PRELIMINARY ENVIRONMENTAL INFORMATION REPORT...by Cambridge Environmental Research Consultants (CERC, 2016), which developed the ADMS models. For the scenarios with the DCO Project,

Heathrow Expansion PRELIMINARY ENVIRONMENTAL INFORMATION REPORT

© Heathrow Airport Limited 2019

Volume 3, Chapter 7: Air quality and odour

Appendices

Appendix 7.1 Technical appendix on dispersion modelling

Heathrow Expansion PRELIMINARY ENVIRONMENTAL INFORMATION REPORT: Chapter 7: Appendix 7.1

Appendix 7.1 © Heathrow Airport Limited 2019

Appendix 7.1 Technical appendix on dispersion modelling



Contents

1. Introduction 1

2. Meteorological parameters 2

3. Aircraft emissions 3

3.1 General procedure 3

3.2 The dispersion model 4

3.3 Emissions sources: Aircraft emissions 5 Modes of runway operation 5 Aircraft activity 7 TAAM modelling 12 Reduced engine taxi 12 Main engine emissions: Emission rates 13 Main engine emissions: Future emission factors 17 Main engine emissions: Times in mode 17 Main engine emissions: Thrust settings 21 Auxiliary power units (APU) emissions 22 Brake and tyre wear emissions 24 Aircraft emissions: Spatial disaggregation 24 Aircraft emissions: Runway assignments and temporal variation 29

3.4 Emissions sources: On-airport, non-aircraft emissions 30 Ground support equipment (GSE) 30 Fire training 32

4. Approach to modelling emissions from road traffic 33

4.1 Model domain 33

4.2 Traffic data 34

4.3 Calculating road traffic emissions 35

4.4 Sensitivity test 36

4.5 Additional features 36

5. Approach to modelling car park emissions 38

6. Approach to modelling Lakeside Waste Management Facility emissions 39

7. Approach to predicting background concentrations 40

7.2 Background NO2 and NOX concentrations for sensitivity test 42

7.3 Modelling the background concentration field for traffic emissions 42



8. Calculation of total NO2 concentrations 46

8.2 EU limit value compliance 47

9. Model verification methodology 48

9.2 NOX and NO2 48

9.3 PM10 and PM2.5 57

10. Effect on EU limit value compliance methodology 63

11. CURED sensitivity test results 65

11.1 Community area results 65

11.2 Summary results across the CAQOAA 73

11.3 Effect on EU limit value compliance results 75

12. Bibliography 80

TABLE OF TABLES

Table 3.1: Movements per year by aircraft type: baseline scenarios 8 Table 3.2: Movements per year by aircraft type: ‘Future Baseline’ scenarios 9 Table 3.3: Movements per year by aircraft type: ‘DCO Project’ scenarios 10 Table 3.4: Aircraft data 14 Table 3.5: Approach and landing roll times by aircraft type 18 Table 3.6: Take-off roll and climb times 20 Table 3.7: Take-off thrusts for future scenarios 22 Table 3.8: APU data 22 Table 6.1: Emission data for the existing facility 39 Table 7.1: Sectors included in the background maps 41 Table 9.1: Sites used in the model verification 48 Table 9.2: Calculated factors for each year 49 Table 9.3: Statistical performance of each of the models 50 Table 11.1: CURED sensitivity test NO2 results at selected representative receptors within each community area 66 Table 11.2: Count of properties by impact descriptor 73 Table 11.3: Count of properties by impact descriptor (CURED sensitivity test) 73 Table 11.4: Count of properties by magnitude of change in concentration 74 Table 11.5: Count of properties by magnitude of change in concentration (CURED sensitivity test)74 Table 11.6: Local air quality receptors informing DCO Project significance 74 Table 11.7: Local air quality receptors informing DCO Project significance (CURED sensitivity test)75 Table 11.8: NO2 annual mean EU limit value compliance in 2022 76 Table 11.9: NO2 annual mean EU limit value compliance in 2027 77 Table 11.10: NO2 annual mean EU limit value compliance in 2030 78



TABLE OF GRAPHICS

Graphic 3.1: Approach NOX emissions density plot 25 Graphic 3.2: Landing roll NOX emissions density plot 26 Graphic 3.3: Taxi-in NOX emissions density plot 26 Graphic 3.4: Taxi-out NOX emissions density plot 27 Graphic 3.5: Take-off roll NOX emissions density plot 27 Graphic 3.6: Initial climb NOX emissions density plot 28 Graphic 3.7: APU NOX emissions density plot 28 Graphic 3.8: Brake PM10 emissions density plot 29 Graphic 3.9: Tyre PM10 emissions density plot 29 Graphic 3.10: GSE NOX emissions density plot 31 Graphic 4.1: 1km x 1km receptor cell and corresponding explicitly modelled roads 34 Graphic 7.1: Air quality core assessment area and HHASAM Fully Modelled Area 44 Graphic 7.2: Receptor grid example 45 Graphic 9.1: 2015 Model performance – simple verification approach 50 Graphic 9.2: 2015 Model performance – LOOCV verification approach 51 Graphic 9.3: 2016 Model performance – simple verification approach 51 Graphic 9.4: 2016 model performance – LOOCV verification approach 52 Graphic 9.5: 2017 Model performance – simple verification approach 52 Graphic 9.6: 2017 Model performance – LOOCV verification approach 53 Graphic 9.7: LOOCV primary verification factors 54 Graphic 9.8: Modelled vs measured NO2 at diffusion tube sites close to busy junctions 55 Graphic 9.9: 2015 Modelled vs measured NO2 at diffusion tube sites 56 Graphic 9.10: 2016 Modelled vs measured NO2 at diffusion tube sites 56 Graphic 9.11: 2017 Modelled vs measured NO2 at diffusion tube sites 57 Graphic 9.12: 2015 model performance – PM10 58 Graphic 9.13: 2015 model performance – PM2.5 58 Graphic 9.14: 2016 model performance – PM10 59 Graphic 9.15: 2016 model performance – PM2.5 59 Graphic 9.16: 2017 model performance – PM10 60 Graphic 9.17: 2017 model performance – PM2.5 60


Appendix 7.1-1 © Heathrow Airport Limited 2019

1. INTRODUCTION

1.1.1 This Appendix describes in detail the methodology applied in the dispersion

modelling carried out, as discussed in Chapter 7: Air quality and odour. The

assessment has focussed on the prediction of total pollutant concentrations under

a variety of scenarios, both with and without the DCO Project, at receptors

throughout the illustrative ‘Core Air Quality Objective Assessment Area’

(CAQOAA), a defined 12 x 11km primarily rectangular area centred on the Airport.

The CAQOAA has been extended slightly to the east to enable the prediction of

concentrations at all residential properties within the Heston Community Area (to

inform the Health Impact Assessment work) and is shown in Graphic 7.1.

Dispersion modelling has been used to predict pollutant concentrations in Windsor

and Old Windsor, as the tiered assessment using traffic data outside of the

CAQOAA indicated that impacts in this area may be greater than negligible.

1.1.2 In order to determine total pollutant concentrations, a number of individual

contributions have been determined, and combined. The assessment has

combined the contributions of several sources. The approach taken to determining

these contributions is set out in detail for each source in this Appendix. The

sources are:

1. The emissions from aircraft and all airside sources (including airside vehicles,

machinery and energy plant), modelled using the ADMS-Airport model

2. Road traffic emissions modelled using models from the ADMS-Urban suite

3. Emissions from vehicles using car parks at Heathrow Airport, including

additional cold start emissions

4. The contribution of the Lakeside Waste Management Facility in baseline and

future baseline scenarios

5. The contribution from all other sources (i.e. background concentrations).



2. METEOROLOGICAL PARAMETERS

2.1.1 Meteorological data from the monitoring station at Heathrow has been purchased

from the Met Office for use in the assessment. The surface roughness for the area

has been set in the model at 0.5m, with the minimum Monin-Obukhov length set at

30m. Given that the meteorological monitoring station is located in the centre of

the model domain, in a setting similar to much of the study area, these parameters

have been applied for both the dispersion site and meteorological measurement

site.

2.1.2 Wherever possible, the urban canopy flow module has been utilised to better

represent the effects of buildings on the flow of air throughout the model domain.

This module cannot be used when modelling certain sources (e.g. jet sources), but

has been used when modelling road sources, car park sources, the Lakeside

Waste Management Facility (in the baseline and future baseline scenarios) and

some airside sources. For baseline and future baseline scenarios, the input data

used for the urban canopy flow module was the 1km resolution dataset published

by Cambridge Environmental Research Consultants (CERC, 2016), which

developed the ADMS models. For the scenarios with the DCO Project, this input

file has been edited to reflect the changes to building coverage, average building

height and cross-wind vertical area of buildings associated with the DCO Project.



3. AIRCRAFT EMISSIONS

3.1 General procedure

3.1.1 There are two principal sets of recommendations for carrying out an airport air

quality study. The first arises from the Project for the Sustainable Development of

Heathrow (PSDH) (DfT, 2006). The objective of PSDH was to develop the best

practical methodology for assessing the air quality impacts of a third runway at

Heathrow. This came up with a number of specific recommendations but contains

some areas where the best approach depends on data availability. For example,

PSDH does not make any recommendations about how to determine how long

aircraft spend operating in various modes as there are various potential data

sources, and it is left to the analyst to use their judgement as to the best way of

extracting suitable operating durations.

3.1.2 The PSDH methodology was implemented by Heathrow Airport for its 2008/9

emissions inventory (Underwood et al, 2010a), modelling study (Underwood et al,

2010b) and model evaluation study (Underwood et al, 2010c). The reports give a

detailed description of the methodology used and forms a useful reference. The

model evaluation found that modelled concentrations generally agreed well with

the extensive monitoring data around Heathrow and formed a suitable basis for

evaluating the impacts of future airport developments. Subsequent inventories

produced for Heathrow have used essentially the same methodology, with some

updates where new airport-specific data has become available (e.g. for taxiing

times).

3.1.3 The second methodology was published by the International Civil Aviation

Organization (ICAO) in 2011. This document considers production of emission

inventories for historic years, with very little attention paid to how inventories for

future years might be produced. As such it is less directly relevant to the present

work for the DCO Project.

3.1.4 The ICAO methodology offers different levels of assessment, described as

‘simple’, ‘advanced’ and ‘sophisticated’, each requiring increasingly detailed data.

The sophisticated approach generally requires detailed data on times, engine

settings and so forth for each individual aircraft movement, so it is unsuitable for

modelling future scenarios where assumptions on future operation must be made.

The advanced approach is similar to the PSDH recommendations in terms of data

requirements and can generally be adapted to future scenarios given suitable

forecast data. Much of the detail of the methodology is the same or similar

between PSDH and ICAO.



3.1.5 A third ‘standard’ is the Aviation Environmental Design Tool (AEDT), published by

the US Federal Aviation Administration (FAA) for airport air quality inventories and

noise studies. Detailed documentation of the methodology used by the tool is not

readily available.

3.1.6 While various research groups have suggested ways in which parts of the

inventory calculation can be improved, few of these have been generally

incorporated into received methodologies. One notable exception is the FOA 3a

method (Kinsey and Wayson, 2009) for calculating PM10 emissions from smoke

number emissions.

3.1.7 Defra issues technical guidance on Local Air Quality Management (LAQM) (Defra,

2016), which is an important source of guidance on approaching common sources

of air pollution. However, other than providing a screening threshold of 10 million

passengers per annum or 1 million tonnes of freight, it does not provide

recommendations on the technical issues of modelling air quality around large

airports.

3.1.8 The methodology used in this assessment follows the general approach of the

ICAO advanced and PSDH approaches, and implements many of their specific

recommendations, with decisions about the best approach being led by the

availability of data.

3.2 The dispersion model

3.2.1 The PSDH carried out a model inter-comparison study to compare the use of

various dispersion modelling tools for airport air quality modelling. As a result, the

PSDH endorsed the use of ADMS-Airport, a version of the long-established

dispersion modelling tool ADMS adapted to account for the momentum and

buoyancy fluxes from jet engines. However, the use of the regular version of

ADMS with suitable initial dispersion characteristics was also found to be

acceptable.

3.2.2 AEDT uses AERMOD for the dispersion modelling. AERMOD was developed in

the United States by the American Meteorological Society (AMS)/United States

Environmental Protection Agency (USEPA) Regulatory Model Improvement

Committee (AERMIC). ADMS was developed in the UK by CERC in collaboration

with the Meteorological Office, National Power and the University of Surrey. Both

AERMOD and ADMS are termed ‘new generation’ models, parameterising stability

and turbulence in the planetary boundary layer by the Monin-Obukhov length and

the boundary layer depth. This approach allows the vertical structure of the

planetary boundary layer to be more accurately defined than by the stability

classification methods of earlier dispersion models such as the R91 Gaussian

plume model or Industrial Source Complex (ISC).



3.2.3 Numerous model inter-comparison studies have demonstrated little difference

between the output of ADMS and AERMOD, except in certain complex terrain

scenarios. The principal difference between ADMS and ADMS-Airport is the jet

engine module, which tends to reduce modelled ground-level concentrations from

aircraft engines, especially at high thrust settings, as a result of the heat of the

plume.

3.2.4 Taking the jet engine module into consideration, ADMS-Airport (Version 4.1.1) has

been selected as the most appropriate model to use for the purposes of this

particular study.

3.3 Emissions sources: Aircraft emissions

Modes of runway operation

3.3.1 Each of the runways at the Airport can be used in two directions, with aircraft

moving along it either eastwards or westwards. This means there are two distinct

and independent aspects to the way that the runways are used for departures,

landings or both (‘mixed’):

1. Whether aircraft take off and land facing in a westerly direction or in an easterly

direction

2. Which physical runways are used for departures and which for landings.

3.3.2 Whether the Airport operates in westerly or easterly mode at any given time

depends on the weather. It is safer and more efficient for aircraft to take off and

land facing into the wind, although at Heathrow there is a preference to use

westerly operations as long as the tailwind is only slight. Since the wind direction

also affects the dispersion of pollutants, it is essential to ensure that runway

assignments are aligned with the met data used for the dispersion modelling. The

Airport changes between easterly and westerly operations at unpredictable times,

since it depends on the weather. At Heathrow, westerly operations are more

common than easterlies, by a ratio of about 70:30.

3.3.3 The choice of which runway is used for departures and which for landings is called

the runway operating mode. These modes change regularly, in what is called a

runway alternation pattern, in order to provide noise respite to residents near the

Airport. There are three alternation patterns relevant to this assessment, as

follows.

3.3.4 The alternation pattern currently operated and used for the assessment of

baseline years (2015–2017), is as follows. In westerly operations (i.e. when aircraft

are landing and departing facing in a westerly direction), there are two modes of

operation, identified by a two-letter abbreviation:



1. DL: Departures on the northern runway (27R), landings on the southern

runway (27L)

2. LD: Landings on the northern runway (27R), departures on the southern

runway (27L).

3.3.5 The Airport alternates between these two modes at regularly scheduled times, so

that they are used equally often. The result is that over the course of a year, there

is an equal number of departures on each runway, and an equal number of

landings, for westerly operations.

3.3.6 In easterly operations, there is a single mode of operation: Landings on the

northern runway (09L), departures on the southern runway (09R) (Some landings

also take place on the southern runway early in the morning when there are a

large number of arrivals).

