WG2 Prague 111006 Zachariadis simul sustain transp

Engineering-economic simulations of sustainable transport policies: 

Engineering-economic simulations of sustainable transport policies Theodoros Zachariadis Economics Research Centre, University of Cyprus P.O. Box 20537, 1678 Nicosia, Cyprus t.zachariadis@ucy.ac.cy COST 355 meeting Prague, October 2006

Environmental impact of energy systems: the “engineering approach”: 

Environmental impact of energy systems: the “engineering approach” Emphasis on technological dimension “Bottom-up” approach Detailed simulation of physical/chemical processes (flows, chemical reactions, mass/energy/momentum balances) and/or experimental determination of system properties Evaluation of future technologies based on their technical potential (ΒΑΤ – Best Available Technology)

“Engineering approach” for assessment of vehicle emission abatement strategies: 

“Engineering approach” for assessment of vehicle emission abatement strategies Experimental determination of emissions (chassis/engine dynamometer, exhaust gas analysers, mass balances) Emission factors (g pollutant / km) as a function of average vehicle speed/acceleration Extra emissions per vehicle due to engine/catalyst cold start & fuel evaporation Future evolution of basic variables (vehicle population, distance travelled per vehicle, average driving speed) are simulated phenomenologically Evaluation of future technologies on the basis of research results & engineering knowledge of their technical potential

However, decision-making requires to know:: 

However, decision-making requires to know: Cost constraints Current costs (investment, operation & maintenance, fuel) Economies of scale Learning processes Infrastructure development costs Subjective costs (e.g. discomfort) Consumer/producer behaviour Disposable income Substitution effects Inertia & myopia Rebound effects Overall economic background (e.g. GDP, fuel prices, taxes/subsidies)  Simulations are necessary that account for fundamental (micro)economic principles

A long-term engineering-economic model for the EU transport sector: 

A long-term engineering-economic model for the EU transport sector - Model was developed: at the National Technical University of Athens, within the MINIMA-SUD project (Methodologies to Integrate Impact Assessment in the Field of Sustainable Development) funded by the EC (5th Framework Programme) for each EU 15 country for all transport sectors (passenger/freight, road/rail/air/sea) - Runs year by year up to 2030 - Is calibrated so as to fit official statistics in base year and partly reproduce existing forecasts - Calculates transportation energy consumption, pollutant & greenhouse gas emissions + noise, congestion & road fatalities indicators

Model development – 1: 

Model development – 1 Total expenditure on transport depends on private income (for passenger transport) or weighted industrial+agricultural value added (for freight transport) and average user price of transport A microeconomic optimisation framework is assumed for the allocation of total expenditure between transport modes: Maximisation of consumer utility for passenger transport Minimisation of transport costs for freight transport

Model structure – 1 : 

Model structure – 1

Model structure – 2 : 

Model structure – 2

Model development – 2: 

Model development – 2 Consumer and producer choices are described as a series of separable choices, which create a nesting structure (decision tree). Utility/cost functions at each level of the decision trees are Constant Elasticity of Substitution (CES) functions: q: quantity (pkm/tkm), σ: elasticity of substitution, Y: income, p: generalised price (Euro’00 per pkm/tkm), αi: share parameter

Utility tree for non-urban passenger transport: 

Utility tree for non-urban passenger transport

Model development – 3: 

Model development – 3 Aim: Maximise U subject to budget constraint Y Solution for CES utility/cost function assuming l levels of utility tree: σl available from TREMOVE Model calibration: determination of αi From exogenous reference case, qi, pi are available  αi are calculated  model can reproduce reference case and perform scenario runs

Generalised price concept: 

Generalised price concept Generalised price reflects monetary + time costs, i.e.: Vehicle purchase costs Registration and circulation taxes Maintenance costs Insurance costs Fuel costs Public transport fares Time costs = [(travel time)+(waiting time)] / (avg. distance travelled) * (value of time) (Travel time) = (speed)-1 [min] Value of Time (Euro’00 per passenger/tonne per hour): different for each transport mode, road type, peak/off-peak travel

Generalised price concept – 2 : 

