IMPACT OF THE INTRODUCTION OF ELECTRIC BASED VEHICLES IN SÃO MIGUEL ISLAND

Transcription

1 IMPACT OF THE INTRODUCTION OF ELECTRIC BASED VEHICLES IN SÃO MIGUEL ISLAND P. Baptista 1*, C. Camus 2, C. Silva 1, T. Farias 1 1 IDMEC, Instituto Superior Técnico Universidade Técnica de Lisboa 2 DEEA/Secção de Sistemas de Energia, Instituto Superior de Engenharia de Lisboa *Corresponding author: patricia.baptista@ist.utl.pt Copyright 2009 by P. Baptista. Published and used by MIT ESD and CESUN with permission. Abstract: In a context of volatile fuel prices and rising concerns in terms of energy security of supply and climate change issues, one of the discussed technological alternatives for the transportation sector are electric based vehicles. They should be particularly suited for regions with a high renewable energy potential, such as São Miguel Island in Azores, Portugal. In this paper, the potential for the introduction of electric based vehicles in São Miguel in terms of energy consumption and CO 2 emissions was studied. That allowed quantifying the additional electricity needs and the impacts on the grid management, according to different rates of vehicles replacement. Keywords: Integrated energy analysis, life cycle analysis, alternative vehicle technologies 1. Introduction The urge for energy security of supply, air quality improvement in urban areas and CO 2 emissions reduction are pressing decision makers/manufacturers to act on the road transportation sector, introducing more efficient vehicles on the market and spanning the energy sources. For that reason, it is expected in the near future that the market share of hybrid vehicles will raise and that, in the long term future, the market share of hybrid plug-in vehicles will be significant. Accordingly, this paper intends to exploit the benefits of introducing electricity based vehicles (plug-in vehicles, both pure electric and hybrid) in the light-duty/bus market for the island of São Miguel in Azores, Portugal. These alternatives are particularly suitable for regions with a high renewable energy potential, such as the island of São Miguel. In order to evaluate the environmental impacts of these options, a life-cycle approach was considered for the São Miguel s fleet, covering the materials cradle-to-grave life-cycle (including vehicle assembling, maintenance, dismantling and recycling) and fuel life-cycle: Well-To-Tank (WTT) and Tank-To-Wheel (TTW). The materials life-cycle energy consumption and CO 2 emissions are spread along the vehicle expected lifetime (in this case around 12 years). WTT accounts for energy consumption and CO 2 emissions from primary energy resource extraction through the delivery of the fuel/energy source to the vehicle tank, while TTW accounts for the 1

2 emissions and fuel consumption resulting from moving the vehicle through its drive cycle [1]. Additionally, in order to assess the impact of the penetration of the alternative technologies in the São Miguel s fleet, three distinct scenarios with increasing substitution ratios were evaluated, in this life-cycle analysis framework [2]. As a result, the main goal of this study was to identify the potential for the introduction of electric based vehicles in São Miguel, emphasizing the impacts on energy consumption and CO 2 emissions. That allowed quantifying the additional electricity needs and the impacts on the grid management, according to different rates of vehicles replacement [3]. 2. São Mguel Fleet Characterization After describing the daily/annual vehicle usage in São Miguel and characterizing the existing fleet, the resulting energy consumption patterns and emissions (of CO 2 and local pollutants) were determined (using simulation tools such as ADVISOR [4], SIEMP [5], DEMB [5] and Copert 4 [6]). The 2006 São Miguel road transportation sector by type of vehicle is presented in Table 1. Table 1. São Miguel road transportation sector characterization for Vehicles # Energy Consumption (TJ) Gasoline Diesel LDV Urban bus HDV W Total For validating the results, the statistical data from fuel sales [7] were used for comparison. To estimate local pollutants emissions by different vehicles, the average Portuguese emissions factors (in g.km -1 ) were used [8]. The next step was to determine which were the most adequate technologies for introduction in São Miguel and then use vehicle micro-simulation tools (such as EB-RVS [5] and ADVISOR) to estimate its energy consumption and emissions (if existing). An important aspect is that all vehicle technology configurations considered had to maintain similar performance than existing conventional technologies (e.g. power/weight, acceleration...). Two full electric vehicles and one plug-in hybrid were considered for the light-duty fleet (LDV). For the urban buses, the only market available technology was considered (mini-bus Gulliver [9]). For the two-wheelers (2W), a market available model was adopted. It is worth mentioning that these vehicles range (in km) is very dependent on the tested driving cycle and topography. The values presented in Table 3 were obtained using experimentally measured Portuguese journeys. 2

