An essay on the Delft transport network in the Netherlands

1 About the Delft network
Delft is a city located in the Netherlands, having a population of 96168. It is spread over an area of 24 sq.km. (Wikipedia 2009a). The Delft University of Technology is the largest technical university in the Netherlands, with over 15,000 students (Wikipedia 2009b). Due to the significant student population, cycling is an important mode of transport.

Get Help With Your Essay
If you need assistance with writing your essay, our professional essay writing service is here to help!
Essay Writing Service

Delft is bound by freeways on all sides, except the southern, which is delimited by a highway. The hierarchy of roads in Delft is freeways, highways, main roads, and local roads. The railway line cuts across the city, running in the north-south direction from the centre of the city. A main road and a highway are the main linkages between the two freeways on the either side of the railway line.
2 Transport Problem in Delft
The railway line passing through Delft is a component of one of the principal railway connections in the Netherlands, connecting Rotterdam and The Hague. The existing 2-track railway line is insufficient, and it has been planned to be upgraded to a 4-track line. However, due to lack of space to build another viaduct, and environmental problems caused by the railway line, it has been proposed to rebuild the railway line in a tunnel. 2020 is envisaged to be a typical year while the construction is in operation. Thus, Links 10486, 9873, and 9827 will be blocked for all traffic. Due to the detours which would result due to the blocked links, cyclists and cars would have to deal with increased travel times. It is anticipated that Mercuriusweg/Abtwoudse Pad or the Ruys de Berenbrouckstraat links would be used in lieu of the blocked links (Chen 2009).
3 Description of Current scenario and Alternate Scenarios
3.1 Description of the Current Situation 2003
Zones 1-7 are the external zones, whereas the rest of the zones are considered as the internal zones. The internal zones are the primary areas of concern. It has also been given that in 2003, zones 23, 24, and 25 are still under development and there is no travel demand from and to these zones (Chen 2009).
3.2 Description of the Zero-Alternative Situation 2020
By 2020, zones 24 and 25 will have been completely developed. However, Zone 23 would still not have been developed. Infrastructure has been upgraded accordingly to connect zones 24 and 25 with the rest of the city. Zone 12 remains the highest trip generating and trip attracting internal zone, as does Zone 1 amongst the external zones.
3.3 Description of Future Situation during Construction 2020
Considering the future scenario for the city of Delft in 2020 during the expansion of the rail line from 2-track to 4-track, it shall be assumed that this project will be in progress and 2020 will be a typical year. It has been given that during the expansion, Links 10486, 9873, and 9827 will be blocked (Chen 2009). The trip generation remains the same as for the zero-alternate scenario for 2020.
4 Modelling for Current Situation 2003
4.1 Trip Generation
For performing the trip generation modelling function, the Zonal-based Multiple Regression model has been employed. It takes into account the linear relationships of the socio-economic characteristics of the households in the zones, which affects the trips produced and attracted for each zone (Ortúzar and Willumsen 1999).
The following functions have been used for the trip generating model in the case of internal zones:
Pi = 0.3RESIDENTSi + 0.06JOBSi + 0.03RESEARCHi + 0.03EDUCATIONi
Ai = 0.05RESIDENTSi + 0.75JOBSi + 0.3RESEARCHi + 0.3EDUCATIONi
Where
Pi = production of zone i,
Ai = attraction of zone i,
RESIDENTSi = number of residents in zone i,
JOBSi = number of jobs in zone i,
RESEARCHi = research facility space in zone i,
EDUCATIONi = amount of educational services offered in zone i.
Source: Chen 2009
For the case of external zones, the numbers of trips generated and attracted have been assessed based on the traffic volume counts. These are as indicated in Appendix 1.
For the internal zones, it can be seen that the variable of the number of residents living in each zone is the most dominant one affecting the number of trips produced. The variables which are most dominant for trip attraction are number of jobs, and the research facility space in that particular zone respectively.
Appendix 2 indicates the trip generation for 2003, including both the internal as well as the external zones.
2003 histogram.JPGFigure 4-1: Productions and Attractions for the Current Scenario 2003
It can be seen from the Appendix 2 and figure 4-1, Zone 12 is the largest trip generator and attracting external zone. Zone 1 is the highest trip generating and attracting internal zone.
