TVE 20 015 juni
Examensarbete 15 hp
Juni 2012
Carpool in Östra Sala backe
Case study on how the parking standard is
affected
Jonas Andersson
David Jakobsson
Magnus Larsson
Teknisk- naturvetenskaplig fakultet UTH-enheten Besöksadress: Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress: Box 536 751 21 Uppsala Telefon: 018 – 471 30 03 Telefax: 018 – 471 30 00 Hemsida: http://www.teknat.uu.se/student
Abstract
Carpool in Östra Sala backe
Jonas Andersson, David Jakobsson and Magnus Larsson
In this case study the area of Östra Sala backe, that will be built in the eastern part of Uppsala, has been studied. Uppsala County wants to lower the area needed for parking from todays parking standard of 1,1 parking per household. Östra Sala backe has a vision of becoming an environmental friendly area and a living neighborhood. The way the inhabitants use cars for transports currently is not sustainable. Östra Sala backe wants to change the behavior to become sustainable and to encourage usage of alternative transport, without affecting the freedom the car provides. If fewer
journeys is to be made by car or the car use is more efficient the parking norm will decrease.
In the case study a model has been developed using data about population distribution and travel patterns. Data has been obtained from two similar projects in Sweden, Västra Hamnen in Malmö and Hammarby Sjöstad in Stockholm. These projects have been chosen due to the similarities in size, located in regions with a high growth of population and they are all environmental pilot projects. Data from reports about travel behavior and patterns among the Swedish population has been used to simulate the model developed. A few assumptions have been made about Östra Sala backe due to lack of statistic and the fact that the area has not yet been built. This data is the basis for the simulation to find a suitable size and mix of cars in the carpool of Östra Sala backe.
ISSN: 1650-8319, TVE 20 015 juni Examinator: Joakim Widén
Ämnesgranskare: Joakim Munkhammar Handledare: Anders Hollinder
1
Table of content
1. Introduction... 3
1.1 Aim of the case study ... 3
1.2 Limitations ... 3
1.3 Structure of the case study ... 4
2. Background... 4
2.1 The basic idea of a carpool ... 5
2.2 Three areas of expansion ... 5
2.2.1 Östra Sala backe ... 6
2.2.2 Hammarby Sjöstad in Stockholm ... 8
2.2.3 Västra Hamnen in Malmö ... 10
2.3 Similar but different ... 14
2.3.1 The behavior and conditions ... 14
2.3.2 The areas sizes their distributions of age ... 14
2.3.3 The parking standard ... 15
2.4 Economy and environmental aspects ... 15
2.4.1 Electric car ... 15
2.4.2 Fuel driven cars ... 15
2.5 Cost of establishing parking lot ... 16
3. Methodology ... 17
3.1 The gathering of data ... 17
3.2 The model... 18
3.3 Criticism of the sources ... 19
3.4 Statistics... 19
3.4.1 Distributions ... 20
3.4.2 Long way journeys ... 22
3.4.3 Distance of different type travels ... 23
3.4.4 Work related journeys ... 23
3.4.5 Time of rent ... 24
3.4.6 Time of rent related to work and studies ... 24
3.4.7 Time of rent related to other trips ... 25
3.4.8 Long way journeys and time of rent ... 25
3.5 The purchase cost of the cars ... 25
3.6 Passengers and car travels ... 26
4. The simulation and data ... 26
4.1 Limitations of the simulation ... 27
4.2 Different scenarios ... 28
2
4.3 The simulation of Östra Sala backe ... 29
4.3.1 Deciding the number of travels by car ... 29
4.3.2 Type of trip and means of conveyance ... 30
4.3.3 Calculating the parking norm ... 31
4.3.4 Choosing a car- example calculation ... 31
4.3.5 Deciding the parking norm of Östra Sala backe – example ... 32
4.3.6 Calculating economical impact – example ... 33
4.3.7 Calculating environmental impact-example ... 34
5. The results of the simulation ... 36
5.1 Sensitivity analysis ... 42
6. Discussion ... 43
6.1 Efficiency and incitement ... 44
6.2 Environmental affect and costs of the carpool ... 45
7. Conclusion ... 46
8. Sources ... 47
8.1 Literature ... 47
8.2 Internet-based sources ... 47
8.3 Figures and tables ... 50
9. Appendix ... 51
9.1 Appendix A - Code used for plots in MATLAB ... 51
9.2 Appendix B - Java code of the simulation ... 56
9.2.1 The Main class ... 56
9.2.2 The class Systemet ... 57
9.2.3 The class Queue ... 67
9.2.4 The class GraphicData ... 68
9.2.5 The class Inhabitants ... 69
3
1. Introduction
1.1 Aim of the case study
The purpose of the study is to investigate the impact a carpool will have in Östra Sala backe when built. The aim is further to estimate how a carpool can be used as a substitute for car ownership for each resident without any drawback in terms of car availability. If this requirement can be met the area needed for parking could decrease and the case study will investigate by how much. Furthermore a minor aim of the study is to examine the environmental as well as the economical effect a carpool will have to Östra Sala backe. The economical aspects taken into consideration will be how much capital that could be saved if building fewer parking lots compared to the purchase of cars when establishing the carpool. When examining the environmental aspects the amount of carbon dioxide emissions when driving cars of the carpool will be considered. The question formulation of the case study is:
Is it possible to reduce the area needed for parking by integrating a carpool in Östra Sala backe and what will the economical and environmental impact be?
1.2 Limitations
In this case study the aspects of how the carpool will be managed by its’ managers and how employees of the carpool will behave will not be taken into consideration. The study will not consider that different inhabitants are not equally susceptible to be environmental friendly.
The carpool of the case study is not integrated in the surrounding society hence it is only located and used by inhabitants of Östra Sala backe. It is possible that the carpool of Östra Sala backe could more effective and used by a larger population if a population outside of the areas boundaries. This scenario will not be taken into consideration in the case study.
In the case study it will not be regarded that some trips to work and studies demand special types of cars because of the nature of the errand. The sorts of travels are assumed to be made by a regular medium sized car, which is not entirely realistic. Travels related to the category other errands has a share of fuel driven cars that answers to special needs. Notable is that there are electric cars with a large load capacity and such that might answer to many needs that in the study will be partly handled by fuel driven cars.
The environmental aspect of this study is rather limited since the only thing taken into considerations is the amount of carbon dioxide emissions related to driving the
simulated distances, thus only the combustion of fuel and an average emission of producing electricity. Obviously a car will have other impacts on the environment
4
during the production as well as when used in the carpool and sold on a second-hand market. If a life cycle analyze would be used to investigate a wider impact on the environment the quota of impact of the two different types of cars (electric and fuel driven) is expected to be higher than in this case study since the emissions of driving an electric car is much lower than for driving a fuel driven car. The comparison of
emissions of this study will only reflect the emissions related to driving the cars, thus the emission related to the usage of Östra Sala backes carpool.
In the case study the private economy of the population will not be considered. The purchase of private cars will not be included in economical calculations nor will parking tariffs related to private cars be considered. The economic aspect of the carpool will be an comparison of the capital to be saved when establishing fewer parking lots and the purchase of cars to the carpool. The case study will not regard aspects as administration, service and management of the carpool.