3 Table 3. Electric and plug-in hybrid technologies characterization. Electric vehicles Weight Battery Power Range (kg) (kwh) (kw) (km) Full electric vehicle Full electric vehicle Plug-in series hybrid Gulliver (EV) W (EV) The following step was to establish different scenarios of vehicle replacement (according to Table 4 and Table 5). It is worth noting that, in this preliminary study, it was only considered that a certain part of the fleet would be replaced by electricity based technologies, not a continuous vehicle analysis along time. Table 4. Established scenarios for electric vehicles replacement in São Miguel Scenarios Bus 2W LDV Fleet replaced (%) Basecase 0% 0% 0% - Scenario 1 25% 0% 0% 0.2% Scenario 2 50% 50% 30% 31.9% Scenario 3 100% 100% 50% 55.5% Table 5. Number of vehicles replaced in the considered scenarios. Vehicles Scenario 1 Scenario 2 Scenario 3 LDV Urban bus W These scenarios correspond to an increasing replacement of conventional technologies by electric and plug-in hybrid vehicles. In the first scenario only a percentage of urban buses and taxis are replaced, while in scenarios 2 and 3 the electricity based two-wheelers and light-duty vehicles are also introduced. After having the base analysis for the reference situation, in this case for São Miguel in 2006, future scenarios were considered. More specifically, the years of 2013 and 2020 were studied. To consider scenarios for the future, the evolution of the fleet along time had to be considered. No growth curves are available for São Miguel, nor the fleet size along time. Therefore, it was considered that the fleet of São Miguel would follow a similar behavior of the observed for Portugal. The 2006 fleet corresponds to a vehicle density of 352 vehicles per 1000 inhabitants, which was observed in Portugal in As a result, in 2013 and 2020 the resulting LDV fleet is a result of the combination of the Portuguese vehicle density evolution along time and of the predicted population evolution for these years. Additionally, with future scenarios different 3

4 electricity generation mixes had to be accounted for. The 2006 electricity generation mix corresponds to the real mix observed in São Miguel for that year. For 2013 the considered mix (Table 7) is the expected EDA mix after installing additional geothermal (13 MW) and wind power (9 MW) generation. The same happens to 2020, with additional geothermal and wind power installed. Table 7. Electricity generation mix for São Miguel in 2006, 2013 and Energy type Fueloil 53% 43% 31% Geothermal 42% 51% 63% Wind power 0% 3% 4% Hidro power 5% 3% 2% 3. Fleet Impacts 3. A. Energy consumption and CO 2 emissions impacts The first analysis was done considering only the vehicle fuel use tank-to-wheel (TTW) stage in São Miguel for the different scenarios (Figure 1). As expected increasing electric vehicle penetration leads to lower energy consumption (in the TTW stage). For scenarios 1, 2 and 3 reductions of 1.3, 26 and 44% were observed. In terms of lowering CO 2 emissions, these scenarios lead to 1.8, 31 and 52% reductions for scenarios 1, 2 and 3 respectively. *"!!#$!( )"%!#$!( )"!!#$!( '"%!#$!( '"!!#$!( %"!!#$!& 6758, #:$675#9./;23/3;< =41>.1;341,95?3.-.9$8, ;@.25A3.-.95B7?:C!"!!#$!! +,-./,-. 0/.1,2345' 0/.1,2345) 0/.1,2345* Figure 1. TTW energy consumption for the different vehicle technologies. As explained earlier the well-to-tank stage and the materials cradle-to-grave must also be included. The fact that electricity based vehicles are introduced shifts these vehicle s energy consumption from the traditional fossil liquid fuels to electricity. Therefore, electricity needs rise and emissions due to these vehicles are now allocated to its electricity production. Consequently, the WTT energy consumption and emissions increase. As for the materials cradle-to-grave, producing electric and plug-in hybrids is a more energy intensive process. Therefore, the materials contribution to the full life cycle slightly increases with higher electric vehicles penetration. 4

