 13

(iv) Saturday daytime FG: hours beginning 8 to 5: day type 3; and months 1 to 13

2.37 The second type covers the remaining hours in a year and is grouped by flow level. For example:

(i) the next busiest 1,000 hours

(ii) the next busiest 2,000 hours

(iii) all remaining hours

2.38 The user is referred to Part 5 Chapter 4 for further discussion of the use and calculation of flow groupsin NESA.

Traffic Flow Profiles

2.39 NESA uses traffic flow profiles to calculate the level of flow in each flow group. These profilesdescribe how traffic flows vary through the day (daily profile) and through the year (annual profile),and are defined by user class.

2.40 The profiles were derived using data from Transport Scotland’s Automatic Traffic Classifier Sites(ATCS) and Roadside Interview (RSI) surveys.

Table 5/2/8: Default Hours in each Flow Group

Flow Group (FG) Number of Hours in FG1 50002 25003 7604 5005 06 0



2.41 The user is referred to Part 5 Chapter 4 for a further discussion of the use of traffic flow profiles inNESA.

Annual Average Hourly Traffic

2.42 Annual Average Hourly Traffic flows (AAHT) in conjunction with the traffic flow profiles form thebasis of the calculation of the flow in each flow group, unless of course the trip matrices are specifiedby flow group (see Part 5 Chapter 4).

2.43 To calculate AAHT flows from the input matrices NESA uses the F, E and M factors described earlierin this chapter.

2.44 The user is referred to Part 5 Chapter 4 for a more detailed discussion of the conversion of the inputmatrices to AAHT.

The Use of Local Data

2.45 The use of standard values and relationships is central to cost-benefit analysis within NESA, however,NESA has the flexibility for a number of input parameters to be modified by the user.

2.46 Where local data is both reliable and significantly different from national NESA values, it should beinput to NESA and details of the local data should be supplied. A sensitivity test using default values asa benchmark may also need to be carried out; this is to ascertain the importance of the local variationsand to allow comparison of schemes on a similar basis.

2.47 Data input to NESA can be classified in three categories:

(i) Data which is always local, for example, link and junction geometric characteristic(see Parts 7 and 8 respectively);

(ii) Data which should be local if values are reliable and differ significantly from nationalvalues are as follows:

• Network Classification Recommended• Seasonality Index (SI) Recommended• Car-in-Work Time Recommended (not required for trip matrix

input by user class)• Vehicle Category Proportions Recommended (not required for trip matrix

input by user class or vehicle category)• F-Factor (FFAC) Optional• E-Factor (EFAC) Recommended • M-Factor (MFAC) Recommended• Month of Count Optional• Local traffic growth Optional (see Part 5 Chapter 6)• Local accident data Recommended (see Part 6)

(iii) Data which should always be national are, for example, values of time, values ofvehicle operating costs and GDP growth assumptions (see Part 6). Some site specificvariation in accident costs is, however, allowed.

2.48 Reference must be made to Transport Scotland if there is doubt concerning the interpretation ofprocedures in applying NESA.

Volume 15 Section 1 Chapter 3Part 5 Traffic Modelling in NESA Road Network


3 ROAD NETWORK3.1 The NESA user is required to provide descriptions of the road networks for both the existing situation

and all proposed improvements. For modelling purposes the network can be viewed as providing thesupply constraints. For computer coding purposes a network is described as a series of links joining at,or terminating in, nodes. DMRB Volume 12 gives guidance on the definition of a study area, theinclusion of roads in the network and an overview of trip matrix(ces) which contain a summary of allzone to zone movements for the base year. Below is a summary of the main points that should beconsidered when preparing a network description.

Extent of Network

3.2 The network should extend far enough from the improvement to include all links on which there islikely to be significant differences in traffic flows between the Base and Design situations. If a schemeis expected to result in a significant change in the flow level on a competing/complementary route, thecompeting/complementary route should be included in the network. This concept should be balancedby the consideration that, as the network spreads, benefits arising at a distance from the scheme areinherently less plausible and more difficult to assess than local benefits.

Zones and Zone Connectors

3.3 The demand for travel is aggregated into manageable units by splitting the study area into zones, eachzone being represented by a centroid situated at its centre of gravity with respect to traffic flow. Zonesare linked to the main network by zone connectors, which serve to introduce traffic flow onto thenetwork. They are positioned at points within the network where significant amounts of trafficcommence or terminate journeys. NESA requires that the flow from and to each zone should be thesame for the base network and for all the design networks (see the Fixed Trip Matrix assumption Part 3Chapter 5). Zone connectors are excluded from the economic assessment.

Nodes

3.4 Nodes should be placed at significant junctions, where speed limits change and where the road layoutchanges (for example, from single to dual carriageway, but excluding local dualling at junctions).

3.5 Nodes perform slightly different roles in NESA’s assignment and economic evaluation routines. In theassignment model they allow vehicles to pass from one link to another, whilst in the economicassessment model they form the basis of junction delay and accident cost calculations. Consequently, itis important to ensure that the assignment road network coding is detailed enough to produce anunbiased economic assessment, i.e. an assessment that includes all relevant junction delays andaccident costs. To do this for example, a grade separated junction rather than being represented by onenode (as may be sufficient for the assignment model), may require the use of several nodes toadequately capture the junction delays and accident costs that occur.

Links

3.6 A link is a length of road joining two nodes. A link should be a length of road of consistent layoutalong which the volume of traffic flow does not vary substantially; it is usual to model both trafficdirections as a single, two way link.

Chapter 3 Volume 15 Section 1Road Network Part 5 Traffic Modelling in NESA


3.7 For traffic modelling purposes links are described by:

• length

• average vehicle speed

3.8 It should be noted that in addition to the above two properties links are also described using up to amaximum of eight geometric variables, such as hilliness, bendiness, visibility, etc. (see Part 7 Table 7/1/3). Despite being input on the same line in the network data file these properties have no influence inthe traffic model, their role being confined to the calculation of link speeds in NESA’s evaluationmodel. Part 7 contains a detailed description of these additional geometric variables.

3.9 It should be noted that the geometric parameters can be used to ensure consistency between coded andcalculated speeds. Consistency between coded and calculated speeds is an important part of theeconomic assessment validation (see Part 9 chapters 3 and 5).

LENGTH

3.10 Link distance measured to the nearest 10m.

AVERAGE VEHICLE SPEED (KPH)

3.11 The speeds (kph) allocated to each link represent average speeds (by light and heavy vehicles). Eachlink speed is fixed, that is it does not vary with the volume of traffic on a link.

3.12 Ideally all link speeds should be derived from local survey data. However, resource constraints oftenlimit surveys to critical parts of the network. DMRB Volume 12 describes the survey methods involvedin collecting this data. It should be noted that the emphasis on local data is an underlying themethroughout any traffic and economic assessment (see Part 5 Chapter 2).

3.13 In the absence of journey time information for any particular link a NESA default speed can be used.The default speeds are activated through the specification of a road category and a speed limit. The useof a central flag will lower the default speed by 10kph for light vehicles and 5kph for heavies (urbanlink, road categories 1 to 7). Table 5/3/1 lists default link speeds and capacities for each of the fortyNESA road categories. The default speeds could be useful in a situation where speeds on the criticalsections of the network have been surveyed, but resources are not available to make a complete surveyof all link speeds in the study area.

Junction Delays and Banned Turns

3.14 NESA allows the user to specify junction delays, by turning movement, at any node. Banned turns aredefined by setting a very large delay on the banned movement (see Part 10 Chapter 6). These delays areonly used in the traffic model.

3.15 If junction delays are deemed an important influence on route choice then they should be observedbefore being coded into the road network. DMRB Volume 12 gives advice on methods of surveyingjunction delay.

3.16 NESA offers the user facility to apply default junction delays, known as Junction Indices. Theseindices are rural only link properties representing average geometric delay (i.e. delay due todeceleration and acceleration) experienced at roundabouts, traffic signals and on non-prioritymovements at priority junctions. Junction Indices do not represent junction delay caused by queuingtraffic. These indices should only be used where detailed junction information is not available andwhere the collection of such data is not considered to be worthwhile, e.g. for junctions where noqueuing delays are anticipated in future years.

Volume 15 Section 1 Chapter 3Part 5 Traffic Modelling in NESA Road Network


Climbing Lanes

3.17 Roads with climbing lanes should be modelled using two way links with the appropriate RoadCategory as detailed in Table 5/3/1 e.g. 30, 33, 36 or 41. Hilliness rise (HR) and hilliness fall (HF) mustbe coded so that NESA can determine the direction of the climbing lane.

C - Central flag (see Part 7)ST - Small Town flag (see Part 7)Note: The central flag and small town flag are important in defining speed flow types (see Part 7 Chapter 1).

Table 5/3/1: NESA Road Categories, Link Speeds and Link Capacities

Road Speed Limit Default Link Speed CapacityCategory Description Lights Heavies (veh/hr/

(mph) (kph) (kph) (kph) direction)1 Urban - single 6.0m 1 lane/direction 30C 48C 20 20 800

30 48 30 25 80040/50 64/80 40/50 35/45 1500

30/40ST 48/64ST 30/40 25/35 1200

2 Urban - single 7.3m 1 lane/direction 30C 48C 20 20 80030 48 30 25 800

40/50 64/80 40/50 35/45 150030/40ST 48/64ST 30/40 25/35 1200

3 Urban - single 10.0m 1 lane/direction 30C 48C 20 20 80030 48 30 25 800

40/50 64/80 40/50 35/45 150030/40ST 48/64ST 30/40 25/35 1200

4 Urban - single 4 2 lanes/direction 30C 48C 20 20 160030 48 30 25 1600

40/50 64/80 40/50 35/45 300030/40ST 48/64ST 30/40 25/35 2400

5 Urban - one way 6.0m 1 lane 30C 48C 20 20 80030 48 30 25 800

40/50 64/80 40/50 35/45 150030/40ST 48/64ST 30/40 25/35 1200

6 Urban - one way 7.3m 2 lanes 30C 48C 20 20 160030 48 30 25 1600

40/50 64/80 40/50 35/45 300030/40ST 48/64ST 30/40 25/35 2400

7 Urban - one way 10.0m 3 lanes 30C 48C 20 20 240030 48 30 25 2400

40/50 64/80 40/50 35/45 450030/40ST 48/64ST 30/40 25/35 3600

Chapter 3 Volume 15 Section 1Road Network Part 5 Traffic Modelling in NESA


Table 5/3/2: [Contd] NESA Road Categories, Link Speeds and Link Capacities

Road Speed Limit Default Link Speed CapacityCategory Description Lights Heavies (veh/hr/

(mph) (kph) (kph) (kph) direction)8 Urban - dual 2 30/40/50 48/64/80 40/50/65 35/45/60 30009 Urban - dual 3 30/40/50 48/64/80 40/50/65 35/45/60 450010 Urban - dual 4 30/40/50 48/64/80 40/50/65 35/45/60 6000

11 Urban - Expressway, 2 or more lanes 30/40/50 48/64/80 45/55/70 40/50/65 3000

12 Urban - Motorway and dual ramps, 1 lane 30/40/50 48/64/80 45/55/70 40/50/65 190013 Urban - Motorway and dual ramps, 2 lanes 30/40/50 48/64/80 45/55/70 40/50/65 3800

14 Urban - Motorway - D2 50/60/70 80/96/113 80/90/100 75/80/90 380015 Urban - Motorway - D3 50/60/70 80/96/113 80/95/105 75/85/95 570016 Urban - Motorway - D4 50/60 80/96 80/95 75/85 760017 Urban - Motorway - D5 50/60 80/96 80/95 75/85 9500

20 Rural - single 4 lane 60 96 95 80 2300

21 Rural - poor single 4.0m 60 96 50 45 14022 Rural - poor single 5.5m 60 96 55 45 80023 Rural - poor single 6.0m 60 96 60 50 90024 Rural - typical single 6.0m 60 96 65 55 900

25 Rural - poor single 7.3m 60 96 65 55 120026 Rural - typical single 7.3m 60 96 70 60 120027 Rural - good single 7.3m 60 96 80 70 1200

28 Rural - typical single 10.0m 60 96 80 70 150029 Rural - good single 10.0m 60 96 85 75 150030 Rural - single with climbing lane 60 96 80 70 1500

31 Rural - dual 2 lanes 70 113 100 85 340032 Rural - dual 3 lanes 70 113 105 90 510033 Rural - dual 2 with climbing lane 70 113 100 85 5100

34 Rural - dual 2 lanes with grade separation 70 113 105 90 340035 Rural - dual 3 lanes with grade separation 70 113 110 95 510036 Rural - dual 2 lanes with GS and climbing lane 70 113 105 90 5100

37 Rural - Motorway and dual ramps, 1 lane 70 113 90 80 190038 Rural - Motorway and dual ramps, 2 lanes 70 113 90 80 3800

39 Rural - Motorway - D2 70 113 110 95 380040 Rural - Motorway - D3 70 113 115 100 570041 Rural - Motorway - D2 with climbing lane 70 113 110 95 5700

50 Zone connector - - - -

Volume 15 Section 1 Chapter 4Part 5 Traffic Modelling in NESA Trip Matrices


4 TRIP MATRICES4.1 The trip matrix provides the demand in a traffic model. There are no trip matrix estimation facilities

within NESA, thus a previously derived trip matrix has to be input to NESA (see DMRB Volume 12).This matrix typically represents the demand over a user specified time period, for example 12 hourAAWDT (Annual Average Weekday Traffic).

Trip Matrix Input

4.2 NESA allows the user to specify trip matrices by:

• User class - there are fifteen user classes in NESA (see Part 5 Chapter 2), and

• Flow group - there can be up to 6 flow groups (see Part 5 Chapter 2)

4.3 The input of trip matrices by user class allows for a better representation of seasonal effects, whilst theinput of trip matrices by flow group allows for a better representation of peak hour tidal effects.

4.4 The particular choice of matrix input is very much scheme dependent:

(i) If the scheme is subject to seasonal effects (e.g. a tourist route) then annual averagematrix(ces) (e.g. AAHT, AADT, AAWDT) should be input. This traffic data shouldbe disaggregated into as many vehicle categories/user classes as the quality of the dataallows (e.g. vehicle categories if only MCC counts have been made or user classes ifRSI data has been collected). AAHT is the ideal form of input.

(ii) If the scheme is subject to morning and evening peak hour effects (e.g. an urban area)then consideration should be given to the input of traffic data by flow group. NESAallows the user to disaggregate each flow group by user class. If this option is chosenmatrices may typically be defined for four flow groups:

e.g. FG4 AM peak (input matrix) Jan - DecFG3 PM peak (input matrix) Jan - DecFG2 Between Peak (input matrix) Jan - DecFG1 Rest

However, NESA offers the user the facility to define up to six flow groups. Thus it ispossible for the user to input matrices for the AM, BP and PM peaks and let NESAuse the traffic flow profiles (see Part 5 Chapter 2) to define Saturday, Sunday and offpeak flows.

e.g. FG6 AM peak (input matrix) Jan - DecFG5 PM peak (input matrix) Jan - DecFG4 BP (input matrix) Jan - DecFG3 Saturday 7am - 7pm Jan - DecFG2 Sunday 7am - 7pm May - SepFG1 Rest

Trip Matrix Conversion to AAHT (Seasonality Modelling Only, see Paragraph 4.4 (i))

4.5 NESA requires AAHT matrices by user class for the purpose of modelling seasonality.

4.6 Unless AAHT matrices by user class are provided, the input matrix(ces) are disaggregated to the fifteenuser classes using local vehicle category proportions and/or the default user class proportions detailedin Table 5/2/7.

Chapter 4 Volume 15 Section 1Trip Matrices Part 5 Traffic Modelling in NESA


4.7 The fifteen matrices (by user class) are converted to AAHT flows through the application of the F, Eand M factors (annualisation).

Day Types

4.8 Travel demand for a particular journey purpose varies with the day of the week. For example, at theweekend home-based to work trips fall to approximately 20% of their weekday level, whilst other(leisure) trips experience an increase in demand. In addition to this variation, travel demand for aparticular journey purpose can also occur at different times on different days of the week.

4.9 To model these daily variations NESA uses four day types:

Day Type 1 - Monday to Thursday

Day Type 2 - Friday

Day Type 3 - Saturday

Day Type 4 - Sunday

Traffic Flow Profiles

4.10 NESA uses daily profiles by user class and day type and annual profiles by user class to model daily,weekly and seasonal variations in traffic flow. These variations represent the demand to undertake aparticular activity at a certain time of the day on a certain day. They are independent of NetworkClassification. The profiles were derived using data from Transport Scotland’s Automatic TrafficClassifier sites (ATCs) and Roadside Interview (RSI) surveys.

DAILY PROFILES AND HOURLY FACTORS (HF)

4.11 The daily traffic profiles in NESA describe the hourly travel demand variations by user class and daytype. Figure 5/4/1 indicates the very different daily profiles of three of the fifteen user classes; carhome to work, car work to home and OGV2 for two day types.

4.12 NESA uses 24 Hourly Factors (HF) to describe each daily profile. These are defined as follows:

ANNUAL PROFILES AND DAILY FACTORS (DF)

4.13 The annual profiles in NESA describe both the seasonal and daily variations in travel demand by userclass. As with daily profiles, the annual profiles represent the demand to undertake a particular activityon a certain day in a year. Figure 5/4/2 indicates the very different annual profiles for three of the userclasses; car home to work, car non-home based other and OGV2.

Equation 5/4/1

where HFcht = Hourly Factor for user class (c) hour (h) and day type (t)Tcht = Volume of traffic of user class (c) in hour (h) on day type (t)DTc = Volume of Daily Traffic of user class (c)

HFchtTchtDTc----------=



Figure 5/4/1: Day Type 1 & 3 Daily Profiles for Three User Classes



4.14 NESA uses 365 Daily Factors (DF) to describe each annual profile. These are defined as follows:

Mean Factor Values

4.15 By extracting all the Daily Factors for each day type (by user class), it is possible to define a meanfactor value for each day type and user class. This results in 60(=4*15) mean factor values in total. Acomparison of mean factor values within each user class indicates how travel demand variesthroughout a week.

Figure 5/4/2: Annual Profiles for Three User Classes

Equation 5/4/2

where DFcd = Daily Factor for user class (c) and day (d)DTcd = Volume of traffic of user class (c) on day (d)AADTc = Annual Average Daily Traffic flow for user class (c)

DFcdDTcd

AADTc--------------------=



Adjustment of Annual Profiles and Daily Factors

4.16 As discussed in Part 5 Chapter 2, a local Seasonality Index should normally be input to NESA. Toensure NESA captures local seasonal travel demand variations the program synthesises a SeasonalityIndex from the annual profiles and compares it to the observed one.

4.17 If the synthesised and observed SIs do not agree within acceptable limits (+/-1%), NESA automaticallyadjusts the annual profiles for Other trips (user classes 5-8 and 10-11) until the observed andsynthesised SIs are sufficiently close. The rationale for only adjusting the profiles for Other trips is thatthese trips are the major source of seasonal variation.

4.18 The adjustment routine modifies the other Daily Factors (DFs) using a scaling factor based on twovariables; firstly the difference between the observed and synthesised SI; and secondly on theproportion of Other trips on the link(s) used to define the synthesised SI.

4.19 This scaling factor is applied to the difference between each Daily Factor and its corresponding meanfactor value to produce a new Daily Factor. This method ensures that both the total annual traffic flowper user class remains unchanged and the mean factor values for each day type remain unchanged.

4.20 The synthesised SI is then re-calculated and compared with the observed value. If the synthesised andobserved SIs do not agree within the prescribed limits, a new scaling factor is calculated and the Otherfactors revised. This process is repeated until convergence is obtained.

4.21 An amended annual profile resulting from the adjustment process is shown graphically in Figure 5/4/3.

4.22 If the observed seasonality is too high, this adjustment procedure can produce unrealistically low flowsfor winter months. To address this problem, a maximum adjustment factor has been incorporatedwithin the program to restrict the amount by which the Other Daily Factors can be scaled DOWN.

Figure 5/4/3: An Example of an Adjusted Profile



Hourly Flow Group Multipliers (Seasonality Modelling Only, see Paragraph 4.4 (i))

4.23 Hourly Flow Group Multipliers (HFGM) are used to calculate the level of flow by user class in eachflow group from the AAHT flows.

4.24 Hourly Flow Group Multipliers by user class are calculated by the following procedure:

(i) Specify the number of flow groups and the numbers of hours in each group; theNESA defaults are given in Table 5/2/8.

(ii) Calculate the total network demand in each of the 8760 hours of a year (Tripschd)using hourly and daily factors derived from the daily and adjusted annual profiles (seeEquation 5/4/3).

Figure 5/4/4 shows a typical weekly profile calculated using Equation 5/4/3. Forclarity the fifteen user classes have been aggregated into four user groups in thisfigure (cars (non-work), cars (in-work), LGV, HGV).

(iii) Rank the hours by order of magnitude.

(iv) Group the hours into their respective flow groups.

(v) The Hourly Flow Group Multipliers (by flow group and user class) are defined by theratio of average flow by user class in each flow group to the AAHT flow by user class(see Figure 5/4/4).

4.25 It should be noted that the average demand level in each flow group is set to ensure the total annualflow is correct.

4.26 The hours that tend to make up a flow group are dependent upon the area being modelled. For example,the peak flow group will mainly contain hours from the holiday season on a rural tourist route, whereasfor an urban area it mainly contains AM and PM peak hours.

Flow Group Traffic Levels

4.27 For each network link, the flow group travel demand levels are determined by applying Hourly FlowGroup Multipliers (by user class) to the AAHT travel demands by user class if seasonality is beingmodelled or directly if flow group matrices are input.

Equation 5/4/3

where AAHTc = Annual Average Hourly Traffic flow for user class (c)HFcht = Hourly factor for user class (c), hour (h) and day type (t)DFcd = Annual daily factor for user class (c) and day (d)

Tripschd AAHTc * HFcht * DFcd=



Figure 5/4/4: Example Total Weekly Traffic Flow Profile (January)

Volume 15 Section 1 Chapter 5Part 5 Traffic Modelling in NESA Route Choice


5 ROUTE CHOICEIntroduction

5.1 The route choice component of NESA assigns the trip matrix to the road network according to userdefined instructions. NESA offers two assignment techniques, a single route All-Or-Nothing (AON)method and a multiple routeing method based upon Burrell’s technique.

Generalised Cost

5.2 Many factors influence drivers’ choice of route, amongst which are time, petrol costs, tolls, distance,congestion, signposting, ease of travel, safety and scenery. However, various studies have consistentlyshown that the most statistically significant factors are time, distance and out of pocket costs. Bycombining these factors into a weighted generalised cost (usually expressed in units of generalisedtime) NESA is able to explain most route choice behaviour (see Equation 5/5/1).

5.3 Economic theory suggests that by giving behavioural values of time and distance to coefficients a andb, respectively, and a value of 1.00 to c, the generalised cost of routes can be measured and optimalroutes then chosen.

5.4 However, in modelling terms it is the ratio of a to b that is the most important factor in defining routechoice. It is therefore common place to set the value of a to 1.00 and use only the parameters b and c ascalibration tools. The recommended procedure is to base these route choice coefficients on theguidance presented in TAG Unit 3.1.2 which should be the starting point for a number of sensitivitytests that should demonstrate the robustness of the resulting assignment model (ensuring that smallchanges in time and distance coefficients do not lead to large changes in forecast flows).

5.5 For NESA models which have All Vehicle matrices a suggested starting point is a generalised costequation of the form (t + 0.5d). It should, however, be noted that TAG Unit 3.1.2 also providesguidance on separate time and vehicle operating cost values for working and non-working trippurposes, and for different vehicle types and these can be incorporated into the assignment process inNESA (see Part 10 Chapter 6).

Equation 5/5/1

where a, b, c = coefficientst = link travel time (minutes) calculated from coded link distances

and speedsd = link distance (kilometres)TOLL = out of pocket expenses (pence)

Cost at bd cTOLL+ +=

Chapter 5 Volume 15 Section 1Route Choice Part 5 Traffic Modelling in NESA


Tree Building

5.6 An integral part of any assignment routine is the method employed to define the minimum cost pathbetween each O-D pair. Since there is usually only one minimum cost route to any node from a givenorigin, the optimal route to zones which pass through that node must share a common route at least asfar as that node (unless of course there are some banned turning movements existing at the node). Thusthe pattern of optimal routes from a given zone is one of continuous branching, hence the analogybetween minimum cost route building and a tree.

5.7 To account for banned turns, NESA expands each node at which there is a turning delay (see Figure 5/5/1). In the expanded node each turning movement is simulated by a turning link. To prevent anytraffic using a banned turn its associated turning link is given a very high delay (see Part 5 Chapter 3and see Part 10 Chapter 6).

All-Or-Nothing Assignment

5.8 The simplest route choice and assignment method is the All-Or-Nothing assignment. This form ofassignment finds the minimum cost path between each origin destination pair and assigns all the traffic(demand) to that route and nothing to the less attractive routes.

5.9 All-Or-Nothing assignments tend to produce realistic link flows in sparse and uncongested networks,and consequently are suitable for many inter-urban trunk road schemes. However, in areas in whichthere are at least two or more competing routes an All-Or-Nothing assignment can produce a situationwhere one route is overloaded and the others have little or zero flow. In this situation, the multiplerouteing assignment facility in NESA should be used.

Multiple Routeing

5.10 It is an over-simplification to suppose that every driver making a particular zone-to-zone movementwould use an identical route, or that the same individual driver would use the same route on every

Figure 5/5/1: Node Expansion

Volume 15 Section 1 Chapter 5Part 5 Traffic Modelling in NESA Route Choice


occasion. Drivers do not have perfect knowledge of journey times and distances, and the former in anycase are variable both randomly and, for example, by time of day. Some drivers are more averse toparticular junctions, bends and gradients than others, and, in the context of generalised cost, both thecost per kilometre and value of time will vary amongst vehicles and drivers.

5.11 NESA therefore allows for multiple routeing between each pair of zones, using a method based uponBurrell’s technique. Multiple routeing is achieved by allowing the user to specify the number of trees,n, built for each zone. The first round of tree building uses link journey costs as input by the user,however, in the second and subsequent rounds of tree building, the link journey costs are subject torandom variation from their original values. The randomly generated cost C' is obtained as:

Note: DMRB Volume 12.1.9 describes a method whereby the perturbation factor (P) can be estimatedfrom roadside interview origin and destination data.

5.12 Following n rounds of tree building, the multiple routeing procedure allocates 1/nth of the demand foreach O-D pair to each tree building round. The link flows accumulated from each round are thenfurther aggregated to produce the total modelled flows. In general, the trips assigned to a particularroute between two zones will not all be on their optimal routes, though if a particular route betweentwo zones is substantially better than the alternatives there is a strong possibility that most or even allthe flow between the zones will be assigned to that optimal route. This is quite reasonable, since driverswill be much more likely to select the same route if its advantages are obvious, than if it is onlymarginally better than an alternative.

5.13 If only one set of trees per origin is built and the perturbation factor is non-zero, then the link journeycosts will be randomly varied. However, if one set of trees per origin is built and a perturbation factorof zero is specified an All-Or-Nothing assignment will be produced.

5.14 Compared to a single All-Or-Nothing assignment, multiple routeing will naturally tend to increase themodelled estimates of total vehicle mileage and travel time in the study area. Because the results fromthe single and the multiple routeing assignments are used in exactly the same manner in the subsequentstage of economic evaluation it is important to ensure that the type of tree building used in the base anddesign networks is the same.

Multiple User Class Assignment

5.15 In some instances commuting, leisure, business and heavy goods trips may need different treatmentduring the assignment process. For example, it might be expected that business trips will be timesensitive (implying low distance and cost coefficients), whilst leisure trips will be more cost sensitive(implying high distance and cost coefficients).

Equation 5/5/2

where C = original link time (minutes)= randomised link time (minutes)

N = a random number (0 <= N <= 10)P = Perturbation Factor (P >= 0)

C C N 5– P500

--------------------- C+=

C

Chapter 5 Volume 15 Section 1Route Choice Part 5 Traffic Modelling in NESA


5.16 From the fifteen user classes defined in NESA (see Table 5/2/6), the user can specify up to four groupsof user classes for assignment purposes according to expected route choice behaviour. For example, thefollowing user groups could be specified:

1. Commuting - Car (user classes = 1, 2)2. Leisure - Car /Coach (user classes = 5, 6, 7, 8, 10, 11, 15)3. Business - Car/LGV (user classes = 3, 4, 9, 12)4. Heavy Goods - OGV1/OGV2 (user classes = 13, 14)

5.17 To model the behavioural differences between user groups, the coefficients in the generalised costfunction (Equation 5/5/1) and the perturbation factor in Equation 5/5/2 can be specified by user group(see Table 5/5/1). Within each user group, there may also be some variation in the reaction to out ofpocket costs, such as tolls, because, for example of different time constraints or income levels ofdrivers within that group. NESA allows the user to model this variation, by specifying the generalisedcost toll coefficient, c, for each tree built in a user group. Table 5/5/1 details some acceptable values fora, b, c and the perturbation factor, P, for all vehicles and the above four user groups (four trees beingbuilt per group).

Tolls

5.18 NESA can model the effect that road tolls may have on traffic flows. Users are advised to contact theTransport Scotland if this procedure is required.

Volume 15 Section 1 Chapter 6Part 5 Traffic Modelling in NESA Traffic Forecasts


6 TRAFFIC FORECASTSFixed/Variable Trip Matrices

6.1 In situations where the Fixed Trip Matrix assumption (see Part 3 Chapter 5) is deemed applicable (seeDMRB 5.1.4 SH1/97), future year traffic flows can be estimated in NESA by assigning the matrix tothe Do-Something trees. This is the standard NESA assignment procedure and is done automatically.

6.2 However, if analysis has shown that the FTM assumption is inappropriate (see DMRB 5.1.4 SH1/97)then a variable matrix assessment must be undertaken and Transport Scotland advice should be sought.

Standard Case

6.3 In the standard case, output from the NESA assignment model is used to produce future year trafficflow forecasts by applying external growth rates to the base year assigned flows. The future year trafficflows can then be used in the operational appraisal of the future year network (see DMRB Volume 12).

6.4 The underlying assumption behind the simple growth factor method of forecasting is that the pattern oftrip making and route choices remains fixed, and that all sectors of the trip matrix experience the samegrowth rate.

6.5 NESA outputs user requested traffic flows for the base year and any forecast year that coincides with afive year interval from the base year.

6.6 The base year for scheme forecasting should be the most recent year for which observed local trafficfigures are available. Base years, should not, however be more than 3 years earlier than the currentyear. If this is the case, local data should be collected to update the model.

Traffic Growth

6.7 Users should refer to the Overseeing Organisation’s guidance on economic assessment when selectingthe appropriate Traffic Growth. Local growth, or growth based on national forecasts, is generally used.In Scotland for trunk road schemes, Transport Scotland’s advice is to consider using TMfS as theprinciple source of growth forecasts; or alternatively Scottish Trip End Program (STEP); or NationalRoad Traffic Forecasts (NRTF97). Where NRTF97 forecasts are used, users should note that theseassume zero traffic growth post 2031.

National Road Traffic Forecasts (NRTF)

6.8 The default traffic growth factors in NESA are the National Road Traffic Forecasts (NRTF). A fullaccount of the forecasts and the basis of their derivation is contained in National Road Traffic Forecasts(Great Britain) (Department for Transport, 1997).

6.9 Traffic growth is subject to a great deal of uncertainty. This uncertainty is modelled through the use ofdifferent growth scenarios: High, Low and Central growth. These forecasts reflect differentassumptions concerning, for example economic growth, fuel prices, car ownership levels and goodsvehicle traffic growth.

6.10 Table 5/6/1 gives actual national road traffic growth from 1981 to 1996 and the NRTF for Low, Highand Central growth expressed as annual percentage growth rates for each year up to 2031. Zero growthis assumed post 2031.

Chapter 6 Volume 15 Section 1Traffic Forecasts Part 5 Traffic Modelling in NESA


Non - Standard Cases

6.11 In certain situations, the use of NRTF and the simple growth factor method of forecasting (i.e. fixed tripdistribution and fixed routes) may be deemed inappropriate, particularly where there are:

(i) local network constraints which will limit traffic growth;

(ii) future developments which will result in local trip end growth increasing at differential rateswithin the study area;

(iii) changes to the road network outwith the study area which will affect the number of tripspassing through the study area;

(iv) internal network changes which will alter the routeing of traffic;

(v) sections of network which become heavily congested and affect the routeing of traffic.

Table 5/6/1: National Road Traffic Forecasts (NRTF 1997) - Annual Percentage Growth Rates

PERIOD ACTUAL GROWTH (% per year)(see note) CARS LGV OGV1 OGV2 PSV

1981 2.1 1.1 -5.0 -2.2 -0.61982 3.6 -0.9 -3.7 0.9 0.61983 1.7 -0.1 1.8 3.3 5.11984 5.5 5.7 4.5 3.6 4.31985 2.6 2.8 1.4 2.3 -5.21986 5.6 5.3 1.7 4.2 0.91987 7.6 9.4 7.5 15.4 10.51988 7.3 10.6 5.8 9.2 5.71989 8.5 10.4 4.6 9.7 5.01990 1.4 0.8 -1.9 -2.3 1.01991 -0.2 4.3 0.4 -1.8 4.51992 0.8 -1.3 -1.4 -4.1 -3.81993 0.2 -0.7 0.2 1.3 0.31994 2.1 4.9 2.1 7.0 2.31995 2.2 2.2 -0.2 2.3 -1.31996 2.6 3.2 0.1 7.1 3.6

LOW GROWTH FORECASTS (% PER YEAR)RANGE OF

YEARS CARS LGV OGV1 OGV2 PSV

1997 to 2001 0.56 1.77 -0.35 1.53 -0.392002 to 2006 1.31 1.93 0.43 2.21 0.342007 to 2011 1.11 1.82 0.46 2.09 0.342012 to 2016 1.01 1.92 0.55 2.15 0.412017 to 2021 0.64 1.83 0.57 2.01 0.502022 to 2026 0.30 1.67 0.60 1.89 0.592027 to 2031 0.28 1.49 0.60 1.73 0.662032 onwards 0.00 0.00 0.00 0.00 0.00

HIGH GROWTH FORECASTS (% PER YEAR)RANGE OF


Volume 15 Section 1 Chapter 6Part 5 Traffic Modelling in NESA Traffic Forecasts


Note: The 1991 growth factor is that used to convert a 1990 flow to a 1991 flow.

6.12 In NESA there are a number of facilities which can be used to address the above situations. Theseinclude the use of local growth factors (for case (i)), future year matrices (for cases (i), (ii) and (iii)) anda reassignment facility (for cases (iv) and (v)).

6.13 NESA allows the user to input local annual growth rates by vehicle type in a format consistent withTable 5/6/1. Local growth rates are suitable for a situation where local growth is expected to bedifferent from national growth or a situation where local network constraints are expected to limittraffic growth. These growth rates may be derived from a variety of sources including local planningdata, a higher tier traffic model such as the Transport Model for Scotland (TMfS) or the Scottish TripEnd Program (STEP). Although NRTF (1997) allows forecasts to be disaggregated by road type, areatype and time period, these should not be used without prior consultation with Transport Scotland.

6.14 Future year matrices are an assignment and evaluation tool that allows NESA to take account of localtraffic growth that departs from NRTF. NESA offers the user the facility of up to five future yearmatrices to cover differential growth during the evaluation period. Each of these matrices is assigned tothe same routes as used by the base year traffic. Future year matrices can be input in any year from thebase year. For each year that future year matrices are input an assignment is output.

6.15 The future year matrices should be consistent with the assumptions underlying NRTF for low, centraland high growth (see Table 5/6/1). Part 10 Chapter 5 provides further details of the operational aspectsof the future year matrix facility.

6.16 If changes to the internal road network occur during the evaluation period, these can be handled inNESA by a process known as reassignment. This process involves assigning the base year trip matricesto modified Do-Minimum and Do-Something networks representing the latter phase of the evaluationperiod. The user undertakes assignments to the modified Do-Minimum and Do-Something networks inthe standard way then simply specifies the future year from which the updated assignments apply. Part10 Chapter 14 provides further details of the operational aspects of the reassignment option.

1997 to 2001 2.69 3.93 1.76 3.68 1.722002 to 2006 1.95 2.58 1.07 2.85 0.982007 to 2011 1.76 2.48 1.10 2.74 0.992012 to 2016 1.67 2.59 1.20 2.82 1.072017 to 2021 1.31 2.51 1.24 2.69 1.172022 to 2026 0.98 2.36 1.29 2.58 1.272027 to 2031 0.97 2.19 1.30 2.44 1.362032 onwards 0.00 0.00 0.00 0.00 0.00

CENTRAL GROWTH FORECASTS (% PER YEAR)RANGE OF


1997 to 2001 1.65 2.87 0.72 2.63 0.682002 to 2006 1.65 2.27 0.77 2.55 0.682007 to 2011 1.46 2.17 0.80 2.44 0.692012 to 2016 1.37 2.29 0.91 2.52 0.772017 to 2021 1.01 2.21 0.94 2.39 0.872022 to 2026 0.69 2.06 0.99 2.27 0.972027 to 2031 0.67 1.89 1.00 2.13 1.062032 onwards 0.00 0.00 0.00 0.00 0.00

Table 5/6/1: (Continued)National Road Traffic Forecasts (NRTF 1997) - Annual Percentage Growth

Chapter 6 Volume 15 Section 1Traffic Forecasts Part 5 Traffic Modelling in NESA


6.17 While the reassignment facility could also be used to reflect routeing changes due to increasingcongestion over sections of the internal network, such circumstances would be more accuratelyrepresented through the application of a congested assignment model.



VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


VALUATION OF COSTS AND BENEFITS

Contents

Chapter

1. Introduction

2. The Valuation of Time Savings

3. The Valuation of Vehicle Operating Costs

4. The Valuation of Accidents

5. The Valuation of Accidents on Links

6. The Valuation of Accidents at Junctions

7. Carbon Emissions

8. Construction Costs

9. The Preparation of Scheme Cost Data forUse in NESA

10. An Example of Scheme Cost Inputs

11. Road Maintenance

12. Delays During Construction

PART 6

Volume 15 Section 1Part 6 Valuation of Costs and Benefits


Volume 15 Section 1 Chapter 1Part 6 Valuation of Costs and Benefits Introduction


1 INTRODUCTION1.1 A basic principle of cost-benefit analysis is that a comparison is made between the sum of the benefits

generated by a road improvement and the sum of the costs incurred (see Part 3 Chapter 1).

1.2 To facilitate this comparison NESA converts all costs and benefits to monetary units. This monetaryvaluation process is based on the willingness to pay principle and is an important feature ofcost-benefit analysis as practiced in the UK (see Part 3 Chapter 1). Valuing costs and benefits using thisprinciple places practical limitations on the types of cost and benefit that can be incorporated within theeconomic assessment. NESA therefore only considers the following cost types:

• travel time costs (see Part 6 Chapter 2)

• vehicle operating costs (see Part 6 Chapter 3)

• accident costs (see Part 6 Chapters 4, 5 and 6)

• carbon costs (see Part 6 Chapter 7)

• construction costs (see Part 6 Chapters 8, 9 and 10)

• maintenance costs (see Part 6 Chapters 11 and 12)

1.3 Travel time, vehicle operating costs, accident costs and carbon costs are incurred by the road user andare consequently known as user costs. A scheme’s Present Value of Benefit (PVB) is defined as thechange in the discounted value of the user costs between the Do-Minimum and Do-Somethingnetworks (see Part 3 Chapter 2). The capital costs of construction and routine maintenance are incurredby society as a whole and are incorporated into the Present Value of Costs for the scheme (see Part 3Chapter 1). User costs incurred during construction and routine maintenance (e.g. travel time delaysand accidents) can be calculated using the Highways Agency program QUADRO.

RESOURCE COSTS

1.4 To avoid the distorting effects of taxation and government subsidies all the costs used in NESA andpresented in this chapter are valued at their resource value (see Part 3 Chapter 6).

The NESA Method

1.5 NESA uses a link based method to calculate the network wide user costs on the Do-Minimum andDo-Something networks for every year of the evaluation period (see Figure 6/1/1).

1.6 Using the assigned link flows and turning movements output by the traffic assignment model by flowgroup and user class (see Part 5), NESA calculates link travel times, junction delays and the number ofaccidents for each link/junction. The methodology used to do this is outlined in Part 7 (for link traveltimes), Part 8 (for junction delays) and Part 6 Chapters 5 and 6 (for the number of accidents).

1.7 By applying values of time, accident costs and the vehicle operating cost formulae to the evaluated linktravel times, junction delays and the number of accidents, NESA estimates the user costs by flow groupfor each year of the evaluation period.

Chapter 1 Volume 15 Section 1Introduction Part 6 Valuation of Costs and Benefits


Figure 6/1/1: Calculation of User Costs by Flow Group and Forecast Year from Assigned Link Flows

Volume 15 Section 1 Chapter 2Part 6 Valuation of Costs and Benefits The Valuation of Time Savings


2 THE VALUATION OF TIME SAVINGS2.1 Changes in the time taken by traffic to pass along links and to pass through junctions in the NESA

network are the major items of the calculated benefit resulting from a road improvement. It is necessaryto put a money value on time savings in order to compare these with construction costs and accidentand vehicle operating cost savings.

2.2 Three distinct purposes of travel are distinguished in NESA; travel in the course of Work, Commuting(travel to and from normal place of work) and Other (travel for other non-work purposes).

2.3 Travel in the course of work (working time) is valued at its cost to the employer of the travellingemployee, on the grounds that the value of the output produced in working time must be at least equalto the cost to the employer of hiring labour for that time. This assumes that all savings in working timecan be used for the production of output by the employee, and that the value of this output is measuredby the cost to the employer. This is the resource value of the time savings used in NESA, and is alsotaken to be the behavioural value perceived by the employee. The cost to the employer of an hour’sworking time is given by the gross wage rate, increased by National Insurance and pensioncontributions and by overheads which vary according to the type of labour employed. At present, anaddition of 21.2% is made to the gross wage rate to cover these items. The basic sources of wage ratedata are New Earnings Survey of the Office for National Statistics, National Travel Survey (NTS) ofthe Department for Transport and the 2000 labour cost survey.

2.4 Values of working time are estimated for different types of vehicle occupant and are given in Table 6/2/1. PSVs include coaches; LGVs and OGVs are defined in Part 5 Chapter 2. OGV values apply toOGV1 and OGV2 classes as defined in Part 5 Chapter 2. For car drivers and passengers and buspassengers the value of working time is based on mileage-weighted incomes of a sample of workers totake account of the variation in mileage travelled by workers with different earnings. The data sourcefor mileage weighting is the National Travel Survey.

2.5 There are several other assumptions implicit in the valuation of working time as described above. Thevaluation uses average pay, rather than marginal pay at overtime rates. It is assumed that in the long runemployers will tend to adjust to a saving of working time by increasing output, with the level ofemployment being maintained, rather than reducing either employment or the number of workinghours per employee. It is generally assumed that an employee when travelling always acts in the bestinterest of his employer rather than that, for example, the employee uses a saving in travelling time tohave a longer meal break.

2.6 Another assumption is that all travel in course of work is carried out in the employer’s time rather thanin the employee’s time so that time savings are thus working rather than non-working time savings. Afinal assumption for travellers who are not transport workers is that no productive work is possibleduring travel.

2.7 Commuting is for journeys made to and from the normal place of work and Other non-workembraces all purposes of travel except travel in the course of work and commuting. There is no directmarket in which non-working time be bought and sold in the same way that an employer can buy theworking time of an employee. A more indirect approach is thus necessary for the problem of valuingnon-working time.

2.8 The value of non-working purposes used in NESA has been derived from studies of how people chooseto travel when faced with a choice between a slow, cheap mode and a fast, expensive mode or betweena short, expensive car route (such as over a tolled bridge) and a long, cheaper car route. If there are noother significant factors to consider, such as relative degree of comfort, then if a person chooses to payan extra X pence to save Y minutes, he is revealing an implicit valuation of his time of at least X/Y

Chapter 2 Volume 15 Section 1The Valuation of Time Savings Part 6 Valuation of Costs and Benefits


pence per minute. If he does not pay X pence to save Y minutes, his value of time is less than X/Ypence per minute.

2.9 The unit value given to time savings does not vary with the length of those savings. Thus a saving of 1minute for a particular journey purpose has the same unit value as a saving of 30 minutes.

2.10 This practice was endorsed by the Advisory Committee on Trunk Road Assessment. Arguments in itssupport are:

(i) empirical studies have shown that people place the same value on time savings ofbetween 3 and 30 minutes (it is generally impossible to test the value of savings ofless than 3 minutes).

(ii) empirical evidence also suggests that time savings of a given length have adistribution of values. The practice of assuming that all time savings have the one unitvalue therefore only implies that the mean of the distribution (the average value perunit of time) is equal to the one unit value used. This value is the same for any lengthof time saving (see (i) above).

(iii) development of the national road network over time produces a series of smallsavings which when summed together have a large positive value. It is inconsistentwith the cost-benefit analysis approach to forego an eventual large time saving byrefusing to undertake schemes yielding the small time savings.

2.11 The NESA program has fifteen user classes (see Part 5, Table 5/2/5) each of which can be categorisedas either working or non-working (see Paragraph 2.7). All goods vehicles and their occupants areappraised in the working time mode only.

2.12 Table 6/2/1 shows the default average vehicle occupancies in NESA which have been derived from theNational Travel Survey (NTS) - see also TAG data book Table A 1.3.1. In combination with data fromother sources it has also been possible to estimate the proportion of light goods vehicle travel not inworking time (but not split between Commuting and Other non-work). Combining the values of timeand occupancies produces values of time per vehicle (based on 2000 national average occupancies).

2.13 Trends identified in the NTS indicate that the occupancy of non-working cars is reducing as incomerises and more car passengers become car drivers. Similarly, the occupancy of working cars is alsoreducing, albeit to a lesser extent. These trends are expected to continue for as long as predicted trafficgrowth, that is 2031, and are subsequently taken into account when calculating link transit costs.



2.14 Taking the elasticity of occupancy with respect to car ownership as the best way of forecasting thepredicted decline in car occupancies, the expected growth rates in car occupancies were derived and arereproduced in Table 6/2/2. The occupancy of all other vehicle types should be assumed to remainunchanged over time.

2.15 Average vehicle occupancies are combined with the values of working time by type of vehicleoccupant and with the value of in-vehicle non-working time to produce values of time per vehicle, inpence per hour in 2010 values and resource prices (see Table 6/2/1).

2.16 The National Travel Survey (NTS) indicates how the proportion of car mileage in work time varies byhour of the week. The NESA default car-in-work time (CWT) proportion is 0.131,car-in-commuting-time proportion is 0.253 and LGV-in-non-work-time proportion is 0.120. NESA

Table 6/2/1: Values of Time per Person and per Vehicle in NESA (2010 prices and values)

Type of Vehicleand Purpose

Weekly Average Occupancy

Occupant Purpose Resource Cost (pence/hour)per occupant

Working Car 1.00 driver Working 22740.20 passengers Working 1725

Non-Working CarCommuting 1.00 driver Commuting 572

0.14 passengers Commuting 572Other 1.00 driver Other Non-Work 508

0.85 passengers Other Non-Work 508Working Light Goods Vehicle (LGV)

1.00 driver Working 10240.20 passengers Working 1024

Non-Working Light Goods Vehicle (LGV)

Commuting 1.00 driver Commuting 5720.59 passengers Commuting 572

Other 1.00 driver Other Non-Work 5080.59 passengers Other Non-Work 508

Other Goods Vehicle (OGV1 & OGV2)

1.00 drivers Working 1206

Public Service Vehicle (PSV)

1.00 driver Working 12320.35 passengers Working 13972.50 passengers Commuting 5729.35 passengers Other 508

Table 6/2/2: Compound Annual Rates of Change in Car Occupancies (%)

Low and High Traffic ForecastRange of Years Working Cars Non-Working Cars

(Commuting & Other)(% pa) (% pa)

2002 - 2036 -0.45 -0.562036 onwards 0.00 0.00



offers the user the facility to input local car-in-work time proportions by flow group where there isstatistically reliable data.

2.17 The real value of average employee earnings is assumed to reflect the growth in the real value of bothworking time and non-working time. Future growth is expected to grow in line with real GDP per head.Forecasts of growth in the real value of time are given in Table 6/2/3. In accordance with the TreasuryGreen Book guidance these growth rates are decreased after thirty years when the discount rate isreduced. The reduction is the same proportional reduction as the change in the discount rate.

Table 6/2/3: Assumed Compound Annual Rates of Growth in the Real Value of Time (%)

Low and High Economic Forecast (%pa)Range of Years Working Time Non-Working Time

(Commuting & Other)

2002 - 2003 3.83 3.832003 - 2004 1.92 1.922004 - 2005 2.02 2.022005 - 2006 2.34 2.342006 - 2007 1.73 1.732007 - 2008 -1.15 -1.152008 - 2009 -4.98 -4.982009 - 2010 1.10 1.102010 - 2011 0.08 0.082011 - 2012 0.00 0.002012 - 2013 1.09 1.092013 - 2014 2.05 2.052014 - 2015 1.67 1.672015 - 2016 1.95 1.952016 - 2017 1.99 1.992017 - 2018 1.90 1.902018 - 2019 1.91 1.912019 - 2020 1.90 1.902020 - 2021 1.88 1.882021 - 2022 1.87 1.872022 - 2023 1.89 1.892023 - 2024 1.90 1.902024 - 2025 1.92 1.922025 - 2026 1.94 1.942026 - 2027 1.95 1.952027 - 2028 1.97 1.972028 - 2029 1.99 1.992029 - 2030 2.01 2.012030 - 2031 2.02 2.022031 - 2032 2.04 2.042032 - 2033 2.05 2.052033 - 2034 2.06 2.062034 - 2035 2.07 2.072035 - 2036 2.08 2.082036 - 2037 2.09 2.092037 - 2038 2.10 2.10



2038 - 2039 2.10 2.102039 - 2040 2.10 2.102040 - 2041 2.10 2.102041 - 2042 2.12 2.122042 - 2043 2.12 2.122043 - 2044 2.12 2.122044 - 2045 2.12 2.122045 - 2046 2.12 2.122046 - 2047 2.15 2.152047 - 2048 2.15 2.152048 - 2049 2.15 2.152049 - 2050 2.15 2.152050 - 2051 2.15 2.152051 - 2052 2.19 2.192052 - 2053 2.19 2.192053 - 2054 2.19 2.192054 - 2055 2.19 2.192055 - 2056 2.19 2.192056 - 2057 2.21 2.212057 - 2058 2.21 2.212058 - 2059 2.21 2.212059 - 2060 2.21 2.212060-2061 2.21 2.212061-2062 2.22 2.222062-2063 2.21 2.212063-2064 2.21 2.212064-2065 2.21 2.212065-2066 2.21 2.212066-2067 2.20 2.202067-2068 2.20 2.202068-2069 2.20 2.202069-2070 2.20 2.202070-2071 2.20 2.202071-2072 2.17 2.172072-2073 2.17 2.172073-2074 2.17 2.172074-2075 2.17 2.172075-2076 2.17 2.172076-2077 2.17 2.172077-2078 2.17 2.172078-2079 2.17 2.172079-2080 2.17 2.172080-2081 2.17 2.172081-2082 2.17 2.172082-2083 2.17 2.172083-2084 2.17 2.172084-2085 2.17 2.172085-2086 2.17 2.17




2.18 These assumptions relate to long term forecasts only, and undue weight should not be given to shortterm fluctuations.

Conversion from Resource Costs to Market Prices

2.19 NESA works in resource costs which have to be converted to market prices to be consistent with theTransport Economic Efficiency (TEE) process, see Scottish Transport Appraisal Guidance (STAG)Chapter 8. The market price of time is obtained by multiplying the resource value by (1 + t) where t isthe average rate of indirect taxation in the economy; in 2010 this was 19%.

2086-2087 2.17 2.172087-2088 2.18 2.182088-2089 2.18 2.182089-2090 2.18 2.182090-2091 2.18 2.182091-2092 2.18 2.182092-2093 2.17 2.172093-2094 2.17 2.172094-2095 2.17 2.172095-2096 2.17 2.172096-2097 2.17 2.172097-2098 2.17 2.172098-2099 2.17 2.172099-2100 2.17 2.17


Volume 15 Section 1 Chapter 3Part 6 Valuation of Costs and Benefits The Valuation of Vehicle Operating Costs


3 THE VALUATION OF VEHICLE OPERATING COSTS

3.1 Differences in the vehicle operating costs (VOC) incurred by traffic using the Do-Something roadnetwork compared to the VOC incurred by traffic using the Do-Minimum network are recorded amongthe benefits resulting from a road improvement.

3.2 The change in total VOC over all links depends on changes in the distance travelled by vehicles and onaverage link speeds. In most schemes the aggregate time saving is positive, as is the overall saving incosts of accidents, but the change in overall VOC can be either negative or positive depending on thebalance of changes in distance travelled and speeds.

3.3 VOC in NESA comprises six items: fuel, oil, tyres, maintenance, depreciation, and size of vehiclefleets. Only items which vary with the use of the vehicle are measured so, for example, vehicle exciseduty, insurance and garaging are excluded from VOC. More precisely, it is the mileage-related resourceVOC that is included in NESA. All resource costs exclude any element of indirect taxation asexplained in Part 3 Chapter 6.

3.4 The fuel consumption is estimated by vehicle category and fuel type using a function of the form:

This function gives a higher consumption at low speeds, reflecting the effects of stop-start motoring incongested conditions.Fuel costs are subsequently derived using the costs presented in Table 6/3/2.

3.5 The non-fuel elements of the vehicle operating costs are combined in a formula of the form:

3.6 The marginal resource costs of oil, tyres, mileage and maintenance related depreciation are assumed tobe fixed costs per kilometre and appear in the a1 term. The b1 term in the non-fuel costs representschanges in the productivity of commercial vehicles and cars in working time, all goods vehicles andPSVs.

3.7 The time component of depreciation is excluded since it does not vary with distance or speed. ForOGVs and PSVs depreciation is assumed to be totally time related; this is based on evidence from tradesources which suggests that factors such as obsolescence and condition are more importantdeterminants of vehicle value than mileage per se. For cars and LGVs evidence from second hand

Equation 6/3/1

where L = consumption in litres per kilometre per vehicle

V = average link speed in kilometres per houra, b, c and d are parameters defined for each vehicle category

(see Table 6/3/1)

Equation 6/3/2

where C = cost in pence per kilometre per vehicleV = average link speed in kilometres per houra1, b1 are parameters defined for each vehicle

category (see Table 6/3/1)

L aV---- b cV dV2+ + +=

C a1 b1/V+=

Chapter 3 Volume 15 Section 1The Valuation of Vehicle Operating Costs Part 6 Valuation of Costs and Benefits


prices indicates that part of their depreciation is related to mileage; therefore this element is recorded asa marginal resource cost.

3.8 There is no specific allowance in NESA for fuel used at junctions.

3.9 The VOC formulae parameter values in 2010 prices and values by type of vehicle and type of fuel(defined in Part 5 Chapter 2) are given in Table 6/3/1.

3.10 The annual rates in fuel resource costs, fuel duty and VAT are given in Table 6/3/2. They are the samefor both low and high economic forecasts. The vehicle fleet using either petrol or diesel is shown inTable 6/3/3.

Table 6/3/1: VOC Formulae Parameter Values (2010 prices and values)

Vehicle Category

Fuel (litres/km, 2010 prices)a b c d

Petrol Car 1.119233393 0.044004770 -0.000081383 0.00000244908Diesel Car 0.492145560 0.062181967 -0.000590984 0.00000464689Petrol LGV 1.950832769 0.034527979 0.000067987 0.00000371490Diesel LGV 1.396883496 0.033477400 -0.000229978 0.00000767320

OGV1 1.431445529 0.258021379 -0.003906637 0.00003362306OGV2 2.670111055 0.557155643 -0.007976139 0.00006003527PSV 5.980054953 0.245278327 -0.003064986 0.00003061478

Vehicle Category

Energy Consumption (kWh/km, 2011 prices)a b c d

Electric Car 0.12564236

Vehicle Category

Non-Fuel (pence/km,2010 prices)a1 (pence/km) b1 (pence/hr)

Work Car 4.966 135.946Non-work Car 3.846 0.000Work Electric 1.157 135.946

Non-work Electric 1.157 0.000Work LGV 7.213 47.113

Non-work LGV 7.213 0.000OGV1 6.714 263.817OGV2 13.061 508.525PSV 30.461 694.547

Table 6/3/2: Fuel Resource Costs, Fuel Duty and VAT Rates (2010 prices and values)

Year Resource Cost Duty VATPetrol Diesel Electricity Petrol Diesel Electricity Standard Electric

(p/litre) (p/litre) (p/KWh) (p/litre) (p/litre) (p/KWh) % %2010 42.82 44.57 11.88 57.53 57.53 0.00 17.5 5.02011 52.21 56.43 12.45 57.22 57.22 0.00 20.0 5.02012 53.24 58.23 13.02 56.02 56.02 0.00 20.0 5.02013 51.50 56.19 13.74 55.13 55.13 0.00 20.0 5.02014 47.87 53.18 14.27 53.89 53.89 0.00 20.0 5.02015 44.65 49.49 14.03 53.50 53.50 0.00 20.0 5.02016 43.19 47.82 16.02 56.12 56.12 0.00 20.0 5.0



2017 42.74 47.30 15.90 55.15 55.15 0.00 20.0 5.02018 43.19 47.82 16.02 56.12 56.12 0.00 20.0 5.02019 43.60 48.29 17.05 57.16 57.16 0.00 20.0 5.02020 44.57 49.40 17.05 58.00 58.00 0.00 20.0 5.02021 45.40 50.35 17.67 58.74 58.74 0.00 20.0 5.02022 46.26 51.33 17.98 59.37 59.37 0.00 20.0 5.02023 47.12 52.32 17.97 60.01 60.01 0.00 20.0 5.02024 47.98 53.31 18.34 60.66 60.66 0.00 20.0 5.02025 48.88 54.33 18.82 61.31 61.31 0.00 20.0 5.02026 49.82 55.41 19.24 61.97 61.97 0.00 20.0 5.02027 50.75 56.48 19.09 62.64 62.64 0.00 20.0 5.02028 51.73 57.59 19.08 63.31 63.31 0.00 20.0 5.02029 52.70 58.71 18.98 63.99 63.99 0.00 20.0 5.02030 53.71 59.87 19.20 64.68 64.68 0.00 20.0 5.02031 54.76 61.07 19.20 65.38 65.38 0.00 20.0 5.02032 55.81 62.27 19.19 66.08 66.08 0.00 20.0 5.02033 56.86 63.47 19.17 66.79 66.79 0.00 20.0 5.02034 57.98 64.76 19.15 67.51 67.51 0.00 20.0 5.02035 59.11 66.04 19.13 68.24 68.24 0.00 20.0 5.02036 59.11 66.04 19.10 68.97 68.97 0.00 20.0 5.02037 59.11 66.04 19.08 69.72 69.72 0.00 20.0 5.02038 59.11 66.04 19.05 70.47 70.47 0.00 20.0 5.02039 59.11 66.04 19.02 71.23 71.23 0.00 20.0 5.02040 59.11 66.04 18.99 71.99 71.99 0.00 20.0 5.02041 59.11 66.04 18.94 72.77 72.77 0.00 20.0 5.02042 59.11 66.04 18.95 73.55 73.55 0.00 20.0 5.02043 59.11 66.04 18.90 74.34 74.34 0.00 20.0 5.02044 59.11 66.04 18.86 75.14 75.14 0.00 20.0 5.02045 59.11 66.04 18.90 75.95 75.95 0.00 20.0 5.02046 59.11 66.04 18.88 76.77 76.77 0.00 20.0 5.02047 59.11 66.04 18.84 77.59 77.59 0.00 20.0 5.02048 59.11 66.04 18.97 78.43 78.43 0.00 20.0 5.02049 59.11 66.04 18.91 79.27 79.27 0.00 20.0 5.02050 59.11 66.04 18.93 80.13 80.13 0.00 20.0 5.02051 59.11 66.04 18.96 80.99 80.99 0.00 20.0 5.02052 59.11 66.04 18.98 81.86 81.86 0.00 20.0 5.02053 59.11 66.04 19.00 82.74 82.74 0.00 20.0 5.02054 59.11 66.04 19.03 83.63 83.63 0.00 20.0 5.02055 59.11 66.04 19.05 84.53 84.53 0.00 20.0 5.02056 59.11 66.04 19.07 85.44 85.44 0.00 20.0 5.02057 59.11 66.04 19.09 86.36 86.36 0.00 20.0 5.02058 59.11 66.04 19.11 87.29 87.29 0.00 20.0 5.02059 59.11 66.04 19.14 88.23 88.23 0.00 20.0 5.02060 59.11 66.04 19.16 89.18 89.18 0.00 20.0 5.02061 59.11 66.04 19.17 90.14 90.14 0.00 20.0 5.02062 59.11 66.04 19.19 91.11 91.11 0.00 20.0 5.02063 59.11 66.04 19.20 92.09 92.09 0.00 20.0 5.02064 59.11 66.04 19.22 93.08 93.08 0.00 20.0 5.0




2065 59.11 66.04 19.23 94.08 94.08 0.00 20.0 5.02066 59.11 66.04 19.24 95.10 95.10 0.00 20.0 5.02067 59.11 66.04 19.25 96.12 96.12 0.00 20.0 5.02068 59.11 66.04 19.26 97.16 97.16 0.00 20.0 5.02069 59.11 66.04 19.27 98.20 98.20 0.00 20.0 5.02070 59.11 66.04 19.27 99.26 99.26 0.00 20.0 5.02071 59.11 66.04 19.28 100.33 100.33 0.00 20.0 5.02072 59.11 66.04 19.28 101.41 101.41 0.00 20.0 5.02073 59.11 66.04 19.29 102.50 102.50 0.00 20.0 5.02074 59.11 66.04 19.29 103.60 103.60 0.00 20.0 5.02075 59.11 66.04 19.29 104.72 104.72 0.00 20.0 5.02076 59.11 66.04 19.29 105.84 105.84 0.00 20.0 5.02077 59.11 66.04 19.29 106.98 106.98 0.00 20.0 5.02078 59.11 66.04 19.29 108.13 108.13 0.00 20.0 5.02079 59.11 66.04 19.29 109.30 109.30 0.00 20.0 5.02080 59.11 66.04 19.29 110.47 110.47 0.00 20.0 5.02081 59.11 66.04 19.29 111.66 111.66 0.00 20.0 5.02082 59.11 66.04 19.29 112.87 112.87 0.00 20.0 5.02083 59.11 66.04 19.28 114.08 114.08 0.00 20.0 5.02084 59.11 66.04 19.28 115.31 115.31 0.00 20.0 5.02085 59.11 66.04 19.28 116.55 116.55 0.00 20.0 5.02086 59.11 66.04 19.27 117.80 117.80 0.00 20.0 5.02087 59.11 66.04 19.26 119.07 119.07 0.00 20.0 5.02088 59.11 66.04 19.26 120.35 120.35 0.00 20.0 5.02089 59.11 66.04 19.25 121.65 121.65 0.00 20.0 5.02090 59.11 66.04 19.24 122.96 122.96 0.00 20.0 5.02091 59.11 66.04 19.24 124.28 124.28 0.00 20.0 5.02092 59.11 66.04 19.23 125.62 125.62 0.00 20.0 5.02093 59.11 66.04 19.22 126.97 126.97 0.00 20.0 5.02094 59.11 66.04 19.21 128.34 128.34 0.00 20.0 5.02095 59.11 66.04 19.21 129.72 129.72 0.00 20.0 5.02096 59.11 66.04 19.20 131.11 131.11 0.00 20.0 5.02097 59.11 66.04 19.19 132.53 132.53 0.00 20.0 5.02098 59.11 66.04 19.18 133.95 133.95 0.00 20.0 5.02099 59.11 66.04 19.17 135.39 135.39 0.00 20.0 5.0

2100 onwards 59.11 66.04 19.16 136.84 136.84 0.00 20.0 5.0

Table 6/3/3: Proportion of Cars and LGVs Using Petrol/Diesel/Electric by Vehicle Kms (%)

Year Cars LGVsPetrol Diesel Electric Petrol Diesel

2004 73.280 26.720 0.000 11.070 88.9302005 70.945 29.055 0.000 10.202 89.7982006 68.610 31.390 0.000 9.333 90.6672007 66.275 33.725 0.000 8.465 91.5352008 63.940 36.060 0.000 7.597 92.4032009 61.605 38.395 0.000 6.728 93.2722010 59.270 40.730 0.000 5.860 94.140




3.11 The change in vehicle efficiency over time is shown in Table 6/3/4.

3.12 Table A 1.3.10, TAG data book, November 2014 only presents vehicle efficiency changes to 2035.NESA assumes that no further changes occur post 2035.

3.13 NESA adjusts the fuel consumption in accordance with the changes in vehicle efficiency presented inTable 6/3/4. NESA then applies the appropriate fuel cost from Table 6/3/2 after taking into account thepetrol/diesel/electric split presented in Table 6/3/3.

2011 57.010 42.958 0.032 5.416 94.5842012 54.750 45.186 0.064 4.972 95.0282013 52.490 47.414 0.096 4.528 95.4722014 50.230 49.642 0.128 4.084 95.9162015 47.970 51.870 0.160 3.640 96.3602016 47.116 52.564 0.320 3.290 96.7102017 46.262 53.258 0.480 2.940 97.0602018 45.408 53.952 0.640 2.590 97.4102019 44.554 54.646 0.800 2.240 97.7602020 43.700 55.340 0.960 1.890 98.1102021 43.842 54.882 1.276 1.720 98.2802022 43.984 54.424 1.592 1.550 98.4502023 44.126 53.966 1.908 1.380 98.6202024 44.268 53.508 2.224 1.210 98.7902025 44.410 53.050 2.540 1.040 98.9602026 44.420 52.486 3.094 0.990 99.0102027 44.430 51.922 3.648 0.940 99.0602028 44.440 51.398 4.202 0.890 99.1102029 44.450 50.794 4.756 0.840 99.1602030 44.460 50.230 5.310 0.790 99.210

2031 onwards 44.460 50.230 5.310 0.790 99.210

Table 6/3/4: Change in Vehicle Efficiency (%PA)

Range of Years Cars LGVs OGV1 OGV2 PSVPetrol Diesel Electric Petrol Diesel

2006 - 2007 -0.42 -0.49 0.00 -0.01 0.00 -1.23 -1.23 0.002007 - 2008 -1.05 -1.07 0.00 -0.01 0.00 -1.23 -1.23 0.002008 - 2009 -1.78 -0.92 0.00 -1.35 -1.23 -1.23 -1.23 0.002009 - 2010 -1.43 -1.63 0.00 -0.34 -1.80 -1.23 -1.23 0.002010 - 2011 -1.81 -2.23 0.00 -0.11 -2.71 0.00 0.00 0.002011 - 2015 -1.81 -2.23 0.10 -0.11 -2.71 0.00 0.00 0.002015 - 2020 -3.32 -2.22 -0.02 -2.35 -2.35 0.00 0.00 0.002020 - 2025 -3.16 -2.02 -0.12 -2.85 -1.65 0.00 0.00 0.002025 - 2030 -1.56 -1.19 0.00 -2.40 -0.74 0.00 0.00 0.002030 - 2035 -0.57 -0.51 0.08 -0.54 -0.11 0.00 0.00 0.002035 - 2100 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Table 6/3/3: Proportion of Cars and LGVs Using Petrol/Diesel/Electric by Vehicle Kms (%)



Conversion from Resource Costs to Market Prices

3.14 NESA works in resource costs which need to be converted to market prices for inclusion in theTransport Economic Efficiency (TEE) process, see Scottish Transport Appraisal Guidance (STAG)Chapter 8. See Table 9/4/1 in Part 9 Chapter 4 for further detail of conversion to market prices.

Volume 15 Section 1 Chapter 4Part 6 Valuation of Costs and Benefits The Valuation of Accidents


4 THE VALUATION OF ACCIDENTS4.1 The benefits from a reduction in the number and severity of accidents constitute an important element

in the economic appraisal of trunk road schemes. It is necessary to put a money value on accidentsavings so that they are given an appropriate valuation relative to that given to construction costs and totime and vehicle operating cost savings. Although average accident costs in Scotland are generallyhigher than for Great Britain as a whole, it is not possible to provide Scottish accident costs at thedegree of disaggregation required for NESA. The costs used are therefore equivalent to those inCOBALT. However, adjustments to the accident cost input to NESA may be made if there is substantialstatistical evidence to support it. Transport Scotland approval must, however, be sought. The basis foraccident cost valuation is given in the TAG data book, November 2014.

4.2 Table 6/4/1 details the components of accident costs which in addition to the casualty cost includecosts for police and administration, damage to property and an allowance for damage only accidents.

4.3 In NESA the average accident severity split (i.e. the number of fatal, serious and slight casualties peraccident) for a particular road/junction category is used to determine overall accident costs.Information on the severity split by class of road (links and junctions) is given in Part 6 Chapter 5 forlinks and Part 6 Chapter 6 for junctions, incorporating differential severity splits. Note that accidentsare classified according to the most seriously injured casualty and that for accident coding purposesrural roads are defined as those with a speed limit of more than 40 mph (64 kph). Those with speedlimits of 40 mph or below are defined as urban roads.

4.4 Details of the average proportion of fatal/serious/slight accidents on links is given in Table 6/4/2 and atjunctions in Table 6/4/3. The proportions are based on 1999-2001 data. These proportions are not basedon the most recent COBALT accident proportions (2009, TAG data book, November 2014) but rather,

Table 6/4/1: Components of Accident Costs (2010 prices and values)

COST PER CASUALTY, £Fatal casualty 1,640,134

Serious casualty 184,305Slight casualty 14,208

COST PER ACCIDENT, £Insurance Damage to Property

Administration Urban Rural MotorwayFatal accident 301 7,842 13,301 16,919

Serious accident 187 4,203 6,064 14,437Slight accident 114 2,479 4,019 7,304Damage only 54 1,773 2,651 2,547

Police CostUrban Rural Motorway

Fatal accident 17,012 17,469 17,673Serious accident 1,878 2,345 2,477Slight accident 486 666 556Damage only 36 20 17

Number of damage only accidents per PIA 17.7 7.8 7.6

Chapter 4 Volume 15 Section 1The Valuation of Accidents Part 6 Valuation of Costs and Benefits


on previous COBALT accident proportions (2000, TAG data book, January 2014). See paragraph 5.5 inChapter 5 for more details.

4.5 Accident rates and severity have been reducing over recent years and this trend is expected to continueinto the future; see paragraph 5.6 in Chapter 5. The forecasting of the proportion of fatal/serious/slightaccidents is based on a similar methodology. First, the fatal and serious proportions are forecast byapplying the accident rate change coefficient (see paragraphs 5.6 and 5.7 and Tables 6/5/1 and 6/5/2 inChapter 5). The sum of these is subtracted from unity to determine the slight proportion.

4.6 The average accident costs used in NESA will normally be appropriate even where local accident ratesdiffer from the average. In some circumstances the severity split may differ with a consequent change

Table 6/4/2: Proportions of Fatal, Serious and Slight Accidents on Links (Average 1999- 2001)

LINK ONLY ACCIDENT PROPORTIONS (2000 Base)Accident Type Road Type Accident Proportions

Casualty Severity Fatal (f) Serious (se) Slight (sl)1-3 Motorways 0.014 0.096 0.890

Speed Limit 30/40 mph (48/64 kph) >40 mph (>64 kph)Casualty Severity f se sl f se sl

4-8 S2 A Roads 0.013 0.146 0.841 0.050 0.232 0.7189 Other S2 Roads 0.009 0.147 0.844 0.027 0.208 0.765

10-15 Dual Carriageways 0.014 0.140 0.847 0.028 0.159 0.813LINK AND JUNCTION COMBINED ACCIDENT PROPORTIONS (2009 Base)




10-15 Dual Carriageways 0.009 0.113 0.878 0.025 0.148 0.827

Table 6/4/3: Proportions of Fatal, Serious and Slight Accidents at Junctions (Average 1999 - 2001)

PROPORTIONS OF ACCIDENTS AT JUNCTIONS (2000 Base)Speed Limit 30/40 mph (48/64 kph) >40 mph (>64 kph)Casualty Severity f se sl f se sl3 Arm Priority 0.0072 0.1240 0.8688 0.0243 0.1883 0.78744 and 5 Arm Priority 0.0060 0.1162 0.8778 0.0271 0.2064 0.76653 Arm Traffic Signals 0.0060 0.1066 0.8874 0.0092 0.1171 0.87374 and 5 Arm Traffic Signals 0.0060 0.1069 0.8871 0.0088 0.1153 0.8759Roundabouts (Standard & Small) 0.0027 0.0746 0.9227 0.0056 0.0912 0.9032Mini Roundabout 0.0028 0.0868 0.9104 0.0056 0.0912 0.9032Signalised Roundabouts 0.0031 0.0643 0.9326 0.0039 0.0624 0.9337



in average accident costs, but this is seldom likely to be significant. Transport Scotland should beconsulted if it is considered necessary to use local severity splits. Transport Scotland policy is todiscourage the use of a local severity split unless it can be shown that exceptional conditions genuinelyarise which are unlikely to be corrected by modest remedial works.

4.7 In order to accept a local severity split the user must:

(i) demonstrate that the severity split is significantly different in statistical terms from theNESA value, and also that this does not result from one or two particularly badaccidents, the effect of which will be evened out by less extreme accidents as timegoes by. Data covering all available accident history, with a minimum of five yearsmust be supplied;

(ii) arrange an Accident Investigation and Prevention Study to identify the causes of thesafety problem and recommend remedial safety measures. Where this studyconcludes that modest remedial works are unlikely to correct the problem then a localseverity split may be used. However where modest remedial works are recommended,the cost of these works should be included in the do minimum and the revised NESAseverity split used.

4.8 Statistics of damage only accidents are not generally available. On the basis of some surveyinformation, these are taken in NESA to occur at the rates given in Table 6/4/1. The accident costsgiven in Part 6 Chapters 5 and 6 include allowance for damage only accidents at these rates.

4.9 For the purposes of appraisal it is necessary to form a view on how costs will vary over future years.The value of most elements of accident costs are proportional to national income, and for this reason itis assumed that values change in line with GDP per head. In accordance with the Treasury Green Bookguidance these growth rates are decreased after thirty years when the discount rate is reduced. Thereduction is the same proportional reduction as the change in the discount rate.

Table 6/4/4: Assumed Compound Annual Rates of Growth in Accident Values (%)

Range of Years Rates of Growth in Accident Values (% pa)

2010-2011 0.082011-2012 0.002012-2013 1.092013-2014 2.052014-2015 1.672015-2016 1.952016-2017 1.992017-2018 1.902018-2019 1.912019-2020 1.902020-2021 1.882021-2022 1.872022-2023 1.892023-2024 1.902024-2025 1.922025-2026 1.942026-2027 1.952027-2028 1.972028-2029 1.99



2029-2030 2.012030-2031 2.022031-2032 2.042032-2033 2.052033-2034 2.062034-2035 2.072035-2036 2.082036-2037 2.092037-2038 2.102038-2039 2.102039-2040 2.102040-2041 2.102041-2042 2.122042-2043 2.122043-2044 2.122044-2045 2.122045-2046 2.122046-2047 2.152047-2048 2.152048-2049 2.152049-2050 2.152050-2051 2.152051-2052 2.192052-2053 2.192053-2054 2.192054-2055 2.192055-2056 2.192056-2057 2.212057-2058 2.212058-2059 2.212059-2060 2.212060-2061 2.212061-2062 2.222062-2063 2.212063-2064 2.212064-2065 2.212065-2066 2.212066-2067 2.202067-2068 2.202068-2069 2.202069-2070 2.202070-2071 2.202071-2072 2.172072-2073 2.172073-2074 2.172074-2075 2.172075-2076 2.172076-2077 2.17




4.10 The total cost of accidents on a road network is calculated by multiplying the number of accidentspredicted to occur on the network by the cost per accident. As explained above, the cost per accidentvaries by type and area of road. The number of accidents on a given length of road is expressed as anaccident rate, defined as so many Personal Injury Accidents per million vehicle kilometres, so thatdoubling either the length or the traffic flow on the road will double the number of accidents. Apartfrom length and flow level, in NESA there are two other determinants of the number of accidents: thenumber and type of junctions and the type of links.

4.11 NESA incorporates a method of separating out the effects of links and junctions on accidents. Wherejunctions are coded for delay calculation, they should also be coded for accident calculation. Inaddition, where there are junctions which are subsumed in links for speed calculations (in particular inurban areas), but which are likely to be associated with accidents, these should be coded asaccident-only nodes. Finally, where either a very large link-only network is used and accident-onlynodes are difficult to identify, or local data on existing accidents are difficult to split between links andjunctions, combined link and junction accident rates can be attributed to links. The treatment ofaccidents on links and junctions is described in detail in the following two chapters.

4.12 NESA presents total numbers of injury accidents for both the Do-Minimum and Do-Somethingnetworks plus their associated costs and benefits (ref. NESA output Tables 4 and 5). The total numberof casualties by severity (fatal, serious and slight) are also presented for both the Do-Minimum andDo-Something networks.

4.13 NESA has the ability to run accident only assessments. These assessments can provide the accidentbenefits (or costs) associated with a scheme where the main economic assessment has been undertakenusing another program which does not include an assessment of accidents such as TUBA or PEARS

2077-2078 2.172078-2079 2.172079-2080 2.172080-2081 2.172081-2082 2.172082-2083 2.172083-2084 2.172084-2085 2.172085-2086 2.172086-2087 2.172087-2088 2.182088-2089 2.182089-2090 2.182090-2091 2.182091-2092 2.182092-2093 2.172093-2094 2.172094-2095 2.172095-2096 2.172096-2097 2.172097-2098 2.172098-2099 2.172099-2100 2.17




(refer STAG Technical Database Section 8.2.1). Details of how to run an accident only assessment areprovided in Part 10 Chapter 18.

Volume 15 Section 1 Chapter 5Part 6 Valuation of Costs and Benefits The Valuation of Accidents on Links


5 THE VALUATION OF ACCIDENTS ON LINKS5.1 The NESA user has to decide whether to code junction accidents separately from link accidents.

Combined accident rates attribute all accidents to links. The link-only rates exclude junction accidents(that is, those occurring within 20 metres of a junction). They are therefore lower than the combinedrates for a particular link. If link only rates are used then accidents at junctions are modelled separately(see Part 6 Chapter 6).

5.2 The preferred method of evaluating accidents is to separate link and junction accidents, using localaccident data to define the Do-Minimum rates and the default rates for new links and junctions in theDo-Something. However, Combined accident rates should be selected in NESA when:

(i) local data for the Do-Minimum are not available;

(ii) local data have already been collected in combined form and resources todisaggregate the data are not available; or,

(iii) a large network is being used and accident-only nodes would be difficult to identify.

5.3 Accident rates and severities have been falling steadily over time and the trend is expected to continuein the future. The Government has also announced National Casualty Reduction targets and themethodology and parameters in NESA are consistent with those targets.

5.4 Local data can be obtained from the appropriate police or local authority and should relate to a periodwhen conditions on the road have been broadly unchanged (for example, no abnormal changes intraffic flow, no changes in junction design or road geometry, etc.). Local data should normally coverthe five years previous to the NESA assessment. The number of accidents for this five year period orthe average accident rate (accidents per million vehicle kilometres) must be input, including zero forthose links where no accidents occurred. Where the number of accidents has been input NESA willinternally produce a local accident rate (accidents per million vehicle kilometres) for each link.

5.5 For existing links where local accident data is not available and for new links, the program will use theNESA default accident rates shown in Table 6/5/1 (Link Only) and Table 6/5/2 (Link & JunctionCombined). These rates are not based on the most recent COBALT accident rates (2009, TAG databook, November 2014) but rather, on previous COBALT accident rates (2000, TAG data book, January2014). This is because the most recent COBALT accident rates (2009) result in fewer link categoryrates. Where there were previously six accident rates covering different link types, there are now onlythree. Where previous versions of NESA would capture a step change between, for example RC26(Rural - typical single 7.3m) and RC27 (Rural - good single 7.3m), the new COBALT accident rates(2009) would no longer reflect this. Therefore accident rates and casualty rates, shown in Table 6/5/3and Table 6/5/4, have been retained at 2000 values.

5.6 Accident rates have been selected on the basis of the COBALT accident type which best fits the NESAroad category description. The user should, however, consider the appropriateness of these default rateswith respect to the standard of the link being considered, any available local accident data and the ratesand classifications given in Table 6/5/3 and Table 6/5/4. Changes to defaults should always be justified.

Chapter 5 Volume 15 Section 1The Valuation of Accidents on Links Part 6 Valuation of Costs and Benefits


5.7 The declining trend in accident rates was examined in TRL Report 382 and at a more disaggregatelevel in later work for the DTLR. It was found that the changes in accident rates and the number ofseverities per accident are explained by the relationship:

5.8 The values for the accident rate change coefficient incorporated in the NESA program for thedifferent link accident types are given in Tables 6/5/1 and 6/5/2. They are the same for Link-Only andLink and Junction Combined analyses and should be applied for any year from 1995 until year 2010.Between 2011 and 2020 the change is assumed to be one half of the 1995 to 2010 rate, and between2021 and 2030 the change is assumed to be one quarter of the 1995 to 2010 rate. Zero change isassumed post 2030.

5.9 For design networks, the user also has the facility to code link specific accident rates. Clearjustification for changes to the default Do-Something rates or costs should be presented to TransportScotland.

Equation 6/5/1

where AN = the accident rate or number of casualties per accident N years after base year= the accident rate or number of casualties per accident in the base year

N = accident rate change coefficient raised to the power N (the number of years after base year)

AN AO N=

AO



Table 6/5/1: Default Link Only Accident Rates (2000 Base)

Link Only

NESA Road

Category

Accident Type

Description AccidentRates

(P.I.A./MVkm)

Accident Rate Change

Coefficient1 9 Urban - single 6.0m 0.297 0.9832 8 Urban - single 7.3m 0.297 0.9843 8 Urban - single 10.0m 0.297 0.9844 8 Urban - single 4 lanes 0.297 0.984

5 9 Urban - one way 6.0m 0.297 0.9836 8 Urban - one way 7.3m 0.297 0.9847 8 Urban - one way 10.0m 0.297 0.984

8 12 Urban - dual 2 0.295 0.9849 15 Urban - dual 3 0.295 0.98410 15 Urban - dual 4 0.295 0.984

11 12 Urban - Expressway, 2 or more lanes 0.295 0.98412 12 Urban - Motorway and dual ramps, 1 lane 0.295 0.98413 12 Urban - Motorway and dual ramps, 2 lanes 0.295 0.984

14 1 Urban - Motorway - D2 0.089 1.00115 2 Urban - Motorway - D3 0.089 1.00116 3 Urban - Motorway - D4 0.089 1.00117 3 Urban - Motorway - D5 0.089 1.001

20 9 Rural - single 4 lane 0.297 0.998

21 9 Rural - poor single 4.0m 0.297 0.99822 9 Rural - poor single 5.5m 0.297 0.99823 9 Rural - poor single 6.0m 0.297 0.99824 8 Rural - typical single 6.0m 0.226 0.973

25 9 Rural - poor single 7.3m 0.297 0.99826 8 Rural - typical single 7.3m 0.226 0.97327 4 Rural - good single 7.3m 0.174 0.973

28 5 Rural - typical single 10.0m 0.138 0.97329 6 Rural - good single 10.0m 0.113 0.97330 6 Rural - single with climbing lane 0.113 0.973

31 12 Rural - dual 2 lanes 0.154 0.97332 15 Rural - dual 3 lanes 0.154 0.97333 15 Rural - dual 2 with climbing lane 0.154 0.973

34 11 Rural - dual 2 lanes with grade separation 0.089 0.97335 14 Rural - dual 3 lanes with grade separation 0.089 0.97336 11 Rural - dual 2 lanes with GS and climbing lane 0.089 0.973

37 10 Rural - Motorway and dual ramps, 1 lane 0.119 0.97338 10 Rural - Motorway and dual ramps, 2 lanes 0.119 0.973

39 1 Rural - Motorway - D2 0.089 1.00140 2 Rural - Motorway - D3 0.089 1.00141 2 Rural - Motorway - D2 with climbing lane 0.089 1.001



Table 6/5/2: Default Link and Junction Combined Accident Rates (2000 Base)

Link & Junction Combined

NESA Road

Category

Accident Type

Description AccidentRates

(P.I.A./MVkm)

Accident Rate Change

Coefficient1 9 Urban - single 6.0m 0.844 0.9832 8 Urban - single 7.3m 0.844 0.9843 8 Urban - single 10.0m 0.844 0.9844 8 Urban - single 4 lanes 0.844 0.984

5 9 Urban - one way 6.0m 0.844 0.9836 8 Urban - one way 7.3m 0.844 0.9847 8 Urban - one way 10.0m 0.844 0.984

8 12 Urban - dual 2 1.004 0.9849 15 Urban - dual 3 1.004 0.98410 15 Urban - dual 4 1.004 0.984

11 12 Urban - Expressway, 2 or more lanes 1.004 0.98412 12 Urban - Motorway and dual ramps, 1 lane 1.004 0.98413 12 Urban - Motorway and dual ramps, 2 lanes 1.004 0.984

14 1 Urban - Motorway - D2 0.098 1.00115 2 Urban - Motorway - D3 0.098 1.00116 3 Urban - Motorway - D4 0.098 1.00117 3 Urban - Motorway - D5 0.098 1.001

20 9 Rural - single 4 lane 0.404 0.998

21 9 Rural - poor single 4.0m 0.404 0.99822 9 Rural - poor single 5.5m 0.404 0.99823 9 Rural - poor single 6.0m 0.404 0.99824 8 Rural - typical single 6.0m 0.381 0.973

25 9 Rural - poor single 7.3m 0.404 0.99826 8 Rural - typical single 7.3m 0.381 0.97327 4 Rural - good single 7.3m 0.293 0.973

28 5 Rural - typical single 10.0m 0.232 0.97329 6 Rural - good single 10.0m 0.190 0.97330 6 Rural - single with climbing lane 0.190 0.973

31 12 Rural - dual 2 lanes 0.226 0.97332 15 Rural - dual 3 lanes 0.226 0.97333 15 Rural - dual 2 with climbing lane 0.226 0.973

34 11 Rural - dual 2 lanes with grade separation 0.131 0.97335 14 Rural - dual 3 lanes with grade separation 0.131 0.97336 11 Rural - dual 2 lanes with GS and climbing lane 0.131 0.973

37 10 Rural - Motorway and dual ramps, 1 lane 0.174 0.97338 10 Rural - Motorway and dual ramps, 2 lanes 0.174 0.973

39 1 Rural - Motorway - D2 0.098 1.00140 2 Rural - Motorway - D3 0.098 1.00141 2 Rural - Motorway - D2 with climbing lane 0.098 1.001



5.10 Table 6/5/4 gives the casualty rate reduction factor for each link type. The changes are assumed toapply up to 2010 with zero change thereafter.

Table 6/5/3: Average Number of Casualties per Accident (2000 Base)

LINK ONLY CASUALTIES (2000 Base)Accident

TypeROAD TYPE CASUALTIES PER P.I.A.




10-15 Dual Carriageways 0.0143 0.1546 1.145 0.0314 0.2005 1.312LINK AND JUNCTION COMBINED CASUALTIES (2000 Base)





Table 6/5/4: Casualties per Accident Change Coefficients

LINK ONLY Change Coefficients Accident

TypeROAD TYPE BETA FACTORS




10-15 Dual Carriageways 0.949 0.965 1.013 0.947 0.967 1.007LINK AND JUNCTION COMBINED Change Coefficients





Volume 15 Section 1 Chapter 6Part 6 Valuation of Costs and Benefits The Valuation of Accidents at Junctions


6 THE VALUATION OF ACCIDENTS AT JUNCTIONS

6.1 NESA estimates numbers and costs of accidents at specified junctions. This is the recommendedmethod for NESA appraisal and should be used unless information on junction characteristics islacking. All new junctions should be coded for accident appraisal, as should all existing junctionswhere there are significant forecast traffic flow changes and where accidents are likely to occur.Junctions which are coded for accident but not traffic delay purposes are called accident-only nodes. Inthe absence of local data on junction-attributable accidents, NESA default values for junctions shouldbe used.

6.2 NESA incorporates two models which relate accidents at junctions to given flow configurations. Theseallow forecasts of future accident numbers to be derived for existing and new junctions. For existingjunctions, the use of local accident data is recommended, with default values being used in the absenceof such data. The number of accidents occurring at (that is within 20 metres of) each junction asrecorded by the appropriate police or local authority should be used. Local data should normally coverthe five years previous to the NESA assessment and must cover at least three years.

6.3 The models are of two types, both of the basic form:

They have been estimated by reference to accidents and flows at existing junctions. The choice offunction varies according to junction type as indicated in Table 6/6/1 which also shows the associatedvalues of a and b.

6.4 In the Cross Product (C) model, (f) is the value produced by multiplying the combined inflow from thetwo major opposing links by the sum of the inflows on the other one or two minor links. Inflows aremeasured in thousands of vehicles per annual average day. In the Inflow (I) model, (f) is the value ofthe total inflow from all links in thousands of vehicles per annual average day. Where the user inputslocal accident numbers for existing junctions, the NESA program calculates a local value for a, with bbeing fixed at the national value.

6.5 As with links, accident rates and their severity at junctions have been falling steadily over time and thetrend is expected to continue in the future. The Government has also announced National CasualtyReduction targets and the methodology and parameters in NESA are consistent with those targets.

6.6 NESA uses 96 junction types for accident assessment. These types can be aggregated into three broadcategories, namely, major/minor, signals and roundabouts (subdivided into standard, small and mini).Major/minor junctions include staggered, standard priority, multiple, Y-junctions and cross-roads.Staggered junctions can be treated either as crossroads or as pairs of three-arm junctions. Normally,where the stagger is significant and the junction effectively operates as two T-junctions, it should becoded as two three-arm junctions. This also applies to the coding of junctions for delay purposes (seePart 8 Chapter 8).

Equation 6/6/1

where A is the annual number of accidentsf is a function of traffic flowa, b vary among junction types

A a f b=

Chapter 6 Volume 15 Section 1The Valuation of Accidents at Junctions Part 6 Valuation of Costs and Benefits


Table 6/6/1: Junction Accident Parameters (2000 Base)

NON-BUILT-UP: ABOVE 40 MPH (>64 KPH)

BUILT-UP: UP TO 40 MPH(<= 40 KPH)

Junction Types Number of Arms

Highest Link Standard

(Single or Dual)

Formula Type

Junction Type

Coeff ‘a’ Power ‘b’ Junction Type

Coeff ‘a’ Power ‘b’

PRIORITY 3 S C 1 0.195 0.460 2 0.195 0.4603 D C 3 0.195 0.460 4 0.195 0.4604 S I 5 0.361 0.440 6 0.361 0.4404 D C 7 0.240 0.710 8 0.240 0.7105 S I 9 0.361 0.440 10 0.361 0.4405 D I 11 0.361 0.440 12 0.361 0.440

PRIORITY WITH GHOST

ISLANDS

3 S C 13 0.195 0.460 14 0.195 0.4603 D C 15 0.195 0.460 16 0.195 0.4604 S I 17 0.361 0.440 18 0.361 0.4404 D C 19 0.240 0.710 20 0.240 0.7105 S I 21 0.361 0.440 22 0.361 0.4405 D I 23 0.361 0.440 24 0.361 0.440

PRIORITY WITH SINGLE

LANE DUALLING

3 S C 25 0.195 0.460 26 0.195 0.4603 D C 27 0.195 0.460 28 0.195 0.4604 S I 29 0.361 0.440 30 0.361 0.4404 D C 31 0.240 0.710 32 0.240 0.7105 S I 33 0.361 0.440 34 0.361 0.4405 D I 35 0.361 0.440 36 0.361 0.440

SIGNALS 3 S I 37 0.223 0.610 38 0.223 0.6103 D C 39 0.494 0.420 40 0.291 0.5104 S C 41 1.378 0.200 42 1.378 0.2004 D C 43 0.494 0.420 44 0.291 0.5105 S I 45 0.254 0.620 46 0.254 0.6205 D I 47 0.238 0.850 48 0.160 0.970

ROUNDABOUTS - STANDARD

3 S C 49 0.033 0.760 50 0.033 0.7603 D C 51 0.033 0.760 52 0.033 0.7604 S C 53 0.024 0.890 54 0.048 0.7404 D C 55 0.063 0.690 56 0.022 0.8505 S I 57 0.007 1.770 58 0.014 1.5305 D I 59 0.019 1.420 60 0.006 1.730

- SMALL 3 S C 61 0.033 0.760 62 0.033 0.7603 D C 63 0.033 0.760 64 0.033 0.7604 S C 65 0.101 0.660 66 0.263 0.5404 D C 67 0.101 0.660 68 0.263 0.5405 S I 69 0.044 1.280 70 0.095 1.1405 D I 71 0.044 1.280 72 0.095 1.140

- MINI 3 S C 73 0.012 1.040 74 0.012 1.0403 D C 75 0.012 1.040 76 0.012 1.0404 S C 77 0.070 0.640 78 0.070 0.6404 D C 79 0.070 0.640 80 0.070 0.6405 S I 81 0.013 1.470 82 0.013 1.4705 D I 83 0.013 1.470 84 0.013 1.470

- SIGNALLED 3 S C 85 0.033 0.760 86 0.033 0.7603 D C 87 0.033 0.760 88 0.033 0.7604 S C 89 0.024 0.890 90 0.048 0.7404 D C 91 0.063 0.690 92 0.022 0.8505 S I 93 0.007 1.770 94 0.014 1.5305 D I 95 0.019 1.420 96 0.006 1.730



6.7 As an example, imagine a rural 4-arm major / minor junction with inflows as shown in Figure 6/6/1below (AADT in 000s):

The major road in this case is a dual carriageway and so the relevant formula is for junction type 7, is ofthe Cross Product type, and is:

Given the above flows, the value of f is 54 (that is, (8+10) x (1+2)) and the predicted number ofpersonal injury accidents per annum is:

In all applications of the Cross Product model, any combined inflow from the one or two minoropposing links that amounts to less than 1 (that is, 1000 AADT) will be taken by NESA to be 1 becausesuch low flows were rarely encountered in the research which produced the accident formulae;consequently little evidence of the effect of changes in very low combined inflows was found. In suchcases the formulae are sensitive only to changes in combined minor or major link inflows whichinvolve inflow levels over 1000 (AADT). This cut-off applies also to the combined inflow from thetwo major flows. However it will usually be inappropriate to model such low flow junctions asparagraph 6.8 and Table 6/6/2 make clear.

If all junction approaches had been single carriageways, and there was no ghost island at the junction,the number of accidents would be calculated using an Inflow formula with parameters for junction type5. In this case, the value of the flow function (f) would be (10+8+2+1)=21, and the number of accidentswould be:

Figure 6/6/1: Rural 4-Arm Major/Minor Junction

A 0.24 f 0.710=

A 0.24 54 0.710 4.08= =

A 0.361 21 0.440 1.38= =



6.8 It must be stressed that these junction accident formulae have been derived from actual records ofaccidents and flows at junctions. The accident predictions for given junction types with flow levels andconfigurations outside the ranges recorded in the research are not a reliable guide to design, althoughthey provide useful indications when used in conjunction with other methods of operational analysis.They should be used with great care when considering individual junctions, for example, junctionaccident rates may vary to some extent with the local spacing of junctions and other characteristics notincluded in the NESA accident formulae. When comparing the NESA accident benefits of differentjunction types (as part of an economic appraisal of preferred junction type), the user should check thatthe ranking by accident benefits accords with engineering judgement. Sensitivity tests using a range oflikely accident rates may be appropriate. In particular, comparisons based on flow levels outside theobserved ranges of the NESA junction formulae set out in Table 6/6/2 should be scrutinised carefully,especially where the combined inflow minimum value of 1 is imposed in a Cross Product accidentformula. In such cases accident predictions may not be sensitive to small changes in flows. TransportScotland should be consulted where the assessment of preferred junction type depends critically onaccident benefits.

6.9 As with links, accident rates and accident severity at junctions have been falling steadily over time andthe trend is expected to continue in the future. The declining trend in accident rates was examined inTRL Report 382 and at a more disaggregate level in later work undertaken for the DfT. It was foundthat the changes in accident rates and the number of severities per accident at junctions are explainedby the relationship:

6.10 The values for the change coefficient found by the research are given in Table 6/6/3 for Major andMinor junctions in Built-up (BU = 30 or 40mph (48 or 64 kph) speed limits) and Non Built-up (NBU =above 40 mph (>64 kph) speed limit) locations. They are the same for each junction type. Major meansthat a dual carriageway or motorway is the highest carriageway standard approaching the junction andMinor means that only single carriageway roads approach the junction (this is a slight change toprevious versions of NESA (NESA05 and earlier) where junctions were specified as Major when amotorway or A-class road was the highest carriageway standard approaching the junction and specifiedas Minor when only lower class roads approached the junction). The change in methodology ensures

Table 6/6/2: Observed Ranges of Flow in NESA Junction Accident Model Calibration

Junction Type Thousands Veh/Day3-Arm

Major / Minor 5 - 10Signal & Roundabout 15 - 20

4 or more ArmMajor / Minor 5 - 10 (Single) 15 - 20 (Dual)Signal 10 - 20 (Single) 25 - 35 (Dual)Roundabout- Standard 25 - 30 (Single) 30 - 40 (Dual)- Small 25 - 35- Mini 15 - 20

Equation 6/6/2

where AN = the accident rate or number of casualties per accident N years after base yearAO = the accident rate or number of casualties per accident in the base yearN = accident rate change coefficient raised to the power N (the number of years

after base year)

AN AO N=



consistency with the approach adopted by the Department for Transport's COBALT program. TheMajor values of are held as default within the program and are assumed to apply for all years from1995 to 2010. As with links, the number of casualties per accident post 2010 is held constant at the2010 level. But between 2011 and 2020 and 2021 and 2030 the accident rate change is assumed to beone half and one quarter respectively of the 1995 to 2010 change. Zero change is assumed post 2030.

Table 6/6/3: Accident and Casualties per Accident Change Coefficient for Junctions

Junction Classification

Accident Rate Number of Casualties per AccidentFatal Serious Slight

Major, BU 0.991 0.949 0.962 1.010Minor, BU 0.976 0.968 0.958 1.006

Major, NBU 0.984 0.961 0.959 1.011Minor, NBU 0.996 0.976 0.972 1.011

Volume 15 Section 1 Chapter 7Part 6 Valuation of Costs and Benefits Carbon Emissions


7 CARBON EMISSIONS7.1 NESA includes the calculation and valuation of Carbon Dioxide equivalent (CO2e).

7.2 The calculation of Carbon Dioxide equivalent (CO2e) is considered to be the key indicator in assessingthe impacts of transport options on climate change. The valuation of Carbon Dioxide equivalent(CO2e) considers the additional global damage arising from an additional tonne of carbon beingemitted.

7.3 Carbon Dioxide equivalent (CO2e) is estimated per litre of fuel burnt and as such are calculated fromthe amount of fuel consumed. The change in fuel consumption will depend on changes in both thedistance travelled and vehicle speeds. The change in Carbon Dioxide equivalent (CO2e) can be eithernegative or positive depending on the changes in vehicle speeds and distance travelled.

7.4 Carbon Dioxide equivalent (CO2e) per litre of fuel burnt and the shadow price of carbon per tonne,plus changes in these over time, are detailed further in the TAG data book, November 2014.

7.5 Table 6/7/1 presents carbon emissions per litre of fuel burnt and Table 6/7/2 presents the low, centraland high non-traded values of carbon per tonne of carbon.

Table 6/7/1: Carbon Dioxide equivalent (CO2e) per Litre of Fuel Burnt / KWh Used

Year Petrol(KgCO2/l)

Diesel(KgCO2/l)

Electricity(KgCO2/KWh)

2010 2.2299 2.5617 0.35402011 2.2111 2.5665 0.34732012 2.2106 2.6087 0.34032013 2.2009 2.5973 0.33282014 2.1891 2.6013 0.32502015 2.1891 2.6014 0.31672016 2.1891 2.6015 0.30802017 2.1595 2.5561 0.29882018 2.1299 2.5107 0.28902019 2.1003 2.4654 0.27882020 2.0707 2.4200 0.26792021 2.0707 2.4200 0.25652022 2.0707 2.4200 0.24442023 2.0707 2.4200 0.23162024 2.0707 2.4200 0.21822025 2.0707 2.4200 0.20402026 2.0707 2.4200 0.18902027 2.0707 2.4200 0.17312028 2.0707 2.4200 0.15642029 2.0707 2.4200 0.13882030 2.0707 2.4200 0.12022031 2.0707 2.4200 0.10972032 2.0707 2.4200 0.10012033 2.0707 2.4200 0.09142034 2.0707 2.4200 0.08342035 2.0707 2.4200 0.07622036 2.0707 2.4200 0.0695

Chapter 7 Volume 15 Section 1Carbon Emissions Part 6 Valuation of Costs and Benefits


2037 2.0707 2.4200 0.06352038 2.0707 2.4200 0.05792039 2.0707 2.4200 0.05292040 2.0707 2.4200 0.04832041 2.0707 2.4200 0.04302042 2.0707 2.4200 0.04202043 2.0707 2.4200 0.03732044 2.0707 2.4200 0.03332045 2.0707 2.4200 0.03422046 2.0707 2.4200 0.03162047 2.0707 2.4200 0.02892048 2.0707 2.4200 0.03402049 2.0707 2.4200 0.0302

2050 onwards 2.0707 2.4200 0.0302

Table 6/7/2: Non Traded Values of Carbon Dioxide equivalent (CO2e) (£ per Tonne)

Year Low Central High

2010 26.64 53.28 79.922011 27.04 54.08 81.122012 27.45 54.89 82.342013 27.86 55.72 83.582014 28.28 56.55 84.832015 28.70 57.40 86.102016 29.13 58.26 87.392017 29.57 59.14 88.702018 30.01 60.02 90.032019 30.46 60.92 91.392020 30.92 61.84 92.762021 31.43 62.87 94.302022 31.95 63.90 95.852023 32.46 64.93 97.392024 32.98 65.96 98.942025 33.50 66.99 100.492026 34.01 68.02 102.032027 34.53 69.05 103.582028 35.04 70.08 105.122029 35.56 71.11 106.672030 36.07 72.14 108.222031 39.42 78.84 118.262032 42.77 85.54 128.312033 46.12 92.24 138.362034 49.47 98.94 148.412035 52.82 105.64 158.462036 56.17 112.34 168.51

Table 6/7/1: Carbon Dioxide equivalent (CO2e) per Litre of Fuel Burnt / KWh Used

Volume 15 Section 1 Chapter 7Part 6 Valuation of Costs and Benefits Carbon Emissions


2037 59.52 119.04 178.562038 62.87 125.74 188.602039 66.22 132.44 198.652040 69.57 139.13 208.702041 72.92 145.83 218.752042 76.27 152.53 228.802043 79.62 159.23 238.852044 82.97 165.93 248.902045 86.31 172.63 258.942046 89.66 179.33 268.992047 93.01 186.03 279.042048 96.36 192.73 289.092049 99.71 199.43 299.142050 103.06 206.12 309.192051 105.64 213.41 321.192052 108.03 220.48 332.922053 110.37 227.57 344.772054 112.64 234.67 356.702055 114.73 241.54 368.352056 116.77 248.45 380.142057 118.63 255.12 391.602058 120.34 261.62 402.892059 121.95 268.01 414.082060 123.41 274.25 425.082061 124.24 279.19 434.152062 125.01 284.11 443.212063 125.51 288.53 451.552064 125.87 292.71 459.562065 125.97 296.39 466.812066 126.01 300.02 474.032067 125.76 303.05 480.332068 125.38 305.82 486.252069 124.81 308.16 491.522070 124.08 310.19 496.302071 123.32 312.20 501.092072 122.41 313.87 505.332073 121.39 315.29 509.192074 120.13 316.14 512.142075 118.89 317.03 515.172076 117.33 317.10 516.882077 115.78 317.22 518.652078 114.06 316.82 519.592079 112.26 316.24 520.212080 110.28 315.08 519.892081 108.65 314.92 521.202082 106.83 314.21 521.592083 104.94 313.26 521.582084 103.00 312.11 521.22


Chapter 7 Volume 15 Section 1Carbon Emissions Part 6 Valuation of Costs and Benefits


2085 101.11 311.10 521.092086 99.03 309.48 519.932087 96.90 307.62 518.342088 94.77 305.71 516.652089 92.57 303.51 514.45


Volume 15 Section 1 Chapter 8Part 6 Valuation of Costs and Benefits Construction Costs


8 CONSTRUCTION COSTS8.1 Cost-benefit analysis compares the benefit of a project with its costs. The benefits in NESA are the

streams of user cost savings appraised over the nominal life of the scheme (see Part 3 Chapter 4) anddiscounted to the present value year. The costs taken into account in NESA are principally theconstruction, land and property costs involved in carrying out the road improvement, includingpreparation and supervision costs. Value Added Tax should be excluded from the scheme cost estimateused in NESA assessment (see Part 3 Chapter 6). All costs are allocated to the sector incurring thecosts, that is, Central Government, Local Government or the Private Sector (for example Developers).

8.2 In addition, NESA will road calculate non-traffic related maintenance cost savings. Where delay anddiversion is imposed on road users during the construction period, an estimate of the traffic delay costsshould be input to the NESA assessment. These are treated as negative benefits in NESA. Theprocedures explained in Part 6 Chapters 8 to 10 apply to both Do-Minimum and Do-Something costs.

8.3 The costs of a scheme are usually incurred over a period of years in which preparation and constructiontakes place. It is necessary, as with the stream of benefits, to discount these to the present value year. Itis important that construction, land and property delay costs should be allocated accurately to theperiod in which they are incurred.

8.4 It is an important principle of cost-benefit analysis that costs should be recorded when they are incurredrather than when payment is made. As soon as resources are used on a project they cease to beavailable for alternative uses; the fact that there may be a lag in payment is immaterial. However it isnot until it is taken out of other uses that it can be charged against the cost-benefit analysis of theproject. For example, it may be the case that agricultural land has been acquired for a scheme wellbefore construction begins. Despite this, the acquisition cost should not be charged in the cost-benefitanalysis, until the land is no longer farmed. Land and property costs are discussed further in paragraphs8.5 to 8.10.

8.5 The NESA appraisal should concern itself only with costs which will be incurred as a result of adecision to go ahead with the scheme. It should not include bygones, that is expenditures incurred priorto economic appraisal which cannot be retrieved as a result of any subsequent decision. A distinctionshould be made between expenditure which is genuinely a bygone and that which is retrievable. Anexample of the former might be the preparation costs which have been incurred in the designing of ascheme prior to the date of the NESA appraisal. However, land and property costs are generallyretrievable, since it is usually possible to recoup them if the scheme does not go ahead. Such costsshould not therefore be treated as a bygone, unless the property has actually been demolished. In suchcircumstances, however, the land on which the property once stood will still have a re-sale value.

8.6 Economic appraisal should precede the undertaking of advance works. Advance works are notjustifiable if the scheme itself cannot be justified. Properties likely to be required for a scheme shouldideally be managed and kept in use rather then demolished prematurely. However, where the economicappraisal is being updated after part of the scheme has been undertaken, irretrievable costs should beexcluded from the cost of the remaining sections being appraised.

Land and Property Costs

8.7 The following paragraphs consider in greater detail how some of the principles described above applyto land and property costs in NESA. These principles apply to all land and property affected by thescheme, that is, not just land and property directly on the line of the final route of the scheme but alsoall land and property off the line of the route which may become involved with the scheme in someway. This would include land and property acquired through discretionary purchase and compensationpayments. Such costs are often difficult to predict, since they can depend on the outcome of arbitration

Chapter 8 Volume 15 Section 1Construction Costs Part 6 Valuation of Costs and Benefits


or other factors. However, information may be obtained from the out-turn costs of previous schemes toindicate the likely scale of costs involved. It is important that all such relevant land and property costsare considered in the appraisal and included in the scheme cost estimates.

8.8 Payment for land and property may be made at various times before, during and after construction.Where land has been purchased in advance of its use for a scheme, the value of the land may havechanged either upwards or downwards in the interval. This change reflects a change in the opportunitycost of the land, that is the value of the land when put to its best alternative use. Irrespective of thepurchase, it is the current value, as estimated by the District Valuer, which should be debited against thescheme.

8.9 For many schemes it would probably be sufficient to use the simplifying assumption that land andproperty is taken out of its previous use in the first year of scheme construction and hence all land andproperty costs would be entered in the year they are purchased (probably the first year of schemeconstruction). The situations where this is likely to be an inappropriate assumption are:

• where land/property is taken out of use before the first year of construction;

• where the land and property costs of a scheme are unusually large;

• for all schemes going to Order Publication the simplifying assumption should be questioned.

8.10 Land and property which is purchased before scheme construction should always, as far as ispracticable, be made available for continued agricultural use/occupation. However, in some cases thismay not be possible. For example, a residential property which the Department owns may becomeunsafe for occupation. If formal appraisal of the alternative options then shows that demolition ispreferred to the maintenance which would allow occupancy to continue for the short period beforeconstruction then the property may be taken out of use. In such a case the value of the property shouldbe entered in the appraisal in the year in which it would be taken out of use. This reflects the loss of theuse of the property, avoidable in the Do-Minimum, prior to construction where this is significant. Itshould be noted the decision to remove land and property from use before construction must be basedon a formal appraisal, this prevents any perverse application of the concept of bygones (see paragraph7.4) to reduce scheme costs.

8.11 Land and property costs consist of a number of different resource elements and it is important toconsider when each resource cost is likely to be incurred. This may mean in certain circumstances thatinstead of using the simplifying assumption described in paragraph 8.7, land and property costs aredisaggregated and each element is input at different stages of the appraisal period. The recommendedtreatment for land and property costs for some situations is set out in Table 6/8/1.

8.12 Table 6/8/1 distinguishes the following elements of the land and property costs of road construction:

• acquisition cost, not necessarily the money paid to the previous owner, but the District Valuer’s estimate of the current value of the land and property (converted to 2010 prices);

• legal transaction costs, the amount paid to estate agents/solicitors to deal with the paper work relating to the acquisition (excluding stamp duty);

• property management costs, the costs of managing and maintaining the land and property before it is used for the scheme. It is usually desirable to keep agricultural land farmed and properties occupied rather than let them go unused so there may also be a transaction cost of renting out the land or property to be included;

• resale value of surplus (that is off-line) land and property. The District Valuer’s estimate of the current value of land and property which Transport Scotland might



buy but would be surplus to the scheme is entered in the appraisal as a negative cost at the point in time when it is likely to be re-sold.

Notes:

(i) If the property is taken out of use (either demolished or boarded up) long beforeconstruction begins the land it stood/stands on still has a potential value for alternativeuse. This may be significant in areas where property values are high. In such a case ashare of the acquisition cost should be input when the property is taken out of use andthe remainder input when construction begins.

8.13 In addition to the costs set out above there will be compensation costs which have to be included in theappraisal. There will be compensation payments linked to all the types of land purchase set out in Table6/8/1. Moreover, compensation for injurious affection is also paid in respect of depreciation in value ofland and property which Transport Scotland does not purchase but which is affected by the scheme. Inaddition, third party claims may arise, during the construction period, which are borne by the ScottishExecutive.

8.14 For land and property purchased by Transport Scotland a compensation payment may be made to theprevious owners to reflect the loss in value of the land and property caused by the scheme. This wouldbe equivalent to change in the market value of the land and property caused by the scheme. In addition,various compensation payments may be made to the previous owners to reflect the costs andinconvenience associated with adjusting to the changes brought about by the scheme.

8.15 Transport Scotland will also pay compensation in respect of land and property it does not purchase.Compensation may be paid to land and property owners to offset the costs of noise insulation which isinstalled during scheme construction or on the opening of the road. Transport Scotland is also liable topay claims under Part I of the Land Compensation Act between one and seven years after the roadopens. Note that the resource cost of all Part I payments is incurred when the road opens.

Table 6/8/1: The Treatment of Land and Property Costs

Type of Purchase and Treatment of Land and

Property

Cost to be Input to NESA Appraisal

Land and Property - on line• not farmed/unoccupied

from purchase to construction

all costs input at time of purchase(consisting of acquisition cost, legal transaction costs and any property management costs)(i)

Land and Property - on line• farmed/occupied until

construction begins

costs input at time of purchase:• legal transaction costs of purchase• estimated transaction costs of renting, if available• property management costs

acquisition costs to be input when taken out of use for demolition(i)

Land and Property - off line• not farmed/unoccupied and

later resold

cost input when taken out of use:• full purchase costs (i.e. acquisition & legal transaction costs)

cost input when resold and returned to use:• resale value

Land and Property - off line• farmed/occupied and later

resold

costs input at time of acquisition:• transaction costs of purchase• property management costs



8.16 The likely compensation costs associated with a scheme may be estimated using examples ofcompensation paid in previous schemes. Strictly, cost benefit analysis requires each compensationpayment should be entered in the appraisal at the point when the resource cost was incurred. However,it is probably reasonable to simply the analysis by entering the estimate of all the elements ofcompensation at a single point during scheme construction.

Optimism Bias and Risk Assessment

8.17 One of the fundamental considerations of the assessment process is that of Optimism Bias. OptimismBias is the term used to reflect a tendency for the true capital cost, operational cost or works duration ofschemes in the public sector to be underestimated thereby overestimating the benefits of the scheme.

8.18 Consequently, there is a need, particularly at the initial stages of the assessment, to provide morerealistic cost estimates. To this end, the appraisal process now includes the requirement to apply anuplift to the capital costs of the works and an extension to the project programme. Optimism Bias willgenerally be higher at the early stages of an assessment when details of any proposed scheme, and theassociated risks, will generally be less well defined and lower at the later stages of assessment when thedetails of the proposed scheme and associated risks are better defined. Optimism Bias must be appliedto all scheme assessments. Optimism Bias is not a sensitivity test.

8.19 For Transport Scotland trunk road schemes, the levels of Optimism Bias generally applied are outlinedin Table 6/8/2. Users should confirm the levels of Optimism Bias to be applied at each stage of ascheme's assessment with their Overseeing Organisation. Further general advice regarding OptimismBias can be found in TAG Unit 3.5.9, The Estimation and Treatment of Scheme Costs (January 2014).

Sources: *Flybjerg (2004) & Mott MacDonald (2002), otherwise Transport Scotland

8.20 All scheme cost estimates should include an assessment of the risks involved. Transport Scotlandexpects a Quantified Risk Assessment (QRA) to be undertaken and encourages this as best practice(see TAG Unit 3.5.9, The Estimation and Treatment of Scheme Costs - January 2014). Where a QRA

Table 6/8/2: Optimism Bias Adjustment Factors Generally Applied for Different Stages of Assessment of Trunk Road Schemes

Stage /Factors

Preliminary Assessment

DMRBStage 1

Route Option Assessment /

Pre-Feasibility DMRBStage 2

Preferred Scheme

Assessment / Pre-Order Publication

DMRBStage 3

Pre-Tender Assessment /

Order Publication

Post-Tender Assessment /

Works Commitment

Cost Factor (Roads)

44%* 25% 15%* 15% 3%

Works’ Duration Factor (Roads)

20%* 10% 10%* 5% 1%*

Cost Factor (Fixed Links i.e.

bridges and tunnels)

66%* 44% 23%* 23% 6%*

Works’ Duration Factor (Fixed

Links i.e. bridges and

tunnels)

25%* Agree with TS Agree with TS Agree with TS 3%*



has been carried out and project costs include an allowance for risk, the size of the Optimism Biasadjustment required may be reduced, but this will require the prior approval of Transport Scotland.Allowance for risk is not however a complete substitute for Optimism Bias adjustment.

Volume 15 Section 1 Chapter 9Part 6 Valuation of Costs and Benefits The Preparation of Scheme Cost Data for Use in NESA


9 THE PREPARATION OF SCHEME COST DATA FOR USE IN NESA

9.1 NESA can accept cost data as either a Total Scheme Cost (TSC) or itemised into four components(construction, land, preparation and supervision). The way in which scheme costs are often spread overseveral years is modelled in NESA through a profile of expenditure input by the user. The remainder ofthis chapter presents the methodology used to calculate total scheme costs, an example of which iscontained in Chapter 10.

9.2 The items described in estimates as making up the total scheme cost are generally grouped as follows:

1. CONSTRUCTION COSTS

(i) Main Works Contract (including preliminaries, roadworks general, earthworks,main carriageways, inter-changes, side roads, signs, etc.)

Roadworks Sub-total

Structures Sub-total

Main Works Totalcontingencies (10%, 15% or 20%)

Main Works Contract Total

(ii) Ancillary Works Contract (including maintenance compounds, lighting, motorwaycommunications, landscaping, noise insulation, etc.)

Ancillary Works Totalcontingencies (10%, 15% or 20%)

Ancillary Works Contract Total

(iii) Works by Other Authorities (including Network Rail, local authorities, statutoryundertakers)

Works by Other Authorities Total contingencies (10%, 15% or 20%)

Works by Other Authorities Total

Total Construction Costs (i) + (ii) + (iii) Adjustment Factor for Optimism Bias (%)

Total Adjusted Construction Costs

2. LAND AND PROPERTY COSTSLand Costs

Adjustment Factor for Optimism Bias (%)

Total Adjusted Land Costs

Chapter 9 Volume 15 Section 1The Preparation of Scheme Cost Data for Use in NESA Part 6 Valuation of Costs and Benefits


3. PREPARATION AND ADMINISTRATION

4. ON SITE SUPERVISION AND TESTING

Note: VAT should be excluded from cost estimates to be input to NESA.

