Delft was the third city in The Netherlands to experiment with modern cycling infrastructure, aided by the national government. After the experiments in Tilburg and The Hague in the 1970s, where they built one very good (but also very expensive) cycle route, that had mixed results but didn’t lead to more cycling overall, Delft took a different and innovative approach. Delft wanted to improve the city’s existing cycle network, which had a lot of missing links. The reason for this area-wide experiment was the increasing modal share of private motor traffic. The city clogged up and couldn’t cope with all those cars, it certainly wouldn’t be able to accommodate even more cars in the future. Cycle expert André Pettinga, who worked for the city of Delft at the time, summarises the need for the Delft Cycle Plan in just a few words: “The local government wished to increase the modal share of cycling!”. This cycle plan was a direct answer to the mainly car-driven Traffic Circulation Plans that had been made for many cities in the Netherlands, including Delft, in the 1960s. The execution of those plans was stopped one after the other, because of opposition of the public and changed ideas regarding urban planning.
André Pettinga is an expert who has been involved in cycling all his working life. He now works as an independent consultant operating under the name Cyclemotions. In the 1970s and 80s he was working for the traffic department of the municipality of Delft. André and I sometimes work very pleasantly together. I helped him guide groups through Utrecht for instance. Recently, I visited André in his home office in Utrecht, where he showed me the extensive archive he has on the Delft Cycle Plan with great enthusiasm. While we browsed through the cardboard boxes containing hand written notes, minutes, maps, leaflets to inform the public in multiple languages, hand typed reports, faded pictures and evaluation reports, André answered all my questions about the important Delft experiment. “We were determined to get a good network of cycle routes, not necessarily all on protected cycle paths, because we knew we couldn’t afford that.” André tells me, “And we already had good experiences with traffic calming of roads and streets.” It may be good to realise that there was no ‘zero’ situation in Delft. The presumption was, there were parts of a network, but with many missing links. The network of the plan already existed for about 75% (this includes traffic calmed streets and cycling infrastructure on distributer roads). The modal share for cycling before the plan was already 38%.
The local government of Delft did an appeal on the regional and national government for financial support. The national government was willing to spend money on this experiment, because it wanted to collect new insights (technical and sociological) on how to promote cycling. The reason for that, in turn, was that the authorities wanted to limit car use. The government expected there would be more cycling when road safety would be improved on an area-wide scale. Why chose Delft? According to André this was the reason: “In the 1970s, Delft had already invented the woonerf-concept. Its succes was based on good governance, an innovative staff and succesful public participation. No other city in The Netherlands had developed such a thorough, fundamental cycle plan.”
To investigate what the people in Delft really wanted, the Ministry of Public Works hired a German sociologist and survey expert. This caused quite a commotion in the world of traffic experts and engineers at the time. Werner Brög from Socialdata in Munich, investigated 4,700 households (Delft was a city of about 80,000 inhabitants at the time). It was an in-depth investigation. People were visited at home and they weren’t just asked how they travelled where, they were specifically asked for their constraints, the reasons why they didn’t cycle to supermarket X for instance. The answers were very concrete: “because intersection Y doesn’t feel safe”, or “because canal Z forces me to make a detour of that many metres”. The response of the survey, 72%, was very high and helped the city to identify the most important physical, financial and mental barriers to cycling in the city.
The city found out that it wasn’t just about the shortest routes for people. André explains: “People want transparency of their neighbourhoods and built up areas. A clear structure of the network that coincides with their mental map of the city. The informal connections, short cuts though parks, are just as important as larger bridges and underpasses. We learnt that we needed to give people multiple route options. People will choose their own routes based on their own motivation and judgement of levels of road safety and directness. That perception is different for every person.”
But how can you provide a complete cycle network of a high quality, when you try to minimise cost at the same time? What can you do differently, compared to the expensive one route solutions in Tilburg and The Hague? The key was to identify three different networks that would be used for a different type of trips and that would require a different type of solutions, varying in cost from very cheap to very expensive. (Which is even then still relatively low, compared to investments in highways and metro systems.)