Generalised price concept – 2 Congestion function: with invex investment expenditure in road infrastructure parkex investment expenditure in parking space m vehicle type, b in the baseline, s in a scenario r1,r2,r3 adjustment factors LF load factors PCU passenger car units p,f indices for passenger and freight transport

Congestion : 

Congestion Congestion-related sustainability indicator: Total travel time (hours spent travelling in a vehicle per year, by road type) with kmv average distance travelled annually per vehicle of each type

Road accidents/fatalities indicator – 1: 

Road accidents/fatalities indicator – 1 Number of road accidents: with ACC road injury accidents in thousands vkm billion road vehicle kilometres a,b country-specific parameters (estimated from statistics of the period 1980-2000) n type of area studied (built-up or non-built-up)

Road accidents/fatalities indicator – 2: 

Road accidents/fatalities indicator – 2 Road fatalities: with F number of deaths in road accidents af,bf country-specific parameters estimated from statistics of the period 1970-2000 t time in years, with t=0 for 1970.

Noise indicator : 

Noise indicator Like air pollution, noise annoyance is addressed through an ‘emissions’ approach, i.e. emitted sonar energy Most common indicator: A-weighted equivalent noise level Leq, expressed in db(A) Base year noise emissions come from the TRENDS project (Keller et al., 2002) Future emissions calculated with UBA Vienna approach: with Leq noise emissions level in db(A) MSV total vehicle kilometres driven p share of heavy duty vehicles in traffic v average driving speed

Running a scenario: 

Running a scenario In a scenario (evaluation of a policy instrument), some transport demand quantities or prices in the model change This changes also generalised prices / demand quantities / congestion This will feed back to a further change in quantities / prices / congestion After some iterations, the new equilibrium prices and quantities are determined for each year; this is the model solution for that scenario

Calculation of road vehicle stock: 

Calculation of road vehicle stock pkm/tkm and prices available from model solution Annual vehicle mileage by vehicle size/road type evolves as a function of income and oil prices Occupancy rates of cars decrease with time as a result of rising income and declining household size With the aid of the above assumptions, vehicle stock is calculated for several fuel/size groups

Vehicle fuel/size groups: 

Vehicle fuel/size groups

Allocation of vehicle stock into vintages: 

Vehicle stock is decomposed into age cohorts, according to an initial age distribution in base year assumptions on evolution of scrapping rates Scrapping is simulated through a modified Weibull function: with φ(k) survival probability, k age in years, b,T parameters with C the total lifetime cost of a new car, b in the baseline, s in a scenario Allocation of vehicle stock into vintages

Determination of technology shares : 

Determination of technology shares Choice of technology in road transport is driven by Emissions legislation (within the same fuel/size group) Relative user prices, determined from vehicle, maintenance and fuel costs The model includes the 113 technology classes of the COPERT III methodology + alternative vehicle technologies/ fuels: CNG, methanol, ethanol, fuel cells, electricity Simpler approach for non-road transport modes New registrations change average technical and economic properties of each vehicle fuel/size group  For subsequent years, technical and economic data are updated with new technology shares Emissions calculated: NOx, NMVOC, SO2, PM, Pb, CO2

Sample list of ‘conventional’ road vehicle technologies: 

Sample list of ‘conventional’ road vehicle technologies

Examples of emission and fuel consumption factors of various vehicle types: 

Examples of emission and fuel consumption factors of various vehicle types

Major data sources for the transport model – 1 : 

Major data sources for the transport model – 1 Eurostat (NewCronos database): energy balances, vehicle stock data, macroeconomic data, energy prices & taxes DG TREN Statistical Pocketbook ‘Energy and Transport in Figures’: pkm/tkm data, total vehicle stock, road fatalities Eurostat/EEA (TERM report): vkm data for all transport modes ECMT/UNECE/Eurostat Pilot Survey on the Road Vehicle Fleet in 55 countries EC TRACE project (1999): data on value of time by country, vehicle type and road type UITP (International Public Transport Union): fares for buses, tram & metro AEA (Association of European Airlines): air transport fares

Major data sources for the transport model – 2 : 