1.3 Structure of the case study
This case study consists of three parts of data, Three area of expansion, Economy and
environmental aspects and Statistics. The structure with three data sections has been
used to make the reading more convenient. Three area of expansion will present statistic about how Östra Sala backe most likely will be when it comes to age
distribution and car use. The statistic is based on data from two other Swedish projects, where the construction has progressed further. In Economy and environmental aspects the driving cost and carbon dioxide emission is presented. Section Statistics will present national statistics about travel patterns for the general Swede.
2. Background
About one in two Swedish trips are made by car as well as a large segment is
sustainable traveling, the segment of walking or riding a bike. The distribution is not equal to those of the areas we have analyzed and the distributions vary amongst them. The way the shares of transport are distributed may depend on local factors as well as what alternatives that are provided both regionally and locally. The general Swede travels as shown in Figure 1.
5
Figure 1, The Swedish distribution of transport
2.1 The basic idea of a carpool
The idea of the carpool is to make both the environment and the users a favor, but it is also supposed to be economical for the individual. The member would not need to do a major economical investment that a car purchase involves. The membership means that all the users share expenses for all the cars such as fees, taxes, insurances and service. The idea is that you pay for the car when you use it. But the carpool will deal with all the time-consuming parts that's need to be done if you own a car, such as cleaning and change of tires. An alternative might be that the members could perform these tasks and get a minor financial compensation. You should also be able to make a reservation easily and whenever you would want to. There are carpools today where you either reserve a car from your cellphone or online. The car gets unlocked using some kind of identification card, smartcard or your own cellphone. (Hammarby Sjöstad, 2012, tag service-bilpool) Just like the reservation needs to be simple so does the payment. By using a carpool the cars will be used more efficient and the space needed for parking may decrease, which is the major aspect of the study to investigate.
2.2 Three areas of expansion
The following three areas are located in regions of expansion, in different proportions but still an expansion.
Hammarby Sjöstad is located in Nacka and a part of the Stockholm region. This region is growing faster than ever. (Stockholms Läns Landsting, 2012) The project started in 1990 and is calculated to be finished in 2017. (Stockholm Bygger, 2012) Focus on the environment has been an important part of the construction of Hammarby Sjöstad. The investment on the environment includes improving the traffic situation and public transportation. (Stockholm Stad, 2009)
52,7% 11,4% 32,7% 3,2% Car Public Transport Walk or bicycle Other transport
6
The city of Malmö has a layout plan to grow notably in the years to come. The plan suggests that 1500 new residences are to be built and as many new jobs established yearly until 2020. (Malmö Stad, 2001, p. 10)
Östra Sala backe is a part of Uppsala city, which also is growing quite rapidly (in comparison to other parts of Sweden else than Malmö and Stockholm). In 2009 the population in Uppsala town was not far from 200 000 inhabitants (Statistiska
Centralbyrån, 2010, p. 95) and the increase was almost 4100 from the previous year. (Statistiska Centralbyrån, 2010, p. 113) The exact increase of the population was just more than 2 % in a year. This makes Uppsala interesting for companies who want to expand. (Uppsala Kommun, 2010a, p. 4) Östra Sala backe is located in the east of Uppsala, one of the most expanding locations in the county. (Uppsala Kommun, 2011c, p. 6) Since Östra Sala backe has not yet been built, data from Västra Hamnen as well as Hammarby Sjöstad has been used in the model to obtain a picture of how the population distribution and car use probably will be in Östra Sala backe. Why these areas have been chosen and used is motivated later on in the study, but the major factor is that they are in many ways comparable. The three following projects are all pilot projects in the sense to make the areas more environmental sustainable. The goal with the projects is to change the inhabitants’ behavior when it comes to transportation.
2.2.1 Östra Sala backe
The vision with Östra Sala backe is a living a neighborhood of squares, parks, and business places. That it will be a neighborhood where people can and do meet.(Uppsala Kommun, 2011c, p. 4) The district should be an important part in Uppsala’s east districts and be innovative. Östra Sala backe should be built with an awareness of the environment and the people living in the surrounding area. The structure of Östra Sala backe will link together the older neighborhoods and create a neighborhood with
character as a smaller downtown.(Uppsala Kommun, 2011c, p. 9) Östra Sala backe will strive for technological innovation and awareness of the environment combined with long-term financial sustainability. The latest in technology and engineering knowledge will be used. There is a vision for the area to become Uppsala's most climate-friendly area (Uppsala Kommun, 2011c, p.11) where sustainability will be the focus of
ecologist, socially and economically sustainable solutions. (Uppsala Kommun, 2010a, p. 7) Östra Sala backe can be seen as Uppsala's pilot project for eco and sustainable construction. (Uppsala Kommun, 2011d, p. 4)
The small distance of 2 km to the core of the city and E4 gives a good opportunity for public transportation and to use sustainable transports such as bicycle and walking to the city core. (Uppsala Kommun, 2011c, p. 6) It should become natural to walk, ride the bike and choose public transportation over the car. (Uppsala Kommun, 2010a, p. 26) Östra Sala backe shall be one of Sweden’s most bicycle dense areas. To meet the goal with sustainable transportation solutions more public transportation lines will go past Östra Sala backe. (Uppsala Kommun, 2011d, p. 10) Sustainable transportation solutions will be encouraged in Östra Sala backe and for those who does not have access to an
7
own car the possibility to one will be provided. (Uppsala Kommun, 2010a, p. 14)It is in line with Uppsala’s high ambition on environmental, energy and climate change
prevention. (Uppsala Kommun, 2011c, p.11)
In 2002 Uppsala had the second largest supply per inhabitant of general route (public line traffic) services in Sweden, with 105 km per inhabitant. (SIKA, 2004, p. 22)A large part of Uppsala is covered by already established public transports.
Sala backe and Årsta is a part of the area surrounding the proximity of Östra Sala backe and today almost 10 000 people live in Sala backe and 8 000 people in Årsta. A large majority of the households are families that live in multifamily houses, 97% in Sala backe and 78% in Årsta. (Uppsala Kommun, 2010b, p.6) The distribution of families with and without children in Sala backe and Årsta including single households is as follows in the diagram. (Uppsala Kommun, 2011a)
Figure 2, Distribution of families with and without child
The large amount of families in Uppsala does affect the distribution of age. There are a relatively large segment of children younger than 16 years old distributed on a not as large segment of adults.
78,3% 21,7%
No child Child
8
Figure 3, Age distribution of the inhabitants in Sala backe and Årsta
The construction of Östra Sala backe is estimated to around 2000 apartments and 4500 new inhabitants. In a further expansion 500 apartments could be built for another 1100 inhabitants. (Uppsala Kommun, 2010b, p. 6) In the model the population minimum will be 4500 inhabitants and the maximum 5600 inhabitants.
