jackal wrote: ↑Thu Mar 11, 2021 08:27
Peter Freeman wrote: ↑Thu Mar 11, 2021 08:06
jackal wrote: ↑Wed Dec 02, 2020 20:51
... and on a road with several major junctions like the Carnavon bypass there usually will be significant changes in volume along the way that can be accommodated better by a mixture of S2 and D2 than by blanket S2+1.
I agreed with this statement when I first read it. However, on deeper thought, I think the opposite is true. To the first order, this alternative strategy of S2/D2 can provide exactly the same number of overtaking opportunities with an equal area of road surface. However, it is arguable that S2+1 might allow more judicious positioning of its 2-lane sections. This because D2 might put a 2-lane carriageway on the side where it is needed, but wastefully impose 2 lanes on the opposite side even though it is not required there. That wasted area of road surface could be better deployed elsewhere.
Regardless, I favour the S2/D2 mix!
My unstated assumption is that traffic volumes are typically similar in both directions on a bypass such as this. Hence
in terms of volumes alone I don't agree that "S2+1 might allow more judicious positioning of its 2-lane sections"
Yes,
on volumes alone (and on a flat, straight road) the two scenarios would be equal in performance. In either case, you'd spend half your distance on a one-lane section (travelling in queues at a lower level of service and lower speed) and half on a two-lane section (with the luxury of being able to overtake and achieve a higher average speed).
jackal wrote: ↑Thu Mar 11, 2021 08:27
- S2+1 by definition has much more capacity in one direction than the other so will be mismatched to volumes.
Agreed. Therefore, since with either design we would have a situation where the same volume of traffic will sometimes be in single file and sometimes on an overtaking section, the volumes on 1-lane and 2-lane sections will be the same but the level of service and hence average speed will be different.
jackal wrote: ↑Thu Mar 11, 2021 08:27
Where S2+1 is useful is for
climbing lanes. So in the case of high gradients I would agree that "S2+1 might allow more judicious positioning of its 2-lane sections".
And the hypothetical, though unrealistic, illustration of this, would be a 1km road between two roundabouts that rises to a crest half way along and then descends to the final roundabout. With an asphalt budget of 3 lane-kilometres, the optimal design would be S2+1, where the uphill pieces are 2-lane, swapping over at the crest to 1-lane going down the hill. Hardly any delays. An S2/D2 mix, depending on where the break points are placed, would leave you, and/or the other direction, struggling in your car behind a slow truck for up to 500m of that short road.
jackal wrote: ↑Thu Mar 11, 2021 08:27
So in short I think S2+1 is fine up a hill but typically not otherwise. (Some exceptions are where volumes in each direction are significantly different, or short sections adjoining junctions to allow stacking etc.)
I tend to agree. It is, of course, the reason why simple overtaking lanes are normally placed on hills. But if you can't afford D2 then S2+1 is better than just S2.
By the way, in my opinion the 2-lane sections of S2+1 should
not occur where the road is curving to the left, for visibility reasons. This design only partly complies with that rule.
In real life, the asphalt budget and my hypothetical up-down road don't exist, and so every design is a compromise and a one-off special. This design is OK - but an S2/D2 mix as you first suggested would be better and, especially, safer.