5 After combining the 3 stages, a full life cycle assessment is obtained (Figure 2). In terms of energy, scenarios 1, 2 and 3 lower the global energy consumption in 1, 9 and 16%. For CO 2 emissions, the reduction corresponds to 1, 16 and 27%. *"!#$!( )"%#$!( )"!#$!( '"%#$!( '"!#$!( %"!#$!&!"!#$!! "!!!!! &%!!!! &$!!!! &#!!!! 667 &"!!!! &!!!!! 766 %!!!! $!!!! 8,9.23,:- #!!!! "!!!!! +,-./,-. 0/.1,2345' 0/.1,2345) 0/.1,2345* '()*+()*,+*-(./01&,+*-(./01",+*-(./012 Figure 2. Energy consumption and CO 2 emissions LCA for the different scenarios (6*./(7) Regarding local pollutants, increasing the penetration of electric vehicles also leads to a TTW emissions reduction (up to 45% for CO, 49% for NO x, 44% for HC and 52% for PM), according to Table 8. Even considering the full life cycle of CO, NO x and HC, their emissions lower by 37, 11 and 28%. Only particulate matter presents a higher life cycle result due to the high increase in the WTT stage. Table 8. Local air pollutants emissions for São Miguel according to the different scenarios comparing to the basecase situation. Basecase deviation (%) Pollutants LCA stage Secenario 1 Secenario 2 Secenario 3 CO NO x HC PM Materials 0% 20% 32% WTT 0% -18% -30% TTW 0% -28% -45% Total 0% -23% -37% Materials 0% 33% 50% WTT 0% 91% 145% TTW -1% -30% -49% Total 0% -6% -11% Materials 0% 9% 8% WTT 0% -24% -41% TTW 0% -28% -44% Total 0% -16% -28% Materials 0% 27% 41% WTT 1% 144% 232% TTW -1% -31% -52% Total 0% 8% 10% For the 2006, 2013 and 2020 comparison in terms of the island energy security of supply and emissions issues only the WTT and TTW were considered (Figure 3), since most of the materials cradle to grave energy consumption and emissions do not occur within the island boundaries. Assuming a similar fleet growth behavior than the global Portuguese one, the existing fleet will 5

6 continue to grow. For the fleets energy consumption that would imply a 41 and 63% increase in the TTW stage for 2013 and 2020 respectively comparing with Consequently, the implementation of either one of the scenarios would lead to lowering energy consumption (up to 16%) and CO 2 emissions (up to 27%). %"!#$!( )"%#$!( )"!#$!( ("%#$!( ("!#$!( &"%#$!( &"!#$!( '"%#$!( '"!#$!( %"!#$!&!"!#$!!!"! &!!* &!'( &!&! +,-./,-. 0/.1,2345' 0/.1,2345& 0/.1,2345( )"%#$!% )"!#$!% ("%#$!% ("!#$!% '"%#$!% '"!#$!% %"!#$!&!"!#$!! (!!* (!') (!(! Figure 3. Energy consumption and CO 2 emissions WTW analysis for the different scenarios in the studied 3 years. 3. B. Energy and technology prices The energy price evolution in Portugal according to DGEG (Portuguese Energy Agency) [10] is as presented in Figure 4.!"! +,-./,-. 0/.1,2345' 0/.1,2345( 0/.1,2345) 0,05 0,045 0,04 0,035 0,03 0,025 0,02 0,015 0,01 0,005 0 Gasoline 95 RON Diesel Electricity Year Figure 4. Energy price evolution (including all taxes, electricity is average domestic). The expenses per travelled km (discarding maintenance costs) can be estimated as a function of electricity price, according to Figure 5. Assuming that the electricity infrastructure is ready, it can be observed that the current electricity price in Azores of around 0.04 /MJ is quite attractive comparatively to gasoline use, both for the PHEV and the EV technological solutions. 6

