Migrating from Oil- to Electricity-Powered Vehicles: Modeling Germany's Transition to the EV until 2040 in System Dynamics


Bachelor Thesis, 2009
122 Pages, Grade: 1.0

Excerpt

Table of content

List of figures

List of tables

List of abbreviations

Executive Summary

1. Introduction
1.1. Types and advantages of electrified vehicles
1.2. Historic overview
1.3. Barriers to entry
1.4. Modeling alternative fuel vehicle introduction dynamics in System Dynamics

2. Model description
2.1. The consumer model
2.2. The industry model
2.3. The infrastructure model

3. Model analysis
3.1. Scenario analysis
3.2. Sensitivity analysis

4. Policy measures in support of electric vehicles
4.1. Tax credits for EV purchases
4.2. Subsidies for recharging stations
4.3. Tolls for city access for high pollution vehicles
4.4. Qualitative assessment of other measures
4.5. Evaluative summary of policy measures

5. Impact of EV adoption on CO2-emissions

6. Conclusion

References

Appendix

List of figures

Figure 1 – Number of published scientific papers on electric vehicles

Figure 2 – Number of published scientific papers on environmental issues

Figure 3 – Range of forecasts for EV and hybrid car market shares of new sales in Germany (2020)

Figure 4 – Range of forecasts for EV and hybrid car market shares of total fleet in Germany (2020)

Figure 5 – Forecast of global energy demand and forecast world oil supply

Figure 6 – Value at Risk and opportunity for car makers

Figure 7 – Current energy density by battery type

Figure 8 – Historical battery improvement rates at Matsushita

Figure 9 – Example model in System Dynamics

Figure 10 – General model overview

Figure 11 – Consumer model overview

Figure 12 – Vehicle age cohort structure

Figure 13 – Conceptual structure of a Bass innovation diffusion model

Figure 14 – Industry model overview

Figure 15 – METI innovation roadmap for energy storage

Figure 16 – Infrastructure model overview

Figure 17 – Destinations selected for 100 km

Figure 18 – Model trip from Ebersberg to Augsburg (100 km)

Figure 19 – Actual trip from Ebersberg to Augsburg (100 km)

Figure 20 – Overview trip destinations starting from region Ebersberg

Figure 21 – Daily trip schedule model overview

Figure 22 – Fuel/ recharging station development model

Figure 23 – Actual and model ICE fuel station density Germany (right side)

Figure 24 – Scenario analysis: Share of purchase

Figure 25 – Scenario analysis: Share of fleet

Figure 26 – Scenario analysis: Price of utility

Figure 27 – Scenario analysis: Familiarity of ICE drivers with the EV platform

Figure 28 – Scenario analysis: Total vehicle cost

Figure 29 – Scenario analysis: Module costs EV

Figure 30 – Scenario analysis: Share of possible trips for the EV platform

Figure 31 – Scenario analysis: EV recharging stations

Figure 32 – EV share of fleet per region in 2015

Figure 33 – EV share of fleet per region in 2020

Figure 34 – EV share of fleet per region in 2030

Figure 35 – EV share of fleet per region in 2040

Figure 36 – Familiarity (1/2) (advertising scenarios)

Figure 37 – Familiarity (2/2) (advertising scenarios)

Figure 38 – Fractional familiarity decay (advertising scenarios)

Figure 39 – Share of purchase (advertising scenarios)

Figure 40 – Total vehicle costs (fuel price scenarios)

Figure 41 – Total vehicle costs (electricity price scenario)

Figure 42 – Share of purchase (fuel and electricity cost scenarios)

Figure 43 – Travel utility (trip weight scenario)

Figure 44 – Share of purchase (trip weight scenario)

Figure 45 – Total vehicle costs (battery improvement scenario)

Figure 46 – Travel utility (battery improvement scenario)

Figure 47 – Share of purchase (battery improvement scenario)

Figure 48 – Total price (cost degression parameter scenario)

Figure 49 – Share of purchase (cost degression parameter scenario)

Figure 50 – Total vehicle costs (tax credit scenarios)

Figure 51 – Total sales (tax credit scenarios)

Figure 52 – Cumulative sales (tax credit scenarios)

Figure 53 – Familiarity (tax credit scenarios)

Figure 54 – Share of purchase (additional tax credit scenarios)

Figure 55 – Number of fuel stations (station subsidy scenarios)

Figure 56 – Share of purchase for different station subsidy scenarios

Figure 57 – Relation between fuel station density and average distance to the next station

Figure 58 – EV fuel station density in 2020

Figure 59 – Share of purchase (city toll scenarios)

Figure 60 – Share of purchase (city toll scenarios)

Figure 61 – Annual toll revenues (city toll scenarios)

Figure 62 – EV share of fleet with 10 cent city toll compared to base case without toll

List of tables

Table 1 – Types of EVs

Table 2 – Status of current hybrid and EV technology

Table 3 – Current status and future trends for EV batteries

Table 4 – Relative frequency of trip length by means of transportation

Table 5 – Relative frequency of trip length by type of region

Table 6 – Model trip types by length and number of trips per year

Table 7 – Kilometers driven in each hour for each type of driver (variable ldr)

Table 8 – Assumptions for ICE fuel station calibration test

Table 9 – Infrastructure scenario assumptions

Table 10 – Overview policy measures

Table 11 – Overview policy effectiveness

Table 12 – CO2 reduction potential by energy-mix and model scenario

List of abbreviations

Abbildung in dieser Leseprobe nicht enthalten

Executive Summary

The electric vehicle (EV) appears poised to make a successful entrance to the personal vehicle mass market as a viable alternative to the traditional internal combustion engine vehicles (ICE): Recent advances in battery technology indicate decreasing production costs and increasing energy densities to levels soon acceptable by broad consumer segments. Moreover, excluding the generation of the electricity, EVs emit no greenhouse gases and could contribute to meeting the strict CO2 emission limits necessary to dampen the effect of global warming. Several countries around the world have therefore initiated measures like consumer tax credits, research grants or recharging station subsidies to support the introduction of the EV. Finally, the success alternative vehicles like the Toyota Prius Hybrid proves a shift in consumer interest towards cleaner cars with lower operating costs.