The population in Uppsala County has increased every year since the fifties and is expected to grow with another 40000 people to 2030. To be able to meet the growth approximately 20 000 new residences need to be built, this means 1000 new residences each year until 2030. (Uppsala Kommun, 2010a, p. 9)
Today's parking standard is 1,1 in Uppsala. This means that each household has an own parking space and one additional parking space for guests on every tenth house. To solve the parking issue in the area of Östra Sala backe an underground garage could be build, or alternatively solve the problem by implementing a carpool that would reduce the parking standard. (Uppsala Kommun, 2010a, p. 27)
2.2.2 Hammarby Sjöstad in Stockholm
In the year of 2009 about 17 000 people lived in Hammarby Sjöstad. When the whole project is finished the population living in Hammarby Sjöstad will be approximately 25 000 individuals, and housing around 10 000 jobs. Hammarby Sjöstad is a project that tries to establish a modern suburb close to the city, but with the closeness to
surrounding water in mind. (Stockholm Stad, 2009)
A large majority of the population of Hammarby Sjöstad is within working age and a considerably segment are children, which are younger than 16 years old. There is just a small segment of elderly, only a few percent.
14,2% 4,3% 61,6% 20,0% 0-15 years 16-19 years 20-64 years 65> years
9
Figure 4, Age distribution of the inhabitants in Hammarby Sjöstad
There are 7 000 parking spots in Hammarby Sjöstad. The planned parking norm is to be 0,7 parking per household when the project is finished. That number includes both private and public parking for the inhabitants of Hammarby Sjöstad. The planned parking standard is significantly lower than current parking norm in Uppsala of 1,1. In fact it is almost 36,4 percentages less parking spots in Hammarby Sjöstad. That means that the inhabitants manage to live and use about two thirds of Uppsala’s parking spots. The lowered parking norms have had an effect on the population’s behavior and car ownership because the parking norm has dropped from 0,75 to 0,7 between the years of 2005 and 2007 and in the same years the household with car has dropped by 4
percentages. (Stockholm Stad, 2010)
During weekdays you need to pay a fee to use the parking lot for guests. If you are a car owner you can be granted to park without to pay a fee, but then any other who also pays can park there. To get a private parking lot you need to pay for it. (Hammarby Sjöstad, 2012, tag service-parkering) That might be an economical incitement to use a carpool if you aren’t a frequent driver of your car, an extra cost to the ownership besides taxes, fees etc.
To be able to build Hammarby Sjöstad the infrastructure has been transformed. Traffic barriers have been removed and old facilities has been redesigned or removed to be able to make reality of the vision of Hammarby Sjöstad. A factor of significance is that both Nacka County and Stockholm City has been able to agree largely and therefore
effectively collaborated on both sides of the municipal borders. (Stockholm Stad, 2009) Hammarby Sjöstad offers several transportation options and this has been an important part in the commitment to the environment, and a part of the vision of Hammarby Sjöstad. There is an opportunity to take the bus to Stockholm, available both day and night. Also there is an opportunity to cross the Hammarby Lake with cheap or free ferry
19,0% 3,0% 73,0% 5,0% 0-15 years 16-19 years 20-64 years 65> years
10
trips. Tvärbanan is a tram and gives the resident one more option to get into Stockholm on most hours of the day. Both Tvärbanan and the bus connect with the many ferry routes. (Hammarby Sjöstad, 2012, tag service-kommunikationer) Hammarby Sjöstad also has carpools that offer the residents more environmental friendly transportation options than an own car. (Hammarby Sjöstad, 2012, tag service-bilpool)
The current usage of transports in Hammarby Sjöstad is not evenly distributed between the possible ways to travel. Just over half of the population uses a private car to travel.
Figure 5, Distribution of traffic in Hammarby Sjöstad 2003 2.2.3 Västra Hamnen in Malmö
Västra Hamnen is experiencing a change, a transformation from being an area of industries to becoming a modern urban settlement. The area’s proximity to the core of Malmö, Malmö C, means that there are good opportunities for a large segment of journeys to be made with environmentally friendly and sustainable transports. To break the trend of car usage cyclists will have plenty of space in reserved bike lanes. This is an action that itself will not solve the future traffic problem and achieve sustainable travel in the area of Västra Hamnen and its’ surroundings. (Malmö Stad, 2010b, p. 3)
There are just above 2500 residences and about 10 000 job sites in Västra Hamnen today. But that’s only a third of the planned residences for the future. When everything is built according to the plans there will be 20 000 inhabitants (Malmö Stad, 2012a, p. 11) distributed on around 8000 apartments (Malmö Stad, 2010b, p. 6)and Västra Hamnen will include somewhere around 17 000 work places. (Malmö Stad, 2012a, p. 11) The university has calculated to establish facilities for 11 000 students in the area of Västra Hamnen. (Malmö Stad, 2010b, p. 6)
52,0%
21,0% 27,0%
Public transport Private Car Walk and bicycle
11
Figure 6, Age distribution in Västra Hamnen, January 2011
The traffic in Västra Hamnen has increased substantially the last couple of years. The high exploitation of new residence in the area are expected to affect the commune traffic as well as the car traffic, which in turn will result in cues during peaks of traffic flows. (Malmö Stad, 2010b, p. 3) The area of Västra Hamnen is particularly sensitive to increasing flows of traffic since there are a few entrances and exits to the half island. (Malmö Stad, 2010b, p. 6) Malmö Stad wants to reallocate the traffic by taking actions to lower the private usage of cars (Malmö Stad, 2010b, p. 3) as well as support the usage of sustainable transports alternatives. (Malmö Stad, 2010b, p. 6) Estimations of future traffic are based on earlier regional studies. The distribution of traffic in Malmö 2003 was as visualized in Figure 8. (Malmö Stad, 2010b, p. 7)
Figure 7, Distribution of traffic in Malmö 2003
9,1% 8,1% 72,9% 9,9% 0-18 years 19-24 years 25-64 years 65- years 52% 21% 14% 10% 3% Car Bicycle Walk Bus Train
12
An estimation of the distribution of traffic in Västra Hamnen in year 2015 if no
interventions are made to favor sustainable travelling is as shown in Figure 9. (Malmö Stad, 2010b, p. 7)
Figure 8, Västra Hamnen in year 2015
Scenario 40 is a future scenario where the distribution of traffic in Malmö will not be
affected markedly, to and around Västra Hamnen. This will also affect Västra Hamnen and its’ traffic distribution. It is supposed that the inhabitants of Västra Hamnen have reduced their car use, but the percentage has not dropped drastically. It is estimated that the labor related journeys with car would be around 25 percent and the transportation by car for the inhabitants somewhere around 40 percentages. The total percentage of car journeys will end up at approximately 34 percent according to this scenario, as presented in Figure 10. A result of the dropping usage of cars means that the
percentages of alternative travelling will raise, especially the commuter train traffic that is believed to have great potential. To reduce the proportion travelling by car to Västra Hamnen will probably require more drastic measures. Many trips are made over weekends when the alternative transport is less available. One possible solution to reduce car usage would be to lower the parking standards by integrating carpooling in the area. (Malmö Stad, 2010b, p. 11)
43% 21% 8% 14% 13% Car Bicycle Walk Bus Train
13
Figure 9, Scenario 40 in year 2015
Scenario 30 is an estimation of how traffic will be distributed if considerable
interventions are made to favor sustainable travelling and the parking standards are lowered. Car traffic and mainly through traffic has in this scenario been reduced and the travelling to Västra Hamnen will in large part be done collectively, along with bicycle or by walking. The result of Scenario 30 will be that the traffic flows in and to Västra Hamnen will be significantly lowered and also that congestion at entrances and exits drastically reduced.(Malmö Stad, 2010b, p. 12) The shares of expected travels are shown in Figure 11.