4.2 Trip Distribution and Modal Split
The Gravity Model has been employed to generate the trip distribution model. This model is a type of Synthetic Model as “it estimates trips for each cell in the matrix without directly using the observed trip pattern” (Ortúzar and Willumsen 1999 p.159). There are three variants for the Gravity Model function, namely:

Exponential Function: f(cij) = exp (-ßcij)
Power Function: f(cij) = cij-n
Combined Function: cnij exp(-ßcij)

Singly constrained versions, in this case destination-constrained, can be produced by making Ai equal to 1, i.e.
Ai = 1 and Bj= 1/?iDif(cij)
The Combined Function variation of the Gravity Model is the most superior one, as it is the best fit with the actual observed values for trip length distributions of cars (Ortúzar and Willumsen 1999). OmniTRANS performs simultaneous modal split, for which the gravity model can be extended to the “simultaneous gravity model” which is given by:
Tijv=p.Qi.Xj.Fv(zijv)
Where
Tijv= Number of trips from zone I to j via mode v
p= Scaling factor
Xj= Column Balancing Factor
Fv(zijv)= Distribution function taking into account the willingness to travel by mode v given impedance z
Appendix 4 indicates the traffic distribution of cars in Delft. 44566 cars have been distributed. As can be seen in Appendix 5, 8635 cyclists have been distributed the entire network of Delft.
The modal split indicates that 83.77 percent of the traffic is constituted by cars and 16.23 percent by bicycles.
The trip length distribution function for cars can be seen in figure 4-2. This indicates the willingness to travel to a certain distance without any impendence, such as congestion. Hence, in this case the maximum uncongested distance travelled is 10 km. Similarly, figure 4-3 indicates the trip length distribution for bicycles, which experiences its peak at 3 km.
The trip time distributions for cars and bicycles have been indicated in Figure 4-4 and Figure 4-5 respectively. These functions, too, follow a similar pattern as that of the Trip Length Distribution. In the case of cars and bicycles, both, the peak is of 12 minutes.
The mean trip distance for bicycles is 3.14 km and the mean trip time is 14.22 minutes. Similarly, for cars, the mean distance is 7.59 km and time is 11.12 minutes.
Intra-zonal car trips constitute only about 0.57 percent of the total number of trips made by cars, and those made by bicycles constitute about 16.55 percent of the total number of bicycle trips.
4.3 Traffic Assignment
4.3.1.1 Assignment for Cars
The General Equilibrium technique of traffic assignment has been used for cars. The condition as given by Wardrop is that “Under equilibrium conditions, traffic arranges itself in congested networks in such a way that no individual trip maker can reduce his path costs by switching routes” (Ortúzar and Willumsen 1999 p.303). Under this traffic assignment model, costs are all perceived in the same way by the trip makers, and hence Stochastic effects are not considered.
The following links have the largest traffic flow (for a single direction) of about 7064 cars assigned: 10972, 10973, 11289, 11427, 11428, and 11477. Figure 4-6 represents the traffic assignment for cars in 2003. The colours of the bands, in the graph, indicate the V/C Ratio and the width indicates the traffic load. It reflects that the main road, the east and west freeways crossing the railway are suffering from V/C ratios greater than 1, which requires attention.
4.3.1.2 Assignment for Bicycles
The Stochastic Method has been employed for the traffic assignment model for bicycles. This method is superior to the All-Or-Nothing Method, as it takes into account the diversity in the road users’ perceptions of distance, travel time, generalised costs; and thus considers alternate routes to the best-route choice, given by the All-Or-Nothing technique. This is a suitable model for the traffic assignment of bicycles as this technique does not take into account the congestion effects, which is not calculable since capacities for bicycles are not specified.
Link 10850 has been assigned the maximum number of bicycle trips for a single direction, which is 941 trips. Figure 4-7 represents the assignment of bicycles in 2003. Most of the trips are confined to the internal zones.
5 Modelling for Future Situation 2020
The modelling stages for the future scenarios of 2020 will remain the same till the Traffic Assignment Stage, which would differ based on the blocked links for the future construction scenario of 2020.