3.3.7 A new alternation pattern will be introduced for 2R operations for all future years.

In this pattern, westerly operations alternate as at present, but in easterly

operations a pattern of alternation similar to that for westerlies will be introduced,

alternating equally between two modes of operation:

1. DL: Departures on the northern runway (09L), landings on the southern runway

(09R)

2. LD: Landings on the northern runway (09L), departures on the southern

runway (09R).

3.3.8 The effect of this is that whether operating in easterlies or westerlies, each runway

will be used for an equal number of departures and an equal number of landings.

3.3.9 Finally, 3R operations will require a new alternation pattern. This will have four

modes of operation, regardless of westerly or easterly operation. These are

identified by a three-letter abbreviation:

1. MDL: Mixed (both landings and departures) on the (new) northern runway

(09L/27R), departures on the centre (currently northern) runway (09C/27C),

and landings on the southern runway (09R/27L)

2. MLD: Mixed on the northern runway, landings on the centre runway, and

departures on the southern runway

3. DLM: Departures on the northern runway, landings on the centre runway, and

mixed on the southern runway

4. LDM: Landings on the northern runway, departures on the centre runway, and

mixed on the southern runway.



3.3.10 The alternation pattern will use each of these four modes equally. This means that

each of the runways will have equal number of departures and landings over the

course of four weeks. However, when any given mode is in operation, each

landing or departure will have two possible runways available, and this choice will

be based on operational parameters such as aircraft size, stand and (for

departures) the airspace routing. This means that the three physical runways will

not necessarily have the same number of movements.

3.3.11 For future scenarios, it is not possible to know which mode will be in operation

when. Therefore, for each scenario, each relevant mode has been modelled for

the full assessment year and the average of the two or four modes taken. This

provides the best estimate of the long-term average emissions.

Aircraft activity

3.3.12 For the baseline modelling, aircraft activity has been taken from movement

records extracted by Heathrow from their Business Objective Search System

(BOSS) database. For each movement in the year, this provides:

1. Hour of the year

2. Whether arrival or departure

3. Aircraft type

4. Stand

5. Runway (including direction).

3.3.13 For the future scenarios, the details of the aircraft movements are taken from the

forecast schedule. This gives a list of movements for each cargo and passenger

aircraft type for a typical busy summer day, for each of the modelled scenarios

(Future Baseline and DCO Project, for each assessment year). For each

movement in the day, this provides:

1. Hour of the day

2. Whether arrival or departure

3. Aircraft type

4. Terminal

5. Preferred physical runway (separate runway assignments were given for each

of the four possible alternation modes, DLM, LDM, MDL and MLD).

3.3.14 For the future scenarios, the same fleet is assumed to operate every day of the

year, but each movement is adjusted by a factor so that the total number of

movements in the year agrees with the number forecast for that year. For



example, in the 2035 3R scenario, there are 2,153 movements in the busy day

schedule, and the total number of movements over the year is 740,000, so each

movement is adjusted by a factor of 740,000 / (2,153 × 365) = 0.942.

3.3.15 It is assumed that there are the same number of movements every day of the

year. Historic data shows that there are slightly more movements per day in

summer than in winter, but the difference is small, and sensitivity modelling

undertaken as part of previous air quality modelling work for Heathrow has shown

this assumption to be conservative.

3.3.16 These movements are summarised in Table 3.1 to Table 3.3. In these tables,

aircraft types that are not currently on the drawing board and are not expected to

enter into service before 2035 are described in terms of their current analogues

with a ‘G+2’ identifier; more details of this and the Generation column for the future

fleets are given in paragraph 3.3.31.

Table 3.1: Movements per year by aircraft type: baseline scenarios

Aircraft

code

Aircraft description 2015 2016 2017

318 Airbus A318 750 862 1,424

319 Airbus A319ceo 83,532 80,282 79,871

320 Airbus A320ceo 138,680 114,518 110,920

32A Airbus A320neo 2,301 25,785 30,421

321 Airbus A321ceo 42,768 40,496 36,386

32B Airbus A321neo 255 2,544 2,159

332 Airbus A330-200 6,403 7,085 4,998

333 Airbus A330-300 11,447 11,582 10,349

343 Airbus A340-300 1,982 2,224 1,142

346 Airbus A340-600 6,818 5,418 4,577

359 Airbus A350-900 58 714 2,809

388 Airbus A380-800 14,784 18,265 18,483

734 Boeing 737-400 1,534 1,390 702

736 Boeing 737-600 2,418 1,988 866

73G Boeing 737-700 4,903 7,205 4,794

738 Boeing 737-800 8,035 7,551 8,026

7M8 Boeing 737 MAX 8 0 0 102

73J Boeing 737-900 54 412 656

744 Boeing 747-400 24,900 19,967 19,650

74Y Boeing 747-400 Freighter 560 520 458

74H Boeing 747-8 10 7 242

752 Boeing 757-200 4,654 3,586 3,465

763 Boeing 767-300 25,296 22,961 23,092

764 Boeing 767-400 2,942 2,976 647

772 Boeing 777-200 31,346 29,718 28,177

77L Boeing 777-200LR 778 1,386 494



Aircraft

code

Aircraft description 2015 2016 2017

77X Boeing 777-200 Freighter 200 206 326

773 Boeing 777-300 392 0 4

77W Boeing 777-300ER 29,867 29,931 32,304

788 Boeing 787-8 12,148 13,992 16,050

789 Boeing 787-9 3,453 13,599 20,435

AB4 Airbus A300 1,013 0 0

ABY Airbus A600-600 Freighter 126 1,581 1,850

AR1 Avro RJ100 690 496 56

CR9 Canadair Regional Jet 900 710 546 180

CS1 Airbus A220-100 0 6 1,258

DH4 Bombardier Dash 8 400Q 42 50 3,442

E90 Embraer E190 1,892 1,454 2,284

E95 Embraer E195 462 670 332

F70 Fokker 70 2,174 1,646 356

Other Other 3,717 1,359 2,128

Total 474,094 474,978 475,915

Table 3.2: Movements per year by aircraft type: ‘Future Baseline’ scenarios

Aircraft

code

Aircraft description Genera

tion

2R 2022

480k

2R 2027

480k

2R 2030

480k

2R 2035

480k

318 Airbus A318 G0 689 692 0 0

319 Airbus A319ceo G0 67,145 25,263 10,382 0

319N Airbus A319neo G+1 0 0 2,076 2,076

320 Airbus A320ceo G0 101,234 90,324 44,989 14,881

32H Airbus A320ceo G0 15,495 15,573 15,573 8,306

320N Airbus A320neo G+1 28,235 82,019 149,849 197,606

320X Airbus A320 G+2 G+2 0 0 0 0

321 Airbus A321ceo G0 40,287 39,798 20,072 3,461

321N Airbus A321neo G+1 14,806 15,573 36,683 53,295

32B Airbus A321neo G0 3,443 3,461 2,076 2,076

321X Airbus A321 G+2 G+2 0 0 0 0

332 Airbus A330-200 G0 6,198 4,845 2,076 692

333 Airbus A330-300 G0 14,118 12,112 7,267 1,384

339 Airbus A330-900 G+1 0 2,769 6,229 6,229

343 Airbus A340-300 G0 0 0 0 0

346 Airbus A340-600 G0 4,132 0 0 0

359 Airbus A350-900 G+1 3,443 9,690 9,690 13,151

359N Airbus A350-900 G+2 G+2 0 0 0 0

351 Airbus A350-1000 G+1 5,854 21,110 27,686 28,378

351N Airbus A350-1000 G+2 G+2 0 0 0 0

388 Airbus A380-800 G0 15,151 15,227 15,227 12,459

73W Boeing 737-700 G0 3,443 3,461 0 0

7M7 Boeing 737 MAX 7 G+1 0 0 4,845 4,845



Aircraft

code

Aircraft description Genera

tion

2R 2022

480k

2R 2027

480k

2R 2030

480k

2R 2035

480k

738 Boeing 737-800 G0 11,019 9,690 1,384 0

73H Boeing 737-800 G0 2,755 2,769 0 0

7M8 Boeing 737 MAX 8 G+1 689 2,076 4,845 5,537

7X8 Boeing 737 MAX 8 G+2 G+2 0 0 0 0

73J Boeing 737-900 G0 689 692 0 0

7M9 Boeing 737 MAX 9 G+1 689 1,384 1,384 1,384

7X9 Boeing 737 MAX 9 G+2 G+2 0 0 0 0

744 Boeing 747-400 G0 11,707 0 0 0

74Y Boeing 747-400

Freighter

G0 689 346 346 346

74H Boeing 747-8 G0 689 692 692 692

763 Boeing 767-300 G0 0 0 0 0

76W Boeing 767-300 G0 3,443 0 0 0

772 Boeing 777-200 G0 26,514 7,267 0 0

77X Boeing 777-200

Freighter

G0 689 0 0 0

773 Boeing 777-300 G0 9,641 9,690 8,998 6,921

77W Boeing 777-300ER G0 25,825 25,955 25,263 16,265

779 Boeing 777-9 G+1 0 0 2,076 11,766

779N Boeing 777-9 G+2 G+2 0 0 0 0

788 Boeing 787-8 G+1 26,514 26,647 26,647 27,340

788N Boeing 787-8 G+2 G+2 0 0 0 0

789 Boeing 787-9 G+1 24,792 42,221 46,373 53,641

789N Boeing 787-9 G+2 G+2 0 0 0 0

781 Boeing 787-10 G+1 2,410 2,422 2,422 2,422

781N Boeing 787-10 G+2 G+2 0 0 0 0

ABY Airbus A600-600

Freighter

G0 1,377 0 0 0

DH4 Bombardier Dash 8

400Q

G0 4,821 0 0 0

E90 Embraer E190 G0 1,377 1,384 0 0

E95 Embraer E195 G0 0 4,845 4,845 0

E95-2 Embraer E195 G+1 G+1 0 0 0 4,845

Total 480,000 480,000 480,000 480,000

Table 3.3: Movements per year by aircraft type: ‘DCO Project’ scenarios

Aircraft

code

Aircraft description Generation 2R 2022

485K

3R 2027

567K

3R 2030

665K

3R 2035

740K

318 Airbus A318 G0 681 690 0 0

319 Airbus A319ceo G0 66,415 25,883 11,037 0

319N Airbus A319neo G+1 0 0 2,070 3,448

320 Airbus A320ceo G0 100,133 90,071 46,219 16,207

32H Airbus A320ceo G0 15,327 15,530 15,521 8,276

320N Airbus A320neo G+1 32,697 139,766 248,685 326,207



Aircraft

code


485K

3R 2027

567K

3R 2030

665K

3R 2035

740K

320X Airbus A320 G+2 G+2 0 0 0 0

321 Airbus A321ceo G0 39,849 39,687 20,695 4,138

321N Airbus A321neo G+1 14,645 16,220 38,631 60,000

32B Airbus A321neo G0 3,406 3,451 2,070 2,069

321X Airbus A321 G+2 G+2 0 0 0 0

332 Airbus A330-200 G0 7,493 6,212 4,139 2,069

333 Airbus A330-300 G0 14,645 14,149 11,382 5,172

339 Airbus A330-900 G+1 0 2,761 6,209 6,897

343 Airbus A340-300 G0 681 0 0 0

346 Airbus A340-600 G0 4,087 0 0 0

359 Airbus A350-900 G+1 3,406 11,733 15,866 26,207

359N Airbus A350-900 G+2 G+2 0 0 0 0

351 Airbus A350-1000 G+1 5,790 22,432 35,871 42,759

351N Airbus A350-1000

G+2

G+2 0 0 0 0

388 Airbus A380-800 G0 14,986 15,875 17,246 15,862

73W Boeing 737-700 G0 3,406 3,451 0 0

7M7 Boeing 737 MAX 7 G+1 0 0 4,829 6,897

738 Boeing 737-800 G0 10,899 10,353 3,449 690

73H Boeing 737-800 G0 2,725 2,761 0 0

7M8 Boeing 737 MAX 8 G+1 681 2,071 8,968 13,103

7X8 Boeing 737 MAX 8

G+2

G+2 0 0 0 0

73J Boeing 737-900 G0 681 690 0 0

7M9 Boeing 737 MAX 9 G+1 681 1,380 2,070 2,069

7X9 Boeing 737 MAX 9

G+2

G+2 0 0 0 0

744 Boeing 747-400 G0 11,580 0 0 0

74Y Boeing 747-400

Freighter

G0 681 345 345 345

74H Boeing 747-8 G0 681 690 690 1,379

763 Boeing 767-300 G0 0 690 690 0

76W Boeing 767-300 G0 3,406 0 0 0

772 Boeing 777-200 G0 28,269 11,388 0 0

77X Boeing 777-200

Freighter

G0 681 0 0 0

773 Boeing 777-300 G0 9,537 11,043 11,727 8,966

77W Boeing 777-300ER G0 25,544 27,263 30,008 19,655

779 Boeing 777-9 G+1 0 0 2,759 14,483

779N Boeing 777-9 G+2 G+2 0 0 0 0

788 Boeing 787-8 G+1 26,225 28,298 33,112 36,207

788N Boeing 787-8 G+2 G+2 0 0 0 0

789 Boeing 787-9 G+1 25,204 52,110 79,676 102,414

789N Boeing 787-9 G+2 G+2 0 0 0 0

781 Boeing 787-10 G+1 2,384 2,416 4,139 5,517

781N Boeing 787-10 G+2 G+2 0 0 0 0



Aircraft

code


485K

3R 2027

567K

3R 2030

665K

3R 2035

740K

ABY Airbus A600-600

Freighter

G0 1,362 0 0 0

DH4 Bombardier Dash 8

400Q

G0 4,768 0 0 0

E90 Embraer E190 G0 1,362 1,380 0 0

E95 Embraer E195 G0 0 6,212 6,898 2,069

E95-2 Embraer E195 G+1 G+1 0 0 0 6,897

Total 485,000 567,000 665,000 740,000

TAAM modelling

3.3.17 Additional information for the future scenarios comes from Total Airspace and

Airport Modeler (TAAM) simulations. TAAM is a software tool which simulates the

movements of every aircraft on the ground and in the air over the course of the

day, taking account of interactions between aircraft (e.g. waiting for another

aircraft to clear a runway or taxiway). The primary purpose of the simulations is to

ensure that the airfield layout and schedule can function properly without

excessive delays.

3.3.18 The TAAM model takes the forecast schedule as input, along with a mode of

operation (easterly or westerly; MDL, MLD, DLM or LDM). Part of the TAAM output

is the time at which each movement passes various gates (exits runway, arrives at

stand, starts pushback, starts taxi, enters hold zone, starts take-off roll, etc.),

which allows various times in mode to be extracted. In addition, it assigns each

movement to a stand (based on the availability of suitably-sized stands at the

required terminal when the inbound flight arrives) and to a runway (which is

sometimes different from the preferred runway given in the forecast schedule) to

optimise taxiway and runway usage within the constraints of the operating mode.

3.3.19 NATS carried out TAAM simulations using the preliminary airfield layout and 740K

ATM schedule, for all four modes of westerly operation (MDL, MLD, DLM and

LDM), and for LDM in easterlies. For the remaining three easterly modes, NATS

provided data in a similar movement-by-movement format but based on judgement

rather than running the TAAM model. These have since been modelled in TAAM

and the new data will be used in the assessment for the ES.