Figure 10, Scenario 30 in year 2015
For the city of Malmö the parking norm vary depending on the type of housing. For an individual house or villa the parking standard is about twice of an apartment building. Although the norm vary for apartments with a minimum and maximum of 0,6 and 1,6 spots per household, the most common range of spots are between 0,7 to 1 spots per household. For groups with less need for a car as student or elderly, the parking
standard is significantly lower with 0,15 to 0,3 sites per household. All these standards include parking sites for visitor parking. (Malmö Stad, 2010a, p. 18)
34% 22% 11% 16% 16% Car Bicycle Walk Bus Train 26% 25% 8% 23% 17% Car Bicycle Walk Bus Train
14
One part of the measures proposed to reduce the parking standard and to achieve sustainable travel is to establish a carpool in Västra Hamnen. (Malmö Stad, 2010b, p. 15)
2.3 Similar but different
The three areas have several things in common, but also some important differences.
2.3.1 The behavior and conditions
All these areas are located in larger cities and regions that are growing. One factor that unites them is that the areas want to change the behavior of their inhabitants. They all want to make their population to travel less often by car and instead use more
sustainable alternatives. They strive to be innovative and to come up with
environmental friendly solutions and to make the neighborhood a living area for all of its population, but with a downtown suburb touch to it. They have adapted or will adapt the infrastructure to fit the commitment to the environment and to be able to build their vision.
If you are to change the inhabitants’ behavior you have to have access to alternative traveling and if you are to be able to walk or ride a bike the areas location must be rather close to the city core. This criterion is in fact true in each project and therefore it is reasonable to say that the way of traveling is at least comparable, if not equal.
2.3.2 The areas sizes their distributions of age
The sizes of the three different populations are comparable, but with one difference is that the new area of Östra Sala backe is just a few thousand but those in the whole area including Sala backe and Årsta is comparable in size.
The distribution of ages in Hammarby Sjöstad and Västra Hamnen is not equal. There are a bigger percentage of inhabitants in the working age in Västra Hamnen than in Hammarby Sjöstad. It is reasonable to assume that there are a larger number of families in Hammarby Sjöstad, since it is not only fewer inhabitants in working age but also a larger number of children. In fact it is around twice as many in children in Hammarby Sjöstad’s age segment. But the distribution of ages are not that different if you see too younger and elderly people. If you take into account those groups that would not use car as often as the other segments the differences are just a few percentages. But then if you compare the distribution of age in Sala backe and Årsta to the two other areas you can notice a considerably difference. That is probably because of todays living standard in Sala backe and Årsta, which more villas and multifamily residences than the other areas. We will in this study assume that the new area to be built with apartments will mirror the areas of Västra Hamnen and Hammarby Sjöstad.
15
2.3.3 The parking standard
One thing that does differ between the areas is the parking standard, not that much but still significantly. The highest parking standard is in Uppsala with 1,1 but elsewhere it is as low as 0,7. The standard in Malmö city is not certain, but the average is somewhere between 0,7 and 1,0 and it is not unreasonable to set a number of 0,8 which is not that different from the number in Hammarby Sjöstad.
2.4 Economy and environmental aspects
In this study the environmental as well as the economical aspect are studied. To be able to make relevant comparisons between different kinds of cars regarding effluent
emissions and economical costs three different cars have been chosen. All these three cars have an environmental profile and will form the basis of all economic and emission calculations.
2.4.1 Electric car
The car of the year 2011 was Nissan Leaf (Dagens Nyheter, 2011), an electric car with a top range of 175 km (Gröna Bilister, 2011a). In the simulation of the study the electric car that will be used as a reference is Nissan Leaf. The facts and numbers of Nissan Leaf is the numbers our calculations will rely on. The purchase price of a new car is estimated to circa 320 000 SEK (Gröna Bilister, 2011a) and the fuel consumption of Nissan Leaf is 34 kWh/100 miles, which is almost equal to 2,11 kWh/10km. (Green Car Congress, 2010) An estimation of total emissions is calculated to 10-30 grams carbon dioxide/km (g CO2/km) and is depending on the mix of electricity. (Gröna Bilister, 2011a) In our study we will use the mean value for our calculations, which is the maximum and minimum of the emission interval divided by two (equals 20 g CO2/km). Since it is unrealistic to use the top range we will use a range of 100 km for the electric car. (Vattenfall, 2012) That is because there are circumstances that will affect the
driving range such as weight, weather, driving, and discharge of battery as well as usage of AC or instruments. Those factors will reduce the driving range and therefore a 100 km as maximum driving range is used in the simulation. Still a fully charged electric car can handle the majority of everyday travels. (Park Charge, 2012)
Charging an electric car will in ordinary Swedish homes, with a regular fuse (10A) and one-phase wall sockets, take about ten hours. (Park Charge, 2012) In the model an assumption is made that an electric car will be fully charge, or near to fully charged whenever it is taken into use. The motivation to the assumption is that you could charge a battery or change to a fully charged at the carpool if needed.
2.4.2 Fuel driven cars
In the study and model two fuel driven cars have been chosen to match Nissan Leaf and to answer to longer journeys as well as special needs. One of the cars is a Volvo V70 Ethanol/Gasoline and the other car is an Audi A3 Diesel.
16
The cost to purchase a new Volvo V70 Ethanol/Gasoline car is circa 255 900 SEK. Since Östra Sala backe and its’ carpool want to have the environment in mind an assumption is made that these cars will use ethanol only. The fuel consumption for a Volvo V70 is 0,9 liters/10km and the emission of carbon dioxide is 80 g/km. (Gröna Bilister, 2011b),
The fuel consumption of diesel cars is generally lower than gasoline and ethanol cars. The diesel consumption of this Audi A3 1.9 TDI EcoPower is said to be 0,45
liters/10km, but when Gröna Bilister tested the cars fuel consumption it showed that the car consumed a bit more. Therefore the consumption of diesel by this Audi A3 is in the simulation set to 0,5 liters/10km. The purchase price of a brand new Audi A3 is circa 226 000 SEK. (Gröna Bilister, 2007)
The affect on the environment from diesel is 2,48 kg carbon dioxide per liter. When calculated with a consummation of 0,5 liters from the Audi A3, the environmental effect is around 124 g carbon dioxide/km. (Konsumentverket, 2011)
2.5 Cost of establishing parking lot
In Östra Sala backe there are 250 parking lots and an extra 70 guest parking curbstones along the streets today. It is obvious that an expansion of Östra Sala backe will increase the need of parking lots. An regional study regarding city planning shows that a
basement garage in the houses of Östra Sala backe will handle that need, given a
parking norm of 1,0. The already established parking lots of Årsta and Sala backe in the area of Östra Sala backe will make place for the construction of Östra Sala backe. Thus the 250 parking lots that exist today will be removed. To solve the higher need of parking lots that will be a result of Östra Sala backe, there will be an 200 extra curbstones parking lots along the streets, that will be guest parking (0,1 parking per household) and the addition need of parking will be covered by basement garages in the houses (1,0 parking per household). There are also other potential areas for parking, on the housing association garden but this is the housing associations own responsibility. (Uppsala Kommun, 2010b)
The cost of a regular basement garage parking lot vary in the interval 200 000-400 000 SEK depending on conditions but 250 000 SEK would probably reflect an average cost. The cost of a normal parking lot in a parking house can be lower but the parking house needs more area. A regular ground parking lot cost only 10 000 SEK. (Stockholm Stad, 2005, p. 1) If a curbstone lot is to be built the second price of ground parking is used.
