Vehicle Emission Measurement and Analysis - Aberdeen City ... · Dr James Tate Remote Sensing Vehicle Emissions - ABERDEEN ... Aberdeen (April 2015). If Clean Air Zones (CAZs) are

Project Report:

Vehicle Emission Measurement and Analysis

- Aberdeen City Council

Final version 1.1: 7th April 2016

Dr James Tate

Remote Sensing Vehicle Emissions - ABERDEEN

INSTITUTE FOR TRANSPORT STUDIES

York LEZ Feasibility Study – Vehicle Emission Modelling

i

REMOTE SENSING VEHICLE EMISSIONSa) King Street (A956)

b) Bridge of Dee (A90)

Vehicle Emission Measurement - ABERDEEN

ii

EXECUTIVE SUMMARY

Key results and findings

The Real Driving Emission (RDE) performance of 24 036 vehicles were measured inAberdeen over 9 week-days in April 2015 (14th – 17th, 20th – 24th). The database of RDEmeasurements, vehicle operation (speed and acceleration) and registration information(date first registered, marque and model, fuel type, Euro standard and engine specification/performance) comprised 79.59% Cars, 2.07% Taxis, 14.57% LGV (vans), 2.14% OGVs(rigid and articulated Goods Vehicles), 1.14% Buses (PSVs) and 0.49% Coaches;

Diesel vehicle NOX emission controls underperform in Aberdeen, as in most Europeanurban environments;

Diesel cars and vans contribute most of the primary nitrogen dioxide (NO2);

Emissions of air quality pollutants from modern (Euro 4, 5 and 6) petrol and petrol-hybridcars are low. Ultra-Low Emission (ULEZ) and Clean Air Zones restrictions should notrestrict the movement of these vehicles as their emissions of air quality pollutants are at alow-level;

Particle emissions (PM – particulate matter) are high from older diesel vehicles, whetherlight- (car and LGV) or heavy-duty (OGV, Bus, Coach). PM emissions from newer Euro5/V and 6/VI diesel vehicles are significantly lower indicating Diesel Particle Filters(DPFs) are effective;

Seven hundred and eighty two (782) Euro 6/VI vehicles were observed, comprised of 376petrol cars, 362 diesel cars and 44 OGVs. Diesel car NOX emissions were roughly halfthose of previous Euro standards (5 and older). NOX emissions from Euro VI OGVs werean order of magnitude (> 10 times) lower than previous generations. This demonstrates thatthe heavy-duty Euro VI regulations that now includes Real Driving Emissions and availabletechnology e.g. SCR (selective catalytic reduction) exhaust after-treatment, has led to asignificant improvement in the NOX performance of OGVs. The NOX after-treatmentsystems on Euro VI are so effective, NOX emissions from a Euro VI 40 ton articulated OGVand a Euro 6 diesel car are on average, at the same level (emission rate, grams.km-1);

One hundred and fifteen hybrid vehicles were observed. The majority were petrol-hybridcars (Euro 4, 5 and 6) whose emissions were very low;

NOX emissions from the twenty three pass-bys (multiple measurements of 11 uniquevehicles) by Euro IV hybrid Buses (all Alexander Dennis Enviro 350 hybrid, first registered2012/ 2013) were relatively low, roughly a quarter of comparative purely diesel powertrainvehicles;

Two diesel-hybrid cars were observed. These had high NOX emissions. Although a verylimited sample, these results suggest the additional degree of freedom the hybrid powertrainoffers has been used to minimise the official, type approval CO2 figures not lower emissionsof air quality pollutants;

LGV (vans) emissions of NOX per unit of fuel burnt are higher than from diesel passengercars. This is expected as there is a high degree of commonality between the engine andemission control technologies (powertrain) of passenger cars and vans, albeit with vansbeing larger (frontal area) and heavier vehicles. With the introduction of Euro 6 emissionstandard legislation for LGVs lagging cars by at least 12-months, the UK and much ofEurope will continue to suffer from relatively high NOX emissions from the generation ofEuro 5 and older LGVs for many years yet. No Euro 6 LGVs were observed in surveys inAberdeen (April 2015). If Clean Air Zones (CAZs) are implemented that restrict access to


iii

Euro 5 and older diesel vehicles, LGV owners and operators will have very limited optionsto purchase Euro 6 compliant vehicles i.e. new registrations;

The estimated Emission Factors (EFs – grams.km-1) from the remote sensing measurementshave been compared with modelled predictions. Modelled results using local (Aberdeen),real speed profiles and a validated Instantaneous Emission Model (IEM – leading EuropeanModel of its type, PHEM) are broadly of the same magnitude and follow the same trends.The fact that the EFs from two approaches using local information, state-of-the-artinstruments and modelling techniques are in close agreement suggests they are robust;

The Autumn/ Winter 2015 “Dieselgate” scandal widened appreciation of the growing gapbetween passenger car emission performance in the legislated test conditions (typeapproval) and in real-driving. These remote sensing measurements offer a uniqueopportunity to contribute to the debate by comparing the RDE NOX performance ofthousands of diesel passenger cars of common marques, models and Euro standards.Although many of the Volkswagen Euro 5 cars are now known to have a “defeat device”(software) that detects when the vehicle is being tested in laboratory conditions over theNEDC (type approval) drive cycle, calling “alternative” engine management and exhaustafter-treatment settings, their measured NOX emissions in real-driving conditions inAberdeen were on average no better or worse than from other marques/ manufacturers.There was a surprising degree of consistency across common marques and models, withthe exception of Vauxhall (UK sister brand of Opel) whose Euro 5 cars were observed toemit ≈40% more NOX than the average diesel car. Vauxhall (Opel) have publically deniedtheir cars have a “defeat device”. It is therefore supposed their diesel cars are increasinglyoptimised i.e. engine management and EGR settings, to emit relatively low amounts of NOX

in type approval test conditions (to meet the standards), but are increasingly disregardingNOX controls in real-driving to improve performance and fuel economy.

Figure | The Real Driving NOX emission performance of common marques/ models of dieselpassenger cars in Aberdeen (April 2015)


iv

ACKNOWLEDGEMENTS

Aberdeen City Council funded the on-road vehicle emission measurement and analysis study.Aberdeen City Council Principal Environmental Officer Aileen Brodie provided comments ona draft version of the report. The remote sensing instrumentation (RSD-4600), support vehicleand equipment was purchased using SRIF2 infrastructure funds (HEFCE SRIF Award, Code:Earth3, LANTERN Phase II).

ITS PhD students Christopher Rushton and David Wyatt delivered the Remote Sensing andVehicle Tracking surveys.

TABLE OF CONTENTS

EXECUTIVE SUMMARY ii

ACKNOWLEDGEMENTS iv

1. Introduction 1

2. Background and purpose 2

3. Materials and Method 5

3.1 ESP Remote Sensing Detector (RSD-4600) operation 5

3.2 Study Sites 7

3.3 Data collection 8

3.4 Vehicle tracking surveys 9

4. Results 11

4.1 Fleet composition 11

4.2 Passenger car characteristics 20

4.3 Emission Measurements 23

4.3.1 Emission Measurement Results – Passenger cars 24

4.3.2 Emission Measurement Results – Taxis 28

4.3.3 Emission Measurement Results – LGVs 28

4.3.4 Emission Measurement Results – Buses and Coaches 33

4.3.5 Emission Measurement Results – OGVs 38

4.4 Passenger car Emission Factors (EFs) 42

4.5 Passenger Car Emission Distributions 46

4.6 Speed, Acceleration and Engine Power demands across the City of Aberdeen –

comparison with the five Vehicle Emission Measurement sites 49

4.7 Instantaneous Emission Modelling (IEM) – Passenger cars 50

4.7.1 The Instantaneous Emission Model PHEM 51

4.7.2 PHEM results 57


ii

4.7.3 Comparison of RSD, PHEM and EFT NOX Emission Factors 60

5. Recommendations for Future Work 61

6. References 62


iii

TABLE OF FIGURES

FIGURE 1: VEHICLE EMISSIONS MEASUREMENT SYSTEM (RSD-4600) SCHEMATIC................................................ 5TABLE 1. SURVEY SITES............................................................................................................................................ 7FIGURE 2: ABERDEEN MEASUREMENT SITES. ......................................................................................................... 7TABLE 2: OVERVIEW OF THE INFORMATION FIELDS IN THE DVLA REGISTRATION DATABASE............................... 8TABLE 3: DATA COLLECTION OVERVIEW................................................................................................................. 8FIGURE 3: VEHICLE TRACKING ROUTES ACROSS ABERDEEN................................................................................... 9TABLE 4: VEHICLE TRACKING DATA COLLECTION.................................................................................................. 10TABLE 5: EMISSION STANDARD DATES FOR CAR AND LIGHT-GOODS VEHICLE CATEGORIES (E.G. LATEST EU

DIRECTIVE FOR EURO 5 AND 6 - 2007/715/EC) ........................................................................................... 11TABLE 6: VEHICLE NUMBER SAMPLES BY VEHICLE TYPE AND YEAR OF FIRST REGISTRATION. ............................ 12TABLE 7: OBSERVED VEHICLE FLEET PROPORTIONS (%): VEHICLE/ FUEL TYPE AND EURO STANDARD. .............. 13FIGURE 4: OBSERVED PASSENGER CAR FLEET PROPORTIONS, BROKEN DOWN BY EURO STANDARD AND FUEL

TYPE. ............................................................................................................................................................ 14FIGURE 5: VEHICLE AGE DISTRIBUTIONS OF THE OBSERVED PASSENGER CARS: (A) PETROL, (B) DIESEL, (C)

HYBRID AND (D) DIESEL HYBRID CARS......................................................................................................... 15FIGURE 6. PROPORTION (%) OF DIESEL AND PETROL CARS FIRST REGISTERED AFTER 1ST SEPTEMBER 2014 THAT

ARE EURO 6 FOR COMMON MARQUES....................................................................................................... 16FIGURE 7: VEHICLE AGE DISTRIBUTIONS OF THE OBSERVED TAXI FLEET: (A) “A” LICENSE, (B) “PH” LICENSE, AND

(C) “T” LICENSE. ........................................................................................................................................... 17FIGURE 8: VEHICLE AGE DISTRIBUTIONS OF THE OBSERVED LGVS CLASS N1-I, N1-II AND N1-III......................... 17FIGURE 9: VEHICLE AGE DISTRIBUTIONS OF THE OBSERVED BUS AND COACH FLEETS........................................ 18FIGURE 10: VEHICLE AGE DISTRIBUTIONS OF THE HEAVY-DUTY FLEET: BUS, COACH AND HGVS. ....................... 19TABLE 8: OBSERVED VEHICLE TYPES AND FUEL TYPE PROPORTIONS (%) AT THE SURVEY LOCATIONS. .............. 19FIGURE 11: THE DISTRIBUTION OF THE PREDICTED CO2 PER KM FOR THE OBSERVED PASSENGER CARS,

CATEGORISED BY FUEL TYPE AND EURO STANDARD. ................................................................................. 20FIGURE 12: THE DISTRIBUTION OF REPORTED VEHICLE KERB WEIGHTS (MASS IN SERVICE) FOR THE OBSERVED

PASSENGER CARS CATEGORISED BY FUEL TYPE AND EURO STANDARD. .................................................... 21FIGURE 13: THE DISTRIBUTION OF FRONTAL AREAS OF OBSERVED PASSENGER CARS CATEGORISED BY FUEL

TYPE AND EURO STANDARD........................................................................................................................ 21FIGURE 14: THE ENGINE CAPACITY DISTRIBUTIONS OF THE OBSERVED PASSENGER CARS CATEGORISED BY FUEL

TYPE AND EURO STANDARD........................................................................................................................ 22TABLE 9: RECOMMENDED VEHICLE NO2 FRACTIONS. ........................................................................................... 24FIGURE 15: THE DISTRIBUTION OF THE FUEL BASED PASSENGER CAR EMISSION MEASUREMENTS CATEGORISED

BY FUEL TYPE AND EURO STANDARD (A) NO, (B) NO2, (C) NOX................................................................... 26FIGURE 16: THE DISTRIBUTION OF THE FUEL BASED PASSENGER CAR EMISSION MEASUREMENTS CATEGORISED

BY FUEL TYPE AND EURO STANDARD (A) CO, (B) HC, (C) PM (INDEX)......................................................... 27FIGURE 17: THE DISTRIBUTION OF THE FUEL BASED TAXI EMISSION MEASUREMENTS CATEGORISED BY VEHICLE

CLASS (CAR, LGV), FUEL TYPE AND EURO STANDARD (A) NO, (B) NO2, (C) NOX.......................................... 29FIGURE 18: THE DISTRIBUTION OF THE FUEL BASED TAXI EMISSION MEASUREMENTS CATEGORISED BY VEHICLE

CLASS (CAR, LGV), FUEL TYPE AND EURO STANDARD (A) CO, (B) HC, (C) PM (INDEX) ................................ 30FIGURE 19: THE DISTRIBUTION OF THE FUEL BASED LGV EMISSION MEASUREMENTS, COMPARED WITH DIESEL

CARS, CATEGORISED BY LGV CLASS, FUEL TYPE AND EURO STANDARD (A) NO, (B) NO2, (C) NOX.............. 31FIGURE 20: THE DISTRIBUTION OF THE FUEL BASED LGV EMISSION MEASUREMENTS, COMPARED WITH DIESEL

CARS, CATEGORISED BY LGV CLASS, FUEL TYPE & EURO STANDARD (A) CO, (B) HC, (C) PM INDEX .............. 32FIGURE 21: THE DISTRIBUTION OF THE FUEL BASED PSV EMISSION MEASUREMENTS CATEGORISED BY BUS /

COACH AND EURO STANDARD (A) NO, (B) NO2, (C) NOX............................................................................. 34FIGURE 22: THE DISTRIBUTION OF THE FUEL BASED PSV EMISSION MEASUREMENTS CATEGORISED BY BUS /

COACH AND EURO STANDARD (A) CO, (B) HC, (C) PM INDEX......................................................................... 35FIGURE 23: THE MEASURED EMISSION RATIOS FOR INDIVIDUAL BUSES ORDERED BY AGE OF FIRST

REGISTRATION (A) NOX, (B) PM10 (INDEX) ................................................................................................... 36FIGURE 24: THE MEASURED NOX EMISSION RATIOS FOR INDIVIDUAL MARQUES/ MODELS OF BUS. ................. 37FIGURE 25: THE MEASURED PM EMISSION RATIOS FOR INDIVIDUAL MARQUES/ MODELS OF BUS ................... 37


iv

FIGURE 26: THE DISTRIBUTION OF THE FUEL BASED OGV EMISSION MEASUREMENTS CATEGORISED BYARTICULATED / RIGID AND EURO STANDARD (A) NO, (B) NO2, (C) NOX...................................................... 39

FIGURE 27: THE DISTRIBUTION OF MEASURED EURO VI OGV NOX MEASUREMENTS, COMPARED WITHTRANSPORT FOR LONDON (2015) EURO VI LABORATORY MEASUREMENTS OVER A REAL LONDON SPEEDPROFILE. THE LABORATORY MEASUREMENTS INCLUDE EMPTY AND FULLY-LADEN RUNNING. ................ 40

FIGURE 28: THE DISTRIBUTION OF THE FUEL BASED OGV EMISSION MEASUREMENTS CATEGORISED BYARTICULATED / RIGID AND EURO STANDARD (A) CO, (B) HC, (C) PM INDEX.................................................. 41

FIGURE 29: THE DIVERGENCE BETWEEN REAL-WORLD AND MANUFACTURERS’ TYPE-APPROVAL CO2 EMISSIONSFOR PETROL AND DIESEL CARS (ADAPTED FROM ICCT, 2015A). ................................................................. 42

FIGURE 30: THE DISTRIBUTION OF THE PREDICTED PASSENGER CAR EMISSION FACTORS (GRAMS PERKILOMETRE) CATEGORISED BY FUEL TYPE AND EURO STANDARD (A) NO2, (B) NOX................................... 43

FIGURE 31: THE DISTRIBUTION OF THE PREDICTED PASSENGER CAR EMISSION FACTORS (GRAMS PERKILOMETRE) CATEGORISED BY FUEL TYPE AND YEAR OF REGISTRATION (A) PETROL NO2, (B) DIESEL NO2,(C) PETROL NOX, (B) DIESEL NOX .................................................................................................................. 44

FIGURE 32: THE PASSENGER CAR EMISSION CONTRIBUTIONS CATEGORISED BY FUEL TYPE AND EUROSTANDARD (A) NO2, (B) NOX, (C) CO, (D) PM10. ........................................................................................... 45

TABLE 10: SUMMARY OF PASSENGER CARS IDENTIFIED AS HAVING ‘ELEVATED’ EMISSIONS. ............................ 47TABLE 11: THE EMISSION CONTRIBUTION FROM PASSENGER CARS IDENTIFIED AS BEING ‘HIGH-EMITTERS’ .... 47FIGURE 33: DENSITY DISTRIBUTION PLOTS OF THE ESTIMATED NOX EMISSION FACTORS OF PASSENGER CARS

