Sub-problem 2b: Moe Road PM peak hour - With Conditions

In Exhibit 2-25 we show the results from four runs (Datasets 18, 19, 20, and 21). The first is a base case condition. In it, we’ve used the PM With intersection volumes along with an assumption that there are no pedestrians. We used an average cycle length that minimizes delay and balanced the delays among the approaches. The delays average 21.2 seconds per vehicle and range from 10.1 seconds to 33.5 seconds. The average queue lengths range from 0.2 to 11.1 vehicles and the 95^th-percentile queues are double that.

The second run looks at the effects of uncertainty. What happens if the site generated traffic is 30% more than predicted. We see that the average delay increases slightly from 21.2 to 23.9 seconds. The three critical moves (as can be seen from the base case) are the eastbound through, the westbound left, and the northbound approach. These have the highest delays. In the second run, we see that two of these moves have major changes in delay: the eastbound through sees an increase from 23.8 to 28.2 seconds and the northbound approach sees a change from 31.9 to 36.4 seconds.