Random transportation musings

[tradephoric]

What are some random musings you've had regarding basically anything transportation related?  My list is heavily favoring practices that would improve arterial road efficiency.

1.   6-lane high speed arterials should be boulevards.
2.   Traffic signals should be limited to 4-phases.
3.   Maximum pedestrian crossing distances shouldn't exceed 100 feet.
4.   No roundabouts should be installed at an intersection that carries over 50,000 ADT.
5.   Roundabouts should be considered at intersections along residential 2-lane roads.
6.   When a partial cloverleaf interchange is considered, a Parclo B4 should be the preferred design over the Parclo A4.
7.   Grid networks that achieve good 2-way progression should be selected.
8.   Left-turns should be eliminated at intersections crossing 3 or more lanes of opposing traffic.  Innovative designs such as the jersey jughandle, Michigan left, CFI's, PFI's should be considered at such intersections.
9.   Superstreets should be widely used along the grid network especially at intersections that would ruin good signal progression (the "half mile"  signals).
10.   A downtown area should utilize one-way streets and town-center intersections to maintain an efficient arterial grid network for commuters.

[Alps]

#2 is a rule of thumb that traffic engineers go by, so it's not just a musing. Of course, there will arise complex situations when that's not possible. You can also have lead/lag phasing on both approaches (example: Bloomfield Ave./Passaic Ave. in West Caldwell, NJ), which gives you six nominal phases (lead green WB, both through green, lag green EB; repeat NB/SB).

#4 is covered by other rules regarding roundabouts. I think 40,000 AADT is when you go to a 3-lane roundabout. And when you have that many lanes, you start thinking about other solutions. But 50,000 is doable in 3 lanes should you choose it.

#6 entirely depends on traffic volumes and directionality.

[roadfro]

#8  Certainly should be a consideration where feasible, but is not doable in many circumstances--i.e. the entirety of Las Vegas where most arterials are 3 lanes each way so there'd be way too many jughandles, Michigan Lefts, etc. to fit in narrow right of ways.

[Revive 755]

3.   Maximum pedestrian crossing distances shouldn't exceed 100 feet.

As in 100 feet for the entire intersection, or for a single section between refugees (for example, crossing one half of the road to the center median, then have a seperate crossing for the other half of the road as happens here http://maps.google.com/maps?hl=en&ll=42.056347,-88.135285&spn=0.003366,0.008256&t=m&z=18&layer=c&cbll=42.056347,-88.135285&panoid=0Di-v5S0GO2LxQU3g38CtA&cbp=12,146.1,,0,2.81)?

10.   A downtown area should utilize one-way streets and town-center intersections to maintain an efficient arterial grid network for commuters.

I skimmed a research paper the other day that said a two-way network with left turns prohibited would actually be more efficient - though I'm not convinced.


A few more candidates

11. SPUIs should be used over compact diamond interchanges unless there is a large skew present between roads.

12. Roundabouts should not be used at intersection with high pedestrian volumes

13. Roundabouts should not be used at intersections near railroad crossings (I'm guessing a spacing of 750 feet between the entry to the roundabout and the railroad crossing, but this would probably be somewhat dependent on the volumes using the roundabout.

14. Traffic signals should not be installed at an intersection where the new signal will create a queue that backs into another signalized intersection.
       A) When the new signal is under study, it shall be assumed that the new signal is not coordinated or has fallen out of coordination.
       B) The analysis for checking the queuing should either use volumes projected for 20 years or at least a 25% increase in traffic volumes.

[Alps]

14. Traffic signals should not be installed at an intersection where the new signal will create a queue that backs into another signalized intersection.
       A) When the new signal is under study, it shall be assumed that the new signal is not coordinated or has fallen out of coordination.

That's shitty traffic engineering and management. You should always assume coordination if it's within the network. Whatever plans the agency has, assume proper operation according to those plans. Otherwise you're hanging yourself and the agency out to dry.

[Scott5114]

That's shitty [...] management.

Is there any other kind of management???

[Alps]

That's shitty [...] management.

Is there any other kind of management???

Traffic management, believe it or not, can be effective. Think incident response, when VMS switch on and radio stations start broadcasting about delays and alternate routes.

[roadman65]

I think that during road work where signal detector loops are inactive and the default timer takes over that at night when hardly any traffic is using the side streets, that the timer be set to only three seconds or maximum five seconds.

Also, if the road has a protected left turn phase and it is not generally used at night, why have it part of the cycle anyway.  Just disconnect it!  We have too many signals that have you sit and wait for a protected left turn arrow where no vehicle is turning left at all and if there is, could make the turn without hesitation.