Reduced engine taxi

3.3.20 Although traditional practice is to have all of an aircraft’s engines running during

taxi-out and taxi-in, it is increasingly common to use reduced engine taxi (RET) for

at least part of the taxi-out or taxi-in stages. In RET, one or more engines is

switched off for part of the taxi, the remaining engine or engines being sufficient to



propel the aircraft. RET reduces emissions and reduces fuel burn, but there are

various operational considerations which constrain its use.

3.3.21 During RET, it is normal to have the Auxiliary Power Unit (APU) operating, to

ensure that there is a redundant power source. Use of RET has been modelled by:

1. Reducing the number of engines operating during periods of RET, from 2 to 1

for twin-engine aircraft and from 4 to 2 for 4 engine aircraft

2. Assuming that the APU is operating at ‘normal running’ load during periods of

RET.

3.3.22 Heathrow Airport’s BOSS system records whether RET is used for departures. For

baseline scenarios, this record is used to determine whether a departure uses

RET. BOSS does not record whether RET is used for arrivals; since RET is used

more commonly for arrivals than departures, the assumption has been made that

RET is used for arrivals corresponding to departures that use RET.

3.3.23 In addition, a survey of airlines has been carried out to find which airlines use

RET, and on which aircraft types. The results of this survey have been used for

future scenarios; it is assumed that if an airline uses RET for a given aircraft type,

then all arrivals and departures of that airline and aircraft type use RET. Due to

their design, Boeing 787 aircraft do not use RET.

3.3.24 For movements that use RET, it is assumed that the final 2 minutes 30 seconds of

taxi-out before start of roll, and the first 2 minutes 30 seconds of taxi-in, operate

with all engines running (and the APU is not running during these periods).

Main engine emissions: Emission rates

3.3.25 For the baseline modelling, engine assignments were taken from the Heathrow

AUWR (All-Up Weight Return) database. These are collected by Heathrow for the

purpose of emissions charging. Engine assignments are provided in the form of

the engine UID, a unique identifier used in the ICAO databank of emissions

certification data; this allows each aircraft’s engines to be indexed directly in the

databank. Assignments from AUWR have been cross-checked against other

published sources (including BuchAir’s JP Airline Fleets product (Flightglobal,

2013)) and found to have good reliability.

3.3.26 Emission factors for jet engines are taken from the ICAO databank, version 24

(ICAO, 2018). The databank provides emission indices for nitrogen oxides (NOX),

carbon monoxide (CO) and hydrocarbons (HC), fuel flow rates and smoke

numbers; each of these is given at four power settings (100%, 85%, 30% and 7%

of rated thrust). Emission indices (in g of pollutant per kg of fuel burned) are

multiplied by fuel flow rates (in kg s−1) to obtain an emission factor in g s−1.



3.3.27 The ICAO databank gives smoke numbers which need to be converted to

emission indices. This is done using the FOA 3a method (Kinsey and Wayson,

2009), with the amendment that the factor of (1 – bypass ratio) in equation 7a is

only applied to mixed turbofan engines (Underwood et al, 2010a). For some

engines, smoke number data points at certain thrust settings are missing, so an

approach originally developed by Qinetiq for PSDH has been used in which factors

are applied to the maximum smoke number.

3.3.28 For turboprop engines, emission factors are taken from the internationally

recognised Swedish FOI (Swedish Defence Research Agency) database (Swedish

FOI, no date).

3.3.29 For the future scenarios, for those aircraft types whose engines have been

certified and are in the ICAO databank, it has been assumed that the same mix of

engines as in the current Heathrow fleet continues into the future. In other words,

for aircraft types that are currently in service, no improvement in engine emissions

is assumed. This is a conservative assumption, since there are likely to be

incremental improvements in combustor technology over the lifetimes of some of

these aircraft types.

3.3.30 The only turboprop aircraft movements in the forecast schedules relate to a small

number (about 1% of total movements) of Bombardier Dash 8 400Q aircraft in

2022 (only). For these aircraft the current engine assignments from AUWR have

been used.

3.3.31 For other aircraft types, namely the Boeing 777-9 and those which are not

currently on the drawing board (i.e. those described as G+2), a single engine was

assigned for each aircraft type, with emission indices and fuel flow rates based on

typical current engines and projected forward based on forecast technology

improvement rates (see paragraph 3.3.35).

3.3.32 The aircraft engine assignments are summarised in Table 3.4. The UID is the

engine identifier used in the ICAO emissions databank. MTOW is maximum take-

off weight, used in the calculation of brake and tyre wear.

Table 3.4: Aircraft data

Aircraft type Aircraft description MTOW (kg) Number of

engines

Most common UIDs

318 Airbus A318 61679 2 7CM048, 8CM060

319 Airbus A319 68323 2 3CM027, 3CM028,

3IA006, 8CM057

319N Airbus A319 NEO 75500 2 18PW119

320 Airbus A320 74717 2 3CM026, 1IA003,

8CM055




engines

Most common UIDs

320N Airbus A320 NEO 79000 2 17CM082, 18PW122

320X Airbus A320 Replacement 79000 2 NE33, NE34

321 Airbus A321 86243 2 3IA008, 3CM023,

1IA005

321N Airbus A321 NEO 97000 2 17CM083, 18PW126

321X Airbus A321 Replacement 97000 2 NE33, NE34

32B Airbus A321 (Sharklets) 89324 2 3IA008, 10IA017,

8CM054

32H Airbus A320 (Sharklets) 75157 2 8CM055, 10IA013,

1IA003, 17CM082

332 Airbus A330-200 234455 2 3RR030, 4GE080

333 Airbus A330-300 234332 2 3RR030, 4GE080,

4PW067

339 Airbus A330-900 NEO 251000 2 19RR098

33X Airbus A330-1000 NEO 251000 2 3RR030, 2RR023

343 Airbus A340-300 269162 2 2CM015, 7CM047,

1CM011

346 Airbus A340-600 370421 2 6RR041, 8RR045

351 Airbus A350-1000 311000 2 18RR080

351N Airbus A350-1000 Replacement 311000 2 NE31

359 Airbus A350-900 271754 2 14RR075

359N Airbus A350-900 Replacement 271754 2 NE31

388 Airbus A380-800 553367 4 9EA001, 8RR046

738 Boeing 737-800 Passenger 78411 2 8CM051, 8CM065,

11CM072

73H Boeing 737-800 (Winglets) 77453 2 11CM072, 8CM051,

8CM065

73J Boeing 737-900 (Winglets) 82966 2 11CM081, 8CM051

73W Boeing 737-700 (Winglets) 66905 2 3CM031, 8CM063,

3CM032

744 Boeing 747-400 384244 4 4RR036, 2GE045,

1PW055, 1RR010

74H Boeing 747-8 Passenger 447306 4 13GE157

74Y Boeing 747-400 Freighter 402578 4 1PW042, 3GE057,

12PW102, 1PW059

763 Boeing 767-300 181834 2 1RR011, 1GE029,

1GE030, 12PW102,

1PW043

76W Boeing 767-300 (Winglets) 183947 2 1GE029, 12PW101,

1PW043, 1GE030

772 Boeing 777-200 282026 2 10PW099, 2RR027,

5RR040

773 Boeing 777-300 299370 2 2RR027

779 Boeing 777-900 351534 2 NE30

779N Boeing 777-900 Replacement 351534 2 NE31

77W Boeing 777-300ER 347006 2 7GE099




engines

Most common UIDs

77X Boeing 777-200 Freighter 347626 2 7GE097

781 Boeing 787-10 254011 2 17GE177, 19RR096


788 Boeing 787-8 227930 2 13GE162, 13GE161,

11GE137


789 Boeing 787-9 251543 2 13GE160, 12RR067,

12RR068


7M7 Boeing 737 MAX 7 (Winglets) 80286 2 18CM087

7M8 Boeing 737 MAX 8 (Winglets) 81498 2 18CM086, 18CM087

7M9 Boeing 737 MAX 9 (Winglets) 88314 2 18CM085

7X8 Boeing 737 MAX 8 (Winglets)

Replacement

81498 2 NE33, NE34

7X9 Boeing 737 MAX 9 (Winglets)

Replacement

88314 2 NE33, NE34

ABY Airbus A300-600 Freighter 157296 2 1PW048

DH4 De Havilland DHC-8-400 29343 2 TUR096

E90 Embraer 190 47821 2 10GE129, 11GE142,

8GE115, 10GE130

E95 Embraer 195 50790 2 10GE130, 8GE115,

8GE117

E95-2 Embraer 195 E2 61500 2 18PW117

3.3.33 The PSDH recommended a procedure for taking into account changes in ambient

temperature, pressure and humidity on aircraft engine emissions, which it found

changed overall aircraft NOX emissions by about 2 or 3% (DfT, 2006). The PSDH

also recommended a methodology for take-off roll, accounting for non-uniform

acceleration, effects of the forward speed on the engine thrust, etc. It found that

these made a difference of between 2 and 7% on average to NOX emissions from

the take-off roll phase. Unfortunately, the engine-specific data that underlie these

methodologies were not published and remain proprietary; moreover, they do not

cover engines introduced since about 2007. Therefore, new factors were derived

using the same approach as used for the PSDH, and have been applied in the

same way as recommended by PSDH.

3.3.34 ICAO databank emission factors are based on new production engines, so in-

service engines are likely to have suffered deterioration which may affect their

emissions. PSDH recommended correction factors to account for this, namely a

4.3% increase in fuel flow and a 4.5% increase in NOX emission rate (the product

of emission index and fuel flow rate). PSDH did not have sufficient data to resolve

these factors into individual engine types, ages or thrust setting, so they have



been applied uniformly across the engine fleet for all phases of the Landing and

Take-Off (LTO) cycle.

Main engine emissions: Future emission factors

3.3.35 As noted in paragraph 3.3.31, emission factors for the Boeing 777-9 and those

which are not currently on the drawing board have been estimated based on

typical current engines and projected forward based on forecast technology

improvement rates. Improvements in emissions of NOx and PM have been

modelled, along with improvements in fuel consumption.

3.3.36 To develop these projections, interviews of industry experts and other secondary

research were carried out in order to identify the potential technologies that can be

adopted to reduce emissions. Potential technologies are identified for the two main

types of combustor design, namely the Lean combustor and the Rich Burn–Quick

Quench–Lean Burn (RQL) combustor. Examples include staged combustion,

advanced premixing nozzles and novel hot zone casing materials.

3.3.37 These are then correlated against historic improvements, both incremental

improvements within an engine generation and step changes between

generations, to estimate the typical emissions reductions due to each potential

technology. This takes into account the trend for increasing operational pressure

ratio. These potential improvements are then adjusted to take into account

technological and commercial feasibility, industry expectations and regulatory

trends.

Main engine emissions: Times in mode

3.3.38 Approach times for the baseline modelling were derived from data from the

Heathrow Noise and Track-Keeping (NTK) system for a sample of historic

movements. The NTK system provides radar squawks for arriving and departing

aircraft giving position, height and elapsed time, but only while aircraft are more

than a few hundred feet above the ground. Approach times for future cases were

derived from data supplied from the Heathrow Operational Planning and

Scheduling (OPAS) database for a sample of historic movements covering five

non-contiguous days in June and September. The OPAS system provides similar

data to NTK but with a higher time resolution and includes data for aircraft on the

ground.

3.3.39 Overall, OPAS data is preferred to NTK data, because of its higher precision and

its availability for aircraft on the ground, which provides greater consistency across

parts of the LTO cycle. However, preparing the OPAS data into a usable form

requires a significant amount of manual work, which is why data is only available

for a relatively small sample of time. The OPAS data was not available until the



modelling of the baseline scenarios was completed, which is why NTK data was

used for these scenarios. It is intended that modelling for the ES will use OPAS

data.

3.3.40 Approach times do not depend significantly on runway, so a single approach time

was used for all runways, for both historic and future 2R and 3R scenarios.

Approach is divided into two phases: Phase 1 is from 3000 feet (914 m) to

2000 feet (610 m) at a constant speed of 160 knots (82 m s−1), and Phase 2 is

from 2000 feet to the runway threshold, at uniform deceleration to landing speed

(which depends on aircraft type). The resulting approach times are summarised in

Table 3.5.

3.3.41 Landing roll times were derived from data supplied from the OPAS database for a

sample of historic movements. These have been extracted as a function of aircraft

type (Table 3.5) and also by wake vortex category. The former is used where the

aircraft type is in the sample data and the latter where the aircraft type is not in the

sample data. Landing roll times are assumed not to depend on runway or on

runway exit taxiway.

Table 3.5: Approach and landing roll times by aircraft type

Aircraft type Approach Phase 1 time

(s)

Approach Phase 2 time

(s)

Landing roll time (s)

319 78 166 54

320 78 160 54

321 78 156 55

32A 78 161 54

32B 81 157 52

32N 76 166 53

32Q 90 167 56

332 79 158 63

333 77 157 64

346 75 151 59

359 96 160 60

388 83 161 65

738 73 157 51

73H 84 153 53

73W 77 159 58

744 74 150 62

75W 75 168 57

764 73 154 55

76W 73 155 58

772 72 157 63

77W 79 154 60

788 79 156 62



Aircraft type Approach Phase 1 time

(s)

Approach Phase 2 time

(s)

Landing roll time (s)

789 76 152 62

7M8 72 154 53

ABY 100 170 60

CS3 93 164 54

DH4 79 168 53

E90 87 165 48

3.3.42 Taxi-in and pushback/taxi-out/hold times for the baseline scenarios are derived

from Heathrow Electronic Flight Processing Strip (EFPS) data. For each

movement of the assessed years, this gives the times at which the aircraft passes

various gates (exits runway, arrives at stand, starts pushback, starts take-off roll).

For the baseline scenarios, it is possible to associate most movements with their

EFPS data and so obtain movement-specific times. Average times by terminal and

runway end are used for movements which cannot be matched with an EFPS

movement. Times are assumed to be independent of aircraft type. The distribution

along the taxi route (including time spent stationary during pushback and hold)

was derived from OPAS data for each taxi route.

3.3.43 Taxi-in and pushback/taxi-out/hold times for the future scenarios are derived from

TAAM simulation data supplied by NATS. For the 2035 DCO Project scenario, it

was possible to match TAAM output with individual movements in the schedule, so

times were applied on a movement-by-movement basis. For other 3R scenarios,

times are calculated from 2035 TAAM modelling as the average taxi time for the

terminal/runway end combination. For 2R scenarios in westerlies and in easterly

LD mode, times are calculated from 2R TAAM model outputs as the average taxi

time for the terminal/runway end combination. For 2R scenarios in easterly DL

mode, taxi times were taken from averages from the 3R TAAM simulations. These

times are assumed to be independent of aircraft type.

3.3.44 The distribution along the taxi route (including time spent stationary during

pushback and hold) was also derived from the TAAM data for each taxi route. For

the 2R configurations, although the airfield layout and activity levels will be similar

to the baseline scenarios, TAAM data is preferred over EFPS partly for

consistency with the DCO Project scenarios, and partly because the introduction

of full runway alternation in easterly operations introduces a new mode of

operation for which EFPS data is not available.

3.3.45 Take-off roll times were derived from the OPAS dataset. A distribution of take-off

times was derived for each aircraft, runway and runway access taxiway

combination. Fall-back distributions were derived for aircraft types which were not

present in sufficient numbers in the sample data, based on weight vortex category.

