A 54m long tram every 3 minutes, replaces roughly an 8 km car queue with 1.1 km of trams.

What length of previously queuing cars woud  putting a tram in that queue likely remove?

(NOTE1- this article is under review – whilst the calculation and assumptions are more or less correct we are checking  we have expressed the conclusion corrently, but the general drift is undoubtedly correct)

(NOTE1 the selection of a 54m tram at  3 minutes imterval is not a reccomendation, it is used for illustrative purposes; in reality possible a shorter / longer tram and more or less frequent might be found to be optimum)

Here is a rough estimate for the likely queue length reduction created by a 54m tram line running in a long, heavily congested single carriageway (ie with a tram in the same traffic queue, not segregated) once every 3 minutes ( at peak only). (If it’s only 27m, then 6 km will likely be removed, if its 27m long and ever 12 minutes then 3 km etc). this estimate is based on the demonstrated shift of car drivers previously making the same journey by car to the tram:

A typical 54m tram can carry 10,000 passengers per hour, and with 18 trams per during peak periods (30 per hour in Budapest), occupies a total of 54 x 18 = 1km worth of road during one hour, whereas the cars on the road can only carry 1,000 cars per hour =  1,300 passengers plus drivers per hour typically), in a slow moving queue occupying a length of say 8 km for one hour ( at one car per 8 m spacing say). Thus the tram lenght and frequency replaces 8km of queue with 1 km of now swift moving tram.

Trams have regularly shown that they have a high modal shift, of previously 20-30% of cars drivers  to a new   tram, unlike buses with a close to 0% modal shift. And the UK experience is that roughly half of that 30% ie 15%*  (ie  0.15 x 10,000 = 1,500) will be  ex driver + passengers = 1,500 / 1.3 = 1,100 cars) who have  shifted from cars on that specific route to the tram, removing a potential queue previously occupying 1,100 x 8m = 9 km which is more than the length of the previous car queue of 8 km.

(for basis of assumptions see further down *)

Thus, potentially the tram could make the traffic queue disappear and both can proceed at higher speed unimpeded by the previous queues.

Another way of looking at it:

The max capacity of the A4 through Saltford in terms of cars per hour is 1,000. With a tram every 2 minutes at peak the number of cars per hour carried in the tram is 20 x 0.15 x 1000 = 3000 which is 3 times the previous capacity.  But this would require a tram instruction length into the “bottleneck” of 56 x 20m /h = 56 x 20 /  8 car lengths per hour = 140 queued car lengths.  Thus, the tram adds 3000 cars per hour, and still leaves road space for 1000 – 140 = 860 cars per hour, meaning the effective car carrying capacity at peak has increased from 1000 / h to 3860/h.


Of course, some people will quite rightly want to or need to drive but they will find the route far less congested than before because so many previous drivers and their passenger will now be in the car.


And this is indeed the case in all of UK 7 previous highly successful re trams – Edinburgh, Newcastle, Birmingham, Sheffield, Nottingham, Croydon, Manchester which have had significant modal shifts from cars to trams and reduced congestion and pollution.


In addition, because of the high numbers of passengers in a tram, the highway authority will grant Green Wave Traffic Light Pre-emption to the tram which ensures that it rarely if ever encounters a traffic queue since the cities’ lights can be manipulated in advance to permit this.  See a video of this on here:


Please sign the petition: https://you.38degrees.org.uk/petitions/bring-back-modern-trams-to-bath-to-cut-congestion-and-regenerate-the-city-1?source=rawlink&utm_medium=socialshare&utm_source=rawlink&share=8935e90e-6860-4afd-86ff-5da9cc64ec3c


Another version of the calculation is here:


Note 2. We are not saying this would apply to the road pictured above, or in any road in Bath – that can only be determined by a proper analysis and modelling, so this is very much hypothetical, but it illustrates the point. And in fact in all re-trammed cities traffic has reduced.
Note 3. Of course due to suppressed demand, those drivers who are deterred from driving at this time will create more traffic so undoubtedly traffic restraint would be applied to ensure the smooth running of the tram.

Sources and Assumptions:

omments from David Walmsley. Chartered Member of the ILT and a Member of the CIHT.  BSc and PhD in physics.

The PTEG report gives 19% for Croydon (16% drivers and 3% passengers).

It also gives figures for Manchester as 18% peak, 30% off-peak and about 44% weekend. I have done some averaging between the Bury and Altrincham lines here.

For Nottingham, it says 21% of passengers use the 5 park-and-ride sites, and I would expect it to be a bit higher when other car drivers are included.

I would take Lewis’s 25% figure. I suspect the oft-quoted 30% contains a bit of “optimism bias.”

Remember that there are also likely to be 15 to 20% of passengers making a trip they didn’t make before. *My rule of thumb is “half to two-thirds of the tram passengers are former bus passengers, about half the rest transfer from car and the others are new journeys.”

Also remember that these are percentages of the numbers of tram passengers. 25% transfer from car to tram means 25% of the tram passengers, not 25% of the car journeys.

David Walmsley

 Further details on modal shift: