For an at-grade intersection, upgrading normally involves adding additional lanes. Such additional lanes can contribute greatly to the capacity of the intersection. It is easy to calculate the Level of Service after upgrading, using computer or manual methods. However in this particular problem at this site, we are faced with many constraints and the additional lane that we want to provide is only expected to be 30 meters in length. So is it worthwhile to provide such a short additional lane? How can we estimate the extra capacity provided by thee additional lane? This article tries to reason out that the short lane is worth providing.
The site is at Junction A, in Georgetown, Penang.
Travel demand and existing configuration of the intersection is shown below. It is proposed that a 30meter additional lane should be provided at the Jalan Burma leg to allow vehicles to wait before heading north.
For J Burma/ J Gottlieb Intersection, lets anaylse traffic figures.
Based on Afternoon figures (Fig 4.3 from Lean), J burma has 124+964+329 pcu = 1417. The queue length is >30 pcu/lane. This is the most congested leg, at LOS F.
Assuming a 120s cycle time, 4 phases, the green time for them is about 30 sec. For 2 lanes, the flow rate is 708 pcu/lane/hr. In 1 hour, there are 3600 sec, 30 periods of 120 sec. for each period of 120 sec, there is only 30 sec ggreen time. For each green time 708/30 pcu must pass or 23 pcu must rush through.
I am not sure if 23 cars can rush through in each green time. If they can, good, if not there will be queue, in this case >30.
So what happens if we add a 3rd lane, say of length 30m? 30m lane will provide space for 30/5 of about 6 pcu. if additional 6 pcu can rush through in each green time, then when we have 30 green times in an hour, that is 180 pcu in an hour. That means 180/1417 or 13% higher flowrate on this leg.
If we add another 30 m lane on J Mount Erskine, that will be another 180 pcu/hr. What is effect of additional 180+180 pcu/hr on the overall performance ? Total traffic here is 4251. So an further 360 means 360/4251 or 8.5% better performance.
Judging from the LOS of E and F, I think flowrate through a traffic light cannot be as high as 700 pcu/hr. To acheive LOS D, I think a flowrate of 350pcu/hr/lane is more likely. (note that the LOS D is on legs with that range of flowrate).. If that is the case, then total flow rate in this 9 lane intersection (total) is only 3150. If this is the case, 360/3500 is more that 11 percent increase.
So we see that it does not make a big difference but it is still an improvement. It is also important to show to public that after spending rm3mill, there is some increase in capacity. In future, further improvement can be made when developers set back and the extra lane is lengthened. Our extra lane will make it possible for them to join to us.