For future scenarios, times were also averaged across runways. Take-off times



are apportioned into five-second intervals, and each distribution consists of the

fraction of movements in the source data for which the take-off time falls within

each bin. Average take-off roll times are summarised in Table 3.6.

Table 3.6: Take-off roll and climb times

Aircraft type Take-off roll

time (s)

Initial climb to

1000 ft time (s)

Initial climb to

1500 ft time (s)

Climb-out

from 1000 ft

time

Climb-out

from 1500 ft

time

318 38 20 33 72 59

319 37 22 36 68 53

320 34 19 33 75 61

321 33 18 30 70 59

32A 34 19 32 72 59

32B 35 18 29 72 61

332 39 18 30 65 53

333 41 19 33 79 64

33X 33 18 32 74 60

343 46 34 54 96 76

346 40 34 60 95 69

359 36 18 33 69 54

388 46 39 63 92 68

734 32 38 47 43 34

736 39 21 36 70 55

738 40 20 34 61 47

73G 38 20 37 59 43

73H 37 20 33 61 48

73J 39 22 35 77 63

73W 39 22 35 63 49

744 41 34 55 64 43

74H 36 26 43 68 51

74N 35 33 54 64 43

74Y 37 18 28 50 39

752 37 17 28 50 39

75F 30 17 28 50 39

75W 36 16 27 49 39

763 37 22 36 62 48

76W 34 18 29 47 36

772 36 21 35 77 63

77L 38 18 28 48 37

77W 35 19 31 62 51

77X 34 17 27 45 35

788 38 24 37 62 48

789 39 26 40 72 58

AB6 41 14 23 38 30

ABY 27 14 22 38 29

CS3 36 19 30 60 49



Aircraft type Take-off roll

time (s)

Initial climb to

1000 ft time (s)

Initial climb to

1500 ft time (s)

Climb-out

from 1000 ft

time

Climb-out

from 1500 ft

time

DH4 32 20 36 67 51

E90 38 19 29 59 49

E95 35 19 29 59 48

3.3.46 Initial climb and climb-out times were derived from NTK data. These do not

depend significantly on runway, so a single time was used for all runways, for both

baseline and future baseline and with DCO Project scenarios. These times have

only a weak dependence on aircraft type (aircraft are designed to take off at

similar speeds to ensure that separation distances are maintained), but this was

retained since data were available (Table 3.6).

Main engine emissions: Thrust settings

3.3.47 Approach thrusts are assumed to be 15% of maximum rated thrust between

3000 feet and 2000 feet, and 30% from 2000 feet to touchdown. (Heights are

relative to runway level.) This is based on PSDH recommendations.

3.3.48 For taxi, PSDH recommendations are again followed for movements that do not

use RET. For these, the fuel flow rate is assumed to be 17.5% or 32.5% lower

than the fuel flow rate at 7% thrust, for non-Rolls Royce and Rolls Royce engines

respectively. The emission indices are set to those for the 7% thrust setting. For

aircraft movements that use RET, the fuel flow rate and emissions indices are

those for 7% thrust.

3.3.49 Aircraft sometimes use reverse thrust on landing, usually where the runway is

short and/or when weather conditions are poor (e.g. wet or icy). It has not been

possible to obtain robust quantitative data on reverse thrust usage at Heathrow.

Advice from the Airline Working Group was that use of reverse thrust above idle

was uncommon. Therefore, it has been assumed that all aircraft use a thrust

setting of 7%, corresponding to idle, during the landing roll.

3.3.50 It is common for aircraft to take-off at less than 100% thrust, sometimes as low as

75%, primarily to reduce wear on the engines. This is possible because engines

are overpowered for routine take-offs since aircraft need to be able to complete

the manoeuvre safely with the loss of one engine. For baseline modelling, take-off

thrust settings were based on survey data compiled before the 2008/9 Heathrow

inventory. For future scenarios, a simpler set of take-off thrust settings were

assumed (Table 3.7), which were intended to ensure the assumptions were

conservative.



Table 3.7: Take-off thrusts for future scenarios

Aircraft type Reduced thrust setting (%) Flights using 100% thrust (%)

Narrow-body, twin engine 80 6

Wide-body, twin engine 80 6

Wide-body, four engine 84 14

Auxiliary power units (APU) emissions

3.3.51 As well as their main engines, many aircraft have APUs, which are small gas

turbines used to generate electrical power for purposes such as starting the main

engines, powering air conditioning and other services.

3.3.52 Emission factors for NOX and Particulate Matter (PM) were taken from work

carried out for the PSDH, which provided representative emission factors for

various groups of aircraft types, including future types. Emission factors for fuel,

SO2 and HC are taken from the FAA Emissions and Dispersion Modeling System

(EDMS), the forerunner of AEDT.

Table 3.8: APU data

Aircraft type APU class for NOX APU class for PM APU model name

318 c A 36-300

319 c A 36-300

319N c A 36-300

320 c A 36-300

320N c A 36-300

320X c A 36-300

321 c A 36-300

321N c A 36-300

321X c A 36-300

32B c A 36-300

32H c A 36-300

332 e A 331-350

333 e A 331-350

339 e A 331-350

33X e A 331-350

343 e A 331-350

346 e A 331-350

351 f A 331-500

351N f A 331-500

359 f A 331-500

359N f A 331-500

388 g A A388

738 b A 131-9



Aircraft type APU class for NOX APU class for PM APU model name

73H b A 131-9

73J b A 131-9

73W b A 131-9

744 e A PW901A

74H e A PW901A

74Y e A PW901A

763 c C 331-200ER

76W c C 331-200ER

772 f A 331-500

773 f A 331-500

779 f A 331-500

779N f A 331-500

77W f A 331-500

77X f A 331-500

781 c C 331-200ER

781N c C 331-200ER

788 c C 331-200ER

788N c C 331-200ER

789 c C 331-200ER

789N c C 331-200ER

7M7 b A 131-9

7M8 b A 131-9

7M9 b A 131-9

7X8 b A 131-9

7X9 b A 131-9

ABY d C 331-200ER

DH4 a B 36-150[ ]

E90 a B 36-150[ ]

E95 a B 36-150[ ]

E95-2 a B 36-150[ ]

3.3.53 Running times for APUs on stand for the baseline modelling are derived from

monitoring undertaken to ensure compliance with Heathrow’s Operational Safety

Instructions (OSIs).

3.3.54 For future scenarios, times are taken from the OSIs. The OSIs specify the

following maximum periods for which APUs may be operated:

1. Narrow-bodied aircraft: 10 minutes after arrival on stand and 15 minutes prior

to departure

2. Wide-bodied aircraft: 10 minutes after arrival on stand and 50 minutes prior to

departure; and

3. A380: 15 minutes after arrival on stand and 90 minutes prior to departure.



Brake and tyre wear emissions

3.3.55 Emissions of PM from brake and tyre wear are calculated using the PSDH

methodology (ICAO omits this source). Brake wear emissions, in g PM10 per

arrival, are calculated as 2.53 × 10–4 × MTOW, where MTOW is the maximum

take-off weight in kg. Tyre wear emissions, in g PM10 per arrival, are calculated as

2.23 × 10–4 × MTOW – 8.74 for aircraft with an MTOW > 50,000 kg, and 2.41 ×

MTOW / 50,000 for smaller aircraft.

3.3.56 PM2.5 emissions are calculated by multiplying the PM10 emission by 0.4 for brake

wear and 0.7 for tyre wear.

3.3.57 For the baseline modelling, MTOW was taken from the AUWR database. For the

future scenarios, MTOW was taken from the AUWR database for the majority of

aircraft types. For G+2 aircraft types, MTOW has been assumed to be the same

as the analogous current aircraft type (with no account taken of likely reductions in

weight due to improvements in design and materials).

Aircraft emissions: Spatial disaggregation

3.3.58 Emissions from approach, landing roll, taxi-in, taxi-out, hold, take-off roll, initial

climb, climb-out and APU during taxi-in and taxi-out are modelled in ADMS-Airport

as jet sources, spread along a series of straight line segments. Landing roll

emissions are assumed to decelerate at a constant rate from 130 knots (67 m s−1)

at touch-down to 15 knots (8 m s−1) when exiting the runway. Take-off roll is

assumed to accelerate in accordance with a speed-emission curve, depending on

aircraft type, using parameters from PSDH.

3.3.59 Emissions from taxi-in and taxi-out are assigned to a set of straight-line segments

making up each of the taxi routes from stand group to runway. Aircraft do not

travel at uniform speed along the taxi routes; for example, during taxi-out there are

commonly delays for pushback and in the hold zone, as well as waiting for other

aircraft to push back and at taxiway crossings. To take this into account, data from

OPAS (for baseline modelling) and TAAM (for future scenarios) has been

analysed to determine average occupancy times for each segment of each taxi

route, and emissions are distributed along the taxi routes in proportion to the

occupancy times.

3.3.60 Emissions from APU usage on stand are modelled as volume sources, of

dimensions 50 m × 50 m horizontally and with vertical extent 12 m, centred on the

respective stands.

3.3.61 Emissions from tyre wear are modelled as volume sources, of length 300 m, width

50 m and vertical extent 15 m, centred on the touchdown point of the respective

runways.



3.3.62 Emissions from brake wear are modelled as volume sources, of width 50 m and

vertical extent 15 m, extending from the touchdown point to the most common exit

taxiway of the respective runways.

3.3.63 For baseline modelling, the stand is known for each movement. For future

scenarios, the forecast schedules provide terminal assignments for each

movement. For the DCO Project 2035 scenario, the TAAM modelling assigns each

movement to an individual stand, and this stand was used for the air quality

modelling. For other future scenarios, movements were assigned probabilistically

to each of the stands on the specified terminal of a suitable size for the aircraft

type.

3.3.64 Example plots of the emissions are shown in Graphic 3.1 to Graphic 3.9. These

show the emissions for the DCO Project 2035 scenario within 100 m squares

aligned with the National Grid. (Note that the runways at Heathrow are aligned

slightly off the National Grid east-west axis, giving rise to ‘steps’ in the plotted

runway emissions.)

Graphic 3.1: Approach NOX emissions density plot

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha



Graphic 3.2: Landing roll NOX emissions density plot

Graphic 3.3: Taxi-in NOX emissions density plot

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha



Graphic 3.4: Taxi-out NOX emissions density plot

Graphic 3.5: Take-off roll NOX emissions density plot

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha



Graphic 3.6: Initial climb NOX emissions density plot

Graphic 3.7: APU NOX emissions density plot

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha



Graphic 3.8: Brake PM10 emissions density plot

Graphic 3.9: Tyre PM10 emissions density plot

Aircraft emissions: Runway assignments and temporal variation

3.3.65 For modelling, each aircraft movement needs to be assigned to a runway. For

baseline modelling, the runway actually used is known, but for future scenarios,

runways are assigned probabilistically. These probabilities need to align with the

met data used for the dispersion modelling. Because meteorological data is only

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 5

5 to 8

8 to 12

12 to 20

20 to 30

30 to 50

50 to 80

80 to 120

120 to 200

200 to 300

300 to 500

500 or more

Emissionsin kg/y/ha

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 5

5 to 8

8 to 12

12 to 20

20 to 30

30 to 50

50 to 80

80 to 120

120 to 200

200 to 300

300 to 500

500 or more

Emissionsin kg/y/ha



available on an hourly basis, it is sufficient to determine runway probabilities for

each hour of the year.

3.3.66 The met year used for the dispersion modelling corresponds to a historic year for

which actual runway usage is available. This historic year is used to obtain the

relative frequency of easterly and westerly operations in each hour of the year.

3.3.67 The probability of using each physical runway in any given hour of the year is

determined by assuming that each of the four operational modes is equally likely

(or each of the two operational modes for 2R scenarios). For the DCO Project

2035 scenario, for each operational mode, the runway for each movement is

determined from TAAM data. For the other DCO Project scenarios, for each

operational mode, the runway for each movement is determined from the forecast

schedule. For 2R scenarios, for each operational mode, there is only one runway

available for each movement.

3.3.68 In addition, the number of aircraft movements varies with hour of the day and the

time of year (although at Heathrow there is little variation from month to month).

Since the weather also varies systematically between hours of the day, and

between seasons of the year, it is therefore desirable for the model to take this

temporal variation in emissions into account.

3.3.69 The hour of day is known for each movement, in the baseline, future baseline and

With DCO Project scenarios. Emissions were calculated for each hour of the year,

taking into account the movements in that hour and the weather conditions (which

affect emissions through temperature, pressure and humidity effects, as described

in paragraph 3.3.33, as well as the runway direction), for each mode of runway

operation. These were used to create an hour-by-hour time-varying emissions

weighting (‘hfc’) file for each emission source. The emissions were fed into the

dispersion model for each mode of runway operation. The resulting concentrations

were then averaged across mode of runway operation to obtain the final modelled

concentrations. This process was carried out for each of the three met years

2015–2017.

3.4 Emissions sources: On-airport, non-aircraft emissions

Ground support equipment (GSE)

3.4.1 GSE emissions are calculated using the equipment fleet mix from the 2017

Heathrow inventory update (Ricardo, 2018), which is assumed to represent the

mix of equipment types and ages in each future scenario. A bottom-up calculation

of fuel consumption is adjusted to fit actual measured fuel consumption, provided

by the Airside Operations team, and emissions of air pollutants are adjusted

accordingly.



3.4.2 Emission factors for road vehicles are taken from COPERT V (Emisia, no date),

which is the dataset used by Defra’s Emission Factors Toolkit version 8 (Defra,

2017a). Emission factors for non-road mobile machinery (NRMM) are assumed to

be equal to the limits in the EU directive on NRMM (European Parliament, 2004).

In each case, emission factors appropriate to the age of the vehicle, and therefore

the assumed emission control standard, are use, based on the age mix from the

2017 inventory.

3.4.3 Emissions are scaled to future scenarios in proportion to forecast aircraft

movements. Emissions are also adjusted to account for projected penetration of

electric equipment into the fleet, provided by the Airside Operations team. Electric

vehicles are assumed to have zero tail-pipe emissions, with only fugitive emissions

of PM10 and PM2.5 from brake and tyre wear.

3.4.4 Emissions are assumed to occur on aircraft stands, and are modelled as 50 m ×

50 m × 3 m volume sources. Total emissions are calculated and then distributed

between stands and between hours of the year in proportion to the sum of the

MTOWs of the aircraft using that stand during that hour.

3.4.5 An example plot of the emissions is shown in Graphic 3.10. This shows the

emissions per 100 m square for the DCO Project 2035 scenario.

Graphic 3.10: GSE NOX emissions density plot

174000

175000

176000

177000

178000

503000 504000 505000 506000 507000 508000 509000 510000

0 to 100

100 to 150

150 to 250

250 to 400

400 to 600

600 to 1000

1000 to 1500

1500 to 2500

2500 to 4000

4000 to 6000

6000 to 10000

10000 or more

Emissionsin kg/y/ha



Heating plant

3.4.6 Emissions from heating plant are included in all scenarios. Data is based on the

2008–9 inventory (Underwood et al, 2010a), but with updates including the closure

of the Thames Valley Power combined heat and power unit, the introduction of a

new energy centre with biomass-powered boilers; and associated changes to the

assumed usage of the Building 448, Terminal 5 and BA Cargo units.

Fire training

3.4.7 Emissions due to fire training were taken from the 2008–9 emissions inventory

(Underwood et al, 2010a). It is not expected that emissions from this source will

increase with the DCO Project. This source is in any case very small.