Nonetheless, the EV will first need to overcome significant barriers that might delay or even prevent a successful mass market adoption.

- In spite of recent advances and high expectations for the near future, the current level of battery technology poses one of the biggest problems. High battery production costs of more than €10,000 render EVs more expensive than their ICE counterparts. Low energy density of batteries limits the range of one battery charge to around 100 km. Moreover uncertainty regarding battery lifetime and safety still exist. These problems can only be overcome if high demand for batteries allows scale effect in mass production.
- Limited EV range might be extended by stations positioned along the way that can quickly recharge the battery or swap a depleted battery for a fully charged one, yet only if EVs indeed make their entry to the mass market will demand be sufficient to justify investments in this recharging infrastructure.
- Low consumer familiarity with the EV will hinder a rapid adoption.
While public expectations are very high at the moment, decision makers from all stakeholders involved require realistic forecasts on potential EV market penetration in order to take the right decisions now. Unfortunately, current studies on the market potential of the EV are very unreliable, for example, different studies predict market shares for EVs in Germany for 2020 between 0.5% and 25%. We therefore propose in our thesis “Migrating from Oil- to Electricity-Powered Vehicles: Modeling Germany’s transition to the EV until 2040 in System Dynamics” a comprehensive model for Germany that accounts for the non-linear relationships involved between customers, science and the industry. The model includes consumer irrationality biases, detailed driving behavior on a regional level, endogenous decisions of firms to build recharging stations based on perceived profitability, economies of scale in production, as well as the most recent industry forecasts on battery technology improvement. The model is able to predict EV sales on a regional level and determine the most efficient measures to promote EV adoption with the help of scenario analysis. Some of the insights we derived from the model include:
- Driven by decreasing production costs and increasing vehicle ranges, we predict a 7% market share of new purchases for the EV in 2020, steadily increasing to almost 40% in 2030 – even without concrete policy measures or infrastructure investments.
- Even today, less than 1% of desired vehicle trips are not possible due to battery range limitations of the EV. Furthermore, range limitations will quickly cease to be a limiting factor as battery energy density increases rapidly.
- Infrastructure to recharge or switch battery during a trip will only have a marginal effect on EV adoption speed; station density will remain low, even in the long run.
- Due to Germany’s coal intensive electricity mix, the introduction of the EV will only lead to low annual CO2 emission: 0.3 - 0.4% of total emissions in 2020 and 2.8 - 3.1% in 2030.
- The introduction of tax credits for EV purchases could more than double EV market share until 2020. Important for the effectiveness of financial consumer incentives is the simultaneous increase in public awareness for the EV.
- City tolls for non-zero emission vehicles can boost EV adoption in cities. Neighboring regions also profit from second round effects.
- Subsidies for recharging stations have low overall effectiveness, but can be used to stimulate EV adoption in otherwise lagging rural regions.

Further expansions of the model could increase the robustness of results, as well as allow analysis on a more detailed level. We see potential in three areas: expansion of the costumer behavior model to allow the evaluation of alternative payment schemes such as pay-for-service, empirical validation of assumptions and generation of additional data, explicit modeling of cross-country and commercial traffic.

1. Introduction

The recent years have shown a surge of public attention dedicated to electric mobility in media and academia. Diverse stakeholders like governments, car makers or research institutes have indicated their interest in exploring possibilities to introduce electric cars on a broad scale – as well as other vehicles using alternative fuels like hydrogen or biogas. One indicator for this strong increase in interest is the number of scientific papers published on electric vehicles over the last 50 years (Figure 1). Even if we compare these figures to other topics that have shown rapid growth in interest like the environment (Figure 2), growth in scientific interest in the EV in recent years appears enormous: In just nine years, from 2000 to today, the number of papers on EVs tripled compared to the number of those papers published in the last two decades of the 20th century – even though the technology for electric vehicles (in one form or the other) has been available since the early decades of the 20th century.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1 – Number of published scientific papers on electric vehicles

Abbildung in dieser Leseprobe nicht enthalten

Figure 2 – Number of published scientific papers on environmental issues

In addition to the scientific interest, governments and private investors have already made substantial investments and commitments in order to foster the market entry of alternative fuel vehicles (AFVs). Germany alone has committed €500M until 2011 for research and development of electric vehicles as part of their stimulus package for the German economy (“Konjunkturpaket II”). Furthermore, the German Federal Government promised to support a consortium of industry leaders1 with €60M for the improvement of lithium-batteries in addition of the consortium’s €360M commitment until 2012. Other measures such as tax credits for purchases of AFVs to speed up the introduction of AFVs aim at the consumer: Great Britain currently offers £5,000 (about €6,500) of tax benefits to drivers of AFVs. Denmark, whose entire transportation system is still almost 100% oil-dependent (Hasberg, Mets 2008) made electric vehicles tax-exempt for vehicle registration tax that may reach 180% of the purchase price of a new car. Israel has initiated cooperation with the two companies Better Place and Renault in order to build electric cars and deploy battery exchange stations on a broad scale across the entire country while in turn promising to offer tax incentives for EVs.

These developments have led to very high expectations regarding a quick success of the EV: Duke and Anderson (2008) in their paper about introduction of battery electric vehicle (BEVs) in New Zealand argue that “BEVs [...] are probably the best compromise between, performance, cost, land use and energy efficiency”. The International Energy Agency with their special commission HEVIA (Hybrid and Electric Vehicle Implementing Agreement) states that electric vehicles and hybrids “are likely to garner support because of their broad ability to consistently reduce negative externalities with respect to energy imports, GHGs, and air quality” (Appendix IEA HEVIA, Status Overview of Hybrid and Electric Vehicle technology (2007)). On the other hand, some skeptics try to dampen the euphoria: Hensley, Knupfer and Pinner (2009), for example, state that the advent of rechargeable vehicles is a “a safe bet“, but also highlight their skepticism with regard to a too short time horizon: “How long will investors have to wait for the bet to pay off?”.