CATEGORISED BY FUEL TYPE AND EURO CATEGORY: (A) PETROL, (B) DIESEL............................................. 48FIGURE 34: DENSITY DISTRIBUTION PLOTS OF THE ESTIMATED PM10 EMISSION FACTORS OF PASSENGER CARS

CATEGORISED BY FUEL TYPE AND EURO CATEGORY: (A) AND (C) PETROL, (B) AND (D) DIESEL. ................ 48FIGURE 35: HISTOGRAM OF THE (RAW) OBSERVED PM10 / CO2 EMISSION RATIO MEASUREMENT FOR EURO 5

AND 6 DIESEL PASSENGER CARS.................................................................................................................. 49FIGURE 36: FREQUENCY DIAGRAMS OF THE OBSERVED SPEED AND ACCELERATIONS DISTRIBUTIONS FROM THE

VEHICLE TRACKING SURVEYS (ALL) AND THOSE RECORDED BY THE VEHICLE EMISSION MEASUREMENTSYSTEM (VEMS). .......................................................................................................................................... 50

FIGURE 37. AVERAGE SPEED (MODELLED) IN EACH SIMULATION HOUR ............................................................. 52TABLE 12. THE LONDON DRIVE CYCLE STATISTICS. ............................................................................................... 52TABLE 13. PASSENGER CAR SPECIFICATIONS TESTED OVER THE LDC................................................................... 53TABLE 14. AVERAGE ABERDEEN PASSENGER CAR SPECIFICATIONS. .................................................................... 54FIGURE 38. SCATTER PLOTS COMPARING MODELLED (PHEM) AND OBSERVED VALUES FOR ALL SECTIONS

(URBAN, SUBURBAN AND MOTORWAY) OF THE LDC. ................................................................................ 56APPENDIX A. SUMMARY ESTIMATED PASSENGER CAR CO2 AND NOX, EMISSION FACTORS

(GRAMS.KM-1): REMOTE SENSING, PHEM (IEM), EFT AND EU EMISSION STANDARDS.... 65APPENDIX B1. FREQUENCY DIAGRAMS OF THE OBSERVED SPEED AND ACCELERATION

DISTRIBUTIONS FROM THE VEHICLE TRACKING ROUTES 1 TO 15 (SEE FIGURE 3 ANDTABLE 4).................................................................................................................................................... 66

APPENDIX B2. FREQUENCY DIAGRAMS OF THE OBSERVED SPEED AND ACCELERATIONDISTRIBUTIONS FROM THE 5 VEHICLE EMISSION REMOTE SENSING MEASUREMENTSITES (SEE TABLE 1): (1) KING STREET (2) BRIDGE OF DEE (3) BEACH BOULEVARD (4)WESTBURN ROAD AND (5) GREAT NORTH ROAD........................................................................ 67

1

1. Introduction

The University of Leeds, Leeds, LS2 9JT (“the University”) acting through TheInstitute for Transport Studies (ITS) was commissioned by Aberdeen City Council todeliver a 10-day Vehicle Emission remote sensing Measurement campaign, supplyingthe instrument and ancillaries/ support vehicle, Project Management, operators andanalysis/ interpretation of the results.

The surveys were undertaken on nine weekdays in school term-time between Tuesday14th April and Friday the 24th of April 2015. The instrument was operated at five sites:

Beach Boulevard; Bridge of Dee (Stonehaven Road, A90); Great Northern Road (A96); King Street (A956); and Westburn Road (A944);

The University of Leeds tasks/ objectives:

Identify and agree five suitable Vehicle Emission Measurement System(VEMS) sites in / adjacent to the Aberdeen AQMAs;

Deliver 10 days of data collection between 0800-1800hrs (weather permitting).This includes instrumentation/ support vehicle hire, operators and time fornumber plate entry/ data checks;

Source annoymised detailed Motor Vehicle registration information fromwww.CarwebUK.co.uk;

Characterise/ specify the observed on-road, operational vehicle fleetproportions, and compare with the vehicle type proportions determined from theAutomatic Number Plate Recognition (ANPR) cameras in routine operationacross the City;

Characterise the distribution of on-road vehicle fleet emissions (NOX, CO, HC,PM10), including estimates of primary NO2 emissions, classified by vehicle type(Car, Taxi, LGV, OGVs, Bus), age, fuel type and emission standard (e.g. Euro0 - 6);

Identify the abundance and significance of high-emitting vehicles;

Quantify the discrepancies between: the on-road emission measurements,available emission factors/ inventories and Euro emission standards; and

Equip and deploy an Instrumented car survey to map the speed/ acceleration/engine power demands across the wider Aberdeen network; and

Data analysis, reporting, project management and press liaison time.

The majority of plots in this report were produced using the open-source software ‘R’(R. Team, 2016). This report includes sections:

2. Background and purpose.

3. Materials and method.

4. Results.

5. Recommendations for Future work.


2

2. Background and purpose

The 60th anniversary of the Great Smog of London in 1952 has passed, which resultedin 4,000 deaths across the capital. The same number of people still die each year inLondon from air pollution. The total number of UK deaths in 2008 due to poor airquality has been put at 30,000. Although the health impacts and costs, estimated to be£8-20 billion per annum are almost twice those of physical inactivity, it fails to receivethe same level of attention. Under closer scrutiny the health evidence is strengthening,with the World Health Organization classifying diesel engine exhaust as carcinogenicto humans in June 2012 (WHO, 2012).

Road transport is the main source of the pollution in UK urban areas. The ever morestringent EU vehicle emission standards were perceived to deliver cleaner air, but levelsof a key pollutant in our busy streets haven’t been falling. As concentrations of thepollutant in question, Nitrogen Dioxide (NO2), are often above EC air quality standards(limit values) in European urban areas, nations are exposed to potential infraction finesfor non-compliance with EU law. So why hasn’t air quality been improving, whentraffic levels have been relatively stable in the central areas of UK Cities since the1990’s?

Modern diesel vehicle emission controls underperform in urban driving conditionswhen exhaust gases from the engine are relatively cool, inhibiting the operation ofcatalysts and filters. Such stop-start traffic motions are common place in the streets ofour towns and cities, but aren’t adequately represented in the legislated vehicle emissionstandard test conditions. Recent UK research that surveyed the emission performanceof large numbers of vehicles on the road has highlighted the deficiencies in dieselvehicle emission controls in urban driving conditions (Carslaw et al, 2011). The oxidesof nitrogen emission performance of diesel cars and vans in urban driving conditionshave been shown to have changed little in the past 15 years. So a brand new diesel carand one that has been driven for over 10 years, in urban driving conditions, emit similaramounts of a critical pollutant. Worryingly from a local air quality perspective dieselcars are more popular than ever. In 2010 sales of diesel cars overtook those with petrolengines for the first time (SMMT, 2011).

European Commission and UK policies are encouraging the purchase of new diesel carsover their petrol-driven counter-parts, due to their lower like-for-like carbon dioxide(CO2) emission ratings. Whilst this shift in purchasing behaviour is helping motormanufacturers meet their initial average car CO2 rating targets (gCO2/km) in-place from2012 onwards, the trade-off has been the halt in urban air quality improvements since2000-2004. Motor manufacturers face a similar trade-off between CO2 and emissionsof local air quality pollutants, but at a vehicle level as they optimize the operation ofthe engine and emission controls. Motor manufacturers are complying with theemissions legislation by developing exhaust after-treatment technologies andconfiguring their operation for the test conditions. The accelerations in the artificial1970’s test cycle still in use are however slight in comparison with those of “normal”driving. Real-world driving with prompter accelerations demands more power from theengine. At higher power demands more emphasis is given to vehicle performance thanthe emissions of local air quality pollutants. As the complexity and sophistication ofengine management and exhaust after-treatment systems has increased, so has thepotential for motor manufacturers to optimize their operation for the legislated testconditions to a greater degree, at the disregard of “normal” on-road operations. There


3

are also concerns about how the performance of the complex, multi-component dieselemission control systems will degrade with time and usage.

Frustratingly although the technology and capability exists, the combination of weakoutdated European regulations and an industry intent on working to legislation, but notin the spirit of the laws, has resulted in air quality not improving as it should. Thesubstantial health and environmental implications, along with the threat of potentiallyunlimited fines from the European Commission for not achieving air quality targets,has raised the importance of air pollution once again. The problem is not the UK’s notalone; most Member States in the EU are not complying with the air quality standards.The motor industry is going to have to do much more to produce cleaner more efficientvehicles as the easy, low-cost options of encouraging a shift to diesel and reducingvehicle weight by losing the spare tyre have been taken.

The revelations that a large number of Volkswagen Euro 5 diesel passenger cars havea “cheat device” fitted (computer code) to detect when a vehicle was being tested in alaboratory has brought the story of the failure of European road transport emissionstandards and policies to the attention of the wider media and public. In laboratorycircumstances the “cheat device” invokes cleaner engine management and emissioncontrol settings. Although many researchers have widely publicised the poor NOX

performance of modern, European diesel cars in academic literature (e.g. Carslaw,Beevers and Tate et al 2011) and wider press, the fact that Volkswagen have confessedto breaking regulations in both the US and in Europe has led to intensive attention ofthe media and more recently policy makers e.g. Commons Select Committee, Vehicletype approval, 14th December 2015.

There is clearly a need to better understand the emission characteristics of vehicles onthe road with “real-world” driving conditions and behaviour. Vehicle emission RemoteSensing Device instruments measure the tailpipe emissions of vehicles as they drive-through a monitoring site. The technology works by scanning the exhaust plume trailingthe vehicle. This approach is commonly used in the U.S. and Canada in large-scale pre-screening testing for vehicle inspection and maintenance emission programs.Manufactured by ESP (http://www.esp-global.com/) the RSD-4600 instrument used inthis study is able to characterise emissions from thousands of vehicles per day. Themeasurements, when combined with detailed Vehicle Registration Information allowthe on-road vehicle fleet emissions to be characterised, broken down by vehicle type(Car, Taxi, LGV, OGV, Bus), age, fuel type, emission standard (e.g. Euro 0 - 6),marques and for more common vehicles model. This combination of RSD emissionmeasurements and vehicle registration information provides a rare opportunity to:

Study the composition of the vehicle fleet being driven on the road in detail. This isimportant as for example, modern diesel passenger cars are known to typicallycomplete more miles per year than comparable petrol cars, or indeed older dieselcars. As it is newer passenger diesel cars that are now understood to be one of themain contributors towards the increase in primary NO2 emissions, and henceworsening of roadside NO2 concentrations at heavily trafficked locations, it isimportant to develop an accurate and up-to-date knowledge of their proportions onthe road;

Study the emission characteristics of each vehicle fuel type and Euro standard onthe road. Whilst the RSD is not able to measure NO2 directly, with knowledge of avehicle’s fuel type, Euro standard, nitrogen oxide (NO) emissions andrecommended primary NO2 fraction (Grice et al, 2009, Carslaw et al, 2013), it ispossible to predict the NO2 contribution.


4

With an improved evidence base and understanding of the proportion of vehicle milesdriven by different vehicle type sub-categories (fuel type and Euro standard) andcharacterisation of on-road vehicle emissions in Air Quality Management Areas,Authorities will be able to:

Design targeted and more effective management strategies; Better specify vehicle fleet proportions in emission models; and Calibrate and validate emission models.


5

3. Materials and Method

3.1 ESP Remote Sensing Detector (RSD-4600) operation

The vehicle emission remote sensing measurements were carried out using aAccuScanTM RSD-4600 instrument supplied by Environmental Systems Products(ESP, Arizona, US) as a dedicated across-road vehicle emissions monitoring system.Individual exhaust plumes trailing vehicles are measured by casting a focused beam ofNon-Dispersive Infrared (NDIR) and Ultraviolet (UV) light across the plume. A cornercube mirror reflects the IR/UV beam back to the remote sensing detector(spectrophotometer and opacity) module. Open-path NDIR spectroscopic techniqueswere first used to measure CO vehicle emissions by Bishop et al (1989). Enhancementsto measure a portion of HC emissions (Cadle & Stephens, 1994), the addition of a high-speed UV spectrometer capable of measuring NO (Popp et al, 1997, Zhang et al, 1996)and a UV opacity meter (wavelength 230nm) to provide a PM10 proxy (index) measure(Stedman et al, 1997) followed. The RSD-4600 measurements include theconcentration ratios of NO, CO, HC and the PM10 proxy (opacity measure) to theconcentration of CO2. The measurement ratios therefore reflect the pollutant emissionsper unit of fuel used. The use of CO2 as a reference gas facilitates quantitativemeasurements of exhaust species without knowledge of the plume location or extent ofdilution. The emission measurements are supported by: a speed/ acceleration module,temperature and barometric pressure sensors, a camera system to capture a clear digitalimage of the vehicles number plate for post-processing, and a control/ data logging PC.This approach is commonly used in the US and Canada in large-scale pre-screeningtesting for vehicle inspection and maintenance emission programs (e.g. Stedman et al,1997). The basic deployment configuration is similar to the Source/Detector and MirrorBox arrangement shown below in Figure 1.

Figure 1: Vehicle Emissions Measurement System (RSD-4600) Schematic

In accordance with the manufacturer operating procedures, the remote sensing beam islocated in a position where it will intersect a significant proportion of exhaust gas, withthe beam aligned between 250 - 300 mm from the road surface. It is important thesource/ detector module (SDM) is well aligned with the corner cube mirror. Analignment laser beam and digital level are initially used to coarsely align the IR/UVbeam before a real-time beam intensity measure is used to fine-tune the setup. Thesystem was powered on for a minimum of 45 minutes prior to an initial on-site

Camera(Number plate)

Vehicle Detector(Speed andAcceleration)

Source/Detector

Mirror Box

Source

Detector

Emissions Analyser(Common

Configurations)

Camera(Number plate)

Vehicle Detector(Speed andAcceleration)

Source/Detector

Mirror Box

Source

Detector

Emissions Analyser(Common

Configurations)


6

calibration to gain thermal stability in the circuitry, source and detector elements. On-site quality assurance procedures include calibration of the SDM to the knownconcentration of gas in a reference cell that is placed (automated) in the beam path. Thecalibration is verified (an audit) by a blended calibration gas (1000ppm propane, 2%CO, 13.6% CO2, 1000ppm NO) released into the sensing beam. These calibrations andaudits were conducted hourly during normal operation. All routine calibrations werewithin normal performance tolerances.

A measurement is defined as a beam block (by a vehicle body) followed by a halfsecond of data collection. If the data collection is interrupted by another beam block,i.e. a following vehicle with a headway less than 0.5 seconds, the measurement attemptis aborted. A measurement is declared ‘valid’ when: the size of the observed CO2

emission plume is sufficient to allow emission ratios to be calculated (i.e. the maximumCO2 concentration in the measurement open-path is > 10% and the mean of 5consecutive 50Hz CO2 measurements is > 5%); a speed/ acceleration measure is presentwith the speed is in the range 5 to 60kmh-1; and a clear ‘static’ digital image of thenumber plate is captured. The 5 consecutive 50Hz measurements prior to a beam blockare considered as the background concentration for that pass-by.

As the digital (static image) camera is triggered when a sufficiently large (>10%) CO2

plume is observed, vehicles with tail-pipes located at a mid-way point on the chassis(i.e. OGVs) and a high ride height (or break in the body panels) can lead to thepremature activation of the camera. In this situation a static image is taken of the sideof the vehicle, without the number plate in field of view. Although the emissionmeasurement is ‘valid’, the number plate and therefore age/ emission category are‘unknown’. This mainly affects OGVs with tail-pipes positioned close to ground-levelbehind the driver cab unit, which also have a break in trailer/ fixed bed body-panels. Inparallel with the instruments’ digital camera, a High Definition (HD) video ‘GoPro’camera was also deployed. The high definition video was used to identify VRMs thatwere not clearly visible from the instruments ‘static’ images. The Aberdeen surveyswere the first time the ‘GoPro’ backup camera was deployed. This is considered to bean important extension, improving the capture rate of VRMs of particularly heavy-dutyvehicles.

The collection of a high proportion of ‘valid’ measurements requires:

Selected monitoring sites are restricted to single lane operation;

The optical beam path distance is limited to less than 10 m;

The majority of vehicle engines are under load as they drive through themeasurement site. This is to ensure significant emission plumes are available formeasurement. Sites are therefore recommended to have a slight uphill grade;

Weather and environmental conditions are favourable as high wind speeds rapidlydisperse exhaust plumes. The equipment is also not weather-proof, so cannot beoperated in rain or snow.

The RSD-4600 was deployment was led by Dr James Tate (ITS), supported byChristopher Rushton and David Wyatt (ITS PhD students) in accordance withguidelines in the operator’s manual, risk assessments, operator/ pedestrian/ driver safetyissues and public rights of ways. At the start of each RSD-4600 sampling session,ambient conditions and road gradients were logged.