4. APPROACH TO MODELLING EMISSIONS FROM ROAD TRAFFIC

4.1 Model domain

4.1.1 A receptor file has been created that covers the CAQOAA, incorporating specific

sensitive receptors as well as a fine grid of receptors to enable contour plots to be

produced1. This receptor file has been subdivided into 1km x 1km squares.

4.1.2 When predicting concentrations at receptors, all roads within a 2km x 2km square

(which overlaps the 1km x 1km receptor square by 500 m in each cardinal

direction) have been modelled explicitly. Graphic 4.1 shows a 1km x 1km

selection of receptors and the 2km x 2km extent of the road traffic network

modelled explicitly for that selection of receptors.

4.1.3 In this way, the explicitly-modelled road network has been re-defined for each 1km

x 1km grid of receptors, but the number of roads modelled explicitly for each

receptor has been minimised, allowing significantly quicker model run times than if

all roads were modelled explicitly (which would have been prohibitively slow).

4.1.4 Emissions from road traffic outside of each 2km x 2km square have not been

ignored; their incorporation into background concentrations is described in

Section 7.

1 Receptors have been modelled over Cartesian grids at 10 m intervals within 100 m of modelled roads, at 20 m intervals between 100 m and 200 m from modelled roads, and at 50 m intervals beyond 200 m from modelled roads.



Graphic 4.1: 1km x 1km receptor cell and corresponding explicitly modelled roads

Contains Ordnance Survey data © Crown copyright and database right 2018. Ordnance Survey licence

number 100046099.

4.2 Traffic data

4.2.1 Traffic data for the assessment have been derived from the Heathrow Highway

Assignment and Surface Access Model (HHASAM), with the outputs processed to

give Annual Average Daily Traffic (AADT) flows for each link, along with the fleet

composition (proportion Cars, LGVs, HGVs and Buses/Coaches) and an average

speed.

4.2.2 HHASAM does not model motorcycle flows, therefore appropriate motorcycle

numbers have been manually added to the flows for each link using road type-

specific proportions provided by the surface access modelling team.

4.2.3 Where traffic will typically be free-flowing (i.e. away from junctions or other

features that will slow traffic on a specific short section of a road), the modelled

average speeds have been used to determine average emissions for sections of

road. Where there are junctions or other features that will slow traffic on a specific

short section of a road, speeds have been reduced using professional judgement



to account for increased emissions in these locations as a result of slow-moving

and/or queuing traffic. Speeds on the remainder of such links have not been

increased to account for this reduction on some sections, as the slow sections are

typically short and unlikely to lead to substantial changes in speed along the

remainder of the link, and, as a result, unlikely to significantly affect modelled

concentrations.

4.2.4 Diurnal traffic flow profiles have been provided by the surface access modelling

team for the DCO Project. Monthly flow profiles have been derived from the

national profiles published by DfT (DfT, 2017).

4.3 Calculating road traffic emissions

4.3.1 The Defra Emission Factor Toolkit (EFT) v8.0.1 has been used to calculate vehicle

emissions. This tool requires that the user enter one of seven Road Types:

1. Urban (not London)

2. Rural (not London)

3. Motorway (not London)

4. London Centre

5. London Inner

6. London Outer

7. London Motorway

4.3.2 It is important to note that these categories describe the vehicle fleet composition

rather than the precise physical location of the road. The fleet composition does

not, for example, change as a road passes from an urban to a rural area in the

absence of any intervening junctions. The Notes for Users section which

accompanies EFT V8.0.1 explains that:

‘The urban categorisation relates to the DfT definition of an urban area with a population of

10,000 or more.

The London road types are consistent with the area categories defined in the London

Atmospheric Emissions Inventory (LAEI). 'Central' corresponds to the Ultra Low Emission

Zone (ULEZ) area, whilst 'Motorway' denotes the M25 - other motorways in London should

be defined as 'London - Inner' or 'London - Outer' as appropriate’.

4.3.3 Rigidly applying these definitions would artificially introduce step changes in

vehicle fleet compositions part-way along links. It would also run the risk of

artificially changing the fleet composition on any links which are realigned as part

of the DCO Project, thus introducing error into the comparative analysis.



4.3.4 The following approach has, therefore, been followed when using the EFT to

calculate vehicle emissions:

1. All sections of the M25 (including all non-M4 slip-roads) have been assigned

Road Type 7

2. Sections of motorway and slip-roads (other than the M25) outside of the Outer

London area have been assigned Road Type 3

3. All other roads have been assigned Road Type 6.

4.3.5 Applying this approach ensures minimal step changes in emission factors, and

consistency throughout the CAQOAA, where emissions are likely to generally be

very similar to the Outer London fleet.

4.3.6 The changes to the Low Emission Zone (LEZ) and Ultra-Low Emission Zone

(ULEZ) announced by the Mayor of London in June 2018 will change the fleet

composition within London, especially within the north and south circular roads,

although these are outside of the CAQOAA. It has not been possible to take

account of these changes; as such, any predictions for years beyond 2020 that are

made using EFT v8.0.1 can be expected to over-estimate nitrogen dioxide

concentrations, and will thus be worst-case.

4.4 Sensitivity test

4.4.1 Air Quality Consultants Ltd. (AQC) has carried out a detailed, peer-recognised

analysis which showed that, whereas previous standards had had limited on-road

success in reducing nitrogen oxides emissions from diesel vehicles, the ‘Euro VI’

and ‘Euro 6’ standards are delivering real on-road improvements (AQC, 2016).

Defra’s EFT v8.0.1 takes account of these observed improvements, but also

makes additional assumptions regarding the performance of diesel cars and vans

that will be produced in the future. In particular, it assumes that diesel cars and

vans registered for type approval after 2020 will, on average, emit significantly less

NOX than earlier models. A sensitivity test has been carried out using AQC’s

CURED v3A model (AQC, 2017), which assumes that this post-2020 technology

does not deliver any benefits (as a worst-case assumption). Further details of

CURED v3A are provided in the supporting report prepared by AQC (AQC,

2018a).

4.5 Additional features

4.5.1 There are a number of road tunnels within the model domain, and the DCO Project

will introduce some new road tunnels. To ensure a robust assessment, these have

been modelled as tunnels using the ADMS Tunnels module. Some of the modelled



tunnel links are long, and would extend beyond the boundary of some of the 2km x

2km model grid areas (see Paragraph 4.1.2). In order to avoid introducing artificial

tunnel ends where this occurs, tunnels have been removed from the primary

modelled road network and modelled explicitly with a separate receptor file. This

receptor file has incorporated all of the specific receptors within the 1km x 1km

grid squares within which the tunnel links are located and within 500m of the

edges of these areas, then a grid of receptors at 100m resolution further afield, to

align with the approach taken elsewhere in the assessment.

4.5.2 Noise barriers have also been incorporated into the model, where currently

present. The design of the DCO Project is not sufficiently advanced as to

incorporate any potential noise barriers along proposed new roads to be built as

part of the DCO Project, thus noise barriers in the with DCO Project scenarios

have only been included along roads that are to be retained and which currently

have noise barriers present.

4.5.3 Where relevant, elevated roads have been modelled at height. Only sections of

road longer than 50m and which have clear air underneath the carriageway have

been modelled at height. This is because the option to give road sources an

elevation is really intended for true bridges, with the initial mixing of the emissions

given an extra downward component to account for the passage of air beneath the

source. As such, it would not be appropriate to model roads without clear air

underneath them as being at height.



5. APPROACH TO MODELLING CAR PARK EMISSIONS

5.1.1 Emissions from vehicles using car parks operated by Heathrow have been

modelled using the following approach:

1. Volume sources of 3 m depth have been defined covering the area of each car

park (sensitivity testing has identified a volume source of 3 m depth at ground

level as best representing ground level vehicle emissions, when compared to

modelling as a line source), with multi-storey car park volume sources given a

total depth assuming each storey to be 3 m deep

2. An average trip length through the car park has been defined based on

mapping data

3. Emissions have been calculated using the EFT and CURED for every vehicle

using the car park (using usage figures provided by the surface access team)

on the assumption that they travel this average distance at 5 kph, the speed

associated with the highest emissions in the EFT

4. The total emission rate for each car park volume source has been calculated

and modelled using ADMS-5, assuming a constant diurnal and seasonal profile

of emissions

5. Cold-start emissions have been defined using the National Atmospheric

Emissions Inventory (NAEI) cold start emission rates (which are derived from

COPERT). 50% of the total cold start emission for each car park has been

applied to the modelled volume source, with the other 50% averaged over 200

m long line sources representing the most likely routes of traffic exiting from the

car parks.

5.1.2 This is considered to represent a suitably robust approach to modelling emissions

from vehicles using car parks, which represent a very small proportion of total

concentrations. In reality, there is likely to be some diurnal and seasonal profile to

these emissions, but these will be different to typical road traffic profiles on the

highway network, thus it would not be appropriate to apply a default highway

network profile. It is considered more appropriate to simply assume a constant

profile, in the knowledge that this is unlikely to lead to significant uncertainty in

annual mean contributions.

5.1.3 Concentrations have been predicted at all of the gridded and discrete receptors

within 200m of a car park or likely route of traffic exiting a car park, and at 100m

resolution beyond this distance.



6. APPROACH TO MODELLING LAKESIDE WASTE MANAGEMENT FACILITY EMISSIONS

6.1.1 Emissions from the Lakeside Waste Management Facility in the baseline and

future baseline scenarios have been modelled using the ADMS-5 model. The

urban canopy module has been used in order for the modelling to align with the

bulk of the other modelling carried out. This has resulted in buildings not being

incorporated into the model, but the urban canopy flow module should itself

account for some of the effects of the presence of the process building, and the

stack at the existing facility is considerably higher than the highest point of the

process building (75 m as opposed to 42 m), thus building effects will be relatively

limited. As such, the approach is deemed appropriate.

6.1.2 Emissions data for the existing facility have been derived from published datasets,

which include average pollutant emissions monitoring data from 2010 to 2016, and

are summarised in Table 6.1, generally to two significant figures. The facility has

three lines of plant, one clinical waste incinerator (CWI) and two energy from

waste (EfW).

Table 6.1: Emission data for the existing facility

Parameter CWI EfW (per line)

Stack Location (x,y) 503900,177341

Temperature 140 145

Exhaust Volume Flow Rate

(Nm3/s)

4.1 66

Exit velocity 15 19

NOX (mg/Nm3) 200 166

NOX (g/s) 0.82 6.6

PM10 (mg/Nm3) 10 1.2

PM10 (g/s) 0.041 0.047

6.1.3 Concentrations have been predicted at each of the monitoring sites used in the

model verification, across a 100 m resolution Cartesian grid of receptors covering

the CAQOAA, and across a coarser nested grid of receptors at 2km and 4km

resolution covering the extent of the HHASAM Fully Modelled Area (FMA). The

FMA is the area in which all trip movements are included in the model. These

concentrations have been added to the total concentrations at receptors in the

future baseline scenarios using the approach described in Section 8. The with

DCO Project scenarios have assumed no Lakeside Waste Management Facility to

be in place.



7. APPROACH TO PREDICTING BACKGROUND CONCENTRATIONS

7.1.1 Defra’s background maps have been used to provide the concentrations

associated with pollutant sources which are not being explicitly modelled.

7.1.2 In order to avoid double-counting, emissions from sources that have been

explicitly modelled (e.g. airport and road traffic) have been removed from Defra’s

maps. Defra provides a method to do this, but this functionality is limited to

removing pre-defined sectors and to removing all ‘in-square’ emissions and/or all

‘out-square’ emissions2. It would, in theory, be possible to configure the dispersion

model in such a way that only ‘in-square’ emissions were being explicitly modelled

for any one receptor. This would allow all ‘out-square’ background components to

be retained. The main problems with such an approach are:

1. it would introduce inconsistency and step-changes in the outputs

2. it would ignore the effect that the change in traffic caused by the DCO Project

over the wider network will have on the ‘background’ concentration field.

7.1.3 The method set out below (in paragraph 7.3.1 onwards) seeks to predict a

spatially-consistent background concentration field which takes account of wider-

scale changes in traffic emissions caused by the DCO Project.

7.1.4 Emissions attributed to Heathrow Airport fall within several sectors, some obvious

(e.g. ‘Aircraft’) and some less so (e.g. ‘Other’). In order to ensure that no double-

counting of Heathrow emissions occurs, Ricardo Energy & Environment were

commissioned, with Defra’s approval, to determine the explicit contribution of

Heathrow Airport to all non-road sectors of the background maps, so that this can

be removed from the mapped backgrounds.

7.1.5 For completeness, the sectors included in the background maps are detailed in

Table 7.1.

2 i.e. all said emissions originating inside or outside of a specific cell of the 1km x 1km grid over which the maps are provided.



Table 7.1: Sectors included in the background maps

7.1.6 The following road traffic sources have also been removed from Defra’s

background maps:

1. Motorway

2. Trunk A Rd

3. Primary A Rd

4. Brake+Tyre Wear (PM Only)

5. Road Abrasion (PM Only)

7.1.7 The contribution of the Minor Rd+Cold Start sector has been retained, given that

most minor roads will not be modelled explicitly due to not being incorporated in

the HHASAM network, and to allow for the contribution of cold start emissions

across the study area. This will result in some double-counting of emissions from

minor roads, and cold start emissions from Heathrow’s car parks, but this will be

very small. Accounting for the additional emissions from the development-

generated traffic across the HHASAM FMA is considered more important than

avoiding a small amount of double-counting of the baseline component.

7.1.8 Cold start emissions for trips associated with the DCO Project originating at the

airport have been modelled explicitly for all Heathrow-controlled car parks. Cold

start emissions for trips associated with the DCO Project originating elsewhere will

not be counted, other than through the inclusion of the contribution of the Minor

Rd+Cold Start sector from the background maps.

Sector

Motorway

Trunk A Rd

Primary A Rd

Minor Rd+Cold Start

Brake+Tyre Wear (PM Only)

Road Abrasion (PM Only)

Industry

Domestic

Rail

Other

PM secondary (PM Only)

Residual+Salt (PM Only)

Point Sources

Rural (NOX Only)

Aircraft (NOX Only)



7.1.9 In determining appropriate mapped background NOX and nitrogen dioxide (NO2)

concentrations, Defra’s sector removal tool has been used to remove the relevant

road and non-road contributions from the background maps. The end result is a

set of background concentrations that do not include the contribution of major

roads, sources within Heathrow Airport or the Lakeside Waste Management

Facility.

7.1.10 Defra’s background maps use 2015 as a base year, and this is widely

acknowledged to have been a low-pollution year. As a result, the maps may tend

to under-predict future year background concentrations, due to having been

validated against 2015. In order to offset any such under-prediction when using

2016 or 2017 as a base year, the residual mapped background concentrations of

NOX and NO2 have been adjusted upward using the factors recommended in

AQC’s report on Calibrating Defra’s 2015-based Background NOX and NO2 Maps

against 2016 and 2017 Measurements (AQC, 2018b).

7.2 Background NO2 and NOX concentrations for sensitivity test

7.2.1 The road-traffic components of nitrogen oxides and nitrogen dioxide in the

background maps (which are limited to Minor Rd+Cold Start, as the other road

sectors have been removed) have been uplifted in order to derive future year

background nitrogen dioxide and nitrogen oxides concentrations for use in the

sensitivity test. Details of the approach are provided in the report prepared by AQC

(AQC, 2018c).