17
3. Methodology
It was from the beginning of the project clear that the major aim was to investigate the parking standard of Östra Sala backe by develop a model and make a simulation. The result of the simulation should then mirror the parking norm when integrating a carpool in the area. Since Östra Sala backe will have the environment in mind it came naturally to extend the boundaries of the case study to be able to make an investigation of how the mix, emission and economic profile of the carpool would be.
The scenario of Östra Sala backe is analyzed and studied through a stochastic simulation of a model. The reason why stochastic processes are used is to represent random variations in reality as well as random human behavior. The model is based on data and statistics obtained from authorities, reports as well as from local and regional surveys. The model has been developed during the time of the case study and limited by lack of relevant data as well as limited time given the study.
The simulation itself is coded in Java, and it is a recursive simulation. This recursive type of code contains a method that repeats itself automatically until it reaches an end, in the case of this study a given time.
In the case study fuel driven cars are those combusting ethanol, gasoline or diesel, thus an electric car will be referenced as electric car not fuel driven car. When calculating and comparing the amount of emissions to be saved if integrating a carpool in Östra Sala backe, the share of emissions saved is the result of a comparison with a scenario were all travels were by fuel driven cars. The mix of cars in that scenario is an equal share of ethanol and diesel cars but not one single electric car.
3.1 The gathering of data
The data this study is based on is statistics of behavior patterns for transportation, varying statistics of the different studied areas as well as average figures of how long journeys are for different types of errands. All this data has been gathered in the purpose of simulating our designed model.
The areas that are studied are related to projects that are in many ways similar to Östra Sala backe. The data from these two other projects could be adopted in Östra Sala backe due to the areas resemblances. Relevant data from these projects have been collected from reports describing visions and progress as well as the expected future. Furthermore has some general facts obtained from each projects website as well as from the counties websites. Figures that have been used are size, age distribution, method of
transportation and parking norm.
Data from two reports published by SIKA, Statens Institut för Kommunikations Analys, has been use to collect data about general Swedish travel patterns. Data and figures
18
about the distribution by type of transportation, of type journeys for each hour of the day and distribution of distance traveled have been obtained and used in the simulation of the model.
To get the economical and environmental aspects of the carpool three cars has been chosen, each one with a relatively environmental friendly profile. Two of the cars are powered by a fuel, diesel or ethanol, and the third is a fully electric car. Carbon dioxide emission and purchase costs for each car have been obtained from websites
organizations and companies as Vattenfall and Gröna Bilister.
3.2 The model
The stochastic model developed in the case study is visualized in Figure 2. Initially there will be some amount of inhabitants, N (0), in the model (1), but with time this value of inhabitants will vary: N = N(t). The parameter inhabitant is not a fixed figure, but refers to the amount of inhabitants that is available and could possibly travel at a given time. The first process (2) in the model is to decide the number of travellers, the amount of the parameter inhabitants that are to make a journey. The number of
travellers is stochastically decided and is a function of inhabitants, N (t), and the distribution of starting a travel, d (t). The stochastic process could be described as: T (N (t), d (t)).When the amount of travellers is decided a new process (3) will calculate what share of the travellers that will be travelling by car. To receive the number of car users, C, the stochastic process (2) is considered in relation to the actual scenario, S. The scenario will answer to a specific share of travellers that will choose a car as the mean of conveyance, thus: C (t) = T (N (t), d (t)) * S. In the category of car users all individuals will be included, not only drivers but also passengers.
In the third (4) stochastic process the type of errand is to be decided. The type of errand is related to a distribution specific for each sort of errand. Furthermore they have an expected time of usage as well as expected distance of travel. The time of usage is needed to stochastically model a queue, Q (t), thus how many users there is a given hour of the day and when users is returned to be available inhabitants. The number of
inhabitants, N (t), is therefore the same as the initial number of inhabitants, N (0), subtracted by the queue size, Q (t) as: N (t) = N (0) – Q (t). When the errand is definite it will affect what type of car that is to be chosen (5). Some errands have higher
probability to choose a fuel driven car than others and vice versa. The final process (6) is to return the users when their expected time is up. The returned individual is once again available and the car is to be used again by some other user.
19
Figure 11, The model
3.3 Criticism of the sources
The reports where the data, about the three areas, have been gather is believed to be reliable since the data have been made by local and regional actors and each projects county. Data gathered about the carbon dioxide emission and cost for each car may not be entirely correct since the figures obtained from the organization might be affected by its’ opinion. The statistic about travel patterns for the general Swede is collected from a report made by an independent Swedish state department and should be considered reliable.
The cost of establishing a parking lot varies. A parking lot built in a garage is more costly than a curbstone lot. Furthermore, the source of the cost of establishing a parking lot used in the case study may not reflect the actual of Östra Sala backe.
3.4 Statistics
The statistic that has been gathered about travel patterns address how an average day looks like, whether it is a weekday or weekend/holiday. This would not affect the result since the simulation is over a large amount of “days” and the figures for an average day are being used.
In the model an assumption has been made that all the car rents is driven by a single driver and have a maximum of one passenger. This data may not be reflecting the
20
average of travellers in Uppsala, but is a figure regarding the travellers per car in Stockholm. This data has been used since similar data related to the average of travellers per car in Uppsala was not obtained.
3.4.1 Distributions
In the simulation several distributions have been crucial. People travel very differently depending on the hour of the day. To be able to decide if individuals will start a journey a general distribution of journeys have been used. There is one major extreme value in the morning, as well as a minor in the evening, which is strongly related to travels to and from the work place. But of course other types of trips, especially around the later extreme value, also affect the maximum. The data is based on a survey made during approximately one year between September 2005 and October 2006 the Swedish
population made about 13,4 million trips each day, less than 5 billion in common during that period. (SIKA, 2007, p. 23) The basis of the statistics gathered from a SIKA study is a sample of 41 225 Swedes between the ages of 6-84 years. (SIKA, 2007, p. 47)
Figure 12, Distribution of journeys for each hour of the day, thousands
During different hours of the day people are more or less prone to make different types of trips as seen in Figure 11. The distributions that has been used to decide what a traveller will do on its’ trip are those in Figure 13. The distributions regarding the errand of trip are not evenly distributed with several peaks. Travels and errands related to work and studies have peaks in the morning hours as well as in the afternoon and early evening. But the other distributions have their peaks at lunch and then reducing, except for the spare time journeys that has a second major peak just after the second work related peak.