5.1 Trip Generation
The Zonal-based Multiple Regression model has been employed. The functions are the same as those for 2003. However, forecasts for the number of residents, jobs, and the research facility space and education building space in 2020 are different from the current year. Also, it has been given that the productions and attractions generated by the external zones is to be assumed to grow by 15 percent from 2020 to 2003 (Chen 2009).
The above figure 5-1 illustrates that Zone 1 still remains the highest trip production and attraction external zone; and Zone 12 continues to remain the highest trip producing and attracting internal zone. The same is reflected through Appendix 3.
5.2 Trip Distribution and Modal Split
The Gravity Model has been employed for the trip distribution. A total of about 54380 cars have been distributed, and about 10413 bicycles. The modal split for 2020 demonstrates that about 83.5 percent of the trips would be by cars, and the remaining 16.5 percent by bicycles. For cars, the intra-zonal trips constitute about 0.54 percent, whereas for the bicycles, intra-zonal trips constitute about 15.71 percent of the trips.
Trip length distribution function (indicated in Figure 5-2 and 5-3) and the trip time distribution functions (indicated in Figure 5-4 and 5-5) have been generated. The maximum uncongested distance remains the same for both cars and bicycles, as in 2003, as also the maximum uncongested time in the case for cars. For bicycles, the maximum time has reduced to 8 minutes.
The mean trip distance for bicycles is 3.26 km and the mean time is 14.75 minutes. Similarly, for cars, the mean distance is 7.56 km and time is 11 minutes.
5.3 Traffic Assignment for Zero-Alternative Situation
5.3.1 Traffic Assignment for Cars
The General Equilibrium model has been employed. The following links have the maximum numbers of cars assigned, i.e. about 8346 cars, to them for one direction: Links 10972, 10973, 11289, 11427, 11428, 11477, and 11478. Figure 5-6 represents the assignment for cars. The colours of the bands, in the graph, indicate the V/C Ratio and the width indicates the load of traffic. The freeways continue to have high V/C ratios, as well as the main roads crossing the railway.
5.3.2 Traffic Assignment for Bicycles
For the traffic assignment modelling for bicycles, the Stochastic Method has been used again. Link 10850 has been assigned the maximum number of bicycle trips (114 trips). Figure 5-7 represents the bicycle assignment.
5.4 Traffic Assignment for Future Construction Situation
In the future situation during construction, the decision to upgrade the railways from two tracks to four tracks has been made. This would result in the closing of Links 9827, 9873, and 10486.
5.4.1 Traffic Assignment for Cars
It has been calculated that links 11283, 11407, 11463, and 11464 have the maximum load, of 10723 cars in one direction. Figure 5-8 represents the car assignment considering future construction. The cross-railway link south of zone 17 experiences a greater load, whereby there is a reduction in the northern cross-link.
5.4.2 Traffic Assignment for Bicycles
The Stochastic model employed for assigning the bicycle traffic that Link 10452 the maximum load, in one direction, of 1446.54 bicycles. Figure 5-9 represents the assignment for bicycles.
6 Analysis
6.1 Trip Generation
Amongst the internal zones, which are of primary concern, Zone 12 is the zone which remains the highest trip producing and attracting zone. This is supported by the fact that Zone 12 has the highest number of residents living in it, and the maximum number of jobs existing in this zone. This zone is located at the heart of the city, adjacent to the railway line and the main road connecting the two freeways. All the internal zones have experienced a growth rate for the number of trips generated and produced by about 11 percent.
Zone 1 also remains the highest trip producing and attracting external zone. This may be accounted by the fact that Zone 1 is directly connected to Delft by the railway line, and thus could perhaps be an important junction or location.