Major data sources for the transport model – 2 TREMOVE base case results of Auto-Oil II application: vehicle costs, evolution of traffic activity by fuel/size group up to 2020, urban/non-urban split, peak/off-peak split up to 2020 COPERT III methodology & computer model: emission factors and overall calculation scheme for road vehicle emissions (conventional technologies/fuels only) TRENDS database: age & technology distribution of road vehicles in base year, emission and fuel consumption factors for non-road vehicles MEET project: emission and fuel consumption factors for alternative vehicle technologies/fuels and for future non-road vehicles Other studies for costs and fuel consumption of alternative vehicle technologies/fuels

Base year calibration procedure : 

Base year calibration procedure

Baseline assumptions : 

Baseline assumptions Demographic, economic and energy price developments in line with those of the PRIMES model used in DG TREN’s “European Energy and Transport – Trends to 2030” (September 2003) Emissions and fuel consumption of future transport modes (cars, trucks, trains, aircraft, sea vessels) in line with the relevant literature Higher share of new diesel car sales up to 2010 (partly due to ACEA agreement)

Baseline results in EU 15 – 1: 

Baseline results in EU 15 – 1

Cost of passenger transport –1: 

Cost of passenger transport –1

Cost of passenger transport – 2: 

Cost of passenger transport – 2

Cost of freight transport – 1: 

Cost of freight transport – 1

Cost of freight transport – 2: 

Cost of freight transport – 2

Evolution of passenger activity by mode EU 15: 

Evolution of passenger activity by mode EU 15

Evolution of freight activity by modeEU 15: 

Evolution of freight activity by mode EU 15

Baseline results in EU 15 – 2: 

Baseline results in EU 15 – 2

Urban vehicle kilometres as a fraction of total vehicle kilometres by country (%): 

Urban vehicle kilometres as a fraction of total vehicle kilometres by country (%)

Baseline evolution of NOx emissions (kt)EU 15: 

Baseline evolution of NOx emissions (kt) EU 15

Baseline evolution of NOx emission intensityEU 15: 

Baseline evolution of NOx emission intensity EU 15

Baseline evolution of PM emissions (kt)EU 15: 

Baseline evolution of PM emissions (kt) EU 15

Baseline evolution of CO2 emissions (Mt)EU 15: 

Baseline evolution of CO2 emissions (Mt) EU 15

Baseline evolution of other sustainability indicators: 

Baseline evolution of other sustainability indicators

Policy exercises applied : 

Policy exercises applied Subsidies to CNG and fuel cell vehicles (50% of their pre-tax purchase cost) Double tax on automotive diesel fuel for cars/trucks Advanced emission standards from 2006 onwards (‘Euro V’), but at 40% higher purchase costs Double investment expenditure for road infrastructure (current figures: 55 billion Euros’00 in 2000, 69 billion Euros’00 in 2010) Subsidies to public transport fares (50% lower fares) Road pricing: 3 Euros for each urban trip on average Subsidies for scrapping old cars: 50% lower purchase cost for each new car replacing an old one Combination of policies 3 & 6 Combination of policies 1, 3 & 6 Combination of policies 3, 5 & 6

Impacts of policies on sustainability indicators in year 2020 (% change from baseline): 

Impacts of policies on sustainability indicators in year 2020 (% change from baseline)

Impact of policy exercise 1 (subsidies to alternative fuel vehicles) : 

Impact of policy exercise 1 (subsidies to alternative fuel vehicles) Negligible changes in generalised prices, aggregate transport activity and travelling speeds Major change in the fuel mix in the medium term Alternative fuel vehicles will account for 3%, 18%, 11% and 29% of the total fleet of cars, buses, light trucks and heavy trucks respectively in 2020 Total energy consumption falls by 1.8% in 2020 and 3.6% in 2030, and CO2 emissions fall by 1.5% in 2020 and 2.8% in 2030 Demand for energy and methanol rises by 13 and 5 times respectively in 2020 Pollutant emissions decline moderately

Impact of policy exercise 2 (double diesel tax) : 

Impact of policy exercise 2 (double diesel tax) Total user costs of passenger cars and trucks increase by less than 2% in urban areas and by 2% and 7% in non-urban areas respectively Diesel car pkm fall by ~10% in urban peak driving Total car pkm fall by only 1% (shift from diesel to gasoline cars) Truck tkm decrease by ~4% to the benefit of rail Gasoline-related pollutants (NMVOC, Pb) increase Diesel-related pollutants (NOx, PM and SO2) decline