The ‘static’ digital image of each ‘valid’ drive-through measurement was viewed off-line and the number plate (Vehicle Registration Mark – VRM) alphanumeric character


7

string added to the measurement dataset. This record was cross-referenced with thewww.carweb.co.uk UK detailed vehicle registration database, and the Aberdeen taxilicense register (21st April 2015) so the vehicle fleet composition could becharacterised, broken down by vehicle type (Car, Taxi, LGV, OGV, Bus), age, fuel typeand emission standard (e.g. Euro 0 - 6).

3.2 Study Sites

The five survey locations in the Aberdeen are illustrated in Figure 2, with thecoordinates of the sites documented in table 1.

Table 1. Survey sites

Location Coordinates Altitude (m)

Beach Boulevard 57°09'00.73"N

2°05'16.39"W

12m

Bridge of Dee(Stonehaven Road, A90)

57°07'24.57"N

2°07'14.46"W

10m

Great North Road(A96)

57°10'06.45"N

2°07'18.95"W

48m

King Street(A956)

57°09'3527"N

2°05'45.05"W

22m

Westburn Road(A944)

57°09'08.80"N

2°07'4642"W

47m

Figure 2: Aberdeen Measurement Sites.{Background ©Copyright GoogleTM 2015}


8

3.3 Data collection

The surveys were undertaken on nine consecutive weekdays in school term-timebetween Tuesday 14th April and Friday the 24th of April 2015. CarwebUK Ltd(www.carweb.co.uk) matched the recorded number plate records (VRMs) with the UKMotor Vehicle Registration Information database. The information fields madeavailable include details of the vehicle make/ model, engine specification andperformance. A summary of the Motor Vehicle Registration Information available foreach vehicle record are listed in table 2.

Table 2: Overview of the information fields in the DVLA Registration database.

Vehicle details Engine detailsPerformance/transmission Environmental details

VRM Fuel typeAcceleration to100kmh-1 Fuel Cons. - Combined

Manufacturer/ modelEnginecapacity Number forward gears Fuel Cons. - ExtraUrban

Date 1st registration No. cylinders Drive axle(s) Fuel Cons. - UrbanCold

Body style Euro status Maximum powerType approval CO2

(grams.km-1)

Country Of Origin Maximum torqueSoundLevel –EngineSpeed

Size (Height, Length,Width)

SoundLevel –DriveBy

Gross weightSoundLevel –Stationary

The number of ‘valid’ records (emission measurement and detailed vehicle informationsourced from the number plate/ VRM) for each survey date and site are presented inTable 3.

Table 3: Data Collection Overview.

Site Day Date(dd/mm/yyyy)

Time(BST, hh:mm)

No ValidRecords

King Street

(A956)Tuesday 14/04/2015 08:15 – 18:00 2525

Wednesday 15/04/2015 08:02 – 18:00 2375

Bridge of Dee

(Stonehaven Rd,A90)

Thursday 16/04/2015 08:46 – 18:00 6442

Friday 17/04/2015 08:04 – 16:41* 6687

Beach Boulevard Saturday 18/04/2015 Data logging failure 0

Westburn Road

(A944)Monday 20/04/2015 08:24 – 18:00 1957

Tuesday 21/04/2015 08:10 – 12:09* 521

Great North Road

(A96)Wednesday 22/04/2015 07:54 – 12:47* 669

Thursday 23/04/2015 08:24 – 17:00 2291

Beach Boulevard Friday 24/04/2015 08:04 – 12:00 569

TOTAL 24 036

Note: * Surveys ended early due to power supply failure. A fault with one of thecharging units was identified and repaired on 22/04/2015.


9

3.4 Vehicle tracking surveys

A Ford Fiesta Zetec ( Euro 5, 1242cc petrol, kerb weight 1490 kg, 120 grams CO2 perkm, VRM YG14 NTW ) hatchback car (Super-mini market segment) was equippedwith a VBox II Lite (www.velocitybox.co.uk) survey grade, fast up-date (10Hz) GPSand CAN interface (CAN002 module) instrument/ data-logger. Connected to the car’sCAN Bus (information network), the vehicle’s position, GPS and road (wheelrevolutions) speed, and engine speed were recorded at 10Hz. The vehicle trackingsurvey took place on Wednesday the 22nd, Thursday 24th and Friday 25th of April 2015.The times and routes covered are illustrated in Figure 3 and documented in Table 4.

Figure 3: Vehicle tracking routes across Aberdeen.{Background ©Copyright GoogleTM 2015}

The primary aim of the vehicle tracking surveys, was to assess the typical distributionof speed, acceleration and engine power demands (accounting for road grade) acrossthe Aberdeen road network (main routes), and compare this with the measurements atthe five RSD survey locations. The speed profiles were also fed into an InstantaneousEmission Model (PHEM, Hausberger et al, 2015) to establish predicted emission rates/factors for real-driving conditions across Aberdeen. These emission rates/ factors areused to verify and corroborate the RSD results.

As GPS altitude data is not reliable, particularly in urban areas with slower vehiclespeeds and restricted view of the sky, the GPS position was fitted to a Digital TerrainModel (DTM) that maps the earth's terrain excluding building and vegetation with a 5mresolution. The DTM was created by Bluesky International Limited, made available foreducation and research via the Mimas, Landmap Service(http://landmap.mimas.ac.uk/). If the vehicle is travelling above 12ms-1 then thegradient calculation is simply the distance travelled in the measured second by thechange in height in that measured second. At slower speeds an average grade over thepreceding and proceeding 5m sections is calculated helps to smooth out errors caused


10

by inaccurate GPS longitude/ latitude measurements, and therefore inaccurate heightderivation from the DTM. The method is described in Wyatt, Li and Tate (2014).

Table 4: Vehicle tracking data collection.

N DATE

(dd/mm/yyyy)START[BST](hh:mm)

END[BST](hh:mm)

ROUTEdescription

LINEcolour

(Fig 3)

AverageSPEED

(km.h-1)DISTANCE

(km)

1 22/04/2015 14:36 14:45 GNR 27.4 4.25

2 22/04/2015 14:47 14:55 GNR 34.5 4.27

3 22/04/2015 15:34 16:35 LOOP 16.2 16.50

4 23/04/2015 08:18 08:59 WESTBURN 20.7 14.35

5 23/04/2015 09:05 09:39 CITYCENTRE 17.9 10.09

6 23/04/2015 11:20 11:59 GNR2GWR 21.5 13.83

7 23/04/2015 12:04 12:30 GWR2GNR 19.5 8.46

8 23/04/2015 12:53 13:07 SMALL LOOP 23.6 5.38

9 23/04/2015 13:11 13:40 BoDON 14.1 6.90

10 23/04/2015 14:58 15:52 REVERSE LOOP 20.7 18.48

11 23/04/2015 16:02 16:52 CITYCENTRE2 1.5 1.23

12 24/04/2015 08:14 08:32 BEACH 31.9 9.48

13 24/04/2015 08:37 09:18 WELLROAD 31.4 21.26

14 24/04/2015 10:42 11:28 HAUDIGAN 25.9 19.71

15 24/04/2015 11:45 12:14 CENTRE EAST 10.6 5.22

GRAND TOTAL 159.4 km


11

4. Results

4.1 Fleet composition

The detailed vehicle registration information allows the on-road vehicle fleetproportions to be studied in detail, including: vehicle type, age, Euro emission standardand fuel type. Vehicle emission standards are defined in a series of European UnionDirectives (98/69/EC and 1999/96/EC) staging the progressive introduction ofincreasingly stringent standards. Currently, emissions of NOX, HC, carbon monoxide(CO), and particulate matter are regulated for most vehicle types. For each vehicle typedifferent standards apply. Compliance is determined by running the vehicle on astandardised drive cycle (speed profile). Noncompliant vehicles cannot be sold in theEU. No use of specific technologies is mandated to meet the standards, though availabletechnology is considered when setting the standards. The stages are typically referredto as Euro 1, 2, 3, 4, 5 and 6 for light-duty vehicles; Euro I, . The emission standardscome into force on a set date for new type approvals. For example Euro 4 emissionlimits for cars came into force on 1st January 2005. New car models first registered afterthis date should therefore comply with the Euro 4 emission standard. Several MemberStates, including the UK have used tax incentives to encourage motor manufacturers toaccelerate the introduction of cleaner vehicles. Table 5 presents the dates Euro emissionstandards came into effect for passenger cars and Light-Goods Vehicle categories(LGVs).

Table 5: Emission standard dates for Car and Light-Goods Vehicle Categories (e.g.Latest EU directive for Euro 5 and 6 - 2007/715/EC)

Vehicletype

Fueltype

Euroclass.

Dates Emission standards (g/km)

CO NOx HC HC+NOx PM

Car Diesel 1 1992-1995 2.72 - - 0.97 0.142 1996-1999 1.0 - - 0.7 0.083 2000-2004 0.64 0.5 - 0.56 0.054 2005-Aug2009 0.5 0.25 - 0.3 0.0255 Sep2009 –

Aug20140.5 0.18 - 0.23 0.005

6 Sep2014 – 0.5 0.08 - 0.17 0.005Petrol 1 1992-1995 2.72 - - 0.97 -

2 1996-1999 2.2 - - 0.5 -3 2000-2004 2.3 0.15 0.2 - -4 2005-Aug2009 1.0 0.08 0.1 - -5 Sep2009 –

Aug20141.0 0.06 0.1 - 0.005

6 Sep2014 – 1.0 0.06 0.1 - 0.005LGV(1760-3500kg)

Diesel 1 1994-1997 6.9 - - 1.7 0.252 1998-2000 1.5 - - 1.2 0.173 2001-2005 0.95 0.78 - 0.86 0.104 2006-Aug2010 0.74 0.39 - 0.46 0.065 Sep2010 –

Aug20150.74 0.28 - 0.35 0.005

6 Sep2015 – 0.74 0.125 - 0.215 0.005


12

Table 6 presents the vehicle number samples for each vehicle type surveyed by theRemote Sensing instrument, split by year of first registration. This database of ‘valid’records comprised 79.59% Cars, 2.07% Taxis, 14.57% LGV, 2.14% OGVs (rigid andarticulated), 1.14% Buses (Passenger Service Vehicles – PSVs) and 0.49% Coaches.

Table 6: Vehicle number samples by vehicle type and year of first registration.

YEAR Car Taxi LGV OGV Bus Coach

Pre-1980 1 0 0 0 0 0

1980 0 0 0 0 0 0

1981 0 0 0 0 0 0

1982 1 0 0 0 0 0

1983 0 0 0 0 0 0

1984 2 0 0 0 0 0

1985 0 0 0 0 0 0

1986 1 0 0 0 0 0

1987 0 0 0 0 0 0

1988 2 0 0 0 0 0

1989 2 0 0 0 0 0

1990 2 0 0 0 0 0

1991 1 0 0 0 0 0

1992 5 0 1 0 0 0

1993 2 0 1 0 0 0

1994 5 0 0 0 0 0

1995 6 0 1 0 0 0

1996 12 0 0 0 0 0

1997 25 0 2 0 0 2

1998 39 0 3 0 0 0

1999 79 0 4 2 6 0

2000 163 0 7 1 68 0

2001 235 0 17 5 22 0

2002 413 0 36 3 0 0

2003 615 1 72 18 0 0

2004 735 1 43 7 0 4

2005 894 6 110 28 77 25

2006 1065 17 148 25 5 10

2007 1275 19 164 42 15 16

2008 1240 32 231 46 3 5

2009 1329 39 174 31 0 50

2010 1521 59 237 35 10 1

2011 1702 91 401 50 0 4

2012 2026 89 442 59 25 3

2013 2404 106 595 89 24 0

2014 2535 32 658 62 18 0

2015 783 4 153 12 0 0

TOTAL 19120 496 3500 515 273 120

13

Table 7: Observed vehicle fleet proportions (%): vehicle/ fuel type and Euro standard.

EUROStandard

CAR Taxi LGV OGV Bus Coach

Pe

tro

l

Die

sel

Hyb

rid

Hyb

rid

Die

sel

Pe

tro

l

Die

sel

Hyb

rid

Hyb

rid

Die

sel

Pe

tro

l

Die

sel

Die

sel

Die

sel

Hyb

rid

Die

sel

Die

sel

Euro 0 0.083 0.012 0.008 0.017 0.004

Euro 1 0.075 0.050 0.029 0.008

Euro 2 1.078 0.171 0.008 0.250 0.133 0.258

Euro 3 7.072 2.618 0.008 0.017 0.017 1.315 0.408 0.466 0.029

Euro 4 15.480 9.603 0.071 0.054 0.620 4.853 0.412 0.100 0.092 0.237

Euro 5 18.119 21.782 0.296 0.008 0.121 1.228 0.012 0.004 8.067 1.003 0.221 0.225

Euro 6 1.557 1.503 0.008 0.004 0.183

Total (%) 43.465 35.739 0.375 0.008 0.183 1.869 0.012 0.000 0.037 14.531

2.14

1.045 0.092

0.49GRANDTOTAL (%) 79.59 2.06 14.57 1.14

14

Table 7 specifies the observed vehicle fleet proportions, broken down by vehicle/ fueltype, and their Euro category. Figure 4 illustrates the observed passenger car fleetproportions, broken down by Euro standard and fuel type. The ‘on-road’ passenger carfleet is dominated by Euro 4 classification or better vehicles. Cars registered in 2005onwards (Euro 4 classification or better) comprise more than 86% of the ‘on-road’passenger car fleet. Whilst petrol cars currently comprise 54.6% of the operationalpassenger car fleet in Aberdeen, diesel cars are becoming more popular. In 2010 salesof diesel cars overtook those with petrol engines in the UK for the first time (SMMT,2011).

Figure 4: Observed passenger car fleet proportions, broken down by Euro standardand fuel type.

0

10

20

30

40

50

60

Euro 0 Euro 1 Euro 2 Euro 3 Euro 4 Euro 5 Euro 6

Pas

sen

ger

car

fle

et

(%)

Euro Standard

Petrol Diesel

Hybrid-Petrol Hybrid-Diesel

Dieselshare13.0%

Dieselshare13.7%

Dieselshare40.0%

Dieselshare27.0%

Dieselshare54.2%

Dieselshare48.0%

Dieselshare38.2%


15

Figure 5 presents the age distribution (histogram) of the observed petrol, diesel, hybrid(petrol-electric) and diesel-hybrid cars. The number of registered CNG and LPGpassenger cars is low. The popularity of Gas bi-fuel vehicles that promised to be a lowcarbon and emission technology appears to have waned. Only 12 cars were registeredas CNG/ LPG and it is expected many of these may not be running on gas, so have notbeen explicitly analysed. There are early signs that the hybrid (petrol-electric and morerecently diesel-electric powertrains) car market is beginning to flourish.

ABERDEEN 2015 – Passenger cars

(a) Petrol (b) Diesel

(c) Hybrid (d) Diesel Hybrid

Figure 5: Vehicle age distributions of the observed passenger cars: (a) petrol, (b)diesel, (c) hybrid and (d) diesel hybrid cars.

Although the passenger car Euro 6 legislation came into force in 1st September 2014,not all cars sold after this date need to comply. The legislation has recently (October2015) been derogated by 12 months to give manufacturers more time to develop andproduce Euro 6 compliant vehicles. The regulations are now that any new modelsintroduced after 1st September 2015 must meet Euro 6 standards. Individual vehiclesalready on sale that were built by, and dispatched from, the manufacturer before 1st June

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

500

1000

1500

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

500

1000

1500

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

5

10

15

20

25

30

Densi

ty

1980

1985

1990

1995

2000

2005

2010

2015

0

5

10

15

20

25


16

2015 can continue to be sold until 1st September 2016. However, the manufacturermust apply for derogation in these instances. Manufacturers are responding differentlyto the legislation with some introducing and selling cars ahead of the legislative date,others waiting until the regulations force them to supply Euro 6 compliant vehicles.

As the remote sensing and VRM look-up approach is surveying the composition (andemission performance of the active fleet), the proportion of cars Euro 6 cars on-the-road in Aberdeen can be identified. Figure 6 illustrates the (ranked) proportion ofobserved diesel and petrol cars first registered after the 1st September 2014 (date theEuro 6 legislation came into force) that are Euro 6 for common manufacturers. Germanmarques are leading the introduction of Euro 6 for both diesel and petrol passenger cars.No Euro 6 cars were observed from common marques such as Honda, Jaguar, LandRover, Mazda, Nissan and Toyota.

This analysis demonstrates that although the Euro 6 legislation was initially planned tocome into force on 1st of September 2014, it is now expected that only after 1st ofSeptember 2016 will the majority of new cars sold meet the new, more stringentemission standards. The expected benefits of cleaner Euro 6 diesel cars will thereforetake longer to infiltrate the fleet, delaying air quality improvements.

Figure 6. Proportion (%) of diesel and petrol cars first registered after 1st September2014 that are Euro 6 for common marques.

Figure 7 presents the age distribution (histograms) of the observed taxis with licensetypes “A”, “PH” and “T”. License restrictions mean the majority of taxis operating inAberdeen are less than 10 years old. The taxi fleet is comprised mainly of passengercars and small proportion Light Commercial Vehicles (2.2%). Diesel engine vehiclesdominate (92.7%), with a small proportion of petrol (6.7%) and hybrid (petrol-hybrid,0.6%) vehicles.

