> the cost to the nation of traffic arriving 5 minutes later?
Traffic delays are dominated by congestion and road capacity does not increase with higher speeds. Once you hit the peak for that road section, trafic grinds. If, for example, you have a 30 minute comute that spends 15 minutes in congestion-limited trafic and 15 minutes on a speed limited empty highway, you would need to drive 50% faster on those sections to recoup a full 5 minutes, so something like 105 mph on a 70 mph section for a full 10 minutes. That doesn't sound realistic.
There is a cost to low speed limits, clearly, but it is usually marginal - and when it's not, you should address that specific problem of a too low speed limit instead of making speed limits unenforceable throughout. If a section generates lots of speed tickets and no even minor incidents, it's clearly a sign to increase speed.
> drivers de-rating the risks.
True, but the risk rating they assigned to higher speeds was probably not proportional to the actual exponential increase to begin with.
> We don't have more accidents on the fastest roads in the UK.
That's moving the goal posts. The increase in risk is exponential on any given single road. Some roads are exceptionally safe at low speeds so you can increase the speed quite a bit and keep the risks low, hence the higher limits.
>> road capacity does not increase with higher speeds
Fundamentally untrue. If you reduce the average speed of a section of road, say for example in heavy rain, the capacity or throughput of that section of road is reduced since less vehicles can pass through in the same time.
If you can keep drivers attentive enough to maintain safe distance between vehicles, you can increase the speed of a section of road.
>> Once you hit the peak for that road section
You can increase the practical peak in an imperfect world by increasing attentiveness of drivers.
For example, you could defer the onset of gridlock by X minutes by making more efficient use of the available road space. In some cases it will be possible to increase X such that gridlock is never encountered for the same traffic volume.
>> Some roads are exceptionally safe at low speeds
And yet the focus on speed means that roads which are not exceptionally safe at low speed have reduced incentive to be improved. Why improve a road (hard in some cases) when the impact is not measurable and speed is an easy band aid.
Better to pretend the immeasurable doesn’t exist and simply focus on speed. Even if other approaches could have a greater impact on road safety.
The capacity of a sufficiently long road section does not increase with speed because the safety distance must also increase with speed. A good rule of thumb is to keep a 2 second (or 3 second) distance between vehicles, and that means that the peak throughput of any single lane is 0.33-0.5 vehicles per second regardless of speed.
The total time spent on the road increases and the utility of the road decreases since it gets you to your destination slower, but the capacity does not increase, it's the same capacity with an added delay line. In fact, the safety distance increases more than linearly with speed, precisely because the combined effect of reaction time, braking distance and severity of incidents is more than linear. The 2 second rule is just a rough approximation.
These are really "fundamentally true" issues that are well known in traffic engineering so there is nothing much to debate. The rain example is disingenuous since in rain the braking distance increases dramatically, drivers instinctively know this and maintain larger clearances. And the demand of the car traffic increases too in bad weather, for example people avoid using bicycles or walking and opt for a car, which injects additional vehicles into a road network and pushes it over into disequilibrium and gridlock.
>> The capacity of a sufficiently long road section does not increase with speed
Most roads aren't sufficiently long and there's plenty of other factors and practical road usage is nowhere near this perfect maximum of safe usage but to point out your mathematical error:
For 10 minutes a section of road 1 mile long has traffic pass through it. The cars have 3 second gaps between them.
At 30 mph: 160 cars max throughput
At 60 mph: 180 cars max throughput
In the 30mph example, for the first 2 minutes, 0 cars have completed the 1 mile stretch, for the period 2-10 mins, cars pass the finishing line every 3 seconds. 160 cars complete the 1 mile stretch in 10 mins.
In the 60mph example, only the 1st minute is spent without traffic completing the throughput analysis. For the period 1-10 mins, cars pass the finishing line every 3 seconds. 180 cars complete the 1 mile stretch in 10 mins.
Yes, you need to double the speed and construct an artificial trafic burst to notice a marginal increase in capacity that is strictly due to the increased delay (and storage capacity) of the slower road. That was my point. In practice trafic ramps up slow enough so that we can ignore the lost capacity due to empty roads in the early hours. After that, it is steady state for both scenarios.
> there's plenty of other factors and practical road usage is nowhere near this perfect maximum of safe usage
Precisely, but this even further argues against high speed limits. If the road tends to be congested, windy, has trafic lights and is poorly maintained, the areas where you can increase the speeds are very limited and the total impact in trip time is low. So high speeds not only not deliver higher throughputs but they fail to even deliver speed.
Traffic delays are dominated by congestion and road capacity does not increase with higher speeds. Once you hit the peak for that road section, trafic grinds. If, for example, you have a 30 minute comute that spends 15 minutes in congestion-limited trafic and 15 minutes on a speed limited empty highway, you would need to drive 50% faster on those sections to recoup a full 5 minutes, so something like 105 mph on a 70 mph section for a full 10 minutes. That doesn't sound realistic.
There is a cost to low speed limits, clearly, but it is usually marginal - and when it's not, you should address that specific problem of a too low speed limit instead of making speed limits unenforceable throughout. If a section generates lots of speed tickets and no even minor incidents, it's clearly a sign to increase speed.
> drivers de-rating the risks.
True, but the risk rating they assigned to higher speeds was probably not proportional to the actual exponential increase to begin with.
> We don't have more accidents on the fastest roads in the UK.
That's moving the goal posts. The increase in risk is exponential on any given single road. Some roads are exceptionally safe at low speeds so you can increase the speed quite a bit and keep the risks low, hence the higher limits.