9.3 The construction cost groupings are largely self explanatory and should not give rise to difficulties ofinterpretation or valuation when preparing NESA input. Land costs, however, can present particularproblems (see Part 6 Chapter 8).

Preparation Costs

9.4 Preparation costs include Consulting Engineers’/Agent Authorities’ fees, actual costs of pursuingalternative routes (if any) in the early stages of the scheme, Public Consultation, Public Inquiry and thecost of any surveys carried out during scheme preparation. It is recognised that preparation costs cannotbe predicted with any accuracy in a scheme’s early stages and in the absence of specific costs thefollowing defaults should be used:

(1) Stage 1 Preliminary Assessment 12% of scheme construction, landand property costs

(2) Stage 2 Route Option Assessment 9% of scheme construction, landand property costs

(3) Stage 3 Preferred Scheme Assessment 6% of scheme construction, landand property costs

(4) Works Commitment Stage (pre-tender) 2% of scheme construction, landand property costs

The above defaults have been estimated from data collected from the roads programme overall. Thesevalues are therefore representative for the average scheme. Preparation costs are assumed to occurevenly from the present day to the start of construction.

Supervision Costs

9.5 Supervision costs are those associated with the cost of site staff and include a percentage for on sitetesting of materials. Recent research has shown that these costs are, on average, 5% of schemeconstruction, land and property costs. Supervision costs should be spread evenly throughout thecontract period.

Delays During Construction and During Future Routine Maintenance

9.6 In addition to the direct costs of construction, allowances may have to be made for delays imposedupon existing traffic during construction of a scheme (see Part 6 Chapter 12). Where such delays arelikely to be significant, it may be appropriate to model them. For urban schemes only, it maysometimes be possible to compute the delays from the same mathematical model that was used topredict the overall traffic effects of the scheme. The delays may be costed using a separate NESA run,comparing user costs with reduced capacity or diversionary routes during construction, with user costsin the absence of such delays. Alternatively, QUADRO may be used to model and estimatediversionary delays (see Part 6 Chapter 11). Where diversion and network effects are not likely toprove significant, a simpler manual calculation should suffice. This will involve an estimate of totaldelays (by vehicle category) and costing of these delays using the values of time given in the TransportAppraisal Guidance (TAG data book, November 2014).

Volume 15 Section 1 Chapter 9Part 6 Valuation of Costs and Benefits The Preparation of Scheme Cost Data for Use in NESA


Present Value Year

9.7 All items in NESA, including scheme costs, are measured at the constant prices of a given base year,the Present Value Year. In NESA the Present Value Year is 2010. This means the effect of generalinflation, that is the increase in the average prices of all goods and services, is excluded. In NESA theConsumer Price Index (CPI) is used to measure general inflation and should be used to convert usercosts/benefits to different price bases.

9.8 For construction costs, account must be taken of any change in the cost of road construction relative tothe general price level. This was previously done in NESA using the Relative Price Factor. TAG Unit3.5.9 - The Estimation and Treatment of Scheme Costs, January 2014 recommends the use of inflationrates relevant to the delivery of transport schemes and this should be used in the preparation of basecost inputs for NESA from now on.

9.9 The construction costs input to NESA should reflect the most up to date estimate available. The date ofthe estimate and the equivalent CPI value should also be input. For example for a total constructioncost estimate of £2.34M at April 2014 (CPI=128.1) the inputs to NESA would be:

SCHEME_COSTS, TSC=2340/4/2014/128.1, END.

9.10 See Part 10, Section 1 Chapter 15 for further details of inputting scheme construction costs into NESA.The latest CPI values are available from the Office for National Statistics website. See www.ons.gov.ukfor detailed CPI Reference Tables.

9.11 Construction costs can be converted by means of the CPI to the average 2010 prices:

latest available cost x114.5=construction cost in average 2010 pricesCPIM

where CPIM represents the CPI at the date of the latest available cost e.g. if latest cost is at April 2014,CPIM=128.1 (www.ons.gov.uk). Table 6/9/1 gives the CPI values from 2010 to 2014 (seewww.ons.gov.uk for CPI values from 1998).

9.12 The allocation of construction costs in the correct proportion for each year during which constructiontakes place cannot always be precise, especially when a scheme is in its early stages and several optionsare being considered. However, an accurate profile of construction costs covering the contract periodshould be used whenever possible, particularly when the contract period stretches over three or fouryears. Table 6/9/2 shows the default distribution of construction costs (excluding land costs) overvarying construction periods based on the average of a large number of schemes.

Table 6/9/1: CPI Values from 2010 to 2014

Year CPI Values by MonthJan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2010 112.4 112.9 113.5 114.2 114.4 114.6 114.3 114.9 114.9 115.2 115.6 116.82011 116.9 117.8 118.1 119.3 119.5 119.4 119.4 120.1 120.9 121.0 121.2 121.72012 121.1 121.8 122.2 122.9 122.8 122.3 122.5 123.1 123.5 124.2 124.4 125.02013 124.4 125.2 125.6 125.9 126.1 125.9 125.8 126.4 126.8 126.9 127.0 127.52014 126.7 127.4 127.7 128.1

Chapter 9 Volume 15 Section 1The Preparation of Scheme Cost Data for Use in NESA Part 6 Valuation of Costs and Benefits


9.13 The final result of a NESA analysis is expressed in terms of its Net Present Value (NPV) and Benefit toCost Ratio (BCR). The NPV is the Present Value of Benefits (PVB) minus the Present Value of Costs(PVC)

NPV = PVB - PVC

and the BCR is the PVB divided by the PVC

BCR = PVB / PVC

9.14 During the course of a scheme’s design, the cost estimates may be revised, without an alteration in theestimate of benefits (PVB). If this is the case, it is not necessary to re-run NESA. The revised NPV andBCR can be calculated using the original PVB and revised PVC

Table 6/9/2: Default Profile for Construction Costs (Excluding Land Costs)

Percentage of Total CostContract Period

YearsYears Before Opening First Scheme

YearTotal

4 3 2 11.5 29 68 3 1002.0 47 50 3 1002.5 16 42 39 3 1003.0 30 34 33 3 1003.5 11 29 30 27 3 1004.0 22 25 25 25 3 100

Volume 15 Section 1 Chapter 10Part 6 Valuation of Costs and Benefits An Example of Scheme Cost Inputs


10 AN EXAMPLE OF SCHEME COST INPUTS10.1 The calculation of Total Scheme Costs can be broken down into four stages:

(i) converting the construction and land cost estimates to 2010 prices

(ii) calculating preparation and supervision costs

(iii) creating a profile of scheme costs over the preparation and construction period

(iv) discounting these costs to give a Present Value of Costs at 2010 prices

The fourth stage (the discounting of total scheme costs to a PVC at 2010 prices) can be undertakenwithin NESA.

10.2 A standard calculation sheet (SHEET 1) and a summary sheet (SHEET 2) for construction costs can befound at the end of this chapter, a copy of which should be included with the economic assessmentreport (see Part 9 Chapter 3). In the following example numbers in brackets refer to the items [1 to 4]on the standard calculation sheet (e.g. [2]).

10.3 In this example, the latest available estimate for a scheme is based on evidence of rates from March2012. This estimate is £14.30m excluding VAT [1]. From Table 6/9/1, the CPI for March 2012 is 122.2.Hence, scheme cost in 2010 prices is:

£14.30m x 114.5/122.2 = £13.40m ....[2]

where 114.5 is the average value of the CPI in 2010.

10.4 The District Valuer’s estimate (December 2013) puts land costs at £1m. [3]. This estimate has to beconverted to a 2010 price by multiplying by the change in the CPI between December 2013 and 2010.The value of the index in December 2013 was 127.5. The estimate of land costs in 2010 prices istherefore:

£1m x 114.5/127.5 = £0.90m ...[4]

Entry to land is made in the first year of construction, so this amount must be added to the constructioncosts for that year.

10.5 The scheme is currently at Order Publication Report stage (Stage 3 Preferred Scheme Assessment). Asexplained in Part 6 paragraph 9.4, 6% of scheme construction and land costs must be input to NESA forpreparation costs as follows:

(£13.40m + £0.90m) x 0.06 = £858k

The projected start of construction date is 2017 and scheme preparation costs must be spread evenlythroughout the period from now (in this example 2014) to the start of construction as follows:

2014 = £286k2015 = £286k2016 = £286k

Chapter 10 Volume 15 Section 1An Example of Scheme Cost Inputs Part 6 Valuation of Costs and Benefits


10.6 Supervision costs are assumed to be on average 5% of the total scheme cost in 2010 prices. Supervisioncosts are therefore:

(£13.40m + £0.90m) x 0.05 = £715k

These are spread evenly throughout the contract period and opening year of the scheme as follows:

1st year (2017) = £238k2nd year (2018) = £238kopening year (and maintenance period) (2019) = £238k

10.7 Reference may now be made to Table 6/9/2 to estimate the distribution of construction costs over theconstruction period. This assumes that no distribution more appropriate to the specific scheme can beobtained. Assuming that the contract period is 18 months, it can be seen that 29% of construction costscan be assumed to be incurred in the first year of construction, 68% in the second year, and 3% in thefirst year of opening of the scheme. Thus we have:

1st year of construction = 0.29 x £13.40m = £3.886m2nd year of construction = 0.68 x £13.40m = £9.112mopening year = 0.03 x £13.40m = £0.402m

10.8 Having calculated the total undiscounted cost profile for the scheme, these costs should be discountedat 3.5% to give the PVC of the scheme at 2010 prices, discounted to 2010. Table 6/10/1 summarises theresults of this example.

Table 6/10/1: Example of Scheme Cost Profile (£ thousands in 2010 prices)

Programme Year Construction Cost

Land & Property

Cost

Preparation Cost

Supervision Cost

Total Undiscounted

Cost

Discount Factor

Total Discounted

CostPresent Year 2014 286 286 0.871 249

2015 286 286 0.842 2412016 286 286 0.814 233

Start Const. 2017 3,886 900 238 5,024 0.786 3,9492018 9,112 238 9,350 0.759 7,097

Year Open 2019 402 238 640 0.734 470Total 13,400 900 858 715 15,872 12,239

Volume 15 Section 1 Chapter 10Part 6 Valuation of Costs and Benefits An Example of Scheme Cost Inputs


SHEET 1 - CALCULATION SHEET FOR SCHEME COST

SCHEME____________________________________________________DATE_______________________

OPTION_________________________________________________________________________________

1. CONSTRUCTION COSTS (see Part 6 Table 6/9/2 for phasing of costs)

Latest available estimate of construction costs (excluding VAT):= £ ………… [1]

Year and quarter correspondingto rates used in estimate:year: …….quarter: …….CPI for above year and quarter: CPI [Constr] = ............

Estimate of construction = (114.5 x [1]) = £ ………… [2]costs in 2010 prices CPI [Constr]

2. LAND AND PROPERTY COSTS (see Part 6 Chapter 8 for phasing of costs)

Current DV estimate: = £ ………… [3]

Year and quarter of estimate: year: ……..quarter: ……..

CPI for date of estimate: CPI [land] = .............

Estimate of land costs in 2010 prices =114.5 x [3] = £ ………… [4] CPI [land]

3. PREPARATION COSTS (see Part 6 paragraph 9.4 for phasing of costs)

Proportion of construction and land cost dependingon stage of scheme: Prop = .......... %

Preparation cost = Prop x ([2] + [4]) = £ …………

These costs are incurred from the date of this calculation to start of construction.

4. SUPERVISION COSTS (see Part 6 paragraph 9.5 for allocation of costs)

5% of construction and land cost

Supervision cost = 0.05 x ([2] + [4]) = £ …………

These costs are incurred evenly from the start of construction to the end of the maintenance period (see Part 6 paragraph 8.5).

Chapter 10 Volume 15 Section 1An Example of Scheme Cost Inputs Part 6 Valuation of Costs and Benefits


SHEET 2 -

SCHEME____________________________________________________DATE_______________________

OPTION_________________________________________________________________________________

NOTES: i) Column 7 values are the undiscounted total costs at 2010 prices.ii) Columns 8 and 9 are for a manual calculation of total discounted costs.

Programme

(1)

Year

(2)

Construction Cost(3)

Land and Property Cost(4)

Preparation Cost(5)

Supervision Cost(6)

TOTAL COST

(7)

Discount Factor

(8)

Discounted Cost(9)

TOTALS TOTAL

Volume 15 Section 1 Chapter 11Part 6 Valuation of Costs and Benefits Road Maintenance


11 ROAD MAINTENANCE11.1 Since NESA is concerned with evaluating differences in the overall costs of operating a Do-Minimum

and any number of Do-Something networks it is important to appraise the change in the costs to boththe highway authority and the road user of maintaining each network under consideration.

11.2 Maintenance costs can be divided into two broad categories for economic appraisal purposes, referredto as non-traffic related costs (Group 1) and traffic related costs (Group 2).

Non-Traffic Related Costs (Group 1)

11.3 Group 1 costs are comprised of the following items: drainage, street lighting footway/cycle tracks,safety fence/barrier, boundary fences, bridges/culverts/subways, remedial earthworks, vergemaintenance, sweeping, gulley emptying, signals/signs/crossings, road markings, salt/snow-plough/fencing and motorway compounds.

11.4 Although some of the above Group 1 items are strictly speaking traffic related (Group 2) they aregenerally judged to incur relatively small amounts of expenditure and as such have been included inGroup 1.

11.5 Group 1 costs can generally be described as an annual charge per unit length of road. The differencebetween the Do-Minimum and Do-Something Group 1 costs is usually small, and tends to favour theDo-Minimum because the Do-Something adds to the length of road which has to be maintained.

11.6 These costs are irregular items in that higher expenditure is incurred in some years than others. Forcomputational convenience, NESA treats this uneven flow of costs as an equivalent stream of constantcosts arising in each year of the appraisal period.

11.7 The non-traffic related maintenance costs used in NESA are detailed in Table 6/11/1.

Traffic Related Costs (Group 2)

11.8 Group 2 costs are comprised of both the works and traffic delay costs associated with the followingitems: reconstruction, overlay, resurfacing, surface dressing and patching.

11.9 It can be assumed that when a vehicle reassigns from an old road to a new road, it ceases to imposewear and tear on the old road and instead imposes it on the new road. For the purposes of the economicassessment of road schemes, net change in Group 2 maintenance costs between the Do-Minimum andDo-Something situations is assumed to be insignificant. This is still a reasonable assumption in mostcases where the old road is known to be in a sound structural condition.

11.10 However, this is unlikely to be a reasonable assumption where different major maintenance works areplanned in the foreseeable future in the Do-Minimum compared with the Do-Something. This could bethe case if the existing road is in, or is deteriorating into, a poor structural condition. In this case morecostly maintenance works may be needed in the Do-Minimum compared to the Do-Something withlower traffic flows on the old road, and a different kind of assessment needs to be made. The NESAuser should set out explicitly the assumptions made about future maintenance work.

11.11 In addition to these works costs, there are user costs associated with the delay to traffic whilst theworks are carried out. These are the time delays forecast to occur at future maintenance works sites,and the associated vehicle operating and accident costs.

Chapter 11 Volume 15 Section 1Road Maintenance Part 6 Valuation of Costs and Benefits


Table 6/11/1: Non-Traffic Related Maintenance Costs (2010 values and prices)

MaintenanceNESA Road

Category

Description Type Standard Cost (£/km)

1 Urban - single 6.0m 1 S2 9,4002 Urban - single 7.3m 1 S2 9,4003 Urban - single 10.0m 1 S2 9,4004 Urban - single 4 lanes 1 S2 9,400

5 Urban - one way 6.0m 1 S2 9,4006 Urban - one way 7.3m 1 S2 9,4007 Urban - one way 10.0m 1 S2 9,400

8 Urban - dual 2 2 D2AP 13,2009 Urban - dual 3 3 D3AP 16,40010 Urban - dual 4 3 D3AP 16,400

11 Urban - Expressway, 2 or more lanes 3 D3AP 16,40012 Urban - Motorway and dual ramps, 1 lane 4 D2M 21,70013 Urban - Motorway and dual ramps, 2 lanes 4 D2M 21,700

14 Urban - Motorway - D2 4 D2M 21,70015 Urban - Motorway - D3 5 D3M 24,70016 Urban - Motorway - D4 6 D4M 24,70017 Urban - Motorway - D5 6 D4M 24,700

20 Rural - single 4 lane 2 D2AP 13,200

21 Rural - poor single 4.0m 1 S2 9,40022 Rural - poor single 5.5m 1 S2 9,40023 Rural - poor single 6.0m 1 S2 9,40024 Rural - typical single 6.0m 1 S2 9,400

25 Rural - poor single 7.3m 1 S2 9,40026 Rural - typical single 7.3m 1 S2 9,40027 Rural - good single 7.3m 1 S2 9,400

28 Rural - typical single 10.0m 1 S2 9,40029 Rural - good single 10.0m 1 S2 9,40030 Rural - single with climbing lane 1 S2 9,400

31 Rural - dual 2 lanes 2 D2AP 13,20032 Rural - dual 3 lanes 3 D3AP 16,40033 Rural - dual 2 with climbing lane 3 D3AP 16,400

34 Rural - dual 2 lanes with grade separation 2 D2AP 13,20035 Rural - dual 3 lanes with grade separation 3 D3AP 16,40036 Rural - dual 2 lanes with GS and climbing lane 3 D3AP 16,400

37 Rural - Motorway and dual ramps, 1 lane 4 D2M 21,70038 Rural - Motorway and dual ramps, 2 lanes 4 D2M 21,700

39 Rural - Motorway - D2 4 D2M 21,70040 Rural - Motorway - D3 5 D3M 24,70041 Rural - Motorway - D2 with climbing lane 5 D3M 24,700

Volume 15 Section 1 Chapter 11Part 6 Valuation of Costs and Benefits Road Maintenance


Delays During Roadworks

11.12 The computer program QUADRO should normally be used to assess traffic related user delay costsduring roadworks (externally to the NESA program) and the results manually incorporated with theNESA results, to produce aggregate Net Present Values for selected traffic forecasts.

11.13 The QUADRO Manual (DMRB 14.1) describes in detail how Group 2 costs should be assessed.However, the basic QUADRO method is described below so that in those cases where maintenancecosts for Group 2 are significantly different between the Do-Minimum and Do-Something, users cansee how the two programs can be used in conjunction with each other.

11.14 Given the physical layout of a site which is to be the subject of extensive maintenance work, togetherwith predicted future traffic flows in each direction, a 24 hour flow profile, the duration of the works,information about any diversion routes and a works cost profile, the QUADRO program will computethe Present Value of both works and traffic delay costs of any future maintenance profile defined by theuser. Clearly the results obtained will depend upon the validity of the assumptions made about the needfor and the type of maintenance to be carried out during the 60 year appraisal period and the likelytraffic management arrangements during maintenance works. Approval from Transport Scotland forthe proposed maintenance type and traffic management arrangements must be obtained before theprogram is used. In all cases the program should be run for both the Do-Minimum and Do-Somethingoptions. Care will be needed when deciding how many existing links in the Do-Minimum option areaffected by substantial reassignment of traffic to the new link(s), since in general only a fairlysubstantial change in traffic flow will affect predicted maintenance profiles. Therefore only existinglinks with substantial changes should be considered for the purpose of this exercise.

11.15 Maintenance operations should be estimated on the assumption that they will be conducted on acost-effective basis in relation to both works costs and traffic delay costs. Clearly, works which areexpensive in terms of both expenditure and user costs should not be undertaken until all alternativeoptions (for example, night time working) have been investigated.

11.16 Future traffic flows should be consistent with those used in the appropriate NESA analysis. This isparticularly relevant where traffic flows and journey times derived from the traffic appraisal are to beused directly in the economic appraisal (see also Part 2 Chapter 3). Values of economic parameters andthe discount rate built into the QUADRO program are consistent with NESA.

11.17 The principles of economic appraisal and the QUADRO method are best illustrated by way of anexample. Suppose the road scheme in question bypasses an existing stretch of road which is in a poorstate of repair. In the Do-Minimum situation it is estimated that a major reconstruction of the existingcarriageway will be required in 2015 followed by resurfacing in 20 years time and again in 2055.

11.18 In the Do-Something situation traffic volumes would be reduced on the existing road so thatresurfacing need not be considered for 10 years. It is estimated that the new road will need resurfacingin 20 years time and again in 2065. The Do-Minimum and Do-Something options display the followingestimated maintenance profiles:

Chapter 11 Volume 15 Section 1Road Maintenance Part 6 Valuation of Costs and Benefits


11.19 All costs used in the maintenance profile should use current estimates, deflated to 2010 prices using theConsumer Price Index (see Table 6/9/1).

11.20 The effects of the preferred maintenance profile (that is, usually the lowest maintenance PVC) iscombined with the scheme capital costs and benefits from the NESA program by manual adjustment.This will allow the final scheme NPV and BCR to be calculated. It should be noted that the capital costof the maintenance program is part of a scheme’s Present Value of Costs (PVC), whilst the user costsimposed (e.g. delays and accidents) during construction and routine maintenance are part of the PresentValue of Benefits (PVB).

Table 6/11/2: Estimated Maintenance Works Cost Profiles (average 2010 prices)

DO-MINIMUMExisting Road 2015

3 month reconstruction£1,269,000

20351 month resurfacing

£380,700


£1,269,000DO-SOMETHING

Existing Road 20251 month resurfacing

£380,700


£634,500New Road 2035

1 month resurfacing£380,700

20652 month overlay

£951,750

Volume 15 Section 1 Chapter 12Part 6 Valuation of Costs and Benefits Delays During Construction


12 DELAYS DURING CONSTRUCTION12.1 Cost benefit analysis is concerned with evaluating differences in the overall network costs of

Do-Something options against the Do-Minimum and any assessment therefore must include user costsimposed by the road construction itself. For a simple village by-pass where the only disruption totraffic could be at the tie-ins with the existing road the construction delay may be minimal. However,with schemes such as on-line motorway widening, construction delays or their avoidance, form a majorpart in the decision making process.

12.2 The assessment of construction delays is important in both the estimation of overall benefits and theranking of options. Therefore, an analysis or reasons for assuming insignificant dis-benefits, must beincluded in all economic appraisals.

12.3 The assessment of delays during construction is usually carried out using QUADRO (DMRB 14.1).The discounted QUADRO results can then be manually incorporated with the NESA results to give theoverall scheme NPV and BCR.

Chapter 12 Volume 15 Section 1Delays During Construction Part 6 Valuation of Costs and Benefits




VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


SPEEDS ON LINKS

Contents

Chapter

1. The NESA Speed/Flow Types

2. Rural Single Carriageways (Speed/FlowType 1)

3. Rural All-Purpose Dual Carriageways andMotorways (Speed/Flow Types 2-6)

4. Urban Roads (Speed/Flow Types 7 and 8)

5. Suburban Roads (Speed/Flow Types 9 and10)

6. Small Town Roads (Speed/Flow Type 11)

7. Single Track Roads (Speed/Flow Type 12)

8. Treatment of Overcapacity on Links

9. Representative Diagrams of Speed/FlowRelationships

PART 7

Volume 15 Section 1Part 7 Speeds on Links


Volume 15 Section 1 Chapter 1Part 7 Speeds on Links The NESA Speed/Flow Types


1 THE NESA SPEED/FLOW TYPES1.1 Savings in journey time are generally the main source of NESA benefits following a road

improvement. The assessment of time savings requires knowledge of likely traffic flows andknowledge of the behaviour of the network under varying traffic loadings. It is the flow dependence oftraffic speeds, and their variability from one occasion to another, that limits the direct use ofobservations of speeds on existing roads in the evaluation process. The NESA approach is therefore touse nationally-derived speed/flow relationships wherever possible to predict traffic speeds.

1.2 The fixed link speeds that form the basis of the assignment (see Part 5 Chapter 3) are not used in theevaluation process, because though they are accurate enough to produce realistic routeings and trafficflows the speeds are too coarse to produce accurate estimates of travel time changes. When validating aNESA evaluation it is therefore essential to ensure a close match between the link speeds used in theassignment model and the speeds calculated from the NESA speed/flow relationships.

1.3 There are twelve types of speed/flow relationship used in NESA (see Table 7/1/1). Types 1 to 6 areused for all-purpose roads and motorways that are generally not subject to a local speed limit, whileTypes 7 and 8 are used for roads in towns or conurbations subject to 30 mph (48 kph) speed limits only.Types 9 and 10 are used for major suburban routes in towns and cities that are generally subject to a 40mph (64 kph) speed limit. Type 11 is used in small towns or villages for routes subject to a 30 mph (48kph) or 40 mph (64 kph) speed limit. Type 12 relates to single track roads subject to 60 mph (96 kph)speed limit.

The NESA Speed/Flow Types and the Modelling of Junction Delays

1.4 The method of predicting speeds in NESA differs between rural roads, where there is little interactionbetween different links and junctions on the network, and urban and suburban roads, where delays atjunctions tend to be inter-dependent. On rural roads and motorways, relationships are used to predictthe speed of traffic on each link according to link geometry and traffic flow; delays at junctions areseparately assessed using junction delay models (see Part 8). In urban areas, the road network has to beconsidered as a system rather than as a set of links and junctions. Accordingly NESA uses area wideurban and suburban speed relationships based on average journey speeds observed in towns andconurbations in the UK. For small town and single track roads NESA models an average speed for thelink. The relationships used also specifically exclude junction delays.

Table 7/1/1: NESA Speed/Flow Types

Speed / Flow Type

Description

1 Rural single carriageway2 Rural all-purpose dual 2-lane carriageway (D2AP)3 Rural all-purpose dual 3 or more land carriageway (D3AP)4 Motorway, dual 2-lanes (D2M)5 Motorway, dual 3-lanes (D3M)6 Motorway, dual 4 or more lanes (D4M+)7 Urban, non-central8 Urban, central9 Suburban single carriageway10 Suburban dual carriageway11 Small town12 Single track

Chapter 1 Volume 15 Section 1The NESA Speed/Flow Types Part 7 Speeds on Links


The Functional Form of the NESA Speed/Flow Types

1.5 The basic form of the speed/flow relationships varies between the NESA speed/flow types. For rural,suburban and small town roads the speed of vehicles reduces as flow increases until a critical flow levelis reached at which the rate of speed reduction increases until a minimum speed cut-off is reached. Therelationships for urban roads have a uniform negative speed/flow slope for all flow levels above theminimum speed constraint. The other major difference is that rural and suburban relationships provideseparate estimates of the average journey speeds of light vehicles and heavy vehicles, while the urban,small town and single track relationships provide a single estimate of the average vehicle speed. Lightvehicles are defined as cars and light goods vehicles (LGV); heavy goods vehicles are defined as othergoods vehicles (OGV1 and OGV2), buses and coaches (COACH).

1.6 The relationships can predict speeds above the legal speed limit for the particular road categoryconsidered. If this occurs the speed of the vehicle type considered is reduced to the legal speed limitbefore any economic calculations are made. All relationships are subject to a minimum speed cut-offwhich varies by the type of the speed/flow curve.

1.7 When flows reach a particular level on a link NESA can produce an overcapacity report (SEE Part 10Chapter 16). It is a signal to the user that flows are about the highest levels that could normally beexpected on a link of this standard. The levels of the capacity flags (QC) for each road class aredetailed in the following chapters. The treatment of overcapacity links is described in Part 7 Chapter 8together with advice to the user regarding what action is required.

Defining a Link’s Speed/Flow Relationship

1.8 NESA allocates an individual speed/flow relationship to every link in the network. This is donethrough the allocation and refinement of a speed/flow type for each link. The refinement process isbased upon the geometric properties of the link.

1.9 Table 7/1/2 shows the process NESA uses to allocate a speed/flow type to a link. As can be seen fromthis table the determining factors are the road category, the speed limit and the presence of either acentral (C) or small town (ST) flag.

1.10 Table 7/1/3 indicates the geometric properties that affect the speed/flow type. As can be seen from thistable these differ by speed/flow type.

1.11 All the speed/flow types used in NESA are based on empirical data. To ensure the validity of aspeed/flow relationship it is important that a link’s geometric properties do not exceed the valueswhich were observed when defining that relationship. The range of acceptable values for eachgeometric variable are defined in a table in the relevant speed/flow type chapter. Therelationships cannot necessarily be taken to apply outside the given ranges of the variables. Thegeometric variables should all be averaged over a reasonable length of link, generally not lessthan two kilometres.



C - Denotes Central flag in the link descriptionST - Denotes Small Town flag in the link description

Note: The Central flag and Small Town flag are important in defining speed flowtypes (see also Paragraph 5.3.13 and Table 5/3/1)

Table 7/1/2: NESA Road Categories and Speed/Flow Types

NESA Road

Category

Description Speed Limit NESA Speed /

flow type

Speed/Flow Description

(mph) (kph)1 Urban - single 6.0 m 30C 48C 8 Urban Central

30 48 7 Urban Non-Central40/50 48/64 9 Suburban Single

Carriageway30/40ST 48/64ST 11 Small Town

2 Urban - single 7.3 m 30C 48C 8 Urban Central30 48 7 Urban Non-Central

40/50 64/80 9 Suburban Single Carriageway

30/40ST 48/64ST 11 Small Town

3 Urban - single 10.0 m 30C 48C 8 Urban Central30 48 7 Urban Non-Central



4 Urban - single 4 lane 30C 48C 8 Urban Central30 48 7 Urban Non-Central



5 Urban - one way 6.0 m 30C 48C 8 Urban Central30 48 7 Urban Non-Central











Table 7/1/3: [Contd] NESA Road Categories, Link Speeds and Link Capacities

NESA Speed Limit NESARoad Description Speed / Speed / Flow Description

Category (mph) (kph) flow type8 Urban - dual 2 30/40/50 48/64/80 10 Suburban Dual Carriageway9 Urban - dual 3 30/40/50 48/64/80 10 Suburban Dual Carriageway10 Urban - dual 4 30/40/50 48/64/80 10 Suburban Dual Carriageway

11 Urban - Expressway, 2 or more lanes 30/40/50 48/64/80 10 Suburban Dual Carriageway12 Urban - Motorway and dual ramps, 1 lane 30/40/50 48/64/80 1 Rural Single Carriageway13 Urban - Motorway and dual ramps, 2 lanes 30/40/50 48/64/80 2 D2AP

14 Urban - Motorway - D2 50/60/70 80/96/113 4 D2M15 Urban - Motorway - D3 50/60/70 80/96/113 5 D3M16 Urban - Motorway - D4 50/60 80/96 6 D4M+17 Urban - Motorway - D5 50/60 80/96 6 D4M+

20 Rural - single 4 lane 60 96 1 Rural Single Carriageway

21 Rural - poor single 4.0m 60 96 12 Single Track22 Rural - poor single 5.5m 60 96 1 Rural Single Carriageway23 Rural - poor single 6.0m 60 96 1 Rural Single Carriageway24 Rural - typical single 6.0m 60 96 1 Rural Single Carriageway

25 Rural - poor single 7.3m 60 96 1 Rural Single Carriageway26 Rural - typical single 7.3m 60 96 1 Rural Single Carriageway27 Rural - good single 7.3m 60 96 1 Rural Single Carriageway

28 Rural - typical single 10.0m 60 96 1 Rural Single Carriageway29 Rural - good single 10.0m 60 96 1 Rural Single Carriageway30 Rural - single with climbing lane 60 96 1 Rural Single Carriageway

31 Rural - dual 2 lanes 70 113 2 D2AP32 Rural - dual 3 lanes 70 113 3 D3AP33 Rural - dual 2 with climbing lane 70 113 3 D3AP

34 Rural - dual 2 lanes with grade separation 70 113 2 D2AP35 Rural - dual 3 lanes with grade separation 70 113 3 D3AP36 Rural - dual 2 lanes with GS and climbing lane 70 113 3 D3AP

37 Rural - Motorway and dual ramps, 1 lane 70 113 1 Rural Single Carriageway38 Rural - Motorway and dual ramps, 2 lanes 70 113 2 D2AP

39 Rural - Motorway - D2 70 113 4 D2M40 Rural - Motorway - D3 70 113 5 D3M41 Rural - Motorway - D2 with climbing lane 70 113 5 D3M

50 Zone connector - - -



* Calculated by NESA

DESIGN STANDARD (DES)

1.12 If the road is designed to TD9 Standards (DMRB 6.1.1), rural single carriageway links only, then DESshould be set to YES.

BENDINESS (BEND)

1.13 Figure 7/1/1 shows how the variable for road bendiness is defined.

HILLINESS (HR,HF, NG)

1.14 Figure 7/1/1 shows how the hilliness variables HR and HF are defined. Net gradient NG is defined as(HR - HF). As the speed/flow relationships in NESA are applied in a directional manner, the codedvalues of HR and HF must be consistent with the A-node to B-node link specification.

Table 7/1/4: Link Geometric Properties Used in Speed/Flow Relationships

VariableName

Description Speed / Flow TypeRural Urban Suburban Small TownSingle Track

Single Dual c/wayc/way and m/ways

(Type 1) (Type 2-6) (Type 7-8) (Type 9-10) (Type 11) (Type 12)DES Is road designed to TD9 (DMRB 6.1.1)

Standards?

BEND Bendiness; total change of direction (deg/km)

HR Sum of rises per unit distance

HF Sum of falls per unit distance

NG* Net gradient (m/km) [NG=HR-HF]

CWID Average carriageway width between white line edge markings (m)

SWID Average width of hard strip on both sides, including width of white line (m)

N/A

VISI Average sight distance (m)

JUNC Side roads intersection, both directions (no/km)

INT Frequency of major intersections averaged over the main road network (no/km)

AXS Number of minor intersections and private drives (no/km)

VW Average verge width, both sides (m)

DEVEL Percentage of route with frontage development (%)

P30 Percentage of route subject to a 30mph (48kph) speed limit



Figure 7/1/1: Measurement of Road Geometry on Rural Roads



CARRIAGEWAY AND SHOULDER WIDTH (CWID, SWID)

1.15 The actual width of surfaced road is defined by two parameters. The first (CWID) is the width ofcarriageway between any continuous white lines which may or may not be delineating a hard strip. Thesecond (SWID) is the total width of any continuous edge line and hard strip.

VISIBILITY (VISI)

1.16 The average sight distance VISI is the harmonic mean of individual observations (see Figure 7/1/1),such that:

The harmonic mean is used because this reduces the effect on VISI of a few measurements of very highvisibility.

1.17 For proposed new roads, VISI should normally be calculated from engineering drawings.Measurements of sight distance should be taken in both directions at regular intervals (50 metres forsites of uneven visibility, 100 metres for sites with good visibility) measured from an eye height of 1.05metres to an object height of 1.05 metres, both points being above the road surface at the centre line ofthe carriageway. Sight distance should be the true sight distance available at any location, includingany sight distance available across verges, and outside of the highway boundary wherever sightdistance is available across embankment slopes or adjoining land. On the approach to a junction wherethe link being considered loses priority it is unnecessary to make further measurements once thejunction is within the sight distance.

1.18 For existing roads, an empirical relationship has been derived which provides estimates of VISI givenbendiness and edge details. This relationship provides the NESA default visibility parameters:

This relationship should normally be used for all existing roads for which bendiness and verge widthhave been measured. Note that the relationship is non-directional (that is it will give the same value ofVISI for both directions of travel). On long straight roads or where sight distance is available outsidethe highway boundary, the formula will often significantly underestimate VISI. VISI should then be setto 700 metres for roads with high visibility; otherwise estimates should be made from plans or sitemeasurements.

NUMBER OF JUNCTIONS (JUNC, INT, AXS)

1.19 The parameters JUNC, INT and AXS describe the number of side roads, major intersections and minorintersections on a link, respectively. It must be noted that INT has a different definition for urban andsuburban links. For suburban roads INT is specific to each section of route, whilst for urban areas INTmeasures the number of major intersections over the whole main urban road network (see Part 7Chapters 4 and 5). Major intersections will generally be roundabouts or traffic signals.

Equation 7/1/1

where n = number of observationsXi = sight distance at point i

Equation 7/1/2

VISI n/ 1/X1 1/X2 ... 1/Xn+ + + =

Log VISI 2.46 VW SWID+ /25 BEND/400–+=



VERGE WIDTH (VW))

1.20 This is defined as the average verge width along both sides of a link.

FRONTAGE DEVELOPMENT (DEVEL, P30)

1.21 The parameters DEVEL and P30 describe the level of frontage development along a link/route.DEVEL is defined as the percentage of the length of a link that has frontage development, countingbusiness and residential development as 100% and open space as 0%. P30 is defined as the percentageof a route that is subject to a 30 mph (48 kph) speed limit. For urban links (speed/flow types 7 and 8)the value of DEVEL should be common throughout a particular area, whilst for small town links(speed/flow type 11) the values of DEVEL and P30 should be common along a route.

Default Link Geometric Variables

1.22 The default link geometric variables used in NESA are detailed in Table 7/1/5.



Notes: (1) For road category = 1, 2, 3 and 5, lanes = 1For road category = 4 and 6, lanes = 2For road category = 7, lanes = 3

Table 7/1/5: Default Link Geometric Variables used in Speed/Flow relationships by Road Category

Road Description

Category No. of Lanes

DES BEND HR HF CWID SWID VWID VISI JUNC INT AXS DEVEL P30 Min Speed

1 to 7 Central (urban, 30mph (48kph)) (note 1) - - - - - - - - - 4.5 - - - 15

1 to 7 Non-cent (urban, 30mph (48kph)) (note 1) - - - - - - - - - - - 80 0 25

1 to 7 Non-cent (suburban, 40/50mph (64/80kph)) (note 1) - - - - - - - - - 1 40 80 0 25

1 to 7 Non-cent (small town, 30/40mph (48/64kph)) (note 1) - - - - - - - - - - - 80 0 30

8 Urban - dual 2 2 - - - - - - - - - 1 40 - - 35

9 Urban - dual 3 3 - - - - - - - - - 1 40 - - 35

10 Urban - dual 4 4 - - - - - - - - - 1 40 - - 35

11 Urban - Expressway, 2 or more lanes 2 - - - - - - - - - 1 40 - - 35

12 Urban - Motorway and dual ramps, 1 lane 1 No 20 5 5 7.3 1 4 calc 0.6 - - - - 45

13 Urban - Motorway and dual ramps, 2 lanes 2 - 20 5 5 - - - - - - - - - 45

14 Urban - Motorway - D2 2 - 20 5 5 - - - - - - - - - 45

15 Urban - Motorway - D3 3 - 20 5 5 - - - - - - - - - 45

16 Urban - Motorway - D4 4 - 20 5 5 - - - - - - - - - 45

17 Urban - Motorway - D5 5 - 20 5 5 - - - - - - - - - 45

20 Rural - single 4 lane 2 No 40 15 15 11 0 1 calc 2 - - - - 45

21 Rural - poor single 4.0m 1 No 110 40 40 4 0 0 calc 2 - - - - 20

22 Rural - poor single 5.5m 1 No 100 35 35 5.5 0 0 calc 2 - - - - 45

23 Rural - poor single 6.0m 1 No 90 25 25 6 0 0 calc 2 - - - - 45

24 Rural - typical single 6.0m 1 No 90 25 25 6 0 0 calc 2 - - - - 45

25 Rural - poor single 7.3m 1 No 80 25 25 7.3 0 0.5 calc 2 - - - - 45

26 Rural - typical single 7.3m 1 No 50 15 15 7.3 0 2 calc 2 - - - - 45

27 Rural - good single 7.3m 1 No 50 15 15 7.3 0 2 calc 2 - - - - 45

28 Rural - typical single 10.0m 1 No 40 15 15 10 0 3 calc 2 - - - - 45

29 Rural - good single 10.0m 1 No 40 15 15 10 1 3 calc 2 - - - - 45

30 Rural - single with climbing lane 2 - 40 30 30 10 - - - - - - - - 45

31 Rural - dual 2 lanes 2 - 30 10 10 - - - - - - - - - 45

32 Rural - dual 3 lanes 3 - 30 10 10 - - - - - - - - - 45

33 Rural - dual 2 with climbing lane 3 - 30 30 30 - - - - - - - - - 45

34 Rural - dual 2 lanes with grade separation 2 - 30 10 10 - - - - - - - - - 45

35 Rural - dual 3 lanes with grade separation 3 - 30 10 10 - - - - - - - - - 45

36 Rural - dual 2 lanes with GS and climbing lane 3 - 30 30 30 - - - - - - - - - 45

37 Rural - Motorway and dual ramps, 1 lane 1 No 20 5 5 7.3 1 4 calc 0.6 - - - - 45

38 Rural - Motorway and dual ramps, 2 lanes 2 - 20 5 5 - - - - - - - - - 45

39 Rural - Motorway - D2 2 - 20 5 5 - - - - - - - - - 45

40 Rural - Motorway - D3 3 - 20 5 5 - - - - - - - - - 45

41 Rural - Motorway - D2 with climbing lane 3 - 20 15 15 - - - - - - - - - 45

50 Zone connector - - - - - - - - - - - - - - --

Volume 15 Section 1 Chapter 2Part 7 Speeds on Links Rural Single Carriageways (Speed/Flow Type 1)


2 RURAL SINGLE CARRIAGEWAYS (SPEED/FLOW TYPE 1)

2.1 The rural single carriageway speed relationships were derived from a study carried out in 1991 (Speed/Flow/Geometry Relationships for Rural Single Carriageways, TRRL Contractors Report No. 319).None of the sections studied was subject to a local speed limit, and none included a junction where theroad being studied lost priority. Observed journey speeds varied from 30 to 95 kph, and the proportionof heavy vehicles varied from 0 to 50% (both figures give the range of 10 minute averages observed).

2.2 Table 7/2/1 defines the geometric parameters and variables used in the relationships and gives theranges of typical values over which the relationships should apply. The relationships cannot necessarilybe taken to apply outside the given ranges of the variables. The geometric variables should all beaveraged over a reasonable length of link not less than two kilometres.

2.3 The rural single carriageway speed/flow relationship was derived using two way traffic flows. NESA,however, applies the speed/flow curves in a directional manner. The use of directional rather than twoway flows allows for a more realistic representation of peak hour tidality. This gain in realism morethan offsets the fact that the speed/flow curves are applied in a manner that conflicts with the way theywere derived.

2.4 The capacity flag of a single carriageway is set at:

It therefore varies by flow group as the proportion of heavy vehicles changes. When calculating thecapacity flag, CWID has a minimum value of 5.5 metres.

2.5 The point of change of slope (QB) is given by the relationship:

2.6 For flow levels less than the breakpoint QB the speed prediction formulae for light vehicles in kph is:

Equation 7/2/1

Equation 7/2/2

- 0.09 x BEND or -0.015 x BEND for roads designed to TD9/81 Equation 7/2/3

- 0.0007 x BEND x HR- 0.11 x NG- 1.9 x JUNC+ 2.0 x CWID[+ SWID x [(1.6/SWID) + 1.1]if SWID > 0][+0 if SWID = 0]+ 0.3 x VW+ 0.005 x VISI- (0.015 + (0.00027 x PHV)) x Q

QC2400 CWID 3.65– ·

CWID-------------------------------------------------- 92 PHV–

80----------------------------- vehs/hour/dir=

QB 0.8QC=

VL 72.1=

Chapter 2 Volume 15 Section 1Rural Single Carriageways (Speed/Flow Type 1) Part 7 Speeds on Links


2.7 For flow values greater than QB the speed prediction formula for light vehicles is:

2.8 For all flow levels the speed prediction formula for heavy vehicles in kph is:

Equation 7/2/4

where

Table 7/2/1: Definition of Variables Used in Speed Prediction Formulae for Rural Single Carriageways

SYMBOL VARIABLE DESCRIPTION TYPICAL VALUESMin Max

Bend Bendiness: total change of direction (deg/km) 0 150

HR Sum of rises per unit distance one-way links only (m/km) 0 45HF Sum of falls per unit distance one-way links only (m/km) 0 45NG Net Gradient, HR - HF (m/km) -45 45

JUNC Side roads intersection, both directions (no/km) 0 5

CWID Average carriageway width between white line edge markings (m)

6 11

SWID Average width of hard strip on both sides, including width of white line (m)

0 1

VW Average verge width, both sides (m) 0 7

VISI Average sight distance (m) 100 550

PHV Percentage of heavy vehicles (OGV1 + OGV2 + PSV) 2 30

VLVH Speed of light and heavy vehicles (kph) 45 speed limit

Q Flow, all vehicles, two-way or one-way (vehs/hour/lane) - -

QB Breakpoint: the value of Q at which the speed/flow slope of light vehicles changes (vehs/hour/lane)

0.8QC

QC Capacity flag: defined as the maximum realistic value of Q (vehs/hour/lane)

900 1600

- 0.1 x BEND or ZERO for roads designed to TD9/81 Equation 7/2/5

- 0.07 x HR- 0.13 x NG- 1.1 x JUNC+ 0.007 x VISI+ 0.3 x VW- 0.0052 x Q

VL VB 0.05 Q QB– –=

VB speed at Q QB= =

VH 78.2=

Volume 15 Section 1 Chapter 2Part 7 Speeds on Links Rural Single Carriageways (Speed/Flow Type 1)


subject to the constraint that if the calculated value of VH is greater than VL then VH is set equal toVL.

2.9 The maximum speed is defined by the speed limit (60mph = 96.0 km/h). There is a minimum speedcut-off of 45 kph.

One Way Rural Single Carriageway Roads and One Lane Motorway/Dual Ramps (RC 12 and RC 37)

2.10 One way rural single carriageway roads and one lane ramps (RC 12 and RC 37) should be coded with acarriageway width (CWID) as though they were two way. For example, a one way rural singlecarriageway (or 1 lane motorway ramp) with a carriageway width of 4.0m should be coded withCWID=8.0 (=4.0*2).

Climbing Lanes on Single Carriageway Roads (Road Category 30)

2.11 Climbing lanes on single carriageway roads should be coded as 2-way links with Road Category 30.Hilliness rise (HR) and hilliness fall (HF) must both be coded to enable NESA to determine thedirection of the climbing lane. If HR=HF a fatal error occurs. For climbing lanes on single carriagewayroads (RC=30) CWID should be coded as the full carriageway width, typically 10.0m. NESAcalculates pseudo CWIDs for each direction based on the input CWID - 1.33*CWID for the uphilldirection and 0.67*CWID for the downhill direction.

2.12 The effect of the climbing lane is to increase speeds uphill (due to increased carriageway width,reduced flow per lane, reduced access friction and possibly increased visibility) while speeds on thedownhill link remain unaltered. For coding purposes the length of the climbing lane should be taken asthe length of road over which the full width of the climbing lane is provided plus half the taper lengths.

Chapter 2 Volume 15 Section 1Rural Single Carriageways (Speed/Flow Type 1) Part 7 Speeds on Links


Volume 15 Section 1 Chapter 3Part 7 Speeds on Links Rural All-Purpose Dual Carriageways and Motorways (Speed/Flow Types 2-6)


3 RURAL ALL-PURPOSE DUAL CARRIAGEWAYS AND MOTORWAYS (SPEED/FLOW TYPES 2-6)

3.1 The speed relationships for rural all-purpose dual carriageways and motorways were derived from astudy carried out in 1990 (Speed/Flow/Geometry Relationships for Rural Dual Carriageways andMotorways, TRRL Contractors Report No. 279), where individual site lengths varied from 1.5 to 5kilometres. None of the sections studied was subject to a local speed limit, and none included ajunction where the road being studied lost priority. Observed journey speeds varied from 40 to 125 kphover flow ranges of 0-2000 vehs/hour/lane. (Both figures give the range of 10 minute averagesobserved).

3.2 Table 7/3/1 below defines the variables used in the relationships and gives the ranges of values overwhich the relationships should apply. The relationships cannot necessarily be taken to apply outside thegiven ranges of the variables. The geometric variables (BEND, HR and HF) should all be averagedover a reasonable length of road, generally not less than two kilometres.

3.3 The NESA dual carriageway speed/flow relationships are expressed in flow per lane and NOT flow perdirection as is the case with single carriageways. The number of lanes is determined from the roadcategory.

3.4 The majority of all-purpose dual carriageways and motorways are built with standard 3.65 metre widthlanes. Consequently the research to develop the speed/flow relationships was undertaken on links withlanes of, or close to, the standard width and was not able to detect a significant width parameter for usein the speed prediction formulae. Also, unlike single carriageway links, the average speed of lightvehicles on all-purpose dual carriageways is not influenced by the presence of a hard strip. If theaverage lane width of the proposed scheme is significantly less than the standard 3.65 metres then itmay be necessary to use a different road category. Transport Scotland may be contacted for advice.

3.5 QC, the maximum realistic value of Q, is taken as:

3.6 QB, the value of Q at which the speed/flow slope changes, is taken as 1200 and 1080 vehs/hour/lane formotorways and all-purpose dual carriageways respectively.

Equation 7/3/1

and

2330 / 1 0.015 PHV+ for motorways

2110 / 1 0.015 PHV+ for all-purpose dual carriageways

Chapter 3 Volume 15 Section 1Rural All-Purpose Dual Carriageways and Motorways (Speed/Flow Types 2-6) Part 7 Speeds on Links


3.7 For flow levels less than the breakpoint (QB) the speed prediction formula for light vehicles, in kph is:

3.8 The speed prediction formula for heavy vehicles which is applied at all flow levels, in kph is:

subject to the constraint that if the calculated value of VH is greater than VL then VH is set equal to VL.

3.9 As indicated in paragraph 3.5 the maximum realistic flows on dual carriageways and motorways varieswith the percentage of heavy vehicles (PHV). It therefore follows that the NESA representation of

Equation 7/3/2

where 108 for dual 2-lane all-purpose (NESA Speed/Flow Type 2)

115 for dual 3-lane all-purpose (NESA Speed/Flow Type 3)111 for dual 2-lane motorways (NESA Speed/Flow Type 4)118 for dual 3-lane motorways (NESA Speed/Flow Type 5)118 for dual 4-lane motorways (NESA Speed/Flow Type 6)the speed/flow slope for light vehicles, 6 kph per 1000 vehicles

Table 7/3/1: Definition of Variables Used in Speed Prediction Formulae for Rural All-Purpose Dual Carriageways and Motorways


BEND Bendiness: total change of direction (deg/km) 0 60

HR Sum of rises per unit distance one-way links only (m/km) 0 45


VLVH Speed of light and heavy vehicles (kph) 45 speed limit

SLSH Speed/flow slope of light and heavy vehicles (kph reduction per 1000 increase in Q)

0 55

Q Flow, all vehicles, two-way or one-way (vehs/hour/lane) 0 2300

QB Breakpoint: the value of Q at which the speed/flow slope of light vehicles changes (vehs/hour/lane)

1080 or 1200

VB Speed of vehicles at flow QB (kph) 50 75

QC Capacity flag: defined as the maximum realistic value of Q (vehs/hour/lane)

1400 2250

Equation 7/3/3

where 86 for all-purpose(NESA Speed/Flow Types 2 and 3)93 for motorways(NESA Speed/Flow Types 4, 5 and 6)

VL KL= 0.1 BEND–

0.28 HR–

SLQ–

KL is

SL is

VH KH= 0.1 BEND–

0.5 HR–

KH is

Volume 15 Section 1 Chapter 3Part 7 Speeds on Links Rural All-Purpose Dual Carriageways and Motorways (Speed/Flow Types 2-6)


speed/flow between the breakpoint QB and the minimum speed cut-off should also vary with PHV. Itwas found that the best representation of light vehicle speeds in this flow range was given by:

Heavy vehicle speed/flow relationships do not change at the breakpoint.

3.10 The maximum speed is defined by the local speed limit (70mph = 113kph). There is a minimum speedcut-off of 45 kph.

Climbing Lanes on Dual Carriageway Roads (RC 33, 36 and 41)

3.11 Climbing lanes on dual carriageway roads should be coded as 2-way links with the appropriate roadcategory (RC 33, 36 41) as detailed in Table 7/1/2. Hilliness rise (HR) and hilliness fall (HF) mustboth be coded to enable NESA to determine the direction of the climbing lanes. If HR=HF a fatal erroroccurs. For climbing lanes on dual carriageway roads the user is not required to input the carriagewaywidth (CWID) as NESA assumes three lanes uphill and two lanes downhill. The effect of the extra laneuphill will be to increase speeds due to the reduced flow per lane.

Equation 7/3/4VL VB 33 Q QB– /1000–=

Chapter 3 Volume 15 Section 1Rural All-Purpose Dual Carriageways and Motorways (Speed/Flow Types 2-6) Part 7 Speeds on Links


Volume 15 Section 1 Chapter 4Part 7 Speeds on Links Urban Roads (Speed/Flow Types 7 and 8)


4 URBAN ROADS (SPEED/FLOW TYPES 7 AND 8)

4.1 In built-up areas, generally subject to a local speed limit, the road network becomes more dense andintersections play a more significant role in determining speeds. The speed/flow relationships that havebeen developed for urban areas apply to the main road network in the cities and larger towns wherethere is a 30 mph (48 kph) speed limit. The relationships are designed to represent average-speeds intowns on the roads that function as traffic links. A distinction is made between central and non-centralareas, and they include an allowance for junctions. They are linear relationships of fixed negativeslope with a minimum speed cut-off.

4.2 The relationships were derived from data obtained in the Urban Congestion Survey 1976 (TRRL SR438) and in similar surveys in 1971, 1967 and 1963. The surveys were carried out in eight towns(population range 77,000 to 560,000) and in the principal cities of five conurbations (population range300,000 to 1,100,000). The surveys were not intended to provide relationships to estimate speeds onindividual links in urban areas; in practice urban speeds depend more on junction geometry and turningmovements than on link characteristics; and complex relationships would be required to providereasonable estimates of speeds on each link. However, the surveys do provide sufficient information toestimate average speeds in towns, distinguishing central and non-central areas.

4.3 Central areas are defined as those including the main shops, offices and central railway stations, with ahigh density of land use and frequent multi-storey developments, as in the widely used classificationcentral business district (CBD). Conurbations may have several CBDs (for example, central areas ofPaisley and Motherwell are CBDs within the Glasgow conurbation) whilst most free-standing townswill normally have only one. Streets which have commercial or industrial development but are not of ahigh-density CBD nature should not be included in the central area. Non-central areas comprise theremainder of the urban area. With this classification, the central areas constituted between 4 per centand 22 per cent (average 11 per cent) of the total street length in the networks of the 13 towns studied.

4.4 The urban speed/flow relationships were derived using two way traffic flows. NESA, however, appliesthe speed/flow curves in a directional manner. The use of directional rather than two way flows allowsfor a more realistic representation of peak hour tidality. This gain in realism more than offsets the factthat the speed/flow curves are applied in a manner that conflicts with the way they were derived.

4.5 The definition and range of the variables used in the speed prediction formulae are given in Table 7/4/1.

4.6 The average vehicle speed V kph at flow Q vehs/hour/3.65m lane is given by the relationship:

V0 is defined below for Central and Non-central areas. The maximum vehicle speed is limited to thelegal speed limit and there is a minimum speed cut-off (Paragraph 4.11).

4.7 The maximum realistic flow (QC) is 800 vehicles/hour/3.65m lane. For urban links this value is notaffected by the proportion of goods vehicles.

Equation 7/4/1

where V0 is the speed at zero flow

V V0 30Q / 1000–=

Chapter 4 Volume 15 Section 1Urban Roads (Speed/Flow Types 7 and 8) Part 7 Speeds on Links


4.8 Generally the use of area wide speed/flow relationships which include an allowance for junction delayswill be satisfactory. However in some cases schemes are designed to relieve a particularly congestedjunction and there may be a good case for modelling the junction explicitly; the comparison of localjourney time information with NESA modelled times will demonstrate whether this approach isnecessary. Where a scheme is designed to relieve a set of junctions that are clearly interdependent,consideration should be given to using dynamic traffic modelling techniques (for exampleCONTRAM, SATURN or TRAFFICQ) as a supplement to the NESA evaluation, and as an aid todetermine which junctions to model in the NESA evaluation. Whenever individual junctions aremodelled in the urban network, the case for so doing should be clearly set out and sent with the NESAanalysis for validation.

4.9 All links in a particular area of a particular town should have the same value for INT or DEVEL, andtherefore the same speed/flow relationship per standard lane will apply. This emphasises the area-widenature of the urban speed relationships. Changes in speeds in urban areas are predicted on alink-by-link basis, but the absolute level of speed is set by reference to the network-average speedsobserved in central and non-central areas.

4.10 With large traffic models it is also important that a reasonable balance is struck between the networkstructure and zoning system, so that flows on individual links are representative of flows expected onthe ground. In some cases this may not be possible, in which case the speed/flow effect describedabove for individual links may not be appropriate. Such cases should be referred to Transport Scotland.

4.11 NESA imposes the constraint that urban speeds should not fall below the slowest average speedsobserved in towns in practice. This puts a reasonable lower limit to the fall in speeds due to trafficgrowth. The minimum speeds in NESA are 15 kph for central areas and 25 kph for non-central areas.

Non - Central Areas - (Road Speed/Flow Type 7)

4.12 Non-central areas are defined as all those areas not included in the central area definition (seeParagraph 4.3). The speed V0 in kph at zero flow is given by the relationship:

Table 7/4/1: Definition of Variables Used in Speed Prediction Formulae for Urban Roads


INT Frequency of major intersections averaged over the main road network (no/km)

2 9

DEVEL Percentage of road network with frontage development (%) 50 90

V Average vehicle speed (kph) 15 48

V0 Speed at zero flow (kph) 28 48

Q Total flow, all vehicles, per standard land (vehs/hr/3.65m lane)

0 1200

QC Capacity flag: defined as the maximum realistic value of Q (vehs/hr/3.65m lane)

800

Equation 7/4/2V0 64.5 DEVEL /5 kph–=

Volume 15 Section 1 Chapter 4Part 7 Speeds on Links Urban Roads (Speed/Flow Types 7 and 8)


where DEVEL is defined as the percentage of the non-central road network of the town that hasfrontage development, counting business and residential development as 100% and open space as 0%.DEVEL is normally in the range 50-90% with average values about 80%.


Central Areas - (Road Speed/Flow Type 8)

4.14 A definition of central areas is given in paragraph 4.3. The speed V0 (in such areas) in kph at zero flowis given by the relationship:

INT is calculated by dividing the total number of lengths of road between major intersections in thecentral area by the total length of main road in the central area. Major intersections will generally beroundabouts or traffic signals, but they may also be uncontrolled junctions where a significant trafficmovement loses priority. This network based value is not directly comparable with a route based valueused for suburban links. INT should generally be in the range 2 to 9 per km, with an average of about4.5. A value less than 2 is not appropriate for Central areas; if this occurs part or all of the areaclassified as Central should be re-classified as Non-central.


