I have a roadway that invloves several closely spaced merge and diverge areas but no traffic signals. All the merge points operate as free movements. What is the best MOE to llok for to determine the LOS for the roadway. The operating speed on the roadway is 35 mph. Synchro doesn't seem to report arterial LOS if there is no signal along the roadway. Any thoughts..

Thanks

I think average travel speed is a good indicator in your situation, assuming its an urban environment? If I recall correctly, arterial LOS is based on speed and road class from the HCM.
Since you can't get an arterial report without a signal, you may wind up having to do it the old fashioned way, and determine the sum total average speed manually.
You can get the speeds directionally, for each link, in the simtraffic report (report bend nodes if you need to)
set yourself up a spreadsheet that will weight the speed with the length of the link, and then take a weighted average of all the links added up. Use the link length as reported in the Synchro link properties window, and remember the speed reported at the node is for the link the traffic came from. So, the EB and WB (for example) speed at a node are for the vehicles as they approach the node, with will be two different link lengths in your spreadsheet.

Let me know if this makes sense to you.

Eric Royer
Caltrans

Eric Thanks for the reply.
How about using the delay per vehicle and linking it to LOS based on HCS criteria. Also when you take the MOEs out of simtraffic does the interval length makes a difference? In otherwords if I simulate the network for one interval (15 mins) and if I simulate it for one hour (four 15 mins intervals), will the MOEs be same on the average basis? or do I have to simulate it for one hour no matter what?

Thanks.

i am fairly sure the simulation time will make a difference. you can have the travel speed at free flow conditions at exisitng condition for different movement/path and that will be your LOS A(if you want it). and you can come up with LOS B to F based on speed. I used that in a small analysis project.

when I simulate a project my usual practice is to have 3 intervals.
interval 1) seed time, at 100% volume rate (no PHF or anti PHF), the length of this interval varies depending ont the network, but generally I like to have enough time for a car to take the longest path from one side to the other, or 3 minutes, whichever is longer.
interval 2) peak interval (recording)- 15 minutes using PHF adjust. this loads up the network at the maximum rate, and I can see where the congestion occurs
interval 3) off peak interval (recording)- 45 minutes, using anti-PHF adjust. this runs the rest of the hour at a rate that will cause the hourly rate to match my total volume, accounting for the high volume at the first 15 minutes. This period allows me to see how long it takes congestion to clear up, or indeed, if it does at all!
my MOE's are than taken from the "total of recorded intervals" which will be the whole recorded hour.

In my opinion, shorter simulation times (15 minutes) are fine if the network isn't experiencing major congestion, as the LOS probably won't be a whole lot different, and you've probably see the worst.
if there is congestion, its useful to see how queues spill back onto the rest of your network and affect upstream facilities - ramps, signals, driveways, etc.
sometimes an MOE can be "well, in this scenario the queues clear out 15 minutes faster than in the previous scenario" so its clues in how your network can affect peak demand/saturated conditions.

Eric Royer
Caltrans