7.3 Modelling the background concentration field for traffic emissions

7.3.1 While the bulk of the road traffic emissions in the background maps have been

removed, the explicit modelling of road sources only extends to at least 500m from

any given receptor, thus it is necessary to re-introduce the contribution of roads

not explicitly modelled as road sources to the background concentration at each

receptor.

7.3.2 Graphic 7.1 shows the CAQOAA, which is the area over which total

concentrations have been predicted. It also shows the FMA considered in

HHASAM. The extent of the FMA is the boundary beyond which not all trip

movements are included in the model and there is only skeletal coverage of the

road network.

7.3.3 Emissions from all roads within the FMA have been used to predict the road

component of the background concentration field, using the following approach:



1. Step 1) A single 100 m x 100 m volume source3 has been modelled in ADMS

with a unitary emission rate4, with concentrations predicted for a large 100 m x

100 m Cartesian receptor grid.

2. Step 2) The FMA has been divided into 100 m x 100 m grid cells, and those

grid cells which contain roads have been identified. For each traffic model

scenario, link-specific traffic emissions have been calculated and assigned to

the respective grid cells (based on the length of road within each cell).

3. Step 3) The outputs from Step 1 and Step 2 have been combined for each

source-cell, and summed for each receptor on the Cartesian grid, to give the

total concentration at each gridded receptor. This has been done in such a way

that only those roads which are not being modelled explicitly (as described in

Section 4, and including relevant tunnels modelled explicitly) are included.

3 Sensitivity testing by AQC has shown that using a volume source more accurately reproduces the concentrations predicted when modelling road traffic emissions as line sources, when compared to using an area source. This sensitivity testing has identified an optimal volume source of 3 m depth at ground level. 4 A unitary emission rate is a default assumed value, e.g. 1 g/m3/s. Its use allows the actual emission rate to be applied in post-processing, by multiplying the output concentration by the actual emission rate divided by the unitary emission rate (a value of one is used to remove the requirement for this division). The relationship between primary pollutant model inputs and model outputs is linear, and rounding will not introduce significant error, thus this approach results in largely identical modelled concentrations when compared to modelling the actual emission rate.



Graphic 7.1: Air quality core assessment area and HHASAM Fully Modelled Area

Imagery © Google

7.3.4 Graphic 7.2 provides an example subset of the 100m x 100m Cartesian receptor

grid within a defined 1km x 1km square. It also presents the nearest of the

surrounding 100m x 100m volume sources (located outside of the area within

which roads are being modelled explicitly for that 1km x 1km square), whose

contribution to background concentrations at this subsect of receptors will be

calculated following the approach set out above.



Graphic 7.2: Receptor grid example

Contains Ordnance Survey data © Crown copyright and database right 2018. Ordnance

Survey licence number 100046099.

7.3.5 This approach ensures that changes in traffic emissions as a result of the DCO

Project throughout the FMA are incorporated into the total concentrations

predicted. It does result in some minor double-counting of traffic emissions, as

traffic emissions greater than 16km away are effectively included in the ‘Rural’

sector of the background maps. However, these contributions are small, and it is

considered more important that the change as a result of the DCO Project is

accounted for than this minor double-counting is avoided.

7.3.6 The spatial variation in NOX concentrations predicted using this methodology is

limited, as the concentrations primarily relate to distant sources. As such, a

nearest-neighbour approach has been taken to determining the concentration at

discrete receptors, with the road traffic background contribution taken to be that of

the nearest 100m x 100m grid point.

7.3.7 The output NOX concentrations have been adjusted by applying the primary

adjustment factor for road traffic emissions (see Section 9), in order to avoid step

changes in concentrations at the edges of the 1km x 1km grid squares, and to

ensure that background concentrations are not under-predicted.



8. CALCULATION OF TOTAL NO2 CONCENTRATIONS

8.1.1 NOX is emitted as a mixture of nitric oxide (NO) and NO2 (primary NO2), and

reactions in the atmosphere convert NO to NO2, and vice versa. Concentrations of

NOX are conserved5, so are straightforward to calculate through dispersion

modelling. Concentrations of NO2 at receptors are, however, a complex function of

emissions of NO and primary NO2, concentrations of oxidants (principally NO2 and

O3) in the air, the magnitude of incoming solar radiation and travel time. Modellers,

therefore, require a procedure for calculating NO2 concentrations from the NOX

concentrations calculated by dispersion modelling.

8.1.2 The NOX to NO2 Calculator available from the LAQM website

(https://laqm.defra.gov.uk/review-and-assessment/tools/background-

maps.html#NOXNO2calc) has been used to predict total NO2 concentrations. It

takes a semi-empirical approach (Abbott, J, 2005) which ‘uses a one-dimensional

finite difference model of the reactions and mixing of NO, NO2 and O3 in the

surface stress layer of the atmospheric boundary layer.’

8.1.3 The NOX to NO2 calculator requires the user to define a specific local planning

authority area, which is used to estimate regional concentrations of O3, NOX and

NO2 above the surface layer. Hillingdon has been used throughout this

assessment, as it is considered to appropriately represent conditions in the vicinity

of Heathrow Airport, and using a varied selection would result in unrealistic step

changes in concentrations at local planning authority boundaries.

8.1.4 The user is also required to define the traffic mix, which is used to define the

appropriate fNO2 value for road traffic emissions. ‘All London traffic’ has been used

throughout this assessment, as it is considered the most representative option for

the area of interest while ensuring no step changes in concentrations.

8.1.5 In order to determine total pollutant concentrations, all of the individual

contributions have been combined. The process for this is described below.

8.1.6 The first step has been to determine a receptor-specific background NO2

concentration. The mapped background concentrations with the contribution of

Heathrow Airport and major road sources removed (see Section 7) have been

interpolated to provide receptor specific Total NOX, Road NOX, Non-Road NOX

and Total NO2 concentrations. These have been input to Defra’s sector removal

tool, and the adjusted modelled road traffic background contribution added to them

5 Small losses through deposition are ignored for the purposes of modelling air concentrations, giving rise to a small degree of double counting of the deposited NOx.

https://laqm.defra.gov.uk/review-and-assessment/tools/background-maps.html

https://laqm.defra.gov.uk/review-and-assessment/tools/background-maps.html



by entering a negative value in column K, to give a final ‘background’ annual mean

NO2 concentration for every receptor.

8.1.7 The modelled road NOX (including the tunnels contribution) has then also been

adjusted by applying the primary adjustment factor for road traffic emissions (see

Section 9), with the interpolated modelled car park and airport contributions then

added to give a total modelled NOX concentration for every receptor. Receptor-

specific fNO2 values have then been calculated by multiplying each constituent

part by its source-specific fNO2 value, to determine an appropriate overall fNO2

value.

8.1.8 The total modelled NOX, total background NO2 and fNO2 values for each receptor

have then been run through Defra’s NOX to NO2 calculator to define a total

modelled NO2 concentration, to which the secondary adjustment factor has been

applied to give a final total NO2 concentration at every receptor. The application of

a secondary adjustment factor in order to bring the modelled NO2 concentrations

into alignment with the local monitoring is considered appropriate, as the bulk of

the total concentration has been modelled, either explicitly or through a kernel

approach. The need for such a factor is thought to relate to the over-estimation of

road vehicle fNO2 in the NOX to NO2 calculator.

8.2 EU limit value compliance

8.2.1 It should be noted that the approach to the conversion of NOX to NO2 adopted

when assessing EU limit value compliance was different to that described above,

utilising the Jenkin approach set out in Defra’s SL_PCM concentration tool

spreadsheets (Defra, 2017b). This is in order to ensure that the conversion aligns

with the PCM modelling carried out by Defra, which is necessary when considering

limit value compliance. The approach to assessing limit value compliance is

discussed further in Section Error! Reference source not found. of this note.



9. MODEL VERIFICATION METHODOLOGY

9.1.1 The model output concentrations have been verified against measured

concentrations from suitable automatic monitoring sites within the CAQOAA. The

sites used in the model verification are listed in Table 9.1.

Table 9.1: Sites used in the model verification

9.2 NOX and NO2

9.2.1 Most NO2 is produced in the atmosphere by reaction of NO with ozone (O3). It is

therefore most appropriate to verify the model in terms of primary pollutant

emissions of NOX (NOx = NO + NO2).

9.2.2 The model output of road-NOx (i.e. the component of total NOx coming from road

traffic) has been compared with the ‘measured’ road-NOx. Measured road-NOx

has been calculated by subtracting the following components from the measured

NOx concentration at each monitor:

1. background NOx

2. airside NOx (including emissions from aircraft, ground support equipment and

heating and cooling plant)

3. the modelled NOx contribution of Heathrow’s car parks

4. the modelled NOx contribution of the Lakeside Waste Management Facility.

Site Name Years Used Pollutants Measured

London Hillingdon 2015, 2016, 2017 NOX, NO2

London Harlington 2015, 2016, 2017 NOX, NO2, PM10, PM2.5

LHR2 2015, 2016, 2017 NOX, NO2, PM10, PM2.5

Oaks Road 2015, 2016, 2017 NOX, NO2, PM10, PM2.5

Green Gates 2015, 2016, 2017 NOX, NO2, PM10, PM2.5

Oxford Avenue 2015, 2016, 2017 NOX, NO2, PM10 (2016 and 2017 only)

Harmondsworth 2015, 2016, 2017 NOX, NO2, PM10

Hayes 2015, 2016, 2017 NOX, NO2, PM10

Feltham 2015, 2016, 2017 NOX, NO2, PM10

Heston Road 2015, 2016, 2017 NOX, NO2, PM10

Southall 2015 NOX, NO2

Sipson 2015, 2016, 2017 NOX, NO2

Cranford 2015, 2016, 2017 NOX, NO2, PM10

Hatton Cross 2015, 2016, 2017 NOX, NO2, PM10

SLH 3 & SLH6 2015, 2016, 2017 NOX, NO2, PM10



9.2.3 An adjustment factor has been determined as the slope of the best-fit line between

the ‘measured’ road contribution and the model derived road contribution, forced

through zero. The total nitrogen dioxide concentrations have then been

determined by combining the adjusted total NOx concentrations with the predicted

background NO2 concentration within the NOx to NO2 calculator. A secondary

adjustment factor has then been calculated as the slope of the best-fit line applied

to the adjusted total NO2 concentrations and forced through zero.

9.2.4 Model verification factors for NOX and NO2 have been determined for the years

2015, 2016 and 2017 using two approaches, a ‘Simple’ calculation of the factors

when using data from all of the automatic monitoring sites included in the model

verification, and alternative factors calculated using the Leave One Out Cross

Validation (LOOCV) approach. The modelled Road NOX has been adjusted,

followed by total NO2. There is no strong justification for adjusting any other

contributions. The calculated factors for each year are presented in Table 9.2.

Table 9.2: Calculated factors for each year

Year Method Primary Road NOX

Adjustment Factor

Secondary Total NO2

Adjustment Factor

2015 Simple 3.2585 0.9779

LOOCV 3.2668 0.9773

2016 Simple 3.4487 0.8935

LOOCV 3.4615 0.8928

2017 Simple 3.5843 0.9235

LOOCV 3.5915 0.9230

9.2.5 The factors calculated using the two approaches are very similar in each year. The

primary adjustment factors increase year-on-year; this is likely due to projected

emissions reductions in the EFT not having been realised in terms of an obvious

reduction in measured concentrations. 2015 having been a ‘low-pollution year’ in

terms of measured concentrations UK-wide will also have influenced the lower

primary adjustment factor for this year, as well as the secondary adjustment factor,

which is much closer to one.

9.2.6 The secondary adjustment factors calculated demonstrate that secondary

adjustment is necessary to avoid significant over-estimation of annual mean

nitrogen dioxide concentrations. This is thought to be at least partially due to the

over-estimation of primary NO2 from road traffic in Defra’s NOX to NO2 calculator.

Further sensitivity testing around this is proposed for inclusion in the ES.

9.2.7 The statistical performance of each of the models is presented in Table 9.3. LAQM

TG.16 advises that ‘ideally an RMSE within 10% of the air quality objective would

be derived, which equates to 4µg/m3 for the annual average NO2 objective’.



However, it is only recommended that model inputs and verification should be

revisited if RMSE values are higher than ±25% of the objective.

Table 9.3: Statistical performance of each of the models

Year Method Correlation

Coefficient

Root Mean Square Error

Fractional Bias

2015 Simple 0.88 4.15 0.01

LOOCV 0.84 4.93 0.01

2016 Simple 0.80 5.06 0.02

LOOCV 0.75 5.93 0.01

2017 Simple 0.87 4.40 0.01

LOOCV 0.83 5.03 0.01

‘Ideal’ Value 1 0 0

9.2.8 Graphic 9.1 to Graphic 9.6 plot the final modelled annual mean nitrogen dioxide

concentrations against the measured concentration for each verification scenario.

Graphic 9.1: 2015 Model performance – simple verification approach



Graphic 9.2: 2015 Model performance – LOOCV verification approach




Graphic 9.4: 2016 model performance – LOOCV verification approach




Graphic 9.6: 2017 Model performance – LOOCV verification approach

9.2.9 The graphs generally show good agreement. There is one specific outlier, the

Heston automatic monitor in Hounslow, which falls outside of the 25% lines in

some of the graphs. Graphic 9.7 presents the primary adjustment factors

calculated for each site using the data from all of the other sites in the LOOCV

analyses, and highlights that this monitor is a clear outlier, with the factors at the

other sites all being very consistent.



Graphic 9.7: LOOCV primary verification factors

9.2.10 The Heston automatic monitor is located adjacent to the traffic-light controlled

junction of Great West Road (A4) and Heston Road (A3005), while also being

some 150 m west of the traffic-light controlled junction of Great West Road (A4)

and Jersey Road. The A4 here has three lanes in each direction, carries a large

volume of traffic (> daily vehicles), and will see regular idling traffic and quick

acceleration, due to the presence of the traffic lights. The average speeds

assumed in the modelling for the roads at these junctions are low (generally

around 20 kph) and are considered appropriate bearing in mind the above. It is

unsurprising, therefore, that the model is predicting a relatively large road-NOX

contribution.

9.2.11 Further consideration has been given to model predictions at busy junctions, to

determine whether there is any systematic over-prediction of road-NOX. A

comparison of total modelled annual mean NO2 against measured NO2 at 19

diffusion tube monitoring sites close to busy junctions is shown in Graphic 9.8.

This shows a range of over-and under-prediction of concentrations in each year,

which is to be expected and demonstrates that there is no systematic issue with

the model’s performance close to busy junctions. Heston is simply an outlier, a

location where concentrations could reasonably be expected to be higher than

measured, and thus not one that invalidates the overall model performance in any

way.



Graphic 9.8: Modelled vs measured NO2 at diffusion tube sites close to busy junctions

9.2.12 While the removal of the Heston monitor from the verification would reduce the

RMSE to below 10% of the air quality objective for all three years, doing so would

result in a higher primary adjustment factor being applied, and higher total

concentrations being predicted. Graphic 9.9 to Graphic 9.11 plots demonstrate

the model performance against diffusion tube monitoring sites within the

CAQOAA. They show a relatively large scatter, which is to be expected when

comparing against diffusion tube data, which itself has a relatively large

measurement error margin. Overall, the plots suggest that the model is performing

well on average and, if anything, is slightly over-predicting total concentrations,

which is worst-case.