6.2 Trip Distribution and Modal Split
The trip distribution can be reflected through the study of the matrices given in Appendix 4 to 7. For 2003, the maximum number of car trips has been of the nature Internal-to-External, with the most dominant one being from Zone 15 to Zone 1. This has not changed for 2020. For 2003 and 2020, the most dominant character of bicycle trips is Internal-to-Internal, with maximum trips being made within Zone 12. This is as should be expected, since long distance travel by bicycles is not likely to be made except in rare circumstances. The maximum number of bicycle trips may also be attributed to the fact that Zone 12 is the Delft City Centre. The maximum growth (4.8 times) for car trips has been seen for the pair Zone 7-Zone 1, and the maximum for bicycles (0.15 times) has been for the pair Zone 5-Zone 12.
On studying the zone-to-zone average travel distance and time, it can be seen from Appendix 8 and 9 that the maximum increase for trip distance has been from zone 7 to 15 (46.86 percent) and maximum increase in time has been for the pair zone 18 to 22 (48.67 percent). The maximum increase (48.78 percent) in cost has been experienced for zone 11 to 22.
The modal split indicates a marginal change from 2003 to 2005, whereby there is a decrease from approximately 83.77 to 83.5 percent for cars. Intra-zonal trips made in 2020 have experienced a fall from 2003. This may indicate greater travel distances and time, and thus greater generalized costs in the future.
6.3 Traffic Assignment
On comparing the figures indicating the traffic assignments for the three scenarios, it can be seen that the maximum car trips load remains on the freeways defining Delft on the eastern and western sides. The main difference can be seen between the zero-alternative and construction situation for 2020.
Due to the blocked links, car traffic has been assigned to the links 1164, 11645, and Westvest-Hooikade particularly have experienced a significant increase, whereas the Westlandseweg links have experienced a decrease in the load of trips. Also, as was expected, there was an increased use of the Ruys-de-Berenbrouckstraat, and Abtwoudse links. However, the Mercuriusweg saw a decrease in the traffic load as compared to the zero-alternative scenario, which goes against as was anticipated. In the case of assigned bicycle trips, due to the blocked links, the Hof-van-Delftlaan link has seen a substantial increase in the traffic load.
The share of the railway crossing traffic has been seen to change for all the three scenarios. This has been indicated in figure 6-1. There is a decrease in the share from 2003 to the Zero-Alternative scenario. However, comparing the future construction scenario with the Zero-Alternative, there has been an increase in the share, for both cars and bicycles.
7 Conclusions
Zones 1 and 12 being of prime importance, also supported by their connectivity will continue to dominate as the zones attracting and generating the maximum traffic. Since maximum bicycle traffic load is concentrated in Zone 12, measures may be taken to restrict the entry of cars in this area so as to ensure a more safe environment, which would even benefit the households, as the maximum percentage are located in this zone. Increased generalised costs must be taken into account, based on the above analysis. Links which are experiencing greater traffic loads due to construction works have been identified, and relief measures for these should be planned, particularly for the freeways on the eastern and western side of the city.
8 Strengths and Weaknesses of OmniTRANS
The strength of OmniTRANS is that it helps in the rapid and accurate transport modeling procedures. Also, the data can be very easily examined on the transport network map, rather than just being confined to tables and numbers, which makes analysis tedious. Also, a number of variables can be studied, which can be further sorted out based on the direction for any particular link.
The weaknesses could be cited as only being able to employ the use of the Simultaneous Gravity Model being available for trip distribution stage. Also, only limited types of traffic assignment models can be performed (such as Stochastic, All-or-Nothing, and Equilibrium). Public transport modelling is also not performed by the software (OmniTRANS 2009).
References
CHEN, H. 2009. Handout: Coursework Description, lecture notes distributed in TRAN5020 Principles of Transport Modelling. University of Leeds, 3 November 2009.
OMNITRANS. 2009. What’s New in OT5 [online]. [Accessed on 1st December 2009]. Available from http://www.omnitrans-international.com/resources/brochures/what’s%20new%202008.pdf.
ORTÚZAR, J. & WILLUMSEN, L.G. 1999. Modelling Transport. West Sussex: Wiley.
WIKIPEDIA. 2009a. Delft [online]. [Accessed on 10 November 2009]. Available from: http://en.wikipedia.org/wiki/Delft.
WIKIPEDIA. 2009b. Delft University of Technology [online]. [Accessed on 10 November 2009]. Available from: http://en.wikipedia.org/wiki/Delft_University_of_Technology.