0

10

20

30

40

50

60

70

80

90

100

MER

CED

ES

MIN

I

PO

RSC

HE

BM

W

AU

DI

VO

LKSW

AG

EN

FOR

D

VO

LVO

HYU

ND

I

CIT

RO

EN

VA

UX

HA

LL

SEA

T

PEU

GEO

T

SKO

DA

KIA

HO

ND

A

JAG

UA

R

LAN

DR

OV

ER

MA

ZDA

NIS

SAN

TOYO

TA

%Eu

ro6

Diesel Petrol


17

ABERDEEN 2015 – Taxis

“A” license “PH” license “T” license

Figure 7: Vehicle age distributions of the observed taxi fleet: (a) “A” license, (b)“PH” license, and (c) “T” license.

Figure 8 presents the age distribution (histogram) of the observed van (LGV – LightGoods Vehicle) fleet. LGVs are pre-dominantly (>99%) diesel fuelled vehicles inAberdeen. Vans are classified as Light-Goods Vehicles (N1) with three sizes:

N1-I < 1305 kg (GVW); N1-II | 1305 kg to 1760 kg (GVW); and N1-III > 1760 kg (GVW).

The majority (69%) of LGVs detected by the remote sensing system in Aberdeen werethe larger N1-III class vehicles.

ABERDEEN 2015 – LGVsN1-I N1-II N1-III

Figure 8: Vehicle age distributions of the observed LGVs class N1-I, N1-II and N1-III

Figure 9 presents the age distribution (histogram) of the Single-decker / Double-deckerBuses and Coaches observed at the five Aberdeen measurement sites. The observedAberdeen Bus (Passenger Service Vehicles) and Coach Fleets are predominantlysingle-decker vehicles. As Buses make repeated circuits of their routes, with publictransport operators also often purchasing a fleet of identical vehicles to service acorridor/ area, their ages can be clustered around specific dates/ years. In Aberdeen themajority of “valid” emission measurements with cross-referenced VRM/ detailed

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

100

200

300

400

500

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

100

200

300

400

500

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

100

200

300

400

500


18

registration information are either ~2000 or ~2005 vintage. Coaches however are moreevenly distributed, with the majority of vehicles first registered in the 2004-2010 period.

ABERDEEN 2015 – Buses & Coaches

(a) Singledecker BUS (b) Doubledecker BUS

(c) Singledecker COACH (d) Doubledecker COACH

Figure 9: Vehicle age distributions of the observed Bus and Coach fleets.

Other Goods Vehicles (OGVs or Heavy-Goods Vehicles) all run on diesel fuel, withthe majority (>99%) of vehicles registered after 2000. More than three-quarters of theobserved OGVs met the Euro IV, V or VI emission standards. A small proportion (9%)of observed OGVs were registered as meeting the latest Euro VI standards.

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80


19

ABERDEEN 2015 – OGVs

Rigid Articulated

Figure 10: Vehicle age distributions of the heavy-duty fleet: Bus, Coach and HGVs.

The observed vehicle fleets at each of the five survey locations are documented in Table8. The fleet proportions were broadly similar at the King Street, Westburn Road andGreat North Road sites i.e. the Aberdeen arterials. Cars and LGVs dominated theheavily trafficked Bridge of Dee site, in part due to the 7’-0” width restriction.

Table 8: Observed vehicle types and fuel type proportions (%) at the survey locations.

Vehicletype

SITE

KingStreet(A956)

Bridge ofDee (A90)

WestburnRoad

(A944)

GreatNorthRoad(A96)

BeachBoulevar

d

N (%) N (%) N (%) N (%) N (%)

Car 3518 (71.9) 11007 (83.9) 1916 (77.4) 2234 (75.5) 445 (78.3)

Taxi 174 (3.6) 140 (1.1) 88 (3.6) 88 3.0 6 (1.1)

LGV 710 (14.5) 1940 (14.8) 309 (12.5) 441 (14.9) 100 (17.6)

Bus 208 (4.2) 40 (0.3) 106 (4.3) 146 (4.9) 15 (2.6)

Coach 189 (3.9) 0 (0.0) 45 (1.8) 37 (1.3) 2 (0.4)

OGV 97 (2.0) 0 (0.0) 11 (0.4) 12 (0.4) 0 (0.0)

TOTAL 4896 13127 2475 2958 568

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80

Fre

quency

1980

1985

1990

1995

2000

2005

2010

2015

0

20

40

60

80


20

4.2 Passenger car characteristics

The detailed Vehicle Registration Information (VRI) obtained includes the vehiclesspecifications (as listed in Table 2) and the reported vehicle specific (type approval)fuel consumption for the New European Drive Cycle (NEDC). The NEDC consists oftwo parts: the urban ECE (duration: 780s, maximum speed 50kmh-1) and the extra-urban EUDC (duration: 400s, maximum speed 120kmh-1). The urban ECE starts witha cold engine. The VRI includes fuel consumption figures for the combined (urban ECEand EUDC), extra-urban and urban (cold) drive cycles. The VRI fuel consumptionfigures (units - litres per 100km) have been converted into carbon dioxide (CO2)emissions per unit distance (grams per km). The distribution of carbon dioxide emissionper unit distance for each fuel type and Euro standard combination, for the NEDCcombined drive cycles is presented in the Figure 11 boxplot. The bold horizontal lineshows the median value. The bottom and top of the box show the 25th and 75thpercentiles respectively. The vertical dashed lines, termed ‘whiskers’, present the 1.5times the interquartile range of the data. Outliers are not plotted to truncate the y-axisscale. The notches drawn as a ‘waist’ on either side of the median give an indication ofthe 95% confidence interval of the median. A broad notch indicates uncertainty in themedian. A narrow (tight) notch illustrates higher confidence in the median figure.Uncertainties are higher for categories with a smaller sample size i.e. the older Eurostandards.

The fuel and CO2 efficiency of petrol and diesel passenger cars over the ‘type approval’test are progressively improving with time and the Euro standards. It is perhapssurprising that there’s little difference between the median CO2 per km for the observedpetrol and diesel cars. The observed diesel passenger cars are however typically larger,with a higher kerb weight, larger frontal area and engine capacity (illustrated in Figures12, 13 and 14 respectively).

Figure 11: The distribution of the predicted CO2 per km for the observed passengercars, categorised by fuel type and Euro standard.

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

az

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bz

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dz

zH

ybrid-D

iesel_

E5

50

100

150

200

250

300

350

400

CO

2(

gra

ms.k

m1

)


21

Figure 12: The distribution of reported vehicle kerb weights (mass in service) for theobserved passenger cars categorised by fuel type and Euro standard.

Figure 13: The distribution of frontal areas of observed passenger cars categorisedby fuel type and Euro standard.

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

az

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bz

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dz

zH

ybrid-D

iese

l_E

5

1000

1500

2000

2500

Mass

inS

erv

ice

(kg)

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

az

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bz

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dz

zH

ybrid-D

iese

l_E

5

2.0

2.5

3.0

3.5

4.0

Fro

nta

lA

rea

(m2

)


22

Figure 14: The engine capacity distributions of the observed passenger carscategorised by fuel type and Euro standard.

The median engine capacity for Euro 2 onwards petrol or diesel engine passenger carsis relatively constant at 1.5litres and 2.0litres respectively. The kerb (un-laden) weightof both the observed petrol and diesel engine passenger cars has risen steadily,increasing with each successive Euro standard. These increases in weight are driven bymany factors such as improved safety requirements and equipment (air-bags, crumplezones), higher specification vehicles and a growing demand for larger vehicles such assports utility vehicles. This is illustrated by the rising trend in frontal area. Paradoxicallyincreases in vehicle weight have also been caused by the addition of pollution controlequipment such as catalysts and diesel particle filters.

The Euro 4 hybrid-petrol cars are dominated by two contrasting model series, the:

Honda Civic Hybrid (6 vehicles with an engine capacity of1.4l, ≈1400kg); and Lexus RX400H (11 vehicles with an engine capacity of 3.3l, ≈2115kg).

Contrastingly, more than half of the Euro 5 standard hybrid-petrol cars (total 71) aremanufactured by Toyota:

Auris (22 vehicles with an engine capacity of 1.8l, ≈1480kg); Prius (16 vehicles with an engine capacity of 1.5l, ≈1460kg); and Yaris (8 vehicles with an engine capacity of 1.5l, ≈1190kg).

Current hybrid cars can be split into two broad categories: those of moderate size/performance, and larger higher powered vehicles.

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

az

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bz

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dz

zH

ybrid-D

iese

l_E

5

1000

1500

2000

2500

3000

3500

4000

Engin

ecapacity

(cm

3)


23

4.3 Emission Measurements

The Aberdeen database of ‘valid’ RSD measurements (speed, acceleration and emissionrecordings), which must also include a number plate record (VRM) successfullymatched with SMMT Vehicle Registration information, was 24 036 records (see Table3). The RSD-4600 instrument measures the NO (nitric oxide)/CO2 ratio in the exhaustplume. Although it doesn’t measure NOX and NO2 directly, with knowledge of avehicles’ fuel type, Euro standard, nitrogen oxide (NO) emissions and recommendedprimary NO2 fractions (Carslaw et al, 2014, Hausberger et al, 2011; Grice et al, 2009and Jerksjö et al, 2008, see Table 9), it is possible to predict the NOX and NO2 emissionfactors and contributions. The addition of Diesel Oxidisation Catalysts (DOCs) onlight-duty diesel vehicles from Euro 3 onwards has led to a dramatic increase inemissions of NO2 directly from vehicles (Carslaw, 2005).

Hausberger et al. (2011) results have been used in previous studies/ analysis, whichsuggest NO2 fractions of Euro 5 light-duty diesel vehicles are lower than Euro 4. Thisis because Euro 5 DOCs have a higher palladium content and are known to have betterf-NO2 characteristics (Hausberger, 2011). The NO2 fractions applied are the averagesreported by Carslaw et al (2013) who carried out vehicle emission remote sensingmeasurements in the UK using the University of Denver vehicle emission remotesensing instrument (the only instrument of its’ type and capability) that can speciateNO, NO2 and ammonia (NH3). This study did observe considerable variation withinvehicle categories and between vehicles, so NO2 and NOX assessments of individualvehicles from this study are not considered robust. The averages for vehicle sub-categories e.g. Euro 4 diesel passenger cars, are however considered to be reliable. Forcurrent petrol car technologies the f-NO2 is low. This however may not always be thecase. The emergence of ultra-lean burn engine technology may in future lead to anincrease in NOX and NO2 emissions from petrol engines.

A recently published UK remote sensing study led by Carlsaw et al (2014) concludedthat as diesel emission controls degrade with time and usage, f-NO2 characteristic of aEuro category are observed to fall. The deterioration of emission controls and emissionperformance is rarely studied by laboratories as the limited funding they manage toattract is typically targeted to assessing the performance of new vehicles andtechnologies e.g. Euro 6/ VI, rather than going back and reviewing emissionperformance after years in-service. It is these latest, UK measurements of f-NO2 thatare used in this study.

Heavy diesel (Bus, Coach and OGV) average f-NO2 values are significantly lower thanthose reported for light-duty diesels.


24

Table 9: Recommended vehicle NO2 fractions.

Vehicle class Euro class Carslaw et al(2013)

FEAT remotesensing (UK)

PHEM –Hausberger et al

(2011)

% NO2 (byvolume)

Grice et al. (2009)

% NO2 (byvolume)

Jerksjö et al.(2008)

% NO2 (byvolume)

Petrol carsAll

Euro 1 & earlierEuro 2Euro 3Euro 4Euro 5Euro 6

23471212

5 3 ≈ 1

Diesel carsEuro 1 & earlierEuro 2Euro 3Euro 4Euro 5Euro 6

8816262634

81135

42.636.3

1111305555

14 – 2014 – 2030 – 4755 – 60

VansEuro I & earlierEuro IIEuro IIIEuro IVEuro V

13.17.512.727.327.3

811354535

1111305555

14 – 2014 – 2030 – 4755 – 60

HGVsEuro I & earlierEuro IIEuro IIIEuro IVEuro V

N/a18.518.66.69.4

1.72.76.8

11.07.4

1111141010

779

1313

BusesEuro I & earlierEuro IIEuro IIIEuro IVEuro V

N/a 2.83.26.68.57

1111143510

101030

25 - 5248

4.3.1 Emission Measurement Results – Passenger cars

The distribution of the NO, predicted NO2, predicted NOX, CO, HC and PM (index)fuel based (ratio to CO2) passenger car emission measurements, categorised by Eurostandard and fuel/ technology type are presented in the Figure 15 and 16 ‘boxplots’.The NO emissions per unit of fuel used (CO2 emissions) are higher for the older petrolEuro standard categories (Figure 15a). As documented in table 9, the primary NO2 ishigh for Euro 4 and newer light-duty diesel vehicles. Figure 15(b) illustrates thedistribution of the predicted NO2 emissions per unit of fuel used, which is typically low,except for the Euro 3 – 5 diesel sub-categories. The distribution of observed/ predictedNOX emissions per unit of fuel used in Figure 15(c), illustrates:

The NOX performance per unit of fuel consumed of the Euro 3 – 5 petrol carsthat form a large part of the fleet is relatively low; and

The NOX performance of diesel cars per unit of fuel consumed is high andrelatively stable through all Euro standards, only falling for the latest generationof Euro 6 vehicles.

The CO emissions per unit of fuel used are higher for the older petrol Euro standardcategories, as illustrated in Figure 16(a). The measured CO emissions from all otherpassenger vehicle sub-types with a reasonable sample size are low in comparison. Thedistribution of HC emissions per unit of fuel consumed are similar to the CO species,


25

only being elevated for older petrol cars. The PM10 (index) proxy measurement (opacitymethod) shows that only older Euro 0 – 2 diesel cars emit relatively high amounts ofPM10. The introduction and development of diesel particle filter technologies throughEuro standards 3 to 5 has seen progressive reductions in the amount of PM emitted. ThePM10 (index) emissions from Euro 5/6 diesel vehicles are only slightly above those ofthe petrol counter-parts (Euro 5/6). The PM10 (index) measure for newer petrol cars isperhaps higher than one might expect. The instrument is scanning the exhaust plumetrailing the vehicle, which therefore may contain primary tail-pipe emissions but alsoparticles from brake/ tyre wear and those re-suspended from the road surface.


26

EMISSION MEASUREMENTS – Passenger cars(a) NO(%)/CO2(%)

(b) Predicted NO2(%)/CO2(%)

(c) Predicted NOX(%)/CO2(%)

Figure 15: The distribution of the fuel based passenger car emission measurementscategorised by fuel type and Euro standard (a) NO, (b) NO2, (c) NOX

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0

50

100

150

NO/

CO

2ra

tio

*10000

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0

10

20

30

40

NO

2/

CO

2ra

tio

*10000

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0

50

100

150

NO

X/

CO

2ra

tio*

10000


27

EMISSION MEASUREMENTS – Passenger cars(a) CO(%)/CO2(%)

(b) HC(%)/CO2(%)

(c) Particle Mass (PM) index / CO2(%)

Figure 16: The distribution of the fuel based passenger car emission measurementscategorised by fuel type and Euro standard (a) CO, (b) HC, (c) PM (index)

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0.00

0.05

0.10

0.15

0.20

CO/

CO

2ra

tio

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0

20

40

6080

100

120

HC/

CO

2ra

tio

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iesel_

E5

0.0

0.1

0.2

0.3

0.4

PM

Index

/C

O2ra

tio


28

4.3.2 Emission Measurement Results – Taxis

The emission measurement results for diesel taxis are compared directly against thoseof diesel passenger cars in Figures 17 and 18. The number of taxis with petrol enginesor derived from LGVs was too small for meaningful analysis, so are excluded from thefigures.

The emission results are broadly in-line with those of passenger cars, albeit withnominally higher fuel/ CO2 based NO, NO2 and NOx measurements as taxis aretypically larger, more intensively operated light-duty vehicles, having higher annualmileages. It is suggested the degradation of their engine and exhaust after-treatmentsystems is greater than those of a comparable (make, model, year of registration) privatepassenger cars. The stop-start driving conditions in urban areas, with dense traffic signalcontrol and high traffic demand is also not well suited to the efficient and cleanoperation of standard diesel taxis/ cars. Without regular periods of sustained, higherengine power operation (i.e. motorway driving/ speeds) the conditions needed forDiesel Particle Filters (DPFs) to regenerate are absent. DPF faults are thereforecommon on taxis and expensive to rectify as they are often not under warranty.

These results however suggest that the intensive operation of taxis is not significantlyaccelerating the degradation of their emission controls and emission performance asthey are broadly follow the trends and absolute levels of general diesel passenger cars.

4.3.3 Emission Measurement Results – LGVs

The emission measurement results for all three categories of LGVs (namely N1-I, N1-IIand N1-III) are presented in Figures 19 and 20. The results are also compared directlyagainst those of diesel (general) passenger cars.