7 ICEV Diesel ICEV Gasoline EV (Eletricity) HEV Gasoline PHEV Gasoline ,02 0,04 0,06 0,08 0,1 Electricity!/MJ Figure 5. Refueling cost evolution for electricity based vehicles as a function of electricity price. For the light duty solutions presented, a main factor regarding user preference for vehicle technology, besides fuel availability and price, is the acquisition cost. This reflects directly the manufacturing costs. Table 9 shows cost differences having ICEV gasoline as the base case, assuming the cost of for the ICEV gasoline vehicle (base case) [11], a cost of 30 /kw for engine plus transmission, a cost of 300 for gasoline exhaust aftertreatment and 700 for diesel with particle filter, a cost of 462 /kwh [12] for the NiMH battery, a cost of 600 /kwh for the Li-ion battery, a cost of 27 /kw for the electric motor plus controller and 1500 for diesel direct injection system. To have a sense of how many kilometers vehicles must be driven to pay off for initial purchase cost, the 2007 average energy was assumed. It is important to note that it is possible that alternative vehicle technologies costs will be comparable with internal combustion engine cost if sufficient market penetration is attained. Table 9. Cost differences in comparison with ICEV gasoline. Vehicle technology % difference km breaktrough ICEV Gasoline - - ICEV Diesel HEV PHEV EV These values mean that hybrid plug-in and pure electric vehicles pay off in terms of cost only if long distances are driven (higher than km). This fact is important when calculating eventual tax incentives to purchase these kinds of technologies, having in mine that the final consumer is extremely sensitive to the km for breakthrough. In terms of the other solutions presented, the electric mini-bus and the electric 2W, similar conclusions are applicable. These solutions have higher acquisition costs that can be mitigated with long utilization and if technology prices lower. 7

8 4. Impacts on Load Profiles 4. A. Electric Power consumption and power plants technologies One of the main features of power consumption is the difference in demand along the day hours. In São Miguel, electricity consumption has this typical evolution along the day with a valley during the night representing 50-60% of the peak consumption. To provide for this demand, electricity production in São Miguel is around 50% renewable composed with geothermal power plants as base load a few hydro run-of-river and regulation is done mainly with fuel. The typical load profiles depending on the time of year are depicted in Figure 6. 8

9 Spring Load Profile Winter Load Profile Figure 6. Typical load profile in S. Miguel for Spring and Winter seasons As it can be seen in Figure 6, in São Miguel Island, Azores, the geothermal energy production (renewable energy without CO 2 emissions that should be used as base load due to its impossibility of production variation) penetration is limited by the valley consumption. If base load electricity generation is much higher than actual demand, excess electricity will be wasted unless it is coupled to a storage system. This difference in power demand along the day has also great financial consequences with the need of having several power plants that are useless and an underutilized network during the night. On the other side, if there is no increase in renewable production, predictable demand increase should be provided with more fuel. This is an example, where the use of EVs, charged during off-peak hours, allow the development of a production technology from a renewable, endogenous source, with no CO 2 emissions, against the systematic fuel imports (for vehicles and for electricity production). 4. B. Electric Vehicles leveling the power consumption diagram As stated in last section, the fact that the valley demand is about half the value of peak demand, gives the opportunity for electric vehicles contribution for levelling the power consumption diagram. The extra demand for charging the vehicles at each hour of the day was computed using equation 1. P i ( 1+ p ) EVavg loss pcharge = N (1) h charge Where E Vavg is the daily average energy needed for charging a vehicle, p loss is the percentage of energy lost in the transmission lines, p charge and h charge are respectively the percentage that is charged in each period (valley and off-valley) and the length of the considered period and N the number of vehicles. Considering that 85% of the vehicles will uniformly recharge at night, the introduction of these vehicles will create and additional electricity demand that, in the context of today s electricity generation mix, will be fulfilled by the running fuel oil power plants (Figure 7). 9