Equation 7/4/3

where INT is a measure of the frequency of major intersections averaged over the main road network

V0 39.5 5INT / 4 kph–=

Chapter 4 Volume 15 Section 1Urban Roads (Speed/Flow Types 7 and 8) Part 7 Speeds on Links


Volume 15 Section 1 Chapter 5Part 7 Speeds on Links Suburban Roads (Speed/Flow Types 9 and 10)


5 SUBURBAN ROADS (SPEED/FLOW TYPES 9 AND 10)

5.1 The suburban speed relationships apply to the major suburban routes in towns and cities where thespeed limit is generally 40 mph (64 kph). The relationships were derived from a study carried out in1969/70 of the major suburban radial routes of London, Manchester, Liverpool and Birmingham(Speed Flow Relationships on Suburban Main Roads, 1972 - Freeman Fox and Associates). Sixty-ninesections of single and dual carriageway were selected for study, with individual section lengths varyingfrom 1.6 to 8.3 km. The predominant speed limit was 40 mph (64 kph), although some sections had nolocal speed limit and some had a 30 mph (48 kph) speed limit. Observed journey speeds varied from 15to 80 kph.

5.2 The suburban relationships provide estimates of the average journey speed of light and heavy vehiclesseparately, including delays at junctions. Table 7/5/1 below defines the variables used in therelationships and gives the ranges of values over which the relationships apply. The relationshipscannot necessarily be taken to apply outside the given ranges of the variables. The geometric variablesINT and AXS should be averaged over a reasonable length of link, generally not less than twokilometres. Congested junctions should be modelled separately and not included in thecalculation of the value of INT.

5.3 The suburban speed/flow relationships were derived using two way traffic flows. NESA, however,applies the speed/flow curves in a directional manner. The use of directional rather than two way flowsallows for a more realistic representation of peak hour tidality. This gain in realism more than offsetsthe fact that the speed/flow curves are applied in a manner that conflicts with the way they werederived.

Table 7/5/1: Definition of Variables Used in Speed Prediction Formulae for Suburban Roads


INT Frequency of major intersections (no/km) 0 2

AXS Number of minor intersections and private drives (no/km) 5 75


VLVH Speed of light and heavy vehicles (kph) 25 64

SLSH Speed/flow slope of light and heavy vehicles (kph) reduction per 1000 increase in Q

0 45

V0 Speed at zero flow (kph) 48 64

Q Total flow, all vehicles, per standard land (vehs/hr/3.65m lane)

0 1500

QB Breakpoint: the value of Q at which the speed/flow slope changes (vehs/hr/3.65m lane)

1050

QC Capacity: defined as the maximum realistic value of Q (vehs/hr/3.65m lane)

1350 1700

Chapter 5 Volume 15 Section 1Suburban Roads (Speed/Flow Types 9 and 10) Part 7 Speeds on Links


5.4 The basic form of the relationships is that speed reduces as flow increases. The initial speed isdependent upon the road standard, the number of major intersections and the number of minorintersections and private drives. The rate of decrease in speed is dependent upon the number of majorintersections until a critical flow is reached at which point the rate of speed decrease changes to a fixedvalue until a minimum speed cut-off is reached.

5.5 There are important differences between the definition of the variable INT for suburban compared tourban roads. In suburban roads INT is specific to each section of route and major intersections areeither roundabouts or traffic signals. Junctions between consecutive links should not be doublecounted, and classified junctions, whose delays are separately assessed, should be excluded from INT.The number of minor intersections and private drives, AXS, should be the total for both sides of theroad (even for dual carriageways).

5.6 The maximum realistic flow (QC), is the same for both single and dual carriageways and is calculatedby the relationship:

5.7 The point of change of slope (QB) for light vehicles is defined by the relationship:

5.8 The speed of light vehicles (VL) in kph is defined by:

5.9 The speed of heavy vehicles (VH) in kph is defined by:

5.10 It should be noted that the rate of decrease in speed (SL and SH) is the same for both single and dualcarriageways. Below breakpoint (QB) SL and SH are also the same. Above the breakpoint SL increasesto 45 kph/1000vehs, whilst SH remains the same. Because the speed/flow slope for heavy vehicles (SH)does not increase when flow levels exceed the breakpoint, the calculated speed of heavy vehicles can

Equation 7/5/1

Equation 7/5/2

Equation 7/5/3

where V0 = C - (5 x INT) - (3 x AXS/20)C = 70 for single carriageways (Speed/Flow Type 9) C = 80 for dual carriageways (Speed/Flow Type 10)VB = speed at breakpointSL = 12 + (50 x INT/3) kph/1000vehs

Equation 7/5/4

where V0 = C - (5 x INT) - (3 x AXS/20)C = 64 for single carriageways (Speed/Flow Type 9)C = 74 for dual carriageways (Speed/Flow Type 10)SH = 12 + (50 x INT/3) kph/1000vehs

QC 1500 92 PHV– /80 vehs/hour/3.65m lane=

QB 0.7 QC vehs/hour/3.65m lane=

VL V0 SL Q/ 1000 if flow (Q) breakpoint (QB –=

VL VB 45 Q QB– /1000 if flow (Q) breakpoint (QB –=

VH V0 SH Q /1000 –=

Volume 15 Section 1 Chapter 5Part 7 Speeds on Links Suburban Roads (Speed/Flow Types 9 and 10)


exceed the speed of light vehicles. When this occurs the speed of heavy vehicles (VH) is set to thespeed of light vehicles (VL).

5.11 The minimum speed cut-off is 25 kph for single carriageways and 35 kph for dual carriageways and themaximum speed is limited by the legal speed limit (40mph = 64kph, 50mph = 80kph).

Chapter 5 Volume 15 Section 1Suburban Roads (Speed/Flow Types 9 and 10) Part 7 Speeds on Links


Volume 15 Section 1 Chapter 6Part 7 Speeds on Links Small Town Roads (Speed/Flow Type 11)


6 SMALL TOWN ROADS (SPEED/FLOW TYPE 11)

6.1 The small town speed/flow relationship has been developed for areas where it is difficult to define acentral area at peak times and for villages or short stretches of development. In these situations neitherthe main urban relationships nor the suburban relationships apply. In a similar manner to the suburbanspeed/flow relationships the small town relationships do not apply to individual links, they modeltraffic speeds over the whole of a route that is subject to a speed limit of 30 or 40 mph (48 or 64 kph).Unlike the suburban relationships, however, they do not distinguish between light and heavy vehicles,and they specifically exclude junction delays. Junctions where the route loses priority must bemodelled separately. The small town relationship is route specific. The user should therefore ensurethat all links that comprise a route have the same geometric variables. Table 7/6/1 defines the variablesused in the relationships and ranges of typical values.

6.2 The small town speed/flow relationships were derived using two way traffic flows (Journey Speedsthrough Small Towns 1992 - Halcrow Fox and Associates). NESA, however, applies the speed/flowcurves in a directional manner. The use of directional rather than two way flows allows for a morerealistic representation of peak hour tidality. This gain in realism more than offsets the fact that thespeed/flow curves are applied in a manner that conflicts with the way they were derived.

6.3 The breakpoint flow QB is taken as 700 veh/hour/3.65 metre lane. The maximum realistic flow (QC) is1200 veh/hour/3.65 metre lane. As with the main urban formulae, there is no correction for trafficcomposition.

6.4 The average speed in kph of all vehicles for flows below the breakpoint (QB) is given by:

where DEVEL is the percentage of the length of route that has frontage development, countingbusiness and residential development as 100% and open space as 0%: the value will normally lie in therange 35% - 90%.

Table 7/6/1: Definition of Variables Used in Speed Prediction Formulae for Small Town Roads


DEVEL Percentage of road network with frontage development(%) 35 90

P30 Percentage of route subject to a 30 mph (48 kph) speed limit (%) 0 100


VB Average vehicle speed at QB 38 57

Q Total flow, all vehicles, per standard land (vehs/hr/3.65m lane) 0 1200

QB Breakpoint: the value of Q at which the speed/flow slope changes (vehs/hr/3.65m lane)

700

QC Capacity flag: defined as the maximum realistic value of Q (vehs/hr/3.65m lane)

1200

Equation 7/6/1V 70 DEVEL/8 – P30/8 – 12Q/1000 –=

Chapter 6 Volume 15 Section 1Small Town Roads (Speed/Flow Type 11) Part 7 Speeds on Links


6.5 For flows greater than QB the average speed in kph of vehicles is given by:

6.6 The maximum speed is defined by the variable P30. This is because P30 defines the proportion of theroute subject to a 30mph (48kph) speed limit, whilst (100-P30) defines the proportion of the routesubject to a 40mph (64kph) speed limit. The maximum speed is therefore calculated using thefollowing equation:

6.7 There is a minimum speed cut off of 30kph.

6.8 These relationships should not be used for routes with P30 < 10% (that is for routes with an almostcontinuous 40 mph (64kph) limit), DEVEL < 65 (that is with less than 65% development), and accessfriction less than 3; access friction being defined as the total number, both sides, of lay-bys, side roadsand accesses per km (excluding house and field entrances) divided by the carriageway width in metres.In such cases, the route should be split into links, as appropriate, and the standard rural relationshipsshould be used instead.

Equation 7/6/2

Equation 7/6/3

where 1.609 is a factor to convert from mph to kph

V VB 45 Q QB– /1000–=

MaxSpeed kph 1.609 12000P30 300+

-----------------------------=

Volume 15 Section 1 Chapter 7Part 7 Speeds on Links Single Track Roads (Speed/Flow Type 12)


7 SINGLE TRACK ROADS (SPEED/FLOW TYPE 12)

7.1 The single track speed/flow relationship was derived from a study carried out in 1964 on single trackroads situated in the Scottish Highlands (Single Track Roads in the Scottish Highlands, RRL ReportLR 71). Four lengths of three single track roads were selected for the study, individual section lengthsvarying from 4.51 km to 7.89 km. Observed journey speeds varied from 24kph to 48kph over flowranges of 0-280 vehs/hour, two way flow.

7.2 Table 7/7/1 defines the variables used in the relationships and gives the ranges of typical values overwhich the relationships should apply. The relationships cannot necessarily be taken to apply outside thegiven ranges of the variables.

7.3 The single track road speed/flow relationship was derived using two way traffic flows. NESA appliesthe speed/flow curve in this manner as balanced flows are expected on single track roads. Therelationship does not distinguish between light and heavy vehicles and specifically excludes junctiondelays.

7.4 The average vehicle speed V kph at flow Q vehs/hour, two way flow is given by the relationship:

The maximum vehicle speed is limited to the legal speed limit and there is a minimum cut-off speed of20kph.

7.5 The maximum realistic flow (QC) is 280 vehicles/hour, two way flow for a well aligned road. As can beseen from Equation 7/7/1 and Table 7/7/1 the geometric properties of the single track link in questiondo not influence average vehicle speeds.

Equation 7/7/1

Table 7/7/1: Definition of Variables Used in Speed Prediction Formulae for Single Track Roads



Q Total flow, all vehicles, per standard land (vehs/hr/4m lane) 0 280

QC Capacity flag: defined as the maximum realistic value of Q (vehs/hr/4m lane)

280

V 36.5 17*Q200

-------------– * 1.609=

Chapter 7 Volume 15 Section 1Single Track Roads (Speed/Flow Type 12) Part 7 Speeds on Links


Volume 15 Section 1 Chapter 8Part 7 Speeds on Links Treatment of Overcapacity on Links


8 TREATMENT OF OVERCAPACITY ON LINKS 8.1 When the traffic flow on a link in a particular flow group exceeds the capacity for a lane, this can be

reported in the NESA output. Table 5/3/1 details the link capacities by road category. The overcapacityreport is a signal to the user that the NESA evaluation is dealing with flow levels above those normallyexperienced on links of similar standard. Consequently the modelled situation may not be realistic andthe benefits calculated by the program for the period of overcapacity will become less meaningful thehigher the degree of overcapacity.

8.2 NESA’s evaluation module has no control over the traffic flows. It is therefore possible for links in theevaluation to have flows allocated, either in the base year but more likely at some point in the future,that are not operationally feasible. In this situation NESA is attempting to assess a situation whichbreaks the bounds of common sense - that is a situation where future traffic levels exceed physicallypracticable flows per lane. If a link’s capacity is exceeded NESA uses a minimum speed cut-off toprevent the calculated speed becoming unrealistically low.

8.3 The concept of minimum speeds on links is similar to that of maximum delays at junctions. Thejustification for using cut-offs of this nature is that in practice, one expects traffic behaviour to changeto avoid bottlenecks.

8.4 Generally traffic flows input to NESA should not greatly exceed the calculated capacities, however,when overcapacity is reported (see Part 10 Chapter 16) the user should consider three possible coursesof action:

(i) if overcapacity is limited to a few links or junctions only and to the peak flow grouponly, the minimum speed cut-offs in NESA may be used as a proxy for thedelay-avoiding behaviour that would be expected to occur. This amounts to assumingthat the economic cost of alternative action is equal to the cost of travelling on theoverloaded section of a route at the NESA minimum speed (or, on junctions, with theNESA maximum delay). Provided that benefits from the peak flow group are not acrucial factor in the economic case for the scheme, this assumption may beacceptable, but see (ii) below;

(ii) otherwise the user should consider whether simple changes to the assignment orDo-Minimum improvements to the network may be devised to bring link or junctionflows into better balance with link or junction capacities. If the Do-Minimuminvolves a significant proportion of the Do-Something expenditure (that is more than20%), the Do-Minimum expenditure should itself be economically justified against aDo-Nothing (see Part 3 Chapter 2);

(iii) exceptionally, in some large conurbations, predicted levels of congestion may besufficiently severe to warrant modification to the NESA assumption that observedminimum speeds will be maintained in the future. In such cases the NESA cut-offsmay not represent the full economic cost of trip suppression or alteration of time,route, destination or mode of travel. However, simply lowering the cut-off speeds (orincreasing maximum delays) will ultimately overestimate the economic effect ofcongestion, because traffic will tend to take action to avoid or mitigate the effect ofcongestion. In these cases the advice of Transport Scotland should be sought to ensurethat a consistent approach to the problem is adopted.

Chapter 8 Volume 15 Section 1Treatment of Overcapacity on Links Part 7 Speeds on Links


Volume 15 Section 1 Chapter 9Part 7 Speeds on Links Representative Diagrams of Speed/Flow Relationships


9 REPRESENTATIVE DIAGRAMS OF SPEED/FLOW RELATIONSHIPS

9.1 This section gives examples of the speed/flow relationships built into NESA for different types of road.In all cases the examples plotted should be seen as giving the form of a family of curves, differentiatedaccording to the user defined geometric characteristics of each road.

9.2 The parameters used to define the example relationships for rural, urban, small town and suburbanroads are given in Tables 7/9/1 to 7/9/4.

Table 7/9/1: Values of Variables Used in Representative Speed/Flow Relationships - Rural Roads

RURAL ROADSSINGLE CARRIAGEWAYS: SPEED/FLOW TYPE 1

Typical Designed to TD9/81S2 S2 WS2

CWID (m) 7.3 7.3 10HR 7.5 7.5 7.5HF 7.5 7.5 7.5BEND (deg/km) 75 75 75SWID (m) 0 1 1VW (m) 1 4 4JUNC (no/km) 2 0.6 0.6VISI (m) 300 400 400

Maximum Speed 96 kphALL-PURPOSE DUAL CARRIAGEWAYS: SPEED/FLOW TYPES 2 & 3

HR 7.5BEND 30 deg/kmMaximum Speed 113 kph

MOTORWAYS: SPEED/FLOW TYPES 4, 5 & 6HR 7.5BEND 20 deg/kmMaximum Speed 113 kph

ALL SPEED/FLOW TYPESPHV 15%Minimum Speed 45 kph

(see Figures 7/9/1 and 7/9/2)

Chapter 9 Volume 15 Section 1Representative Diagrams of Speed/Flow Relationships Part 7 Speeds on Links


Table 7/9/2: Values of Variables Used in Representative Speed/Flow Relationships - Urban Roads

URBAN ROADSNON-CENTRAL AREAS: SPEED/FLOW TYPE 7

DEVEL = 50% good road conditions= 80% typical road conditions= 90% poor road conditions

Minimum Speed = 25kphCENTRAL AREAS: SPEED/FLOW TYPE 8

INT = 2 good road conditions= 4.5 typical road conditions= 9 poor road conditions

Minimum Speed = 15kphBOTH SPEED/FLOW TYPES

Maximum Speed = 48kph(See Figure 7/9/3)

Table 7/9/3: Values of Variables Used in Representative Speed/Flow Relationships - Suburban Roads

SUBURBAN ROADSSINGLE CARRIAGEWAYS: SPEED/FLOW TYPE 9DUAL CARRIAGEWAYS: SPEED/FLOW TYPE 10

Conditions Good Typical PoorINT (no/km) 0.4 0.8 1.2AXS (no/km) 15 30 40

PHV 12%

Maximum Speed = 64kph

Minimum Speed = 25kph single carriageways= 35kph dual carriageways

(See Figure 7/9/4)



9.3 Figure 7/9/1 shows the average speeds at different levels of traffic flow per lane for rural roads. Therelationships cannot be applied to predict different speeds for different lanes of the same carriageway.Because average vehicle speeds are calculated from light and heavy vehicle speeds weighted in thecorrect proportions there is a double knee in the relationships. Figure 7/9/2 shows average speeds atdifferent levels of total directional traffic flow, the advantage of this presentation is that, given apredicted traffic flow, the difference in predicted traffic speeds on different standards of road can beread directly from the diagram (without calculating the different levels of flow per lane). The speeddifferences will of course depend upon the precise geometric characteristics of the roads and theproportion of heavy vehicles.

9.4 The relationships for urban, small town and suburban roads are plotted on Figures 7/9/3 to 7/9/5. Itmust be remembered that the suburban and small town relationships were developed to give averagespeeds on routes and the urban relationships for areas of a town, consequently individual link speedsare unlikely to be modelled accurately.

9.5 Figure 7/9/6 represents single track speed/flow curves, which are independent of link geometry.

Table 7/9/4: Values of Variables Used in Representative Speed/Flow Relationships - Small Town Roads

SMALL TOWN ROADSSPEED/FLOW TYPE 11

Degree of development Light Typical HeavyDEVEL 35% 60% 90%

P30 20% 50% 100%

Maximum Speed = 64kph

Minimum Speed = 30kph(See Figure 7/9/5)



Figure 7/9/1: Typical Rural Speed/Flow Relationships - Vehicles/Hour/Lane



Figure 7/9/2: Typical Rural Speed/Flow Relationships - Vehicles/Hour/Direction



Figure 7/9/3: Typical Urban Speed/Flow Relationships



Figure 7/9/4: Typical Suburban Speed/Flow Relationships



Figure 7/9/5: Typical Small Town Speed/Flow Relationships



Figure 7/9/6: Single Track Roads Speed/Flow Relationships



VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


JUNCTIONS IN NESA

Contents

Chapter

1. When to Model Junctions

2. Junction Choice

3. Junction Types Modelled

4. The Concept of Maximum Delay

5. Geometric Delay

6. Queuing Delay

7. Formulae for Junction Capacity

8. Geometric Parameters

PART 8

Volume 15 Section 1Part 8 Junctions in NESA


Volume 15 Section 1 Chapter 1Part 8 Junctions in NESA When to Model Junctions


1 WHEN TO MODEL JUNCTIONS1.1 One of the benefits associated with the construction of a new road scheme can be the relief of the delay

which vehicles experience at congested road junctions in the existing road network. The amount ofdelay which vehicles suffer when queuing at a junction is dependent on the capacity of that junctionand the traffic flow level. On the existing network congestion delay occurs when junctions areoperating close to or above their capacity for long periods. There will, of course, also be delayassociated with junctions on the new road scheme.

RURAL AND SMALL TOWN ROADS

1.2 For rural roads and roads in small towns the speed/flow relationships in NESA do not account forjunction delays. It is generally only necessary to model the junctions where there is a significant flowchange but there are no fixed rules relating to the flow change that necessitates a junction beingmodelled. The user will need to make a judgement regarding the significance of each junction to theassessment of the scheme being studied.

URBAN AND SUBURBAN ROADS

1.3 For urban and suburban roads the speed prediction formulae within NESA do take account of delays atjunctions. In general, therefore, most junctions in urban areas do not need to be modelled explicitly.The exception is the junction which is heavily congested (either now or at some time during the life ofthe scheme) to such an extent that the journey times through the network resulting from the applicationof the standard urban speed/flow relationships are incompatible with actual timed runs. Such a junctionshould therefore be modelled explicitly in conjunction with speed/flow information from timed runs.Whenever individual junctions are modelled separately in the urban network, the case for doing somust be clearly set out in the NESA analysis.

The Interaction of Junctions

1.4 NESA models all junctions in isolation, with all arrivals assumed to be random. No allowance is madefor any junction interaction. Therefore, great care must be taken when modelling junctions in the samearea of the network where the capacity of one junction controls the flow at another, either now or atsome time in the future. If this occurs it may only be necessary to model the controlling junction.Double counting of delay must be avoided.

1.5 Turning flows for all junctions in the NESA network are taken directly from the assignment (see Part5). These flows can represent tidal conditions if flow group matrices are used (see Part 5 Chapter 4).

Chapter 1 Volume 15 Section 1When to Model Junctions Part 8 Junctions in NESA


Volume 15 Section 1 Chapter 2Part 8 Junctions in NESA Junction Choice


2 JUNCTION CHOICE2.1 The formulae in NESA for determining junction capacity and delays are based on work carried out by

TRL. The NESA program makes an estimate of the delays to traffic flow due to the presence ofjunctions in the network and calculates associated user costs. The level of detail is sufficient to enablethe program to compare the user costs of different junction types (for example, grade-separated versusat-grade solutions at a particular location), but it must not be used for detailed junction design.Junctions can be modelled in more detail using the programs PICADY, ARCADY and OSCADYwhich are available from the Transport Research Laboratory. NESA uses formulae compatible withthose programs to calculate capacities and delays.

2.2 The programs ARCADY, PICADY, and OSCADY model capacities, queues and delays atroundabouts, major/minor and signalled junctions respectively. They incorporate the high definitionqueuing theory (see Part 8 Chapter 6) and should be used for the detailed design of junctions.

2.3 It is suggested that the following approach should be used for junction choice, at an appropriate stagein scheme appraisal:

• design the junction alternatives using programs such as PICADY, ARCADY andOSCADY, to see which options are feasible from an operational viewpoint;

• test alternative junction types using NESA. This may be done using a reducednetwork taking one Do-Something option as a base and the other options asalternatives ensuring that the fixed trip matrix assumption is not violated;

• identify the optimum junction type in economic terms: it is that which creates themaximum NPV when construction costs, time and accident costs are taken intoaccount. Given the sensitivity of junction formulae and uncertainty of turningmovements, small differences in NPV should not be given too much weight;

• junction choice should recognise uncertainty (see Part 4 Chapter 2). Designs shouldbe sensitivity tested using traffic growth forecasts agreed with Transport Scotland.Where the choice of junction layout is particularly sensitive to the forecast turningmovements, this can be tested by assigning a range of matrices to the set of trees thatpass through the junction;

• the choice of junction type should be considered in the wider context provided by theframework approach to environmental evaluation (see Part 9 Chapter 4). Economic,environmental and operational trade-offs should be identified where junction typesdiffer significantly in non economic terms. Judgement should be exercised where, forexample, long delays would be imposed on traffic on minor arms by the optiondetermined with economic criteria alone.

Chapter 2 Volume 15 Section 1Junction Choice Part 8 Junctions in NESA


Volume 15 Section 1 Chapter 3Part 8 Junctions in NESA Junction Types Modelled


3 JUNCTION TYPES MODELLED3.1 This section gives a brief description of each of the junction types modelled in NESA. The detailed

formulae for calculating queuing delay and junction capacity are given in Part 8 Chapters 6 and 7respectively and are applied on an arm specific basis.

Major/Minor Priority Junctions

3.2 Major/Minor priority junctions are by far the most common form of junction control. They are mostappropriately used where it is desirable to give priority to one route, usually that carrying the greatertraffic volumes. The NESA formulae are based on the calibration of observed relationships between thegeometric characteristics of junctions, saturation and turning flow levels and capacity. They applyprimarily to three armed junctions but the results have been extended within the program to apply tofour armed junctions.

3.3 The designs of major/minor junctions vary widely from one junction to another and it has beennecessary to distinguish those features which have appreciable effect on capacity and those that havelittle or no effect. The parameters used in the NESA program are:

• the width of the major road (and kerbed central reserve if applicable)

• the lane width for non-priority streams

• the visibility

3.4 The delay to through traffic on the major route of a major/minor junction caused by right turning trafficfrom the major road blocking traffic behind it is not considered significant and is therefore notmodelled by NESA.

Roundabouts

3.5 The capacity is predicted entry by entry (up to six) and is dependent on the geometry of the roundaboutand demand flows on each arm. The geometric parameters fall into a distinct hierarchy with the entrywidth and length of flare having by far the most important effect on capacity and with the inscribedcircle diameter having a lesser but still important effect. The angle and radius of entry are also relevant.The entry capacity is defined as the inflow from an entry when the demand flow is sufficiently high asto cause steady queuing on the approach. The inflow is dependent on the circulation flow across theentry, to which it has to give way. The entry capacity therefore decreases as the circulating flowincreases and vice versa, consequently the capacity changes by flow group. This is also found to be truewith each junction type.

Signalised Roundabouts

3.6 The facility to model signalised roundabouts is not available in the current release of NESA. If thisfacility is required Transport Scotland advice should be sought.

3.7 Signalised roundabouts are becoming increasingly common, especially at grade separated roundaboutsat motorway exits. A signalised roundabout is a system of one-way traffic circulating round a centralisland with at least one of the entries controlled by traffic signals. However NESA can only modeljunctions where all entries are controlled by signals. Cases where only one arm is signalised areprobably best treated as conventional roundabouts. Intermediate cases with two or more entry armscould be considered as fully signalised as they would be likely to become so at some time during theassessment period. At some four arm at-grade roundabouts through traffic on the main road passesstraight through the centre of the junction and minor road and turning traffic circulates. Thesehamburger roundabouts have a different coordination strategy from circulating roundabouts, as they

Chapter 3 Volume 15 Section 1Junction Types Modelled Part 8 Junctions in NESA


are designed for minimum delay at the main road traffic passing straight through the junction, andcannot be modelled using NESA.

3.8 The method of assessing delays at signalised roundabouts in NESA was developed from theTRANSYT program, see TRRL Research Report RR 274 - The Use of TRANSYT at SignalisedRoundabouts. It is assumed that the system operates on a common cycle time irrespective of how it iscontrolled on-street. By an iterative procedure NESA determines a suitable cycle time and green timesfor each approach for each flow group for each year of the appraisal. Then using queuing delayformulae predicts mean delay per vehicle, including geometric delay, for each entry.

3.9 The circulating link system necessary to model a signalised roundabouts is set up automatically byNESA. The circulating carriageway typically has between two and four lanes. Each link carries trafficoriginating from a particular entry arm and hence the shared stop-line models the arrival of differentplatoons of traffic at the stop-line at different times in the cycle. All nodes are assumed to be two-stageand there is no double-cycling or double-greening. Inter-green periods are fixed at five secondsalthough at some sites this may need to be increased for operational reasons. Excess queue weightingon the circulating links applies only if the distance between stop-lines is less than100m. The NESAmodel uses a relatively low weighting but takes account of excess queuing once the link is predicted tobe half full on average, rather that the TRANSYT recommended two-third full.

3.10 Modelling signalised roundabouts in NESA can significantly increase the program run time and theuser may prefer to have built separate NESA models for individual junctions until the data is known tobe correct.

Traffic Signals

3.11 Calculations of capacity and delay are performed to reflect OSCADY theory and to use algorithms aspublished in TRRL RR105.

3.12 Designs of signalised junctions vary considerably and NESA can model junctions with three or fourarms and a maximum of six stages. Transport Scotland advice should be sought if the user wishes tomodel either more arms or more stages. The user inputs the permissible movements from each lane andstaging details. The only geometric details affecting the capacity of signals are the width of the stopline and approach gradient. The program assumes that traffic will act in a reasonable way and attemptto equalise congestion between lanes. It also performs an iterative process to reach the optimal signalsettings for each flow group in each year of the evaluation.

Merges at Grade Separated Junctions

3.13 When the total traffic demand on the main line and slip road at a grade separated junction exceed thecapacity of the downstream link flow breakdown can occur with resultant delays to both the mainlineand merging traffic queues. The delays to the total traffic stream can be modelled in NESA.

Volume 15 Section 1 Chapter 4Part 8 Junctions in NESA The Concept of Maximum Delay


4 THE CONCEPT OF MAXIMUM DELAY4.1 The concept of maximum delay has an important bearing on junction evaluation in NESA. The form of

the NESA junction delay formulae imply that, as the capacity of the junction is reached, delaysincrease very rapidly. NESA includes, as default, a maximum delay at junctions of 300 seconds for allflow groups. If journey time evidence warrants a local adjustment for all the junctions in the networkthe user may change the value upwards (maximum 900 seconds) or downwards. The maximum delay(MAXDEL) is attributed to all vehicles but on an arm by arm or stream by stream basis.

4.2 Without a maximum delay cut off, the formulae would predict much higher delays in the future atheavily congested junctions than have been generally observed at existing congested junctions. Suchhigh delays are considered unrealistic bearing in mind that they apply to all traffic in the flow group.The 300 second maximum delay presumes that some vehicles will queue for considerably longer thanthis, some for a shorter period, with the average for all vehicles being 300 seconds. In practice somedrivers will act to prevent such long delays occurring by travelling by a different route or travelling at adifferent time. The maximum delay substitutes for the detailed modelling and evaluation of suchbehaviour for which NESA is not designed.

4.3 Where junctions are explicitly modelled and NESA junction delay benefits are an important element,the realism of the magnitude of the junction delay benefits should be examined by considering thefollowing:

• Do-Minimum improvements (see Part 3 Chapter 2) or, where Do-Something delaysare large, Do-Something junction optimisation. Small changes in the coding ofjunction lay outs can sometimes yield significant changes in junction delay costs.

• Comparison of NESA and measured journey times, for example, where junctionsinteract. In practice, it is more common for local maximum average delays to be lessthan 300 seconds rather than more.

• Explicit modelling of critical junctions outside NESA. In some instances the NESAuser may wish to consider whether in-depth analysis of a critical junction or a set ofinterdependent junctions is necessary (see Part 8 Chapter 1).

4.4 A particular problem with calibrating maximum delays using timed runs is that it is impossible tocalibrate the future, for example, junctions may be coded at 300 seconds maximum delay on the basisof present journey time evidence, but may give rise to longer delays in the future with traffic growth.Again this may be tested using SATURN or more sophisticated junction modelling techniques such asmicrosimulation, which explicitly model local diversionary reassignment; and inputting the resultinglink flows and delays directly into NESA (see Part 2 Chapter 3). Transport Scotland advice should besought in such circumstances.

Chapter 4 Volume 15 Section 1The Concept of Maximum Delay Part 8 Junctions in NESA


Volume 15 Section 1 Chapter 5Part 8 Junctions in NESA Geometric Delay


5 GEOMETRIC DELAY5.1 Geometric delay is considered as the delay suffered in the absence of queues, as a result of the need for

vehicles to slow down, negotiate the junction and accelerate back to normal speed. Geometric delaycan be modelled in the assignment process through the use of junction indices (see Part 5 Chapter 3).Junction indices are at a level of detail sufficient enough to produce realistic routeings and trafficflows, but are too coarse to give accurate estimates of travel time changes for use in the evaluationprocess. Consequently NESA uses the following methodology to calculate geometric delay at priorityjunctions, roundabouts and merges.

Geometric Delay at Major/Minor Priority Junctions

5.2 Vehicles travelling straight through the junction on the major arms do not suffer any significantgeometric delay, but all other vehicle movements are subject to delay dependent upon the turningmovements. Geometric delays are shown in Table 8/5/1 although it must be understood that delays canchange subject to speed-related amendments.

Notes:

(i) Geometric delays vary with entry and exit link speeds. They are significantly higherat intersections where at least one link speed is greater than 64 kph (generally ruralroads without speed limits) for which NESA adds 2 secs delay per vehicle, than thosewith link speed less than 64 kph (generally suburban roads with speed limits).

(ii) Major ahead delay is still zero for poor visibility and high speed junctions.

(iii) Delays at major/minor junctions which do not meet the visibility requirements ofDepartmental Advice Note TD42 (DMRB 6.2) are significantly higher than those atjunctions which do. For this substandard visibility NESA adds 1.4 secs per vehicle.

Table 8/5/1: Geometric Delays at Major/Minor Priority Junctions

Vehicle Category Delay per Movement (seconds/vehicle)Major

leftMajorahead

Major right

Minor left

Minorahead

Minor right

Cars, LGV 5.7 0.0 6.5 7.8 12.2 10.6OGV, PSV 7.8 0.0 8.6 9.9 14.3 12.7

Chapter 5 Volume 15 Section 1Geometric Delay Part 8 Junctions in NESA


Geometric Delay at Roundabouts - Conventional and Signalised

5.3 NESA calculates geometric delay for light vehicles at all types of roundabout, including gradeseparated, using the formula:

5.4 Geometric delay per vehicle for heavy vehicles is calculated at being 15% greater than for lightvehicles.

Geometric Delay at Traffic Signals

5.5 The program assumes no geometric delay at traffic signals.

Merges at Grade Separated Junctions

5.6 Even under light traffic flow conditions there will be a slight geometric delay at all merges at gradeseparated junctions but these are not modelled in NESA. Under heavy flow conditions the formula usedto calculate the total delay includes the geometric delay.

Delay per vehicle Equation 8/5/1

where js

D = inscribed circle diameter in metresdbc = distance travelled on roundabout (= (D-7) metres) = proportion of roundabout travelled. It is assumed that arms are

equally spaced around the roundabout. (For a right turn at a 4 arm roundabout = 0.75.)

V = the average of the approach speed coded on the entry link and the departure speed on the exit link, in kph, and will not vary by flow group. The entry and exit speeds are those coded for light vehicles.

= dbcjs

-------- 0.23V 5.62– 0.12D– 0.000367V.D secs+ +

= 0.96 D 2.03 metres/sec+

Volume 15 Section 1 Chapter 6Part 8 Junctions in NESA Queuing Delay


6 QUEUING DELAY 6.1 Queuing delay can be considered as that delay resulting from the presence of other vehicles. Traffic at

road junctions can usually be divided into a number of separately identifiable streams each with acapacity and a demand flow. Whether or not queuing occurs in a given stream depends on the ratio ofthese two items, known as the Traffic Intensity. The capacity has been described in Part 8 Chapter 3for the different junction types. The demand flow is inherently variable, beside random fluctuations inthe arrivals at a junction the average level of flow varies during the day. At peak times the trafficdemand in a stream may approach or even exceed the capacity available to it, whereas in off peakperiods there is usually a substantial margin of spare capacity. This varying demand is represented inthe NESA flow group structure (see Part 5, Chapter 2). For NESA purposes it is necessary to employflow/delay relationships which reflect these varying flow group conditions.

6.2 Traditionally flow/delay relationships are derived from a steady state theory which provides goodpredictions when there is spare capacity. Steady state queuing theory, however, predicts infinite queuesand delays when the demand flow reaches the capacity available to it and therefore cannot be applied todetermine delay at high traffic intensities which can occur in NESA peak flow groups. Instead, timedependent queuing theory is used when the flow/capacity ratio exceeds 0.80 with a maximum delay cutoff (see Part 8 Chapter 4). The exception is for grade separated merges where a simple linear flow/delay relationship was found to represent observed delays.

6.3 In its most refined state (high definition) this theory makes possible the modelling of the growth anddecay of queues of vehicles in very small time intervals. For NESA purposes, it is sufficient to employflow/delay relationships which give only average delay associated with average flows over the peakflow group hours (a low definition approach) since the uncertainty associated with traffic forecastsmake it inappropriate to predict traffic flows for the life of a scheme at the detailed level required for ahigh definition approach. Assumptions have been made about the shape of the traffic demand profile,the approximate length of peaks and as a result the average intensity and capacity. This is illustrated inFigure 8/6/1. In junction delay formulae the length of each peak is called the block time (T). For NESApurposes the length of each peak has been taken as 1 hour. The flow in the block time adjacent to thepeak is assumed to be 80% of the peak flow.

6.4 The flow/delay relationships used in the program are based on a coordinate transformation techniquewhich smooths the steady state relationship for vehicle delays into the over capacity deterministicresults obtained by integrating demand minus capacity. The relationships depend on the flows andcapacities during the peak and adjacent off peak periods, and on the peak duration. Figure 8/6/2illustrates schematically the form of the curve. The value of the maximum delay cut off includesgeometric delay where appropriate.

6.5 The peak period flow conditions generally correspond to the traditional morning and eveningcommuting peaks (the NESA default peak hour is flow group 4 with a duration of 500 hours).However, in recreational areas the peaks are more likely to occur in the summer months at weekends.

Chapter 6 Volume 15 Section 1Queuing Delay Part 8 Junctions in NESA


Figure 8/6/1: High and Low Definition Peak Models

Figure 8/6/2: Queuing Delay Curves

Volume 15 Section 1 Chapter 6Part 8 Junctions in NESA Queuing Delay


Queuing Delay Formulae (except Grade Separated Merges)

6.6 The formulae used in NESA to calculate delay are given below.

6.7 For steady state queuing (v/c < 0.80):

6.8 For time dependent queuing (v/c > 0.80):

Delay = Equation 8/6/1

where is the traffic intensity for the non-peak flow group being analysed

is the randomness factor, a constant depending on arrival and service patternsis the capacity for the non-peak flow group being analysed

is the low flow queuing delay

For roundabouts and major/minor junctions

For signals

where c is the cycle timeis green time

cycle time


where F =

G =

h =

e =

L is the low flow queuing delay (defined in Equation 8/6/1)q is the demand flow in flow group being analysedq0 is 80% of the demand flow in flow group below that being analysed is the capacity in flow group being analysed is the capacity for 80% of the demand flowT is the Block Time representing length of peakC is the Randomness factor, a constant depending on arrival and service

patterns

C00 1 0– -------------------------- L+

0

C

0

L

L 10------=

L c 1 – 2

2 1 0– --------------------------=

12--- F2 G+

12---

F– e L+ +

10 q0–----------------- 1

2---T q– 1 h

q---–

2C 1 h 1q--- 1

---+

–

+ e+

2T0 q0–----------------- 2C q

--- q– e– 1 h

q---–

0– q0+

2Cq00 0 q0– ----------------------------

Chapter 6 Volume 15 Section 1Queuing Delay Part 8 Junctions in NESA


6.9 At give way junctions (major/minor and roundabouts) it is reasonable to assume that both vehiclearrivals to the rear of the queue and departures across the give way line are approximately random andtherefore C can be taken as 1.0. Also, as before (see Paragraph 6.7) L, the low flow queuing delay = 1/ ( = capacity of flow group being analysed).

6.10 Signalised junctions differ in that the departures across the stop line are clearly non random. Because ofcyclic operation traffic from each approach is released in dense platoons with substantial gaps betweenthe platoons. However, within each platoon vehicles are spaced almost regularly. For this case C istherefore not equal to unity and the most appropriate value for C has been taken as 0.60. L, the lowflow queuing delay is the same as that for the steady state (see Paragraph 6.7).

Queuing Delay Formula for Grade Separated Merges

6.11 The research into merge delays showed that delays varied considerably with time even within a singlepeak period. Day to day variations were also marked under apparently similar flow conditions. It wasfound that the average delays, that is to both main line and merging streams, over and above speed/flowdelays could be represented by a straight line flow/delay relationship of the form:


where the capacity ratio is defined as the total upstream demand divided by the capacity of the downstream link

227 Capacity Ratio - 0.75 seconds per vehicle for v/c > 0.75

Volume 15 Section 1 Chapter 7Part 8 Junctions in NESA Formulae for Junction Capacity


7 FORMULAE FOR JUNCTION CAPACITY7.1 Junction capacity is calculated for each traffic stream entering the junction and can therefore be

thought of as an entry capacity. A summary of the capacity formulae for the various junction types isgiven below. Details of how to determine the various geometric parameters are given in Part 8 Chapter8. The user is, however, referred to the relevant ARCADY, OSCADY and PICADY or Department forTransport guidance for further information.

Major/Minor Priority Junction Capacity

7.2 In general for 3 and 4 arm major/minor junctions, the entry capacity of a stream has the followingform:

7.3 Research has shown that even with extremely high traffic flows on the major road some traffic willalways force a way into the traffic stream. Therefore there is a minimum capacity value of 60 pcu/hourfor the minor road and major right to minor road movement. This value is then divided equally betweenthe number of approach lanes.

Roundabout Capacity

7.4 The capacity (Qe) of each arm of the roundabout in pcu/hr is given by the equation:

Capacity Q Equation 8/7/1

where X is a geometric term incorporating widths and visibility (see PICADY manual)

C is the saturation flow, and= 745 for the left turn out of and the right turn into the minor road (pcu/hr)

or = [627 + 14 x central reservation width (m)] for the right turn into the major road (pcu/hr)

Y = [1 - 0.0345 x major road width (m)]. It determines the overall effect of the flow factor

Flow Factor comprises the flows which inhibit the movement in question weighted according to the effect of each. The flows are in pcu/hr.

Equation 8/7/2

where the terms are defined as:

k =

F = 303 x2Qc = the circulating flow across the entry in pcu/hrfc =

tD =

x2 =S =

X C Y Flow Factor–

Qe k F f cQc– =

1 0.000347 30– – 0.978 1r--- 0.05– –

0.210tD 1 0.2x2+

1 0.5 / 1 D 60–10

---------------- exp+

+

v e v– / 1 2S+ +1.6 e v– / 1

Chapter 7 Volume 15 Section 1Formulae for Junction Capacity Part 8 Junctions in NESA


For grade separated roundabouts, the entry capacity of a slip road into the roundabout is calculatedfrom a modified form of the above equation, where capacity:

is the entry capacity flow in pcus for a motorway slip road at a grade separated roundabout.

7.5 This relationship is not the most recently derived one for grade separated roundabouts (as reported inTRRL Research Report 142) which has been incorporated into ARCADY. This is because the latestformula may give unrealistic results at the low circulating flows which may be found within NESAduring certain periods of the day. ARCADY should of course be used for junction design (seeParagraph 2.3) when peak hour flows will be used, and the user can decide on the realism of the results.Although the formula used in NESA is not the most recent it remains perfectly good, for assessmentpurposes. RR142 states that “the practical implications of the relationship (that is, the latest formula) ascompared to existing predictive methods are not great”.

7.6 Geometric items e, v, l', D, r, are in metres and is in degrees. They are defined below and details ofhow they are measured are given in Part 8 Chapter 9. The following list gives the equivalent NESAdata record headers:

v AWID- Approach road half widthe EWID- Entry width measured along the normal to the outer kerb liner ERAD- Entry radius measured as minimum radius of curvature of the

nearside kerb at entry FI - Entry angle (geometric proxy for the conflict angle between

entering and circulating streams)l´ FLEN - Average effective length over which the flare is developedD DIAM- Inscribed circle diameter

Capacity of Traffic Signals

7.7 Saturation flows in pcu/hr are calculated by movement within a lane. They are dependent upon thegeometric parameters stop line width and gradient. Allowances for unopposed movements, opposedmovements with hooking and opposed movements without hooking (see OSCADY User GuideChapter 4.6.4) are made, using formulae detailed in TRRL Research report 67.

7.8 The cycle and effective green times are derived using a Webster and Cobbe type algorithm.Calculations are based on the inflows and saturation flows on each lane, and the user can input losttime per cycle, where generally lost time (L) is equal to the sum of each intergreen (IG) minus 1 secondper intergreen, that is:

Practical capacity is determined for each lane using the formula:

Equation 8/7/3

Equation 8/7/4

Equation 8/7/5

where is the theoretical capacityS is the saturation flow (see OSCADY manual)g is the green timec is the cycle time

Qe k 1.1F 1.4f cQc– =

L IG 1– =

0.9 S.g/c=

Volume 15 Section 1 Chapter 7Part 8 Junctions in NESA Formulae for Junction Capacity


Note: the cycle time and green times are calculated separately for each year and flow group, and cycletime increases up to a limit of 120 seconds as the flow increases.

Capacity of Merges at Grade Separated Junctions

7.9 Research has shown that the capacity of high standard links is

2330/(1 + 0.015P) vehicles per hour per lane for motorways, and2100/(1 + 0.015P) vehicles per hour per lane for all-purpose dual carriageways

where P is the proportion of heavy vehicles.

Capacity of Signalised Roundabouts

7.10 The facility to model signalised intersections is not available in the current release of NESA. If thisfacility is required Transport Scotland advice should be sought.

7.11 The NESA model has been developed for fully signalised roundabouts with between three and sixnodes, which operate on a common cycle time. It is assumed that all nodes can be modelled astwo-stage junctions irrespective of how they are controlled on-street.

7.12 The capacity of a signalised link is determined using the formula:

7.13 Signal timings are selected such that entry arms are more saturated than the circulating carriagewaywhich therefore has short queues. The capacity of each arm to the roundabout is determined entirely bythe entry arm and not by the downstream links.

7.14 The saturation flow for an entry arm (excluding any flare) is calculated according to the formulae inTRL Research Report RR 67 from the relationship:

Curvature is ignored and no account is taken of the proportion of turning traffic.

7.15 In order to estimate the saturation flow of the flare on an entry arm it is necessary to calculate thenumber of vehicles that can be accommodated in the flare by assuming that each pcu occupies 5.75metres. Then, providing the clearance time is less than the effective green time, the additionalsaturation flow is calculated by assuming cycle time. The estimate is then reviewed at the next iterationof the optimisation procedure. If the approach has been widened over such a length that the flare will

Equation 8/7/6

where S is the saturation flow of the linkg is its effective green timec is the cycle time of the junction

Equation 8/7/7

where LANES is the number of lanes in the approach (excluding flare)FLAG is 0 for downhill approaches and 1 for uphill approachesGRAD Gradient of approach over last 60m if uphill (%)WIDTH is the stopline width (excluding flare) in metres

Capacity S.g/c=

SATFLOW LANES 1940 42 FLAG GRAD – 325– WIDTH 100 pcu/hour+=

Chapter 7 Volume 15 Section 1Formulae for Junction Capacity Part 8 Junctions in NESA


not be cleared during the effective green time then the lanes in the flare should be included in theapproach width.

7.16 The saturation flow for the circulating stop lines is calculated from the relationship given in paragraph7.14 and reduced by 10% as it has been found that the capacity of circulatory arms are usually lowerthan that for entry arms. It is assumed that approaches to circulating stop lines are not flared and havezero gradient. In circumstances where these assumptions are not valid the user is able to inputsaturation flows directly.

7.17 The number of vehicles which can be accommodated on a circulatory link is calculated from:

Number (pcu) = link length (m) x saturation flow (pcu/hour) 5.75 1850

Volume 15 Section 1 Chapter 8Part 8 Junctions in NESA Geometric Parameters


8 GEOMETRIC PARAMETERS8.1 The junction capacity formulae detailed in Part 8 Chapter 7 require the user to input geometric

parameters of each junction modelled. The determination of these is given below:

Major/Minor Junctions - Geometric Parameters

8.2 The measurement of the geometric parameters used in the geometric term X, given in Part 8 Chapter 7,is illustrated in the following figures.

8.3 Major road width (WP). The four parts of Figure 8/8/1 show the main components of major road width.

Figure 8/8/1: Major Road Width

Chapter 8 Volume 15 Section 1Geometric Parameters Part 8 Junctions in NESA


8.4 Lane width for non priority streams, w

Where there are clear lane markings the width is measured directly. The average of measurementstaken at 5 metre intervals over a distance of 20 metre upstream from the give way point is used. Anymeasurement exceeding 5 metres is reduced to 5 metres before the average is taken. Where lanemarkings are unclear (or absent), measurements are taken as illustrated in Figure 8/8/2 and the lanewidth calculated according to:

8.5 Where the width of minor road is insufficient to accommodate separate lanes for left and right turningvehicles measurements a to e are made across the shared lane from the kerb line to the centre line of theminor road.

Equation 8/8/1

Figure 8/8/2: Lane Width for Non-Priority Streams

w a b c d e+ + + + / 5=



8.6 Visibility distances for the minor road streams, VL and VR minor

These are measured from points 10m back from the give way line on lines bisecting each lane.Visibility to the left VL is measured from the offside lane to a line bisecting the far major roadcarriageway. Visibility to the right, VR, is an average of the measurements made from each lane to aline bisecting the near major road carriageway (See Figure 8/8/3). All measurements are made at aheight of 1.05m above the carriageway surface.

8.7 Visibility distance for the major road right turning stream, VR major

One visibility measurement is made, from the mid point of the right turning lane, or the positionassumed by vehicles waiting to turn right, towards the oncoming major road traffic, at a height of 1.05metres. Note: VR major = FT.

8.8 Advice Note TD42

The user must also indicate whether or not the visibility requirements detailed in paragraph 8.2 ofTD42, are met or not, as this effects geometric delay (see Part 8 Chapter 5)

Roundabouts - Geometric Parameters

8.9 The measurement of the geometric features required for the program to calculate capacities and delays(see Paragraph 8.7.4) is described in the following figures.

8.10 The approach half width v (AWID) is measured at a point in the approach upstream from any entryflare, from the median line to the nearside kerb along a normal. See Figure 8/8/4.

8.11 The entry width, e (E-WID), is measured from the point A along the normal to the nearside kerb inFigure 8/8/4.

8.12 The entry radius, r (E-RAD), is measured as the minimum radius of curvature of the nearside kerb lineat entry (see Figure 8/8/4). For some designs the arc of minimum radius may extend into the followingexit, but this is not important provided that a half or more of the arc length is within the entry region.

8.13 The entry angle, Ø (FI), serves as a geometric proxy for the conflict angle between entering andcirculating streams. Three constructions are used; the first two apply to well defined conventional

Figure 8/8/3: Measurement of Visibility Distances VL and VR



roundabouts and the third to all other types. For conventional roundabouts (i.e. those with identifiablyparallel sided weaving sections) the construction is illustrated in Figures 8/8/5 and 8/8/6.

8.14 Figure 8/8/5 refers to roundabouts with approximately straight weaving sections in which the lineparallel to the weaving section is AD where the point A is as in Figure 8/8/4 and D is the point nearestto A on the median island (or marking) of the following entry. Figure 8/8/6 shows the equivalentconstruction for roundabouts with curved weaving sections (for which the line AD is clearly notparallel with the weaving section). A' D' replaces AD as the line parallel to the weaving section.

Figure 8/8/4: Geometric Parameters of Roundabouts

Figure 8/8/5: Entry Angle on Straight Weaving Sections



8.15 In both cases the line BC is a tangent to the line EF, which is midway between the nearside kerb lineand nearside edge of any median island. Ø is the angle between BC and AD (Figure 8/8/5) or the anglebetween BC and the tangent to the line A'D' at the point of intersection of the two lines (Figure 8/8/6).

8.16 On roundabouts with short weaving areas the construction is as in Figure 8/8/7.

8.17 Here, the line BC is as in Figure 8/8/7 and the line GH is the tangent to the line JK. JK is as EF but forthe following exit i.e. midway between the nearside kerb and nearside edge of any median island whereK is a point on the outer boundary of the roundabout circulation. L is the intersection of BC and GH,and Ø is defined by the following:

Ø = 90 - ½ (angle GLB)

This applies if the right hand side is positive. But Ø = 0 if the right hand side is zero or negative. GLBis the angle measured on the outside of the roundabout, that is on the side facing away from the centralisland.

8.18 The practical difference between this and the previous constructions is that in Figures 8/8/5 and 8/8/6 Øis independent of the angle at which the following exit joins the roundabout whereas in Figure 8/8/7 Øtakes account of this angle. The reason is that for roundabouts with appreciable separation betweenentry and following exit (conventional roundabouts) the direction of circulating traffic depends on thealignment of the weaving section and is largely independent of the geometry of the following exit, but

Figure 8/8/6: Entry Angle on Curved Weaving Sections

Figure 8/8/7: Entry Angle on Short Weaving Areas



when the separation is smaller (as for off side priority roundabouts) circulating traffic which leaves atthe following exit traces a path determined in part by the angle at which that exit joins the roundabout.The conflict angle reflects this difference.

8.19 The flare length (F-LEN) is defined as half the length over which the flare is developed. The line CF' isparallel to BG and distance (e-v)/2 from it. Usually CF' is therefore curved and its length is measuredalong the curve. The construction is illustrated in Figure 8/8/8.

8.20 The inscribed circle diameter (DIAM) is the diameter of the largest circle that can be inscribed withinthe junction outline (see Figure 8/8/4). In cases where the outline is asymmetric, the local value in theregion of the entry considered is taken.

Figure 8/8/8: Flare Lengths for Roundabouts



Signalised Roundabouts - Geometric Parameters

8.21 The measurement of the geometric features required by the program to calculate capacities and delaysis described below and shown on Figure 8/8/9.

8.22 The entry width (EWID) is the width of the approach measured normal to the offside kerb and isentered to the program as the average width per lane. It should exclude any flare width (FWID) whichis entered separately.

8.23 The flare width (FWID) is the width of the flare at the stop line measured normal to the kerb and isentered to the program as the average width per lane. It should exclude the entry width (EWID) whichis entered separately.

8.24 The flare length (FLEN) is defined as for conventional roundabouts as half the length over which theflare is developed. Flares less than 5m in length should be ignored.

8.25 The distance from entry link stop-line to next downstream stop-line (JDIST) is the distance travelled onthe ground and not the straight line distance.

8.26 The circulating width (CWID) is the width of the approach to the circulating stopline measured normalto the offside kerb and is entered to the program as the average width per lane. No allowance is madefor any gradient that may exist.

8.27 The inscribed circle diameter (DIAM) is the diameter of the largest circle that can be inscribed withinthe junction outline. This parameter is used for the calculation of geometric delay, and in cases wherethe outline is asymmetric an average value should be used.

Figure 8/8/9: Geometric Parameters for Signalised Roundabouts



VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


VALIDATING A NESA ASSESSMENT

Contents

Chapter

1. NESA Traffic Model Calibration andValidation

2. Goodness of Fit Statistics

3. Economic Assessment Validation

4. A Summary of the Items of Cost andBenefit in NESA

5. The Timing and Documentation for NESAValidations

PART 9

Volume 15 Section 1Part 9 Validating a NESA Assessment


Volume 15 Section 1 Chapter 1Part 9 Validating a NESA Assessment NESA Traffic Model Calibration and Validation


1 NESA TRAFFIC MODEL CALIBRATION AND VALIDATION

1.1 In a manner similar to most traffic models, the output of the NESA assignment model should beassessed more critically than that of any other models external to NESA (e.g. a trip distribution model).This is to be expected since it is the assignment model which produces the link flows by which theperformance of the complete model is ultimately judged, and by which the effects of road proposals areassessed in operational and environmental terms and, indirectly, in economic terms. A close match (inthe statistical sense) between base year assigned and observed flows and journey times, will not onlylend credibility to the forecasting ability of the model, but to the whole modelling process in general.

1.2 Traffic model calibration is the first stage of the process that ensures the traffic model adequatelyrepresents the base year situation. Traffic model validation is the second stage. Often the calibrationand validation of networks occur simultaneously with trip matrix calibration and validation, principallybecause the validity of the trip matrix can be assessed within the assignment model. The local trafficmodel validation report should cover both the calibration and the validation processes for both thenetwork and the trip matrices (see DMRB Volume 12).

1.3 The principal difference between calibration and validation is that during calibration the traffic modelcan be altered so that the overall fit is improved. Model validation acts as an independent check on thecalibration.

Model Calibration

1.4 Model calibration is generally undertaken through thorough network checking, followed by a finetuning of the network and assignment process (e.g. adjustment of the tree building parameters b, P andthe number of trees (see Part 5 Chapter 5)). Goodness of fit statistics (see Part 9 Chapter 2) are used toassess how well a model fits the base year situation. If model fit in certain areas is still poor after allattempts at calibration it is often worth assessing the effect that trip matrix errors have on the fit.

NETWORK CHECKING

1.5 The road network coding is subject to:

(i) network coding mistakes (e.g. coding a link length as 15m when it has been measuredas 150m)

(ii) measurement errors in the characteristics of the links included (e.g. speed anddistance)

(iii) specification errors in the structure of the network (e.g. the selection of and locationof links, and the number of and location of centroid connectors)

1.6 Comprehensive data checks should be made on all aspects of the data. A check for detecting any errorsin link lengths is to compare coded link lengths with crow-fly distances. The crow-fly distances can becalculated using the grid references of nodes at each end of a link. It is recommended that link lengthsand grid references should be examined for any link where the ratio of coded to crow-fly distance isless than 0.98 or greater than 1.4. It should be noted that some straight short links can produceacceptable ratios of less than 1.0 where link lengths are coded to an accuracy of 0.1 kilometres.

1.7 An important check is to ensure that all banned turns have been coded correctly. Analysis of the routesbetween key O-D pairs may identify the omission or mis-coding of a banned turn (see PRINT PATH,Part 10 Chapter 16).

Chapter 1 Volume 15 Section 1NESA Traffic Model Calibration and Validation Part 9 Validating a NESA Assessment


1.8 Manually produced assignment plots and goodness of fit statistics (see Part 9 Chapter 2) are also usefulin identifying parts of the network where data errors may exist. Goodness of fit statistics give ameasure of the fit between model traffic flows on links and across screenlines (see DMRB Volume 12)and traffic counts. Table 9/2/1 gives advice on acceptable values of the main validation measuresmentioned in Part 9 Chapter 2.

1.9 A comparison of observed journey times to link travel times will also assist in highlighting any linkspeed errors within the traffic model. Cumulative time-distance graphs can aid this comparison.DMRB Volume 12 gives advice on journey time data collection. It is recommended that where journeytime surveys are undertaken they should distinguish between link travel times and junction delays.Table 9/2/1 gives advice on acceptable validation measures for journey time comparisons.

1.10 An analysis of routes for key trip movements within the model may also help in this network checkingprocedure. Unlikely routes may indicate the presence of network data errors.

FINE TUNING THE NETWORK AND ASSIGNMENT PROCESS

1.11 Once all network data has been checked thoroughly and the user is satisfied that it is correct, modelcalibration can proceed to the next stage, that of fine tuning the network. Fine tuning is defined as theprocess where slight alterations are made to the network and tree building parameters to achieve abetter model fit. Network alterations may include such things as altering the point at which a zoneconnects on to the network, altering a link’s coded speed or adding in an extra link.

1.12 To achieve acceptable calibration it is necessary to have a statistically close fit between observed andmodelled traffic conditions. DMRB Volume 12 gives advice on how model calibration results shouldbe presented in the traffic model validation report. It should be noted that goodness of fit statistics (seePart 9 Chapter 2) may be used to support model calibration comparisons.

1.13 In addition to goodness of fit statistics, assignment plots, link volume to capacity ratios, journey timecomparisons and checks on the realism of O-D pair routeings are useful at identifying areas of thenetwork that may need altering to improve the model’s fit.

1.14 Once an area of a model has been identified as having a poor fit, a number of options are available tothe user. The particular solution will vary according to circumstances. Listed below are a number ofcommon problems and possible solutions. For the solutions suggested it is assumed that all networkdata has been checked for accuracy. It should also be noted that there may be other solutions, thoughthese will be model and area dependent.