0 200 400 600 800 1000 1200 1400 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
21
Figure 13, Distribution of journeys by type for each hour of the day, thousands
A large majority of journeys that begin during the later afternoon and evening are assumed to be journeys back home, which means that those journeys start at work and not at the carpool. Furthermore that a new car for the journey back home will not be rented once again. Therefore the model has been built in a way that an individual at work or in school will not rent a new car from the carpool during this period.
In Sweden, 73 percent of the working population is working full time, which is 40 hours per week. The figures are measured during year 2009 and it based on information from circa 60 000 persons between 16 and 64 years. (Landsorganisationen i Sverige, 2009) An assumption is made that this study and statistics mirror the area of Östra Sala backe as well as the whole population of Östra Sala backe. These 73 percent are assumed to already have rented a car when the trip goes from work to home during hours afternoon or early evening hours, thus been removed from the distribution. But the other 27 percent assumed not to be working nor studying is able to rent a car for work and studies related journeys during hours 15-18.
The figures of Figure 13 has been recalculated to a new distribution that reflects how many cars that need to be rented for work and school related trips from during hours 15-18, when the working and studying population of 73 percent are kept in mind. The second peak of work and studies travels during the later hours still exist but the quota between work and school related trips versus total trips has decreased with this
assumption, therefore fewer travels with this errand will be made during the afternoon and evening. 0 250 500 750 1000 1250 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Work and studies Service and shopping Spare time journeys Other journeys
22
Figure 14, Calculated distribution of journeys by type for each hour of the day, thousands
To visualize the amount of trips made on the different hours it is useful to see to the same statistics as above, but in percentage instead. Small numbers causes the non-continuity in the early hours. But the differences shows that a large majority of the trips in the night and morning is made related to work and studies, but later on in the day the distributions are more evenly distributed. During lunch you can see that a large majority of the trips made are for daily errands and such.
Figure 15, Distribution of journeys for each hour of the day 3.4.2 Long way journeys
Each day in Sweden there are circa 200 000 long distance journeys made, that is longer then 100 km. That many journeys a day makes around 73 million long distance trips a year in Sweden. (SIKA, 2007, p. 34) That number should be compared to 5 billion, which is the sum of trips the Swedish population made during a duration of a year between fall 2005 to fall 2006. That conclusion of the comparison is that about 1,5 percentages of all trips made by Swedes during a year are long journey trips. The
dominating mean of conveyance was the car with a segment of 68 percentages, followed by airplane and train at 11 percentages together. (SIKA, 2007, p. 22)
0 250 500 750 1000 1250 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Work and studies Service and shopping Spare time journeys Other journeys 0,0% 20,0% 40,0% 60,0% 80,0% 100,0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Work and studies Service and shopping Spare time journeys Other journeys
23
3.4.3 Distance of different type travels
The distance travelled is not equal amongst the different type of journeys. Therefore statistics related to each sort of journey has been used in the simulation. Lack of data regarding the time consumed by each kind of errand has forced assumptions.
3.4.4 Work related journeys
In 2005 the distribution over distance to work for inhabitants of Uppsala were not equally distributed. (Statens väg- och transportforskningsinstitut, 2009, p. 68) One reason could be that a noticeably segment travel out of county to Stockholm to work or travel in other directions to other cities. Of course a share of the population also travel within the county to work.
Figure 16, Distribution of the distance to work I Uppsala County in 2005
The statistics of length to work is used to determine what sort of car to be used in the simulation of the model.
Travels associated with services or shopping errands have an average distance of 30,5 km (SIKA, 2007, p. 66), thus within the range of an electric car. But it is unrealistic that all those travels could be made with an electric car. Therefore an assumption has been made in the study that fuel driven cars will answer to 10 percent of the trips associated with these tasks and errands.
The average distance for journeys related with spare time and other purposes is just more than 51 km. The critical total distance of journey by an electric car is in this study 100 km, as discussed in the chapter named Electric car. An assumption has been made that an electric car is to be chosen if possible, that is if the one-way distance is below or equal to 50 km. The reason why that assumption is made is because Sunfleet, an
established carpool, recommends its’ members to use a fossil fueled car at distances over 50 km. (Sunfleet, 2012) Furthermore that assumption will in the model mean that there is almost a 50-50 chance to choose an electric car instead of a fossil car for these errands. To be certain that a electric car will not be chosen when it is doubtful that the
6,6% 7,8% 20,0% 9,8% 9,8% 21,9% 10,4% 11,3% 0,7% 0,3% 0,4% 1,0% 0,0% 5,0% 10,0% 15,0% 20,0% 25,0%
24
driving range is enough, the probability has been set to 45 percentages to chose an electric car and 55 percent to chose a fuel driven.
The distribution of length of long way journeys is not even. The average distance is 185 km but the median is 155 km. There is a 25 percentages percentile at 110 km and a 75 percentages percentile at 260 km. (SIKA, 2007, p. 34) Therefore the median has been used as a average distance mark with the motivation that most journeys will be around 155 km.
The assumptions regarding the choice of car is not affecting the parking standards of the simulation, only the results related to emissions and economy. Those figures are not supposed to be exact but rough indicators.
3.4.5 Time of rent
A row of assumptions has been made to be able to decide the time individuals will rent their cars, that is the time being away. Furthermore is every travel type associated with an average driving time (SIKA, 2007, p. 68), back and forth. Those numbers have calculated into hours from minutes. Those driving times have then been added to the time of the rent.
Table 1, Average driving time for each sort of errand
Work and studies Service and shop Spare time Other trips
0,5 hours 0,725 hours 1,1 hours 1,01 hours
3.4.6 Time of rent related to work and studies
The foundation of the reasoning about times of rent in the study is that the majority of individuals, the car users, will begin to return home with the cars in the late afternoon or evening. Therefore, an assumption have been made that a large segment of those
leaving in the morning hours will have an average day away of eight hours, but those who leave for work later in the morning will not work a full day. Further assumptions are made that those who leave for work later in the morning will work at least two hours, but a time are randomly added of up to four hours to their workday. That will make those individuals reach the second distribution top of work travels in the later hours, those hours where most people return home. But for all the other hours of the day as well as night the minimum time of a rent is the driving time but the maximum is the driving time plus up to an extra three hours. The adding is in this case a randomly chosen number with an even distribution, which is that every hour between three and one has the same probability to be added.
It has earlier been presented that the distribution of work and studies travels has been slightly modified. That modification is in line with the reasoning rent time of work and
25
studies journeys. If an individual have a workday of eight hours, that individual is not supposed to pick a second car from the pool but drive the already rented car home.
3.4.7 Time of rent related to other trips
The time of travelling associated with the different type of errands is not that unlike, they all are relatively close to an hour. But the time gone on the different types of errands are not equally alike where an assumption is made that travels associated with individuals spare time have a maximum of five hours and all other trips have a
maximum of three hours. The actual time of being away is decided by multiplying a random number to the specific time associated with the type of errand so that the
number varies between the time of travel and the maximum time of being away plus the time of travel.
The model’s time away related for spare time errands, shop and service errands and other errands, thus that is not long way journeys nor work or study related, are estimated. It is not possible to justify time gone in these types of errands with the statistics and data obtained in this case study. Further it is not possible to motivate the time away in the same ways as for work and study trips since there are no distinguished peaks for the type of travels. Therefore the time of errands are estimated.