10 Spring Load Profile Winter Load Profile Figure 7. Load profile in S. Miguel for Spring and Winter seasons with scenario 2 of EVs extra demand As expected, the percentage of renewable energy decreases and in terms of marginal emissions, they are simply being transferred from the tail pipes to the fuel power plants chimneys. Considering an emissions rate of 830 gco 2 /kwh for the fuel power plants, the additional CO 2 emissions to daily charge EVs should be in average 75 tons of CO 2. The TTW daily emissions avoided if these same vehicles were conventional was 142 tons. However, if additional renewable electricity installed capacity is deployed, then the vehicles will also be running on renewable electricity. 4. C. Electric Vehicles Impact for 2013 Scenarios S. Miguel has 24 MW of existing geothermal capacity and the government wants to expand this capacity to meet future demand [13]. Expanding existing capacity would mean that geothermal production will meet or exceed 37 MW. Current base load electricity demand is nowadays less than 40 MW (Figure 6). On the other side investments in wind energy are also expected so that for 2013 a 9 MW in wind power installed is expected in the Island [14]. For the geothermal energy production operating as base load a load factor of 90% was assumed plus a small Gaussian randomness of 1%. For the wind power, an average 35% load factor was considered more strongly during the night. For this situation only scenarios 2 and 3 have a considerable impact in increasing electricity demand. In terms of emissions from electricity generation to charge EVs we can compute either marginal or average emissions. If no increase in renewable energy is considered the charging of vehicles should be done with extra fuel and the marginal emissions should be computed. If EVs, charged during off-valley hours, enable the increase of renewable energy capacity in the Island, then the average emissions rate should be computed for the EVs charge. 10

11 Spring Load Profile Winter Load Profile Figure 8. Expected load profile in S. Miguel for year 2013, 4% demand increase for Spring and Winter seasons with scenario 2 of EVs extra demand. Computing marginal and average emissions, the daily emissions for electricity and EVs charging are distributed as presented in Figure 9 and 10 respectively. Spring (tons of CO 2 ) Winter (tons of CO 2 ) Figure 9. Expected daily marginal emissions from electricity generation for EVs and the normal load in S. Miguel for year 2013, for Spring and Winter seasons and scenario 2. Spring (tons of CO 2 ) Winter (tons of CO 2 ) Figure 10. Expected daily average emissions from electricity generation for EVs and the normal load in S. Miguel for year 2013, for Spring and Winter seasons and scenario 2. 11

12 Average emissions distribute total emissions equally between EVs and the other load so in the accounts of CO 2 emissions for vehicles are less penalized. 4. D. Electric Power Average Unit Costs Electricity production in São Miguel is expected to be around 65% renewable in 2013 composed mainly with geothermal power plants (37 MW capacity installed), wind (9 MW, capacity), hydro (5 MW capacity) and fuel (98 MW) [15]. Geothermal power plants are characterized by high capital investments for exploration, drilling wells and plant installation but low cost for operation and maintenance. EDA investments plan indicates an increase of 3 MW capacity in 2010, 3.5 M of investment to an average annual production of 20 GWh and 10 MW in 2013, 30 M of investment that includes also a new substation to connect to the 60 kv grid leading to an annual increase in 83 GWh [13]. Considered that these plant lifetimes range typically from years, the annualized average cost of these investments are (considering that the annual maintenance and operation costs 2% of total investment and 10% rate of return) between cents/kwh for the first investment and cents/kwh for the other. Wind power investments are expected to be 14.5 M in 9 MW wind power capacity and an annual expected production of 22.5 GWh. For a 25 years life time, annual maintenance and operation costs 2% of investment and 10% rate of return, the annualized average cost is 8.4 cents/kwh. Table 10. Average unit cost of expected new electricity production in São Miguel. Power plant Capacit Annual Investme Average unit y Energy nt cost Geothermal 3MW 20GWh 3.5M 2.1c /kwh (2010) Geothermal 10MW 83GWh 30M 4.4c /kwh (2013) Wind (2011) 9MW 22,5GWh 14.5M 8.4c /kwh Hydro plants are more than 18 years old except the most powerful (1658 kw) that began to produce in They are all small run of river hydro plants [15]. These projects have a 30/35 years life time and the average unit costs are around 8 cents/kwh [16]. The average unit cost of the thermal power plants is highly dependent on the fuel costs, whose values in 2007 were in average 0.36 /kg for fuel oil [15]. Considering 36% of plant performance (4 kwh/kg), fuel costs were around 9 cents/kwh. Apart from the emissions problem, that obliges the thermal power plants to buy emissions allowances whenever the annual allowances allocated by the National Allowances Plan are all used [17], fuel power plants have the highest variable unit costs, another reason to adopt strategies to reduce its use, increasing the renewable sources that prove to be more efficient as soon as the storage and regulation issues could be solved. 12