1.15 Typical problems are:

(i) A link’s assigned flow may exceed its capacity/observed count.

This problem may be remedied by:

- lowering the attractiveness of the over-capacity link by lowering its coded speed. For certain urban links (road categories 1 to 7) it may be appropriate to specify the central flag (see Part 5 Chapter 3) which automatically lowers the default link speed by 10kph (for lights).

- checking to see if a zone’s connector loads onto the problem link. If it does either moving the connector or adding in an additional connector to link the zone to a different part of the network may help solve the problem.

- checking to see if the model accurately represents the road network in the problem area. For example, it may be found that a road in the actual road

Volume 15 Section 1 Chapter 1Part 9 Validating a NESA Assessment NESA Traffic Model Calibration and Validation


network, used by say rat-runners has been omitted from the model. The inclusion of this link into the model may solve the problem.

(ii) Too much traffic is assigned to certain routes or links and little or none to others.

The alteration of a number of assignment variables (distance coefficient, perturbationfactor and the number of trees per origin) or changing a link’s description may helpsolve this problem:

- the higher the value of the distance coefficient b the more route choice is dependent on distance rather than time. Evidence does however suggest that the best fit for an assignment model (as a whole) is not particularly sensitive to coefficients in the correct range (see DMRB Volume 12).

- the higher the perturbation factor P the greater the variance from the strict minimum path is likely to be for any tree built. A high perturbation factor therefore has the effect of spreading traffic more evenly over the network.

- increasing the sets of trees built per origin will decrease the chance of a high probability route being selected and increase the chance of a low probability route being selected.

- a long link with a large travel cost (C) can experience significant variations in cost during the assignment process (see Equation 5/5/2). This can lead to either too much or too little traffic being assigned to it. Splitting the link into two or more shorter ones can reduce the random variation experienced and therefore solve the problem. Links that suffer from this problem will usually only be found on the periphery of the model.

(iii) Not enough traffic is assigned to a high quality route (e.g. a motorway) and toomuch traffic is assigned to parallel (competing) routes with lower designstandards.

This may occur where for example there is little difference between the cost of thehigh quality route and the competing routes. If this is the situation there is a highprobability that after all link costs have been randomised the minimum cost route willlie in the dense all purpose road network (i.e. on a competing route).

A number of solutions may be possible in this situation such as:

- lowering the perturbation factor (the preferred route will lie closer to the strict minimum path)

- lowering the distance coefficient (because the high quality route may not be the most direct route)

- increasing the cost of the competing routes, through for example lowering link speeds, and/or increasing junction delays (see also junction indices).

Chapter 1 Volume 15 Section 1NESA Traffic Model Calibration and Validation Part 9 Validating a NESA Assessment


TRIP MATRIX ERRORS AND MODEL CALIBRATION

1.16 If model fit problems still persist even after all obvious network coding errors have been corrected andattempts have been made at fine tuning the network and assignment process, then it is suggested thatthe trip matrix data should be re-assessed. Errors in the trip matrix may lead to significant variations inassigned link flows making the model very difficult to calibrate. DMRB Volume 12 describes a methodthat could be used to test the sensitivity of modelled link flows to trip matrix errors. If the results of thistest indicate the NESA traffic model is sensitive to trip matrix errors, then the model fit might beimproved through better trip matrix calibration and validation (see DMRB Volume 12).

Traffic Model Validation

1.17 The importance of traffic model validation cannot be overstressed. DMRB Volume 12 provides athorough description of the procedures involved and the reporting required. It is not intended that theadvice contained in DMRB is duplicated here, therefore only a description of the principle aims oftraffic model validation is described.

1.18 Traffic model validation is the name given to the procedure that is concerned with the assessment of thevalidity of a traffic model and of the adequacy of its output as a base for forecasting. The maincomponent of any validation procedure is to compare the assigned link flows estimated by a calibratedmodel against independent observations (generally traffic counts).

1.19 As such traffic model validation is a repeat of traffic model calibration with independent data. The onlydifference is that no network changes can be undertaken. DMRB Volume 12 gives advice on modelvalidation procedures. It should be noted that if a model performs adequately at validation, then theindependent data may be absorbed into the traffic model to improve calibration (see DMRB Volume12).

Local Model Validation Report

1.20 A local model validation report, covering all aspects of model calibration and validation, shouldbe submitted as part of every trunk road appraisal. The user is again referred to DMRB Volume12 for a full description of the elements required.

Volume 15 Section 1 Chapter 2Part 9 Validating a NESA Assessment Goodness of Fit Statistics


2 GOODNESS OF FIT STATISTICS2.1 A number of factors are likely to help decide whether a particular modelled link flow is acceptably

close to an observed one; for example, the position of the link in the network, its importance in thecontext of the study and its proximity to the test scheme. It is also helpful to have some kind ofobjective guideline for indicating when there may be a problem.

2.2 DMRB Volume 12 suggests that such a guideline can be obtained through a comparison betweenmodelled flows and the 95% confidence interval for observed flows. In fact a requirement of a trafficmodel validation report is that a tabular and graphical presentation of modelled link flows and 95%confidence intervals of observed flows is made (see Equation 9/2/1). If the degree of accuracy in themodelled flow is known (e.g. a 95% confidence interval of +/-5%) DMRB also suggests that this isincorporated in the presentation.

TRAFFIC COUNT 95% CONFIDENCE INTERVAL

2.3 DMRB Volume 12 gives advice on the accuracy of the data collection methods used to derive theobserved flow. This information can be used to calculate 95% confidence intervals for the observeddata.

2.4 Other goodness of fit statistics, detailed in Equations 9/2/2, 9/2/3 and 9/2/4 can also provide suchguidelines. These guidelines allow the user to quickly identify links where a close match betweenmodelled and observed flows has been obtained. The goodness of fit statistics should be used to makeindividual and screenline comparisons between modelled and observed flows (see DMRB Volume 12).

ABSOLUTE DIFFERENCE

RELATIVE DIFFERENCE

GEH SUMMARY STATISTIC

Equation 9/2/1

where X% is calculated using information from DMRB

Equation 9/2/2

Equation 9/2/3

Equation 9/2/4

where GEH is the GEH statisticM is the modelled flowC is the observed flow

95% Confidence Interval Observed Flow +/- X%=

Absolute Difference = Modelled Flow - Observed Flow

Relative Difference Modelled FlowObserved Flow------------------------------------=

GEH M C– 2

M C+ /2-------------------------=

Chapter 2 Volume 15 Section 1Goodness of Fit Statistics Part 9 Validating a NESA Assessment


2.5 The above goodness of fit statistics reflect two kinds of error. Firstly, that associated with the modelledflow (reflecting the fit of the model) and secondly that associated with the observed flow. Errors in themodelled flows arise through traffic modelling errors (see DMRB Volume 12), whilst errors in theobserved flows arise through data collection methods (e.g. manual traffic counts) and day to dayvariability in traffic flow (e.g. +/- 15%). The presence of this second error means that the strictapplication of statistical tests (e.g. the Chi-squared test, see DMRB Volume 12) will nearly alwaysindicate that observed and modelled flows are significantly different. The statistical reason for this isthat tests such as the Chi-squared implicitly assume that observed flows are good estimates, when infact they are subject to a great deal of error. A suitable statistical test would be one that assumed theobserved flows were poor estimates.

2.6 The goodness of fit measures detailed above are not statistical tests, but do give some indication of thelevel of fit. It is also recommended that if several observed flows relating to a specific link are availablethen they are all used to assess the model fit. This is again due to the presence of the error in theobserved flows.

2.7 The use of the GEH statistic stems from the inability of either the absolute difference or the relativedifference statistics to cope with flows over a wide range. For example, an absolute difference of 1000vph may be considered a big difference if the flows are of the order of 1000 vph, but would beunimportant for flows of the order of several thousand vph. Equally, a 10% error in 1000 vph would notbe important, whereas a 10% error in say, 30,000 vph might mean the difference between building anextra lane or not.

2.8 A generally acceptable value for the GEH statistic is less than about 5.0, for individual network links,but this will depend to a certain extent on average flow levels and the location of the link(s) concernedin relation to the scheme being appraised. It should also be noted that the GEH statistic has thecharacteristic of weighting the same differences of flow more highly on low flow links than on highflow links. For example, a difference of 2,000 vehicles on a link with an observed flow of 6,000 gives aGEH of 24, whilst a link with an observed flow of 60,000 has a GEH of 8.

2.9 It is recommended that a combination of the above goodness of fit measures is used to assess the fit ofa model. This recommendation is made for two reasons, firstly because of the error associated with theobserved flows and secondly, because each of the above goodness of fit measures exhibit their ownindividual characteristics.

2.10 A further form of presentation that is sometimes used is to plot modelled values against observedvalues and to carry out a correlation analysis between the two sets of values. The correlation coefficient(R) gives some measure of the goodness of model fit. The slope of the best-fit regression line throughthe origin will also indicate the extent to which observed flows are over or under estimated. In the mainarea of influence of the scheme, acceptable values of the correlation coefficient (R) are above 0.95,whilst acceptable values of the slope of the best fit line lie between 0.9 and 1.10. A value of 1.0 for bothstatistics represents a perfect fit. However, misleading results can be obtained where there is a widerange of flows.

2.11 The above statistics (see Equations 9/2/1 to 9/2/4) and any others that are considered appropriate, maybe used to support model calibration comparisons.(see DMRB Volume 12).

2.12 Table 9/2/1 gives advice on acceptable values of the main validation measures mentioned, and suggestsdifferences in emphasis depending on the magnitude of the values being compared. Table 9/2/1 alsogives advice on acceptable validation measures for journey time comparisons (see paragraph 1.10). Amodel that does not meet these guidelines may still be acceptable for appraisal of a given schemeif the larger discrepancies are concentrated away from the area of greatest importance to thatscheme. Conversely, a model that passes the guidelines but has significant discrepancies on themost crucial links may be unacceptable.

Volume 15 Section 1 Chapter 2Part 9 Validating a NESA Assessment Goodness of Fit Statistics


Note: All comparisons are for directional flowsSource: DMRB Volume 12, Section 2, Part 1 - Traffic Appraisal in Urban Areas

Table 9/2/1: Assignment Calibration/Validation: Acceptability Guidelines

Criteria and Measures Acceptability GuidelineAssigned hourly flows compared with observed flows-

1. Individual flows within 15% for flows 700 - 2,700 vph ))

2. Individual flows within 100 vph for flows < 700 vph ) > 85% of cases)

3. Individual flows within 400 vph for flows > 2,700 vph )

4. Total screenline flows (normally > 5 links) to be within 5% All (or nearly all) screenlines

5. GEH statistic:i) individual flows: GEH < 5 > 85% of cases

ii) screenline totals: GEH < 4 All (or nearly all) screenlinesModelled journey times compared with observed times

6. Times within 15% (or 1 minute, if higher) > 85% of routes

Chapter 2 Volume 15 Section 1Goodness of Fit Statistics Part 9 Validating a NESA Assessment


Volume 15 Section 1 Chapter 3Part 9 Validating a NESA Assessment Economic Assessment Validation


3 ECONOMIC ASSESSMENT VALIDATION3.1 A local model validation report will have been produced during traffic model development (see Part 9

Chapter 1). However, this is not sufficient for validation of the economic assessment as this is acompletely separate procedure. This chapter therefore provides guidance on what to check whenvalidating a NESA evaluation printout.

3.2 The documentation required to validate a proposed scheme is listed in Part 9 Chapter 5. The checkingof cost estimates is described in Paragraph 3.9 below.

3.3 The NESA printout is structured in terms of the commands the user specifies. These commands can besplit into three broad categories; Network commands, Assignment commands and Evaluationcommands. The economic validation procedure assumes that the traffic model is valid and therefore isonly concerned with the information contained in the Network and Evaluation parts of the printout. Thefollowing paragraphs identify what checks should be made under each of the command categories.

Error and Warning Messages

3.4 Where necessary error and warning messages are detailed in the NESA printout. Their location in theprintout is determined by the command under which the error occurs. Errors must be corrected sincethey cause the program to terminate prematurely, whilst warning messages should be checked to see ifinput data are either wrongly coded or inappropriate.

Network Commands

3.5 This part of the printout contains information about the network coding which relates to both the trafficmodel and the evaluation model. In validating the economic assessment the user should check thefollowing points:

(i) The base matrix year;

(ii) The traffic and economic growth assumptions;

(iii) The scheme opening year (in the design network title) - is the opening year consistentwith the scheme’s position in the National Road Directorate’s trunk road programme?

(iv) Updates to the base network (i.e. the design network assignment coding) - cross referto the node link diagram. Check link length, road categories and coded speeds.

(v) Link geometric properties - emphasis should focus on the appropriateness of defaultgeometric values and high values for hilliness, bendiness etc.

(vi) Accident rates - is there consistency between the treatment of accident rates in thebase and design networks? Have observed and default accident rates been appliedconsistently throughout each network?

(vii) Total distance by road classification - all differences between the Do-Minimum andDo-Something networks should be rationalised.

Chapter 3 Volume 15 Section 1Economic Assessment Validation Part 9 Validating a NESA Assessment


Assignment Commands

3.6 This part of the printout contains information that principally refers to the operation of the trafficmodel. However, for evaluation purposes it is important to check:

(i) The design network banned turns are coded correctly, particularly at grade separatedjunctions.

(ii) The realism of routeings between O-D pairs which in the Do-Something situationshould experience significant traffic relief.

(iii) Do-Something traffic assignments. Are there obvious assignment errors/inconsistentcoding?

Evaluation Commands

3.7 This part of the printout contains information that is solely concerned with the evaluation. The usershould check the following points:

BASE NETWORK

(i) Network Classification;

(ii) E and M Factors - note use of exceptionally high or low M Factors (compare todefaults by network classification);

(iii) Seasonality Index (compare defaults by network classification);

(iv) Local vehicle category proportions, as a yearly average or by flow group ifstatistically reliable data exists (compare to defaults by network classification);

(v) Local car in work time proportions (compare to defaults by network classification);

(vi) The inclusion of any Do-Minimum scheme costs.

BASE AND DESIGN NETWORKS

(i) Total travel time and vehicle kilometres by road classification. All differencesbetween the Do-Minimum and Do-Something networks should be rationalised.

(ii) Link speeds - compare coded speeds to calculated (by flow group) and determinewhat effect flow is having. A reasonable match between coded and calculated speedsin the base year engenders confidence in the economic results. Discrepancies mayarise through a link being overcapacity (see (v) below). If large discrepancies existbetween coded and calculated speeds, further validation of the calculated speeds canbe derived through journey time information (see Paragraph 3.10).

(iii) Junction coding - note that if junctions are modelled in the base network they shouldalso be modelled in the design network, unless they will no longer exist.

(iv) Junction delays - excessive junction delays should be examined closely. Wherejunctions are close to or overcapacity in early years or outside the peak flow groups,further validation should be obtained through journey time runs (see Paragraph 3.10).The maximum delay (MAXDEL) a junction can experience should also be checked.



(v) Overcapacity links and junctions - this is an important section to check. It showswhich links and nodes are going overcapacity, by year and flow group.

Overcapacity in a scheme’s early years or outside the peak flow groups should bequestioned, by checking if:

(a) the node/link descriptions are correct;

(b) the traffic assignment is reasonable (see Local Model Validation Report);

(c) Do-Minimum improvements are included.

Widespread overcapacity in the Do-Something indicates that the scheme is notsolving the traffic problems efficiently.

DESIGN NETWORK

(i) Total Scheme Costs, including the price base at which the costs have been assessedand the appropriateness of the CPI;

(ii) Scheme cost profile should equal 100%. Is the programme of expenditure realistic?

3.8 The remaining evaluation information in the printout is contained in a series of tables described indetail in Part 10 Chapter 19. These tables should be checked for the following points:

TABLE 1

This shows the NESA user class proportions by flow group. A check should be made to ensure thatthey accord with common sense. For example, Other trips should dominate the peak flow group for atourist route subject to high seasonal flows.

TABLE 2

Check links which have high benefits and/or disbenefits. Are these benefits reasonable? (Cross refer tolink geometric properties, flow reductions, speed changes and v/c ratios).

TABLE 3

Check junctions with large benefits. Are these benefits reasonable? (Cross refer to a junction’sgeometric properties, turning flow changes and v/c ratios). Have Do-Minimum improvements beengiven proper consideration? Also cross refer to Table 5 to assess junction delay and accident benefitsjointly.

TABLES 4 AND 5

Check whether large accident reductions are realistic. Are they the problems which the scheme isdesigned to solve?

TABLES 7 AND 8

Check that the split of benefits by vehicle category and flow group is reasonable (compare with userclass proportions in Table 1) and accords with the main aims of the scheme. Note that in relativelyuncongested situations, benefits will tend to be related to the proportion of traffic in each flow group(duration weighted by intensity) whereas with heavy congestion, benefits will be disproportionatelyhigh in the peak flow groups (particularly with junction relief schemes). Check the type of vehicle that



most of the benefits can be attributed to - does this seem reasonable when compared with localknowledge?

TABLE 10

Check whether the differences between the base and design years are realistic. Check the profile ofemissions benefits over time.

TABLES 11 AND 12

These are the most important tables in the printout, summarising all the cost and benefit items. Checkwhether zero items are justified, for example, maintenance delay savings and construction delaysavings (which will normally be negative and estimated by a QUADRO analysis).

Check the profile of benefits over time. If benefits are relatively low in early years, this could suggestthat the scheme opening date may be too early.

Check relative proportion of accident and link transit/junction delay benefits (accident benefits arenormally 10 to 20% of total benefits).

Check that timing of expenditure and first scheme year of benefits is correct (that is, a smallexpenditure item in the first scheme year). Is the spend profile appropriate to the specific scheme.

TABLES 13 IS NO LONGER INCLUDED IN THE OUTPUT

TABLES 14 AND 15

This summarises the components of the three key items, PVB, PVC and NPV.

Checking Cost Input Data

3.9 The scheme costs input to NESA should be carefully checked to ensure that:

(i) all items have been included (construction, land, preparation and supervision costs)?

(ii) the original local estimates are as up to date as possible?

(iii) the appropriate values of CPI have been used?

(iv) the cost profile is correct?

Journey Time Surveys

3.10 Journey time surveys may have been undertaken as part of the traffic model calibration and validationprocedures (see Part 9 Chapters 1 and 2). However, they may also be necessary to validate a NESAevaluation when:

• there is a large variation in link coded and calculated speeds;

• junctions are close to capacity;

• there is significant junction interaction; and

• travel time savings are a major source of benefit.



3.11 If a close match cannot be obtained between observed local journey time measurements and NESAcalculated link and junction delays, consideration should be given to using more sophisticated trafficmodelling suites, such as SATURN or microsimulation. Transport Scotland advice should be sought.

Conclusion

3.12 When checking a NESA printout, it is recommended that the printout be read backwards! The user costand benefit tables will indicate which are the critical elements in the benefit calculation and theseshould be scrutinised in the input data summaries.

3.13 A suggested order for examining a NESA printout is:

(i) Error and Warning messages.

(ii) Tables 14and 15- what are the most important items of benefit and cost? What is thescheme achieving?

(iii) Network Commands printout information - check coding against node link diagramand scheme plan. Does the network represent the scheme reasonably?

(iv) Evaluation Commands - check in detail overcapacity reports, journey times anddistances for representative routes (and compare with actual or expected), andjunction coding.

Volume 15 Section 1 Chapter 4Part 9 Validating a NESA Assessment A Summary of the Items of Cost and Benefit in NESA


4 A SUMMARY OF THE ITEMS OF COST AND BENEFIT IN NESA

Reporting Requirements

4.1 The requirements of Transport Scotland for the reporting of road scheme assessments are documentedin DMRB 5.1.2 and 5.1.4 (TD 37/93 Scheme Assessment Reporting and SH 1/97 The Traffic &Economic Assessments of Road Schemes in Scotland respectively). The economic results form onlyone part of the assessment summaries documented in TD 37/93 and SH 1/97, the other componentsbeing environmental and operational assessments.

4.2 In addition, the Scottish Transport Appraisal Guidance (STAG) contains guidance on the appraisal ofall transport policies and projects throughout Scotland and details additional reporting requirementsincluding the completion of Appraisal Summary Tables (ASTs). The STAG appraisal centres on theScottish Government’s five objectives, namely: environment; safety; economy; integration andaccessibility. The results from a NESA assessment (NESA02 onwards) form part of the input to theeconomy and safety sections of the STAG ASTs.

4.3 The reader is directed to TD 37/93 and SH 1/97 plus STAG for full details of Transport Scotland’sreporting requirements for trunk road scheme assessments.

Transport Appraisal Guidance

4.4 Following the introduction of the (then) DTLR’s Transport Economics Note (TEN) (March 2001) themethod of cost-benefit analysis changed from one based on social costs and benefits to one based onwillingness to pay. This meant a move away from reporting the results of economic appraisals inresource costs to one where the results are now expressed in the market price unit of account.

4.5 The reader is directed to TAG Unit 3.5.4, Cost Benefit Analysis (January 2014), for a full discussion onthe methodology.

Tables 14 & 15

4.6 This chapter describes how NESA brings together the various elements of the appraisal and presentsthe results in summary tables.

4.7 Tables 14 and 15A, B and C are discussed in more detail below.

Table 14 - Conversion of Travel Costs to Market Prices by Vehicle Category

4.8 Table 14 of the NESA output (see Table 9/4/1 below) shows the calculations necessary to convert thetime and vehicle operating cost changes calculated in resource costs to market prices. The individualcomponents are presented under the TEE categories and converted to market prices by the appropriatetax correction factors.

Table 15A - The Economic Efficiency of the Road System in Market Prices

4.9 Table 15A of the NESA output (see Table 9/4/2) is an adaptation of the TEE Table (see TAG Unit 3.5.2,The Transport Economic Efficiency Sub Objective). Table 9/4/2 shows how the elements of the TEEtable calculated by NESA are transferred from Table 14 and combined with any Delays DuringConstruction and Maintenance Delay Savings to produce the Net Consumer User Benefit and NetBusiness Impact.

Chapter 4 Volume 15 Section 1A Summary of the Items of Cost and Benefit in NESA Part 9 Validating a NESA Assessment


4.10 Any travel time and vehicle operating costs during construction or maintenance (calculated externallyby users via QUADRO, see DMRB Volume 14) should be input manually by users to Table 15A inmarket prices and allocated between Consumers and Business in proportion to the Consumer andBusiness User (Time and VOC) benefits of the scheme under normal operating conditions.

Table 15B - Public Accounts

4.11 Table 15B of the NESA output (see Table 9/4/3) shows the summary of Public Accounts (see TAG Unit3.5.1, Public Accounts).

Table 15C - Analysis of Monetised Costs and Benefits

4.12 Table 15C of the NESA output (see Table 9/4/4) summarises the monetised costs and benefits,including any accident benefits and savings in carbon emissions.

4.13 In many cases there may also be other significant costs and benefits associated with a scheme, some ofwhich cannot be presented in monetised form (see TAG Unit 3.5.1, Public Accounts and the ScottishTransport Appraisal Guidance). Where this is the case, the analysis presented in Table 15C does NOTprovide a good measure of value for money and should not be used as the sole basis for decisions.

Local Government Funding

4.14 Table 15B, Public Accounts, includes entries for both Local Government Funding and CentralGovernment Funding. NESA assumes Central Government Funding only and therefore only completesentries in Table 15B under this heading. If a scheme receives any Local Government Funding, in theform of Investment Costs, Operating Costs, Maintenance Costs, Developer Contributions or Grant/Subsidy Payment these need to be manually input to Table 15B.

Developer Contributions

4.15 Where schemes receive Developer Contributions these should be treated as negative grants and shouldbe recorded both as a cost to the private sector and a benefit to the public sector (see STAG Chapter 8).

4.16 Developer and Other Contributions therefore appear in both Table 15A and Table 15B and should beentered manually by users in both tables as negative numbers.

4.17 Any Developer (or Other) Contributions will therefore have no effect on a scheme's overall Net PresentValue (as they will cancel each other out as benefits and costs). However, they will increase a scheme'sBenefit to Cost (to government) Ratio as the final cost to government will be reduced.

To Calculate the Indirect Tax Revenues

4.18 The calculation of the change in Indirect Tax Revenues is given below. For an explanation of theelements DD, NN etc. see Table 9/4/1. The tax rates used in the calculation, and the 2002 defaults, are:

Because tax rates change over time the calculation must be carried out for each year of the appraisal.

For Work Trips the Indirect Tax Revenue is:

= DD x tF' Cars / (1 + tF' Cars) + (NN + QQ) x tF' OGV / (1 + tF' OGV)

Using the 2002 defaults the third term is zero and the equation is reduced to:

= DDx 0.734 + (NN+ QQ) x 0.713



For Commuting Trips the Indirect Tax Revenue is:

= EE x (tF - t) / (1 + tF) + HH x (tN - t) / (1 + tN)

Using the 2002 defaults the second term is negative and the equation reduces to:

= EE x 0.728 - HH x 0.029

For Other Trips the Indirect Tax Revenue is:

= FF x (tF - t) / (1 + tF) + II x (tN - t) / (1 + tN)

Using the 2002 defaults the second term is negative and the equation reduces to:

= FF x 0.728 - II x 0.029

The entry in Table 15B is the sum of the Work, Commuting and Other calculations above. If morefuel is used the value calculated will be negative and the Government receives more tax revenue.

The adjustment for Work Operating Fuel is:

= x (1 + tF') x (1 + t) -

In NESA, Indirect Tax Revenues contribute to the scheme benefits (PVB), rather than the scheme costs(PVC). However, they are still presented in the Public Accounts table (Table 15B) - see ‘TAG Unit3.5.1 - The Public Accounts Sub-Objective, January 2014’ and ‘TAG Unit 2.7.1 - Appraisal And TheTreasury Green Book, January 2014’ for further details.

t 0.209 average rate of indirect tax on final consumption in the economytF Car 3.45 rate of indirect tax on car fuel as a final consumption goodtF OGV 3.10 rate of indirect tax on OGV fuel as a final consumption goodtF' Car 2.77 rate of indirect tax on car fuel as an intermediate consumption goodtF' OGV 2.49 rate of indirect tax on OGV fuel as an intermediate consumption goodtN 0.175 rate of indirect tax on non-fuel vehicle operating costs as a final consumption goodtN' 0 rate of indirect tax on non-fuel vehicle operating costs as an intermediate

consumption good



Notes:(i) t is the average rate of indirect tax on final consumption in the economy(ii) tF is the rate of indirect tax on fuel as a final consumption good(iii) tF' is the rate of indirect tax on fuel as an intermediate consumption good(iv) tN is the rate of indirect tax on non-fuel vehicle operating costs as a final consumption good(v) tN' is the rate of indirect tax on non-fuel vehicle operating costs as an intermediate consumption good(vi) The adjustment for Work Operating Fuel is: x (1 + tF') x (1 + t) -

Table 9/4/1: Conversion of Travel Costs to Market Prices by Vehicle Category (Table 14 of the NESA output)

Vehicle Category Time TotalTime

Operating Fuel Operating Non-fuel TotalOperating

CostsWork Commute Others Work Commute Others Work Commute Others

Personal Travel

Car and Private LGVTotal Adjustment x t x t x t (row) see note

vi x tF x tF x t x tN x tN (row)

Market Price = AA = BB = CC = DD = EE = FF = GG = HH = II

Bus & CoachAdjustment x t x t x t (row)Market Price = JJ = KK = LL - - - - - - -

Freight

Freight LGVOGV1OGV2Total Adjustment x t (row) see note

vi x t (row)

Market Price = MM = NN = OO

Private Sector

Bus & CoachAdjustment x t (row) see note

vi x t (row)

Market Price = PP = QQ = RR

Market PriceTotals This analysis is based on <growth> traffic growthCosts are in <PV year> prices in multiples of a million pounds and are discounted to <PV year>Evaluation period <evaluation period> yearsFirst scheme year <first scheme year>Current year <current year>Discount rate 3.5% for first 30 years, thereafter 3.0% for 45 years, thereafter 2.5%



Table 9/4/2 : The Economic Efficiency of the Road System in Market Prices (Table 15A of the NESA output)

IMPACT Reference Total Cars & LGVs

OGVs Buses & Coaches

Non-Business User BenefitsTravel Time

Commuting Travel Time 1 (row) BB - KKOther Travel Time 2 (row) CC - LLNon-business Travel Time 3 1+2

Vehicle Operating CostsCommuting Fuel VOC 4 (row) EE - -Commuting Non-fuel VOC 5 (row) HH - -Other Fuel VOC 6 (row) FF - -Other Non-fuel VOC 7 (row) II - -Non-business Vehicle Operating Costs 8 4+5+6+7

During Construction and MaintenanceCommuting During Construction and Maintenance 9 See note (ii) - - -Other During Construction and Maintenance 10 See note (ii) - - -

Net Non-Business Benefits: Commuting 11 1+4+5+9 S - SNet Non-Business Benefits: Other 12 2+6+7+10Net Non-Business Benefits: Total 13 11+12Business User BenefitsUser Benefits

Business Travel Time 14 (row) AA MM JJ+PPFuel VOC 15 (row) DD NN -Non Fuel VOC 16 (row) GG OO -Business Vehicle Operating Costs 17 15+16

During Construction and MaintenanceDuring Construction 18 See note (ii)During Maintenance 19 See note (ii)During Construction and Maintenance 20 18+19

Net Business User Benefits 21 14+17+20Private Sector Provider Impacts

Revenue 22 See note (iii) - - -Fuel VOC 23 (row) - - QQNon Fuel VOC 24 (row) - - RRPrivate Sector Vehicle Operating Costs 25 23+24 - - -Investment Costs 26 See note (iii) - - -Grant/Subsidy 27 See note (iii) - - -

28 22+25+26+27Other Business Impacts

Developer & Other Contributions 29 See note (iv) - - -Net Business Impact 30 21+28+29 S S STotal Present Value of TEE Impacts 31 13+30 S S SThis analysis is based on <growth> traffic growthCosts are in <PV year> prices in multiples of a million pounds and are discounted to <PV year>Evaluation period <evaluation period> yearsFirst scheme year <first scheme year>Current year <current year>Discount rate 3.5% for first 30 years, thereafter 3.0% for 45 years, thereafter 2.5%



Notes:

(i) Items AA to RR referenced in Table 9/4/1.

(ii) Any travel time and vehicle operating costs during construction or maintenance (calculatedexternally using QUADRO, see DMRB Volume 14) should be manually input to Table 15A inmarket prices, and allocated between Non-business Users and Business Users in proportion to theNon-business User and Business User (Time and VOC) benefits of the scheme under normaloperating conditions.

(iii) Revenue, Investment Costs and Grant/Subsidy are manually input to Table 15A, where appropriate.

(iv) Developer & Other Contributions are manually input to Table 15A as negative values. They are alsoinput to Table 15B - Public Accounts as negative values, see paragraph 4.24.



Notes:

(i) Revenue, Investment Costs, Operating Costs and Grant/Subsidy are manually input to Table 15A,where appropriate.

(ii) These costs are usually estimated using QUADRO (see DMRB Volume 14) and manually input intothe NESA results in market prices.

(iii) Developer & Other Contributions are manually input to Table 15A as negative values. They are alsoinput to Table 15B - Public Accounts as negative values, see paragraph 4.24.

Table 9/4/3 : Public Accounts in Market Prices (Table 15B of the NESA output)

IMPACT Reference TotalLocal Government Funding Manually input by user (see paragraph 4.22)Revenue 32 See note (i)Investment Costs 33 Including any Optimism Bias (see Part 3 Chapter 3)

See note (i)Operating Costs 34 See note (i)Maintenance Costs

Non-Traffic (Group 1) 35 See note (ii)Traffic Related (Group 2) 36 See note (ii)

Developer & Other Contributions 37 See note (iii)Grant/Subsidy 38 See note (i)Net Impact 39 32+33+34+35+36+37+38Central Government Funding: TransportRevenue 40 See note (i)Investment Costs 41 Including any Optimism Bias (see Part 3 Chapter 3)Operating Costs 42 See note (i)Maintenance Costs

Non-Traffic (Group 1) 43Traffic Related (Group 2) 44 See note (ii)

Developer & Other Contributions 45 See note (iii)Grant/Subsidy 46 See note (i)Net Impact 47 40+41+42+43+44+45+46Central Government Funding: Non-TransportIndirect Tax Revenues 48 See calculation paragraph 4.26TotalsBroad Transport Budget 49 39+47Wider Public Finances 50 48This analysis is based on <growth> traffic growthCosts are in <PV year> prices in multiples of a million pounds and are discounted to <PV year>Evaluation period <evaluation period> yearsFirst scheme year <first scheme year>Current year <current year>Discount rate 3.5% for first 30 years, thereafter 3.0% for 45 years, thereafter 2.5%



Notes:

(i) Noise, Local Air Quality, Journey Ambience and Option Values are manually input to Table 15A,where appropriate.

Table 9/4/4 : Analysis of Monetised Costs and Benefits in Market Prices (Table 15C of the NESA output)

IMPACT Reference TotalTEE Impacts

Noise 51 See note (i)Local Air Quality 52 See note (i)Greenhouse Gases (Emissions) 53Journey Ambience 54 See note (i)Accident Benefits 55Non-Business User Benefits: Commuting 56 =11Non-Business User Benefits: Other 57 =12Business Users& Provider Impacts 58 =30Wider Public Finance (Indirect Tax Revenues) 59 =-50Option Values 60 See note (i)

Present Value of Benefits (PVB) 61 51+52+53+54+55+56+57+58+59+60Broad Transport Budget 62 =49

Present Value of Costs (PVC) 63 =62Overall Impact

Net Present Value (NPV) 64 =61-63Benefit to Cost Ratio (BCR) 65 =61/63

This analysis is based on <growth> traffic growthCosts are in <PV year> prices in multiples of a million pounds and are discounted to <PV year>Evaluation period <evaluation period> yearsFirst scheme year <first scheme year>Current year <current year>Discount rate 3.5% for first 30 years, thereafter 3.0% for 45 years, thereafter 2.5%

Note: There may also be other significant costs and benefits associated with a scheme, some of which cannot be presented in monetised form. Where this is the case, the analysis presented above does NOT provide a good measure of value for money and should not be used as the sole basis for decisions.

Volume 15 Section 1 Chapter 5Part 9 Validating a NESA Assessment The Timing and Documentation for NESA Validations


5 THE TIMING AND DOCUMENTATION FOR NESA VALIDATIONS

5.1 For trunk road schemes in Scotland, NESA (or an equivalent economic evaluation technique) should beused throughout scheme preparation (see Part 2 Chapter 3). Within Transport Scotland, theresponsibility for ensuring value for money criteria have been correctly applied is met by validation ofeconomic assessments at the following stages:

(a) Stage 1 Preliminary Assessment

(b) Stage 2 Route Option Assessment

(c) Stage 3 Preferred Scheme Assessment

Validation at these stages is normally sufficient to discharge Transport Scotland’s financialresponsibility, but exceptionally further confirmation may be required at any time up to contract letstage if there are material changes which affect the economic analysis (e.g. major cost changes, majordevelopments affecting traffic flows, etc.).

5.2 Guidance on Transport Scotland’s reporting requirements is given in DMRB 5.1.2 and 5.1.4 (TD 37/93and SH 1/97). However, the specific requirements of the reporting of an economic appraisal undertakenusing NESA are detailed below in paragraph 5.4. If an alternative methodology is adopted then similardocumentation is required.

Economic Assessment Report

5.3 The purpose of the Economic Assessment Report (EAR) is to detail and justify the methodology andthe data inputs and to present the results of the economic assessment. The report should include thefollowing:

(i) a plan and description of all options under consideration in the local context. Plans ofthe Do-Nothing, Do-Minimum and all Option(s) at sufficient detail to enable link andjunction details to be checked (these should normally be at 1:2,500 scale).

(ii) a node-link diagram (showing Do-Minimum and Do-Something base year and designyear traffic flows) and a means of relating this to an Ordnance Survey map.

(iii) cost estimates for each option, broken down as detailed on sheets 1 and 2 (see Part 7Chapter 9).

(iv) A report on the NESA and QUADRO inputs. Where local values have been input toNESA, the data on which the local values are based should be supplied and explainedas necessary. In particular, this applies to the following:

(a) seasonality index

(b) vehicle category proportions

(c) car in work time proportions

(d) E, F and M factors

(e) flow group structure

(f) local traffic growth assumptions (e.g. planned developments)

(g) junction maximum delay adjustments

Chapter 5 Volume 15 Section 1The Timing and Documentation for NESA Validations Part 9 Validating a NESA Assessment


(v) evidence that NESA coded and calculated speeds are comparable for both rural andurban links.

(vi) evidence of link speed validation (e.g. journey time comparisons), particularly when ahigh proportion of the benefits accrue from urban areas.

(vii) justification for the presence of overcapacity links and junctions in the Do-Minimumand/or the Option(s) and the affect that these overcapacity links and junctions mayhave on link travel time and junction delay benefits.

(viii) rationale for the inclusion/exclusion of junctions in the economic assessment.

(ix) accident modelling methodology, including the rationale for the use of local or defaultaccident rates and costs in the Do-Minimum and/or the Option(s).

(x) a QUADRO analysis assessing delays during construction or reasons why an analysishas not been carried out.

(xi) a QUADRO analysis assessing the changes in delays during future routinemaintenance or reasons for assuming zero traffic-related maintenance benefits/dis-benefits.

(xii) Unusual features of the NESA input or output data should be explained. For example,if a large proportion of a scheme’s benefits accrue from the network becomingovercapacity in the Do-Minimum, the NESA user should comment on the realism ofthe coded network and flows and the scope for reassignment or Do-Minimumimprovements. Another example is the treatment of junctions in urban areas, wherethe modelling of junctions should be justified.

(xiii) a summary of results and conclusions including an economic summary table (seeTable 9/4/1).

(xiv) conformance with reporting standards detailed in DMRB 5.1.2 and 5.1.4 (TD 37/93and SH 1/97) (including completion of the Traffic and Economic Evaluation Reportform).

(xv) copies of full NESA printouts on computer disk for low, central and high trafficgrowth for all options under consideration.

(xvi) reports of any sensitivity tests that may be necessary and an incremental analysis oflink and junction standards if required.

Fixed Trip Matrix

5.4 If a standard NESA appraisal is undertaken, the justification of the Fixed Trip Matrix assumption mustbe documented. The Scheme Classification Report would suffice (see DMRB 5.1.4 SH 1/97).

Traffic Model Validation and Traffic Forecasting

5.5 Given the importance of the traffic flow input to NESA, the basis of this input should normally bedocumented in the NESA material. The Local Model Validation Report and Traffic Forecasting Report(see DMRB 5.1.4 SH1/97), will normally suffice.



VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


HOW TO USE THE NESA PROGRAM

Contents

Chapter1. Introduction2. Using NESA3. Building the Base Network4. Amending the Base Network5. Building the Trip Matrices6. Building the Base Network Trees7. Assignment8. Evaluating the Base Network Model9. Modelling Junctions for Delay10. The Modelling of Accidents11. Assessment of Alternative Design Schemes12. Building the Design Network13. Building the Design Network Trees14. Reassignment of Traffic15. Evaluating the Design Network16. NESA Print Command Options17. Amalgamation of Stages in NESA18. Accident only assessments19. Example Input Files and Output Tables20. NESA File Formats21. NESA Error Messages

PART 10

Volume 15 Section 1Part 10 How To Use the NESA Program


Volume 15 Section 1 Chapter 1Part 10 How To Use the NESA Program Introduction


1 INTRODUCTION1.1 NESA15 is a Microsoft Windows application and can be run directly from within Windows. Full

instructions on installing and running the program are contained within the README.TXT file whichaccompanies the software.

1.2 This part of the manual describes the data input required to run the program. Where information hasalready been documented in earlier parts of the manual, then the appropriate cross reference is givenrather than repeating the information. This reduces the risk of confusion and possible conflict.

1.3 From within NESA users can select an input file via:

File>>Select Input File

Input files usually have .DAT or .TXT filename extensions.

1.4 NESA will attempt to run the selected input file and will automatically create an output file of the samename but with a .LIS extension.

1.5 The remaining chapters of this part of the manual may be grouped as follows, and are referred to asappropriate in Figure 10/1/1:

• Chapter 2 describes how to operate NESA

• Chapters 3 to 7 describe the traffic model

• Chapters 8 to 10 describe the evaluation of the base model

• Chapter 11 describes the assessment of alternative design schemes

• Chapters 12 to 15 describe the evaluation of the design model

• Chapter 16 describes the NESA print commands

• Chapter 17 describes how stages can be combined in NESA

• Chapter 18 describes the accident only assessment

• Chapter 19 gives examples of input files and output tables

• Chapters 20 and 21 describe the NESA output file formats and give details of the error messages given by the program.

Chapter 1 Volume 15 Section 1Introduction Part 10 How To Use the NESA Program


Figure 10/1/1: The NESA Procedure Summarised



NESA Command File Hierarchical Structure

1.6 The following describes the NESA commands in their hierarchical structure.

(i) All top level commands are optional, except for the first (i.e. directory command) andthe last (i.e. END JOB). Otherwise, optional commands within a level are indicatedby square brackets. Sub-commands of an optional command are only squarebracketed if they in turn are optional within the previous level command.

(ii) UPPER CASE words indicate commands, though they may be input as lower casebecause NESA automatically converts all lower case characters to upper case. Lowercase characters indicate parameters etc.

(iii) "….." indicates repetition as necessary.

(iv) The maximum number of characters that can be entered on a single line is 255. (cont)indicates that the command may be continued on the following line(s) at some breakpoint, normally a comma. These commands need an END to terminate the command.Some other commands (e.g. YEAR=) do not need an END but will not cause a fatalerror if one is included.

(v) (base) indicates the command is only valid in a base run.

(vi) (design) indicates the command is only valid in a design run.

(vii) " / " indicates an either/or.

(viii) All commands should be input in the order shown, except PRINT commands.

1.7 The format is as follows. The indented layout is for illustrative purposes only: NEW_BASE=dir / OLD_BASE=dir[NEW_DESIGN=dir / OLD_DESIGN=dir]BUILD NETWORK (base) BASE NETWORK=”title” year [ind] ZONES=zone,…..,END(cont) [TABLE=headings] network data ….. END UPDATE NETWORK BASE NETWORK=”title” year[ind] / DESIGN NETWORK=”title” year [DELETE] node node ….. END [ADD] [TABLE=headings] network data ….. END [CHANGE] [TABLE=headings] network data …..



END BUILD MATRICES(base) [MATRIX_OPTIONS,parameter=value,…..,END](cont) [DEFINE FLOW GROUPS] FGn=definition ….. MATRIX [fg] =uc,….. FROM origin TO destination, trips ….. END(cont) ….. END MATRIX ….. [BUILD FUTURE YEAR MATRIX year] BUILD TREES [TREE_OPTIONS,parameter=value,…..,END](cont)(base) [ZONE SELECTIONS=zone,.....,END] [TURNING DELAYS=] node,node,node,minutes ….. END SUPPRESS TABLES GROWTH RATES YEAR headings data ….. END ASSIGN EVALUATECURRENT_YEAR=year BASE_COSTS=parameters(base) SCHEME_COSTS,parameter=values,…..,END(cont)(design) PROFILE=year,percentage …..(design) [JUNCTIONS=node,…..,END](cont) [MAXDEL=seconds] [PRIORITY=anode,bnode,cnode,parameter=value,…..] WIDTHS,parameter=value,..... VISIBILITY,parameter=value,..... ….. [ROUNDABOUT=node,ENTRIES=node,…..,DIAM=metres] ENTRY=node,parameter=value / EXIT=node ….. ….. [SIGNALS=node,ENTRIES=node,…..,parameter=value,…..] ENTRY=node,LANES=value LEFT/STRAIGHT/RIGHT=node,…..,parameter=value,..... [FROM=node,TO=node] ….. ….. ….. STAGE=value[ ,FIXED=seconds,MIN=seconds] ….. ….. [MERGE=node,node,node,SLIP=node] ….. [ACCIDENT_COST_FACTORS=rc, value ..... END](base)



…. [ACCIDENT_YEAR=year](base)[ACCIDENT_RATES=rc, value .....[YEAR=year] (base) END](base) ….. [ACCIDENT_LINK=node,node[,RATE=value/NACC=number][,COST_FACTOR=factor]] ..... [ACCIDENT_JUNCTION=node,TYPE=value[,NACC=value][,COST_FACTOR=factor][,CLASS=MAJOR/MINOR]] ….. [ACCIDENT_DEFAULT=LINK_ONLY]END [STOP AFTER DELAY CALCULATIONS] [STOP AFTER FIRST YEAR] PRINT NETWORK PRINT MATRIX [ZONE SELECTIONS=zone,…..,END](cont) PRINT TREES [ZONE SELECTIONS=zone,…..,END](cont) PRINT ASSIGNMENT [YEAR=year,…..] PRINT EVALUATION[TABLES] [TABLES=table,…..] [FUTURE_BASE=directory year] PRINT HFGM PRINT JOURNEY [FLOW_GROUPS=fg,…..] [YEAR=year,…..] node,…..,END(cont) PRINT JUNC_DELAY [JUNCTION_SELECTIONS=node,…..,END](cont) [YEAR=year,…..] PRINT LINK_FLOWS [YEAR=year,…..] node,node,dir ….. ENDPRINT LINK_FLOWS_BY_UC[YEAR=year,…..] node,node,dir ….. ENDPRINT OVER_CAPACITY [YEAR=year,…..] PRINT PATH zone,zone ..... END PRINT SPEEDS [DIFFERENCE [KPH=speed] [PC=percentage]] [YEAR=year,.....]PRINT TURNING [MOVEMENTS] [JUNCTION_SELECTIONS=node,…..,END](cont) [YEAR=year,…..] END JOB



Data Errors

1.8 Three types of error can be produced by NESA:

• Warning

• Fatal

• Stop

and are summarised at the end of the output. These errors are described fully in Part 10, Chapter 21.

1.9 Free format data input is very difficult to error trap, so some program terminations finish with verygeneral errors.

1.10 Fatal errors are sometimes caused by earlier data input errors, which NESA may have only output as awarning. The general advice to solve a fatal error is to go to the first warning in the input file and workforwards to the fatal error.

Examples

1.11 Figures 10/1/2 & 10/1/3 show examples of simple NESA base and design networks respectively.Throughout this part of the manual the worked examples, wherever possible, relate to these networks.Both networks have 5 zones. The base network contains 11 nodes, a priority and a signalisedintersection. The design network contains 13 nodes, with a roundabout, staggered priority intersection,a priority intersection with ghost island and a signalised intersection.

Figure 10/1/2: Example Base Network



Figure 10/1/3: Example Design Network

Volume 15 Section 1 Chapter 2Part 10 How To Use the NESA Program Using NESA


2 USING NESASystem Requirements

2.1 NESA is a Microsoft Windows application. It has been tested on

• Windows 7

• Windows 8.1

2.2 NESA will handle network sizes of up to 2000 directional links, 800 nodes and 120 zones.

Installing and Starting NESA

2.3 NESA is normally released on a CD which will contain some or all of:

• nesann.exe NESA application

• nesa.ico NESA icon

• profile.dat essential day and year profile data

• readme.txt a readme file

• base.dat example base job deck

• design.dat example design job deck

• base.nodes node coordinate file for base.dat

• design.nodes node coordinate file for design.dat

• multiple.txt example multiple job deck input file

2.4 Copy nesann.exe and nesa.ico from the release CD to a chosen program location.

2.5 As with other Windows applications users may wish to create a shortcut on the desktop. Having createda shortcut, right click to select its properties, and use the Change Icon button to select nesa.ico. TheNESA shortcut will then be displayed with the new NESA icon.

2.6 Copy profile.data from the release CD to a chosen location and create a system environment variablecalled NESA_PROFILE, identifying the directory containing the file, e.g.

c:\nesa\profile\

Changes to system environment variables do not take effect until after the next reboot.

2.7 Double click on the file or the NESA icon to run NESAnn.

RUNNING NESA

2.8 NESA can be used to run one or more input files (job decks), and to interrogate the resulting models ina variety of ways, using a menu-driven Windows user interface. It includes a graphical displaycapability which allows users to plot their networks, providing a suitable co-ordinates/nodes file isavailable.

Chapter 2 Volume 15 Section 1Using NESA Part 10 How To Use the NESA Program


Single Input Files

2.9 Select the menu option File>>Select input file to select a job deck, using the standard Windows filespecification dialog. This is equivalent to the old way of running NESA from the DOS command line.

2.10 The output will be written to a file with the same name as the input file but with a .lis extension, in thesame directory as the input file. This output file can be viewed with Wordpad or any other filer viewer/editor.

2.11 NESA data files are written to subdirectories of NESADATA, and a run will fail if the subdirectoryalready exists. Use Windows Explorer to delete any subdirectories which already exist, and then selectthe input file again.

Multiple Input Files

2.12 It is no longer possible to run DOS batch files, but this functionality is replaced by the menu optionFile>>Select multiple input file. This option expects a text file containing a list of job decks, eachwith the FULL file specification, e.g.

c:\nesadecks\base.dat

\\walrus\tpnesa\test decks\design.dat

2.13 This option runs all the specified job decks one after the other, within the current NESA window. It willnot terminate if there is an error in one of the job decks. It is not possible to interrupt the process.

Selecting Networks

2.14 As well as running job decks in NESA it is possible to select networks which already exist, for exampleto view the NESA tables on screen and/or print them.

BASE NETWORK

2.15 The menu option File>>Select base network enables users to select a NESA base network directory.This should be a directory 1 level below NESADATA, e.g.

c:\nesadata\base

2.16 Selecting a base network clears any previous design network setting.

DESIGN NETWORK

2.17 The menu option File>>Select design network enables users to select a NESA design networkdirectory. This should be a directory 2 levels below NESADATA, e.g.

c:\nesadata\base\design

2.18 Selecting a design network automatically also selects the associated base network.

Loading Node Coordinates

2.19 Loading a file of node coordinates extends the functionality of NESA to:

• annotate/interrogate the networks

• compare base and design networks

Volume 15 Section 1 Chapter 2Part 10 How To Use the NESA Program Using NESA


• capture images to include in reports

2.20 It is essential that these node files contain rows for all nodes in each network.

2.21 Users are strongly recommended to create these files by taking a copy of the netnames.and file for thenetwork and edit these. Do not add or delete any rows. The files are basic comma separated text fileswith entries as follows:

node or zone number, easting, northing, comment or text as required, e.g.

Z3, 3688, 6777,A1 to Edinburgh

4, 3687, 6774,

5, 4000, 6800,

2.22 The last line should be blank and there is NO requirement for END at the end of the file. OSGRs can beI4 or I6.

2.23 Node files, for each base and design network, should be located in the same folder as the base anddesign NESA input files.

2.24 When a node coordinate file has been successfully loaded users can view and annotate the network.

• Base networks are displayed in green and design networks in red.

• Zones are drawn with square symbols, and nodes with circular symbols.

• Roundabouts are drawn with circular symbols with a ring round them.

• One-way links are drawn with a single line, and two-way with double lines.

• Zone centroid connectors are drawn with dashed lines.

BASE COORDINATES

2.25 The menu option File>>Select base coordinates enables users to select a file containing coordinatesfor all the nodes in the base network.

DESIGN COORDINATES

2.26 The menu option File>>Select design coordinates enables users to select a file containing coordinatesfor all the nodes in the design network.

NESA Output

2.27 Users can request a significant amount of information from NESA using the PRINT commands (seePart 10 Chapter 16). To keep the output to a manageable size users are recommended to carefully selectthe required information.

2.28 Incorporation of the following commands (including sub-commands) can help reduce output:

SUPPRESS TABLES

PRINT EVALUATION TABLES

TABLES=2,3,4,5,11,12,13

Chapter 2 Volume 15 Section 1Using NESA Part 10 How To Use the NESA Program


Data Format

2.29 Most NESA data and commands are free format. Where column numbers are given in the sections thatfollow, these are advisory, not mandatory.

2.30 In conventional use, NESA data input files must be in ASCII format. If a file is being imported fromother transportation software for use with NESA, e.g. a matrix file, then the file will need to complywith the appropriate NESA file format. A full description of these file formats is given in Part 10,Chapter 20.

2.31 Comment lines can be inserted in the data file. These must have a ! in the first column.For example:-

ZONES=Z01,Z02,Z03,END

! ZONE 3 REPRESENTS A999 EAST

Defining NESA Models

2.32 Each NESA model is identified by a unique name, which identifies the directory in which NESA storesits working files. The first line of any NESA file must define the names of the base model and/or thedesign model as appropriate. Examples are:-

(i) new_base=exist99 (to build a new base model)

(ii) old_base=exist99 (to continue work on an existing base model)

(iii) old_base=exist99 new_design=a999 (to start a new design model)

(iv) old_base=exist99 old_design=a999 (to continue work on an existingdesign model)

2.33 Note that blanks are not permitted in model names, which should not exceed 10 characters in length.Model names become 3rd and 4th level directory names within the user’s disk allocation, and the aboveexamples would create directories called:

[.NESADATA.EXIST99] as a base model and[.NESADATA.EXIST99.A999] as a design model

2.34 Under normal circumstances users will not need to access the files within these directories, which havepredetermined names and are created and accessed by NESA. It is not advisable to place other files(e.g. the NESA input data) within these directories.

Reporting

2.35 A record of the traffic and economic assessment containing all the relevant data should be reported toTransport Scotland.

NESA Support

2.36 NESA is supported by Transport Scotland. If users have any problems with NESA, or require anyfurther information, please email [email protected].

2.37 If users have a problem running NESA, please email [email protected], attachingthe relevant job deck where appropriate.

Volume 15 Section 1 Chapter 3Part 10 How To Use the NESA Program Building the Base Network


3 BUILDING THE BASE NETWORK3.1 This is the first stage in any NESA evaluation. Once the base model is defined, as described in Part 10,

Chapter 2 then the base network can be built. The data sequence required to build a base network isshown in paragraph 1.6 and further described in the paragraphs that follow.

BUILD NETWORK command

This must be the second line in the data file (excluding any comment lines), immediately after the filedefinitions. The format is:

Network Title

3.2 This line of data defines the network title, the base network year and the traffic growth to be used. Thefirst 13 columns are fixed format as follows:

NOTES

(i) The title must be contained within double quotes and may consist of anyalphanumeric characters.

(ii) Where NRTF(97) default traffic growth is being used, the traffic growth indicator canbe either H to invoke High traffic growth, C to invoke Central traffic growth, or L toinvoke Low traffic growth. If omitted the program will default to Central trafficgrowth. The traffic growth indicator is only required in the Base network. There mustbe a space between the base network year and the H, C or L parameter. Since theintroduction of NESA02 the program no longer includes low and high economicgrowth. A single economic growth is now used no matter what traffic growth is used.

(iii) Each item of data should be separated by one or more spaces.

For example:

BASE NETWORK="ANY ROAD SCHEME EXAMPLE NETWORK FOR USER MANUAL" 2014 C

ZONES= command

3.3 The ZONES= command defines the zone names to be used in the NESA model. Each zone name canconsist of up to 8 alphanumeric characters and may be specified in any order. The command has theformat:

Columns 1 - 5 BUILD7 - 13 NETWORK

Columns 1- 4 BASE6 - 13 NETWORK=14 - 93 “network title” base network year traffic growth indicator (if required)

Columns 1 - 6 ZONES=7 - 255 the list of zone names, each separated by a comma

Chapter 3 Volume 15 Section 1Building the Base Network Part 10 How To Use the NESA Program


NOTES

(i) The names can be continued onto additional lines, which do not require the ZONES=command.

(ii) The final zone in the list must be followed by END to close the list.

For example:

ZONES=Z01,Z02,Z03,Z04,Z05,END

Link Data Description

3.4 The detailed description of the highway network is defined using link data. One line of data is used foreach link, in which variables are used to define various physical and geometric attributes.

3.5 These variables can be input to the program in either fixed or free format at the discretion of the user,and this is described in paragraphs 3.10 to 3.15. Irrespective of which method is adopted, the datarequirements are identical.

3.6 The variables are split into two types, mandatory and optional. As the name suggests, mandatoryvariables must appear for each link. If any are omitted then the link will not be accepted by the programand an error will result. The omission of an optional parameter, however, will cause the program toadopt a default value.

3.7 Link variables are also classified as assignment variables or evaluation variables.

(i) Assignment variables are used by NESA to derive the fixed link speeds and linklengths required in the assignment module. Some of these variables may also be usedin the economic analysis, but not necessarily so.

(ii) Evaluation variables are used only in the economic evaluation module. Generallyspeaking these are geometric variables and are used to refine the speed/flowrelationship for each individual link. This is discussed in more detail in Part 7,Chapter 1.

3.8 Tables 10/3/1 and 10/3/2 define the mandatory and optional link variables respectively. The defaultvalue for each of the link variables is given in Table 7/1/4.



Table 10/3/1: Mandatory Link Variables

Variable Title Units Type Description/CommentA-nodeB-node

Link Definition N/A Assignment These are always the first two values on the link data line, irrespective of the method of data entry, and have no associated parameter code. The initial choice of A and B-node is immaterial but once chosen it is important that the adopted convention is observed for the coding of other directional variables. The order in which they are keyed is largely unimportant, except in the case of one-way links or junction indicators.

D Link Length km Assignment Link distance between nodes.

R Road Category N/A Assignment Refer to Part 7 Chapter 1 Table 7/1/2.

SL Speed Limit kph Assignment In kilometres per hour. The speed limit specifies the default light and heavy vehicle speeds used by NESA in the assignment process. Refer to Part 5 Chapter 3 Table 5/3/1.Code 30mph as 48kph, 40mph as 64kph, 50mph as 80kph, 60mph as 96kph, 70mph as 113kph.

Table 10/3/2: Optional Link Variables

Variable Title Units Type Description/CommentJ Junction Index No. Assignment This parameter indicates the number of junctions encountered along

a link where a delay may be expected due to the nature of the junction i.e. roundabout, traffic signals or the non priority movement at a priority junction.

The junction at the start node of the link is not included, therefore the index may have different values in each direction. In this case, the junction index should be specified twice, the second value referring to the B to A-node direction. Refer to Part 5, Chapter 3.

O One Way Flag N/A Assignment This flag indicates a one-way link. The program assumes that the direction of travel on the link is the A to B-node direction. Not required if the road category implies a one-way link.

SLC Coded Speed (Lights)

kph Assignment User defined coded speed for light vehicles. This variable can be used to overwrite the default coded speed invoked by the SL variable (see Table 5/3/1).

SHC Coded Speed (Heavies)

kph Assignment User defined coded speed for heavy vehicles. This variable can be used to overwrite the default coded speeds invoked by the SL variable (see Table 5/3/1).

X External Link Flag

N/A Evaluation This flag causes the link to be included in the tree building and assignment stages but excluded from the economic evaluation.

C Central Flag N/A Assignment Denotes that the default coded speeds for Central urban links are to be invoked (for urban links only) and also determines the speed/flow curve to be used (see Part 7, Chapter 4).