My pet peeve, is the side streets that have separate phases instead of the left turn signals to move traffic during peak hours on the secondary as well as the primary roads there, still operate when it is overnight and no traffic is using the side road.  Then you sit and wait for the max time first on one side, than the other, and then last the left turn signal/ protected left before you can go.  You waited over a minuet, burned some gasoline, and for what?  The possibility that one car or truck might enter that particular intersection!

Orange Blossom Trail and Sun Life Path is currently set up that particular way now during median installation along OBT for the next 5 months.  It has one long green for the empty Florida Mall, then one long green again for the opposing side of the mall followed by the OBT left turn signals.  Why not even, combine both the mall and other side with a simultanius green and let people yield to another like the old fashioned way for the night hours?   That would cut out one phase.  Then shorten the greens to five seconds except OBT, being the main artery, having at least thirty seconds. 

Engineers make enough money to pay the high fuel costs, many of us do not. 

Also, if a left turn lane is over 200 feet long, then the green time on the left turn signal should allow for all vehicles in that spacing to make the signal as well.  Many road agencies, use the left turn vehicles as the sacrificial lamb as the elongated turn lane is keeping the vehicles off the primary general purpose lanes for other traffic flow.  Meanwhile, you could wait 2 or 3 cycles just to make a simple left turn and creating a catch 22 where the lane is filled due to shortened left turn signals.

[Revive 755]

14. Traffic signals should not be installed at an intersection where the new signal will create a queue that backs into another signalized intersection.
       A) When the new signal is under study, it shall be assumed that the new signal is not coordinated or has fallen out of coordination.

That's shitty traffic engineering and management. You should always assume coordination if it's within the network. Whatever plans the agency has, assume proper operation according to those plans. Otherwise you're hanging yourself and the agency out to dry.

I see it as a realistic fail safe method since in some areas the traffic signals seem to fall out of sync every other week.

[tradephoric]

The Orlando area has a lot of wide intersections with long pedestrian crossings.    Possibly the city chooses to run high cycle lengths at these traffic signals (even in the middle of the night) to mask the effects of pedestrian actuations.  When you have pedestrian crossings over 150 feet long it can become quite disruptive for the main arterial when a side-street ped runs.  Maybe John Young Parkway could get away with running a 80 second cycle for the vehicle movements in the middle of the night and during the weekends... but then a pedestrian crosses at Deerfield Blvd and Deerfield suddenly requires 51 of that 80 second dial (153 ft crossing / 3.5 feet per second + 7 second walk time).  It's not so bad if you have light pedestrian traffic but once the pushbutton breaks it looks like complete hell.

Running long cycle lengths leads to a new set of problems; mainly poor signal progression   The travel time along a 45 mph arterial between two traffic signals spaced ½ mile apart comes to 40 seconds.  Good 2-way signal progression can be achieved when the cycle length of the signals either matches the travel time between the signals or is 2X the travel time.  In this example, the theoretical perfect cycle for 2-way progression would be either 40 seconds or 80 seconds.  How many traffic signals run an 80 second dial in the Orlando area at ANY time of day?   Anything over an 80 second dial will start to lead to poor signal progression in at least one direction of travel.

Solutions for Orlando?  #3, #8, #9.

[tradephoric]

3.   Maximum pedestrian crossing distances shouldn't exceed 100 feet.

As in 100 feet for the entire intersection, or for a single section between refugees (for example, crossing one half of the road to the center median, then have a seperate crossing for the other half of the road as happens here http://maps.google.com/maps?hl=en&ll=42.056347,-88.135285&spn=0.003366,0.008256&t=m&z=18&layer=c&cbll=42.056347,-88.135285&panoid=0Di-v5S0GO2LxQU3g38CtA&cbp=12,146.1,,0,2.81)?

I meant 100 feet to the nearest pedestrian refuge.  Here are some examples of crossing a 6-lane road:

6-lane boulevard in Troy, Michigan (required ped time = 23 seconds)


6-lane road crossing in Orlando (required ped time = 59 seconds)


Sure, you only get the pedestrians half way across with the pedestrian refuge design but they only wait for two short cycles as opposed to one long one.   At least the vehicles along the major arterial aren't waiting 59 seconds for a sole pedestrian to cross the road (or worse yet, waiting 59 seconds for a phantom pedestrian to cross the road because the pushbutton is broken).

[roadman65]

One thing, though, the pedestrian signal call does not need to be deactivated during construction.  If someone needs to cross the street, and they are law abiding as many do not push the button and wait, they can call for the light.

Many people on OBT do not use crosswalks.  I see many folks even walk two feet away from the striping, litterally! Then when good ole Lynx drops off a passenger, even though the crosswalk is only a few feet away, the person will walk behind the bus and cross the street even though he or she has to zig zag around stopped cars especially when the signal is green in their favor and a direct straight walk is better for them than to walk a crooked line.