The measured NO and estimated NO2 and NOX ratios to CO2 are all consistently higherthan those of diesel cars. This is expected as there is a high degree of commonalitybetween the engine and emission control technologies (powertrain) of passenger carsand vans, albeit with vans being larger (frontal area) and heavier vehicles. Thepowertrains of LGVs are also considered to be less sophisticated than on an averagediesel car, as the relative unit cost of the engine and exhaust after-treatments systems islower. With larger frontal areas, kerb and loading weights, they will also consume dieselat a higher rate than cars so the amount of an air pollutant per kilometre driven(Emission Factor) will be higher.

With the introduction of Euro 6 emission standard legislation lagging cars by at least12-months, the UK and much of Europe will continue to suffer from relatively highNOX emissions from the generation of Euro 5 and older LGVs for many years yet. NoEuro 6 LGVs were observed in surveys in Aberdeen (April 2015). With LGV Eurostandards lagging passenger cars by 12-months, significant numbers of Euro 6 vans arenot expected to enter the fleet until September 2017 at the earliest.

The emission ratios for other pollutants i.e. CO, HC and PM, were in-line with those ofdiesel passenger cars, falling to low levels for Euro 5 vehicles. It is particularlyimportant for urban air quality and health that emissions of PM fall. It is encouragingthat diesel engine developments and diesel particle filter (DPF) technology iseffectively cutting emissions.


29

EMISSION MEASUREMENTS – Taxis(a) NO(%)/CO2(%)



Figure 17: The distribution of the fuel based taxi emission measurements categorised byvehicle class (car, LGV), fuel type and Euro standard (a) NO, (b) NO2, (c) NOX

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi

_E

3

Taxi

_E

4

Taxi

_E

5

Taxi

_E

6

0

50

100

150

NO/

CO

2ra

tio*

10000

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi_

E3

Taxi_

E4

Taxi_

E5

Taxi_

E6

0

10

20

30

40

50

NO

2/

CO

2ra

tio*

10000

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi

_E

3

Taxi

_E

4

Taxi

_E

5

Taxi

_E

6

0

50

100

150

200

NO

X/

CO

2ra

tio*

10000


30

EMISSION MEASUREMENTS – Taxis(a) CO(%)/CO2(%)

(b) HC(%)/CO2(%)


Figure 18: The distribution of the fuel based taxi emission measurements categorised byvehicle class (car, LGV), fuel type and Euro standard (a) CO, (b) HC, (c) PM (index)

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi_

E3

Taxi_

E4

Taxi_

E5

Taxi_

E6

0.000

0.005

0.010

0.015C

O/

CO

2ra

tio

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi

_E

3

Taxi

_E

4

Taxi

_E

5

Taxi

_E

6

0

5

10

15

20

HC/

CO

2ra

tio

Die

selc

ar_

E0

Die

selc

ar_

E1

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

Taxi

_E

3

Taxi

_E

4

Taxi

_E

5

Taxi

_E

6

0.0

0.1

0.2

0.3

0.4

PM

Index

/C

O2ra

tio


31

EMISSION MEASUREMENTS – LGVs(a) NO(%)/CO2(%)



Figure 19: The distribution of the fuel based LGV emission measurements, compared withdiesel cars, categorised by LGV class, fuel type and Euro standard (a) NO, (b) NO2, (c) NOX

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0

50

100

150N

O/

CO

2ra

tio*

10000

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0

10

20

30

40

50

60

NO

2/

CO

2ra

tio

*10000

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0

50

100

150

200

NO

X/

CO

2ra

tio*

10000


32

EMISSION MEASUREMENTS – LGVs(a) CO(%)/CO2(%)

(b) HC(%)/CO2(%)


Figure 20: The distribution of the fuel based LGV emission measurements, compared withdiesel cars, categorised by LGV class, fuel type & Euro standard (a) CO, (b) HC, (c) PM Index

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0.000

0.005

0.010

0.015C

O/

CO

2ra

tio

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0

5

10

15

20

HC/

CO

2ra

tio

Die

selc

ar_

E2

Die

selc

ar_

E3

Die

selc

ar_

E4

Die

selc

ar_

E5

Die

selc

ar_

E6

dx

LG

V_N

1-I

_E

2

LG

V_N

1-I

_E

3

LG

V_N

1-I

_E

4

LG

V_N

1-I

_E

5

LG

V_N

1-I

_F

2

LG

V_N

1-I

I_E

2

LG

V_N

1-I

I_E

3

LG

V_N

1-I

I_E

4

LG

V_N

1-I

I_E

5

LG

V_N

1-I

I_F

2

LG

V_N

1-I

II_E

2

LG

V_N

1-I

II_E

3

LG

V_N

1-I

II_E

4

LG

V_N

1-I

II_E

5

0.0

0.2

0.4

0.6

0.8

PM

Index

/C

O2ra

tio


33

4.3.4 Emission Measurement Results – Buses and Coaches

The NO, CO, HC, PM (index), estimated NO2 and NOX measured emission ratios(pollutant over CO2 concentration) for the observed single- and double-decker Busesand Coaches are presented in the Figure 21 and 22 boxplots.

Buses (Passenger Service Vehicles – PSVs) are relatively heavy vehicles that have alarge frontal area, so have to overcome higher aerodynamic drag forces. ScheduledBuses are also intensively operated, heavy-duty vehicles with many vehicles operating18+hours per day, 7 days per week throughout the year, over fixed urban driving routeswith frequent stops and starts, with elongated periods of idling at lay-overs at start/ endof routes and at stops with a large number of passengers alighting. These are extremelychallenging operating conditions for exhaust after-treatment systems, particularly NOX

emission controls.

There was little discernible difference in the oxides of nitrogen measurements throughthe Euro standards II to V, nor between vehicle types (Bus or Coach, single- or double-decker). This is consistent with the findings of Carslaw et al (2011) that found the light-and heavy-duty Euro standards and associated technologies were ineffective.

Euro V Buses and Coaches were found to emit less PM than their predecessors, but theproportion of PM in their exhaust gases i.e. per unit of diesel burnt, was higher thanlight-duty diesel vehicles.

A policy of renewing the Bus fleet with Euro 5 specification vehicles is therefore notexpected to lower emissions of NOX. The schedule Bus fleet is assumed to be of interestto Aberdeen City Council as the authority has more influence these vehicles than otherheavy-duty, and indeed light-duty, vehicle types through Bus Quality partnerships,Local Transport Plans and Bus operator engagements.

Preliminary laboratory and on-road testing of Euro 6 Buses with SCR (SelectiveCatalytic Reduction) and DPF (Diesel Particle Filter) exhaust after-treatmenttechnology is encouraging, as the technologies and emission controls can now beconfigured to control and significantly lower NOX emissions.

As scheduled Buses repeat services/ journeys/ routes many times a day, multiple pass-bys/ measurements of many vehicles were available. Figure 23 presents the estimatedNOX and PM emission ratio measurements for individual Buses, where five or morepass-by measurements were available. Although there is some variation in the estimatedemission factors for an individual vehicle, there is greater variation between vehicles ofthe same Euro classification or otherwise. This analysis again demonstrates that thesuccessive III to V Euro standards have not led to consistent improvement in the NOX

emissions from Buses. PM emissions from the Euro V Buses are lower.

Bus emission measurements have also been grouped by common marques/ models/Euro standard. This analysis again illustrates that standard Buses powered by dieselengines have broadly similar NOX characteristics, irrespective of their Euro standard.Encouragingly the diesel-electric (Alexander Dennis Enviro 350 Hybrid) had the lowestof all the common make and model of Bus.


34

EMISSION MEASUREMENTS – PSVs(a) NO(%)/CO2(%)



Figure 21: The distribution of the fuel based PSV emission measurements categorised by Bus/ Coach and Euro standard (a) NO, (b) NO2, (c) NOX

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0

50

100

150

200

250

NO

/C

O2ra

tio*

1000

0

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0

5

10

15

20

NO

2/

CO

2ra

tio

*10000

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0

50

100

150

200

250

NO

X/

CO

2ra

tio*

10000


35

EMISSION MEASUREMENTS – PSVs(a) CO(%)/CO2(%)

(b) HC(%)/CO2(%)


Figure 22: The distribution of the fuel based PSV emission measurements categorised by Bus/ Coach and Euro standard (a) CO, (b) HC, (c) PM Index

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0.000

0.005

0.010

0.015

0.020

0.025

0.030

CO/

CO

2ra

tio

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0

2

4

6

8

10

12

HC/

CO

2ra

tio

Bus_D

D_E

2

Bus_D

D_E

3

Bu

s_D

D_E

6

Bus_

SD

_E

2

Bus_

SD

_E

3

Bus_

SD

_E

4

Bus_

SD

_E

5

Bus_S

D_E

6

Coach_D

D_E

3

Coach_D

D_E

5

Coa

ch_D

D_E

6

Coach

_S

D_E

1

Coach

_S

D_E

3

Coach

_S

D_E

4

Coach

_S

D_E

5

0.0

0.1

0.2

0.3

0.4

0.5

0.6

PM

Index

/C

O2ra

tio

36

EMISSION MEASUREMENTS – Individual PSVs ( > 5 pass-bys)(a) NOX(%)/CO2(%)


Figure 23: The measured emission ratios for individual Buses ordered by age of first registration (a) NOX, (b) PM10 (index)[Note the boxplots are colour coded according to the vehicles respective Euro standard: Euro 2 – red, Euro 3 – orange, Euro 4 – yellow, Euro 5 – blue]

E2_V

601G

GB

E2_V

610G

GB

E2_W

2F

AL

E2_W

4F

AL

E2_W

577R

FS

E2_W

578R

FS

E2_W

582R

FS

E2_W

583R

FS

E2_W

5F

AL

E2_W

7F

AL

E2_X

136F

PO

E3_S

V05D

XC

E3_S

V05D

XE

E3_S

V05D

XG

E3_S

V05D

XH

E3_S

V05D

XJ

E3_S

V05D

XL

E3_S

V05D

XP

E3_S

V05D

XW

E3_S

V05D

XX

E3_S

V05D

XY

E3_X

141F

PO

E3_X

144F

PO

E3_X

971H

LT

E4_S

F07LC

G

E4_S

V07E

HG

E4_S

V55E

EP

E4_S

V55E

EU

E5_S

V59C

GX

E5_S

V59C

HD

E5_S

V59C

HG

E5_S

V59C

HN

E5_S

V59C

HO

0

50

100

150

200

NO

X/

CO

2ra

tio*

10000

E2_V

601G

GB

E2_V

610G

GB

E2_W

2F

AL

E2_W

4F

AL

E2_W

577R

FS

E2_W

578R

FS

E2_W

582R

FS

E2_W

583R

FS

E2_W

5F

AL

E2_W

7F

AL

E2_X

136F

PO

E3_S

V05D

XC

E3_S

V05D

XE

E3_S

V05D

XG

E3_S

V05D

XH

E3_S

V05D

XJ

E3_S

V05D

XL

E3_S

V05D

XP

E3_S

V05D

XW

E3_S

V05D

XX

E3_S

V05D

XY

E3_X

141F

PO

E3_X

144F

PO

E3_X

971H

LT

E4_S

F07LC

G

E4_S

V07E

HG

E4_S

V55E

EP

E4_S

V55E

EU

E5_S

V59C

GX

E5_S

V59C

HD

E5_S

V59C

HG

E5_S

V59C

HN

E5_S

V59C

HO

0.0

0.2

0.4

0.6

0.8

PM

Index

/C

O2ra

tio

37

Figure 24: The measured NOX emission ratios for individual marques/ models of Bus.

Figure 25: The measured PM emission ratios for individual marques/ models of Bus[Note the boxplots are colour coded according to the vehicles respective Euro standard: Euro 2 – red, Euro 3 –

orange, Euro 4 – yellow, Euro 5 – blue]

E2_V

OLV

O_B

SE

RIE

SB

10

BLE

E3_V

OLV

O_B

SE

RIE

SB

10

BLE

E3_V

OLV

O_B

SE

RIE

SB

7L

E3_V

OLV

O_B

SE

RIE

SB

9T

L

E4_A

LE

XD

EN

NIS

_E

NV

IRO

350

HY

BR

ID

E4_V

OLV

O_B

SE

RIE

SB

12B

E4_V

OLV

O_B

SE

RIE

SB

7R

E4_V

OLV

O_B

SE

RIE

SB

7R

LE

E5_A

LE

XD

EN

NIS

_E

NV

IRO

300

12.0

M

E5_O

PT

AR

E_V

ER

SA

V1110

E5_V

OLV

O_B

SE

RIE

SB

12B

0

50

100

150

200

250

NO

X/

CO

2ra

tio

*10000

E2_V

OLV

O_B

SE

RIE

SB

10

BLE

E3_V

OLV

O_B

SE

RIE

SB

10

BLE

E3_V

OLV

O_B

SE

RIE

SB

7L

E3_V

OLV

O_B

SE

RIE

SB

9T

L

E4_A

LE

XD

EN

NIS

_E

NV

IRO

350

HY

BR

ID

E4_V

OLV

O_B

SE

RIE

SB

12B

E4_V

OLV

O_B

SE

RIE

SB

7R

E4_V

OLV

O_B

SE

RIE

SB

7R

LE

E5_A

LE

XD

EN

NIS

_E

NV

IRO

300

12.0

M

E5_O

PT

AR

E_V

ER

SA

V1110

E5_V

OLV

O_B

SE

RIE

SB

12B

0.0

0.1

0.2

0.3

0.4

0.5

0.6

PM

Index

/C

O2ra

tio


38

4.3.5 Emission Measurement Results – OGVs

The NO, CO, HC, PM (index), estimated NO2 and NOX measured emission ratios (pollutantover CO2 concentration) for the observed articulated (artic) and rigid Ordinary Goods Vehicles(OGVs) are presented in the Figure 26 and 28 boxplots. The distribution of the observed (RSD)NOX measurements from Euro VI OGVs are also corroborated published Transport for London(TfL, 2015) laboratory data over a real driving profile in Figure 27 (see section 4.8.1).

The number of OGVs recorded is low in comparison with the other vehicle categories.However, Euro VI RDE legislation was effective from January 2014, ahead of regulations forlight-duty vehicles. A reasonable sample of 22 Euro VI articulated and 22 Euro VI rigid OGVswere surveyed in Aberdeen. This is important as the heavy-duty Euro VI RDE regulations andassociated technologies are significantly cleaner for NOX. This is seen in the step-change fallin observed NOX to CO2 ratios in Figure 26. The distribution of the RSD NOX to CO2 ratioscover the range in the TfL (2015) published figures. NOX emission controls, particularly SCRsystems need catalysts to be at operational temperatures, to be effective. Temperatures inemission controls are higher when engines are under load. At lighter engine power demandsduring periods of idling for example, exhaust gas flow rate and temperatures can be lower,cooling catalysts and their ability to mitigate NOX emissions. In urban driving, the developmentand integration of heavy-duty engines and their emission controls means NOX control is nowgood in all but the lowest power demand conditions. This has been achieved through thedevelopment of components (both engines and exhaust after-treatment systems) and managing/controlling them in synchrony. Only in extended periods of light power demands are NOX

emissions elevated, for example an OGV driving without any cargo load (empty running) inurban driving conditions. This is clearly illustrated in the TfL (2015) laboratory results, whereboth rigid and articulated OGVs emit considerably more NOX over the TfL drive-cycles whenrunning empty (0% load). The RSD results lie between the results for the two rigid and singlearticulated trucks tested. This is to be expected as ≈30% of OGV vehicle kilometres are knownto be running empty. The average loading factor is ≈50% for smaller OGVs, rising to >60%for larger (25ton rigid and articulated) vehicles (DfT, 2015).

There’s a clear reduction in the ratio of PM to CO2 observed in the Euro II through to Euro VIOGV exhaust plumes. This again illustrates the success of Euro standards and associatedtechnologies, notably DPFs, removing coarser PM.


39

EMISSION MEASUREMENTS – OGVs(a) NO(%)/CO2(%)



Figure 26: The distribution of the fuel based OGV emission measurements categorised by Articulated/ Rigid and Euro standard (a) NO, (b) NO2, (c) NOX

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0

50

100

150

200

250

NO/

CO

2ra

tio*

10000

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0

10

20

30

40

50

60

NO

2/

CO

2ra

tio

*10000

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0

50

100

150

200

250

300

350

NO

X/

CO

2ra

tio*

10000


40

Figure 27: The distribution of measured Euro VI OGV NOX measurements, compared with Transportfor London (2015) Euro VI laboratory measurements over a real London Speed profile. The

laboratory measurements include empty and fully-laden running.

aN

2rigid

0%

load

aN

3rigid

0%

load

bN

2rigid

100%

load

bN

3rigid

100%

load

cR

SD

rigid

d

dN

3art

ic0%

load

dN

3art

ic100%

load

zR

SD

art

ic

0

20

40

60

80N

OX/

CO

2ra

tio*

10000


41

EMISSION MEASUREMENTS – OGVs(a) CO(%)/CO2(%)

(b) HC(%)/CO2(%)


Figure 28: The distribution of the fuel based OGV emission measurements categorised by Articulated/ Rigid and Euro standard (a) CO, (b) HC, (c) PM Index

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0.000

0.002

0.004

0.006

0.008

CO/

CO

2ra

tio

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0

5

10

15

20

25

HC/

CO

2ra

tio

Art

ic_E

2

Art

ic_E

3

Art

ic_E

4

Art

ic_E

5

Art

ic_E

6

Az

Rig

id_E

2

Rig

id_E

3

Rig

id_E

4

Rig

id_E

5

Rig

id_E

6

0.00

0.05

0.10

0.15

0.20

0.25

PM

Index

/C

O2ra

tio


42

4.4 Passenger car Emission Factors (EFs)

Passenger car Emission Factors (EFS, grams.km-1) can be estimated by weighting the rawemission measurements (ratios of air pollutants to CO2 in the exhaust plume trailing a vehicle)by each individual vehicle’s known fuel efficiency / CO2 performance over the NEDC drive-cycle (see Figure 11 for the variation and trend in the Aberdeen car fleet official CO2

performance). The ICCT (2015) have however shown that the discrepancy between the officialCO2 performance figures and those observed in real driving (RDE) has increased through theperiod 2000 to 2015. The official testing regime, legislation and ‘type-approval’ is returningincreasingly unrealistic CO2 figures. This trend has been accounted for by up-lifting the officialCO2 figures by the published divergence between real-world and manufacturers’ type-approvalCO2 emissions year on year for petrol and diesel cars (ICCT, 2015). The trend in thediscrepancy is illustrated in Figure 29.