9.2.13 Several of the diffusion tube monitoring sites at which the model is tending to over-

predict, when compared to measured concentrations, are located close to the M25

motorway. Without these sites included, the comparison would be much closer to

a 1:1 relationship. This suggests that, in general, the model is performing well, but

that over-predictions close to the M25 are likely. This is somewhat unsurprising,

given that the model adjustment factors calculated have been applied to all road

traffic contributions, including motorways, when it is common knowledge among

air quality professionals that models tend to perform better when it comes to



motorway emissions, requiring less or no upward adjustment. In considering the

results of the air quality modelling it should be borne in mind that concentrations

close to the M25 are likely to have been over-predicted, and are thus worst-case,

especially in the case of the CURED sensitivity test predictions.

Graphic 9.9: 2015 Modelled vs measured NO2 at diffusion tube sites





9.2.14 The comparison to diffusion tube monitoring has demonstrated that the model is

already tending towards over-predicting concentrations, and this over-prediction

would be made worse without Heston incorporated in the verification. As such, it is

considered that there is no robust justification for removing the Heston monitor

from the verification.

9.2.15 Bearing in mind this tendency to over-predict, the adjustment factors calculated

using the ‘Simple’ approach have been applied, rather than the averages from the

LOOCV scenario. The LOOCV scenarios provide valuable insight into the model

performance, and produce factors very similar to the ‘Simple’ approach, albeit

ones that would produce very slightly higher total concentrations. It has been

judged best to apply the factors derived using all of the monitoring sites in each

year.

9.3 PM10 and PM2.5

9.3.1 The model performance has been tested against monitoring data for PM10 and

PM2.5. Graphic 9.12 to Graphic 9.17 present graphs of the modelled annual mean

concentrations plotted against the measured concentrations at all of the

appropriate monitoring sites in the CAQOAA.



Graphic 9.12: 2015 model performance – PM10

Graphic 9.13: 2015 model performance – PM2.5











9.3.2 The graphs for PM2.5 show a very good alignment with the monitoring data. In

reality, concentrations of this pollutant are dominated by the background

concentration; the graphs imply that the background concentrations used are

reasonable, and that the model is doing an adequate job of predicting the road

traffic and airport contributions, without the need for adjustment.

9.3.3 The graphs for PM10 do not present such a strong alignment, suggesting that,

while the model is performing well at some sites, concentrations of this pollutant

are being under-predicted at a number of sites. The locations of these sites have

been carefully considered to establish whether there is any reason why the model

might systematically under-predict concentrations at them, but in reality they share

little in common with one another with regards to the key modelled sources (i.e.

road traffic and airport emissions). This analysis has, however, identified that there

may well be specific local sources affecting these sites. The following have been

identified:

1. The Hayes automatic monitor is located adjacent to the junction of North Hyde

Road and North Hyde Gardens, with the latter being an access road to a

number of industrial premises, some of which are known to handle aggregates

(e.g. Conway Asphalt Plant). Historic photography of the junction shows that

the roads are routinely very dusty, a localised issue likely to lead to higher

measured concentrations that might reasonably be expected, and not one that

can readily be allowed for in the modelling

2. Historic photography of the junction adjacent to the Heston automatic monitor

also routinely shows very dusty roads. There are no obvious local sources for

this dust, but its presence on the roads will lead to higher particulate

concentrations than might otherwise be expected. The fact that the Hayes and

Heston monitors measured considerably higher concentrations in every year

when compared to the other sites suggests that these persistently dirty roads

are likely to be driving the high concentrations

3. The Harmondsworth automatic monitor is located adjacent to a quiet, dead-end

road, and the high measured concentrations are unlikely to relate to road traffic

emissions. There are no obvious local sources that might be causing the high

concentrations measured here. It is possible that agricultural practices on the

adjacent fields might be leading to dust generation, or the industrial sites on the

opposite side of the M25 (upwind under prevailing wind conditions) might

represent a significant source, but this is merely speculation. Airport-related

sources will not be contributing significantly to the measured concentrations,

which are clearly above the wider background levels, and there is no

justification for adjusting any of the modelled contributions



4. Roads close to the Oxford Avenue monitor do not appear to routinely be dusty,

and there are no other obvious sources of PM nearby. The higher

concentrations measured here are unlikely to be airport-related, given how

much lower concentrations measured at nearby LHR2 are. It is possible that

they relate to road traffic sources, but the modelled road traffic contribution

would have to be increased by a factor of around 15 to match the measured

values. Such an increase seems unrealistic, and does not inspire confidence

that adjustment is appropriate

5. The other sites mostly fall within the 25% error margins, but those closer to the

upper 25% bound are likely to also be influenced by more minor, local sources,

although these are not obvious. One example is Slough’s Colnbrook monitor,

located in Poyle. Poyle is surrounded by industrial estates to the north, east

and south, and these are likely the reason why measured concentrations are

above wider background levels, although the measured concentrations

probably do represent an appropriate background concentration for the village

of Poyle. Given that the background maps represent 1km square averages,

they will inevitably slightly over or under-predict concentrations at some

locations within each square.

9.3.4 It is considered most likely that local sources are leading to the high measured

concentrations at some sites, and these are not sources that could readily be

incorporated into the model. As such, it would not be appropriate to adjust the

model outputs, which are judged to represent an appropriate prediction of PM10

concentrations across the CAQOAA in the absence of specific local sources of

particulate matter.



10. EFFECT ON EU LIMIT VALUE COMPLIANCE METHODOLOGY

10.1.1 In order to assess the impact of the DCO Project on compliance with the NO2

annual mean EU limit value, emissions associated with the additional traffic

generated by the DCO Project have been modelled on a selection of the worst-

case roads for potential limit value impacts within the CAQOAA. These are those

roads where the DCO Project causes the largest changes in traffic volumes or

fleet composition, and those with the highest predicted concentrations in future-

year PCM mapping.

10.1.2 Concentrations have been predicted at 4 m from the kerb on stretches of road that

are representative of 100 m of road and not within 25 m of major junction, in line

with historical PCM modelling.

10.1.3 A NOX increment associated with the DCO Project has been calculated at each of

these PCM receptors using the same approach as described above for the

emissions modelling, by subtracting the total NOX concentration in the with DCO

Project scenarios from those in the future baseline scenarios. This increment has

incorporated the additional emissions from road traffic changes throughout the

FMA and changes in aircraft and other airside source emissions.

10.1.4 The NOX increment calculated has been added to the NOX concentration in

Column Y of the ‘Calcs_Scenario’ tab of the SL_PCM concentration tool

spreadsheets available from the Defra website (Defra, 2017b). The local oxidant

value in column AA has been adjusted to account for these additional

contributions, with the link-specific fNO2 value from the spreadsheet applied for

road-NOX and an appropriate receptor-specific fNO2 value applied for airside NOX,

derived based on the relative contribution of the sources and their relative fNO2

values. The spreadsheet has then derived a new total NO2 concentration (in

Column AC).

10.1.5 The change in NO2 as a result of the DCO Project has been determined by

comparing this new total NO2 concentration to the PCM baseline, and compliance

with the limit value considered. It must be noted that this analysis utilises Defra’s

2015 reference year modelling, which is the only modelling for which SL_PCM

concentration tool spreadsheets are available. In order to consider the impacts

against Defra’s recently published 2017 reference year modelling (Defra, 2019),

the change in concentrations calculated using the 2015 reference year

spreadsheets has simply been added to the 2017 reference year baseline

concentrations. While this method will not have resulted in exactly the right

conversion ratio of NOx to NO2 being applied, the baseline concentrations vary



relatively little between the two scenarios, thus any inconsistency will be extremely

small, and highly unlikely to affect the conclusions drawn. It is also the best

possible solution without 2017 reference year SL_PCM concentration tool

spreadsheets, which are not available.

10.1.6 Outside of the CAQOAA, the key limit value compliance locations of concern are

those in Central London; impacts here have been considered on the basis of traffic

data within the FMA and qualitatively using demand forecast information from the

surface access modelling.



11. CURED SENSITIVITY TEST RESULTS

11.1 Community area results

11.1.1 The NO2 results produced for each community area using the CURED sensitivity

test emission factors are presented in Table 11.1.



Table 11.1: CURED sensitivity test NO2 results at selected representative receptors within each community area

Year Maximum

Concentration

(µg/m3)

Magnitude (µg/m3) Impact Descriptors at Selected Representative Receptors

Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

Harmondsworth

2022 35.6 36.7 0.0 1.2 0 0 0 28 7 0 0

2027 30.0 29.4 -4.1 5.0 0 0 2 24 6 3 0

2030 28.1 26.3 -3.2 5.4 0 0 2 24 5 4 0

2035 28.6 25.8 -5.2 5.9 0 1 8 17 4 5 0

West Drayton

2022 46.8 47.2 0.1 1.6 0 0 0 15 6 5 0

2027 39.6 41.1 0.3 2.4 0 0 0 10 14 2 0

2030 37.3 38.2 0.3 1.9 0 0 0 16 9 1 0

2035 37.8 38.7 0.2 2.5 0 0 0 15 5 6 0

Sipson

2022 29.3 29.7 0.1 0.6 0 0 0 31 0 0 0

2027 24.8 26.3 0.9 5.0 0 0 0 8 19 4 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 23.3 24.4 0.8 4.8 0 0 0 8 20 3 0

2035 23.7 25.2 1.0 4.9 0 0 0 7 19 5 0

Harlington

2022 44.5 45.1 0.2 1.0 0 0 0 14 5 0 1

2027 36.9 36.9 -0.4 5.0 0 0 0 17 1 2 0

2030 34.2 34.4 0.1 4.4 0 0 0 17 2 1 0

2035 34.6 34.5 -0.1 4.1 0 0 0 14 6 0 0

Hayes

2022 44.3 44.9 0.2 0.6 0 0 0 7 3 5 0

2027 36.7 36.5 -0.1 0.6 0 0 0 15 0 0 0

2030 34.0 34.1 -0.1 0.3 0 0 0 15 0 0 0

2035 34.4 34.2 -0.1 0.4 0 0 0 15 0 0 0

Cranford Cross

2022 29.1 29.5 0.2 0.4 0 0 0 13 0 0 0

2027 24.3 23.7 1.4 1.4 0 0 0 13 0 0 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 22.9 22.6 1.7 1.7 0 0 0 13 0 0 0

2035 22.8 23.1 2.2 2.2 0 0 0 13 0 0 0

Cranford

2022 39.4 39.6 0.3 0.4 0 0 0 13 1 0 0

2027 32.8 33.7 0.2 0.9 0 0 0 12 2 0 0

2030 30.6 31.0 0.3 0.4 0 0 0 14 0 0 0

2035 30.9 31.3 0.3 0.4 0 0 0 14 0 0 0

Heston

2022 45.9 46.3 0.2 0.5 0 0 0 9 6 2 0

2027 37.7 37.7 0.2 0.4 0 0 0 16 0 0 0

2030 34.9 35.0 0.3 0.3 0 0 0 17 0 0 0

2035 35.2 35.5 0.3 0.3 0 0 0 17 0 0 0

Hounslow (Central and South)

2022 36.0 36.3 0.0 0.2 0 0 0 6 0 0 0

2027 29.7 30.7 0.0 1.0 0 0 0 5 1 0 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 27.9 28.9 0.0 1.0 0 0 0 6 0 0 0

2035 28.3 28.7 0.0 0.4 0 0 0 6 0 0 0

Hounslow (West and Heath)

2022 38.5 38.7 0.0 0.7 0 0 0 22 2 0 0

2027 32.3 33.5 0.0 1.2 0 0 0 21 3 0 0

2030 30.3 31.0 0.0 0.7 0 0 0 23 1 0 0

2035 30.6 31.3 0.0 0.7 0 0 0 23 1 0 0

Feltham North

2022 38.0 38.3 0.3 0.5 0 0 0 11 1 0 0

2027 31.8 32.3 -0.1 0.7 0 0 0 11 1 0 0

2030 29.6 29.9 -0.1 0.4 0 0 0 12 0 0 0

2035 29.9 30.6 0.6 0.7 0 0 0 11 1 0 0

Bedfont

2022 37.2 37.5 0.2 0.4 0 0 0 12 0 0 0

2027 31.2 32.1 -0.6 1.3 0 0 0 10 2 0 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 29.1 30.0 -0.4 1.2 0 0 0 12 0 0 0

2035 29.4 30.6 0.6 1.5 0 0 0 11 1 0 0

Stanwell

2022 32.9 33.3 0.3 0.5 0 0 0 18 0 0 0

2027 27.2 28.2 0.0 1.0 0 0 0 18 0 0 0

2030 25.4 27.4 0.2 2.1 0 0 0 18 0 0 0

2035 25.6 27.6 1.1 2.3 0 0 0 17 1 0 0

Stanwell Moor

2022 31.2 31.7 0.4 1.1 0 0 0 12 0 0 0

2027 26.3 26.1 -0.2 2.2 0 0 0 12 0 0 0

2030 24.7 26.2 -0.5 4.3 0 0 0 7 4 1 0

2035 24.8 26.6 -0.1 4.6 0 0 0 7 3 2 0

Poyle

2022 30.4 30.7 0.2 0.3 0 0 0 12 0 0 0

2027 25.8 25.6 -0.1 2.4 0 0 0 11 1 0 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 24.2 24.8 0.1 2.3 0 0 0 11 1 0 0

2035 24.3 25.4 0.3 2.4 0 0 0 10 2 0 0

Colnbrook

2022 32.3 32.8 0.1 0.5 0 0 0 12 0 0 0

2027 26.7 26.8 0.4 1.3 0 0 0 12 0 0 0

2030 24.6 24.8 0.4 1.1 0 0 0 12 0 0 0

2035 24.7 25.2 0.4 1.2 0 0 0 12 0 0 0

Brands Hill

2022 45.8 47.4 0.3 3.0 0 0 0 2 2 4 8

2027 39.1 39.3 -1.2 0.6 0 1 4 9 2 0 0

2030 36.5 36.6 -1.4 0.6 0 0 4 12 0 0 0

2035 37.1 37.0 -1.7 0.9 0 0 4 11 1 0 0

Iver and Richings Park

2022 52.2 52.9 0.3 0.7 0 0 0 12 0 0 2

2027 44.7 44.9 0.5 0.8 0 0 0 13 1 0 0



Year Maximum

Concentration

(µg/m3)


Future

Baseline

With

DCO

Project

Maximu

m

decrease

or

minimu

m

increase

Maximu

m

Increase

Subs.

Ben.

Mod.

Ben.

Slight

Ben.

Neg. Slight

Adv.

Mod.

Adv.

Subs.

Adv.

2030 41.8 41.8 0.6 0.6 0 0 0 14 0 0 0

2035 42.3 42.3 0.6 0.6 0 0 0 14 0 0 0

Notes:

Impacts described as beneficial or adverse and substantial, moderate, slight or negligible.

The maximum pollutant concentrations show the maximum at a receptor in a community area for a particular scenario. The maximum concentration

does not necessarily occur at the same receptor in the Future Baseline and With DCO Project scenario in each year.



11.2 Summary results across the CAQOAA

11.2.1 The property counts for NO2 concentrations produced using Defra EFT road traffic

emissions factors and CURED sensitivity test emission factors are presented in

Table 11.2 to Table 11.7.