This identification of three types of networks was completely new.
- The Urban network; a grid with a mesh width of approximately 400 to 600 metres. This part of the network would also give access to regional routes. This network is used for longer interurban and urban trips of about 2 to 3 kilometres. With this mesh width a missing link means you have to make a detour of about 1 kilometre, which some people are not willing to do. The links must therefore be provided to persuade these people to take up cycling again. For this network the most elaborate and expensive solutions needed to be chosen, because this is the network that carries most people riding their bicycles. Their volume requires larger bridges and railways and main roads should be crossed without conflict at a different level (which means: viaducts and underpasses).
- The District network; a grid with a mesh width of approximately 200 to 300 metres. The main function of this network level is to make it possible for people to cycle to local destinations in their own neighbourhood or the rest of the city; to schools, work, shops etc. A missing link in this network would lead to a detour of about 500 metres. For this network you can remedy missing links by smaller bridges over canals, shortcuts through parks, filtering car traffic by blocking off car traffic through residential areas, making one-way streets two-way for cycling, adjusting traffic light phasing, etc. Measures which are considerably cheaper than for the urban network.
- The Neighbourhood network: a grid with a mesh width of approximately 100 to 150 metres. Its main function is making it possible for children to cycle to primary school, typically short trips. Measures can be cheap and simple. They are mainly in the range of removing through (motor) traffic out of residential streets so that they become safer or perhaps building a small bridge over a ditch.
For each of the networks all barriers and missing links were identified, and a solution was considered. These solutions were prioritized, considering the number of people which would use it and the cost. Providing links that shorten often used routes got a higher priority than measures that would make cycling “only” more convenient. Missing links can be very different. They can range from a large bridge, an expensive underpass or maybe even a stretch of 1 kilometre of missing cycleway. Changes to signalised intersections were also needed, e.g. by introducing special traffic lights for cycling. The (then experimental) introduction of a right-turn on red for cycling. Readjusting traffic light phases with separated or longer green phases, especially for the benefit of more vulnerable groups (in other words: longer green for the elderly and children). Sometimes it was possible to achieve the same result (a missing link solved) with fewer cost. Removing a few shrubs for a shortcut is obviously preferable over building a new bridge. “Most important in this weighing of measures was first and foremost the objective to remove smaller and larger barriers”, says André.
The planning phase took place in the period of 1976 to 1979. A lot of stakeholders participated, the Chamber of Commerce, the Shopkeepers Federation, Schools, Homes for the elderly, the Cyclists’ Union, the Traffic Safety Societies, a society for pedestrians and of course the public. After the Cycle Plan was approved in 1979, a subsidy was requested from the national government which was granted in 1981. The entire project would cost 70 million Guilders, but since this was an experiment, only the western and southern part would be executed. It would then be possible to investigate the differences between the area where the plan was implemented (and where the missing links were remedied) and the part of the city where the gaps remained. The plans that were executed cost circa 27 million guilders in total (which would be the equivalent of 26.3 million euros today, considering inflation). The list of larger projects was as follows:
- Two large tunnels, to access the new development of Tanthof
- Three cycle bridges
- 3 km of new connecting cycle links
- 6 km of one-way streets converted into two-way for cycling
- 5 km of new cycle tracks (both separate and on-street lanes)
- 10 km of cycle tracks resurfaced in asphalt
Apart from these larger projects there was a special budget for numerous smaller interventions. Ranging from improved surface brickwork to bicycle parking racks which were placed at the request of shopkeepers. Every municipality in the Netherlands has such a (maintenance) budget and so did Delft at the time, but thanks to the Government subsidy that budget was expanded considerably and it was possible to have many more of these smaller – but very visible – improvements. The actual building phase of the Cycle Plan was roughly between 1982 and 1987. The Delft Cycle Network was festively opened on 19 September 1987 with the re-opening of the expanded underpass to cross the elevated railway near the railway station of Delft (which had been a pedestrian tunnel before). The Delft Cycle Plan was received with much enthusiasm, possibly thanks to the large stakeholder participation and the relatively low-cost for the many improvements. The Cyclists’ Union, which had been very critical of the experiments in Tilburg and The Hague, was also very positive about the Delft experiment. This is quite a contrast to the fierce opposition that was met in nearby The Hague in their experiment just a few years earlier. To objectively measure the effectiveness of the plan the “before situation” was monitored in the years 1982/1983. A study of the “after-situation” was carried out in 1985/1986. Obviously at locations where the measures of the plan had already been finished. The results were presented at many conferences, amongst which ECF-VeloCity Groningen-Delft in 1987 and Copenhagen in 1989.