This uncertainty is also reflected in the wide range of expert estimates on the number of electric and hybrid electric vehicles on German roads at a certain point in time of the future. We have collected some of their estimates for 2020 both for the market share of new car sales and the share of the entire vehicle fleet in Figure 3 and Figure 4 below.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3 – Range of forecasts for EV and hybrid car market shares of new sales in Germany (2020)

Abbildung in dieser Leseprobe nicht enthalten

Figure 4 – Range of forecasts for EV and hybrid car market shares of total fleet in Germany (2020)

These wide ranges of estimates are due to different assumptions about technology improvements, consumer preferences etc. These differences are compounded by the complex feedback mechanisms between consumer, infrastructure and industry that are probably not properly taken into account in these often simple models. These uncertainties led us to examine the dynamics of electric vehicles in Germany in our research. Our work explores possible scenarios for the German market for the next 30 years and places electric vehicles into a broad socio-economic system that takes into account the decisions and actions of agents, changes thereof and interdependencies between technological changes, market penetration and infrastructure density.

1.1. Types and advantages of electrified vehicles

In a narrow sense, the term electric vehicle exclusively describes vehicles entirely powered by electricity. Used in a broader sense, a wider variety of electrified vehicles can be subsumed under the term “EV” (McKinsey Quarterly 03/2009), some of which (Hybrid EVs) would not fall within the narrow definition. The following table taken from Chan (2002) gives an overview of the different types of electric vehicles:

Abbildung in dieser Leseprobe nicht enthalten

Table 1 – Types of EVs

The term “Battery electric vehicle” or “BEV” describes a vehicle that is powered by an installed battery that serves as energy source to the electric motor. Energy can only be transferred to the vehicle by recharging the battery by connecting it to the electric grid or some other electricity supply or by swapping the battery for a fully charged one. BEVs are currently more expensive than their ICE peers due to high battery costs. Moreover, due to limitations on battery size and energy density, vehicle ranges of a single battery charge are still relatively limited.

The next type of EV is the hybrid electric vehicle (also called plug-in hybrid-electric vehicle (PHEV) or full hybrid (Pehnt, Höpfner, Merten 2007)). They need to be distinguished from so called “Hybrids”, a word often used in the media. These “Hybrids”, such as the Toyota Prius, have two propulsion and two energy storage systems: One regular internal combustion engine connected to a fuel tank as its primary system and an electric engine with a battery as its secondary system. Due to battery capacity limitations, these vehicles can drive just on electricity for very short trips only. We call these vehicles “mild hybrid EV”. True Hybrid EVs on the contrary are more similar to BEVs in that they have their electric motor as their main propulsion system. In contrast to the BEV however, they have two energy storage systems: a battery and a regular fuel tank connected to an energy generator recharges the battery. No propulsion system for regular fuel exists. The primary advantage of the PHEV over the BEV is the extended range and the possibility to refuel at regular gas stations, greatly reducing the risk of running out of fuel during the trip.

Lastly, fuel cell vehicles also rely on electric motors as their propulsion system, but instead of batteries, fuel cell vehicles use hydrogen or methanol based fuel cells as energy storage system. Like the ICE, the fuel cell vehicle is dependent on a suitable infrastructure to refuel. This reliance on a suitable infrastructure is one of the major disadvantages of this type of EV compared to the BEV that can be recharged at home on a regular electricity outlet. Like in the case of batteries, production costs of fuel cells are still high. Even though we focus on BEVs in our work, our results may also be applicable to other EV types as market entry barriers and industry dynamics may be very similar.

Advantages of EVs

The increased interest in electric vehicles is without doubt due to the many advantages of EVs compared to ICEs or even to competing AFV technologies. At Germany’s National Strategic Conference for Electric Mobility on November 25th/26th 2008 five main arguments were brought forward:

1. EVs protect the environment by reducing CO 2 emissions in the transportation sector

Neither atmospheric CO2 as a cause of global warming, nor the anthropogenic increase in CO2 is any longer questioned or even denied (Reckien et al. 2006). The high level of annual CO2 emissions constitutes “a compounding problem” (Fontaine 2008), as yearly emissions of green house gases are so much higher than amounts that dissolve per year so that it may take decades to return to normal levels – even if no further green house gas were emitted as of today. Scientists agree that the maximum increase in temperature tolerable for the delicate balances on Earth is around two degrees Celsius. In order not to exceed this level, emissions of carbon dioxide and its equivalents of other green house gases need to be reduced drastically. Even with concentrations of 450 ppm of CO2 in the atmosphere which is within reach of the current level of nearly 400 ppm, only a 50% chance exists that the temperature increase will not exceed the crucial level of 2 degree.2 A temperature increase above that level would have unpredictable consequences on nature’s equilibria, as certain tipping points might be reached such as the melting of pole caps that will lead to irreversible changes that may further accelerate the increase in temperature.

Economies are still very reliant on the generation of energy from fossil fuels – the biggest contributor to global CO2-emissions: According to estimates, 95% of German transportation is still dependent on oil and its derivatives as a source of fuel (Pehnt, Höpfner, Merten 2007) and thus contributes significantly to Germany’s total CO2 emissions. In addition to the existing vehicle stock in developed countries, the rising numbers of car owners in developing countries poses another problem. In India and China only 1.2% and 2.0% of the population own a car (Fontaine 2008). Compared to Germany where the ratio of car owners to the entire population is about 50%3, we can thus expect a massive surge in car sales in these countries as an increasing share of the population will join the middle class due to strong economic growth.

In order to make the transport sector more environmentally friendly, a transition towards electricity powered vehicles seems a viable option. Electricity is a multi-sourced form of energy (Grove 2009), as it can be generated from a variety of fuels and by a multitude of techniques. For an assessment of GHG emitted by EVs it is therefore necessary to determine the sources from which the electricity has been generated (Katz 2009). A comparison between a regular ICE running on gasoline and electric vehicle using the current electricity mix in the United Stated results in superior performance for the EV; however that is easily reversed if the ICE runs on E85 (Samaras, Meisterling 2008).4 Conversely, if the EV solely used electricity generated from wind or solar power where no green house gases are emitted during the production, the EV would be CO2 neutral, something impossible for any regular combustion engine.

2. EVs reduce local emissions of CO 2 , pollution, noise and energy consumption

Even if the available electricity mix would not enable the EV to reduce overall CO2 emissions, other advantages may exist from an environmental point of view.

