Topics: Signal Installation, Signal Warrants, How to Annoy a Traffic Engineer
Talking Traffic Episode 12 – How to Get a Signal Installed
Hello and welcome to Talking Traffic, the podcast about bringing traffic engineering and transportation topics down to earth so regular people can talk about them. By regular, of course, I’m talking about everyone who’s not the people I work with every day. *We* like to throw jargon back and forth in a confusing spray that is merely one-upmanship at times. Nope, this podcast tries to talk about the details of transportation in a way that you don’t need to have years of experience in a design office or road crew to understand. Today is February 11, 2008, and this is episode 12. Today’s topic is “How to Get a Signal Installed.” A secondary, but no less applicable, title is “How to Annoy a Traffic Engineer.” Yes! Not only am I going to teach you the best ways to go about having your local transportation authority place a traffic signal where you want one, but I’m also showing you some of the most devastating hot buttons that we traffic engineers just hate to have pushed. If you use one of these tactics on us, be ready for the instant face-freeze followed by some sort of weasel wording. Trust me.
Before I tear into the topic of signals, I’ll preemptively point out that at the end of that last episode I said I was going to do this one on Speed, speeding and speed limits. Hahahahahahahahaha! Thus you witness the beauty of a self-published podcast, with no ties to anyone! I can change my mind whenever I want to, and there’s nothing you can do about it!
Lest you think I’m only being arbitrary, I’ll explain. There are several topics I want to cover on this podcast that take a bit more than the average amount of preparation and research. Just to give you some sort of idea, I spend about 4-5 hours a week on the episodes. That’s everything from writing it up to production of the audio and final formatting of the post. For the speed episode, as well as one I’m researching for safety, I estimate it will take about triple that amount of time in research alone. So, until I have the literature search complete, I can’t really start the writing. Yes, speed and safety are both that complicated. I promise they will come, if not soon, within the next 10 or so episodes. Remember, this is not my day job.
On to this week’s topic: Signals, and how to get one installed. I’m *positive* that if you’re a driver in America, you have, at one point or another, sat at some stop-controlled intersection thinking, “They really need to install a signal here!” If you are not one of this majority of drivers, I congratulate you. You must be more phlegmatic than the rest of us. It is a common belief that traffic signals will automatically fix all that ails an intersection. It is the panacea that can address the congestion and collision history that is perceived by the local drivers. Unfortunately, this common belief is generally dead dead dead wrong.
There are many good reasons for installing traffic signals, but there are even more bad ones. Having an exceptionally large sidestreet delay is one good reason. A signal can be installed to address an intersection with a high rate of collisions, too. Another reason to install a signal is to improve the progression of vehicle platoons in a coordinated signal system. We talked about that a bit in Episode 1 where I explained the word “coordination”. More about that in a second. A fourth reason is because the sheer volume of vehicles on the mainline and sidestreets requires one because otherwise the flows break down and the traffic delays mount. Also, if there is a large peak hour traffic generator that hits the road with an exceptionally large flow all at once, that can be a reason to install a signal. An example in my experience was a large state penitentiary that released a great deal of traffic at shift change. Or, if you have a significant pedestrian volume crossing a street at a defined location, a pedestrian-actuated signal may be necessary. These are all good reasons.
Some *bad* reasons include a perceived traffic problem by drivers approaching an arterial from a sidestreet. A few cars waiting a few minutes for a gap in traffic does not necessitate a signal. Similarly, one collision at an intersection, no matter how terrible, does not ncecessariy require a signal. Another bad reason is because someone influential decides that a signal should be installed for political reasons. Some communities are better at resisting these sorts of pressures than others. Traffic calming is also a poor reason to install a signal. Traffic signals are *not* traffic calming devices. They shout not be installed in an attempt to control the speed of a roadway. It won’t work and it’s liable to make the situation worse.