13 5. Conclusions The penetration up to 55% of electric based vehicles allows an energy consumption, CO 2 and pollutants emissions reduction, even in a life-cycle framework (up to 15, 20 and 37% respectively). As was presented, the impacts of scenario 1, which corresponds to solely a low public transportation fleet replacement, is clearly insufficient. Only if the light-duty fleet is considered will the consequences in terms of energy and emissions be significant. Another important issue is that, based on today s electricity generation mix, the electricity based vehicles will mostly be recharged by fossil fuels, since fuel oil will be used to fulfill demand needs. This leads to lower reductions in terms of WTT energy consumption and emissions than if the electricity was produced from renewable sources. As expected, scenarios of higher renewable integration in electricity generation are favorable for the introduction of these types of vehicles. Due to the characteristics of the renewable resources in São Miguel, and the typical load profiles there are synergies between EVs and electricity production that will, if fitted together, reduce overall emissions in the Island and fuel imports. The mass penetration of EVs, with the guarantee that they charge mainly during off peak hours, has the advantage of capturing the geothermal and wind energy produced in excess, in those hours, and avoid the investment in water pump power plants that should be needed to accommodate the excess renewable production during valley hours. Acknowledgements Thanks are due to the MIT Portugal Program and Fundação para a Ciência e Tecnologia for the PhD financial support (SFRH/BD/35191/2007) POS_Conhecimento. The authors would like to acknowledge FCT-Fundação para a Ciência e Tecnologia through the national project MMSAFU-Microssimulation Model to Simulate Alternative Fuel Usage (POCI/ENR/57450/2004), and national project Power demand estimation and power system impacts resulting of fleet penetration of electric/plug-in vehicles (MIT-Pt/SES-GI/0008/2008). References 1. Moon, P., Burnham, A. and Wang, M., Vehicle-Cycle Energy and Emission Effects of Conventional and Advanced Vehicles. SAE TECHNICAL PAPER SERIES, Baptista, P., Tomás, M., Silva, C., HYBRID PLUG-IN FUEL CELL VEHICLES MARKET PENETRATION SCENARIOS, in HYPOTHESIS VIII: HYdrogen POwer - THeoretical and Engineering Solutions - International Symposium. 2009: Lisbon - Portugal. 3. C. Camus, T. L. Farias, J. Esteves, Impact of Plug-in Hybrid Electric Vehicles in the Portuguese Electric Utility System, in 2nd International Conference on Power Engineering, Energy and Electrical Drives,. 2009: Lisbon, Portugal. 4. Wipke, K., Cuddy, M., Burch, S., ADVISOR 2.1: A User Friendly Advanced Powertrain Simulation Using a Combined Backward/Forward Approach. IEEE Transactions on Vehicular Tecnology, (48): p Silva CM, Farias T., Analysis and simulation of "low-cost" strategies to reduce fuel consumption and emissions in conventional gasoline light-duty vehicles. Energy Conversion and Management, (2): p

14 6. Dimitrios Gkatzoflias, C.K., Leonidas Ntziachristos and Zissis Samaras, COPERT 4, computer programme to calculate emissions from road transport. 2007, ETC/AEM. 7. DGEG, Balanço energético J. A. Peças Lopes, F.J. Soares, P. M. Rocha Almeida, P. C. Baptista, C. M. Silva and T. L. Farias, Quantification of Technical Impacts and Environmental Benefits of Electric Vehicles Integration on Electricity Grids, in 8th International Symposium on Advanced Electromechanical Motion Systems. 2009: Lille - France. 9. Stüssi R., S.I., Esteves J., Resende M., Developing monitoring and maintenance agreement and tools for 5 Portuguese cities, in APVE report DGEG, Estatísticas-Preços de Energia Edwards, R., Well-to-Wheels analysis of future automotive fuels and powertrains in the European context M.A Kromer, J.B. Heywood, Electric Powertrains: Opportunities and challenges in the U.S. light-duty fleet, in MIT report number LFEE RP Informa, E., O presente e o futuro das energias renováveis Botelho, F. 2008: Public presentation at ERSE. 15. [cited last accessed 16/04/2009]; Available from: [cited last accessed 17/04/2009]; Available from: Plano Nacional de Atribuição de Licenças de Emissão de CO 2 (PNALE) Versão para Consulta Pública, Junho de