Table 11.2: Count of properties by impact descriptor

Impact Descriptor Scenario

2022 2027 2030 2035

Substantial Adverse 1 0 0 0

Moderate Adverse 106 14 27 70

Slight Adverse 315 327 378 371

Negligible 145030 145105 145047 145011

Slight Beneficial 0 6 0 0

Moderate Beneficial 0 0 0 0

Substantial Beneficial 0 0 0 0

Table 11.3: Count of properties by impact descriptor (CURED sensitivity test)

Impact Descriptor Scenario

2022 2027 2030 2035

Substantial Adverse 19 0 0 0

Moderate Adverse 186 18 16 51

Slight Adverse 387 483 438 437

Negligible 144860 144918 144961 144925

Slight Beneficial 0 30 35 39

Moderate Beneficial 0 3 2 0

Substantial Beneficial 0 0 0 0



Table 11.4: Count of properties by magnitude of change in concentration

Magnitude of Change in Annual Mean NO2 (µg/m³)

Scenario

2022 2027 2030 2035

>4.0 0 29 51 93

2.0 – 4.0 3 404 512 759

0.4 – 2.0 1701 11033 18157 43330

0.0 – 0.4 135638 132264 126318 100847

-0.4 – 0.0 8110 1552 376 367

-2.0 – -0.4 0 170 38 56

-4.0 – -2.0 0 0 0 0

<-4.0 0 0 0 0

Table 11.5: Count of properties by magnitude of change in concentration (CURED sensitivity test)


Scenario

2022 2027 2030 2035

>4.0 0 15 19 53

2.0 – 4.0 3 403 546 804

0.4 – 2.0 2017 12580 17802 39264

0.0 – 0.4 135795 130101 125918 104050

-0.4 – 0.0 7637 2125 975 988

-2.0 – -0.4 0 224 191 280

-4.0 – -2.0 0 4 1 13

<-4.0 0 0 0 0

Table 11.6: Local air quality receptors informing DCO Project significance


Total Number of Receptors

2022 2027 2030 2035

A6 B7 A B A B A B

Large (>4.0) 0 0 0 0 0 0 0 0

Medium (2.0 – 4.0) 0 0 0 0 0 0 0 0

Small (0.4 – 2.0) 33 0 0 0 0 0 0 0

6 Worsening of air quality objective already above objective or creation of a new exceedance. 7 Improvement of an air quality objective already above objective or the removal of an existing exceedance.



Table 11.7: Local air quality receptors informing DCO Project significance (CURED sensitivity test)


Total Number of Receptors

2022 2027 2030 2035

A B A B A B A B

Large (>4.0) 0 0 0 0 0 0 0 0

Medium (2.0 – 4.0) 0 0 0 0 0 0 0 0

Small (0.4 – 2.0) 104 0 3 0 1 2 0 0

11.3 Effect on EU limit value compliance results

11.3.1 The maximum predicted total annual mean NO2 concentrations along the key PCM

links within the CAQOAA, using both the EFT and the CURED sensitivity test, are

presented in Table 11.8 to Table 11.10.



Table 11.8: NO2 annual mean EU limit value compliance in 2022

Census ID

Road Annual Mean NO2 Concentration (µg/m3)

Future Baseline With DCO Project

EFT Emissions CURED Emissions

PCM151 PCM172 Increment PCM15 PCM17

Increment

PCM15 PCM17

6123 A4 Bath Road East of Waggoners Roundabout 32.7 32.4 0.3 33.0 32.7 0.3 33.0 32.8

16112 A4 Bath Road (Heathrow Boulevard to Nobel Drive)

37.4 37.8 0.9 38.3 38.7 0.9 38.3 38.7

18727 A312 between M4 and Hayes Road 44.1 44.5 0.7 44.8 45.2 0.8 44.8 45.3

26914 A312 between High Street and M4 41.1 41.0 0.2 41.3 41.2 0.2 41.3 41.2

36309 A30 past Hatton Cross 37.0 39.3 1.2 38.3 40.6 1.2 38.3 40.6

48810 A312 between Pump Lane and A4020 45.6 44.5 0.5 46.1 45.0 0.5 46.1 45.1

49028 A312 North of A4020 40.9 40.0 0.5 41.4 40.5 0.5 41.4 40.5

56686 A312 between A30 and High Street 39.2 39.7 0.2 39.4 39.9 0.2 39.4 39.9

59008 A4 near Henlys Roundabout 39.5 38.7 0.5 39.9 39.2 0.5 40.0 39.2

73633 A30 West of A312 34.4 36.4 0.0 34.4 36.4 0.0 34.4 36.4

73636 A312 between Hayes Road and Pump Lane 45.5 44.5 0.6 46.1 45.1 0.6 46.2 45.1

36013 M4 Spur 32.5 36.2 2.1 34.6 38.4 2.1 34.6 38.4

75071 A4 Great West Road East of Jersey Road 37.1 37.7 0.1 37.2 37.9 0.1 37.2 37.9

78344 A4 Brands Hill 30.3 30.4 2.4 32.7 32.8 2.5 32.9 32.9

18487 M4 Near J3 34.6 36.2 1.1 35.7 37.2 1.2 35.8 37.3

26012 M4 East of Heston Road 36.4 41.5 0.9 37.4 42.4 1.0 37.4 42.5

6013 M4 East of J4 34.2 35.5 1.2 35.4 36.7 1.3 35.5 36.8



73446 A4020 The Broadway 33.3 31.3 0.2 33.5 31.5 0.2 33.6 31.5 1 This is the predicted baseline concentration from Defra’s 2017 NO2 projections data (2015 reference year) 2 This is the predicted baseline concentration from Defra’s 2019 NO2 projections data (2017 reference year)


Census ID Road Annual Mean NO2 Concentration (µg/m3)



PCM15 PCM17 Increment PCM15 PCM17 Increment PCM15 PCM17



35.0 35.2 -3.4 31.6 31.8 -3.7 31.3 31.5





49028 A312 North of A4020 30.5 29.9 0.3 30.8 30.2 0.4 30.9 30.3



73633 A30 West of A312 27.1 28.7 0.9 28.0 29.6 1.1 28.3 29.9


36013 M4 Spur 26.3 28.5 5.3 31.6 33.7 6.6 32.9 35.1


78344 A4 Brands Hill 23.9 23.9 -0.7 23.2 23.2 -0.9 23.0 23.0



Census ID Road Annual Mean NO2 Concentration (µg/m3)




18487 M4 Near J3 26.9 27.7 -0.1 26.8 27.6 -0.2 26.7 27.5

26012 M4 East of Heston Road 27.5 30.6 -0.3 27.2 30.3 -0.4 27.1 30.2

6013 M4 East of J4 26.8 27.3 0.9 27.7 28.2 1.1 27.9 28.3

73446 A4020 The Broadway 25.1 24.4 0.4 25.6 24.9 0.5 25.6 24.9


Census ID







33.9 34.0 -4.4 29.5 29.6 -5.1 28.9 28.9





49028 A312 North of A4020 26.5 26.1 0.4 26.9 26.5 0.5 27.0 26.6



Census ID







73633 A30 West of A312 24.3 25.8 0.6 24.9 26.4 0.6 25.0 26.5


36013 M4 Spur 23.8 25.3 3.5 27.3 28.9 4.5 28.3 29.8


78344 A4 Brands Hill 21.6 21.5 -0.8 20.8 20.7 -1.1 20.4 20.3

18487 M4 Near J3 23.7 24.2 0.2 23.9 24.5 0.1 23.8 24.3

26012 M4 East of Heston Road 23.6 26.0 0.0 23.5 26.0 -0.2 23.4 25.9

6013 M4 East of J4 23.7 23.9 1.1 24.7 25.0 1.3 24.9 25.2

73446 A4020 The Broadway 22.0 22.0 0.4 22.4 22.4 0.4 22.5 22.4



12. BIBLIOGRAPHY

Full text reference In-text reference

Abbott, J. (2005). Primary nitrogen dioxide emissions from road traffic:

analysis of monitoring data. Netcen/ED48208120/R/1925/Issue 1.

Abbott, 2005

Air Quality Consultants. (2016). Emissions of Nitrogen Oxides from

Modern Diesel Vehicles. [online]. Available at:

http://www.aqconsultants.co.uk/getattachment/Resources/Download-

Reports/Emissions-of-Nitrogen-Oxides-from-Modern-Diesel-Vehicles-

210116.pdf.aspx [Accessed 17 May 2019].

AQC, 2016

Air Quality Consultants. (2017). CURED v3A. [online]. Available at:

http://www.aqconsultants.co.uk/Resources/Download-Reports.aspx

[Accessed 17 May 2019].

AQC, 2017

Air Quality Consultants. (2018). Adjusting Background NO2 Maps for

CURED v3A. [online]. Available at:



AQC, 2018c

Air Quality Consultants. (2018)., Calibrating Defra’s 2015-based

Background NOX and NO2 Maps against 2016 and 2017 Measurements.

[online]. Available at:

http://www.aqconsultants.co.uk/AQC/media/Reports/Background-

Calibration-News-F2.pdf [Accessed 17 May 2019].

AQC, 2018b

Air Quality Consultants. (2018a)., Development of the CURED v3A

Emissions Model. [online] Available at:



AQC, 2018a

CERC. (2016). London Urban Canopy Data. [online]. Available at:

http://www.cerc.co.uk/IJARSG2016 [Accessed 17 May 2019].

CERC, 2016

Defra. (2016). Local Air Quality Management Technical Guidance (TG16).

[online]. Available at: https://laqm.defra.gov.uk/technical-guidance/


Defra, 2016

Defra. (2017). 2017 NO2 projections data (2015 reference year). [online]

Available at:https://uk-air.defra.gov.uk/library/no2ten/2017-no2-projections-

from-2015-data [Accessed 17 May 2019].

Defra, 2017b

Defra. (2017). Emissions Factors Toolkit. [online] Available at:

https://laqm.defra.gov.uk/review-and-assessment/tools/emissions-factors-

toolkit.html [Accessed 17 May 2019].

Defra, 2017a

Defra. (2019). 2019 NO2 projections data (2017 reference year). [online].

Available at: https://uk-air.defra.gov.uk/library/no2ten/2019-no2-pm-

projections-from-2017-data [Accessed 17 May 2019].

Defra, 2019

Department for Transport. (2006). Project for the Sustainable Development

of Heathrow. Report of the Airport Air Quality Technical Panels. [online].

DfT, 2006

http://www.aqconsultants.co.uk/getattachment/Resources/Download-Reports/Emissions-of-Nitrogen-Oxides-from-Modern-Diesel-Vehicles-210116.pdf.aspx





http://www.aqconsultants.co.uk/AQC/media/Reports/Background-Calibration-News-F2.pdf

http://www.aqconsultants.co.uk/AQC/media/Reports/Background-Calibration-News-F2.pdf


http://www.cerc.co.uk/IJARSG2016

https://laqm.defra.gov.uk/technical-guidance/

https://uk-air.defra.gov.uk/library/no2ten/2017-no2-projections-from-2015-data

https://uk-air.defra.gov.uk/library/no2ten/2017-no2-projections-from-2015-data

https://laqm.defra.gov.uk/review-and-assessment/tools/emissions-factors-toolkit.html

https://laqm.defra.gov.uk/review-and-assessment/tools/emissions-factors-toolkit.html

https://uk-air.defra.gov.uk/library/no2ten/2019-no2-pm-projections-from-2017-data

https://uk-air.defra.gov.uk/library/no2ten/2019-no2-pm-projections-from-2017-data



Full text reference In-text reference

Available at:

http://webarchive.nationalarchives.gov.uk/20100513113102/http://www.dft.

gov.uk/pgr/aviation/environmentalissues/heathrowsustain/ [Accessed 17

May 2019].

Department for Transport. (2017). DfT Automatic traffic Counters Table TRA0305-0307, 2017. [online]. Available at: https://www.dft.gov.uk/traffic-counts/ [Accessed 17 May 2019].

DfT, 2017

Emisia. (no date). COPERT: The industry standard emissions calculator. [online]. Available at: https://www.emisia.com/utilities/copert/ [Accessed

17 May 2019].

Emisia, no date

Flightglobal. (2013). JP Airline Fleets International 2013/14, 47th ed. ISBN

978-1-898779-49-0.

Flightglobal, 2013

FOI. (no date). Aircraft Engine Emissions Database. [online]. Available on

request from: https://www.foi.se/en/foi/research/aeronautics-and-air-combat-simulation/environmental-impact-of-aircraft.html [Accessed 17 May 2019].

FOI, no date

ICAO. (2011). Airport Air Quality Manual, Doc 9889. [online]. Available at:

https://www.icao.int/publications/Documents/9889_cons_en.pdf [Accessed

17 May 2019].

ICAO, 2011

ICAO. (2018). Aircraft Engine Emissions Databank, version 24. [online].

Available at: https://www.easa.europa.eu/document-library/icao-aircraft-

engine-emissions-databank [Accessed 17 May 2019].

ICAO, 2018

Kinsey, J. and Wayson, R.L. (October 2009)., Appendix C PM methodology

discussion paper. In: G Ratliff et al., Aircraft Impacts on Local and

Regional Air Quality in the United States. PARTNER Project 15 final report.

PARTNER-COE-2009-002.

Kinsey and Wayson, 2009

Official Journal, Directive 2004/26/EC of the European Parliament and of

the Council of 21 April 2004 amending Directive 97/68/EC on the

approximation of the laws of the Member States relating to measures

against the emission of gaseous and particulate pollutants from internal

combustion engines to be installed in non-road mobile machinery., (2004).

European Parliament, 2004

Ricardo. (2018). Heathrow Airport 2017 Emission Inventory. ED11486 Issue

Number 1.

Ricardo, 2018

Underwood, B.Y., Walker, C.T. and Peirce, M.J. (July 2010a). Heathrow

Airport Emission Inventory 2008/9. AEAT/ENV/R/2906 Issue 1.

Underwood et al., 2010a

Underwood, B.Y., Walker, C.T. and Peirce, M.J. (July 2010b). Air Quality

Modelling for Heathrow Airport 2008/9: Methodology. AEAT/ENV/R/2915

Issue 1.

Underwood et al., 2010b

Underwood, B.Y., Walker, C.T. and Peirce, M.J. (July 2010c). Heathrow

Airport Air Quality Modelling for Heathrow Airport 2008/9: Results and

Model Evaluation. AEAT/ENV/R/2948 Issue 1.

Underwood et al., 2010cb

http://webarchive.nationalarchives.gov.uk/20100513113102/http:/www.dft.gov.uk/pgr/aviation/environmentalissues/heathrowsustain/

http://webarchive.nationalarchives.gov.uk/20100513113102/http:/www.dft.gov.uk/pgr/aviation/environmentalissues/heathrowsustain/

https://www.dft.gov.uk/traffic-counts/

https://www.dft.gov.uk/traffic-counts/

https://www.emisia.com/utilities/copert/

https://www.foi.se/en/foi/research/aeronautics-and-air-combat-simulation/environmental-impact-of-aircraft.html

https://www.foi.se/en/foi/research/aeronautics-and-air-combat-simulation/environmental-impact-of-aircraft.html

https://www.easa.europa.eu/document-library/icao-aircraft-engine-emissions-databank

https://www.easa.europa.eu/document-library/icao-aircraft-engine-emissions-databank