Firstly, even if the sources used for electricity generation lead to a similar carbon footprint similar to that of an ICE, emissions from a power plant would still much easier be captured on-site, for example with better filters. Furthermore, many people expect a major breakthrough in carbon capture and storage (CCS) technology that theoretically allows separating carbon dioxide from the exhaust air, and then compressing and storing it in a suitable storage (Wuppertal Institut 2009)5. Also, electric utility companies are already part of the cap-and-trade system which would certainly facilitate an integration of the transportation sector into this system (Fontaine 2008).

Secondly, urban air quality has been decreasing ever more quickly in high-density population areas, mainly due to high traffic levels. High levels of air pollution from evaporative substances like nitrogen oxides, carbon monoxides, and air toxics could be drastically reduced by introduction of EVs which emit none of these substances (Socolow, Thomas 1997). In fact, this quality of an EV – harming other people less – was the one unique selling proposition when experts thought first on how to market an EV (Gärling, Thøgersen 2001).

Thirdly, lacking an internal combustion engine, EVs run very quietly, a feature that made it popular for inside uses (forklifts in factories) and peaceful outdoor activities (golf cart) (Larminie, Lowry 2003). At higher speed though, the noise reduction is smaller as car noise is mainly due to tire friction which is similar for EVs and gasoline cars.

Lastly, EVs offer increased fuel efficiency compared to ICEs, thereby saving energy (Brauner 2009). Whereas a normal gasoline/ diesel car achieves only a degree of efficiency of 15% and 20%, the EV typically shows levels around 50%. This is consistent with the recent estimate of Werber, Fischer and Schwartz (2009) who claims that efficiency levels are about 2.6 times greater for the EV.

3. EVs strengthen energy independence and thus ensure Germany’s energy supply

Germany and other Western countries are still heavily dependent on oil as a fuel for their transportation sectors, although consumption decreased during the global economic crisis in what is the biggest single-year drop in more than 25 years (International Energy Agency Oil Market Report May 2009). Even though forecasts of the availability of oil show huge discrepancies, many experts agree that increasing energy independence by reducing exposure to oil and its derivates is crucial.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5 – Forecast of global energy demand and forecast world oil supply6

One advantage of higher independence from oil is the ability to better cope with sudden (unintended) supply disruptions in the near future and supply stops in the longer term (Pehnt, Höpfner, Merten 2007). Another advantage of diversifying the energy portfolio is a decreased likelihood to be pulled into energy related conflicts and even wars over resources, as “oil-producing countries flex their muscles more and more openly“ (Grove 2009). Reducing a country’s dependence on oil by shifting the transportation sector from oil to electricity should therefore be a top priority for every net oil importing country.

4. The integration of EVs into the electric grid may be beneficial

The idea of integrating EV batteries into the grid has already been mentioned in 1997: It was even argued that beneficial effects from the EVs to utilities might be so large that these companies should themselves subsidize the EV’s market entry (Kempton, Letendre 1997). The integration of EVs into the grid renders the system more reliable and thus less costly. Moreover, they could “facilitate large-scale integration of intermittent-renewable energy resources” (Kempton, Letendre 1997). The idea behind this is that many renewable energy sources are very volatile and unpredictable in their power output, like wind or solar power for instance, which currently hinders them to contribute decisively to the electricity supply. Sometimes they produce more electricity than needed – energy that cannot be captured without adequate storage systems; sometimes they produce no energy at all – even though demand for energy exists. The problem is particularly severe during the low demand at night (Fontaine 2008). This is where EVs with their vehicle-to-grid (V2G) services can add value by replacing the job of expensive “quick-response, high-value electric services” (Kempton, Tomi 2005). These services normally balance constant fluctuations in load and adapt to unexpected equipment failures by supplying electricity when demand is high, recharging during low demand periods. As personal vehicles are idle on average for 95% of time (Fontaine 2008, Kempton, Tomi 2005) it is probably only economic to look for a vehicle’s second function.

Some fear that the large scale introduction of EVs might lead to an additional electricity demand so high that current plant capacities would not suffice. This fear is nonetheless ungrounded, as estimates predict that one million EVs in Germany would only consume 2 TWh per year of incremental energy which equals just 1/300 of the entire annual German electricity used (Pehnt, Höpfner, Merten 2007). Interestingly, even small fleets of EVs can provide valuable support for the grid and should be integrated from the beginning (Fontaine 2008).

5. The EV can help Germany advance its knowledge in an attractive technology field

As paradigm shifts occur in one of the oldest and most established industries around the world new chances and risks appear. A very interesting and detailed analysis was conducted in 2008 by the Carbon Trust in collaboration with the two consultancies McKinsey & Company and Oxera. The Carbon Trust is an independent company set up in 2001 by the UK government aiming to help find (business) solutions towards a low-carbon future.7 In their 2008 report they assess six industries with regard to future chances and risks for four industry types: Demand up, demand down, volatility and transformation. They classify the automotive industry as a transformational industry, as they see both winners and losers on the side of the manufacturers. They attribute the following risks and opportunities to different scenarios:

Abbildung in dieser Leseprobe nicht enthalten

Figure 6 – Value at Risk and opportunity for car makers

According to their estimates there is overall more to lose than to gain from the transformation in the automotive industry. The far greatest opportunity exists under the Technology scenario which assumes a major breakthrough in the technology of EVs as automakers would likely be able to protect initial advantage like cost advantages or performance margins and use it to grow market share. Especially for the German car manufacturers is might be a viable opportunity to step out of the fiercely competitive environment of ICE manufacturing and further increase their exports (German Federal Government at the National Strategy Conference for Electric Mobility 2008).

These all seem to be – and actually are – powerful arguments in favor of electric vehicles which leads to an important question: Why has the EV not successfully entered the market yet? We will answer that question by first giving a brief historic overview and then describing the barriers that prevented the EV so far from dispersing on a broad scale.