Okay, I’ve covered some of the good and bad reasons for installing traffic signals. How does one go about actually deciding to install a signal? Back in episode 9, I mentioned the manual on uniform traffic control devices, or MUTCD, the traffic engineer’s bible. This bible contains 8 guidelines for determining whether a signal is needed. They are called Signal Warrants, and if they collectively or individually determine that a signal is required, the traffic signal is said to be warranted. It’s important to note, however, that the satisfaction of one or more signal warrants does not necessarily justify or demand the installation of a traffic signal. These are guidelines and worksheets, not requirements. There is no substitute for engineering judgment, which is why I have a job.
On to the signal warrants. As I mentioned, there are 8 of them. The first four are volume warrants, having to deal with the amount of traffic passing through an intersection. “Traffic” in this case, includes pedestrians. The volume warrants focus on specific “hours” in the measured traffic. Warrant number 1, which is primarily intended to address intersections where vehicular volumes are large, examine what is called the eighth highest hour. If you count all of the traffic entering an intersection during a 24-hour period and then rank them in order of traffic, from top to bottom, the one that is Number 8 from the top is the eighth highest hour. Easy! Well, I’m actually glossing over a great deal of detail, but this episode isn’t about teaching you how to conduct a signal warrants analysis, it’s about signal warrants basics. Warrants 2 and 3 deal with the fourth highest hour and the peak hour, respectively, and warrant 4 deals with pedestrians. For warrant one, you merely compare the numbers you have on your eighth highest hour to a table found in the MUTCD. If your numbers are higher, congratulations! Your signal is warranted. If not, you can go to warrant 2 and 3, which uses graphs. If your graphed point is above certain lines, then again, congratulations! Your signal is warranted. Unfortunately, most agencies don’t like to install a signal based solely on warrant number 3, the peak hour warrant. It’s only one hour. It’s better to hit the four hour, or eight hour warrants to justify a signal.
Warrant number 4 is the pedestrian warrant. What it says, in brief, is if you have 100 pedestrians per hour for four hours during the day (or 190 during any single hour), and there aren’t sufficient gaps in the traffic stream to allow pedestrians to cross, you meet the warrant. This sort of pedestrian volume only happens in urban areas, downtown business districts, or maybe near open-campus schools.
Warrant number 5 is the school crossing warrant and as it is named, it is primarily interested in how many students are crossing an unsignalized roadway. This warrant also specifically states that it shouldn’t be considered until you’ve exhausted other possible measures. It’s never good to put in a signal that is only used twice a day. Like the peak hour warrant, this one is used sparingly.
Warrant number 6 is the Coordinated Signal System warrant. As I mentioned in the intro, this one is intended to provide platooning of vehicles inside a coordinated signal system. What is platooning? It’s the deliberate stacking of vehicles into individual groups called platoons that will then flow through the signal network with as little delay as possible. If all your vehicles are stacked up, it’s easier to time the system so they hit the maximum amount of green lights. If your vehicles are randomly distributed, then a lot of them will be hitting reds. Signals sometimes are installed to keep these platoons together, especially when signals are placed far apart; a signal between two distant ones will maintain the optimum progression.
Warrant number 8 (warrant 7 we’ll get to in a moment) is the Roadway Network warrant. It is intended for the times when you may want to install a signal at a major road intersection that doesn’t justify it now, but good practice means you want vehicles to go there to make their turns. This is typically done when you have a major arterial on the state highway network being crossed in multiple locations by roads that might serve the locality as feeder routes to the arterial. It can ease congestion if everyone goes to the same crossing road before getting onto that arterial, and that is the nature of this warrant.
Warrant number 7, is a biggie, like warrant 1. This one is the Crash Experience warrant and it’s used when there is an intersection that might be exhibiting high crash rates. This is the one that will cause traffic engineers to grit their teeth at people in a public information meeting. There is a perception that if there is a fatal or serious collision at an unsignalized intersection, it would have been prevented by a traffic signal. What people fail to realize is that traffic signals will often *increase* the collision rate at an intersection, although they also generally reduce the severity of the resulting collisions. For example, it’s likely that crossing and angle collisions like T-bones and Left turns will be replaced by rear-ends and sideswipes, which are less severe. “Okay,” you say, “We had that fatal collision, so let’s install a signal and be done! You just said that installing a signal will reduce severities!” Yes, I did say that, but if that one fatality is the only collision at an intersection since 10 years ago, are you really warranted in installing a signal? The answer here is No, and I won’t even be wishy washy like I do on other topics. Any one isolated incident does not justify the radical changing of traffic control, whatever form that change may take.