Figure 29: The Divergence between real-world and manufacturers’ type-approval CO2 emissions forpetrol and diesel cars (adapted from ICCT, 2015a).

This method is considered to provide reliable Emission Factors for groups of vehicles i.e. dieselEuro 5 passenger cars (Carslaw et al, 2011). The emission factors are not absolute emissionfactors, as measured in laboratory or PEMS studies (ICCT, 2015b), instead being estimates thatillustrate trends in the RDE performance of the operational fleet. The aggregated, averageemission factors agree well with reputable emission modelling results that attempt to accountfor real-driving conditions. This analysis is presented in section 4.7.

1

1.1

1.2

1.3

1.4

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Div

erge

nce

of

‘rea

l-w

orl

d’f

rom

man

ufa

ctu

rers

’typ

e-ap

pro

valC

O2

(fac

tor)

Year of first registration

Diesel

Petrol


43

The passenger car NOX emission factors and trends illustrated in Figure 30 track those of the‘raw’ emission measurement ratios presented in section 4.3.1, notably:

NOX performance of diesel cars is high and relatively consistent for Euro 2, 3, 4 and 5cars. This is consistent with the findings of Carslaw et al (2011) who highlighted, usingUK remote sensing data, the failure of light-duty diesel emission standards andassociated technologies in lowering NOX emissions in real-driving (RDE).

The latest Euro 6 diesel cars are emitting NOX at roughly half the rate of Euro 5 (andtherefore Euro 2, 3 and 4);

NOX emissions from Euro 4 and newer (Euro 5 and 6) petrol cars are at a low-level.

EMISSION FACTORS – Passenger cars(a) Predicted NO2 (grams.km-1)

(b) Predicted NOX (grams.km-1)

Figure 30: The distribution of the predicted passenger car emission factors (grams perkilometre) categorised by fuel type and Euro standard (a) NO2, (b) NOX.

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hyb

rid-D

iese

l_E

5

0.0

0.2

0.4

0.6

0.8

1.0

NO

2(g

ram

s.k

m1

)

aP

etr

ol_

E0

aP

etr

ol_

E1

aP

etr

ol_

E2

aP

etr

ol_

E3

aP

etr

ol_

E4

aP

etr

ol_

E5

aP

etr

ol_

E6

ax

bH

ybrid-P

etr

ol_

E4

bH

ybrid-P

etr

ol_

E5

bH

ybrid-P

etr

ol_

E6

bx

Die

sel_

E0

Die

sel_

E1

Die

sel_

E2

Die

sel_

E3

Die

sel_

E4

Die

sel_

E5

Die

sel_

E6

dx z

Hybrid-D

iesel_

E5

0.0

0.5

1.0

1.5

2.0

2.5

NO

X(g

ram

s.km

1

)


44

The variation in NO2 and NOX emission factors estimates for petrol and diesel passenger carsby year of registration are also presented in Figure 31, which clearly illustrates that whilst theengine and emissions controls on modern petrol cars have led to an improvement in NOX

emissions, there has been little improvement in the NOX performance of diesel passenger carsin urban areas in the last fifteen years.

EMISSION FACTORS – Passenger cars

Predicted NO2 (grams.km-1)


Predicted NOX (grams.km-1)

(c) Petrol (d) Diesel

Figure 31: The distribution of the predicted passenger car emission factors (grams perkilometre) categorised by Fuel type and Year of registration (a) Petrol NO2, (b) Diesel NO2,

(c) Petrol NOX, (b) Diesel NOX

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

0.0

0.2

0.4

0.6

0.8

1.0


NO

2(g

ram

s.km

1

)

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

0.0

0.2

0.4

0.6

0.8

1.0


NO

2(g

ram

s.km

1

)

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

0.0

0.5

1.0

1.5

2.0

2.5

3.0


NO

X(g

ram

s.km

1

)

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

0.0

0.5

1.0

1.5

2.0

2.5

3.0


NO

X(g

ram

s.km

1

)


45

The predicted emission factors for passenger cars are documented in the appendix Table A1.The predicted NOX emission factor trends are broadly in-line with earlier studies/ findings(Carslaw et al, 2011). The predicted values are however slightly lower. This is partly due tothe lower assumed f-NO2 fractions (Carslaw et al, 2013; also see Table 9), but may also be dueto differences in the local observed traffic situations.

The relative estimated emission contribution from each passenger car fuel type and Eurostandard combination is presented in the Figure 32 bar charts. The contribution accounts forthe passenger car fleet proportions (annotated in [%]) and their emission characteristics/ factors(EFs). Whilst the Euro 1 petrol cars have poor on-road CO and NOX performance, attributedto an absence or degraded exhaust after-treatment system, as they only comprise less than0.09% of the sampled fleet, or one in every thousand cars, their contribution is small. Similarlyalthough the tail-pipe PM performance of pre-Euro, Euro 1 and 2 diesel cars is poor, as thesevehicles are scarce their emission contribution is also low.

EMISSION CONTRIBUTIONS (%) - Passenger cars

(a) NO2 (b) NOX

(c) CO (d) PM10

Figure 32: The passenger car emission contributions categorised by Fuel type and Eurostandard (a) NO2, (b) NOX, (c) CO, (d) PM10.

Ehaust after-treatment systems such as Diesel Particulate Filters (DPFs) and Diesel OxidationCatalysts (DOCs), whilst now leading to a decrease in PM emissions in real-world (on-road)conditions, they are shown to less effective in relation to controlling NOX emissions in urbandriving conditions. As both DPFs (during filter regeneration) and DOCs oxidise exhaust gases,

E0

petr

ol

E1

petr

ol

E2

petr

ol

E3

petr

ol

E4

petr

ol

E5

petr

ol

E6

petr

ol

E0

die

sel

E1

die

sel

E2

die

sel

E3

die

sel

E4

die

sel

E5

die

sel

E6

die

sel

NO

2co

ntr

ibutio

n(%

)

0

20

40

60

80

[0.0

0]

[0.0

9]

[1.3

6]

[8.9

5]

[19.5

8]

[22.9

2]

[1.9

7]

[0.0

0]

[0.0

0]

[0.2

2]

[3.3

1]

[12.1

5]

[27.5

5]

[1.9

0]

E0

petr

ol

E1

petr

ol

E2

petr

ol

E3

petr

ol

E4

petr

ol

E5

petr

ol

E6

petr

ol

E0

die

sel

E1

die

sel

E2

die

sel

E3

die

sel

E4

die

sel

E5

die

sel

E6

die

sel

NO

Xcontr

ibution

(%)

0

10

20

30

40

50

60

70

[0.0

0]

[0.0

9]

[1.3

6]

[8.9

5]

[19.5

8]

[22.9

2]

[1.9

7]

[0.0

0]

[0.0

0]

[0.2

2]

[3.3

1]

[12.1

5]

[27.5

5]

[1.9

0]

E0

petr

ol

E1

petr

ol

E2

petr

ol

E3

petr

ol

E4

petr

ol

E5

petr

ol

E6

petr

ol

E0

die

sel

E1

die

sel

E2

die

sel

E3

die

sel

E4

die

sel

E5

die

sel

E6

die

sel

COc

ontr

ibution

(%)

0

10

20

30

40

[0.0

0]

[0.0

9]

[1.3

6]

[8.9

5]

[19.5

8]

[22.9

2]

[1.9

7]

[0.0

0]

[0.0

0]

[0.2

2]

[3.3

1]

[12.1

5]

[27.5

5]

[1.9

0]

E0

petr

ol

E1

petr

ol

E2

petr

ol

E3

petr

ol

E4

petr

ol

E5

petr

ol

E6

petr

ol

E0

die

sel

E1

die

sel

E2

die

sel

E3

die

sel

E4

die

sel

E5

die

sel

E6

die

sel

PM

10co

ntr

ibutio

n(%

)

0

10

20

30

40

[0.0

0]

[0.0

9]

[1.3

6]

[8.9

5]

[19.5

8]

[22.9

2]

[1.9

7]

[0.0

0]

[0.0

0]

[0.2

2]

[3.3

1]

[12.1

5]

[27.5

5]

[1.9

0]


46

these after-treatment systems lead to higher emissions of NO2 at the tail-pipe (termed primaryNO2). The measurements and analysis presented in Figure 32 clearly illustrate newer Euro 5diesel passenger cars are a major contributor towards the increase in primary NO2 emissions.

4.5 Passenger Car Emission Distributions

An application of vehicle emission RSD technology is to identify the abundance andsignificance of high-emitting vehicles on the road. This could for example be poorlymaintained petrol cars, or those with a fault (three-way) catalytic converter. There are alsoconcerns the removal of a faulty DPF (diesel particle filters), or intentionally removed toimprove fuel economy may be slowing the expected trend of reducing tail-pipe PM emissions.How common-place DPF removal is, is not well-known, just that there are numerous business’soffering DPF removal services e.g. http://www.dpfspecialistlondon.co.uk/. A car willautomatically fail its MoT test if a filter had been fitted as standard (manufacturer) but is nolonger present. This simple visual check however may not be sufficient in identifying anddiscouraging owners from having their DPFs removed. Examining the distributions of remotesensing results is one approach to reviewing and assessing the significance of high-emittingvehicles. It is proposed that high-emitting vehicles will present themselves as outliers indistributions (Borken-Kleefeld et al, 2012).

The density plots in Figure 33 present the changes in the estimated NOX emission factor(continuous) distributions for the passenger car fleet, categorised by the Euro standard and fueltype (Petrol, left-panels; Diesel, right-panels). Petrol cars with elevated NOX emissions in theright-hand tail of the distributions are commonly older vehicles in the Euro 2 category (orolder). The proportion of vehicles with very low emissions also tends to increase with eachsuccessive more stringent Emission standard. The skewed nature (gamma distribution) of thevehicle emission ratios indicates that a small number of more polluting vehicles can contributea significant proportion to the total emissions. The NOX emission distributions of diesel carsEuro 3 to 5 are broadly similar (range and distribution). Euro 6 passenger cars are cleaner, withthe majority of predicted NOX emission factors less than 1 grams.km-1.

The density plots for the PM (index) measure are illustrated in Figure 34. The proportion ofthe diesel cars with elevated PM (index) emissions falls through the Euro 0 -5 emissioncategories. The majority of Euro 5 and 6 diesel cars emit comparatively little PM, but still morethan their petrol counter-parts.

In order to quantify the abundance and emission contribution of high-emitting vehicles, it isnecessary to develop and apply classification criteria. In this study a two-step approach hasbeen used:

1. Identify passenger cars with elevated emissions. The threshold for passenger cars withelevated emissions is when an emission factor estimate is greater than the fleet meanplus 3 standard deviations for their NOX and PM10 (index) emission factor; and

2. Identify high-emitters. A high-emitter is a car that is classified as having elevatedemissions for more than one pollutant species.

Table 10 documents the elevated emission threshold and the number, percentage andcontribution to the estimated total emissions. The small proportion of vehicles classified ashaving elevated emissions, are responsible for a significant proportion of the total NOX (13.1%)and PM10 (19.7%) emitted. The majority of these passenger cars are diesel.

The contribution to the total emissions of the critical species NOX and PM10 from the small(0.05% of the passenger car fleet) subset of high-emitter cars is less pronounced (0.26% and


47

0.51% respectively). This investigation suggests a policy of identifying and removing high-emitters would not have a significant minimal impact on the total emissions of NOX and PM10.

Table 10: Summary of passenger cars identified as having ‘elevated’ emissions.

Pollutant ElevatedEmissionThreshold

(Mean + 3st. dev.)

MeanElevatedVehicles

(grams.km-1)

%Vehicles

%Diesel

% NOX

Contribution

(Emission total)

% PMContribution

(Emission total)

NOX2.24

(grams.km-1) 2.77 2.19% 92.4% 13.1% 3.4%

PM1053.5

(index) 98.1 1.41% 75.8% 2.3% 19.7%

Table 11: The emission contribution from passenger cars identified as being ‘high-emitters’

Numbervehicles

% Vehicles %NOX

emissions%PM10

emissions

9 0.05% 0.26% 0.51%


48

NOX EMISSION DISTRIBUTIONS – Passenger cars

Predicted NOX (grams.km-1)


Figure 33: Density Distribution Plots of the estimated NOX emission factors of passengercars categorised by fuel type and Euro category: (a) petrol, (b) diesel.

PM10 (index) EMISSION DISTRIBUTIONS – Passenger cars

PM10 (index)


Figure 34: Density Distribution Plots of the estimated PM10 emission factors of passengercars categorised by fuel type and Euro category: (a) and (c) petrol, (b) and (d) diesel.

The possible influence of DPF removal is investigated in Figure 35, which presents the ‘raw’PM10 / CO2 emission ratio measurement distribution of Euro 5 and 6 diesel cars. The individualmeasurements are displayed beneath the histogram. Those with ‘elevated’ PM10 / CO2 emissionratios exceed (mean plus 3 standard deviations) a value of 0.11. An extension of the applicationof RSD surveys in future studies is to identify new diesel cars that are observed to be emittinghigh amounts of tail-pipe PM and investigate whether they have had their DPF removed in agarage.

0 1 2 3 4 5 6

0

1

2

3

4

5

6

7

NO X (grams.km1

)

Densi

ty

0 1 2 3 4 5 6

0.0

0.5

1.0

1.5

2.0

NO X (grams.km1

)

Densi

ty

0

50

100

150

0.00

0.05

0.10

0.15

0.20

0.25

0.30

PM 10 (index.km1

)

Densi

ty

0

50

100

150

0.00

0.05

0.10

0.15

0.20

PM 10 (index.km1

)

Densi

ty


49

Figure 35: Histogram of the (raw) observed PM10 / CO2 emission ratio measurement forEuro 5 and 6 diesel passenger cars.

4.6 Speed, Acceleration and Engine Power demands across the City ofAberdeen – comparison with the five Vehicle Emission Measurement sites

A criticism of remote sensing studies is that they only survey a limited range of vehiclesoperation conditions at the fixed sites. The presence of traffic management (signing and cones)is also considered to temper driver’s behaviour. The remote sensing instrument captures themovement (speed and acceleration rate) and an emission signature of thousands of vehiclespassing through the measurement station each day, with any driver behaviour modificationsconsidered to be minor. This is qualitatively evaluated by comparing the speed and accelerationdistributions from the remote sensing instrument with those of the vehicle tracking surveys (seesection 3.4, Table 4).

Vehicle tracking surveys were carried out across the wider Aberdeen road network, recordingthe speed and position of a Super-mini (market segment) as it was driven on the main routes inthe City during peak and off-peak periods. The distribution of second-by-second (1Hz) speedand acceleration, speed and VSP, for the vehicle tracking (all routes) and vehicle emissionmeasurement (remote sensing) system (passenger cars only) across all are compared in Figure36. The frequency of speed/ acceleration occurrences in hexagonal bins are illustrated on agrey-scale. Darker hexagon bins therefore indicate a higher frequency of data points in thatrange than lighter shaded areas.

PM 10 / CO 2 emission ratio

Fre

que

ncy

0.0 0.1 0.2 0.3 0.4

01

00

02

00

03

00

0400

050

00


50

Figure 36: Frequency diagrams of the observed Speed and Accelerations distributions fromthe Vehicle tracking surveys (All) and those recorded by the Vehicle Emission Measurement

System (VEMS).

The range of the vehicle speeds and accelerations observed by the mobile and fixed systemsare broadly similar. The vehicle tracking surveys cover a wider speed range as the vehicle wasdriven in 30 m.p.h. and 40 m.p.h. zones. The tracked car also spent more time at lower speeds,as it responded to traffic signals, surrounding vehicles and minor incidents on the network. Theremote sensing instrument does not consider vehicle pass-by speeds below 5 km.h-1 a validmeasurement and are excluded. The magnitude and distribution of positive acceleration issimilar from both the mobile and fixed measurements suggesting driver behaviour is not undulytempered by the presence of the instrument and traffic management.