1.2. Historic overview

The first electric vehicle prototypes were developed as early as in 1830. Nonetheless, production for consumers started only once batteries were made rechargeable near the end of the 19th century and coincided with the development of ICEs and steam cars (Larminie, Lowry 2003). This corresponds to the first of five phases Cowan and Hultén (1996) define for the evolution of automobiles, called the “formative years of the automobile industry”. It is important to note that in the beginning the different technologies of electric, steam and gasoline cars were approximately equal – each type had some strengths but also its own deficiencies: gasoline cars were noisy and smelly, had high water consumption, low top speed, limited range and were difficult to start; steam cars bore high explosion risk, had even higher water consumption and needed long warming time before start; electric cars had also only limited range and could not easily climb steep hills (Larminie, Lowry 2003). Others even emphasize the initial superiority of the EV over its two peers by arguing that the EV had less numerous and less severe deficiencies. The EV even won the world speed record in 1899 with “la jamais contente” achieving a top speed of 100 km/h (Struben 2006).

After 1905 however the EV started its decline, mainly due to the poor performance of lead-batteries and the long recharging times, as well as the easy and cheap availability of fuel for the ICE. As a consequence the ICE became the predominant form of personal travel and society became locked-in into the ICE technology (Van den Bergh et al. 2005). It might seem at first that a short-term dominance of one technology over alternative technologies should not be able to determine future growth paths, but path dependencies due to the head-start of the ICE technology strongly biased the playing field for a long time (Cowan, Hultén 1996). Firstly, batteries were no longer represented in the R&D budget of companies as demand was low. While between the 1890s and 1911 battery density more than doubled, it took until 1990 to double density again – a huge disadvantage for EVs. Secondly, people lost familiarity with vehicle types other than the ICE which certainly prevented them from choosing the “alien” electric car. Thirdly, infrastructure and related industry environment like refueling stations and fuel suppliers as a whole were veered towards the “one” technology, without any significant infrastructure – be it recharging stations or repair shops – for other platforms. (The latter two arguments and their important feedback dynamics are explored later in detail.) Fourthly, the initial advantage of slightly better performance and lower costs widened rapidly with increasing returns to scale and experience (Urry 2004).

An interesting aspect about these path dependencies is that the dominant technology need not necessarily be the best one (as with the QWERTY keyboard) and might even have negative consequences (for example the introduction and long-lasting application of CFCs in spray cans). However, an escape from a locked-in technology is difficult and is often only possible if one or more of the following six events occur (Cowan, Hultén 1996)8: Crisis in the existing technology, regulation, technological break-through for the challenging technology, changes in taste, niche markets and new scientific results.

It seems as if many of these factors are in place for the EV at the moment to escape the lock-in on the ICE technology: Firstly, the necessity to reduce CO2 emissions and ICEs being a major source thereof could be described as a crisis in the existing technology. Regulations in favor of the EV were also put in place in many countries: As early as in the 1990s the California Air Resources Board (CARB) intended to act against increasing health risks and CO2 emissions from transportation. California embarked on a plan to reduce vehicle emissions to zero through the introduction of the Zero Emission Vehicle (ZEV) Program. Originally it was planed that from 1998 on 2% of the vehicles that large manufacturers produced for sale in California had to be ZEVs, increasing to 5% in 2001 and 10% in 2003. However, the plan failed due to a law suit filed against the CARB in 20029. Currently, the EU is planning to introduce a tax on vehicle carbon emission for vehicles with emissions of more than 120 g/km10 which is still far beyond the reach of many car models. Such a tax would strongly favour the development of zero-emission vehicles (ZEV). Technological breakthroughs like the invention of lithium-ion batteries fully rechargeable within 10 to 15 minutes or systems to swap depleted batteries with fully recharged ones in less than a minute have just recently occurred, on top of significant cost reductions in battery cell production. Changes in taste are already visible as consumer are increasingly aware of environmental issues and take interest in what their consumption means for the environment; an example being the indication of lifecycle CO2 emissions of goods on the good’s label like Tesco does11. Ecologically advantageous cars like the Prius hybrid have made their way into the mass markets. Niche markets could also help EVs reach a broad-scale success: Already today, powerful electric sports cars like the Tesla Roadster or the Swiss Mindset are available, which even begin to catch the general public’s attention. But in spite of this multitude of factors favoring a switch from the lock-in on ICE to the EV platform, there are still significant barriers to entry in place that need to be overcome first.

1.3. Barriers to entry

Some of the major barriers to entry could be described as “chicken-and-egg-problems”, the first of which involves the deployment of recharging infrastructure. As ICEs experienced enormous growth from the 1910s on, an entire industry and infrastructure evolved dedicated to the ICE: Petrol refineries and stations, maintenance and repair facilities. Exactly that infrastructure is missing today for alternative platforms like the EV: As long that there are only few EVs, it is not profitable to open charging or maintenance stations as demand would be low compared to the high investments necessary, but unless there are charging or maintenance stations only very few persons will purchase an EV as range will remain limited and risk to run out of fuel on the way high. Smaller charging stations could be built at parking spots at work or at shopping centers that are highly frequented and have access to high-tension current. Larger stations could allow a complete battery exchange, something the start-up company Better Place suggests and already pursues in Israel and Denmark.

The second “chicken-and-egg-problem” concerns a lack of consumer familiarity with the EV technology. Although it might seem on first sight that an EV does not differ much from an ICE car there are quite substantial differences (Urban, Weinberg, Hauser 1996):

- New technologies, including new composite materials, propulsion control systems, high-pressure low-friction tires and deep-discharge batteries
- New attributes, such as recharging time, limited driving range on a charge, potential stranding, and perceived hazards due to high amperages
- New levels of existing attributes, such as noise (the engine is extremely quiet), smooth acceleration, no need to shift gears, and size (usually small)
- A new dimension of utility is created as consumers fulfill their environmental responsibility (and advertise this to their peers) by driving a car with no on-road emissions

As long as consumers feel not comfortable with their knowledge about this alternative platform they will not consider buying it. But as long as EVs are not bought, information diffusion about the features of EVs by means of word-of-mouth and actual consumer experience will remain low, thereby preventing consumers from buying EVs (Urban, Weinberg, Hauser 1996).