Even school children at Oakridge Road will walk around each individual automobile than use the crosswalk, even with the many nuts on the road nowadays that could run them over during a rage attack. 

Then on to another subject with traffic light spacing being less than one half a mile.  John Young Parkway at LB McCleod with that signal at Clear Way where the  JYP traffic from the LB McCleod signal gridlocks that intersection.  I am sure you know why that signal is there in the first place at Clear Way also.  It was a deal made by the City of Orlando to give the neighborhood behind the Isle of Catlina (or whatever the hotel is now called) to give them their very own traffic light because of the social class of the neighborhood there.  Either the light could be on LB McCleod (as it was many years ago) or at that location.  For some reason the City of Orlando moved it to there.  Anyplace else, it would be a right in and right out!  I think it should be that as the traffic counts for Clear Way does not warrant that signal, and it is not that hard to use LB McCleod to reach JYP SB or I-4.  Also SB JYP could make a u turn at LB McCleod as well!  Rules do not always apply in some cases.

[tradephoric]

Then on to another subject with traffic light spacing being less than one half a mile.  John Young Parkway at LB McCleod with that signal at Clear Way where the  JYP traffic from the LB McCleod signal gridlocks that intersection.  I am sure you know why that signal is there in the first place at Clear Way also.  It was a deal made by the City of Orlando to give the neighborhood behind the Isle of Catlina (or whatever the hotel is now called) to give them their very own traffic light because of the social class of the neighborhood there.  Either the light could be on LB McCleod (as it was many years ago) or at that location.  For some reason the City of Orlando moved it to there.  Anyplace else, it would be a right in and right out!  I think it should be that as the traffic counts for Clear Way does not warrant that signal, and it is not that hard to use LB McCleod to reach JYP SB or I-4.  Also SB JYP could make a u turn at LB McCleod as well!  Rules do not always apply in some cases.

Ahhhh yes, when politics trumps all logic and reason!  Reminds me of a light at the County Center complex along Telegraph that gets a favorable amount of green time even on a Sunday afternoon when hardly anyone is coming out of the complex.  Interestingly enough, the only hard stop I came to along a recent 20 mile drive down Dixie/Telegraph was at that County Center light.

At least after the long wait you get a green and progress quite nicely down the rest of Telegraph Road.  The worst thing is to get stuck at a long red light only to get your green and get stopped at the next red light down the road... that's a pretty huge pet peeve that I'm sure every motorist has experienced.

[Alps]

14. Traffic signals should not be installed at an intersection where the new signal will create a queue that backs into another signalized intersection.
       A) When the new signal is under study, it shall be assumed that the new signal is not coordinated or has fallen out of coordination.

That's shitty traffic engineering and management. You should always assume coordination if it's within the network. Whatever plans the agency has, assume proper operation according to those plans. Otherwise you're hanging yourself and the agency out to dry.

I see it as a realistic fail safe method since in some areas the traffic signals seem to fall out of sync every other week.

That sucks. There are places that can't maintain their signals, but most of the agencies I deal with do a thorough job (NJDOT, Port Authority, Newark are a few examples).

[Brandon]

14. Traffic signals should not be installed at an intersection where the new signal will create a queue that backs into another signalized intersection.
       A) When the new signal is under study, it shall be assumed that the new signal is not coordinated or has fallen out of coordination.

That's shitty traffic engineering and management. You should always assume coordination if it's within the network. Whatever plans the agency has, assume proper operation according to those plans. Otherwise you're hanging yourself and the agency out to dry.

I see it as a realistic fail safe method since in some areas the traffic signals seem to fall out of sync every other week.

That sucks. There are places that can't maintain their signals, but most of the agencies I deal with do a thorough job (NJDOT, Port Authority, Newark are a few examples).

Then there are places that rely solely on actuated signals.  These are never in sync, and make driving very frustrating.  Yes, IDOT, I'm looking at you and IL-59.

[kphoger]

Then there are places that rely solely on actuated signals.  These are never in sync, and make driving very frustrating.  Yes, IDOT, I'm looking at you and IL-59.

Funny, my mind immediately jumped to Illinois when I read the first part of this post.  The first road that came to mind, though, was Naper Blvd/Naperville Road/whatever it's called through Naperville and Warrenville.

[KEK Inc.]

For 8.  If there's a lane wide center lane or turn lane it should be fine to turn left.  In arterials you shouldn't turn left if there's no dedicated lane.  Loop around the block a la 19th Ave in San Fran.

[vdeane]

That must be a west coast thing.  On the east coast, we just make a left turn unless the road is physically divided.