Capacity Analysis for a signalized intersection provides LOS and corresponding queue length 95th percentile for each lane. My question is at some places i have LOS A but very high queue length like 95 ft and at some places LOS C and queue length 36 ft. What do these results imply ?
Thanks in advance for your help.

this issue has come several times in this forum and in my six years of traffic engineering life. you can search the previous discussions.
HCM really does not incude queuing in their capacity analysis.
has it neve been happened to you that you are waiting for minutes to get your gap when you come from a low volume road? this is LOS F with queuing less than 24'.
The other would be a complex thing to explain, in the signal you stopped for few seconds. but you are the 7th car of the progression. that could end up being LOS A with more than 100' queue.
by the way I would not call a 95' a very long queue length. (I know standards vary person to person, area to area)

Yes, Jahid has hit the issue. Queues and delays are not a 1 to 1 relationship. Delays reported by the models are delay PER vehicle.

Take the low queue, high delay scenario. What if 2 vehicles pull up on the side street and the cycle is 120 seconds. These 2 vehicles may wait a long time to get their green (waiting delay). So, the delay/vehicle is very high but since there are 2 vehicles, the queue is quite short.

Next, as Jahid had pointed out, consider a large platoon of vehicles driving down the road. This platoon may slow and queue up at the signal. But, if they arrived right about the time the signal is turning green, they will be delayed very little. That is, long queue, but short delays.

The Time Space Diagram can be very helpful to picture this type of situation.

Thanks,

John.

Well that was a good explanation...... lets say that I have two intersections with about 200 feet apart and the queue almost back up till the other intersections... how do i go about that then .. the intersection still show a level of service B

no HCM methodology will accurately analyze intersections that queue into each other. In this case, you're better off looking at the delay reported from a simulation of the intersections, in Simtraffic, than you are by looking at the calculated delay in Synchro. In order for the Synchro/HCM analysis to be accurate, the intersection would have to release its vehicles in an unrestricted manner. This won't be the case, if they are immediately slowing after crossing the intersection, because of vehicles stopped in the roadway in front of them.

Eric Royer
Caltrans