Both problems have in common that they can form a vicious or a virtuous circle depending on conditions. Once fleet size is beyond a certain level, infrastructure will become profitable enough and more people will purchase these vehicles and consequently, even more infrastructure is built. It’s thus a self-reinforcing process. The same holds for familiarity: If there is familiarity with the EV platform beyond a certain threshold, enough people will consider it for a purchase and eventually (depending on its qualities of course) also purchase it. The more people then possess the EV, the more familiar the general public will be with it and consequently stimulate additional sales of that platform. These thresholds a system needs to reach in order for a trend to become self-sustaining are called tipping points (Sterman 2000) or critical parameter values (Gassmann, Ulli-Beer, Wokaun 2007). They mark points of irreversibility. Systems they are part of are often characterized by multistability, having several equilibria (Gassmann, Ulli-Beer, Wokaun 2007 also call it a discontinuous system.) The tipping point itself marks no stable equilibrium which is quite intuitive as it is located just on the edge between two dispersing dynamics – one towards a higher level and one towards a lower level.

Apart from these chicken-egg problems further barriers for the EV exist. Firstly, battery costs are still extremely high, even for small EVs, ranging from €12,000 to €15,000 for a range of approximately 100 km per charge (Fontaine 2008). Secondly, there is great uncertainty about battery life time and the possible numbers of recharging cycles (Chan 2002). A possible estimation can be derived by looking at the length of battery warranties that EV or hybrid car manufacturers like Toyota offer12. Thirdly, it is difficult to determine current battery cell technology levels due to the huge variety of different battery cells available. In the beginning, EV ran on lead-acid batteries (Larminie, Lowry 2003) and this did not change until even the 1990s when GM presented their EV1 which also used lead-acid batteries (Werber, Fischer, Schwartz 2009). From then on, however, very soon second (alkaline) and third generation (lithium-ion) batteries were tested and eventually employed in EVs (Chan et al. 2007). The following table by Passier et al. (2007) provides an overview of the different generations of battery cells and their characteristics.

Abbildung in dieser Leseprobe nicht enthalten

Table 2 – Status of current hybrid and EV technology

A further uncertainty concerns the potential technology level that can be reached using current battery materials and what other materials provide the best chances of developing batteries with even higher density. As an example, in 1990 Lithium-ion batteries had approximately the same energy density as Nickel-Hydride and Nickel-Cadmium cells. Until 2005 however, Lithium-ion batteries improved by 2.7 times whereas Nickel-Cadmium batteries did not improve at all and Nickel-Hydride batteries improved by a factor of 1.813. The current energy densities thus vary widely among different chemicals (Figure 7). Today many companies consider lithium ion batteries the future battery technology. Zinc-air batteries could potentially store even more energy per kg, yet they are not rechargeable and would thus need some kind of battery exchange system and require reprocessing of the ingredients, potentially consuming more energy in the process than the electricity charged to the battery (Duke, Andrews, Anderson 2008).

Modeling Germany’s transition to the EV until 2040 in System Dynamics

Abbildung in dieser Leseprobe nicht enthalten

Figure 7 – Current energy density by battery type

Abbildung in dieser Leseprobe nicht enthalten

Figure 8 – Historical battery improvement rates at Matsushita

Even among lithium batteries we can see many differences and additional concerns like the security of the batteries which often exhibit elevated fire hazard risks (Li-Tec, Evonik). The following table by Chan et al. (2007) shows major differences of Lithium based batteries.

Abbildung in dieser Leseprobe nicht enthalten

Table 3 – Current status and future trends for EV batteries

Over the whole range of vehicles no battery technology is dominant, as advantages and disadvantages are very often dependent on the use of the vehicle. Whereas lead-acid batteries have rather gotten out of date due to their heavy weight and pollution and are only used in heavy duty hybrid electric vehicles, Nickel metal hydride batteries are the current choice of passenger car OEMs for power assist/full hybrid applications (Passier et al. 2007).

1.4. Modeling alternative fuel vehicle introduction dynamics in System Dynamics

Modeling these dynamics can prove to be demanding and difficult due to inherent complexity: Firstly, many stakeholders are involved, such as governments, vehicle manufacturers, consumers, environmental groups, universities, fuel suppliers and suppliers of service (maintenance and repair), which will each make individual decisions influencing the outcome (Yeh 2007). As a consequence, different relationships, decision making processes and interdependencies have to be modeled. These elements all exhibit highly non-linear behavior, many self-reinforcing feedback loops, thus calling for comprehensive tools that include the entire system with most of its elements (Struben, Sterman 2007). Trying to search for only one factor to be the explanatory variable of an observation is mostly useless as the occurrence of simultaneous developments, however small each, can determine the outcome of a system (Welch 2006).

Model examples

Welch (2006) briefly explores three different models of the transition towards hydrogen cars, as one potential alternative fuel vehicle platform. His selection criteria are that the models be dynamic (i.e. change with time), focus on (hydrogen) transition, not perform point estimates for the future and contain endogenous vehicle demand and refueling infrastructure growth. These are important factors since only models fulfilling these criteria can realistically show the evolution and escape from the lock-in (Cowan, Hultén 1996) on the gasoline ICE to vehicles based on different fuels. Showing the process rather than the mere end results is important in order to derive policy recommendations and create useful insights. Only these models can take into account the interdependencies and feedback systems over time, increasing the model’s accuracy. Finally, it is crucial to learn under which conditions the system is able to overcome these inherent chicken-egg problems, something only achievable by modeling demand and fuel infrastructure endogenously as these two elements are highly interdependent. We quickly describe the three models Welch (2006) evaluates and then argue which model is the most suitable to serve as basis for our investigation of EV introduction dynamics.

1. HyTRANS

The HyTRANS model was developed by Oak Ridge National Laboratory for the transition to hydrogen-powered cars until 2050 in order to “aid in developing credible visions of the transition to hydrogen-powered transportation, to estimate the potential benefits and costs of the transition, and eventually to understand the role of policy as well as technology in achieving the transition” (Greene, Leiby 2007). It is endogenous to the point that investment decisions are market based: Infrastructure and R&D are only conducted if expected pay-off is high enough. Furthermore, it incorporates the perspectives and views of different stakeholders (consumers, fuel suppliers, vehicle manufacturers) and has been made consistent with other US Department of Energy (DOE) forecasts. According to HyTrans, the US AFV fleet will consist of between 2 and 10 million vehicles in 2025, depending on different goals and policy decisions. In each scenario the hydrogen car is sustainable by itself. The source code to this model is not freely available.

2. Complex Adaptive System (CAS) Model

Developed by RCF Economic and Financial Consulting and several partners, among them Argonne National Laboratory, British Petrol and Ford Motor Co., the model describes scenarios of the evolution of hydrogen infrastructure. Motivated by a “lack of understanding of the transition of a hydrocarbon-based economy to a H2-based economy” (Tolley 2005), the model uses agent-based modeling and focuses on the chicken-and-egg problem between supply of hydrogen fuel and purchase of hydrogen vehicles. Agent-based modeling means each agent (consumer and investor) is modeled individually and evolves overtime. Model behavior is therefore dynamic as agents can learn from mistakes and false assumptions, change behavior and decision rules. Their findings are that the price disadvantage is huge impediment to market penetration and recommend that a tax credit – even if only limited to a ten-year period – can significantly increase the share of purchase.

3. Dynamic, Behavioral, Spatial System Dynamics Model

The last model has been developed by MIT using System Dynamics and takes a holistic view on the entire system around hydrogen cars. There is extensive coupling of infrastructure supply and vehicle demand, modeling the chicken-egg-problem of interdependence between fuel stations and vehicle demand endogenously. Welch (2006) describes it as highly dynamic and interactive in its nature, allowing for different feedback dynamics over time. Additionally it is the only model to include a spatial component, i.e. modeling of where the fuel infrastructure will be built. The conclusions include sensitivity tests on the required subsidies size, as well as a forecast on the geographic location of infrastructure for a California dataset which exhibits AFV clusters in high-density population areas like L.A. The model was developed during the Ph.D. thesis of the author; a description is therefore publicly available.

Among the three mentioned models we opted to base our work on the third model, developed by MIT’s Ph.D. student Jeroen Struben: Firstly, the model was developed for AFVs in general and could thus be transformed to be applicable for EVs. Moreover, as mentioned before, it is the only model to include spatial distribution of refueling infrastructure. In order to be able to give adequate policy recommendation this aspect is vital as drivers care about the specific spatial availability instead of the total number of fuel stations in their respective country. Also, demand for fuel at stations may strongly vary from region to region not just as vehicle stock varies from region to region, but also as actual routes driven may lead to higher demand in some regions such as cities with work places and shopping malls and lower demand in rural areas. Additionally, the model is the only one based on System Dynamics, a highly efficient modeling language that we intend to use in our own work.

Overview System Dynamics

Developed by Jay W. Forrester in the 1960s, System Dynamics is both a field of study and a modeling methodology, but could also be described as a problem-solving mindset. It originally addressed industry dynamics and their agents as described in Forrester’s “Industrial Dynamics” (1961). However, applications also evolved for other types of systems, such as governmental issues of healthcare, medicine and ecosystems (Forrester 2007), as well as “urban dynamics” (Forrester 2005). The most important characteristic of system dynamics is that a problem is embedded into a broader socioeconomic system “to guide effective decision making, learning and policy in a world of growing dynamic complexity” (Supple 2007). Dynamic complexity in this view does not refer to the large number of elements involved (which is known as detail complexity) but rather the complexity that arises from the interaction of – even only few – elements over time; i.e. evolution of the system. It is thus based on the understanding that the macro-level behavior of a system is caused by its structure on the micro-level (Radzicki 1999).

Apart from generating concrete policy recommendations it can also aide to develop intuitions about the behavior of a system and thus enhances learning for those within the system (e.g. learning in a business by taking into consideration in which ways the environment influences the business and vice versa). Rather than taking an event-oriented view that is searching for single causes to explain phenomena, System Dynamics takes a feedback-oriented view that not only includes main causes and results but also takes into account side effects, delays and changes in initial situations (Sterman 2000). It is considered by many to generate more accurate predictions and forecasts than mere statistical models that do not consider the underlying structures and determinants (Lyneis 2000).

Variables in System Dynamics models can belong to several types: Stocks (also called levels), flows (also known as rates) and auxiliary variables which interact with each other. Stocks are described by an integral over the time of the system. A certain stock at a certain time can therefore be interpreted as a certain number of objects in the system at a certain time. Flows on the other hand are always defined as the number of items per time unit that decrease or increase a certain stock (outflows or inflows). For example, a stock of widgets at a plant inventory would be described as the integral over widget inflows less widget outflows. Auxiliary variables on the other hand often have no concrete interpretation. They are used to determine other variables like flows or stocks; they can be constants or are calculated endogenously within the model.

The relationships between variables are often circular, such that an increase in a certain variable will lead to changes in other variables that will in turn eventually change the value of the initial variable. There are two basic kind of these circular relationships: positive (or self-reinforcing) ones that are can be represented by a “+”-sign and negative (or self-correcting) ones that can be represented by a “-“-sign. Positive feedbacks basically lead to infinite growth or decline to zero, whereas negative feedbacks take a correcting position, readjusting to an equilibrium.

The following is an exemplary diagram showing the relationship between profitability of fuel stations, number of fuel stations and number of cars on the road.

Abbildung in dieser Leseprobe nicht enthalten

Figure 9 – Example model in System Dynamics

The left hand side shows the reinforcing feedback dynamics between number of vehicles and number of refueling stations. As mentioned before, this is a positive feedback loop (indicated by the large plus-sign in the middle) which can also be derived from the arrows which both have another “+”-sign attached to them – like in mathematics the multiplication of two positive numbers yields a positive product. This representation can be interpreted as “the more cars, the more refueling stations” and vice versa, as well as “the fewer cars, the fewer stations” and vice versa.

The number of fuel stations is however not only determined by the number of vehicles but also by the average profitability of fuel stations. This is illustrated on the right hand side and represents a balancing feedback loop, indicated by the large minus sign in the middle of the circle which could again be “calculated” using the signs of the two arrows. The interpretation is that a higher profitability incentivizes refueling stations to enter the market (indicated by the “+”-sign attached to the arrow) but that more refueling stations reduce average profitability (indicated by the “-“-sign attached to the arrow). Taken together, the two loops could lead to a large variety of behavioral patterns of the system: For example, one loop might be so much stronger than the other loop such that the system grows infinitely or collapses quickly. Or the strength of each of the loops might be dependent upon the number of fuel stations such that it only collapses or grows infinitely dependent upon this initial level.