ST Small Town Flag N/A Evaluation Denotes that the Small Town speed/flow relationship is to be invoked (urban links only).

DES TD/9 Flag N/A Evaluation Indicates whether link is designed to TD/9 standards.



Free Format Input

3.9 For free format data input the following rules apply

(i) The first two items on the line must be the A and B nodes, otherwise variables can beinput in any order.

(ii) Each value input for the link must be preceded by the relevant parameter code.

(iii) Each value must be separated by at least one space.

The final line of link data must be followed by END.

For the example network, a typical free format link data entry may be:-

10 20 D1.5 SL96 R26 HR10 HF10 B10

Table 10/3/3: [Contd] Optional Link Variables

Variable Title Units Type Description/CommentB Bendiness deg/km Evaluation The requirement to specify some or all of the remaining variables in

the table is dependent upon the speed/flow type invoked by the combination of road category and speed limit. The definitions of these variables and their use for different speed/flow types are detailed in Part 7 Chapter 1 paragraphs 1.12 to 1.21 and Tables 7/1/2 and 7/1/3.

HR Hilliness Rise m/km Evaluation

HF Hilliness Fall m/km Evaluation

CWID Carriageway Width

m Evaluation

SWID Strip Width m Evaluation

VISI Sight Distance m

JUNC Side Road Intersections

No./km Evaluation

INT Major Intersections

No./km Evaluation

AXS Minor Intersections

No./km Evaluation

VW Verge Width m Evaluation

DEVEL Frontage Development

% Evaluation

P30 % of route subject to 30mph

speed limit

% Evaluation



Input Using the TABLE= Option

3.10 This option allows link data to be entered in tabular form. The layout of the table is defined using theTABLE= option which identifies the column positions of the variables on the subsequent lines of linkdata. The format of the option is:

3.11 Rules governing the structure of the TABLE= option are as follows

(i) Data must be right justified to the parameter codes listed on TABLE= statement.

(ii) Parameter code positioning must allow spaces between data fields in the link data.

The lines of link data can now follow, with parameter values input without any codes.

3.12 The free format example used above would look like:-

TABLE= D SL R HR HF B10 20 1.5 96 26 10 10 10

As with free format data, the last line of link data must be followed by END.

3.13 If required, the user can define more than one table with varying formats, provided that the relevantlink data follows the corresponding TABLE= option. However all the tables must be input before theEND command.

3.14 Link data can also be entered in both formats, i.e., with and without the TABLE= option. However thefree format data must precede the data in fixed format. The data formats cannot be mixed on a line.

PRINT NETWORK command

3.15 If the user wishes to obtain confirmation of the network description, then the PRINT NETWORKcommand should follow the END command at the end of the link data. The format is:

END JOB command

3.16 No further data is required for building the network. The user now has the option to terminate the runor continue onto the next stage of the evaluation. It is usual practice with a new network to stop at thisstage and check the network integrity before continuing. The command to stop a NESA run is the ENDJOB command which has the format:

Columns 1 - 6 TABLE=8 - 255 parameter codes in any order

Columns 1 - 5 PRINT7 - 13 NETWORK

Columns 1 - 3 END5 - 7 JOB



Special Requirements Relating to Junction Modelling

3.17 Certain rules appropriate to the user defined junctions defined at the evaluation stage (see Part10,Chapter 8) can affect the link descriptions at the network building stage. The cases to note areexplained below:

(i) Zone connectors (see Part 5, Chapter 3) cannot be linked directly to user definedjunctions. If a zone connector can accurately represent the traffic flow on an arm of ajunction, this restriction can be overcome by inserting a dummy link between thejunction and the centroid connector.

(ii) At roundabouts where U-turns are permitted the roundabout node should be prefixedwith the letter R (e.g. R1001) in the network description.

(iii) If pelican crossings are to be represented as signalised junctions, dummy links wouldbe required to meet the programming constraints relating to the number of phases andturning movements (as described in Part10, Chapter 9).

(iv) Geometric delays at roundabouts are calculated using the coded link speeds on theconnecting links. Care should be exercised in ensuring that coded and calculatedspeeds do not vary significantly or else erroneous delays will be obtained (see Part 8,Chapter 5).

Example:

The following example builds the base network shown in Figure 10/1/2, using the TABLE= option

NEW_BASE=BASE

BUILD NETWORK

BASE NETWORK="ANY ROAD SCHEME EXAMPLE NETWORK FOR USER MANUAL " 2014 C


! ZONE 1 IS FROM EDINBURGH

TABLE= D R SL ST B HR HF SWID JUNC VW DEVEL P30

Z01 10 0.10 50 96

Z02 80 0.10 50 96

Z03 60 0.10 50 96

Z04 100 0.10 50 96

Z05 110 0.10 50 96

10 20 1.00 26 96 20 6 0 0.5 2 2.0

20 30 2.00 26 96 47 12 8 0.5 5 2.0

20 110 1.00 26 96 83 5 6 0.5 3 2.0

30 40 1.50 1 48 ST 90 100

40 50 1.50 1 48 ST 90 100

40 70 1.00 1 48 ST 90 100

40 90 1.00 1 48 ST 90 100

50 60 2.00 26 96 40 5 9 0.5 3 2.0

70 80 1.00 26 96 27 0 2 0.5 3 2.0

90 100 1.50 26 96 10 8 2 0.5 1 2.0

END

Volume 15 Section 1 Chapter 4Part 10 How To Use the NESA Program Amending the Base Network


4 AMENDING THE BASE NETWORK4.1 Once the base network has been created the user may discover errors and/or omissions. If a large

number of mistakes are found, it is suggested that the network is removed and re-created using BUILDNETWORK. If zone names need to be changed then BUILD NETWORK must be repeated. Formodest changes to the network, however, the update facility in NESA should be used. The inputsequence is detailed in Paragraph 1.6 and is described in detail in this chapter.

UPDATE NETWORK command

4.2 As for BUILD NETWORK, this line must immediately follow the file definitions, which are describedin paragraph 2.6. The format is:

Network Title

4.3 The format of this line is the same as for building the base network, described in paragraph 3.3. Ifdesired, the title can be the same as the BUILD NETWORK title. However it is recommended that thetitle is annotated in some way to indicate that the network has been amended.

If the only change to the network is to amend the title, then the title line must be followed by END.

Network Alteration Commands

4.4 There are three network alteration commands; DELETE, ADD and CHANGE. All data relating to agiven command must immediately follow the command. While the three commands can be entered inany order it is recommended that the sequence DELETE, ADD, CHANGE be used to avoid errors inrestructuring the network. A command can be omitted if it is not required. The three commands aredescribed below.

DELETE command

4.5 To remove links from the network, the DELETE command must be used to introduce the link data. Theformat is:

The lines immediately following this command must contain the link deletion data. Links to be deletedare defined simply by specifying the A and B nodes. One link should be specified per line. The finalline of link data must be followed by a standard END command to close the list.

4.6 A two way link cannot be deleted in one direction only.

ADD command

4.7 This command is used to introduce wholly new links to the network. The format is:

The lines immediately following this command must contain link addition data. Links to be added aredefined in exactly the same manner as for building the network. The data may be entered in fixed or

Columns 1 - 6 UPDATE8 - 14 NETWORK

Columns 1 - 6 DELETE

Columns 1 - 3 ADD

Chapter 4 Volume 15 Section 1Amending the Base Network Part 10 How To Use the NESA Program


free format as described in Part 10, Chapter 3. As with the DELETE command, the final line of linkdata must be followed by the standard END command.

CHANGE command

4.8 In some cases it is only necessary to change one or two variables on a link. For example, the road classor bendiness may alter. Except for the special cases noted below, it is not necessary to delete the linkand add a new one, the CHANGE command can be used instead. Links to be changed are introducedby the CHANGE command which has the following format:

The links to be changed are input immediately after this command. As with the ADD command, theformat of the link data is exactly the same as for building the network. However, only those variablesthat are being altered need to be specified. Any variables on a link which are not specified will remainat their original values. Again, the TABLE = option can be used to precede the link data if desired, andthe last line of link data must be followed by the standard END command.

4.9 The CHANGE command should not be used to convert a one-way link to two way, or vice versa.Instead the links should be deleted and then added as new links with the appropriate variables altered.


4.10 This command allows the user to print the updated network if desired. The format is the same as thatdescribed in the Part 10, Chapter 3:

END JOB command

4.11 No further commands are required to update the base network. The user now has the choice ofcontinuing onto the next stage of the evaluation or terminating the run. If the latter course is taken thenthe standard END JOB command is required with the format:

Example:

The user is referred to Part 10, Chapter 12, which gives an example of updating a base network, tocreate a design network. The principles that apply to amending a base network are all embodied in thatexample.

Columns 1 - 6 CHANGE


Columns 1 - 3 END5 - 7 JOB

Volume 15 Section 1 Chapter 5Part 10 How To Use the NESA Program Building the Trip Matrices


5 BUILDING THE TRIP MATRICES5.1 Once the user is satisfied that the base network has been correctly built then the assessment can

progress to the next stage. This stage can be either to build the matrices or build the assignment trees.The only stipulation is that both stages must be undertaken before the evaluation, described in Part 10,Chapter 8.

5.2 For the purposes of this manual, the building of the trip matrices is described first, with building of theassignment trees following in Part 10, Chapter 6. The data required to build the trip matrices is shownin Paragraph 1.6

BUILD MATRICES command

5.3 BUILD MATRICES is the first line required for the matrix building stage. Its format is:

Matrix Data Input

5.4 NESA recognises 15 different user classes and matrix data can be submitted in any combination ofthese. The program splits any chosen aggregation using annual and daily profiles by user class (see Part5, Chapter 2). Definitions of the fifteen user classes are given in Part 5 Table 5/2/5.

5.5 Matrix data may also be input in the form of flow group matrices, a feature which allows the user toinput peak period matrices. The time periods should be defined using the DEFINE FLOW GROUPScommand.

5.6 The default network classifications, E Factors and M Factors in NESA imply a default time-frame ofAnnual Average Weekday Traffic (AAWDT), based upon the average 12-hour weekday flow for theaverage working week. If an input trip matrix is not submitted in this form then the appropriate MFACand EFAC parameters must be supplied by the user with the MATRIX_OPTIONS command.Alternatively a factor (FFAC) may be input to convert the matrix to 12hr AAWDT to enable EFAC andMFAC factors to be applied.

MATRIX_OPTIONS command

5.7 The MATRIX_OPTIONS command is an optional command used to define, where appropriate, theFFAC, EFAC, MFAC, traffic composition, network classification and seasonality index. Theparameters and their abbreviations are given in Table 10/5/1.

5.8 Most of the parameters have default values if they are omitted, which are also listed in the table. Theuser should overwrite the default values with local data whenever possible. For trunk road appraisals,these local values must be approved by the Chief Road Engineer, Transport Scotland and evidenceshould be included in the economic submission in support of any local values used.

Columns 1 - 5 BUILD7 - 14 MATRICES

Chapter 5 Volume 15 Section 1Building the Trip Matrices Part 10 How To Use the NESA Program


5.9 The format of the MATRIX_OPTIONS command is:

The data can be continued on additional lines of the format

5.10 Note that the final parameter must be followed by an END command.

Parameters marked thus * may also be input by individual flow group, rather than by all day, by addinga numerical suffix to the respective parameter which represents the appropriate flow group.

e.g. CWK3 The percentage of cars travelling in work mode in flow group 3.

Columns 1 - 15 MATRIX_OPTIONS,16 - parameter=value, parameter=value, etc.

Columns 1 - parameter=value, parameter=value, etc.

Table 10/5/1: Base Network MATRIX_OPTIONS Parameters

Parameter Description Range DefaultNETCLASS Network Classification. See Part 5, Table 5/2/1 1-8 8 (All Roads)

MONTH Month of traffic data (if MFAC not supplied) 1-13 None

S_I User defined Seasonality Index. See Part 5, Table 5/2/2 N/A Varies by NETCLASS

MFAC User defined M Factor. See Part 5, Table 5/2/4 N/A Varies by NETCLASS

EFAC User defined E Factor. See Part 5, Table 5/2/3 N/A Varies by NETCLASS

FFAC Factor to convert non-standard matrices (e.g. 8hr) to 12hr

N/A 1.0

CAR* Percentage of cars in 24hr Annual Average Day.See Part 5 Table 5/2/7

0-100 Varies by NETCLASS

LGV* Percentage of LGVs 0-100 Varies by NETCLASS

OGV1* Percentage of OGV1s 0-100 Varies by NETCLASS

OGV2* Percentage of OGV2s 0-100 Varies by NETCLASS

PSV* Percentage of PSVs 0-100 Varies by NETCLASS

CWK* Percentage of cars in work mode in 24hr Annual Average Day. See Part 5, Paragraph 2.27


LGVNWK Percentage of LGVs in non-work mode in 24hr Annual Average Day




Flow Group Data

5.11 By default NESA assumes the use of four flow groups:-

(i) Flow group 4 The busiest 500 hours of the year

(ii) Flow group 3 The next busiest 760 hours

(iii) Flow group 2 The next busiest 2500 hours

(iv) Flow group 1 The remaining 5000 hours

These flow group definitions may be overridden by user defined flow groups (see Part 5, paragraph2.31)

DEFINE FLOW GROUPS command

5.12 This command invokes the input of user-defined flow group data. This feature can be used to model thetidality of traffic i.e. AM and PM peak periods. The modelling of peak periods should require thedefinition of at least four flow groups.

5.13 The following lines then define each flow group in descending order (i.e. flow group 1 is defined last):the maximum number of flow groups permitted is 6.

then either

Range values are in one of the following forms:

Columns 1 - 18 DEFINE FLOW GROUPS

Columns 1 - 4 FGn=where n= the flow group number

Columns 5 - HBrange/DTrange/MNrange

where HB determines hour beginning (0-23)DT determines day type (1-4) 1=Mon-Thurs

2=Friday3=Saturday4=Sunday

MN determines month (1-13) (month 8 is the holiday month representing mid July to mid August)

i n1 i.e. single numberii n1 - n2 all values between n1 and n2 inclusiveiii n1 + n2 values n1 and n2 (which may be extended indefinitely, e.g.

n1+n2+n3+n4)

or ivColumns 5 - 8 NEXT

9 - n implying the next busiest n hours



5.14 The final command in this stream MUST be:

Example

5.15 The following example defines 2-hour weekday AM / PM periods, the Between Peak period and flowgroup 1 the remaining hours:

DEFINE FLOW GROUPSFG4=HB7+8/DT1+2/MN1-13FG3=HB16+17/DT1+2/MN1-13FG2=HB9+10+11+12+13+14+15+18/DT1+2/MN1-13FG1=REST

MATRIX = command

5.16 Data for each trip matrix submitted must be preceded by a header indicating the user class orcombination of classes. Note that all fifteen user classes must be accounted for and must follow theMATRIX= command. The format is:

For example, permissible matrix data headers would be:

MATRIX=1,2MATRIX=7,8or MATRIX=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15

5.17 When inputting flow group type matrices the format is as follows:

Using the flow groups defined in paragraph 5.15 as an example, the AM peak matrix would be enteredas MATRIX4, the PM peak matrix as MATRIX3, and the adjacent to peak matrix as MATRIX2.MATRIX1 need not be input manually as the program would calculate this using the FFAC, EFAC, andMFAC parameters. As before all fifteen user classes must be accounted for.

5.18 Vehicle type matrices may be entered using the appropriate user classes.

Trip Data

5.19 The trip data for the matrix is entered on a row by row basis for each origin zone. The data for eachzone may be spread over several lines, but the first line for each zone must be of the format:

Columns 1 - 8 FG1=REST implying all remaining hours not already defined.

Columns 1 - 7 MATRIX =8 - user class number, user class number, etc.

Columns 1 - 8 MATRIXn=9 - user class number, user class number, etc.

where n is the flow group number.

Columns 1 - 4 FROM6 - 255 origin zone name

destination zone name, number of tripsdestination zone name, number of tripsetc.......

where is one or more spaces



5.20 The zone data can be continued on subsequent lines and these simply have the format:

The final destination zone value should be followed by END. For example, a zone named Z01 has tripsto Z02, Z03, Z04 and Z05. The format of the input data could be:-

FROM Z01 TO Z02,500 Z03,240 Z04,110 Z05,3020 END

NOTES

(i) If a zone to zone movement is zero, it need not be specified. Any zone to zonemovements which are not specified are assumed to be zero.

(ii) If a matrix origin zone is omitted, the program assumes that there are no trips fromthat zone.

END MATRIX command

5.21 The END MATRIX command must follow the last line of trip data for each matrix being input. Theformat is:

PRINT MATRIX command

5.22 A printout of the matrix may be produced using the PRINT MATRIX command. As with the network,the matrices can be printed at any stage after they are built. The first line of data must be the PRINTMATRIX command and has the format:

5.23 If the user wishes to print trips from all zones then no further instructions are required. However, theuser may print zones selectively by using the following command.

ZONE SELECTIONS = sub command

5.24 This command allows the user to define individual zones for which trip data is to be printed. It has theformat:

5.25 The zone list may be continued on subsequent lines. The last specified zone must be followed by END.

For example, to select zones Z01, Z03 & Z04 for printing:

ZONE SELECTIONS=Z01,Z03,Z04,END

Columns 1 - 255 destination zone name, number of tripsdestination zone name, number of tripsetc.....


Columns 1 - 3 END5 - 10 MATRIX

Columns 1 - 5 PRINT7 - 12 MATRIX

Columns 1 - 4 ZONE6 - 16 SELECTIONS=17 - 255 zone names required, separated by commas



GROWTH RATES command

5.26 This command permits the user to input local growth rates The command must appear beforeEVALUATE, the format is:

The user input years and growth rates must be defined right justified on subsequent lines, introducedusing the following header:

END command

5.27 After all the local growth rates have been input, a standard END command must be input on a separateline. For example:

GROWTH RATESYEAR CARS LGV OGV1 OGV2 PSV2008 3.207 3.125 1.536 3.870 0.0002009 2.848 3.125 1.536 3.740 0.0002010 2.382 3.125 1.536 3.650 0.0002011 2.068 3.125 1.536 3.570 0.000END

Future Year Matrices

5.28 NESA forecasts future year traffic by default using forecasts contained in NRTF (Great Britain) 1997(Department for Transport, 1997). Future year matrices for unspecified years are calculated byapplying NRTF to the previous specified year.

5.29 The purpose of future year matrices is to define the changes that result from changes in land use, e.g.the impact of a proposed development in a particular zone. Future year matrices should not be used tooverride the default growth - the GROWTH RATES command serves that purpose.

BUILD FUTURE YEAR MATRIX command

5.30 Each matrix is introduced by the BUILD FUTURE YEAR MATRIX command, the format of which is:

where future_year is the year the future year network becomes effective.

5.31 Following this, the MATRIX= or MATRIXn= (flow group matrices) command is used to define therelevant user classes using the same format as the base year matrix(ces).

NOTES:

(i) The input data for the matrix follows the same format as for the base year matrix.

(ii) Each input matrix must be terminated using the END MATRIX command.

Columns 1 - 6 GROWTH8 - 12 RATES

Columns 1 - YEAR CARS LGV OGV1OGV2PSVwhere is one or more spaces

Columns 1 - 24 BUILD FUTURE YEAR MATRIX26 - 29 future_year



(iii) Up to five future year matrices can be input.

The commands for building a future year matrix should be positioned before the EVALUATEcommand.

5.32 When using future year matrices the following points should be observed:

(i) NESA calculates traffic growth in the standard manner, using NRTF low, central orhigh growth. This type of growth can be defined by including an 'L', 'C' or 'H' trafficgrowth indicator on the BASE NETWORK = … line within the input deck.Furthermore users can also input opening year link flows and include their owngrowth rates via the GROWTH RATES command.

(ii) NESA can accept future year matrices for any number of years - specifically these doNOT need to be at 5-year intervals. Inputting of a future year matrix overwrites thestandard NESA matrix for that year.

(iii) Flat growth is assumed prior to the first specified year - however, this is likely to beacademic as the first specified year is normally the design opening year.

(iv) After the first specified year, the calculation of growth subsequent to any future yearmatrix input uses NRTF growth (low, central or high) rates until the year before thenext future year matrix input or ad infinitum if no further future year matrices areinput.

(v) Future year matrices should be built in the same run as the base year matrix.

5.33 The PRINT MATRIX command may be used to produce a printout of a future year matrix. However,this command produces a printout of the most recently built matrix, so a PRINT MATRIX commandfor the base matrix must be positioned before a BUILD FUTURE YEAR MATRIX command.

Examples

5.34 The following examples demonstrate the procedure for the building and printing of an all vehicle tripmatrix and matrices for three user classes. The examples are based upon the base network shown inFigure 10/1/2.



Example 1: ALL VEHICLE matrix

12hr April WDT observed all vehicle matrix

Local vehicle proportions

BUILD MATRICESMATRIX_OPTIONS,NETCLASS=4,MONTH=4CAR=67.00,LGV=15.00,OGV1=9.00,OGV2=9.00,PSV=0.00,ENDMATRIX=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,500 Z03,3020 Z04,110 Z05,240 ENDFROM Z02 TO Z01,500 Z03,90 Z04,1000 Z05,10 ENDFROM Z03 TO Z01,3270 Z02,100 Z04,250 Z05,30 ENDFROM Z04 TO Z01,110 Z02,900 Z03,150 Z05,10 ENDFROM Z05 TO Z01,250 Z02,10 Z03,30 Z04,10 ENDEND MATRIXPRINT MATRIXZONE SELECTIONS=Z01,Z03,Z04,END END JOB

Note: The program will assume defaults for Seasonality Index, E-factor, M-factor and proportion ofcars in work time.

Example 2: 3 User Classes - (Car, LGV & HGV/PSV)

AADT observed matrices for 3 user classes

Local SI, E & M-factors

BUILD MATRICESMATRIX_OPTIONS,EFAC=1.16,MFAC=384,S_I=1.44,END! Car matrixMATRIX=1,2,3,4,5,6,7,8,9,10,11FROM Z01 TO Z02,350 Z03,2114 Z04,77 Z05,168 ENDFROM Z02 TO Z01,350 Z03,63 Z04,700 Z05,7 ENDFROM Z03 TO Z01,2289 Z02,70 Z04,175 Z05,21 ENDFROM Z04 TO Z01,77 Z02,630 Z03,105 Z05,7 ENDFROM Z05 TO Z01,175 Z02,7 Z03,21 Z04,7 ENDEND MATRIX! LGV matrixMATRIX=12FROM Z01 TO Z02,100 Z03,604 Z04,22 Z05,48 ENDFROM Z02 TO Z01,100 Z03,18 Z04,200 Z05,2 ENDFROM Z03 TO Z01,654 Z02,20 Z04,50 Z05,6 ENDFROM Z04 TO Z01,22 Z02,180 Z03,30 Z05,2 ENDFROM Z05 TO Z01,50 Z02,2 Z03,6 Z04,2 ENDEND MATRIX! HGV/PSV matrixMATRIX=13,14,15FROM Z01 TO Z02,50 Z03,302 Z04,11 Z05,24 ENDFROM Z02 TO Z01,50 Z03,9 Z04,100 Z05,1 ENDFROM Z03 TO Z01,327 Z02,10 Z04,25 Z05,3 ENDFROM Z04 TO Z01,11 Z02,90 Z03,15 Z05,1 ENDFROM Z05 TO Z01,25 Z02,1 Z03,3 Z04,1 END



END MATRIXPRINT MATRIXZONE SELECTIONS=Z01,Z03,Z04,ENDEND JOB

NOTES

(i) Trips from zones Z01, Z03 and Z04 requested using the ZONE SELECTIONS=command.

(ii) The matrices can be entered in any order.

Example 3: ALL VEHICLE Flow Group Matrices

BUILD MATRICESMATRIX_OPTIONS,EFAC=1.16,MFAC=384,S_I=1.44CAR=67.00,LGV=15.00,OGV1=9.00,OGV2=9.00,PSV=0.00,ENDDEFINE FLOW GROUPSFG4=HB7+8/DT1+2/MN1-13FG3=HB16+17/DT1+2/MN1-13FG2=HB9+10+11+12+13+14+15+18/DT1+2/MN1-13FG1=REST! AM PeakMATRIX4=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,120 Z03,725 Z04,26 Z05,58 ENDFROM Z02 TO Z01,120 Z03,22 Z04,240 Z05,2 ENDFROM Z03 TO Z01,785 Z02,24 Z04,60 Z05,7 ENDFROM Z04 TO Z01,26 Z02,216 Z03,36 Z05,2 ENDFROM Z05 TO Z01,60 Z02,2 Z03,7 Z04,2 ENDEND MATRIX! PM PeakMATRIX3=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,190 Z03,1148 Z04,42 Z05,91 ENDFROM Z02 TO Z01,190 Z03,34 Z04,380 Z05,4 ENDFROM Z03 TO Z01,1242 Z02,38 Z04,95 Z05,11 ENDFROM Z04 TO Z01,42 Z02,342 Z03,57 Z05,4 ENDFROM Z05 TO Z01,95 Z02,4 Z03,11 Z04,4 ENDEND MATRIX! Between PeaksMATRIX2=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,190 Z03,1148 Z04,42 Z05,91 ENDFROM Z02 TO Z01,190 Z03,34 Z04,380 Z05,4 ENDFROM Z03 TO Z01,1242 Z02,38 Z04,95 Z05,11 ENDFROM Z04 TO Z01,42 Z02,342 Z03,57 Z05,4 ENDFROM Z05 TO Z01,95 Z02,4 Z03,11 Z04,4 ENDEND MATRIXPRINT MATRIXZONE SELECTIONS=Z01,Z03,Z04,ENDEND JOB



Example 4: ALL VEHICLE Future Year Matrix

BUILD FUTURE YEAR MATRIX 2017MATRIX=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,515 Z03,3110 Z04,113 Z05,247 ENDFROM Z02 TO Z01,515 Z03,93 Z04,1030 Z05,10 ENDFROM Z03 TO Z01,3368 Z02,103 Z04,258 Z05,31 ENDFROM Z04 TO Z01,113 Z02,927 Z03,154 Z05,10 ENDFROM Z05 TO Z01,258 Z02,10 Z03,31 Z04,10 ENDEND MATRIX

Example 5: ALL VEHICLE Flow Group Future Year Matrices

BUILD FUTURE YEAR MATRIX 2017! AM PeakMATRIX4=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,124 Z03,747 Z04,27 Z05,60 ENDFROM Z02 TO Z01,124 Z03,23 Z04,247 Z05,2 ENDFROM Z03 TO Z01,809 Z02,25 Z04,62 Z05,7 ENDFROM Z04 TO Z01,27 Z02,222 Z03,37 Z05,2 ENDFROM Z05 TO Z01,62 Z02,2 Z03,7 Z04,2 ENDEND MATRIX! PM PeakMATRIX3=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,196 Z03,1182 Z04,43 Z05,94 ENDFROM Z02 TO Z01,196 Z03,36 Z04,391 Z05,4 ENDFROM Z03 TO Z01,1279 Z02,39 Z04,98 Z05,11 ENDFROM Z04 TO Z01,43 Z02,352 Z03,59 Z05,4 ENDFROM Z05 TO Z01,98 Z02,4 Z03,11 Z04,4 ENDEND MATRIX! Between PeaksMATRIX2=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15FROM Z01 TO Z02,196 Z03,1182 Z04,43 Z05,94 ENDFROM Z02 TO Z01,196 Z03,36 Z04,391 Z05,4 ENDFROM Z03 TO Z01,1279 Z02,39 Z04,98 Z05,11 ENDFROM Z04 TO Z01,43 Z02,352 Z03,59 Z05,4 ENDFROM Z05 TO Z01,98 Z02,4 Z03,11 Z04,4 ENDEND MATRIXPRINT MATRIXEND JOB

Volume 15 Section 1 Chapter 6Part 10 How To Use the NESA Program Building the Base Network Trees


6 BUILDING THE BASE NETWORK TREESGeneral Introduction

6.1 Up to four sets of trees may be built, each set being defined by different values of perturbation, timeand distance coefficients, and number of trees per origin. Each of fifteen demand matrix user classes isassigned to one of the (up to) four trees. Note that it makes little sense to disaggregate the assignmentbeyond the travel demand definition described at the matrix-building stage.

6.2 For large and/or complex NESA models, where four trees are being built, a possible combination ofuser classes is:-

(i) Trees No.1: Car + LGV Business

(ii) Trees No.2: Commuting

(iii) Trees No.3: Leisure

(iv) Trees No.4: OGV and PSV

6.3 The program allows tree building to be carried out at two levels. The first level permits the user to buildselected trees from a few zones. These are printed by the program but are not stored in the base model.The printed trees can then be checked by the user and the tree building options altered if required. Oncethe user is satisfied that the trees are following common sense paths, the second stage can be invokedwhich initiates a full round of tree building between all the network zones. These trees are saved in thebase model for use in the subsequent evaluation stages of the program.

6.4 For small networks only it is possible to build and print all trees at the checking stage. For largernetworks it is more efficient to build and print selected trees.

6.5 Paragraph 1.6 shows the data input required to build and print selected trees and to build all trees.

BUILD TREES command

6.6 The BUILD TREES command must be the first line, and the format is:

TREE_OPTIONS command

6.7 The TREE_OPTIONS command defines the number and type of trees to be built from each originzone. A number of parameters can be specified and these are discussed in Part 5, Chapter 5.

(i) NTR(1..4): the number of trees per origin.

(ii) P(1..4): the perturbation factor for each set of trees. This is a constantcontrolling the deviation in link cost.

(iii) COST(1..4): the relative weighting given to time and distance for each linkin the tree building process. The parameters should be input inthe form T value +D value, i.e. T1.0+D0.5.

Columns 1 - 5 BUILD7 - 11 TREES

Chapter 6 Volume 15 Section 1Building the Base Network Trees Part 10 How To Use the NESA Program


(iv) HEAVY: if set equal to TRUE, this denotes trees to be built using heavyvehicle link speeds. If this option is not required then it shouldbe omitted from the input.

(v) CLASS=(1..15),TREE=(1..4):These options allocate the 15 travel demand classes tothe appropriate set of trees. One line is required for eachdemand class. The CLASS= and TREE= options shouldappear on the same line separated by a comma.

6.8 The format of the TREE_OPTIONS command is:

The NTR, P and COST parameters can be in any order, but should be separated by commas. In the costequation there must be no space immediately before or after the + sign.

6.9 For example: for the second set of trees there are to be 3 trees per origin, the P value is to be 50, and thegeneralised cost equation should be (1.0 * TIME) + (0.50 * DISTANCE).

TREE_OPTIONS,NTR2=3,P2=50,COST2=T1.0+D0.50,END

NOTES:

(i) NTR must be integer, the P, T and D values can be real or integer.

(ii) If any parameters are omitted, the default values assumed by the program are:

NTR1 = 1

P1 = 0

COST1 = T1.0+D0.0 (time only trees)

(iii) The TREE_OPTIONS list must be closed with an END command.

6.10 In the event that only one set of trees is sufficient, it is not necessary to define the assignment of traveldemand classes, but in all other cases all 15 travel demand classes must be accounted for.

ZONE SELECTIONS= sub command

6.11 This is a sub command to BUILD TREES and is used to define those zones for which trees are to bebuilt in the testing stage. The format is identical to that used for matrix printing, described in paragraph5.24:

The same rules regarding continuation lines and the END command apply.

6.12 This command is not required if all trees are being built.

Columns 1 - 13 TREE_OPTIONS,14 - 255 NTR(1..4)=value,P(1..4)=value,COST(1..4)=Tvalue+Dvalue,

HEAVY=TRUE,CLASS=xx,TREE=n,END




TURNING DELAYS= command

6.13 This command is used to input turning delays associated with any junction movements in the network.It should be noted that these delays only apply to the tree building stage in NESA, and they are not usedin the evaluation of delays. Junctions which are to be modelled for delay in the economic evaluationmust be defined explicitly as described in Part 10, Chapter 9. The option is introduced by a first line ofthe format:

6.14 The subsequent lines are used to define the required junction turning delays, one delay being specifiedper line. The turning movement is expressed as the appropriate node triplet, followed by the turningdelay expressed in minutes, all separated by commas.

TURNING DELAYS=10,20,110,9.370,40,50,999END

NOTES:

(i) The turning delay can be to 1 d.p., or integer. Delays expressed to more than 1 d.p.will be truncated and not rounded up.

(ii) Banned turns are defined by a turning delay of 999 minutes.

(iii) The final turning delay must be followed by the END command, on its own line.

(iv) Reference should be made to junction indices (see Part 5, Chapter 3).

If no turning delays are to be input, then the command can be omitted altogether.

6.15 At nodes where at least one turning delay is specified, the program will expand the junction into anumber of sub-nodes, which are identified by suffix letters appended to the new node name. Theseexpansions, however, are internal to the program and the user should always refer to the unexpandednode name. It should be noted that if it is required that a tree should pass through a node twice (e.g. ifU-turns at roundabouts are permitted), then that node should be expanded through the use of a dummyturning delay.

PRINT TREES command

6.16 If selected trees are to be printed as part of a network check, or if all trees are being printed in a simplenetwork, then the END command should be followed by the PRINT TREES command of the format:

6.17 The PRINT PATH command can also be used to print the paths/trees between selected zones. The datarequirements are detailed in Part 10, Chapter 16.

END JOB command

6.18 The standard END JOB command should be used to terminate the run.

Columns 1 - 7 TURNING9 - 15 DELAYS=

Columns 1 - 5 PRINT7 - 11 TREES



Updating A Base Model

6.19 Tree building parameters and turning delays can be updated and/or amended in an existing base model.The appropriate details can be re-specified or new delays added. An existing turning delay can bedeleted by assigning a value of 0.0.

6.20 If the tree building parameters and turning delays for an existing base model are unknown, these can beprinted by selecting and printing trees for one or more zones without specifying the TREE_OPTIONScommand.

Examples

6.21 The following three examples demonstrate the building of all trees and building and printing ofselected trees. The first two examples should be read in conjunction with the matrix examples shown inparagraph 5.34.

Example 1: All Vehicle Matrix

BUILD TREESTREE_OPTIONS,NTR1=3,P1=60,COST1=T1.0+D0.5,ENDEND JOB

NOTES:

(i) Only one set of trees is being built so CLASS=xx,Tree=n is not required.

(ii) A ratio of time to distance of 2:1 has been assumed in the generalised cost equation.

Example 2: Three Vehicle Categories (one set of trees each)

BUILD TREES! Tree 1 = cars! Tree 2 = LGVs! Tree 3 = HGV/PSVsTREE_OPTIONS,NTR1=3,P1=60,COST1=T1.0+D0.5NTR2=3,P2=50,COST2=T1.0+D0.5NTR3=3,P3=40,COST3=T1.0+D0.5,HEAVY=TRUECLASS=1,TREE=1CLASS=2,TREE=1CLASS=3,TREE=1CLASS=4,TREE=1CLASS=5,TREE=1CLASS=6,TREE=1CLASS=7,TREE=1CLASS=8,TREE=1CLASS=9,TREE=1CLASS=10,TREE=1CLASS=11,TREE=1CLASS=12,TREE=2 CLASS=13,TREE=3CLASS=14,TREE=3CLASS=15,TREE=3,ENDPRINT TREESZONE SELECTIONS=Z01,Z02,ENDEND JOB



Example 3: Full set of tree building for 15 observed user classes (matrices)

BUILD TREES! Tree 1 = cars/LGVs business! Tree 2 = cars/LGVs commuting! Tree 3 = cars/LGVs leisure! Tree 4 = HGV/PSVsTREE_OPTIONS,NTR1=3,P1=50,COST1=T1.0+D0.5NTR2=3,P2=80,COST2=T1.0+D0.6NTR3=3,P3=60,COST3=T1.0+D0.8NTR4=3,P3=40,COST3=T1.0+D0.5,HEAVY=TRUECLASS=1,TREE=2CLASS=2,TREE=2CLASS=3,TREE=1CLASS=4,TREE=1CLASS=5,TREE=3CLASS=6,TREE=3CLASS=7,TREE=3CLASS=8,TREE=3CLASS=9,TREE=1CLASS=10,TREE=3CLASS=11,TREE=3CLASS=12,TREE=1 CLASS=13,TREE=4CLASS=14,TREE=4CLASS=15,TREE=4,ENDPRINT TREESZONE SELECTIONS=Z01,Z02,ENDEND JOB

Volume 15 Section 1 Chapter 7Part 10 How To Use the NESA Program Assignment


7 ASSIGNMENT7.1 As part of a NESA model calibration a user is required to carry out a validation check of the network

trip assignment. Full details of the procedure are given in Part 9.

7.2 A useful command to note which permits a user to validate the trip assignment without the need tocarry out a full evaluation is the ASSIGN command. The ASSIGN and EVALUATE commandsperform mutually exclusive tasks, although, it should be noted that both force an assignment. No fatalerror will occur if both commands are included in an input deck but computer running times will beincreased as the program performs the assignment procedure twice.

ASSIGN command

7.3 The ASSIGN command invokes the network assignment. The format is:

Columns 1 - 6 ASSIGN

Chapter 7 Volume 15 Section 1Assignment Part 10 How To Use the NESA Program


Volume 15 Section 1 Chapter 8Part 10 How To Use the NESA Program Evaluating the Base Network Model


8 EVALUATING THE BASE NETWORK MODEL8.1 Assignment, forecasting future flows and the calculation of network costs are all part of the NESA

evaluation stage.

8.2 There are two stopping points within the evaluation stage recognised by the program: after the firstyear junction delay estimates and after the first year network costs have been calculated (see paragraph8.11). Once the base year evaluation is verified then the full assessment can be executed. Paragraph 1.6shows the data input required for the two stages of the evaluation.

8.3 By default, NESA assumes traffic growth on every link to be in accordance with NRTF forecasts,published by Transport Scotland (see Part 5, Chapter 6). However, facilities exist to overwrite thisgrowth by:-

(i) inputting local growth rates (detailed in Part 10, Chapter 5)

(ii) using future year matrices (detailed in Part 10, Chapter 5) and

(iii) using the re-assignment option (detailed in Part10, Chapter 14)

EVALUATE command

8.4 This command introduces the evaluation stage and is of the format:

8.5 The user can specify the apprasial period using the evaluate command. For example, to specify a 35year appraisal period the command would be as follows:

8.6 As mentioned in paragraph 8.1, the EVALUATE command invokes, amongst other activities, anetwork assignment which, apart from calculating link flows, is needed for an expansion of the nodesbeing modelled for junction delay. This assignment is reported in two formats:

(i) In the printout file the flows are expressed in the input units (e.g. 12hr) as all vehicles.In addition, flow groups and AAHT are shown.

(ii) In the ASSIGN.AND file in the NESA model (see Part 10, Chapter 20). The flows areexpressed in the input units and by flow group, and are disaggregated by the 15 UserClasses.

BASE_COSTS= command

8.7 If the capital costs associated with the do-minimum scheme are being modelled, then this commandfollows the EVALUATE command and is used to input the relevant cost data.

8.8 Costs can be allocated to various years throughout the economic life of the design scheme. Costsshould not be assigned to any year prior to the opening year of the design scheme. One line of data isrequired for each year that costs are incurred, the format is:

Example:BASE_COSTS=2015,128.4/6/2013/125.9BASE_COSTS=2019,247.3/6/2013/125.9

Columns 1 - EVALUATE

Columns 1 - EVALUATE35

Chapter 8 Volume 15 Section 1Evaluating the Base Network Model Part 10 How To Use the NESA Program


Geometric Data At Junctions

8.9 If any junctions are to be modelled for delay, then the relevant data should be included immediatelyafter the evaluation parameters and the BASE_COSTS= data, if present.

8.10 The data requirements for junctions are discussed in Part 10, Chapter 9, including the input of localjunction accident data.

STOP AFTER commands

8.11 The evaluation procedure can be stopped at two possible interim stages, before full costs are calculated.The two options are described below.

STOP AFTER DELAY CALCULATIONS command

8.12 This option allows the evaluation to calculate turning volumes and associated delays at all theuser-specified junctions. This allows checks to be made on the realism or otherwise of estimatedjunction delays. The format is:

STOP AFTER FIRST YEAR command

8.13 This option allows the computation of first year costs before terminating the run. This option allowsexamination of speed/flow and economic calculations before proceeding to a full evaluation. Theformat is:

END JOB command

8.14 The last line of the evaluation must be END JOB command. This will either follow the END commandwhich terminates the junction data, or the final END MATRIX command of the future year matrices, ifinput.

Columns 1 - 11 BASE_COSTS=12 - year(E), cost/month(P)/year(P)/CPI

whereyear(E) = the year in which the expenditure is incurredcost = the cost in £000smonth(P)/year(P) =

the date defines the price level at which the costs have been assessed

CPI = the consumer price index for month(P)/year(P)

Columns 1 - 29 STOP AFTER DELAY CALCULATIONS

Columns 1 - 21 STOP AFTER FIRST YEAR

Volume 15 Section 1 Chapter 9Part 10 How To Use the NESA Program Modelling Junctions for Delay


9 MODELLING JUNCTIONS FOR DELAY9.1 Four types of junction can be modelled for delay in NESA: priority junctions, roundabouts, signalised

and accident junctions. The junction data, if present, must follow the EVALUATE and BASE COSTScommand. Junction data can be entered in any order, but it is recommended that for clarity all junctiontypes are grouped together and entered in the order (i) priority junctions, (ii) roundabouts, (iii)signalised junctions and (iv) merge junctions. NESA now adopts a modified ARCADY, PICADY andOSCADY approach, thus radically revising the data input requirements from earlier versions of theprogram.

9.2 Unlike previous versions of NESA, junctions are no longer carried forward from the base to the designnetwork by the program. If a junction is to be evaluated for delay in both the base and design networksthen it must be explicitly coded in both input files, even if the junction is unaffected by the schemeproposals.

JUNCTIONS= command

9.3 The nodes in the network which are to be modelled as junctions must first be introduced using theJUNCTIONS= command. This has the format:

The data can be continued onto further lines in the usual way.

Using the design network illustrated as an example:

JUNCTIONS=200,40,210,220,END

Using the base network example:

JUNCTIONS=20,40,END

NOTES

(i) The last node specified must be followed by the END command to close the list.

(ii) The list should only include the junctions likely to be included in the analysis, and forwhich detailed junction coding is being provided. However, the inclusion of nodes forwhich no data is provided does not cause an error.

(iii) Accident junctions must also be included (see Part10, Chapter 10)

MAXDEL= command

9.4 The total delay on an arm at a junction may be limited to a maximum value. This command mustfollow immediately after the JUNCTIONS command. The format of this command is:

If omitted, a maximum delay of 300 seconds is assumed.

Columns 1 - 10 JUNCTIONS=11 - nodes to be modelled in any order, separated by commas

Columns 1 - 7 MAXDEL=8 - value in seconds

Chapter 9 Volume 15 Section 1Modelling Junctions for Delay Part 10 How To Use the NESA Program


PRIORITY JUNCTIONS

9.5 Three lines of data are required for each priority junction to be modelled in NESA.

PRIORITY= command

9.6 The first line specifies the three nodes which defines the priority movement, the turning movementsrelating to the specified priority movement and the junction category. The format is:

where either LH or RH can be omitted if the junction is only a T-junction.

For example, the junction shown in Figure 10/9/1 (which is relevant to the example base network)could be input as:

PRIORITY=30,20,10,LH=110,CAT=1orPRIORITY=10,20,30,RH=110,CAT=1

9.7 The priority junction categories which can be used are as follows: CAT=1 T-JunctionCAT=2 Cross RoadsCAT=3 Right-Left StaggerCAT=4 Left-Right Stagger

WIDTHS sub command

9.8 The next line of the priority junction definition specifies the widths of the minor road queuing lanes.These widths are of crucial importance since they have a fundamental effect on capacity. The lanewidths must be introduced using the WIDTHS command. The format is:

9.9 LT, LF and TL etc. are keywords which define each lane width and relate to a FROM, TO, RH and LHjunction delineation. The complete set of keywords are as follows and are defined in Figure 10/9/2:

Columns 1 - 9 PRIORITY=10 - node(from),node(via),node(to), LH=node,RH=node,CAT=number

Figure 10/9/1: Priority Junction

Columns 1 - 7 WIDTHS,8 - LT=value,LF=value,TL=value, etc.



FR - From RightLT - Left ToLF - Left FromTL - To LeftRF - Right FromRT - Right ToWP - Width of major road carriageway (refer to Part 8, Figure 8/8/1)CR - Width of kerbed central reserve (refer to Part 8, Figure 8/8/1)

9.10 If the priority carriageway is of dual carriageway standard or the junction layout provides single-lanedualling (with a kerbed central reserve), then on the same line of data the keyword CR should be usedto define the width of the central reserve. The keyword WP denotes the total major road carriagewaywidth excluding any central reserve, either by kerbed or ghost island.

Example

9.11 If the junction layout in Figure 10/9/2 were to include a section of single-lane dualling (kerbed island)then the typical coding format would be:

WIDTHS,LT=3.7,LF=3.7,TL=3.5,RT=3.7,RF=3.7,FR=3.5,CR=3.5,WP=7.3

NOTES

(i) All values are in metres and quoted to 1 or 2 d.p.

(ii) If the minor arm has only one lane then the lane width RT or LF should be coded aszero.

VISIBILITY sub command

9.12 The third and final line used in the coding of priority junctions defines the sight line visibility forturning vehicles, and a flag to indicate the adherence to the DMRB Volume 6 design standard TD 42/

Figure 10/9/2: Carriageway Lane Definition



95. Sight line visibility must be introduced using the VISIBILITY command. Refer to Part 8, Figure 8/8/3. The format is:

NOTES

(i) All values are in metres and quoted to 1 or 2 d.p.

(ii) The TD 42/95 parameter is optional.

Examples

9.13 The complete coding of the base network priority junction in Figure 10/1/2 might be as follows:

PRIORITY=10,20,30,RH=110,CAT=1WIDTHS,FR=3.2,RF=3.65,RT=0.0,WP=7.3VISIBILITY,RF=120.0,RT=120.0,FR=120.0,TD42/95=TRUE

Figures 8/8/1 to 8/8/3 in Part 8, Chapter 8 illustrate how these parameters are measured.

9.14 The complete coding of the design network staggered junction in Figure 10/1/3 might be as follows:

PRIORITY=10,200,210,LH=30,RH=110,CAT=3WIDTHS,LT=5.0,LF=0.0,TL=3.5,FR=3.5,RF=3.0,RT=3.0,CR=3.5,WP=18.6VISIBILITY,LF=120.0,LT=120.0,RF=120.0,RT=120.0,FR=120.0,TL=120.0

9.15 The complete coding of the design network priority junction in Figure 10/1/3 (with ghost island) mightbe as follows:

PRIORITY=210,220,60,LH=50,CAT=1WIDTHS,LT=5.00,LF=0.0,TL=3.5,CR=3.5,WP=10.0VISIBILITY,LF=120.0,LT=120.0,TL=120.0,TD42/95=TRUE

ROUNDABOUTS

9.16 Each node to be modelled as a roundabout is introduced by the ROUNDABOUT= command, followedby a line of data for each of the entries in turn.

ROUNDABOUT= command

9.17 The ROUNDABOUT= command has the format:

Columns 1 - 11 VISIBILITY,12 - 255 LF=value,LT=value,RF=value,RT=value,TL=value,FR=value,

TD42/95=TRUE

Columns 1 - 11 ROUNDABOUT=12 - 255 node(R),ENTRIES=node(E),node(E),node(E)etc.,DIAM=value

wherenode (R) = the node to be modelled as a roundaboutnode (E) = the nodes connected to the roundabout node, specified in clockwise

order. Note exit only nodes should also be specifiedDIAM = the inscribed circle diameter of the roundabout



ENTRY= sub command

9.18 The second and subsequent lines of data define each of the roundabout entries in turn, by specifying anumber of geometric parameters as follows:

AWID: The approach half carriageway width (in metres)EWID: The entry width (in metres)ERAD: The radius of curvature between the entry and the following

weaving section (in metres)FI: The entry angle (in degrees)FLEN: The entry flare length (in metres)MSLIP: Motorway slip flag

9.19 Figures 8/8/4 to 8/8/8 in Part 8, Chapter 8 illustrate how these parameters are measured. The format ofeach line is:

columns 1-6 ENTRY=columns 7-255 node(E),AWID=value,EWID=value,ERAD=value,FI=value,

FLEN=value, MSLIP=TRUE

NOTES

(i) The MSLIP parameter may be omitted. Default, MSLIP=FALSE.

(ii) All values must be quoted to 1 or 2 d.p.

(iii) The entries should be specified in the order that they are defined on theROUNDABOUT= line.

EXIT= sub command

9.20 This command is only used for one-way links exiting the roundabout. The format is:

Example:

The roundabout junction in the example design network in Figure 10/1/3 might be input as follows:

ROUNDABOUT=210,ENTRIES=220,100,200,90,DIAM=60ENTRY=220,AWID=7.3,EWID=10.0,ERAD=20,FI=30,FLEN=30ENTRY=100,AWID=3.65,EWID=4.5,ERAD=25,FI=35,FLEN=10ENTRY=200,AWID=7.3,EWID=10.0,ERAD=20,FI=30,FLEN=30ENTRY=90,AWID=3.65,EWID=5.0,ERAD=30,FI=24,FLEN=20

SIGNALISED JUNCTIONS

9.21 Each node to be modelled as a signalised junction is introduced by the SIGNALS= command. Thenumber of subsequent lines is dependent upon the specific characteristics of the junction beingmodelled. Up to a maximum of four arms can be modelled by NESA. The input data is structured in theformat:

SIGNALS= Defines the node to be modelled and general junction

Columns 1 - 5 EXIT=6 - 15 node(E)



characteristicsThen for each entry:

ENTRY= Defines the entry node and number of lanes

And for each lane of the entry:-

Lane data One line of data defining the characteristics of the given lane.

FROM= Defines any opposing traffic movements.Only required if OPP= parameter appears in the lane dataline immediately preceding.

Then after all the entries and lanes have been defined:-

STAGES= One line for each stage of the cycle.

SIGNALS= command

9.22 The format of the SIGNALS= command is:

9.23 For example, the signal junction (Figure 10/9/3) from the base network example could be coded asfollows:

SIGNALS=40,ENTRIES=50,90,30,70,LOSTTIME=10,MAXCYCLE=120.0,STAGES=2

Or

SIGNALS=40,ENTRIES=50,90,30,70,LOSTTIME=1.5,FIXED=250.0,STAGES=2

Columns 1 - 8 SIGNALS=9 - 255 node(S),ENTRIES=node(E),node(E),node(E),node(E),

LOSTTIME=value, MAXCYCLE=value, FIXED=value, STAGES=value

wherenode(S)= The node to be modelled as signals (mandatory)node(E)= The entry nodes connected to the signals node (mandatory)

LOSTTIME= The lost time per cycle as defined for standard traffic signal calculations - see OSCADY manual (optional))

MAXCYCLE= Maximum cycle time in seconds (optional- default=120s)

FIXED= Fixed cycle time in seconds (optional)STAGES= Number of stages (mandatory)

Period of time during the cycle which is green for one or more streams and during which all signal aspects remain unchanged.



ENTRY= command

9.24 Each entry link must be described in turn and introduced by a line of data containing the entry nodename and the number of lanes at the stopline. The format is:

Example

ENTRY=30,LANES=2

9.25 Each lane within the given entry link must now be described in the order defined by the ENTRIES=command. This is achieved by means of a series of keywords, as follows:-

LEFT Node number of the permitted turning movement from theRIGHT lane (one of LEFT, RIGHT, or STRAIGHT is mandatory)STRAIGHT

OPP Number of opposing movements (optional)

WID Lane width in metres, taken as the average width measured at5m intervals for a distance of 20m back from the stop line

RADLEFT The left turning radius in metres

RADRIGHT The right turning radius in metres

Figure 10/9/3: Signalised Intersection

Columns 1 - 6 ENTRY=7 - node number,LANES=value



EXTRAGREEN The length of extra green time in seconds (optional)The equivalent additional green time available to a particularstream of traffic (due perhaps to reduced lost time)

STOP The number of right turning vehicles (prevailing type) whichcan be stored in advance of the stopline (optional)

STAGE A flag indicating on what stage each lane has the green signal

NOTES

(i) The STAGE= parameter may be repeated if the lane has the green signal in more thanone stage. These stages should be consecutive.

Example

RIGHT=50,WID=3.0,RADRIGHT=50,STAGE=2,STAGE=1,STOP=3,OPP=1

The layout in Figure 10/9/4 illustrates how some of the above parameters are measured.

FROM= command

9.26 Any opposing movement must be defined directly after the lane definition data. The opposingmovement description defines which node the opposing movement is coming from and going to. Oneline is required for each movement. The program assumes that only right turning vehicles are opposed.The format is:

FROM=origin node,TO=destination node

Figure 10/9/4: Signalised Intersection Dimensions



Example

FROM=77,TO=79

STAGE= command

9.27 Following the entry link and lane descriptions, the stages are defined in terms of fixed time andminimum green time (both defined in terms of seconds). If signal timings are not known (zero), thestage definition must still be included. The format is:

NOTES

(i) There must be at least two stages defined, but no more than six.

(ii) If only STAGE= value, is entered a minimum greentime of 10 seconds is assumed.

Example 1

A complete example of a traffic signal junction as shown in Figure 10/9/3 might be as follows:


ENTRY=50,LANES=2

LEFT=90,STRAIGHT=30,WID=3.0,RADLEFT=60,STAGE=1

RIGHT=70,STRAIGHT=30,WID=3.0,RADRIGHT=70,STOP=3,STAGE=1,OPP=1

FROM=30,TO=50

ENTRY=90,LANES=2


RIGHT=50,WID=3.0,RADRIGHT=50 STOP=2,STAGE=2,OPP=1

FROM=70,TO=90

ENTRY=30,LANES=2



FROM=50,TO=30

ENTRY=70,LANES=2


RIGHT=30,WID=3.0,RADRIGHT=50,STOP=2,STAGE=2,OPP=1

FROM=90,TO=70

STAGE=1,FIXED=0.0,MIN=10.0


Example 2

Figure 10/9/5 shows a typical layout of a signalised T-junction with specific lane definitions. Thejunction operates with two signal stages as shown, with link 40-50 being provided with a left filter instage 2.

A complete example of the traffic signal coding is as follows:

SIGNALS=40,ENTRIES=50,90,30,LOSTTIME=10,MAXCYCLE=120.0,STAGES=2

ENTRY=50,LANES=2

LEFT=90,WID=3.0,RADLEFT=60,STAGE=1,STAGE=2

STRAIGHT=30,WID=3.0,STAGE=1

ENTRY=90,LANES=2

LEFT=30,WID=3.0,RADLEFT=40,STAGE=2

Columns 1 - 6 STAGE=7 - 255 value,FIXED=value,MIN=value



RIGHT=50,WID=3.0,RADRIGHT=50,STAGE=2

ENTRY=30,LANES=2

STRAIGHT=50,WID=3.0,STAGE=1


FROM=50,TO=30



MOTORWAY MERGES

9.28 Each node to be modelled as a motorway merge is introduced by the MERGE= command. The nodetriplet which defines the upstream to downstream movement, and the node corresponding to the entryto the merge slip must appear on the same line of data. No further data is required.

Figure 10/9/5: Signalised T - Junction



MERGE= command

9.29 The MERGE= command has the format:

Example:

9.30 The motorway merge layout shown in Figure 10/9/6 would be input as follows:

MERGE=30,20,10,SLIP=110

END command

9.31 Accident data should follow after the last line of the junction data. If there is no accident data, astandard END command should be input on a separate line.

Columns 1 - 6 MERGE=7 - node(from),node(via),node(to),SLIP=node

Figure 10/9/6: Motorway Merge

Volume 15 Section 1 Chapter 10Part 10 How To Use the NESA Program The Modelling of Accidents


10 THE MODELLING OF ACCIDENTS10.1 Benefits gained from a reduction in the number and severity of accidents are an important element in

the economic appraisal of trunk road schemes. Details of the valuation of accidents are given in Part 6,Chapters 4, 5 and 6. The following chapter details the NESA input required to evaluate accident costs/benefits. All the accident commands detailed below must follow the junction commands but can bedefined in any order thereafter.

ACCIDENT_COST_FACTORS= command

10.2 Where accident severities are deemed to differ from the default values for individual road categories,the ACCIDENT_COST_FACTORS= command may be used to input the appropriate cost factor data.Accident costs for individual road categories can only be defined in the Base model. The format is:

with an END command to complete the list.

Example:

This example applies accident cost factors for road categories 26 and 40.

ACCIDENT_COST_FACTORS=26,0.8 40,1.2 END

ACCIDENT_YEAR= command

10.3 Where observed local accident values are available, the ACCIDENT_YEAR= command can be used todefine the year of observation. This single, separate command will apply to all values input via theNACC= or RATE= commands. Where omitted, NESA assumes 2009 values.

Example:

ACCIDENT_YEAR=2012

ACCIDENT_RATES= command

10.4 If accident rates are being input for individual road categories, the ACCIDENT_RATES= command isused to input the relevant rates. Accident rates for individual road categories can only be defined in theBase model. Where observed local accident values are available, the YEAR= command can be used todefine the year of observation. The format is:

with an END command to complete the list.

Accident rates i.e. Personal Injury Accidents per Million Vehicle Kilometres (PIA/MVehKm) can bemanually calculated using the following formula:

Note: a minimum of 5 years worth of accident data needs to be considered.

Columns 1 - 22 ACCIDENT_COST_FACTORS=23 - Road category,cost factor Road category,cost factor etc.

Columns 1 - 15 ACCIDENT_RATES=16 - Road category,rate Road category,rate [YEAR=XXXX] etc.

Chapter 10 Volume 15 Section 1The Modelling of Accidents Part 10 How To Use the NESA Program


Example:

This example alters the accident rates for road categories 26 and 40.

ACCIDENT_RATES=26,0.23 40,0.11 END

Example:

This example alters the accident rate for road category 26 where an observed local rate was calculatedbased on five years’ worth of accident data.

ACCIDENT_RATES=26,0.31 YEAR=2013 END

Accident Rate (PIA/MVehKm) =

No. Accidents in X years (usually 5)2-way AADT * 365 days * X years * Length/106

where: the length is the length of route associated with the number of accidentsobserved.

Volume 15 Section 1 Chapter 10Part 10 How To Use the NESA Program The Modelling of Accidents


ACCIDENT_LINK= command

10.5 Accident rates and cost factors can be specified for individual links in either the Base or Design modelsby using the ACCIDENT_LINK= command. Where the ACCIDENT_YEAR= command has beenspecified, this will apply to all observed values via the NACC= or RATE= commands. The format is:

NOTES

(i) If NACC is input, and ACCIDENT_YEAR= is specified, NESA automaticallycalculates an equivalent 2009 accident rate for the specified link. Where theACCIDENT_YEAR= command is omitted, NESA assumes the user defined rates as2009 values.

(ii) When RATE is specified the NACC parameter should be omitted.