3.4.8 Long way journeys and time of rent
For long journeys there is a fifty percent chance of being away one day, if you are away longer the time of being away is 2.8 days (Rese- och turistdatabasen, 2009), equal to just over 62 hours. In the time away of a daytrip is assumed to be between eight and twelve hours. The actual time is randomly decided and each hour has the same
probability of being chosen. For longer trips it will be added an extra few hours as time of travelling back from the long journey. That time should reflect a car driving a
distance of half the median long journey distance (77,5 km) and has been assumed to be an average of nine hours. The total time of several days long journey is then the sum of the hours, a total of 71 hours.
3.5 The purchase cost of the cars
In the model the assumption has been made that the cars are not purchased at a full price. This has been made due to factors such as quantity discount, different tax reliefs’ etcetera. The assumption has been made that the cars are purchased for 90 percent of its total price.
26
3.6 Passengers and car travels
The traffic in the inner city of Stockholm is not equal to the traffic of Uppsala, but in the model the lack of data makes it impossible to see to passengers without making the assumption that the traffic is comparable. One major difference is the traffic tolls, even though they are not affecting the amount of travelling companions other factors could. The average of travellers per car is 1,27 in Stockholm during year 2006, and has not significantly changed since 1994. (Transek, 2006, p. 22f) To be able to see to the passengers of Östra Sala backes car travels the statistics from Stockholm will be used. The figure will not be exact and therefore affect the result slightly, but not to use the statistics would be utterly misleading of the results.
4. The simulation and data
The building of Östra Uppsala backe has in the simulation been simplified in a few ways, mostly because lack of data or statistics. The major objective of the simulation is to generate a picture of the amount of cars needed in Östra Sala backes carpool as well as what mix of cars that is to be required to serve the needs of the inhabitants.
Furthermore it is to gather data viewing how different errands could be made with an environmental friendlier car instead of a fuel driven car and how much of the emissions of carbon dioxide that could be reduced. A third objective of the model is to receive an economical picture. Enough data is to be collected to be able to make a rough
calculation of how much that could be saved or lost by establishing a carpool and therefore reduce the number of parking lots. The model will not reflect the exact distribution of how cars are used, but a rough picture will be received. The exact usage of cars is for the major purpose of this case study not relevant because the sought information is the everyday extreme values and a rough estimation of the length of driving for each type of car.
To get an accurate simulation of the amount of cars and the mix of the carpool the simulations interval must be relatively extensive. When executing the simulation over a few hundred days there are some varies, but when reaching a thousand days those differences are reduced. Therefore our simulation has a limit at 1095 days, equal to the days of three years. To chose a larger amount of days would only give an insignificantly change on the result, if any change at all.
To get an accurate parking standard the simulation is executed five times for each scenario and for different amount of individual members of the carpool. The average value is then calculated from the executions and then plotted in MATLAB. To secure the results the simulations a number of single runs have been compared and plotted with those average values. An assumption has been made that if a percentage of the
inhabitants are members of the carpool it is equal to the same percentage of the
households being members in the carpool. This assumption has been made to calculate the parking per household for Östra Sala backe. There could be a variation that some
27
households consist of more members than others, since this variation can change the result both negative and positive and the simulation is executed over a large amount of days this would only affect result slightly, if at all.
4.1 Limitations of the simulation
This report will only consist of one single carpool. The members of the carpool are only able to pick up and return the car at the carpool in Östra Sala backe. If the major
carpool of Östra Sala backe could be integrated with other carpools the usage of cars would be more effective. Due to lack of time, statistics and insecurity of where to place these minor carpools, no scenarios with several carpools will be simulated.
Figures about the driving time each car user spends for each errand has been found in the unit minutes. These figures have been rounded up to the closest integer to get the figures in the unit hours. This is a reasonable since the cars are rented by the hour. The time each errand are estimations and should be kept in mind.
There are still some years until Östra Sala backe will be built. Because of this there is a possibility that the maximum range an electric car could be driven will improve. If this improvement becomes reality it will affect the mix of our carpool and therefore the affect on the environment as well as the economic aspect. The figure of maximum range an electric car could be driven is lower in reality and in this study we have relied on recommendations from established car pools, their recommend maximum range of an electric car. If the range should increase in the years to come, the results of this report might change.
An aspect that might not exact is the amount of passengers in Östra Sala backe since no reliable figures were found regarding Uppsala. The data and figures used are regarding Stockholm, a larger city with other traffic needs and behaviors. A result of this is that the results of the simulation might be slightly misleading.
When the cost to build all parking lots of Östra Sala backe have been calculated, an assumption that no other parking alternatives then curbstones parking lots and basement garage have been built. Therefore our parking lots probably will be more expensive then the actually cost of parking lots in Östra Sala backe. The reason is that property owners may build parking lots of their own, thus the calculated number of parking lots will be slightly higher than in reality. It is because of this assumption a limit of the simulation that the number of externally built lots is not taken in to mind.
Notable is that electric cars is influenced by weather and seasons, especially in a climate alike Sweden’s with cold winters. Driving an electric car in a cold climate, with an AC and other comfort functions will affect the driving range due to extra stress on the battery. That is a factor outside of the simulation and therefore a limit since it is not regarded in the simulation.
28
The distributions describing when to start a travel, that is specific for the hour of the day, used in the simulation are not mirroring the travel patterns of weekdays and
weekends, nor fluctuations over seasons. The figures are general and will affect extreme values slightly. For example, the distinct peaks during the early morning hours related to work and studies will be slightly lower than in reality since statistics from weekend trips are taken into the calculations and vice versa. Since the simulations extreme values are lower than it would be in reality there will be a slightly shortage of cars during this period.
4.2 Different scenarios
In the simulation two alternatives have been used. One alternative was Östra Sala backe will be built to inhabit 4500 persons and another alternative were Östra Sala backe would inhabit 5600 persons. These population sizes mirror the maximum of possible inhabitants as well as the minimum of Östra Sala backe. Any size variations outside of this interval will not be reflected over in this study.
A few different scenarios have been used in the simulation to investigate how sensitive the parking norm is to variations of share using the car as mean of conveyance. The different scenarios are scenario 26, scenario 34 and scenario 43. Each scenario has a different amount of car usage that should reflect in which extent the population uses alternative transports, the figures in their names. To be able to get relatively fair results from the simulation we have defined these three scenarios. The scenarios are
estimations of the car use for travel in Östra Sala backe, but with statistics taken from reports of Västra Hamnen.
4.2.1 Motivating the different scenarios
Scenario 43 is an estimation of what Västra Hamnen expects of its’ distribution of transports without any major changes in infrastructure or public transport in year 2015. A second scenario is when some changes are made, but not major ones and changes that are very costly. In the case of our second scenario the estimated car usage would be 34 percentages, therefore the name scenario 34. If there are major changes and expensive changes to be made, Västra Hamnen will have an even lower percentage of car usage, as low as 26 percentages. This will in the model be a scenario with least amount of car usage, the scenario named Scenario 26.