The effectiveness was investigated using in-depth household surveys and road side surveys. Resulting figures showed that the primary goal of the Cycle Plan was met: the modal share of cycling increased from 38% to 41%. Those figures deceive. This actually amounts to a relative growth of cycling with 7%, because at the same time the modal share of the private car had stabilised. The growth of local car use was indeed stopped. These results were very satisfying for the authorities. It also became clear that many people chose a different cycle route. The new shortcuts were found, bridges and underpasses were well used. People did not only chose the shortest routes, other factors were also important: comfort, surface quality, protection from wind, visual quality of the route, safety etc. These shifts in chosen routes also relieved the pressure on previously busy points, leading to a better cycling climate overall. The cycling perception was also improved. It was confirmed that the urban network was the most important network. Of all trips 60% took place on this level of the route network that only makes up 30% of the total length of all networks combined. Most improvements were made to this network. Thanks to the improved cycle network destinations further away came into reach of cycling, travel speed increased with 15% and the distance travelled increased with 7%, while the travel time did not increase.
With such a good track record it was only a matter of time before other cities followed this example. The success of Delft spread all over the country and even the world. André remembers guiding hundreds of people from The Netherlands and abroad in those days. Just like he did earlier; taking groups to the traffic calmed residential areas, after the world had discovered the ‘woonerf’, “which”, André says, “is another Delft innovation”. The foreign visitors nicknamed André ‘Mister Delft’. The lessons learnt in Delft were studied nationwide, internalised and they ended in the Dutch guidelines for cycling infrastructure. Unfortunately, the Cycle Plan did not lead to a continued high interest for cycling in Delft. A bit like in The Hague and Tilburg, the interest in cycling was almost lost after this big project. However, if you go to Delft today you can see a lot of new cycling infrastructure, especially in the redeveloped railway zone. This includes the huge bicycle parking facility that I showed you earlier. The Delft experiment especially improved planning and providing for cycling in The Netherlands by demonstrating that a good quality and complete cycle network is a precondition for improving the cycle climate in a city, in such a way that people will start cycling more often and longer distances, in a safe and convenient way. André concludes: “Building urban cycle networks does really work.”
My video about the 1979 Delft Cycle Plan.
BICYCLE DUTCH 
kreeg dit artikel in frankrijk van een fiets organisatie die hard bezig is om de overheid te overtuigen wat te doen… dank voor de inspiratie!
Fantastic informative post. Thanks, as always, for recording this for posterity in an accessible way. Andre P is a hero too! One of my best memories from PCCAMS 2015 was his ‘lecture of two halves’
mooie post, Mark!
klein foutje in de laatste zin:
This post would not have been possible [[with]] the help of André Pettinga …
Dankjewel Marjolein, haha ja daar lees je dan 100x overheen, blind voor je eigen foutjes. Het is aangepast!
Almost half of Dutch people are overweight FYI
Notice how body composition stayed the same from 1987 to 2019? People are still slim and fit. Now go look at how Americans looked in 1987 and look now – all fat and unfit!