The distribution of speed and accelerations for each vehicle tracking route (1-15) and the fiveremote sensing sites are illustrated in appendix B.

4.7 Instantaneous Emission Modelling (IEM) – Passenger cars

The 160 kms (Table 4) of Real Driving vehicle tracking data can also be used to estimatevehicle emissions. Instantaneous Emission Models (IEM) consider the influence of vehicleaccelerations, road gradient, vehicle and engine load when predicting fuel consumption andtail-pipe emission rates. This is clearly highly desirable when attempting to understand vehicleemissions across congested networks or to evaluate environmental traffic managementpolicies.

The leading IEM in Europe is considered to be PHEM (Boulter et al, 2007). The model isdescribed in section 4.7.1. This includes a validation of its performance against a sample ofTransport for London (2015) in-service passenger car emission testing data. The predictedEmission Factors for Euro 0 – 6, petrol and diesel passenger cars, over the 15 sets of vehicletracking data are described in section 4.7.2.

CAR TRACKING REMOTE SENSING

NOTE: Time stationary 26%

0 20 40 60 80 100

-3

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

33

65

98

130

162

194

226

258

291

323

355

387

419

452

484

516

Counts

10 20 30 40 50 60 70

-3

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

12

23

35

46

57

68

79

90

102

113

124

135

146

158

169

180

Counts


51

4.7.1 The Instantaneous Emission Model PHEM

PHEM is a comprehensive power-instantaneous emission model that is able to simulate fuelconsumption and NOX, NO2, HCs, Particulate Mass (PM10), Particle Number (PN), CarbonMonoxide (CO) and Hydrocarbons (HC) tail-pipe emissions of the whole vehicle fleet of Euro0 to Euro 6, petrol/ diesel/ CNG and bio-gas fuels, heavy-duty vehicles, passenger cars andlight commercial vehicles second-by-second. PHEM has been developed at the TechnicalUniversity of Graz (TUG, AU) since 2000 in the research projects ARTEMIS, COST346 andHBEFA. The model is based on light- and heavy-duty vehicle engine speed – power emissionmaps established from engine and chassis dynamometer measurements (Rexeis et al, 2007,Zallinger et al. 2005).

PHEM has a time alignment and correction sub-model to relate engine speed – power eventsto predicted engine-out emissions. Emissions recorded in chassis dynamometer, exhaustemission sampling and analysis facilities have been delayed and engine-out peaks smoothedduring transport through the exhaust, sampling and analysis systems. Dynamic correction andtime alignment functions have been developed through experimental investigation and CFD-simulations of the exhaust gas transport (Zallinger et al, 2005). It is the time corrected emissionmeasurements that are used to populate the engine speed – power emission maps. Thesemethods make PHEM capable of simulating the instantaneous fuel consumption and emissionsfor any speed profile or driving cycle.

The IEM PHEM has been validated with second-by-second Transport for London (TfL) chassisdynamometer (Millbrook) measured data. A sample of Euro 4, Euro 5 and now Euro 6Passenger Cars have been tested over a drive-cycle (speed profile) intended to represent thebroad range of London (real-world) driving conditions. The ‘London Drive Cycle’ for light-duty vehicles has been developed by TfL as part of an on-going Vehicle Emission Study (TfL,2015). The drive cycle was developed in association with www.millbrook.co.ukwho werecommissioned to track a car (VBox GPS and CAN Bus link) making repeated circuits of a setroute in the North-East of London at different times of day: AM peak, Inter-peak and in Free-flow conditions. The route contained sections of (urban) motorway, suburban and urban(central London) driving conditions. The speed profile (time-series) of the London Drive Cycleis illustrated in Figure 37. The drive-cycle is considered to represent typical driving style/behaviour in the UK. The drive-cycle doesn’t consider fluctuations in road gradient. Summary


52

statistics for the different elements of the drive cycle (road type and time period) aredocumented in Table 12.

Figure 37. Average speed (modelled) in each simulation hour

Table 12. The London Drive Cycle statistics.

RoadType

TimePeriod

Duration(seconds)

Distance(km)

AverageSpeed

(km.h-1)

MaximumAcceleration

(m.s-2)

SectionID

Urban Free-flow 1202 8.92 26.73 2.67 7

Urban AM peak 2048 8.93 15.69 1.97 8

Urban Inter-Peak 2311 8.93 13.91 2.48 9

Suburban Free-flow 1036 13.33 46.31 2.4 10

Suburban AM peak 1894 13.33 25.33 2.67 11

Suburban Inter-Peak 1591 13.33 30.16 2.31 12

Motorway Free-flow 1023 24.61 86.60 1.62 13

Motorway AM peak 1884 24.61 47.03 1.69 14

Motorway Inter-Peak 1030 24.61 86.02 2.46 15

The agreement of a sample of TfL second-by-second vehicle emission measurements over theLDC with IEM predictions from PHEM (version 11.4) for four Euro IV Passenger cars hasbeen evaluated. The specification of the four vehicles tested and simulated is documented inTable 13. The PHEM vehicle specifications were adjusted to match the test vehicles. The

Time (seconds)

Sp

ee

d(k

m/h

)

0

20

40

60

80

100

120motorway

0

20

40

60

80

100

120suburban

0

20

40

60

80

100

120

0 500 1000 1500 2000

urban


53

PHEM engine power and emission maps are an average (normalised) of several vehicles of thatcategory. The fuel efficiency and emission performance of broadly comparable (size, weightand engine power) makes and models are variable, particularly for air quality pollutants. Fuelefficiency and emission performance of two identical cars (make, model and year) can alsovary by ≈ ± 10%, so it is not expected the observed and modelled will be in close agreement, rather that the predictions will replicate the trends and dynamics of the measurements.

Table 13. Passenger Car specifications tested over the LDC.

Fuel Type Market

Segment

Euro

Standard

Make &Model

Engine Enginepower(kW)

GVW(kg)

Frontalarea(m2)

Petrol Compact 4 Peugeot 107 1.0 L 50 900 2.4

Small Family 4 Ford Focus 1.6L (Zetec) 75 1370 2.75

Diesel Mini 4 Ford Fiesta 1.4L TDCi 66 1223 2.92

MPV 4 Ford Galaxy 2.0L TDCi 105 1890 3.74

PHEM produces emission predictions for a vehicle trip, totals for road sections (links,segments) and second-by-second results. It is these second-by-second emission predictions thathave been compared with the transient measurements from the chassis dynamometer. Thechassis dynamometer (Constant Volume Sampling – CVS) measurements have been adjustedto account for time delays and shifts in emission peaks, to make them comparable withinstantaneous tail-pipe emission predictions.

The comparison of the observed and modelled values for the entire LDC, for the four passengercars are presented in the scatterplots in Figure 38. The frequency of data points in a hexagonalbin is illustrated on a colour-scale, so both the range in values and where the core of the datalies are visualised. The second-by-second observed and modelled CO2 data are in closeagreement, demonstrating PHEM is replicating the dynamics and magnitude of measuredvalues well, for all four test cars. The CO2 model performance is extremely good for all testcars and driving settings (urban, suburban and motorway).

Predicting tail-pipe emission of air quality pollutants is more challenging as engine/ emissionmaps are more variable and the impact of exhaust after treatment systems also needs to beconsidered. Petrol car NOX emissions are at a low-level and not illustrated. The scatterplots foroxides of nitrogen (NOX) demonstrate that PHEM is also reliably predicting the dynamics andmagnitude of NOX emissions well for diesel cars. The main discrepancy is for the smaller‘Mini’ diesel car at higher emission rates ( > 0.04 grams.sec-1). The model is predicting slightlylower emissions during these more polluting periods that have higher engine power demands.The optimisation of the engine/ emission map of the observed car may be slightly worse thanthe average medium sized diesel car in the model. These differences are to be expected betweenspecific vehicles and fleet averages.

For a given ‘Real Driving’ or ‘micro-simulated’ speed profile (drive-cycle) the PHEM modelis considered to reliably predict the second-by-second tail-pipe emissions of CO2 and NOX forthe Euro 4 passenger car test data. The PHEM model is considered to be an appropriate tool tocouple to ‘Real Driving’ vehicle trajectories to predict CO2 and NOX tail-pipe emissions, andevaluate the impact of policies that adapt the distribution of vehicle speeds and accelerations.


54

It is important to differentiate PHEM from more basic instantaneous models that predict tail-pipe emissions using speed/ acceleration emission maps. The first generation of instantaneousemission models such as MODEM (Jost et al, 1992) derived the speed/ acceleration/ fuelconsumption and emission maps (or tables) from continuous (second-by-second) laboratorymeasurements. The absence of dynamic time alignment and correction functions limited theprediction accuracy of such models, especially for more modern (Euro 2 and newer) vehiclesequipped with exhaust after-treatment systems (Zallinger et al, 2005). More recently the AIRE(Analysis of Instantaneous Road Emissions- Transport Scotland, June 2011) has been launchedwhich uses instantaneous emission modelling tables derived second-hand from PHEMmodelled outputs. SIAS Limited in collaboration with TRL (Transport Research Laboratory)developed AIRE using a version of PHEM released in 2005 (Hausberger, 2011), so is not ableto predict emissions from modern Euro 4 and 5 vehicles that now dominate the fleet. Fewdetails are available of the speed profiles (drive-cycles) supplied to PHEM, to in turn derivethe speed/ acceleration emission tables. If these speed profiles (drive-cycles) are notrepresentative of the driver behaviour and traffic flow conditions of a study network, thenmodel prediction uncertainties will increase. The use of AIRE is not recommended until furtherdetails of the model development and model accuracy are made available.

As input, PHEM requires 1Hz speed data, road gradient and the vehicle specification including:fuel type, Euro emission standard, engine size, exhaust after-treatment systems, vehicle weight,frontal area, drag coefficients and transmission ratios. Although the PHEM model has defaultEU average values available, as this study had a detailed understanding of the local vehiclefleet, key parameters were reflected to represent the Aberdeen passenger car fleet i.e. frontalarea, average vehicle weight and engine power information set for each fuel and Euro standardcategory as documented in Table 14. Insufficient Euro 1 petrol and Euro 0 diesel cars wereobserved, so their specifications were assumed. The Euro 6 passenger cars observed arepredominantly from the ‘upper medium’ and ‘executive’ market segments. This is reflected intheir higher (average) engine power rating.

Table 14. Average Aberdeen passenger car specifications.

FuelType

EuroStandard

Weight (kg) Enginepower (kW)

Frontal area(m2)

Petrol 0 1299.3 96.4 2.36

1 1299.3 96.4 2.36

2 1278.0 94.1 2.63

3 1356.0 92.5 2.67

4 1270.1 89.5 2.69

5 1252.8 86.5 2.77

6 1329.2 123.0 2.65

Diesel 0 1665.0 77.0 2.95

1 1665.0 77.0 2.95

2 1465.9 77.8 2.82

3 1641.1 95.7 2.92

4 1672.8 106.4 2.96

5 1655.0 113.3 3.03

6 1607.8 133.1 2.90


55

The influence of road gradient was also considered. The gradient each one-second time-step ofthe vehicle tracking data established from a Digital Terrain Model described in section 3.4) andcombined with the speed trajectories for the IEM simulations. All the PHEM simulation resultsassumed vehicles were in a ‘hot-running’ condition.

56

Pollutant

PETROL DIESEL

Compact Small Family Mini MPV

CO2

NOX

Figure 38. Scatter plots comparing Modelled (PHEM) and Observed values for all sections (urban, suburban and motorway) of the LDC.(Top-row) CO2 (a) Petrol compact car (left); (b) Petrol small family car (middle-left); (c) Diesel mini car (middle-right); (d) Diesel MPV

(right); (Bottom-row) NOX (e) Diesel mini car (middle-right); (f) Diesel MPV (right). NOTE: Petrol car NOX emissions at a low-level and notillustrated

Modelled_CO2

Ob

serv

ed_

CO

2

0

2

4

6

0 2 4 6

Counts

12346

101625396298156247390618978

1547

Modelled_CO2

Ob

serv

ed_

CO

2

0

2

4

6

8

0 2 4 6 8 10

Counts

12346

101625406399157249394624989

1566

Modelled_CO2

Ob

serv

ed_

CO

2

0

2

4

6

8

0 2 4 6

Counts

12346

101625396298155244386611966

1528

Modelled_CO2

Ob

serv

ed_

CO

2

0

5

10

15

0 5 10

Counts

112358

111726385786129194290435653

Modelled_NOx

Ob

serv

ed_

NO

x

0.00

0.02

0.04

0.06

0.08

0.10

0.00 0.02 0.04 0.06

Counts

12347

1219315082134219357582950

15502530

Modelled_NOx

Ob

serv

ed_

NO

x

0.00

0.05

0.10

0.00 0.02 0.04 0.06 0.08

Counts

122469

1421325077119184285440680

1050

57

4.7.2 PHEM results

The time series plots in Figure 39 presents PHEM’s capability to predict the tail-pipeemissions from a sample trajectory (City Centre – route 5) second-by-second. Thevehicle illustrated is a Euro 5 diesel car. It is evident fuel burnt (correlated with CO2

emissions) and tail-pipe emissions of NOX, NO2 and PM are elevated duringacceleration events.

Figure 39. Time series plot of PHEM results for a sample Euro 5 diesel car circulatingAberdeen City centre (a) Speed, (b) Road gradient, (c) CO2, (d) NOX and NO2, (e)Particle Mass (PM).

Simulations of all diesel and petrol, Euro standard 0 -6 passenger cars over all Aberdeenvehicle tracking routes (N1-15, see Table 4) were undertaken. The average CO2, NOX

and PM Emission Factors are illustrated in Figures 40, 41 and 42 respectively. There isa good deal of variation in the average emission factors over the 15 vehicle trackingdatasets. This is because some vehicle tracking tests contained more fuel intensive stop-start driving conditions (e.g. route 15 annotated in grey) than others where the testvehicle was mainly cruising i.e. route 12 (annotated in brown). The ranking of highestto lowest EFs is consistent across the pollutants CO2, NOX and PM.

The fuel efficiency of diesel cars is broadly flat through the Euro standards 0 to 6. TheEuro 2 category is lower as the average observed Euro 2 weighed ≈200kg less than theother categories of diesel car. Although there have been improvements in the fuelefficiency of diesel engines and powertrains, this have been broadly offset by the fuelpenalty of emission controls that have had to be introduced and installed on the Euro 4,5 and 6 i.e. DPFs. Petrol conversely already emit air quality pollutants at a low leveldue to the success and robustness of three way catalytic (TWC) converters.Development can therefore focus on improving fuel efficiency, reflected in the lowerCO2 emissions per km driven for the Euro 5 and 6 petrol passenger cars.

Diesel passenger car NOX EFs for the 15 routes vary by ±35% the average. Asillustrated in the time series figure (39), NOX emissions like CO2 are highest during

0 500 1000 1500 2000

01

030

50

Sp

ee

d(k

m.h1)

0 500 1000 1500 2000

-6-2

02

46

Gra

die

nt(%

)

0 500 1000 1500 2000

02

46

CO

2(g

.se

c1)

0 500 1000 1500 2000

01

02

030

40

Time (seconds)

NO

X(m

g.s

ec1)

0 500 1000 1500 2000

10

20

30

40

50

Time (seconds)

PM

(µg

.se

c1)

58


acceleration events, so tests with more stop-start driving and less cruising phases resultin the highest EFs. NOX Emission controls on Euro 6 diesel cars and more modern(Euro 4 – 6) petrol cars are predicted to lower emissions of this important air qualitypollutant.

Figure 40. Modelled passenger car CO2 emission factors (a) diesel, (b) petrol.

Figure 41. Modelled passenger car NOX emission factors (a) diesel, (b) petrol

Particle Mass (PM) and Number Count (PNC) are predicted to be low for all generationsof petrol passenger cars. Diesel particle emission controls are shown to reduce

Euro standard

NO

X(g

ram

s.km

1)

0.0

0.2

0.4

0.6

0.8

1.0

E0 E1 E2 E3 E4 E5 E6

DIESEL

E0 E1 E2 E3 E4 E5 E6

PETROL

Euro standard

CO

2(g

ram

s.km

1)

200

300

400

500

600

E0 E1 E2 E3 E4 E5 E6

DIESEL

E0 E1 E2 E3 E4 E5 E6

PETROL

59


emissions, with DPFs lowering PM and PNC to a lower level for Euro 5 and 6generations of cars. This approach of collating real driving behaviour (speeds) andpredicting tail-pipe particle emissions offers a rare opportunity to understand PM andPNC emissions in real-driving. PM and PNC tail-pipe emission measurements arenotoriously challenging to carry-out on-the-road and are limited to laboratory studieswith sophisticated dilution systems and instrumentation.