Impact of policy exercise 3 (emission standards) : 

Impact of policy exercise 3 (emission standards) Car travel becomes more expensive, particularly in non-urban areas  improved congestion levels  reduced costs for trucks because of lower time costs Urban public transport gains ~5% after 2020 High-speed rail and aviation gain ~4% after 2020 Energy demand falls by 8.2% in 2020 and by 9.5% in 2030, with some switch to alternative fuels Urban NOx / PM decline by 18% and 39% in 2030 Higher driving speeds cause a 3% increase in injury accidents and a 4% increase in road fatalities by 2020-2030

Impact of policy exercise 4 (investment expenditure for roads) : 

Impact of policy exercise 4 (investment expenditure for roads) Total time spent in urban driving declines by 6% Driving becomes somewhat cheaper (by ~4% in urban areas and by <1% in motorways) Impact not very remarkable because of ‘rebound effect’: improved congestion makes car travel more attractive  road pkm/tkm & energy intensity increase Largest benefit for freight transport due to higher share of time costs Pollutant emissions change by ±3% Negligible impact on accidents Some increase in noise levels

Impact of policy exercise 5 (subsidies for public transport fares) : 

Impact of policy exercise 5 (subsidies for public transport fares) Public transport use becomes much more attractive, with bus/rail pkm rising by 13% and 28% respectively in 2020 The use of cars is not affected considerably (only -1%) ‘Rebound effect’ again All other indicators remain essentially unchanged

Impact of policy exercise 6 (road pricing) : 

Impact of policy exercise 6 (road pricing) Immediate (short-term) effects are observed Urban travel costs rise by more than 20% for cars and light trucks  average urban peak speeds 20% higher  congestion improves by >4%  pkm of urban buses and tram/metro rise by >25% and >15% respectively in 2010-2020 4.4% lower transportation energy demand and CO2 emissions by 2010 Air pollutants fall by 5-15% in urban areas Fatalities in road accidents increase by >15% due to higher urban travel speeds

Impact of policy exercise 7 (scrapping subsidies) : 

Impact of policy exercise 7 (scrapping subsidies) Travel costs, car ownership and car use are not affected (subsidies address only new cars that replace old ones) Significant acceleration in car renewal: scrapping rates increase by >15% Average age of the passenger car stock falls from ~7.3 years in 2010-2020 in the baseline to ~6.8 years in this scenario Cars up to 5 years of age are ~10% more than in the baseline; old cars fall by up to 60% Energy use and pollutant emissions decrease particularly in the 2006-2015 period Pollutant emissions of cars decrease, but total transportation emissions fall by small amounts Negligible impact on congestion, noise and accident rates from application of this instrument

Impact of combined policy exercises (8, 9, 10) : 

Impact of combined policy exercises (8, 9, 10) Impacts of scenarios 3 and 6 are added up in the case of policy 8 (stricter standards + road pricing): large improvements in energy & CO2 (-12%) and pollutant emissions (-25% in urban PM) but also 20% more accident fatalities In policy 9 (= policy 8 + alternative fuel subsidies) all indicators except fatalities improve more than in policy 8 Effectiveness of policy 10 (= stricter standards + road pricing + public transport subsidies) is a bit lower than policy 9 because of ‘rebound effect’

Cumulative impact of selected policies – 1: 

Cumulative impact of selected policies – 1

Cumulative impact of selected policies – 2: 

Cumulative impact of selected policies – 2

Synopsis: 

Synopsis For the formulation of effective sustainable development strategies it is necessary to combine and reconcile: Engineering approaches (detailed evaluation of technical measures) Economic approaches (costs, international economic context, consumer/producer behaviour, feedback mechanisms)  Development of engineering-economic models Evaluation of costs (direct and indirect) is crucial

Acknowledgements: 

Acknowledgements MINIMA-SUD study (Methodologies to Integrate Impact Assessment in the Field of Sustainable Development) was funded by European Commission – DG Research (5th Framework Programme) Nikos Kouvaritakis & Nikos Stroblos (ICCS/NTUA)