Both Östra Sala backe and Västra Hamnen want to be role models when it comes to environmental aspects and therefore they want to encourage environmental lifestyles of its’ inhabitants. The different areas of Västra Hamnen, Hammarby Sjöstad as well as Östra Sala backe will make different actions to give their populations alternative transport solutions and exactly how the sum of actions will affect the usage of car is uncertain, especially in Östra Sala backe that has not yet been built. The reason why statistics of car usage has been taken from reports of Västra Hamnen is because the areas are similar. Notable is that Hammarby Sjöstad already have made a change, they
29
have a low parking standard as well as good alternative transports and available carpools for their inhabitants.
The scenarios of Västra Hamnen are relatively comparable to what is to be expected in Östra Sala backe. Therefore it is assumed in our simulation that the percentages of Västra Hamden’s scenarios are equal to those in Östra Sala backe. This means that the lowest car usage to be simulated is 26 percentages, which is reasonable since
Hammarby Sjöstad has a car usage of 21 percentages. It would not be fair to use the number of Hammarby Sjöstad’s car usage since there in Stockholm are car tolls and queue problems to mention a few differences that will decrease car usage.
4.3 The simulation of Östra Sala backe
The alternatives introduced in chapter the previous chapter Different scenarios is just simplification for the model of the case study. It is not realistic for all inhabitants to move in all at the same time. But since the purpose of the model and simulation is to get the maximum parking standard when the area is completed the simulation starts at a point when all the inhabitants have moved in.
In this case study a few assumptions and simplification has been made due to lack of data. These assumptions and simplifications will affect the result slightly. The model is simulated over a considerable amount of time, 1095 days. The simulations were then repeated for each scenario five times to receive an average number to use in the plots. The final result is the average value from these five executions. When assumptions and simplifications have been necessary they have been made rounded up to get a higher parking standard. The priority has been made that it is better that the result shows a higher parking norm then in reality instead of a lower to avoid the problems that the carpool is undersized when implemented. Even if the value of maximum of needed cars will differ slightly the quota between cars needed and residents, parking norm, will only vary a few percentage. This reports parking standard will give a reliable result and be a good indicator on what the actually parking norm for Östra Sala backe will be.
4.3.1 Deciding the number of travels by car
To decide how an individual will act in some way a random number is drawn between 0 and 1 and then compared to statistics, which is some figure obtained from a distribution or other data. In the simulation a number of comparisons like this are made to make sure that each individual acts randomly.
One of the probability distributions describes the probability that a journey will begin, regardless of mode of transport and type of trip. Initially all residents are able to chose to begin a trip, but if residents randomly uses a car they become unavailable. The amount of individuals able to choose to travel varies with time.
The process of deciding how many travellers there are every hour is made by randomly choose a number for each available individual and then comparing the obtained number
30
with the specific probability that a journey should start during the hour of the day. The sum of all those randomly decided travellers is the total population travelling that hour, and that sum is not available to make a new trip as long as they have their imaginary car.
4.3.2 Type of trip and means of conveyance
When the amount of travellers for an hour is decided, it is possible to get the share that will choose to travel by car thus the given scenario provides a percentage. But since the percentage is a decimal numeral the product is a decimal number and in the simulation all shares of persons (when there is a decimal number referring to individuals) that number is rounded up. The rounded product of travellers and the percentage given by the scenario is the number of travellers by car specific for the hour. Knowing the process of choosing travellers by car it is obvious that the number of travellers grows with growing population, but it does not mean that the parking norm grows with population since it is a quota of maximum need and residences.
Further, to ensure what errand to be made a new random number is drawn between 0 and 1. This number is then compared to probabilities, obtained from distributions for each type of errand in a given order. Those distributions are mirroring the type of journeys made during specific hours of the day, regardless of type of travel. The categories are work, service or shop, spare time travels, longer trips and other kind of trips. Every type of journeys has an own distribution over the hours of the day except the longer trips that is a percentage of every trip made. If the number is less then the probability it is compared to the next errand probability is tested against. The procedure is aborted if one errand is chosen, thus only one errand per individual can be chosen. Long travel trips are checked against a random number initially before the random number is checked against the other distributions. If a long trip is begun, it is not sure that the traveller will choose a car by travel of means. To decide if the car is to be used for the travel a random number is checked against a probability, and if a car is used the population of car travellers is reduced by one otherwise that individual will start another travel by car since it is already a car user.
There is obviously some chance that there will be more travellers per car then just the driver. To decide whether a passenger should join the driver when a journey begins again a new random number is tested against some probability. If the comparison is true not just the driver is made unavailable for some future but the passenger as well.
When knowledge of what kind of trip that is to be made statistics can be used to decide what kind of car that is to be chosen. The simulation is designed to pick an electric car if it is possible, which it is if the distance of the travel is within the driving range of the electric car. Probability of choosing an electric or fuel driven car is tested by
comparison as earlier described, but the probability varies among the different types of travels.
31
If it is known the number trips, what trips as well as what type of car used it is possible to use the average distances to calculate the total distance travelled for each specific hours. Further more it is possible to calculate the distance driven by each type of car as well as in what type of errand. To determine the cost of all trips and the costs related to the different type of cars the known distances travelled is multiplied with the driving of electric and or fuel driven cars. The cost of driving is very unstable, it shifts almost every day because of the differences in cost of electricity and fuel. These two constants, the cost of electricity and fuel will therefore be estimation for the day of us but it should provide an approximate value of the actual cost. In the same way that driving costs are calculated, an estimation of emissions can be computed.
The simulation will give us the length of all trips together and the total cost for those, but also the length and specific cost for each type of car, electric and fuel driven. Furthermore data needed to decide the mix of the carpool will be provided thus the maximum number of cars needed during on hour will be known as well as the maximum fuel driven cars and electric cars for an specific hour.
4.3.3 Calculating the parking norm
The parking norm of the carpool is calculated in the simulation. The calculation in the simulation is based on the maximum need of cars divided by the number of residences. But given a share of members, the amount on non-members will affect the total parking norm of the whole area of Östra Sala backe. To receive the total parking norm an addition of two products are made. One of these products is the multiplication of the share of members in the carpool and the simulated parking norm of the car pool, the second product is the share of non-members multiplied with parking norm of Uppsala (1,1). The total parking standard of the different scenarios and population sizes is then plotted to be simpler to understand.
4.3.4 Choosing a car- example calculation
This is an example is to show how a car is chosen in the simulation. In this example Östra Sala backe has not been expanded and therefore the population size is 4500, thus the initial parameter inhabitant in Figure 2 is 4500. Furthermore the share of
membership is set to 30 percentages, which means that a share of 30 percent of the parameter inhabitant is available to the carpool (1). The actual amount of the parameter will be 1350 individuals.
The second process (2) in the model is to decide the amount of travellers. During the 9th hour of the day the probability to make a journey is 10,34 percentages. Since every individual is given a unique random number between 0 and 1 that random number is compared to the actual probability to travel. In this example the random number is decided to be 0,0341. Since the random number drawn is lower than the probability to make a journey for that specific hour, random number < probability, a journey will be made by that inhabitant. This process of comparing random numbers to the probability