(iii) Inclusion of ACCIDENT_LINK overrides both ACCIDENT_COST_FACTORS andACCIDENT_RATES.

Example:

This example alters the accident rate and cost factors for link 20-30, based upon observed PIAnumbers.

ACCIDENT_LINK=20,30 NACC=10 COST_FACTOR=1.1

ACCIDENT_JUNCTION= command

10.6 Junctions to be modelled for accidents are specified by using the ACCIDENT_JUNCTION command.Where the ACCIDENT_YEAR= command has been specified, this will apply to all observed valuesvia the NACC= commands. The format is:

One line per junction is required, each parameter separated by at least one space.

Example:

This example alters the accident rate for junction node 40, based upon observed PIA numbers.

ACCIDENT_JUNCTION=40 TYPE=37 NACC=4 COST_FACTOR=1.0 CLASS=MINOR

Columns 1 - 14 ACCIDENT_LINK=15 - A-node, B-node RATE=r NACC=n COST_FACTOR=c

wherer accident rate in PIAs per million veh.kms

(Where specified the ACCIDENT YEAR= command will apply)n observed number of PIAs over the last 5 years

(Where specified the ACCIDENT_YEAR= command will apply)c accident cost factor

Columns 1 - 18 ACCIDENT_JUNCTION=19 - node TYPE=t NACC=n COST_FACTOR=c CLASS=mm

wheret junction type, see Part 6, Table 6/6/1n number of accidents in last five years (optional)

(Where specified the ACCIDENT_YEAR= command will apply)c accident cost factormm MAJOR or MINOR (optional in which case the default is MAJOR)

Chapter 10 Volume 15 Section 1The Modelling of Accidents Part 10 How To Use the NESA Program


ACCIDENT_DEFAULT=LINK_ONLY

10.7 Where a separate link and junction assessment is being undertaken, theACCIDENT_DEFAULT=LINK_ONLY command should be used to set all default accident rates, betavalues and casualty splits to Link-Only values. The command should be input after the EVALUATEand will be applied independently to Base and Design decks.

END command

10.8 After all junction and accident data are input, a standard END command must be input on a separateline.

Volume 15 Section 1 Chapter 11Part 10 How To Use the NESA Program Assessment of Alternative Design Schemes


11 ASSESSMENT OF ALTERNATIVE DESIGN SCHEMES

11.1 On completion of the steps outlined in Part 10, Chapters 3 - 10, the final base network model willcontain the details of the base network evaluation. This model is then used as the datum against whichto compare alternative design schemes.

11.2 For each option, the base network model is then updated as described in the sections that follow toform a design network model, leaving the original base network model unchanged. In this wayalternative schemes can be directly compared with the base network.

Chapter 11 Volume 15 Section 1Assessment of Alternative Design Schemes Part 10 How To Use the NESA Program


Volume 15 Section 1 Chapter 12Part 10 How To Use the NESA Program Building the Design Network


12 BUILDING THE DESIGN NETWORK12.1 The design network is built by updating the base network, thus minimising errors and ensuring

consistency of evaluation parameters between the models. The design network is an entirely separateNESA model and the original base network model is unchanged by the updating process.

Defining the Design Network Model

12.2 Part 10, Chapter 2 describes the commands for defining base and design network model names. Thesecommands must form the first line of any design network input file. A typical example might be:

OLD_BASE=EXBASE NEW_DESIGN=ROUTE1A

This would define the existing base network model to be updated as EXBASE, and the new designnetwork model as ROUTE1A.

UPDATE NETWORK command

12.3 This command is exactly the same as defined in Part 10, Chapter 4. It should immediately follow theline containing the model definitions.

Network Title

12.4 This line defines the title of the design network and the year of opening of the scheme. The format ofthe first 15 columns is fixed as follows:

Columns 16-132 contain the network title and opening year in accordance with the followingrequirements.

(i) The title must be contained in double quotes and immediately follow the = sign incolumns 15.

(ii) The opening year follows the title, and should be separated by at least one space.

(iii) The opening year must not be more than ten years after the base year.

(iv) No traffic/economic growth indicator is required after the design year. The growth isdefined in the base network model.

For example:

DESIGN NETWORK="ROUTE OPTION 1" 2017

Network Alteration Commands

12.5 The base network is updated to the design network in exactly the same fashion as is described in Part10, Chapter 4, by using the DELETE, ADD and/or CHANGE commands.

Columns 1 - 6 DESIGN8 - 15 NETWORK=

Chapter 12 Volume 15 Section 1Building the Design Network Part 10 How To Use the NESA Program



12.6 If the user requires a printout of the updated network, the PRINT NETWORK command is used asdefined in paragraph 4.10.

END JOB command

12.7 The standard END JOB command can be used to terminate the run after updating the network, ifdesired.

Example

The design network shown in Figure 10/1/3 could be coded as:-

OLD_BASE=BASE NEW_DESIGN=DESIGN

UPDATE NETWORK

DESIGN NETWORK= "ROUTE OPTION 1 DESIGN EXAMPLE FOR MANUAL" 2017

DELETE

10 20

20 30

20 110

90 100

50 60

END

ADD

TABLE= D R SL ST B HR HF CWID SWID VISI JUNC VW

10 200 0.90 31 113 10 5 5

200 210 4.15 31 113 20 10 5

210 220 3.25 31 113 20 10 10

220 60 1.00 31 113 5 5 10

200 30 2.00 27 96 20 12 8 7.3 0.5 475 5 2.0

200 110 0.75 27 96 15 5 6 7.3 0.5 550 3 2.0

210 90 0.75 27 96 10 17 8 7.3 0.5 385 1 2.0

210 100 0.75 27 96 10 17 8 7.3 0.5 385 1 2.0

220 50 1.00 27 96 20 5 9 7.3 0.5 457 3 2.0

END

END JOB

Volume 15 Section 1 Chapter 13Part 10 How To Use the NESA Program Building the Design Network Trees


13 BUILDING THE DESIGN NETWORK TREES13.1 This stage immediately follows the updating of the network. It should be noted that the BUILD

MATRICES stage does not occur in the design network evaluation; the matrices are read from the basenetwork model.

13.2 The commands used to build the design network trees are exactly the same as those described inchapter 6 for the base network trees, except that the TREE_OPTIONS command is not used in thedesign network. The tree building parameters are read from the base network model and should not berespecified. This is because these parameters must be the same for both the base and design networks.

13.3 All the other commands, such as TURNING DELAYS=, PRINT TREES, ZONE SELECTIONS= etc.are all used in exactly the same manner as for the base network. The turning delays from the basemodel are automatically included in the design model. If any turning movement is not applicable in thedesign network, the turning delay is rejected by the program with a warning message. The user mayinsert additional turning delays, or amend any existing ones, using the TURNING DELAYS =command in the design run.

13.4 Paragraph 1.6 shows the data input required for building the design network trees.

Chapter 13 Volume 15 Section 1Building the Design Network Trees Part 10 How To Use the NESA Program


Volume 15 Section 1 Chapter 14Part 10 How To Use the NESA Program Reassignment of Traffic


14 REASSIGNMENT OF TRAFFIC14.1 NESA now provides the facility for the user to build future year networks using the procedure

described in this chapter. This procedure should be used if the study area road network alters at anypoint during the evaluation period (see Part 5, Chapter 6).

14.2 Having created the standard base and design evaluations, the user should create base and designevaluations of the future year networks in their own directories. These may be built using the standardbase and design years even though the future network is not applicable until a later year.

14.3 The reassignment facility combines output Tables 11,12,14 and 15 only (see Part 10, Chapter 19). Thecombined tables represent the overall economic performance of the scheme. The user should refer tothe individual base and design output files for traffic and link/junction specific economic output.

14.4 The design sub-directory of the future base network should have the same sub-directory name as thestandard design network. For example, if the standard base directory is EXIST11 and the standarddesign is A999, the future base directory could be EXIST17 but the future design must be A999.

PRINT EVALUATION TABLES command

FUTURE_BASE= sub command

The PRINT EVALUATION TABLES command (see Part 10, Chapter 16) has a sub-command whichshould come after the TABLES= command (if it exists). The format is as follows:

Example:

Step 1 Run base input file EXIST11Step 2 Run design input file A999Step 3 Run future year base EXIST17Step 4 Run future year design A999Step 5 Run the following input file

OLD_BASE=EXIST11 OLD_DESIGN=A999PRINT EVALUATION TABLESFUTURE_BASE=EXIST17 2017END JOB

Columns 1 - 12 FUTURE_BASE=13 - directory_name future_year

wheredirectory_name = the sub-directory of the NESADATA directory that contains the

future base;future_year = the year the future year network becomes effective.

Chapter 14 Volume 15 Section 1Reassignment of Traffic Part 10 How To Use the NESA Program


NOTES:

(i) The evaluation tables will use data from the standard base and design runs untilfuture_year when the future base and design run data is used.

(ii) At present, only one future year network is permissible.

Volume 15 Section 1 Chapter 15Part 10 How To Use the NESA Program Evaluating the Design Network


15 EVALUATING THE DESIGN NETWORK15.1 Although the assignment, forecasting and cost calculation can be carried out in one run, it is again

suggested that the user carries out the evaluation in two steps by making use of the STOP AFTERfacility, explained in Part 10, Chapter 8. If future year matrices were used for the base network, they arenot required for the design network as the program automatically applies them, at the correct time, tothe design networks.

EVALUATE command

15.2 This invokes the evaluation stage and takes exactly the same format as described in Part 10, Chapter 8.User defined flow groups are automatically read from the Base model and used in the Design model.

SCHEME_COSTS command

15.3 Scheme costs may be entered either in the form of total scheme costs (TSC) or sub-divided intoconstruction, land & property, preparation and supervision costs (see Part 6, Chapter 8). The format ofthe command is either:

(i) SCHEME_COSTS,TSC=cost/month/(P)/year(P)/CPI,END

or

(ii) SCHEME_COSTS,CONS=cost/month(P)/year(P)/CPI

LAND=cost/month/(P)/year(P)/CPIPREP=pc/first year(P),END

15.4 Table 10/15/1 shows the data input requirements for Option (i), whereas Table 10/15/2 shows those forOption (ii).

Table 10/15/1: Scheme Cost Data Input

Parameter DescriptionTSC = Input Total Scheme Cost in format

TSC = cost/month(P)/year(P)/CPI where:

cost = the total scheme cost in £000s excluding VAT but including design, supervision, land etc (see Part 6 Chapters 7,8 & 9).

month(P)/year(P) = the month and year which define the price base for the cost estimate

CPI = the consumer price index in month(P)/year(P)

Chapter 15 Volume 15 Section 1Evaluating the Design Network Part 10 How To Use the NESA Program


NOTES

(i) The construction cost is distributed over the years defined by the PROFILE=command.

(ii) Land cost is applied to the first year specified by the PROFILE= command.

(iii) Preparation cost is applied evenly from the first year specified to the year precedingthe first profile year.

(iv) Supervision cost is calculated as 5% of the total construction and land costs and isdistributed evenly over the years specified by the PROFILE= command.

Table 10/15/2: Alternative Scheme Cost Data Input

Parameter DescriptionCONS= Input Construction Cost in format

CONS=cost/month(P)/year(P)/CPI where:

cost=cost of construction works in £000s excluding VAT, design, supervision, land etc (see Part 6 Chapters 7, 8 & 9)



LAND= Input Land and Property Cost in format

LAND=cost/month(P)/year(P)/CPI where:

cost=cost of construction works in £000s excluding VAT (see Part 6 Chapters 7, 8 & 9)



PREP= Input Preparation Cost in format

PREP=pc/first year(P) where:

pc=percentage of total construction and land costs that defines the preparation cost (see Part 6 chapters 7, 8 & 9)

first year(P) = the first year in which these costs are incurred.

Volume 15 Section 1 Chapter 15Part 10 How To Use the NESA Program Evaluating the Design Network


PROFILE= command

15.5 The total scheme cost must be allocated to specific years to form an expenditure profile. Part 6, Table6/8/2 gives default expenditure profiles for varying contract lengths, but actual values should be used ifavailable.

15.6 For example, a scheme is due to open in March 2017 with a 2 year construction program. Theexpenditure profile might be:

2016 65%2017 35%

The PROFILE= command is used to input the expenditure profile and has the form:

So the above example would be input as:-

PROFILE=2016,65 2017,35

Clearly, the total profile should sum to 100%.

Junction Data

15.7 Data relating to all junctions which are to be evaluated for delay must be included in the designnetwork input file. Junctions which are common to both networks must be respecified in the design ifthey are to be evaluated.

Junction Accidents

15.8 These are specified in exactly the same manner as for the base network.

STOP AFTER commands

15.9 The two STOP AFTER commands can be used to stop an evaluation early in exactly the same manneras described in Part 10, Chapter 8.

Columns 1 - 8 PROFILE=9 year, % year, % etc.


Chapter 15 Volume 15 Section 1Evaluating the Design Network Part 10 How To Use the NESA Program


END JOB command

15.10 This must be the last line of any design network data file, irrespective of at what stage the evaluationhas been stopped.

Example

For the example network, a design evaluation might be:

EVALUATE

SCHEME_COSTS,CONS=18550/6/2013/125.9/1.1 LAND=550/6/2013/125.9 PREP=5/

2013,END

PROFILE=2016,65 2017,35


PRIORITY=10,200,210,LH=30,RH=110,CAT=3

WIDTHS,LT=5.00,LF=0.0,TL=3.5,FR=3.5,RF=3.0,RT=3.0,CR=10.0,WP=18.6

VISIBILITY,LF=120,LT=120,TL=120,FR=120,RT=120,RF=120,TD42/95=TRUE

PRIORITY=210,220,60,LH=50,CAT=1

WIDTHS,LT=5.00,LF=0.0,TL=3.5,CR=3.5,WP=18.6

VISIBILITY,LF=120.0,LT=120.0,TL=120.0,TD42/95=TRUE

ROUNDABOUT=210,ENTRIES=220,100,200,90,DIAM=40

ENTRY=220,AWID=7.3,EWID=10.0,ERAD=20,FI=30,FLEN=30





ENTRY=50,LANES=2



FROM=30,TO=50

ENTRY=90,LANES=2



FROM=70,TO=90

ENTRY=30,LANES=2



FROM=50,TO=30

ENTRY=70,LANES=2



FROM=90,TO=70



END

END JOB

Volume 15 Section 1 Chapter 16Part 10 How To Use the NESA Program NESA Print Command Options


16 NESA PRINT COMMAND OPTIONS16.1 NESA contains a range of printing options. Reference has been made previously to PRINT

NETWORK, PRINT MATRIX and PRINT TREES, but the full range is detailed here forcompleteness.

16.2 These commands should in general be positioned at the end of the job deck immediately preceding theEND JOB command, though some commands such as PRINT NETWORK may be convenientlypositioned immediately after the corresponding BUILD command.


16.3 As described in Part 10, Chapter 3, if the user wishes to obtain a printout of the network, then a PRINTNETWORK command of the following format should be included:

PRINT MATRIX command

16.4 A print of the input matrices using the format described in Part 10, Chapter 5. As with the network, thematrices can be printed at any stage after they are built. The first line of data, irrespective of whichoption is required, must be the PRINT MATRIX command and has the format:

16.5 If the user wishes to print trips from all zones then no further instructions are required. However theuser has the option to print zones selectively by using the following command.

ZONE SELECTIONS= sub command

16.6 This command allows the user to define individual zones for which trip data is to be printed. It has theformat:

The zone list may be continued on subsequent lines. As with other options, the full 132 columns on agiven line do not have to be used before moving onto the next one. The last specified zone must befollowed by END.

For example, to select zones Z01, Z03 & Z04 for printing:-

ZONE SELECTIONS=Z01,Z03,Z04,END


Columns 1 - 5 PRINT7 - 12 MATRIX


Chapter 16 Volume 15 Section 1NESA Print Command Options Part 10 How To Use the NESA Program


PRINT TREES command

16.7 As described in paragraph 6.15 the format for the PRINT TREES command is as follows:

16.8 The ZONE SELECTIONS= sub command may also be used with PRINT TREES to print zonesselectively.

PRINT ASSIGNMENT command

16.9 The PRINT ASSIGNMENT command produces a printout of link assigned traffic flows for the baseyear in units of (i) the input matrix, (ii) by AAHT and (iii) by flow group by AAHT. The format for thePRINT ASSIGNMENT command is as follows:

16.10 The user may wish to print the assigned traffic flows for specific future years. This option is evoked bythe YEAR=sub command as follows:

YEAR= sub command

16.11 Any year from the base year may be specified.

PRINT JUNC_DELAY command

16.12 The PRINT JUNC_DELAY command produces, for all or selected junctions, a printout of queuing,low flow and geometric delays, by flow group for light and heavy vehicles and for each turningmovement. The capacities in vehicles for each turning movement are also output by flow group. Theformat for the PRINT JUNC_DELAY command is as follows:

JUNCTION_SELECTIONS= sub command

16.13 Junction delays can be output for specific junctions using the JUNCTION_SELECTIONS= subcommand. The format is as follows:

The last specified node must be followed by END.

16.14 Junction delays and capacities can be printed for specific future years, as detailed in paragraph 16.10. Ifthe command is omitted, all junctions specified in the JUNCTIONS command are printed.

Columns 1 - 5 PRINT7 - 11 TREES

Columns 1 - 5 PRINT7 - 16 ASSIGNMENT

Columns 1 - 5 YEAR=6 - 80 year required, separated by commas

Columns 1 - 5 PRINT7 - 16 JUNC_DELAY

Columns 1 - 20 JUNCTION_SELECTIONS=21 - node,node, etc.



PRINT TURNING MOVEMENTS command

16.15 The PRINT TURNING MOVEMENTS command produces an output of turning movements, invehicles, by node number. The format is as follows:

16.16 Turning movements can be printed for specific junctions, as detailed in paragraph 16.13, and forspecific future years as detailed in paragraph 16.10.

PRINT SPEEDS command

16.17 The PRINT SPEEDS command produces a printout of the coded and calculated light and heavy vehiclelink speeds by flow group. The format for the PRINT SPEEDS command is as follows:

16.18 Differences between coded and calculated speeds can be printed for urban links, rural links, forexample 20kph variance, 20% variance by forecast year. This output is invoked by the DIFFERENCEcommand as follows:

16.19 Link speeds can be printed for specific future years, as detailed in paragraph 16.10.

PRINT OVER_CAPACITY command

16.20 The PRINT OVER CAPACITY command produces a printout of all links or junctions where thevolume to capacity ratio is greater than 1.0. The format for the PRINT OVER_CAPACITY commandis as follows:

16.21 Link flows and volume to capacity (v/c) ratios are printed for each link where the v/c exceeds 1.0 forone or more flow groups. Maximum junction v/c for each flow group for each specified junction arealso output where one or more movements have a v/c over 1.0.

16.22 Over-capacity links and junctions can be printed for specific future years, as detailed in paragraph16.10.

Columns 1 - 5 PRINT7 - 13 TURNING15 - 23 MOVEMENTS

Columns 1 - 5 PRINT7 - 12 SPEEDS

Columns 1 - DIFFERENCEKPH=valuePC=valuewhere is one or spaces

Columns 1 - 5 PRINT7 - 19 OVER_CAPACITY



PRINT LINK_FLOWS command

16.23 The PRINT LINK_FLOWS command allows specific link flows to be printed by user group, flowgroup and year. The link matrix flow, AAHT and hourly flows by flow group for each link specified isoutput together with the sum of all links specified. The format for the PRINT LINK_FLOWScommand is as follows:

16.24 Link flows can optionally be disaggregated by user class using the format:

16.25 Link flows are printed for specific future years, as detailed in paragraph 16.10.

16.26 The specific links are defined on subsequent lines as follows:

Further links can be specified on subsequent lines. The last specified link must be followed by END.

PRINT HFGM

16.27 The PRINT HFGM command allows details of the hourly flow group multipliers to be printed by userclass. The format is as follows:

PRINT JOURNEY command

16.28 The PRINT JOURNEY command allows details of specified journeys to be printed by flow group. Theoutput consists of link lengths, speeds and time for each flow group requested along with turningdelays and junction costs. Cumulative time and distances are printed for lights and heavies. The formatfor the PRINT JOURNEY command is as follows:

16.29 Specific flow groups are defined on subsequent lines, as follows:

16.30 Link flows can be printed for specific future years, as detailed in paragraph 16.10.

16.31 The specific links are defined on subsequent lines as follows:

The last specified node must be followed by END.

Columns 1 - 5 PRINT7 - 16 LINK_FLOWS

Columns 1 - 5 PRINT7 - 22 LINK_FLOWS_BY_UC

Columns 1 - A node, B node, dirwhere

dir = 1 or 2 for one way or two way flows (default = 2)

Columns 1 - 5 PRINT7 - 10 HFGM

Columns 1 - 5 PRINT7 - 13 JOURNEY

Columns 1 - FLOW_GROUP=flow group,flow group, etc.

Columns 1 - node1, node2, etc.



PRINT EVALUATION TABLES command

16.32 The PRINT EVALUATION TABLES command produces a printout of the economic evaluation intabular form. Details of the individual tables are provided in Part 10, Chapter 19, including examples.The format for the PRINT EVALUATION TABLES command is as follows:

16.33 The user may only wish to print specific tables. The option is invoked by the TABLES = sub commandas follows:

TABLES= sub command

NOTE: the selection of Table 14 (table number 14) automatically prints out Tables 4, 5, 14, 15A, 15Band 15C. Table 15 cannot be selected on its own.

16.34 A typical example for printing selected tables is as follows:

TABLES=11,12,14

FUTURE_BASE= sub command

16.35 This is a sub-command which should come after the TABLES= command (if it exists). The format is asfollows:

PRINT PATH command

16.36 The PRINT PATH command prints the paths/trees between selected zones. The output consists of nodesequences plus overall times, distances and generalised costs for each tree and each tree class. Theformat for the PRINT PATH command is as follows:

16.37 The specific zone pairs are defined on subsequent lines as follows:

The last specified zone pairing must be followed by END.

Columns 1 - 5 PRINT7 - 16 EVALUATION18 - 23 TABLES

Columns 1 - 7 TABLES=8 - table number, table number, etc.

Columns 1 - 12 FUTURE_BASE=13 - directory_name future_year

wheredirectory_name= the sub-directory of the NESADATA directory that contains the

future base;future_year= the year the future year network becomes effective.

Columns 1 - 5 PRINT7 - 10 PATH

Columns 1 - zone1,zone2 one pair per line



SUPPRESS TABLES command

16.38 The SUPPRESS TABLES command enables the user to suppress the printing of the resource costtables in the output. This command should be located before the EVALUATE command in the inputfile. If growth factors have been input by the user then the growth rate tables are not suppressed. Theformat is as follows:

Columns 1 - 8 SUPPRESS10 - 15 TABLES

Volume 15 Section 1 Chapter 17Part 10 How To Use the NESA Program Amalgamation of Stages in NESA


17 AMALGAMATION OF STAGES IN NESA17.1 Chapters 3 to 16 have described the NESA procedure as a sequence of stages. However, it is general

practice to amalgamate some if not all of these stages into a single computer run.

17.2 Amalgamation is achieved simply by combining the data input for the various stages and removing anyinterim END JOB commands. The only restriction is that it is not possible to amalgamate base networkand design network evaluations together. A complete base network evaluation can be undertaken in asingle run, as can a complete design network evaluation.

Chapter 17 Volume 15 Section 1Amalgamation of Stages in NESA Part 10 How To Use the NESA Program


Volume 15 Section 1 Chapter 18Part 10 How To Use the NESA Program Accident only assessments


18 ACCIDENT ONLY ASSESSMENTSACCIDENT ONLY Inputs

18.1 The network and matrix / flow data inputs required to run an accident only NESA assessment(NESA10 onwards) are discussed below:

Network Data

18.2 In order to run an accident only assessment users need to code the links to be modelled. Linkinformation should be coded in the standard format but only the Mandatory Link Variables are requiredi.e. A-Node, B-Node, Link Length (D), Road Category (RC) and Speed Limit (SL). Within an accidentonly assessment there is no need to include any of the other Optional Link Variables e.g. Bendiness,Hilliness Rise etc. - see Tables 10/3/1 & 10/3/2 on Page 10-3-3.

Matrix / Flow Data

18.3 Accident only assessments require link flow data (disaggregated into 5 vehicle classes - Cars, LGVs,OGV1s, OGV2s and PSVs) to be input. The disaggregated link flow data is necessary in order to applythe NRTF growth rates that correspond to the 5 vehicle classes.

18.4 Link flows for unspecified years before the first specified year are the same as the first specified year,link flows for unspecified years between specified years are linearly interpolated and link flows forunspecified years after the last specified year are subject to NRTF growth from the last specified year.Users can also input opening year link flows and include their own growth rates via the GROWTHRATES command. Users should note that there is potential for overestimating accident numbers andcosts when using link flows (rather than NRTF) to define traffic growth.

Input Commands

18.5 Users should develop separate Do Minimum and Scheme models, but both should be in the format ofNESA Base models rather than the more traditional NESA Base for the Do Minimum and NESADesign for the Scheme.

18.6 The accident only Do Minimum and Scheme input decks should include the BUILD NETWORKrecords, but not the BUILD MATRIX nor the BUILD TREES records.

18.7 Immediately before the EVALUATE command, the following records should be inserted:

ACCIDENT_ONLYMATRIX_OPTIONS,parameter=value,……,ENDYEAR=yearanode,bnode,carflow,lgvflow,ogv1flow,ogv2flow,psvflow……[YEAR=year][anode,bnode,carflow,lgvflow,ogv1flow,ogv2flow,psvflow]…………END

18.8 The MATRIX_OPTIONS defines the units of flow and should normally be included. The program willnot fail if this command is omitted as the internal default values will be assumed, but these may not beappropriate.

Chapter 18 Volume 15 Section 1Accident only assessments Part 10 How To Use the NESA Program


18.9 An accident only assessment using NESA does not require the NETCLASS command, which isnormally specified under the MATRIX_OPTIONS record.

18.10 After the ACCIDENT_ONLY command, the only PRINT commands that are permissible are PRINTEVALUATION TABLES and PRINT LINK_FLOWS.

Running an Accident Only NESA Assessment

General

18.11 Assuming users have link flow data available, the equivalent accident only NESA assessment shouldbe undertaken as follows:

(i) Users should develop separate Do Minimum and Scheme models - both should be inthe format of NESA Base models rather than the more traditional NESA Base for theDo Minimum and NESA Design for the Scheme.

(ii) Flows and growth rates etc. are input into both the Do Minimum and Scheme models.Different flows and growth rates can be input to the Do Minimum and Schememodels if required, allowing a variable trip matrix accident assessment to be carriedout.

(iii) The flow data input to the Do Minimum model should assume the same base /opening year as the Scheme model to ensure that the equivalent appraisal periods areassessed.

(iv) The Do Minimum and Scheme model files should be run through NESA in the normalway.

(v) The accident numbers and costs associated with the Do Minimum and Schememodels are calculated within NESA. Users should extract the accident numbers andcosts associated with the Do Minimum and Scheme from Tables 4 and 5 of theirrespective .LIS output files. Note: the accident costs in Tables 4 and 5 are in MarketPrices.

(vi) Users should manually subtract the Scheme accident numbers and costs from theequivalent Do Minimum accident numbers and costs, to determine the changes inaccident numbers and costs and hence the benefits (or dis-benefits / costs) attributedto the scheme.

(vii) The accident benefits need to be manually incorporated into the Transport Economic& Efficiency (TEE) table output by the main economic assessment e.g. TUBA orPEARS. Note: as the accident benefits / costs are already in Market Prices no furtheradjustment is required prior to incorporating them into the TEE table.

(viii) The economic indicators e.g. PVB, NPV & BCR, should then be manually updatedand reported accordingly.

18.12 Running the accident only assessment as outlined above allows users to input different link flow dataetc. into the Do Minimum and the Scheme networks, and therefore Variable Trip Matrix accidentassessments can be carried out.

18.13 Examples of Base and Design networks for accident only assessments are shown in Part 10 Chapter 19on Pages 10-19-3 and 10-19-4.

Volume 15 Section 1 Chapter 19Part 10 How To Use the NESA Program Example Input Files and Output Tables


19 EXAMPLE INPUT FILES AND OUTPUT TABLES

19.1 Various economic assessment tables can be output by NESA. Each table is described in this chapter,together with notes on how the information may be used.

19.2 Input files based upon the example base and design networks, described in Part 10, Chapter 1, are givenbelow.

Base Network for a typical NESA assessment

NEW_BASE=BASE

BUILD NETWORK

BASE NETWORK="ANY ROAD SCHEME EXAMPLE NETWORK FOR USER MANUAL " 2014 C


! ZONE 1 IS FROM EDINBURGH

TABLE= D R SL ST B HR HF SWID JUNC VW DEVEL P30

Z01 10 0.10 50 96

Z02 80 0.10 50 96

Z03 60 0.10 50 96

Z04 100 0.10 50 96

Z05 110 0.10 50 96

10 20 1.00 26 96 20 6 0 0.5 2 2.0

20 30 2.00 26 96 47 12 8 0.5 5 2.0

20 110 1.00 26 96 83 5 6 0.5 3 2.0

30 40 1.50 1 48 ST 90 100

40 50 1.50 1 48 ST 90 100

40 70 1.00 1 48 ST 90 100

40 90 1.00 1 48 ST 90 100

50 60 2.00 26 96 40 5 9 0.5 3 2.0

70 80 1.00 26 96 27 0 2 0.5 3 2.0

90 100 1.50 26 96 10 8 2 0.5 1 2.0

END

PRINT NETWORK

BUILD MATRICES

MATRIX_OPTIONS,NETCLASS=4,MONTH=4,

CAR=67.0,LGV=15.0,OGV1=9.0,OGV2=6.0,PSV=3.0,END

MATRIX=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15

FROM Z01 TO Z02,500 Z03,3020 Z04,110 Z05,240 END

FROM Z02 TO Z01,500 Z03,90 Z04,1000 Z05,10 END

FROM Z03 TO Z01,3270 Z02,100 Z04,250 Z05,30 END

FROM Z04 TO Z01,110 Z02,900 Z03,150 Z05,10 END

FROM Z05 TO Z01,250 Z02,10 Z03,30 Z04,10 END

END MATRIX

PRINT MATRIX

ZONE SELECTIONS=Z01,Z02,END

BUILD TREES

TREE_OPTIONS,NTR1=3,P1=50,COST1=T1.0+D0.5,END

TURNING DELAYS=

10,20,110,9.3

!70,40,50,999

END

PRINT TREES

EVALUATE

BASE_COSTS=2017,128.4/6/2013/125.9

JUNCTIONS=20,40,END

PRIORITY=10,20,30,RH=110,CAT=1

WIDTHS,FR=3.2,RF=3.65,RT=0.0,WP=7.3

VISIBILITY,RT=120.0,RF=120.0,FR=120.0,TD42/95=TRUE

Chapter 19 Volume 15 Section 1Example Input Files and Output Tables Part 10 How To Use the NESA Program



ENTRY=50,LANES=2



FROM=30,TO=50

ENTRY=90,LANES=2



FROM=70,TO=90

ENTRY=30,LANES=2



FROM=50,TO=30

ENTRY=70,LANES=2



FROM=90,TO=70



ACCIDENT_JUNCTION=40 TYPE=41 NACC=4 COST_FACTOR=1.0

ACCIDENT_COST_FACTORS=26,0.8 27,1.1 END

ACCIDENT_RATES=26,0.5 END

END


END JOB

Design Network for a typical NESA assessment

OLD_BASE=BASE NEW_DESIGN=DESIGN

UPDATE NETWORK

DESIGN NETWORK= "ROUTE OPTION 1 DESIGN EXAMPLE FOR MANUAL" 2017

DELETE

10 20

20 30

20 110

90 100

50 60

END

ADD

TABLE= D R SL ST B HR HF CWID SWID VISI JUNC VW

10 200 0.90 31 113 10 5 5

200 210 4.15 31 113 20 10 5

210 220 3.25 31 113 20 10 10

220 60 1.00 31 113 5 5 10

200 30 2.00 27 96 20 12 8 7.3 0.5 475 5 2.0

200 110 0.75 27 96 15 5 6 7.3 0.5 550 3 2.0

210 90 0.75 27 96 10 17 8 7.3 0.5 385 1 2.0

210 100 0.75 27 96 10 17 8 7.3 0.5 385 1 2.0

220 50 1.00 27 96 20 5 9 7.3 0.5 457 3 2.0

END

PRINT NETWORK

BUILD TREES

EVALUATE

SCHEME_COSTS,CONS=18550/6/2013/125.9/1.1 LAND=550/6/2013/249.7 PREP=5/2013,END

PROFILE=2016,65 2017,35


PRIORITY=10,200,210,LH=30,RH=110,CAT=3

WIDTHS,LT=5.00,LF=0.0,TL=3.5,FR=3.5,RF=3.0,RT=3.0,CR=10.0,WP=18.6

VISIBILITY,LF=120,LT=120,TL=120,FR=120,RT=120,RF=120,TD42/95=TRUE

PRIORITY=210,220,60,LH=50,CAT=1

WIDTHS,LT=5.00,LF=0.0,TL=3.5,CR=3.5,WP=18.6



VISIBILITY,LF=120.0,LT=120.0,TL=120.0,TD42/95=TRUE

ROUNDABOUT=210,ENTRIES=220,100,200,90,DIAM=40






ENTRY=50,LANES=2



FROM=30,TO=50

ENTRY=90,LANES=2



FROM=70,TO=90

ENTRY=30,LANES=2



FROM=50,TO=30

ENTRY=70,LANES=2



FROM=90,TO=70



END


END JOB

Base Network for an accident only NESA assessment

NEW_BASE=Example2DM

BUILD NETWORK

BASE NETWORK="Accident Only Example 2 DO MINIMUM" 2014

ZONES=Z01,Z02,Z03,Z04,END

TABLE= D R SL B HR HF INT AXS DEVEL DES JUNC CWID VW SWID O ST

Z01 100 0.100 50 60

Z02 120 0.100 50 60

Z03 130 0.100 50 60

Z04 140 0.100 50 60

! mainline

100 110 0.426 26 60 0 0 0 0 7.3 2 0

110 140 0.632 26 60 0 0 0 0 7.3 2 1

! sideroads

110 120 0.296 24 60 90 25 25 DES 2 6.0 0 0

110 130 0.415 24 60 90 25 25 DES 2 6.0 0 0

END

PRINT NETWORK

ACCIDENT_ONLY

MATRIX_OPTIONS,EFAC=1.0,MFAC=365,S_I=1.0,END

YEAR=2014

100,110,4907,550,232,175,68

YEAR=2019

100,110,5103,567,244,179,69

YEAR=2024

100,110,5307,583,356,182,69

YEAR=2029

100,110,5520,601,269,186,70

YEAR=2034

100,110,5740,619,282,189,71

END



EVALUATE

JUNCTIONS=110,END


ACCIDENT_COST_FACTORS=6,0.8 END


PRINT LINK_FLOWS

YEAR=2014,2019,2024,2029,2034

100,110,1

END


END JOB

Design Network for an accident only NESA assessmentNEW_BASE=Example2DS

BUILD NETWORK

BASE NETWORK="Accident Only Example 2 DO SOMETHING" 2014

ZONES=Z01,Z02,Z03,Z04,END

TABLE= D R SL B HR HF INT AXS DEVEL DES JUNC CWID VW SWID O ST

Z01 100 0.100 50 60

Z02 120 0.100 50 60

Z03 130 0.100 50 60

Z04 140 0.100 50 60

! mainline

100 110 0.426 29 60 0 0 0 0 7.3 2 0

110 140 0.632 26 60 0 0 0 0 7.3 2 1

! sideroads

110 120 0.296 24 60 90 25 25 DES 2 6.0 0 0

110 130 0.415 24 60 90 25 25 DES 2 6.0 0 0

END

PRINT NETWORK

ACCIDENT_ONLY

MATRIX_OPTIONS,NETCLASS=6,EFAC=1.0,MFAC=365,S_I=1.0,END

YEAR=2014

100,110,4907,550,232,175,68

YEAR=2019

100,110,5103,567,244,179,69

YEAR=2024

100,110,5307,583,356,182,69

YEAR=2029

100,110,5520,601,269,186,70

YEAR=2034

100,110,5740,619,282,189,71

END

EVALUATE

JUNCTIONS=110,END


ACCIDENT_COST_FACTORS=6,0.8 END


PRINT LINK_FLOWS

YEAR=2014,2019,2024,2029,2034

100,110,1

END


END JOB

User Class Proportions by Flow Group (Table 1)

19.3 The user should ensure that the proportions accord with common sense, for example, Other trips shoulddominate the peak flow group for a tourist route subject to seasonal congestion.



TABLE 1 / Matrix User Class Proportions by Flow Group :

Flow Group No.: 1 (5000 hrs)

Class 1 0.10

Class 2 0.09

Class 3 0.03

Class 4 0.02

Class 5 0.05

Class 6 0.09

Class 7 0.11

Class 8 0.14

Class 9 0.03

Class 10 0.03

Class 11 0.07

Class 12 0.12

Class 13 0.05

Class 14 0.06

Class 15 0.01

Working Car/All Cars : 0.11 Working Car/All Vehs : 0.08

Commuter Car/All Cars : 0.24 Commuter Car/All Vehs : 0.18

Link Transit Costs and Benefits (Table 2)

19.4 This table shows the link transit costs and benefits for each link on the network divided into:

(i) value of time;

(ii) vehicle operating costs (both fuel and non-fuel related).

The user should check which links in the base network have the largest cost reduction in the designnetwork and whether these are realistic (cross refer to link geometric properties, flow reductions, speedchanges and v/c ratios).

TABLE 2

L I N K T R A N S I T C O S T S A N D B E N E F I T S

( 60 YEAR DISCOUNTED / MILLIONS )

------------- VOC/F -------- ------------ VOC/NF -------- --------- TRAVEL TIME ------

LINK BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS

100 R110 1.112 1.112 0.000 2.175 2.175 0.000 8.094 8.094 0.000

120 R110 0.116 0.116 0.000 0.229 0.229 0.000 0.876 0.876 0.000

130 R110 0.037 0.037 0.000 0.073 0.073 0.000 0.276 0.276 0.000

140 150 3.879 3.879 0.000 7.527 7.527 0.000 27.357 27.357 0.000

140 R110 1.552 1.552 0.000 2.997 2.997 0.000 10.771 10.771 0.000

150 160 2.518 2.518 0.000 4.848 4.848 0.000 17.334 17.334 0.000

160 170 0.113 0.113 0.000 0.228 0.228 0.000 0.937 0.937 0.000

160 180 0.173 0.173 0.000 0.350 0.350 0.000 1.444 1.444 0.000

160 190 2.732 2.732 0.000 5.365 5.365 0.000 20.097 20.097 0.000

190 200 4.222 4.222 0.000 8.116 8.116 0.000 28.887 28.887 0.000

200 210 2.070 2.070 0.000 4.034 4.034 0.000 14.815 14.815 0.000

210 220 0.027 0.027 0.000 0.054 0.054 0.000 0.223 0.223 0.000

210 230 2.980 0.000 2.980 5.853 0.000 5.853 21.921 0.000 21.921

210 300 0.000 0.462 -0.462 0.000 0.907 -0.907 0.000 3.399 -3.399

230 240 1.863 1.863 0.000 3.687 3.687 0.000 14.094 14.094 0.000

230 310 0.000 0.231 -0.231 0.000 0.454 -0.454 0.000 1.699 -1.699

240 250 0.194 0.194 0.000 0.393 0.393 0.000 1.620 1.620 0.000

240 255 2.311 2.311 0.000 4.608 4.608 0.000 18.018 18.018 0.000

255 260 0.878 0.878 0.000 1.722 1.722 0.000 6.438 6.438 0.000

300 310 0.000 2.323 -2.323 0.000 4.453 -4.453 0.000 15.871 -15.871

---------------------------- ---------------------------- ----------------------------

TOTAL 26.777 26.812 -0.035 52.259 52.220 0.039 193.204 192.252 0.952



Junction Delay Costs (Table 3)

19.5 The information tabulated shows the junction delay costs for modelled junctions in both the base anddesign networks together with the benefits accrued.

The user is recommended to check whether high design network benefits for existing junctions arerealistic (cross refer to a junction’s geometric properties, turning flow changes and v/c ratios). Wouldminimum improvements be feasible; and could high design network user costs for junctions be reducedby redesign?

TABLE 3

J U N C T I O N D E L A Y C O S T S A N D B E N E F I T S

( 60 YEAR DISCOUNTED / MILLIONS )

NODE -- BASE NETWORK -- -- DESIGN NETWORK -- -- BENEFITS --

160 1.144 1.144 0.000

210 0.075 0.075 0.000

240 0.564 0.564 0.000

R110 9.296 9.296 0.000

-----------------------------------------------------------

TOTAL 11.079 11.079 0.000

Link Accident Numbers and Costs (Table 4) & Junction Accident Numbers and Costs (Table 5)

19.6 The following results are produced for all classified links or accident junctions:

• number of accidents in the opening year

• number of accidents in the opening year + 15;

• total number of accidents in the scheme; and

• total cost of accidents in the scheme.

Costs are presented for both the base and design networks with the overall benefits shown also.

The user is recommended to check whether high accident benefits on particular links or junctions arerealistic, for example, could they be distorted due to atypical recent accident records and are they theproblems which the scheme is designed to solve?

TABLE 4

L I N K A C C I D E N T N U M B E R S A N D C O S T S

( COSTS ARE IN MILLIONS, ACCUMULATED AND DISCOUNTED OVER 60 YEARS )

------- BASE SCHEME ------- ------ DESIGN SCHEME ------ --------- BENEFITS --------

2000 NUMBER OF ACCIDENTS TOTAL NUMBER OF ACCIDENTS TOTAL NUMBER OF ACCIDENTS TOTAL

------- LINK ------- COST FAC RATE 2014 2029 TOTAL COST 2014 2029 TOTAL COST 2014 2029 TOTAL BENEFIT

100 R110 1.000 0.381 0.6 0.6 37.3 2.361 0.6 0.6 37.3 2.361 0.0 0.0 0.0 0.000

120 R110 1.000 0.381 0.1 0.1 3.9 0.248 0.1 0.1 3.9 0.248 0.0 0.0 0.0 0.000

130 R110 1.000 0.381 0.0 0.0 1.3 0.079 0.0 0.0 1.3 0.079 0.0 0.0 0.0 0.000

140 150 1.000 0.381 2.1 2.1 129.4 8.202 2.1 2.1 129.4 8.202 0.0 0.0 0.0 0.000

140 R110 1.000 0.381 0.8 0.9 51.6 3.270 0.8 0.9 51.6 3.270 0.0 0.0 0.0 0.000

150 160 1.000 0.381 1.4 1.4 83.6 5.295 1.4 1.4 83.6 5.295 0.0 0.0 0.0 0.000

160 170 1.000 0.381 0.1 0.1 3.9 0.244 0.1 0.1 3.9 0.244 0.0 0.0 0.0 0.000

160 180 1.000 0.381 0.1 0.1 5.9 0.373 0.1 0.1 5.9 0.373 0.0 0.0 0.0 0.000

160 190 1.000 0.381 1.5 1.5 91.7 5.814 1.5 1.5 91.7 5.814 0.0 0.0 0.0 0.000

190 200 1.000 0.381 2.3 2.3 140.0 8.870 2.3 2.3 140.0 8.870 0.0 0.0 0.0 0.000

200 210 1.000 0.381 1.1 1.1 69.2 4.385 1.1 1.1 69.2 4.385 0.0 0.0 0.0 0.000

210 220 1.000 0.381 0.0 0.0 0.9 0.058 0.0 0.0 0.9 0.058 0.0 0.0 0.0 0.000

210 230 B 1.000 0.381 1.6 1.7 100.1 6.344 1.6 1.7 100.1 6.344

210 300 D 1.000 0.381 0.3 0.3 15.5 0.984 -0.3 -0.3 -15.5 -0.984



230 240 1.000 0.381 1.0 1.0 62.8 3.983 1.0 1.0 62.8 3.983 0.0 0.0 0.0 0.000

230 310 D 1.000 0.381 0.1 0.1 7.7 0.492 -0.1 -0.1 -7.7 -0.492

240 250 1.000 0.381 0.1 0.1 6.7 0.420 0.1 0.1 6.7 0.420 0.0 0.0 0.0 0.000

240 255 1.000 0.381 1.3 1.3 78.3 4.963 1.3 1.3 78.3 4.963 0.0 0.0 0.0 0.000

255 260 1.000 0.381 0.5 0.5 29.5 1.868 0.5 0.5 29.5 1.868 0.0 0.0 0.0 0.000

300 310 D 1.000 0.190 0.6 0.6 38.3 2.428 -0.6 -0.6 -38.3 -2.428

---------------------------------------------------------------------------------------------------------------------------------

-

TOTAL 14.5 14.9 896.0 56.776 13.9 14.2 857.5 54.336 0.6 0.6 38.6 2.441

---------------------------------------------------------------------------------------------------------------------------------

-

Fatal casualties 34.6 33.1 1.5

Serious casualties 198.6 190.1 8.5

Slight casualties 1211.2 1159.0 52.1

TABLE 5

J U N C T I O N A C C I D E N T N U M B E R S A N D C O S T S

( COSTS ARE IN MILLIONS, ACCUMULATED AND DISCOUNTED OVER 60 YEARS )

-------- BASE SCHEME ------- ------ DESIGN SCHEME ------- ---------- BENEFITS ---------

2000 NUMBER OF ACCIDENTS TOTAL NUMBER OF ACCIDENTS TOTAL NUMBER OF ACCIDENTS TOTAL

NODE COST FAC TYPE COEFF A 2014 2029 TOTAL COST 2014 2029 TOTAL COST 2014 2029 TOTAL BENEFIT

40 B 1.000 41 0.400 0.8 0.7 42.0 1.648 0.8 0.7 42.0 1.648

---------------------------------------------------------------------------------------------------------------------------------

TOTAL 0.8 0.7 42.0 1.648 0.0 0.0 0.0 0.000 0.8 0.7 42.0 1.648

---------------------------------------------------------------------------------------------------------------------------------

Fatal casualties 0.3 0.0 0.3

Serious casualties 3.9 0.0 3.9

Slight casualties 65.2 0.0 65.2

Link Transit and Junction Delay Costs by Vehicle Category - Discounted (Table 7)

19.7 For each flow group and subsequent vehicle category the following results are tabulated:

(i) junctions - value of time

(ii) links - value of time

- vehicle operating cost (both fuel and non-fuel related)

This table shows the distribution of time and VOC benefits by flow group and vehicle category. Theuser should check that the split of benefits by vehicle category and flow group is reasonable (comparewith user class proportions in Table 1) and accords with the main aims of the scheme. Note that inrelatively uncongested situations, benefits will tend to be related to the proportion of traffic in eachflow group whereas with heavy congestion, benefits will be disproportionately high in the highest flowgroup (particularly with junction relief schemes). Check in which flow group most of the benefits areaccruing, and whether the pattern is realistic in terms of observed current traffic problems. This shouldbe linked to the analysis of the overcapacity link reports.



TABLE 7

*** D I S C O U N T E D ***

L I N K T R A N S I T A N D J U N C T I O N D E L A Y C O S T S (£M)

B Y V E H I C L E C A T E G O R Y

------ VOC/F ----- ----- VOC/NF ----- --- TRAVEL TIME -- ------ JUNC ----- ----- TOTAL ------

BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS

*** 2014 *** FLOW GROUP 3 ( 760 hrs) ***

CAR 0.11 0.11 0.00 0.20 0.20 0.00 0.63 0.63 0.00 0.04 0.03 0.00 0.97 0.97 0.01

LGV 0.02 0.02 0.00 0.04 0.04 0.00 0.09 0.09 0.00 0.01 0.00 0.00 0.15 0.15 0.00

OGV1 0.02 0.02 0.00 0.02 0.02 0.00 0.03 0.03 0.00 0.00 0.00 0.00 0.07 0.07 0.00

OGV2 0.02 0.02 0.00 0.04 0.03 0.00 0.03 0.03 0.00 0.00 0.00 0.00 0.09 0.08 0.00

PSV 0.01 0.01 0.00 0.02 0.02 0.00 0.05 0.05 0.00 0.00 0.00 0.00 0.08 0.08 0.00

TOTAL 0.17 0.17 0.00 0.32 0.32 0.00 0.82 0.82 0.00 0.05 0.05 0.00 1.36 1.35 0.01

Link Transit and Junction Delay Costs by Vehicle Category - Undiscounted (Table 8)

19.8 Results tabulated in the same format as Table 7, except costs are undiscounted.

TABLE 8

*** U N D I S C O U N T E D ***

L I N K T R A N S I T A N D J U N C T I O N D E L A Y C O S T S (£M)


------ VOC/F ----- ----- VOC/NF ----- --- TRAVEL TIME -- ------ JUNC ----- ----- TOTAL ------

BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS BASE DESGN BFITS

*** 2014 *** FLOW GROUP 3 ( 760 hrs) ***

CAR 0.15 0.15 0.00 0.27 0.27 0.00 0.86 0.86 0.01 0.05 0.05 0.00 1.33 1.32 0.01

LGV 0.03 0.03 0.00 0.06 0.06 0.00 0.12 0.12 0.00 0.01 0.01 0.00 0.21 0.21 0.00

OGV1 0.02 0.02 0.00 0.03 0.03 0.00 0.04 0.04 0.00 0.00 0.00 0.00 0.09 0.09 0.00

OGV2 0.03 0.03 0.00 0.05 0.05 0.00 0.03 0.03 0.00 0.00 0.00 0.00 0.12 0.12 0.00

PSV 0.01 0.01 0.00 0.03 0.03 0.00 0.07 0.07 0.00 0.00 0.00 0.00 0.10 0.10 0.00

TOTAL 0.23 0.23 0.00 0.43 0.43 0.00 1.12 1.12 0.01 0.07 0.06 0.01 1.85 1.84 0.01

Summary of Emissions Costs (Table 10)

19.9 For each year of the design scheme’s economic life the low, central and high carbon emissionsdiscounted and undiscounted costs are shown. Additionally, Table 10 provides the total carbon intonnes.



TABLE 10

S U M M A R Y O F E M I S S I O N S C O S T S (£M)

TOTAL CARBON ----------------- UNDISCOUNTED ---------------- ------------------ DISCOUNTED -----------------

TONNES LOW CENTRAL HIGH LOW CENTRAL HIGH

BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN

2014 25.87 25.87 0.021 0.021 0.041 0.041 0.062 0.062 0.015 0.015 0.030 0.030 0.045 0.045

2015 25.77 25.77 0.021 0.021 0.042 0.042 0.062 0.062 0.015 0.015 0.029 0.029 0.044 0.044

2016 25.73 25.73 0.021 0.021 0.042 0.042 0.063 0.063 0.014 0.014 0.029 0.029 0.043 0.043

----------------------------------------------------------------------------------------------------------------------------

2017 25.80 25.84 0.021 0.021 0.043 0.043 0.064 0.064 0.014 0.014 0.028 0.028 0.043 0.043

2018 25.91 25.95 0.022 0.022 0.044 0.044 0.066 0.066 0.014 0.014 0.028 0.028 0.042 0.042

2019 25.89 25.92 0.022 0.022 0.044 0.044 0.066 0.067 0.014 0.014 0.027 0.027 0.041 0.041

2020 25.82 25.86 0.022 0.022 0.045 0.045 0.067 0.067 0.013 0.013 0.027 0.027 0.040 0.040

etc...

Summary of Benefits (Table 11)

19.10 For each year of the design scheme’s economic life the following discounted and undiscounted costsare shown:

• link transit

• junction

• maintenance

• accidents

• emissions

• total annualAdditionally, the table provides values of NPV, NPV/NPC ratio and First Year Rate of Return for each year.Table 11 is not produced for base runs.

TABLE 11

S U M M A R Y O F B E N E F I T S (£M)

LINK TRANSIT JUNCTION MAINTENANCE ACCIDENTS EMISSIONS TOTAL ANNUAL NPV NPV/NPC SYRR

UNDISC DISC UNDISC DISC UNDISC DISC UNDISC DISC UNDISC DISC UNDISC DISC

2017 0.03 0.02 0.00 0.00 0.00 0.00 0.09 0.06 0.00 0.00 0.12 0.08 -1.52 -0.95 5.11

2018 0.03 0.02 0.00 0.00 0.00 0.00 0.10 0.06 0.00 0.00 0.13 0.08 -1.44 -0.90 5.25

2019 0.03 0.02 0.00 0.00 0.00 0.00 0.10 0.06 0.00 0.00 0.13 0.08 -1.36 -0.85 5.39

2020 0.03 0.02 0.00 0.00 0.00 0.00 0.10 0.06 0.00 0.00 0.13 0.08 -1.28 -0.80 5.50

etc...

TOTAL 0.96 0.00 0.00 2.44 0.00 3.39 1.79

Summary of Costs and Benefits (Table 12)

19.11 This table provides an analysis of Present Value of Costs and an analysis of Present Value of Benefitsby year. The results for both base and design schemes are divided into:

• link transit

• junction

• maintenance

• accidents

• emissions



• total annual

As with Table 11, values of NPV, NPV/NPC ratio and FYRR are provided for each year.

The user should check the following:

Does the profile of total benefits over time show a preponderance of benefits in later years? If so,delaying or staging the scheme should be considered; and are the main benefits coming from one ortwo items, and does this accord with the nature of the scheme? (for example, is a new road beingjustified by junction delay benefits which could be obtained by more direct junction improvements)? Isthe timing of the do-something construction cost profile realistic?

TABLE 12

S U M M A R Y O F C O S T S A N D B E N E F I T S (£M)

----------------------------- U N D I S C O U N T E D ------------------------------ DISCOUNTED

LINK TRANSIT JUNCTION MAINTENANCE ACCIDENTS EMISSIONS TOTAL ANNUAL NPC NPV NPV/NPC SYRR

BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN BASE DESIGN

2014 9.81 9.81 0.37 0.37 0.11 0.11 2.04 2.04 0.04 0.04 12.37 12.37 9.08 9.08 0.00 0.00 0.00

2015 10.05 10.05 0.36 0.36 0.11 0.11 2.08 2.08 0.04 0.04 12.65 12.65 18.04 18.04 0.00 0.00 0.00

2016 10.33 10.33 0.37 0.37 0.11 0.11 2.13 2.13 0.04 0.04 12.98 12.98 26.93 26.93 0.00 0.00 0.00

--------------------------------------------------------------------------------------------------------------------------------

2017 10.62 10.58 0.38 0.38 0.11 0.11 2.17 2.08 0.04 0.04 13.32 13.20 35.75 37.27 -1.52 -0.95 5.11

2018 10.90 10.87 0.40 0.40 0.11 0.11 2.22 2.12 0.04 0.04 13.67 13.54 44.49 45.94 -1.44 -0.90 5.25

2019 11.19 11.16 0.41 0.41 0.11 0.11 2.26 2.16 0.04 0.04 14.02 13.89 53.16 54.52 -1.36 -0.85 5.39

2020 11.44 11.40 0.42 0.42 0.11 0.11 2.30 2.20 0.04 0.04 14.32 14.18 61.70 62.98 -1.28 -0.80 5.50

etc...

----------------------------------------------------------------------------------------

TOT 1303.98 1299.21 55.99 55.99 7.87 7.87 282.83 270.94 9.69 9.70 1660.35 1643.71

Conversion of Travel Costs to Market Prices by Vehicle Category (Table 14)

19.12 This table presents travel costs totals by vehicle category in Market Prices. The individual componentsare presented under the TEE categories and converted to market prices by the appropriate taxcorrection factors. Personal Travel covers travel by Car, Private LGVs and some Bus and Coachjourneys. Freight covers travel by freight LGVs plus OGV1s and OGV2s. The Private Sector coversthe remaining Bus and Coach journeys.



TABLE 14

C O N V E R S I O N O F T R A V E L C O S T S

T O M A R K E T P R I C E S


Time Operating fuel Operating non-fuel Total

Vehicle ------------------------------ ---------------------- ---------------------- Operating

category Work Commute Other Total Work Commute Other Work Commute Other Costs

Personal Travel

---------------

Car 0.33 0.11 0.37 0.81 0.00 0.00 -0.01 0.01 0.00 0.00 -0.01

Private LGV 0.00 0.00 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00

-------------------------------------------------------------------------------------------------------

TOTAL 0.33 0.11 0.38 0.82 0.00 -0.01 -0.01 0.01 0.00 0.00 -0.01

Adjustment 0.07 0.02 0.08 0.17 0.00 -0.01 -0.02 0.00 0.00 0.00 -0.03

-------------------------------------------------------------------------------------------------------

MARKET PRICE 0.40 0.13 0.45 0.99 -0.01 -0.01 -0.03 0.01 0.00 0.00 -0.04

Bus & Coach 0.00 0.00 0.00 0.00

Adjustment 0.00 0.00 0.00 0.00

-------------------------------------------------------------------------------------------------------

MARKET PRICE 0.00 0.00 0.00 -0.01

Freight

-------

Freight LGV 0.15 0.00 0.00 0.15 -0.02 0.00 0.00 0.00 0.00 0.00 -0.01

OGV1 -0.01 0.00 0.00 -0.01 0.00 0.00 0.00 0.01 0.00 0.00 0.01

OGV2 -0.01 0.00 0.00 -0.01 0.00 0.00 0.00 0.02 0.00 0.00 0.02

-------------------------------------------------------------------------------------------------------

TOTAL 0.14 0.00 0.00 0.14 -0.02 0.00 0.00 0.03 0.00 0.00 0.01

Adjustment 0.03 0.00 0.00 0.03 -0.03 0.00 0.00 0.01 0.00 0.00 -0.02

-------------------------------------------------------------------------------------------------------

MARKET PRICE 0.17 0.00 0.00 0.17 -0.04 0.00 0.00 0.03 0.00 0.00 -0.01

Private Sector

--------------

Bus & Coach 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Adjustment 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

-------------------------------------------------------------------------------------------------------

MARKET PRICE 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01

TOTALS 0.57 0.13 0.45 1.15 -0.05 -0.01 -0.03 0.05 0.00 0.00 -0.05

This analysis is based on CENTRAL traffic growth

Costs are in 2010 prices in multiples of a million pounds and are discounted to 2010

Evaluation period 60 years

First scheme year 2017

Current year 2014

Discount rate 3.5% for first 30 years, thereafter 3.0% for 45 years, thereafter 2.5%



Summary of Expenditure and Benefits in Market Prices (Table 15)

Table 15A - The Economic Efficiency of the Road System in Market Prices

19.13 Table 15A is an adaptation of the TEE Table (see TAG Unit 3.5.2, The Transport Economic EfficiencySub Objective). Part 9 Chapter 4 outlines how the elements of the TEE table calculated by NESA aretransferred from Table 14 and combined with any Delays During Construction and Maintenance DelaySavings to produce the Net Consumer user Benefit and Net Business Impact.