This simple example already exhibits clearly the difficulty of determining the number of cars and infrastructure stations to evolve. If now even more factors are included (like consumer purchase criteria) and barriers like delays (for example delays between the increased number of vehicles on the road and the consecutive establishment of more refueling infrastructure) many unexpected nonlinear relationships might be observed that shed light on the dynamics at play.

The general modeling approach takes five steps in order to build an accurate, meaningful and robust model (Sterman 2000):

- Problem Articulation: set the model boundary and define which problem should be modeled and what will certainly be included
- Formulation of dynamic hypothesis: what is the pre-modeling view of the causes of the problem? Here it is important to generate answers endogenously, so that model can explain behavior later.
- Formulation of simulation model: specify elements of the system and compute consistency tests
- Testing: comparison to reference models, robustness and sensitivity tests
- Generation of policy recommendations

In addition to System Dynamics, Struben’s model draws on other fields of study such as behavioral economics and bounded rationality (Simon 1957), product diffusion theory (Bass 1969) and insights of the topic of discrete choice (McFadden 1963) which are also used in our model approach.

2. Model description

Our model consists of three parts, representing consumer, industry and infrastructure dynamics. In the consumer section, we concentrate on dynamics regarding consumer familiarity with the EV as a new technology. As we argued, market entrance of new technologies is often inhibited by low familiarity of potential customers with that platform. Consumers learn from experiences of other users, as well as external influences from media and advertisement. Only if a certain tipping point can be overcome such that the endogenous exposure to that platform by other users is high enough to offset the loss of familiarity with a platform over time, the introduction of a new technology can be successful. In the industry section we concentrate on the degree to which EV production costs will decrease due to economies of scale, as well as technology improvement. Furthermore, as the maximum EV range is a function of battery cell energy density, we model battery cell improvements over time in detail. Lastly in the infrastructure section, we model endogenous fuel/ recharging station entry and exit behavior in order to model the basic chicken-and-egg-problem described above. Additionally, we determine in this section utility levels for vehicle trips on a regional level, based on fuel/ recharging station density and service time.

We calibrate the model to Germany, by using data about German driving behavior, vehicle fleet, vehicles sales and more on a regional basis. That enables us to reach realistic sensitivities as well as parameter estimates.

Abbildung in dieser Leseprobe nicht enthalten

Figure 10 – General model overview

2.1. The consumer model

This first part of the model serves to answer questions about consumer behavior and consumer choice with respect to the diffusion dynamics of EVs. Among these questions are: How do consumers learn about a new product (i.e. in our case the EV)? How do the consumers make purchase decisions? What is the role of public visibility of a certain product?

Abbildung in dieser Leseprobe nicht enthalten

Figure 11 – Consumer model overview

General concepts on the consumer model

Struben (2006) draws in his model on an extended Bass diffusion model (Bass 1969), a model for the diffusion of innovative products. Bass solved one of the major flaws of the classic logistic model of innovation diffusion that knew only endogenous sources of awareness growth by adding an exogenous source of awareness growth that can be interpreted as advertising, media reports or direct sales effort. We follow in our model Struben’s extensions to the original Bass model:

Firstly, contrary to the original Bass model, we separate the purchase decisions and the generation of familiarity with one product. This is important as the purchase decision is also a function of other variables such as the relative utility and cost of one vehicle platform as compared to the other. By separating these two variables, new insights can be gained about their interaction and the dynamics during the transition to the EV.

[...]


1 Consortium includes BASF, BOSCH, EVONIK, Li-Tec, and VW (Nationale Strategiekonferenz zu Elektromobilität, 25./26.11.2008)

2 Carbon Trust Analyses

3 Number of cars: ca. 41M according to the Kraftfahrzeug-Bundesamt, German population ca. 82M according to the Statistisches Bundesamt, leading to a ratio of about 50%

4 E85 is a high concentration ethanol blend with 85% ethanol and 15% gasoline

5 http://www.wupperinst.org/de/projekte/themen_online/carbon_capture_and_storage/index.html http://www.co2-handel.de/article340_10638.html

6 International Energy Agency - World Energy Outlook 2008 of November 17th 2008

7 http://www.carbontrust.com/EN/Home.aspx

8 See Hasberg, Mets 2008 for detailed discussion of these factors for EV technology.

9 http://www.arb.ca.gov/msprog/zevprog/background/background.htm

10 For a reference weight of approx. 1.500kg, CO2 allowance increases and decreases with deviations from that weight; Permitted specific emissions of CO2 = 130 + a × (M – M0) where: M = mass in kg, M0 = 1289.0 and a = 0.0457

11http://www.reuters.com/article/marketsNews/idUSL2876348920080428?sp=true

12 Toyota just recently extended their warranty to 8 years or 160,000 km whichever occurs first (http://www.toyota.com/prius-hybrid/warranty.html)

13 Matsushita Battery Industrial Co., Ltd.

Excerpt out of 122 pages

Details

Title
Migrating from Oil- to Electricity-Powered Vehicles: Modeling Germany's Transition to the EV until 2040 in System Dynamics
College
Otto Beisheim School of Management Vallendar  (Kuehne Foundation Chair in Logistics Management)
Grade
1.0
Authors
Year
2009
Pages
122
Catalog Number
V136810
ISBN (eBook)
9783640439454
File size
2035 KB
Language
English
Notes
Tags
Electric vehicle, EV, alternative energy, Better Place, System dynamics, market forecast, CO2 emissions, hybrid, zero emission vehicle, co2 abatement, battery
Quote paper
Michael Stephan (Author)Anna Feller (Author), 2009, Migrating from Oil- to Electricity-Powered Vehicles: Modeling Germany's Transition to the EV until 2040 in System Dynamics, Munich, GRIN Verlag, https://www.grin.com/document/136810

Comments

  • No comments yet.
Read the ebook
Title: Migrating from Oil- to Electricity-Powered Vehicles: Modeling Germany's Transition to the EV until 2040 in System Dynamics


Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free