Figure 42. Modelled passenger car PM emission factors (a) diesel, (b) petrol

Figure 43. Modelled passenger car Particle Number Count (PNC) emission factors(a) diesel, (b) petrol

Euro standard

Pa

rtic

leN

um

be

r(#

.km

1)

0.0e+00

2.0e+13

4.0e+13

6.0e+13

8.0e+13

1.0e+14

1.2e+14

E0 E1 E2 E3 E4 E5 E6

DIESEL

E0 E1 E2 E3 E4 E5 E6

PETROL

Euro standard

PM

(gra

ms.k

m1)

0.00

0.05

0.10

0.15

E0 E1 E2 E3 E4 E5 E6

DIESEL

E0 E1 E2 E3 E4 E5 E6

PETROL

60


4.7.3 Comparison of RSD, PHEM and EFT NOX Emission Factors

The estimated passenger car Emission Factors from: the Remote Sensing Device (RSD)measurements in Aberdeen, the IEM (PHEM) using tracked Real Driving speed profilesin Aberdeen, are compared with those in the latest version of EFT (Emission FactorToolkit | DEFRA, 2015) in Figure 44.

The RSD and IEM NOX EFs based on local information are broadly of the samemagnitude and follow the same trends. The fact that the EFs from two approaches usinglocal information, state-of-the-art instruments and modelling techniques are in closeagreement suggests they are robust. The EFT NOX EFs for diesel passenger are only≈half those predicted by the RSD and PHEM approaches. EFT is widely considered tounder-estimate diesel NOX emissions (Carslaw et al, 2011). Importantly all methods,whether the local observations or modelled results, indicate NOX emissions from Euro6 diesel cars are ≈half those of earlier generations (Euro 3 to 5). The NOX emissionfactors for Euro 4, 5 and 6 are all higher than the levels set in the European emissionstandards over the NEDC test cycle in laboratory conditions. The RSD NOX EFs forEuro 5 and 6 diesel passenger cars are 4.8 and 5.5 times greater than the standardsrespectively. The discrepancy between the EFT and Euro standard EFs is considerablyless at 2.4 and 1.9 times respectively. NOX EFs for petrol cars from the RSD areconsistently higher than from the models, although all follow the same down trend(emission reduction) through the Euro standards. The RSD may be reporting higherNOX emissions from petrol cars because:

The EFT and IEM (PHEM) modelled results are for vehicles in a hot-runningcondition. i.e. no cold-start effect. NOX emissions from petrol cars are high duringcold-starts when the TWC has not yet reached its operational temperature;

The EFT and IEM (PHEM) modelled results do not account for the in-servicedeterioration of vehicles and their emission controls, with the majority of under-lying emission maps generated from laboratory measurements of new/ low-mileagevehicles. The discrepancy because of the deterioration of emission controls ishowever expected to be greater for older vehicles i.e. Euro 0 -2, and not new Euro5 and 6 cars; and

The RSD measurement of clean petrol cars may be elevated due to contaminationfrom emission plumes from traffic travelling in the opposite direction (lane).

Further research is needed to identify the relative impact of each of these factors.

Figure 44. Comparison of RSD, PHEM and EFT NOX Emission Factors (a) diesel, (b)petrol.

Euro standard

NO

X(g

ram

s.km

1)

0.0

0.2

0.4

0.6

0.8

1.0

E0 E1 E2 E3 E4 E5 E6

DIESEL

E0 E1 E2 E3 E4 E5 E6

PETROL

61


5. Recommendations for Future Work

The RSD/ VRM look-up approach can rapidly detect and provide an early indicationof the performance of emerging vehicle technologies. Historically there has been agreater than one-year time lag between the introduction of a new Euro standard, andthe release of an assessment of their emission performance from laboratories. Thesuccess or otherwise of the Euro 6 emission standard legislation is of criticalimportance to the levels of nitrogen dioxide in the medium-term, and whetherexceedances of the NO2 annual mean limit value will be curbed at heavily traffickedlocations. A continuing programme of vehicle emission remote sensing campaignsin the UK should be conducted so an early assessment of the environmentalperformance of the Euro 6 generation of vehicles, from different manufacturingclusters, can be established. It is important this programme is extended until Euro 6LGVs are part of the operational fleet;

Methods to use the remote sensing emission measurements as a real-time ‘high-emitting’ vehicle screening tool could be explored. With support from the VOSA(http://www.dft.gov.uk/vosa/) Agency, vehicles classified in real-time as ‘high-emitters’ could be pulled in for subsequent testing at a near-by Authorised TestingFacility in an attempt to reduce emissions from the worst offending vehicles. It isrecommended identifying and assessing the significance of DPF removal is apriority for such an exercise;

Emerging research instruments (Burgard et al, 2006) in the US have recentlyextended the RSD capability to sense nitrogen dioxide (NO2) and ammonia (NH3).A study with this research instrument has been carried out in Greater London in thesummer of 2012 and 2013. It is recommended this instrument is again imported anddeployed on UK road’s to identify NO2 fractions in Euro 6 / VI vehicles and reviewthe trends in NO2 fractions (degradation);

Develop a detailed traffic and vehicle emission modelling capability to help designand provide improved, reliable assessments of the Aberdeen City Council AirQuality Action Plans. It is now practical to couple traffic microsimulations, whichmodel the movement of individual vehicles through the well specified roadnetwork, with the instantaneous vehicle emission model (IEM) that providessecond-by-second fuel consumption and tail-pipe emission predictions. As themodelling approach is a step towards a second-by-second “virtual” representationof the “real” traffic network, it naturally encapsulates many events and processesthat effect vehicle emissions. Urban Buses for example have to make additionalstops-and-starts to pick up passengers on their scheduled routes. As Buses are large,heavy-duty diesel vehicles these stop-start motions have a significant fuel andemission penalty. The proposed detailed modelling approach can simulate andconsider this behaviour, more aggregate approaches do not.

Advances in desktop computing mean it is also now feasible to micro-simulatetraffic movements at a detailed level as they negotiate traffic junctions, signals andinteract with each other for complete City networks e.g. Tate, 2012. Microscopictraffic simulation models are frequently used by transport engineers and planners toanalyse traffic operations and evaluate management strategies. It is therefore likelythat traffic microsimulation networks developed for transport planning purposesmay be available to support improved environmental assessments.

62


6. References

Bishop, G., Starkey, J., Ihlenfeldt, A. 1989. IR long-path photometry: a remotesensing tool for automobile emissions. Analytical Chemistry, 61: 671A.

Borken-Kleefeld, J., Kupiainen, K., Chen, Y., Hausberger, S., Rexeis, M., Sjodin, A.,Jerksjo, M., Tate, J. High-emitting vehicles in laboratory and on-roadmeasurements. 19th International Transport and Air Pollution Conference,Thessaloniki, Greece, 26-27 November, 2012

Burgard, D.A.; Dalton, T.R.; Bishop, G.A.; Starkey, J.R.; Stedman, D.H. 2006. ANitrogen Dioxide, Sulfur Dioxide, and Ammonia Detector for Remote Sensingof Vehicle Emissions. Rev. Sci. Instrum., 2006, 77, 014101/1-014101/5 [DOI:10.1063/1.2162432].

Cadle, S., Stephens, R. 1994. Remote sensing of vehicle exhaust emissions.Environmental Science & Technology, 28: 258A-.

Carslaw, D. 2005. Evidence of an increasing NO2/NOX emissions ratio from roadtraffic emissions. Atmospheric Env. 39, pp. 4793 – 4802.

Carslaw, D. and Beevers, S. 2005. Estimations of road vehicle primary NO2 exhaustemission fractions using monitoring data in London. Atmospheric Env. 39, pp167-177

Carslaw, D., Beevers, S. Westmoreland, E. Williams, M. Tate, J. Murrells, T.Stedman, J. Li, Y., Grice, S., Kent, A. and I. Tsagatakis. 2011. Trends in NOX

and NO2 emissions and ambient. [Accessed 9th March 2011] http://uk-air.defra.gov.uk/reports/cat05/1103041401_110303_Draft_NOx_NO2_trends_report.pdf

Carslaw, D., Williams, M., Tate, J., Beevers, S. 2012. The importance of high vehiclepower for passenger car emissions. Atmospheric Environment, In press,10.1016/j.atmosenv.2012.11.033

Carslaw, D., Rhys-Tyler, G. 2013. Remote sensing of NO2 exhaust emissions fromroad vehicles, A report to Defra. King’s College London and NewcastleUniversity, Draft 11th April 2013.

Carslaw, D., Priestman, M. 2014. Analysis of the 2013 vehicle emission remotesensing campaigns data. King’s College London, Final draft, 10th November2014

DfT (2015). Percentage empty running and loading factors by type and weight ofvehicle: annual 2000 – 2014. https://www.gov.uk/government/statistical-data-sets/rfs01-goods-lifted-and-distance-hauled [Accessed 29/02/2016]

Grice, S., Stedman, J., Kent, A., Hobson, M., Norris, J., Abbott, J., Cooke, S. 2009.Recent trends and projections of primary NO2 emissions in Europe.Atmospheric Env. 43, pp. 2154 – 2167.

Hausberger, S. 2010. Fuel Consumption and Emissions of Modern Passenger Cars.Report Nr. I-25/10 Haus-Em 07/10/676 from 29.11.2010. Technical Universityof Graz, AT.

Hausberger, S., Rexeis, M., Zallinger, M., Luz, R. 2011. User Guide to the PHEMEmission Model. Version 10. Technical University of Graz, Austria, January2011.

63


HEBFA. 2010. Handbook emission factors for road transport (HBEFA). Version 3.1,Jan. 2010. [Accessed 18th September 2012] http://www.hbefa.net/

Jerksjö, M., Sjödin, A., Bishop, G.A., Stedman, D.H., 2008. On-road emissionperformance of a European vehicle fleet over the period 1991-2007 as measuredby remote sensing. 18th CRC On-Road Vehicle Emissions Workshop SanDiego, March 31-April 2, 2008.

Jiminez-Palacios, J. 1999. Understanding and Quantifying Motor Vehicle Emissionswith Vehicle Specific Power and TILDAS Remote Sensing. PhD Thesis,Department of Mechanical Engineering, Massachusetts Institute of Technology.

National Audit Office. 2009. Briefing for the House of Commons EnvironmentalAudit Committee. December 2009. [Accessed 12th January 2010]http://www.nao.org.uk/publications/0910/eac_briefing_-_air_quality.aspx

Popp, P., Bishop, G., Stedman, D. 1997. Development of a High-Speed UltravioletSpectrophotometer Capable of Real-Time NO and Aromatic HydrocarbonDetection in Vehicle Exhaust. 7th CRC On-Road Vehicle Emissions Workshop,San Diego, CA.

R Development Core Team. 2006. R: A language and environment for statisticalcomputing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.

SMMT. 2011. New car market betters forecast but was down 4.4% in 2011 to 1.94million. The Society of Motor Manufacturers and Traders (SMMT), 6th ofJanuary 2012. [Accessed 6th January 2012]http://www.smmt.co.uk/2012/01/new-car-market-bettersforecast-but-was-down-4-4-in-2011-to-1-94-million/

Stedman, D., Bishop, G., Aldrete, P. 1997. On-road evaluation of an automobileemission test program. Environmental Science & Technology, 31: 927-31.

Tate, J. 2010. The on-road emission characteristics of UK passenger cars: Theapplication of a remote sensing vehicle emission measurement system.Transport and Air Pollution Conference, Zurich, 2010.

Tate, J. 2012. Assessing Low Emission Zone and Demand Management policies usingdetailed traffic and vehicle emission measurements and coupled models. 19thInternational Transport and Air Pollution Conference, Thessaloniki, Greece, 26-27 November, 2012

Tate, J. 2012. Modelling Low Emission Zone and Demand Management measures inthe City of York using Detailed Traffic-emission Tools. Investigation of AirPollution Standing Conference, June 2012, London. [Accessed 24th July 2012]http://www.iapsc.org.uk/document/0612_J_Tate_for_web.pdf

Volvo Buses. 2011. The only commercially viable hybrid. [Accessed 6th January2012] http://www.volvobuses.com/bus/global/engb/products/City%20buses/Volvo%207700%20Hybrid/Pages/Introduction_new.aspx

Wright Bus. 2011. Hybrid-electric vehicles a clean way forward. [Accessed 6thJanuary 2012]

WHO. 2012. Diesel Engine Exhaust Carcinogenic. International Agency for Researchon Cancer, World Health Organisation. [Accessed 12th July 2012]http://press.iarc.fr/pr213_E.pdf

64


Wyatt, D., Li, H., Tate, J. 2014. The impact of road grade on carbon dioxide (CO2)emission of a passenger vehicle in real-world driving. Transportation ResearchPart D: Transport and Environment, 32, pp 160-170, October 2014, DOI:10.1016/j.trd.2014.07.015

Zhang, Y., Stedman, D., Bishop, G., Beaton, S., Guenther, P., McVey, I. 1996.Enhancement of Remote Sensing for Mobile Source Nitric Oxide. J. Air WasteManage. Assoc, 46: 25-29.

65

Appendix A. Summary estimated Passenger Car CO2 and NOX, emission factors (grams.km-1): Remote Sensing, PHEM (IEM),EFT and EU Emission Standards.

Remote Sensing Estimates PHEM (IEM)

(Aberdeen Real Driving profiles)

EFT EU EmissionStandards

VEHICLEtype

FUELtype

EUROStandard

Numberof

vehicles

CO2

[Scaled ‘Combined’Type Approval figures]

(grams.km-1)

NOX

(grams.km-1)

CO2

(grams.km-1)

NOX

(grams.km-1)

NOX

(grams.km-1)

NOX

(grams.km-1)

Car Petrol Euro 0 - - - 0.65 0.89 -

Euro 1 - - 294 0.37 0.40 -

Euro 2 0.49 200 277 0.26 0.24 -

Euro 3 0.25 206 277 0.08 0.06 0.15

Euro 4 0.15 195 285 0.06 0.05 0.08

Euro 5 0.13 175 232 0.04 0.03 0.06

Euro 6 0.11 180 215 0.04 0.03 0.06

Diesel Euro 0 - - 273 0.72 0.36 -

Euro 1 - - 252 0.75 0.40 -

Euro 2 1.11 202 228 0.84 0.44 -

Euro 3 0.84 196 260 0.82 0.41 0.50

Euro 4 0.82 201 261 0.68 0.35 0.25

Euro 5 0.86 183 240 0.77 0.43 0.18

Euro 6 0.44 169 240 0.27 0.15 0.08

66

Appendix B1. Frequency diagrams of the observed Speed and Acceleration distributions from the Vehicle tracking routes 1 to 15(see Figure 3 and Table 4)

1 2 3 4 5

VE

HIC

LE

TR

AC

KIN

G

6 7 8 9 10

11 12 13 14 15

0 10 20 30 40 50 60

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

2

3

4

5

6

7

8

9

10

11

Counts

0 10 20 30 40 50 60

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

123456789

1011121314151617

Counts

0 10 20 30 40 50 60

-3

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1368

10131517202224262931333638

Counts

0 10 20 30 40 50

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

12345678

101112131415161718

Counts

0 10 20 30 40 50 60

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1245678

10111214151617182021

Counts

0 10 20 30 40 50 60

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

13568

101214161719212325262830

Counts

0 10 20 30 40 50

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

12345678

101112131415161718

Counts

0 10 20 30 40

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

23

45

6

7

8

9

10

11

12

1314

15

Counts

0 10 20 30 40

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

124578

1011121415171820212324

Counts

0 10 20 30 40 50 60

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

13568

101214161719212325262830

Counts

0 10 20 30 40

-3

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

13579

111315171921232527293133

Counts

0 10 20 30 40 50

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1469

11141619222427293234373942

Counts

0 20 40 60 80 100

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1368

10121417192124262830323537

Counts

0 10 20 30 40 50 60

-3

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

148

1114182124283134374144475154

Counts

0 10 20 30 40

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

12356789

101213141516181920

Counts

67


Appendix B2. Frequency diagrams of the observed Speed and Acceleration distributions from the 5 Vehicle Emission remotesensing Measurement Sites (see Table 1): (1) King Street (2) Bridge of Dee (3) Beach Boulevard (4) Westburn Road and (5) GreatNorth Road

1 2 3 4 5

RS

D

10 20 30 40 50

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

4

7

11

14

17

20

23

26

30

33

36

39

42

46

49

52

Counts

10 20 30 40 50

-3

-2

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

9

16

24

32

40

48

55

63

71

78

86

94

102

110

117

125

Counts

10 20 30 40 50 60

-2

-1

0

1

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

2

3

4

6

7

8

9

10

11

12

13

14

16

17

18

19

Counts

10 20 30 40 50

-0.5

0

0.5

1

1.5

2

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

3

4

6

8

9

11

13

14

16

18

20

21

23

25

26

28

Counts

10 20 30 40 50 60 70

-1

0

1

2

3

Speed (km.h1

)

Acce

lera

tio

n(m

.se

c2)

1

6

11

16

22

27

32

37

42

47

52

57

62

68

73

78

83

Counts

Documents

Vehicle Emission Measurement and Analysis - Aberdeen City ... · Dr James Tate Remote Sensing Vehicle Emissions - ABERDEEN ... Aberdeen (April 2015). If Clean Air Zones (CAZs) are