19.14 Any travel time and vehicle operating costs during construction or maintenance (calculated externallyby users via QUADRO, see DBRB Volume 14) need to be input manually by users to Table 15A inmarket prices and allocated between Consumers and Business in proportion to the Consumer andBusiness User (Time and VOC) benefits of the scheme under normal operating conditions.

Table 15B - Public Accounts

19.15 Table 15B shows the summary of Public Accounts (see TAG Unit 3.5.1, Public Accounts).

Table 15C - Analysis of Monetised Costs and Benefits

19.16 Table 15C summarises the monetised costs and benefits as calculated by NESA and includes anyaccident benefits.

TABLE 15A

E C O N O M I C E F F I C I E N C Y O F T H E R O A D S Y S T E M

I N M A R K E T P R I C E S

IMPACT Reference Calc/Source Total Cars LGVs OGVs Bus/Coach

NON-BUSINESS USER BENEFITS

Travel Time

-----------

Commuting Travel Time 1 0.13 0.13 0.00 0.00

Other Travel Time 2 0.45 0.44 0.01 0.00

Non-business Travel Time 3 1+2 0.58 0.58 0.01 -0.01

Vehicle Operating Costs

-----------------------

Commuting Fuel VOC 4 -0.01 -0.01 0.00

Commuting Non-fuel VOC 5 0.00 0.00 0.00

Other Fuel VOC 6 -0.03 -0.03 0.00

Other Non-fuel VOC 7 0.00 0.00 0.00

Non-business Vehicle Operating Costs 8 4+5+6+7 -0.05 -0.04 -0.01

During Construction and Maintenance

-----------------------------------

Commuting During Construction and Maintenance 9 *

Other During Construction and Maintenance 10 *

NET NON-BUSINESS BENEFITS: COMMUTING 11 1+4+5+9 0.12 0.12 0.00 0.00

NET NON-BUSINESS BENEFITS: OTHER 12 2+6+7+10 0.42 0.41 0.01 0.00

NET NON-BUSINESS BENEFITS: TOTAL 13 11+12 0.54 0.53 0.01 -0.01

BUSINESS USER BENEFITS

User Benefits

-------------

Business Travel Time 14 0.57 0.40 0.18 -0.02 0.00



Fuel VOC 15 -0.05 -0.01 -0.05 0.00

Non Fuel VOC 16 0.04 0.01 0.00 0.03

Business Vehicle Operating Costs 17 15+16 -0.01 0.00 -0.04 0.03

During Construction 18 *

During Maintenance 19 *

During Construction and Maintenance 20 18+19

Subtotal 21 14+17+20 0.56 0.40 0.14 0.02 0.00

Private Sector Provider Impacts

-------------------------------

Revenue 22 *

Fuel VOC 23 0.00 0.00

Non Fuel VOC 24 0.00 0.00

Private Sector Vehicle Operating Costs 25 0.01 0.01

Investment Costs 26 *

Grant/Subsidy 27 *

Subtotal 28 22+25+26+27 0.01 0.01

Other Business Impacts

----------------------

Developer & Other Contributions 29 *

NET BUSINESS IMPACT Total 30 21+28+29 0.57 0.40 0.14 0.02 0.00

TOTAL PRESENT VALUE OF TEE IMPACTS 31 13+30 1.10 0.94 0.15 0.02 0.00





Current year 2014


* Impact calculated external to NESA & manually input by the user. Any manual inputs will require the

manual recalculation of the sub-totals/impacts etc as well as the NPV & BCR etc in Table 15C

TABLE 15B

P U B L I C A C C O U N T S

IMPACT Reference Calc/Source Total

Local Government Funding

------------------------

Revenue 32 *

Investment Costs 33 *

Operating Costs 34 *

Maintenance Costs

-----------------

Non-Traffic (Group 1) 35 *

Traffic Related (Group 2) 36 *


Grant Subsidy Payment 38 *

Net Impact 39 sum 32to38

CENTRAL GOVERNMENT FUNDING : Transport

--------------------------------------

Revenue 40 *

Investment Costs 41 1.94

Operating Costs 42 *

Maintenance Costs



-----------------

Non-Traffic (Group 1) 43 0.00

Traffic Related (Group 2) 44 *


Grant Subsidy Payment 46 *

Net Impact 47 sum 40to46 1.94

Central Government Funding: Non-Transport

-----------------------------------------

Indirect Tax Revenues 48 -0.05

TOTALS

------

Broad Transport Budget 49 39+47 1.94

Wider Public Finances 50 -0.05





Current year 2014


Costs appear as positive numbers, while revenues and Developer & Othe Contributions appear as negative numbers



TABLE 15C

A N A L Y S I S O F M O N E T I S E D C O S T S A N D B E N E F I T S

I N M A R K E T P R I C E S

IMPACT Reference Calc/Source Total

TEE Impacts

Noise 51 * ^

Local Air Quality 52 * ^

Greenhouse Gases (Emissions) 53 0.00

Journey Ambience 54 * ^

Accident Benefits 55 2.44

Non-Business User Benefits:Commuting 56 =11 0.12

Non-Business User Benefits:Other 57 =12 0.42

Business Users & Provider Benefits 58 =30 0.57

Wider Public Finances (Indirect Tax Revenues) 59 =-50 0.05

Option Values 60 * ^

Present Value of Benefits (PVB) 61 sum 51to60 3.59

Broad Transport Budget 62 =49 1.94

Present Value of Costs (PVC) 63 =62 1.94

Overall Impacts

---------------

Net Present Value (NPV=PVB-PVC) 64 =61-63 1.65

Benefit to Cost Ratio (BCR=PVB/PVC) 65 =61/63 1.85





Current year 2014






^ Costs & benefits which are regularly or occasionally presented in monetised form in transport appraisals,

together with some where monetisation is in prospect.

In addition to the costs & benefits outlined above, there may also be significant others, some of which

cannot be presented in monetised form. Where this is the case, the analysis presented above does NOT

provide a good measure of value for money and should not be used as the sole basis for decisions.

Volume 15 Section 1 Chapter 20Part 10 How To Use the NESA Program NESA File Formats


20 NESA FILE FORMATS20.1 No knowledge of the following is required in order to be able to run NESA. The details will be of

interest to users wishing to interface with other software. Not all fields are described, since some havesignificance only to the internal working of NESA.

20.2 In general, the NESA model is contained in a set of files generated by the program. These reside off athird-level NESADATA directory, itself generated by NESA, carrying the NESA model name providedby the user.

20.3 Printed output is to an LP file defined by the user. Apart from this file, all NESA-generated files residein the third-level NESADATA directory.

20.4 NESA is written in FORTRAN-77, and occasional references to output formats below reflect this. Filessuffixed by “.and” are ASCII text files, and can be read with a text editor. Files suffixed by “.bnd” arebinary files, and cannot be read with a text editor. The “net_” prefix is attached to files essential to thebasic network description.

File NETHEAD.AND

Contents A global description of the network and key details of the model

Format Variable length text

Length 9-15 records, depending on model configuration

Comments Each record is labelled, to facilitate recognition

Record 1 Chars 2 - 512 - 22

DATEdate of network creation in DD-MMM-YYYY format


TIMEtime of network creation in HH:MM:SS format


NLINKSnumber of links


NODESnumber of nodes + zones


ZONESnumber of zones


BASEYRbase year


IDESGNopening year; set to zero for base networks

Record 8 Chars 2 -1012 - 91

NET_TITLEnetwork title


JFLAGjflag array in format (20i2)jflag(2) = number of future year matricesjflag(8) = 1 for high growth

Chapter 20 Volume 15 Section 1NESA File Formats Part 10 How To Use the NESA Program


Record 10 Chars 2 - 812 - 21

VERSIONversion number (in form v04.nnn.nn)

Record 11if jflag(2) > 0

Chars 2 - 1112 - 16

FUTURE MATfirst future matrix year


Chars 2 - 1112 - 16

FUTURE MATsecond future matrix year


Chars 2 - 1112 - 16

FUTURE MATthird future matrix year


Chars 2 - 1112 - 16

FUTURE MATfourth future matrix year

Record 15if jflag(2) = 5

Chars 2 - 1112 - 16

FUTURE MATfifth future matrix year

File NETNAMES.AND

Contents Zone and node names

Format Fixed length 10 character text

Length NN records, where NN = number of zones + nodes

Comments The first NZ records contain zone names. The next NN-NZ records contain node names.Zones and nodes are alphanumerically sorted. Names are right-adjusted, blank filled.The position of a zone or node name in this file determines its network sequential reference number (transparent to the user, but essential for computation).

File NETNODEA.AND

Contents Cross-reference table to link NETNAMES.AND and NETLINKS.AND

Format Fixed length 10-character text

Length NN records, where NN = number of zones + nodes

Comments In the same order as NETNAMES.AND (i.e. the nth record in NETNODEA.AND relates to the nth record in NETNAMES.AND)

Record Chars 1 - 5

6 - 10

for the nth record, the first record position in NETLINKS.AND for which the nth node in NETNAMES.AND is an A-nodecorresponding last position

File NETHEAD.AND (Continued)



File NETLINKS.AND

Contents Link parameters

Format Fixed length 91-character text

Length NL, where NL = number of network links

Comments Sorted on columns 1-10. - in any character position indicates a default value

Record Chars 1- 56 - 1011 - 1516 - 2021 - 2526 - 2729 - 3031 - 3334 - 3637 - 3940 - 414243444546474849

50 - 52

53 - 54

55 - 5758 - 6061 - 6364 - 6667 - 69

70 - 7276 - 7677 - 7773 - 7580 - 8586 - 9192 - 94

link A-node sequence numberlink B-node sequence numberlength*100speed*100 .. lightsspeed*100 .. heaviesNESA road category-- (not used)hilliness rises (m/km)hilliness falls (m/km)bendiness (m/km)-- (not used)O for one-way flagC for central flagjunction indexX for external flag0 (not used)reserved for internal use .. a design link Dreserved for internal use .. a deleted linkreserved for internal use .. a changed link(ADDED or CHANGED)CWID*10 average carriageway width between white line edge markings (m)SWID*10 average width of hard strip on both sides, including white line (m)VISI average sight distance (m)verge width*10JUNC*10 side road intersections, both directions (no/km)INT*10 frequency of major intersection (no/km)AXS*10 number of minor intersections and private drives (no/km)DEVEL % frontage development DES flag set to D if a TD9 standard linkST flag set to S if a small town linkP30 % @ 30mph (48kph)calculated accident rate *10000input accident cost factorspeed limit kph



File MATHEAD.AND

Contents A description of the contents of the MATRICES.BND file


Length 5 records

Comments Each record is labelled, to facilitate recognition


DATEdate of matrix creation in DD-MMM-YYYY format


TIMEtime of network creation


ZONESnumber of zones in the matrix

Record 4 Chars 2 - 1215

MATRIX TYPES for single matrix inputF for flow group matrices

Record 5 Chars 2 -1215

FLOW GROUPSnumber of flow groups

File MATRICES.BND

Contents Origin/Destination flows for the base year only

Format Fixed length, NZ+3 wordsThe first two words are integer, the rest are real

Length NZ*NFG*15 records, where NZ = number of network zonesFor S type matrices, NFG = 1For F type matrices, NFG = number of flow groups

Comments Each record contains traffic flows to all traffic zones from a single origin zone.Records are sorted on origin zone within matrix. Matrices appear in the NESA matrix number order

Record Word 1 origin zone sequence number

Word 2 (flow group number)*100 + (user class number);for S type matrices flow group number is 0

Word 3 matrix row total

Next NZ words for each destination zone, the total flow for the time period indicated by word 2



File MATRICE2.BNDMATRICE3.BNDMATRICE4.BNDMATRICE5.BNDMATRICE6.BND

Contents Identical to MATRICES.BND

Format Future year matrices

File EXPLINKS.BNDEXPNAMES.BNDEXPNODEA.BND

Contents Identical to the NET_ equivalents

Length EXPNAMES.AND and EXPNODEA.AND contain NN_EXP records, where NN_EXP is held in the TREEHEAD.AND file.EXPLINKS.AND contains NL_EXP records

Comments EXP files describe the expanded network, generated as a result of the presence of junction turning delays. The first NN records of EXPNAMES.AND are identical to those in NETNAMES.AND

File EXPUNEXP.BND

Contents Cross-reference table to link EXPNAMES.AND and NETNAMES.AND, and EXPLINKS.AND and NETLINKS.AND

Format Fixed length, single integer words

Length NN_EXP-NN+NL_EXP records, where NN_EXP and NL_EXP are held in the TREEHEAD.AND file

Record First NN_EXP-NN records contain the cross-reference between records NN+1 to NN_EXP of EXPNAMES.AND, and records 1 to NN of NAMES.AND.(ie the value in the nth record of EXPUNEXP.BND, where n <= NN_EXP-NN, points to the record in NETNAMES.AND from which the (NN+n)th record in EXPNAMES.AND is derived).The remaining NL_EXP records contain the cross-reference between records 1 to NL_EXP of EXPLINKS.AND and records 1 to NL of NETLINKS.AND



File JUNDELS.AND

Contents Junction turning delays, used for tree-building


Length NDEL records, where NDEL = number of junction turning delays input to NESA, and is held in TREEHEAD.AND

Comments Sorted on the from-via-to movement

Record Chars 1 - 1011 - 2021 - 3031 - 40

from node name, right adjustedturning node nameto node namedelay (mins*10.0)

File TREEHEAD.AND

Contents A description of the tree-building parameters and dimensions of the expanded network


Length Minimum 26 records, maximum 41 records

Comments Each record is labelled to facilitate recognition

Record 1 Chars 2 - 512 -2

DATEdate of creation of the trees in DD-MMM-YYYY format


TIMEtime of creation of the trees in HH:MM:SS format


NTREESnumber of sets of trees


NDELnumber of junction turning delays


NL_EXPnumber of links in the expanded network


NN_EXPnumber of nodes in the expanded network


NTRnumber of trees per origin


Puser-input perturbation factor for cost trees


DCOSTuser-input distance factor for cost trees

Record 10 Chars 2 - 614 - 19

TCOSTuser-input time factor for cost trees



File TREEHEAD.AND (Continued)


HEAVY1 denotes trees built using HGV speeds; 0 otherwise

Records 7-11 are repeated for each set of trees

Subsequent 15 records

Chars 2 - 36

user class number (listed sequentially)corresponding tree set number

File TREES.BND

Contents Minimum cost/time/distance routes between all traffic zones

Format NZ*NTR blocks (where NTR=number of trees per origin) as follows:

1st Record:next(NN_EXP)/100 records:last record:

two integer words

100 integer 2 wordsNN_EXP-100*(NN_EXP/100) integer 2 words

Length NZ*NTR(2+NN_EXP/100) records

Comments Blocks of records are sorted on origin zone.Note the NN_EXP=NN if no turning delays are present.

Record contents for each block of records

Record 1 1st word2nd word

origin zone sequence numbertree number

Subsequent records

100 integer 2 words. In total, NN_EXP words are contained in 1+NN_EXP/100 records.

The nth word contains the position of the backlink in the EXPLINKS.AND file (or NETLINKS.AND file, if no turning delays) for the nth node in the EXPNAMES.AND (or NETNAMES.AND) file.

File ASSIGN.AND

Contents Link flows by time period, for the base year


Length NL*(NFG+2) records

Comments In the same order as the NETLINKS.AND file



File ASSIGN.AND (Continued)

Records for each link

Record 1 Chars 1 - 1011 - 20

link A-node namelink B-node name

Record 2 Chars 1 - 89 - 16.....113 -120

flow for user class 1 (input matrix units)flow for user class 2etc.flow for user class 15

Record 3 Chars 1 - 89 - 16.....113 -120

flow group 1 hourly flow for user class 1flow group 1 hourly flow for user class 2etc.flow group 1 hourly flow for user class 15

This record is repeated for each flow group

File JUNTVOLS.AND

Contents Junction turning volumes


Length Equal to 7*(total number of turning movements through all junctions listed in file JUNLIST.AND)

Comments Sorted on junction turning node

Records for each turning movement

Record 1 Chars 12 - 1617 - 2122 - 2930 - 37.....134 -141

sequential number of inbound linksequential number of outbound linkflow for user class 1 (input matrix units)flow for user class 2etc.flow for user class 15

Record 2 Chars 12 - 1617 - 2122 - 2930 - 37.....134 - 141

sequential number of inbound linksequential number of outbound linkflow group 1 hourly flow for user class 1flow group 1 hourly flow for user class 2etc.flow group 1 hourly flow for user class 15

This record is repeated for each flow group.

File JUNLIST.AND

Contents List of node names for which junction delays are calculated


Length Equal to the number of junctions input by the user following the JUNCTIONS= command



File JUNLIST.AND (Continued)

Comments Sorted

Record Chars 1 -10 node name

The last junction record is followed by the following 2 records:

Penultimaterecord

Chars 1 -10 **********

Last record 1 -10 maximum junction delay

File JUNDESCR.AND

Contents A vetted and formatted representation of the junction data input to NESA


Length According to the number and types of junction data input

Comments Stored in same order as input

Record Blocks of data describing a single junction.Priority junctions are preceded by a record with the single character P, roundabouts by R, signals by S and merge junctions by M. Accident junctions are also described in this file and they are preceded by A.

Priority junctions

Record 1 Chars 11 - 2021 - 2526 - 3536 - 4546 - 5556 - 65

junction node namepriority junction categoryfrom node nameleft-hand turn node nameto node nameright-hand turn node name

Record 2 Chars 11 - 1819 - 2627 - 3435 - 4243 - 5051 - 5859 - 6667 - 74

width FRwidth LTwidth LFwidth TLwidth RFwidth RTwidth of kerbed central reservewidth of major road carriageway

Record 3 Chars 11 - 1819 - 2627 - 3435 - 4243 - 5051 - 5860

visibility LFvisibility LTvisibility TLvisibility RFvisibility RTvisibility FRT if TA42/95 is true; F otherwise



File JUNDESCR.AND (Continued)

Roundabout junctions

Record 1 Chars 11 - 2021 - 283537 - 4647 - 56.....

junction node namediameternumber of entry linksnode name of arm 1node name of arm 2etc.

Entry details occur in the order on the first record.

Entry links are described by a single record.

Chars 2021 - 2829 - 3637 - 4445 - 5253 - 60

M if motorway slip; blank otherwiseAWID approach half carriageway widthEWID entry widthERAD entry radiusFI entry angleFLEN entry flare length

Signals junctions

Record 1 Chars 11 - 20242829 - 3637 - 4445 - 52

junction node namenumber of armsnumber of stagesmaximum cycle timefixed cycle timelost time

For each arm


far node namenumber of lanes

For each lane on this arm

Chars 21 - 3031 - 4041 - 5055

far node name of left turn or blankfar node name of straight on or blankfar node name of right turn or blanknumber of opposing movements

For each opposing movement on this lane

Chars 56 - 6566 - 75

FROM far node nameTO far node name

Record 2 Chars 1 - 56 - 10.....

lane width of lane 1lane width of lane 2etc.

Record 3 Chars 1 - 56 - 10.....

left turn radius of lane 1left turn radius of lane 2etc.



File JUNDESCR.AND (Continued)

Record 4 Chars 1 - 56 - 89 -1314 -16.....

right turn radius of lane 1STOP value (right turning vehicles)right turn radius of lane 2STOP value (right turning vehicles)etc.


extra green time for lane 1extra green time for lane 2

For each stage

Record 1For each arm

Chars 1 - 23 - 4.....

non-zero if lane 1 is green in this stagenon-zero if lane 2 is green in this stageetc.


fixed green timeminimum green time

Merge junctions

Single record for each junction

Chars 11 - 2021 - 3031 - 4041 - 50

upstream node namejunction node namedownstream node nameslip link node name

Accident junctions

Single record for each junction

Chars 11 - 2024 - 2627 - 3031 - 37

junction node nameaccident junction typenumber of accidents in last five yearscost factor

File MATPARAM.AND

Contents Matrix parameter values


Length 31 + (number of flow groups) records. There is an additional 82 records if user growth rates are included

Comments There are header records and descriptive text which are not fully detailed

Each record contains

Record 1 Chars 1 - 2 number of flow groups

Record 2 Chars 1 - 6 number of hours in flow group 1 (repeated for each flow group)



File MATPARAM.AND (Continued)

Next record Chars 1 - 8 value of MFAC

Next record Chars 1 - 8 value of EFAC

Next record Chars 1 - 8 value of FFAC

Next record Chars 1 - 8 seasonality index

Next record Chars 8 network classification

Next record Header record

Next record Chars 11 - 1617 - 2223 - 28.....

% cars in 24hr AADT% cars in flow group 1% cars in flow group 2etc.

This record is repeated for LGVs, OGV1s, OGV2s, PSVs and cars in work mode.

Next record Char 1 T or F

If char 1 = T then

Chars 3 - 78 - 12.....73 - 77

user class 1 proportion for flow group 1user class 2 proportion for flow group 1etc.user class 15 proportion for flow group 1

This record is repeated for each flow group

Next record Chars 13 - 18 proportion of LGVs in non-work mode

Next record Header record

Next record Chars 1 - 34 - 1516 - 27.....

user class numberhourly flow group multiplier for FG 1hourly flow group multiplier for FG 2etc.

This record is repeated for each user class.

If user growth rates are used, the following records are appended

First 2 records Header records

Next record Chars 1 - 67 - 1213 - 1819 - 2425 - 30

% growth for cars to 1994% growth for LGVs to 1994% growth for OGV1s to 1994% growth for OGV2s to 1994% growth for PSVs to 1994

This record is repeated for each year up to 2100.



File ACCPARAM.AND

Contents User input accident rates and cost factors by road category


Length 50 records, one for each road category in numerical order

Record Chars 2 - 34 - 910 - 17

road categoryaccident rateaccident cost factor

File TABLES.AND

Contents Assignment data for evaluation. All costs are expressed in millions of pounds

Format Variable length character text

Length Undetermined

Each record commences with

Chars 2 - 45 - 6

TABtable number (2, 3, 5, 7, 12, 13 or 14)

TAB 2 data Chars 8 - 1718 - 2728 - 3738 - 4748 - 5758 - 6768 - 7779 - 8283 - 8990 - 9697 - 103104 - 110111 - 117

A-node nameB-node nametravel time costsoperating costs (fuel)operating costs (non-fuel)total accident coststotal accident numbersyearaccident cost factor for this linkaccident rate for this linknumber of fatal casualtiesnumber of serious casualtiesnumber of slight casualties

TAB 3 data Chars 8 - 1718 - 2729 - 32

junction node namejunction delay costsyear

TAB 5 data Chars 8 - 1718 - 2728 - 3739 - 4244 - 4546 - 5253 - 5960 - 6667 - 7374 - 80

accident junction node nametotal accident coststotal accident numbersyearaccident typecoefficient acost factornumber of fatal casualtiesnumber of serious casualtiesnumber of slight casualties



File TABLES.AND (Continued)

TAB 7 data Chars 101315 - 1823 - 3536 - 4849 - 6162 - 7475 - 8788 - 100101 - 113114 - 126

mode (1=car, 2=LGV, 3=OGV1, 4=OGV2, 5=PSV, 6=all)flow group numberyeardiscounted junction delay costdiscounted travel time costdiscounted operating cost (fuel)discounted operating cost (non-fuel)undiscounted junction delay costundiscounted travel time costundiscounted operating cost (fuel)undiscounted operating cost (non-fuel)

TAB12 data Chars 9 - 12 year

Then one of 5 forms

1 Chars 17 - 3134 - 43

total base costsBASE COSTS

2 Chars 17 - 3132 - 4647 - 6162 - 7679 - 90

total travel time coststotal operating costs (fuel, car)total operating costs (fuel, other)total operating costs (non-fuel)LINK TRANSIT

3 Chars 17 - 3134 - 41

total junction delay costsJUNCTION

4 Chars 17 - 3134 - 44

total maintenance costsMAINTENANCE

5 Chars 17 - 3132 - 4649 - 57

total link accident coststotal junction accident costsACCIDENTS

6 Chars 17 - 3132 - 4647 - 6162 - 76

total carbon emissions (tonnes)total carbon costs (low)total carbon costs (central)total carbon costs (high)

TAB13 data

Record 1 Chars 7 - 1819 - 3034 - 3537 - 4041 - 464849 - 62

compounded scheme cost in design yearscheme cost discounted to 2010price base monthprice base yearCPInumber of profile yearstotal scheme cost

Record 2 Chars 8 - 1113 - 16.....47 - 5051 - 54.....

first profile yearsecond profile yearetc.first profile %second profile %etc.



TAB 14 data Chars 9 - 1012 - 1520 - 3233 - 4546 - 5859 - 7172 - 84

mode (1- 12)yearjunction delay costs by modetravel time costs by modeVOC fuel costs (petrol) by modeVOC fuel costs (diesel) by modeVOC non-fuel costs by mode

File JACCRATE.AND

Contents Junction accident data


Length Number of accident junctions

Record Chars 1 -1011 - 1516 - 1819 - 2526 - 32

junction node namesequential node numberaccident junction typecoefficient acost factor

File BASECOST.AND

Contents Capital costs associated with the economic base


Length Number of base cost years

Record Chars 1 - 45 - 1214 - 1517 - 2021 - 27

base cost yearcostprice base monthprice base yearCPI

File JUNCAPS.AND

Contents Node data for over-capacity junctions


Length One record for each arm in each flow group in each year for each junction that is over-capacity



Record Chars 1

3 - 1213 - 17192122 - 26

junction typeP = priorityR = roundaboutS = signalsM = merge

junction node nameyearflow group numberarm numbervolume/capacity ratio

File LINKSPDS.AND

Contents Calculated link speeds in kph for all links (except zone connectors) for each flow group for base year and every year up to 2080


Length NLX * NFG * 9 records whereNLX is the number of links less the number of zone connectorsNFG is the number of flow groups

Record Chars 1 - 46 - 91113 - 1719 - 23

yearsequential link numberflow group numbercalculated light speed kphcalculated heavy speed kph

File LINKCAPS.AND

Contents Link data for over-capacity links


Length One record for each link in each flow group in each year that is over-capacity

Record Chars 1 - 45 - 91112 - 1718 - 2324 - 29

yearsequential link numberflow group numberflowcapacityvolume/capacity ratio



File EVALUATE.AND

Contents Evaluation period


Length One record

Record 1 Chars 1 - 813 - 1416 - 19

EVALUATEevaluation years (normally 60)current year

Volume 15 Section 1 Chapter 21Part 10 How To Use the NESA Program NESA Error Messages


21 NESA ERROR MESSAGES21.1 This chapter of the manual gives details of the error messages produced by NESA in tabular form.

21.2 Notes on the error messages are as follows:-

(i) Error Type

Warning Job does not terminate

Fatal Job will terminate after it finishes processing the current command

Stop Job terminates immediately

(ii) Message Type

Full Error message is printed to the screen and the LP file

Screen Error message is printed to the screen only (generally because theLP file does not exist at the time)

(iii) Error Message

Some messages are contained in brackets. These are when the program constructs theerror message which includes some variables.

Table 10/21/1: NESA Error Messages

Number Module Routine Error Type

Message Type

Message

1 O0 nesa97 stop full Unrecognised general command2 PARSE2 parse_read_line stop full Premature end-of-file3 SALFROUT open_lp_file stop screen Unable to create printout file4 O0 dimen_check stop full (maximum array dimension exceeded)6 PARSE2 parse_read_line warning full Blank line encountered in input data7 NESA OpenSelectedFile stop screen Input filepath more than 100 characters11 IO io_files_spec stop full Unrecognised first line of data12 IO io_files_spec stop full Invalid combination of base and design14 IO io_files_spec stop full Failed to create NESA directory15 IO io_files_spec stop full NESA directory does not exist16 IO io_files_spec stop full Failed to create NESA base directory17 IO / PRINT io_files_spec /

read_future_basestop full NESA base directory does not exist

18 IO io_files_spec stop full NESA base directory already exists19 IO / PRINT io_files_spec /

read_future_basestop full NESA design directory does not exist

20 IO io_files_spec stop full NESA design directory already exists21 IO io_files_spec stop full One or more essential NESA network files missing22 IO io_files_spec stop full Failed to create NESA design directory31 INITIALISE initialise_si fatal full Error reading PROFILE_DATA.TXT32 INITIALISE initialise_si fatal full Cannot read environment variable NESA_PROFILE50 various various stop screen Error writing to printout file from routine51 various various stop full Program error in routine52 various various stop full Program dimensions exceeded in routine54 O11_95 various stop full Error writing to TABLES.AND file in routine55 PRINT_TABLES overflow_check warning screen Number too large to be printed in table

Chapter 21 Volume 15 Section 1NESA Error Messages Part 10 How To Use the NESA Program


Table 10/21/2: [Contd]: NESA Error Messages


Message Type

Message

101 O1 netbld stop full BUILD NETWORK can only be the first command102 O1 netbld fatal full Error in network title command103 O1 netbld fatal full Error in ZONES= command104 O1 netbld fatal full No END found in ZONES= command105 O1 netbld fatal full Update network should be used in design run106 O1 netbld warning full High growth indicator not required in design run107 O1 netbld warning full Number of zones greater than 100110 O1 yr_check fatal full Specified base year is earlier than 1984111 O1 yr_check fatal full Specified base year is later than 2010112 O1 yr_check fatal full Specified opening year is earlier than the base year113 O1 yr_check fatal full Specified opening year is more than 10 years after the base year114 O1 yr_check warning full Specified base year is earlier than 1994115 PRINT_TA sum_table2_value stop screen Evaluation period incompatible between base and design runs116 O6 read_current_year stop full Specified current year is outside limit 2010-2030120 O1 lnkdat fatal full Error in link data specification121 O1 lnkdat fatal full This link does not exist in the network122 O1 lnkdat fatal full Invalid distance123 O1 lnkdat fatal full Invalid road category124 O1 lnkdat fatal full Road category for this link should be a zone connector125 O1 lnkdat fatal full (Insufficient link data)126 O1 lnkdat warning full Heavy speed is greater than light speed127 O1 lnkdat warning full Light speed coded but no heavy speed128 O1 lnkdat warning full Heavy speed coded but no light speed129 O1 lnkdat fatal full Hilliness rise and fall should differ for climbing lane links130 O1 lnkdat fatal full Hilliness values for climbing lane links must be specified131 O1 lnkdat warning full Zero coded in one-way field132 O1 lnkdat fatal full Central indicator on link with road category greater than 7133 O1 lnkdat fatal full Invalid speed limit for Central link134 O1 lnkdat fatal full Cannot have small town and central indicators on same link135 O1 lnkdat fatal full Small town indicator on link with road category greater than 7136 O1 lnkdat fatal full Invalid speed limit for small town link137 O1 check_speed_limit fatal full Speed limit missing or invalid for this road category138 O1 lnkdat fatal full Climbing lane links must be 2-way150 O1 logchk fatal full This zone has no entry or exit to the network151 O1 logchk warning full This zone is unconnected152 O1 logchk warning full This node is unconnected153 O1 logchk warning full This node is a dummy node in a one way system160 O1 range_check_integer

/range_check_realwarning full (Value of parameter is outside range)

200 O0 nesa98 stop full Matrix building incorrectly positioned201 O2 matbld stop full Error in MATRIX= command202 O2 matbld fatal full Error reading matrix trip data203 O2 matbld fatal full Origin zone does not exist in the network204 O2 matbld fatal full Destination zone does not exist in the network206 O2 matbld stop full Unable to calculate flow group 1 matrix207 O2 matbld fatal full No matrix defined for user class208 O2 matbld warning full Problem encountered in calculating flow group 1 matrix209 O2 matbld fatal full Both CWK* and CCOM* should be set210 PRINT read_zone_selections fatal full Selected zone does not exist in network211 PRINT matprn stop full Error reading matrix file212 O2 matbld fatal full Mix of single and flow group matrices not permitted213 O2 read_matrix_options stop full MATRIX_OPTIONS not permitted in design run214 O2 read_matrix_options stop full Error in MATRIX_OPTIONS command





Message Type

Message

215 O2 set_user_class_proportions

fatal full Vehicle percentages do not sum to 100

216 PRINT read_zone_selections stop full Error in ZONE SELECTIONS command220 O2 future_matrix_input stop full Bad matrix future year221 O2 future_matrix_input stop full Base matrix must be built in same run as future year matrix231 SI profile_adjustment stop full Error in calculation of Seasonality Index402 O4 option stop full Error in TREE_OPTIONS command403 O4 option stop full TREE_OPTIONS not permitted in design run405 O4 option warning full No classes defined for tree class410 O4 dlayrd fatal full Error in turning delay specification411 O4 dlayrd fatal full Invalid node in turning delay specification412 O4 dlayrd fatal full Invalid turning movement in delay specification420 various various stop full Error reading tree file for zone430 O4 unexpand_notes stop full Cannot find unexpanded equivalent of node510 O5 trbld warning full (No path from zone to...)511 O5 trbld warning full Generalised cost from ... to ... exceeds 999.0512 O5 trbld stop full Array overflow in tree building513 O5 trbld stop full Error writing tree files601 O2 read_define_flow_

groupsstop full Error in DEFINE FLOW GROUPS command

602 O2 read_define_flow_groups

stop full DEFINE FLOW GROUPS not permitted in design run

610 O6 read_scheme_costs fatal full Error in SCHEME_COSTS command612 O2 read_matrix_options warning full Both MFAC and MONTH have been specified on

MATRIX_OPTIONS command616 PARSE_EVAL parse_acconly_flows fatal full No year specified for accident flow data617 PARSE_EVAL parse_acconly_flows fatal full Error in accident flow data618 PARSE_EVAL parse_acconly_flows fatal full Accident only year not in range 2001 to 2100619 PARSE_EVAL parse_acconly_flows fatal full Accident only link not found620 PARSE_EVAL parse_acconly_flows fatal full Accident only link flows invalid623 O6 read_stop_after_

commandstop full Error in STOP AFTER command

624 O6 read_profile stop full Error in PROFILE command625 O6 read_accident_costs stop full Error in ACCIDENT_COSTS command626 O6 read_accident_rates stop full Error in ACCIDENT_RATES command627 O6 read_base_costs stop full Error in BASE_COSTS command629 O10P eval_priority_junctions stop screen Invalid turning movement in junction specification630 O6 move_calc fatal full Node in JUNCTIONS command not found in network631 O6 datchk stop full Error in JUNCTIONS command632 O6 move_calc stop full Junctions not yet defined 633 O6 junction_node_not_

foundfatal full (node not found in JUNCTIONS list)

634 O6 prdat/rndat/sigdat/sigrnddat

fatal full Junction parameters already supplied for node

635 O6 link_accdat fatal full Error in ACCIDENT_LINK command636 O6 link_accdat fatal full Accident link does not exist in network637 O6 link_accdat fatal full Accident link contains numbers and rate638 O6 read_accident_data stop full Error in accident data639 O6 link_accdat fatal full Link accident cost factor too large640 O6 prdat fatal full Error in PRIORITY command641 O6 prdat fatal full Entry link to priority junction does not exist642 O6 prdat fatal full Exit link from priority junction does not exist643 O6 prdat fatal full LH link at priority junction does not exist644 O6 prdat fatal full RH link at priority junction does not exist645 O6 prdat fatal full Error in WIDTHS command





Message Type

Message

646 O6 prdat fatal full Missing or invalid value for LT in WIDTHS command647 O6 prdat fatal full Missing or invalid value for TL in WIDTHS command648 O6 prdat fatal full Missing or invalid value for WP in WIDTHS command649 O6 prdat fatal full Missing or invalid value for RF in WIDTHS command650 O6 prdat fatal full Missing or invalid value for FR in WIDTHS command651 O6 prdat fatal full Error in VISIBILITY command652 O6 prdat fatal full Missing or invalid value for LT in VISIBILITY command653 O6 prdat fatal full Missing or invalid value for LF in VISIBILITY command654 O6 prdat fatal full Missing or invalid value for TL in VISIBILITY command655 O6 prdat fatal full Missing or invalid value for RF in VISIBILITY command656 O6 prdat fatal full Missing or invalid value for RT in VISIBILITY command657 O6 prdat fatal full Missing or invalid value for FR in VISIBILITY command658 O6 prdat fatal full Priority CAT parameter outside range 1-4660 O6 rndat fatal full Error in ROUNDABOUT command661 O6 rndat fatal full Roundabout node does not exist in network662 O6 rndat fatal full Error in roundabout ENTRY command663 O6 rndat fatal full Roundabout ENTRY/EXIT arm not consistent with ENTRIES

node list664 O6 rndat fatal full Roundabout EXIT link does not exist665 O6 rndat fatal full Roundabout ENTRY link does not exist666 O6 rndat fatal full AWID parameter missing or set to zero667 O6 rndat fatal full EWID parameter missing or set to zero668 O6 rndat fatal full ERAD parameter missing or set to zero669 O6 rndat fatal full FI parameter missing or set to zero670 O6 rndat fatal full FLEN parameter missing or set to zero671 O6 rndat fatal full No ENTRY/EXIT command for node672 O6 rndat warning full AWID greater than EWID675 O6 check_accident_nod

e_typewarning full Accident node type inconsistent with junction node type for

node676 O6 check_accident_nod

e_typewarning full Accident node type inconsistent with number of arms at

junction for node677 O6 check_accident_nod

e_typewarning full Accident node type inconsistent with highest link standard for

node679 O6 sigdat fatal full Stage number exceeds max number of stages680 O6 sigdat fatal full Error in SIGNALS command681 O6 sigdat fatal full Signals does not exist in network682 O6 sigdat fatal full Signals ENTRY link does not exist in network683 O6 sigdat fatal full Error in signals ENTRY command684 O6 sigdat fatal full Signal ENTRY link not consistent with ENTRIES node list685 O6 sigdat fatal full Error in signals lane data686 O6 sigdat fatal full Lane width not specified687 O6 sigdat fatal full Left turning radius not specified688 O6 sigdat fatal full Right turning radius not specified689 O6 sigdat fatal full Error in signals opposing movement command690 O6 sigdat fatal full Error in signals STAGE command691 O6 sigdat fatal full Both FIXED and MIN green times set692 O6 sigdat fatal full No ENTRY command for node693 O6 sigdat fatal full More than 4 arms specified694 O6 sigdat fatal full Maximum cycle time not large enough - minimum required =695 O6 sigdat fatal full No STAGE data for stage696 O6 mergdat fatal full Error in MERGE command697 O6 mergdat fatal full Merge link does not exist in network698 O6 accdat fatal full Error in ACCIDENT_JUNCTION command699 O6 accdat fatal full Accident node not in JUNCTIONS list





Message Type

Message

700 O6 accdat fatal full Accident junction type outside range 1-96701 O6 accdat fatal full Accident junction cost factor too large702 O6 accdat fatal full Number of accidents too large704 O6 read_accident_data fatal full ACCIDENT_COST_FACTORS not permitted in design run705 O6 read_accident _data fatal full ACCIDENT_RATES not permitted in design run710 O71 assign warning full (Trips have been lost because of unreached trees between zones)711 O71 assign stop full Error reading matrix row from matrix file713 O71 assign stop full Matrix rows out of order in matrix file721 O2 establish_flow_

groupingstop full Flow exceeds program limits in determination of flow groups

722 PARSE_EVAL parse_define_flow_groups

fatal full More than 8760 hours defined in flow group definitions

723 O2 establish_flow_grouping

fatal full Flow group has zero hours

730 O0 growth_rates stop full Error in growth rate headings740 O6 read_profile fatal full Construction costs defined as well as total scheme costs741 O6 read_profile fatal full Land & property costs defined as well as total scheme costs742 O6 read_profile fatal full Preparation costs defined as well as total scheme costs743 O6 read_profile fatal full Construction costs not defined744 O6 read_profile fatal full Land & property costs not defined745 O6 read_profile fatal full Preparation costs not defined746 O6 read_profile fatal full Preparation costs first year not earlier than construction first year747 O6 read_profile fatal full Profile year invalid748 O6 read_profile fatal full Sum of profile percentages not equal to 100751 O10P/R/S eval_priority_

junctions/(rbt)/(sig)warning full (Node is operating beyond the limits of the delay formulae in

year)752 O10A eval_jun_accidents warning full Error in accident junction specification for node753 O0 evaluate warning full Cycle time has had to be extended for some junctions760 O6 prdat warning full Value for TL in WIDTHS command outside range 2.2-5.0 metres761 O6 prdat warning full Value for FR in WIDTHS command outside range 2.2-5.0 metres762 O6 prdat warning full LH and RH nodes input for T-junction763 O6 prdat warning full No LH node input for crossroads764 O6 prdat fatal full No RH node input for crossroads765 O6 prdat fatal full Incorrect parameters used for T-junction770 O6 sigrnddat fatal full Error in SIGROUND command771 O6 sigrnddat fatal full Signalised roundabout node does not exist in network772 O6 sigrnddat fatal full Error in signalised roundabout ENTRY command773 O6 sigrnddat fatal full Signalised roundabout ENTRY/EXIT arm not consistent with

ENTRIES node list774 O6 sigrnddat fatal full Signalised roundabout EXIT link does not exist775 O6 sigrnddat fatal full Signalised roundabout ENTRY link does not exist776 O6 sigrnddat fatal full LANES parameter missing or set to zero777 O6 sigrnddat fatal full EWID parameter missing or set to zero778 O6 sigrnddat fatal full CLANES parameter missing or set to zero779 O6 sigrnddat fatal full CWID parameter missing or set to zero780 O6 sigrnddat fatal full JDIST parameter missing or set to zero781 O6 sigrnddat fatal full No ENTRY/EXIT command for node800 READ_93 read_trees_header /

read_treesstop full The base trees have not yet been created

801 READ_93 read_trees_header stop full The design trees have not yet been created802 READ_93 read_assignment stop full Error reading assignment file803 READ_93 read_assignment stop full Error reading turning volumes file806 READ_93 read_network warning full (The network was written with version - this is version)807 READ_93 read_network stop full This version of NESA is not compatible with networks pre

97.003





Message Type

Message

810 READ_93 read_jun_dels warning full (Delay read from file incompatible with this network)821 READ_93 read_exp_

networkstop full Error reading EXPNAMES.AND

822 READ_93 read_exp_network

stop full Error reading EXPNODEA.AND


stop full Error reading EXPLINKS.AND


stop full Error reading EXPUNEXP.AND

825 READ_93 read_jun_list stop full Error reading JUNLIST.AND826 READ_93 read_jun_dels stop full Error reading JUNDELS.AND827 READ_93 read_trees_header stop full Error reading TREEHEAD.AND828 READ_93 read_matrix_params stop full Error reading MATPARAM.AND829 READ_93 read_acc_params stop full Error reading ACCPARAM.AND830 PRINT process_over_capacit

ystop full Error reading LINKCAPS.AND

840 WRITE_93 write_matrix_header stop full Error writing MATHEAD.AND841 WRITE_93 write_matrix stop full Error sorting matrix prior to writing MATRICES.AND842 WRITE_93 write_assignment stop full Error writing ASSIGN.AND843 WRITE_93 write_assignment stop full Error writing JUNTVOLS.AND844 WRITE_93 write_matrix_paramsstop full Error writing MATPARAM.AND845 WRITE_93 write_exp_network stop full Error writing EXPNAMES.AND846 WRITE_93 write_exp_network stop full Error writing EXPNODEA.AND847 WRITE_93 write_exp_network stop full Error writing EXPLINKS.AND848 WRITE_93 write_exp_network stop full Error writing EXPUNEXP.AND849 WRITE_93 write_links stop full Error writing NETLINKS.AND850 WRITE_93 write_net stop full Error writing NETNAMES.AND851 WRITE_93 write_net stop full Error writing NETNODEA.AND852 WRITE_93 write_jun_dels stop full Error writing JUNDELS.AND853 WRITE_93 write_jun_list stop full Error writing JUNLIST.AND854 WRITE_93 write_net_header stop full Error writing NETHEAD.AND855 WRITE_93 write_trees_header stop full Error writing TREEHEAD.AND856 WRITE_93 write_acc_params stop full Error writing ACCPARAM.AND870 PRINT print_routines warning full Print keyword not recognised871 PRINT read_year stop full Unrecognised report years data873 PRINT hfgm_print warning full Unable to print HFGM due to no assignment at this stage875 PRINT_TABLES read_tab13 stop full Error reading TAB13 data in TABLES.AND876 PRINT link_flow_print fatal full Error in LINK_FLOW data input877 PARSE_93 parse_link_flow warning full LINK_FLOW link does not exist in network878 PRINT journey_print fatal full Error in JOURNEY data input879 PARSE_93 parse_journey fatal full JOURNEY link does not exist in network881 PRINT read_flow_groups stop full Error in FLOW_GROUP command882 PRINT read_tables stop full Error in TABLES command883 PRINT read_future_base stop full Error in FUTURE_BASE command884 PRINT read_junction_

selectionsstop full Error in JUNCTION SELECTIONS command

885 PRINT read_junction_selections

fatal full JUNCTION SELECTIONS node not found in JUNCTIONS list

886 PRINT read_difference stop full Error in DIFFERENCE command887 PRINT process_speeds stop full Error reading LINKSPDS.AND890 PRINT path_print fatal full Error in PATH data input891 PARSE_93 parse_path fatal full Zone does not exist in network892 PRINT path_print warning full No path found between these zones893 PRINT_TABLES table_11 table_12 fatal full Base cost year before opening year901 FILE file_close warning full Unable to close file on unit904 FILE file_open stop full File does not exist905 FILE file_open stop full Error opening file for read906 FILE file_open stop full Error opening file for write



VOLUME 15 ECONOMIC

ASSESSMENT

OF ROAD SCHEMES

IN SCOTLAND


INDEX AND ABBREVIATIONS

Contents

Chapter

1. Index

2. Abbreviations

PART 11

Volume 15 Section 1Part 11 Index and Abbreviations


Volume 15 Section 1 Chapter 1Part 11 Index and Abbreviations Index


1 INDEXSubject Part.ChapterAbbreviations 11.2Accident 6.4

Accident-Only Node 6.6Costs 6.4, 6.5, 6.6Junctions 6.6Links 6.5Rates 6.4, 6.5, 6.6Types 6.5, 6.6

Combined Link and Junction 6.5Junctions 6.6, 6.6Link Only 6.5

ACCIDENT 10.10_COST_FACTORS= 10.10_LINK= 10.10_RATES= 10.10

ADD 10.4Adjustment of Profiles 5.4All or Nothing Assignment 5.5Annual Average Daily Traffic (AADT) 5.4Annual Average Hourly Traffic (AAHT) 5.2, 5.4Annual Profiles 5.4ARCADY 8.2Assessment Period 3.4Assessment Summary 9.4ASSIGN 10.7Assignment -

All or Nothing 5.5Multiple Routeing 5.5Multiple User Class 5.5

Banned Turns 5.3BASE_COSTS= 10.8BASE NETWORK 10.3Behavioural Value of Time 6.2Bendiness 7.1Block Time 8.6BUILD -

FUTURE YEAR MATRIX 10.5MATRICES 10.5NETWORK 10.3TREES 10.6

Calibration and Validation 9.1Calibration Base 2.2, 5.1Capacity -

Junctions 8.7

Chapter 1 Volume 15 Section 1Index Part 11 Index and Abbreviations


Major/Minor Priority 8.7Merges 8.7Roundabouts 8.7Signalised Roundabouts 8.7Traffic Signals 8.7

Links 7.8Carbon Emissions 6.7Carriageway Width 7.1Central Areas 7.4Central Growth 3.3CHANGE 10.4Classification -

Junction 8Links 7Network 5.2

Climbing Lanes (Dual Carriageways) 5.3,7.3Climbing Lanes (Single Carriageways) 5.3,7.2Command File 10.1Competing Schemes 4.4Complementary Schemes 4.4Construction -

Effect of Delaying 4.5Traffic Delays 6.9, 6.12

Construction Costs 6.8Allocation over Time 6.9Calculation Sheets 6.10Items of Scheme Cost 6.10Land Costs 6.8Preparation Costs 6.9Property Costs 6.8Supervision Costs 6.9

Construction Delays 6.9, 6.12Consumer Price Index (CPI) 6.9Consumer Surplus 3.1Contents 1Conversion to AAHT 5.4Cost Benefit Analysis 3.1Costs -

Discounted 3.3Perceived 3.1Resource 3.6Unperceived 3.1

Costs and Benefits 9.4Current Year 3.3

Damage Only Accidents 6.4Data -

Examples 10.19Format 10.2

Subject Part.Chapter



Daily Profiles 5.4Day Types 5.4.8DEFINE FLOW GROUPS 10.5Delaying Construction 4.5Delays During Construction 6.9, 6.12Delays During Maintenance 6.9, 6.11DELETE 10.4DESIGN NETWORK 10.12Design Standard Choice 4.3Developer Contributions 9.4Developer Funded Schemes 4.4Different Opening Years 4.6Discounting 3.3Do-Minimum 3.2Do-Nothing 3.2Do-Something 3.2Documentation for Submissions 9.5Dual Carriageway Speed/Flow -

Rural 7.3Suburban 7.5

E-Factor 5.2Economic Assessment Report (EAR) 9.5Economic Assessment Validation 9.3Economic Decision Criteria 4.1Economic Concepts 3Economic Indicators 3.3END JOB 10.3,10.4,10.6,10.8,10.12,10.15END MATRIX 10.5Enquiries 1.5ENTRY= 10.9Error Messages 10.21Errors -

Model 4.2, 9.1Program 10.21

EVALUATE 10.8, 10.15Evaluation Period 3.4Examples of Data 10.19Exclusion Analysis 4.4EXIT= 10.9Extent of Network 5.3

Factors -E-Factor 5.2Flow Factor 5.2Flow Group Multipliers 5.4Growth Factors -

Economic 6.2, 6.3, 6.4Traffic 5.6




M-Factor 5.2File Formats 10.19First Year Rate of Return 4.5Fixed Trip Matrix 3.5, 5.6, 9.5Flow -

Composition 5.2Factor 5.2Groups 5.2Group Multipliers 5.4Group Traffic Levels 5.4Group Types 5.2Profiles 5.2Variation 5.2

Forecasts -Economic 2.2, DMRB 13.2Traffic 5.6

Free Format data entry 10.3FROM= 10.9Frontage Development 7.1FUTURE_BASE= 10.14

GDP 6.2, 6.4Generalised Cost 5.5Generation 3.5Geometric Delay 5.3, 8.5Geometric Parameters -

Junctions 8.8Links 7.1

Goodness of Fit 9.2Gradient Adjustment to Fuel Consumption 6.3Growth -

Rate, Accident Costs 6.4Rate, Value of Time 6.2Rate, Fuel 6.3Traffic 5.6

GROWTH RATES 10.5

Highway Maintenance 6.11Highways Economic Note No.2 DMRB 13.2Hilliness 7.1

Implicit Valuation Approach 9.4Incremental Analysis 4.3Index 11.1Indirect Tax Revenues 9.4Isolation Analysis 4.4

Journey Time Measurements 9.3Junction 7.1,8




Accidents 6.6Capacity 8.7Choice 8.2Delays (Assignment) 5.3Geometric Delay 8.5Geometric Parameters 8.8Indices 5.3Interaction 8.1Maximum Delay 8.4Queuing Delay 8.6Tidality 8.1Types Modelled 8.3When to Model 8.1

JUNCTIONS= 10.9JUNCTION_SELECTIONS= 10.16

Land Costs 6.8Link Accidents 6.5Links 5.3

Overcapacity 7.8Speeds 7

Local Authority Schemes 4.4Local Data 5.2Local Model Validation Report 9.1, 9.5

M-Factor 5.2Maintenance delays 6.11Major/Minor Priority Junctions 8.3

Capacity 8.7Geometric Delay 8.5Geometric Parameters 8.8

Market PricesMatrix(ces) 5.4

Fixed Trip 3.5, 5.6, 9.5Variable Trip 2.3, 3.5, 5.6

MATRIX 10.5_OPTIONS 10.5

MAXDEL= 10.9Maximum Delay at Junctions 8.4Mean Factor Values 5.4MERGE= 10.9Merges at Grade Separated Junctions 8.3

Capacity 8.7Geometric Delay 8.5

Minimum Speed 7.1, 7.8Modal Transfer 3.5Model Calibration and Validation 9.1Motorways Speed Flow 7.3Multiple Routeing Assignment 5.5




Multiple User Class Assignment 5.5

National Road Traffic Forecasts 5.6Network -

Classification 5.2Road 5.3Size 5.3

NEW_BASE 10.2NEW_DESIGN 10.2Nodes 5.3Node Expansion 5.5Non-Central Areas 7.4Non-Standard Cases 5.6Non-Working Time 6.2

Occupancy (car) 6.2OLD_BASE 10.2OLD_DESIGN 10.2Opening Years 4.6, 9.3Optimism Bias 6.8OSCADY 8.2Output 10.19Overcapacity on Links 7.8

Peak -and Off-Peak Assignments 5.4Flow Groups 5.2

Perceived Cost 3.1Personal Injury Accident 6.4Perturbation Factor 5.5PICADY 8.2Preparation Costs 6.9Present Value 3.3Present Value Year 3.3, 6.9Price Base Year 3.3Principles of Traffic Modelling 5.1PRINT -

ASSIGNMENT 10.16EVALUATION TABLES 10.16HFGM 10.16JOURNEY 10.16JUNC_DELAY 10.16LINK_FLOWS 10.16MATRIX 10.5, 10.16NETWORK 10.3, 10.4, 10.12, 10.16OVER_CAPACITY 10.16PATH 10.16SPEEDS 10.16TREES 10.6, 10.16




TURNING MOVEMENTS 10.16PRIORITY= 10.9PROFILE= 10.15Profiles

Annual 5.4Daily 5.4Flow 5.2

Program Availability 1.4Program Capacity 1.4Property Costs 6.8

QUADRO 6.11, 6.12Queuing Delay 8.6

Steady State 8.6Time Dependant 8.6

Ranking of Scheme Options 4.4Reassignment Process 5.6Reassignment 3.5Redistribution 3.5Reporting 9.5Resource Cost 3.6Risk Assessment 6.8Road Categories 5.3Road Geometry 7.1Road Maintenance 6.11Road Network 5.3ROUNDABOUT= 10.9Roundabouts 8.3


Route Choice 5.5Run the Program 10Rural Roads Speed/Flow 7.2, 7.3, 7.7

Seasonality Index 5.2SCHEME_COSTS 10.15Scheme Cost Input 6.10Scheme Cost Preparation 6.9Scheme in Route 4.4SIGNALS= 10.9Signalised Roundabouts 8.3


Single Carriageway Speed/Flow -Rural 7.2Small Town 7.6




Suburban 7.5Urban 7.4

Single Track Speed/Flow 7.7Small Town Roads Speed/Flow 7.6Speed/Flow Curves 7.1

Diagrams 7.9Dual Carriageway 7.3Minimum Cut-Off 7.1Relationships 7.9Rural Single Carriageway 7.2Single Track 7.7Small Town 7.6Suburban 7.5Urban 7.4

STAG 3.1STAGE= 10.9STOP AFTER DELAY CALCULATIONS 10.8STOP AFTER FIRST YEAR 10.8Strategy Appraisal 4.4Suburban Roads Speed/Flow 7.5Sunk Costs 3.1Supervision Costs 6.9SUPPRESS TABLES 10.16

TABLE= 10.3TABLES= 10.16Taxation 3.6Terminal Year 4.6Tidality 8.1Time Savings 6.2

Work 6.2Non-Work 6.2

Tolls 5.5Traffic Delays During Construction 6.12Traffic Flow Input 5.2, 5.4

Annual Average Daily Traffic 5.4Annual Average Hourly Traffic 5.212 Hour 5.216 Hour 5.2

Traffic Flow Profiles 5.2,5.4Traffic Flow Variation 5.2Traffic Forecasts 5.6

Local 5.6National 5.6

Traffic Model 5Traffic Signals 8.3

Capacity 8.7Geometric Delay 8.5

TRANSYT 8.3




Note: NESA commands in are in capital letters.

Tree Building 5.5TREE_OPTIONS 10.6Trip Matrices 5.4

Fixed 3.5, 5.6, 9.5Variable 2.3, 3.5, 5.6

TURNING DELAYS 10.6Turning Delay (Assignment) 5.3Turning Flows 8.1

Uncertainty 4.2Unperceived Cost 3.1UPDATE NETWORK 10.4, 10.12Urban Roads Speed/Flow 7.4User Classes 5.2User Costs 3.1User Support 10.2

Validation of a NESA Assessment 9.1,9.5Value of -

Accidents 6.4, 6.5, 6.6Accidents at Junctions 6.6Accidents on Links 6.5Time 6.2Fuel 6.3

Variable Trip Matrix 2.3, 3.5, 5.6VAT 3.6Vehicle -

Categories 5.2Category Proportions 5.2Occupancy 6.2Operating Costs 6.3

Visibility 7.1VISIBILITY 10.9

WIDTHS 10.9Working Time 6.2

YEAR= 10.16

Zones 5.3ZONES= 10.3Zone Connectors 5.3ZONE SELECTIONS= 10.5,10.6,10.16


Volume 15 Section 1 Chapter 2Part 11 Index and Abbreviations Abbreviations


2 ABBREVIATIONSAADT Annual Average Daily Traffic FlowAAHT Annual Average Hourly Traffic FlowAAWDT Annual Average Weekday Daily Traffic FlowARCADY Computer program for the design of RoundaboutsATC Automatic Traffic CountATCS Automatic Traffic Classifier Site

BCR Benefit to Cost Ratio

CBD Central Business DistrictCOBA COst Benefit Analysis computer programCPI Consumer Price IndexCWK Car in Work Time

DF Daily FactorDfT Department for TransportDMRB Design Manual for Roads and BridgesDTLR Department for Transport, Local Government and the Regions

FYRR First Year Rate of Return

GDP Gross Domestic Product

LGV Light Goods Vehicle

MCC Manual Classified Count

NPV Net Present ValueNRTF National Road Traffic ForecastsNTS National Travel Survey

OGV Other Goods VehicleOGV1 Other Goods Vehicle - Category 1OGV2 Other Goods Vehicle - Category 2OSCADY Computer program for the design of Traffic Signals

PHV Percentage Heavy Vehicles (OGV1 + OGV2 + PSV/COACH)PIA Personal Injury AccidentPICADY Computer program for the design of Priority JunctionsPSV Passenger Service VehiclePVB Present Value of BenefitsPVC Present Value of Costs

QUADRO QUeues And Delays at ROadworks computer program

RSI Roadside Interview

SACTRA Standing Advisory Committee on Trunk Road Assessment

Chapter 2 Volume 15 Section 1Abbreviations Part 11 Index and Abbreviations


SI Seasonality IndexSTAG Scottish Transport Appraisal GuidanceSYRR Single Year Rate of Return

TAG Transport Appraisal GuidanceTEE Transport Economic EfficiencyTEN Transport Economics NoteTRANSYT Computer program for the Optimisation of Linked Traffic SignalsTRL Transport Research Laboratory (formerly TRRL)TRRL Transport and Road Research LaboratoryTUBA Transport User Benefits Appraisal

VAT Value Added TaxVOC Vehicle Operating CostsVOT Value of Time

dmrb volume 15 - sias.comsias.com/2013/ts/the nesa manual - october 2015.pdf · urban roads...

Documents