Now, I’ve been drawing a picture concerning an intersection with an actual severe collision or death. If you really really really want to annoy traffic engineers, come up and speak passionately about “near misses” and “dangerous intersections.” For example, and this is direct firsthand experience, I had conducted a traffic study on a suburban arterial roadway. There was a cross street that served a high school a numerous subdivisions. This intersection was perceived as extremely dangerous and congested by the majority of the people whom I spoke with. While I agreed that this was a congested intersection during commuting periods, I could find no real evidence of the “danger” of the intersection. Traffic engineers typically look at 3-5 years of collision data while analyzing crash statistics. I had 3 years in hand and there were only 8 collisions. That’s 2 and 2/3 per year none of which qualified as “severe”. This is not the mark of a dangerous intersection, especially given the volume of vehicles passing through every day. Yet the fact that there is a high school up the road trumps engineering judgment. I actually, truly, had a man come up to me at the public information meeting and ask, “Son, how many kids are you going to kill before you install a light?” I was tempted to answer, “Three”, but I did not. I was a good traffic engineer.
In case you care, we did not end up installing a signal there.
The crash experience warrant is designed to try and quantify these factors. It depends on a certain volume of traffic, 80% of the warrant 1 volumes, plus 5 crashes in a 12 month period that are correctible by the installation of a signal. That is key. Rear end crashes are not considered correctible. Side swipes are not, either, nor are head-on crashes, normally. Angle collisions and turning collisions are the ones we look for. Remember, the installation of a traffic signal is likely to increase the number of rear end crashes.
Before you start accusing me of being cold blooded, remember that these warrants are only recommendations. There’s nothing saying that you can’t install a signal at an intersection that doesn’t meet warrants. This happens all the time. It’s up to the engineer in charge to make a final recommendation, and sometimes the local politicos have to okay these decisions, too.
This episode has been about how professionals analyze and intersection to determine if a traffic signal is needed. I didn’t quite deliver on my promise during the intro to teach you how to get that traffic signal that you want…or did I? I’ve taught you the approved process, and there’s no reason you can’t follow it as a private citizen. Crash records are subject to freedom of information act requests, and you can easily pay a traffic counting firm to do traffic counts at an intersection. For that matter, you can grab a lawn chair and count traffic all on your own. Although I warn you that high volume intersections are hard to do all by yourself. As I mentioned earlier, I glossed over a lot of the details of a signal warrants analysis, but all of the procedures are there for the reading in the MUTCD, Section 4c. It’s not exactly an easy read but it’s definitely not rocket science; if you’re willing to study the procedure, there’s no reason you can’t produce a warrants analysis at least as well as I.
Whatever you decide to do, be mindful of what a signal is supposed to do and what it is not. Signals are, by their nature, the most restrictive kind of traffic control, possibly imposing the most delay on an intersection. They are intended to manage the flow of vehicles in a manner that serves the maximum number of travelers while minimizing their delay. In a lot of cases, an all-way stop would do this job better. Signals are not a form of speed control. They will not reduce the speed of vehicles away from the signal. They are not primarily safety treatments, either. They are not the answer to all roadway problems. There are many things that can be done to a roadway that will address issues that might be much more effective than a signal. Also remember that signals are expensive and that it’s *your* tax money that pays for them. It’s not just the initial investment, either; it’s the maintenance and upkeep year round that requires engineers and technicians. So, think about your pocketbook when advocating for your signal.
This episode of talking traffic is issued under a creative commons attribution non-commercial, no derivatives license. The music you hear is by Five Star Fall off of Magnatune.com. Thanks for listening, and I hope you’ll tune in next week for my arbitrarily decided topic. It might be speed, or it might not. I haven’t decided yet. I want to take this moment and wish a Happy Birthday to Steve Eley of Escape Pod.org. Escape Pod is my favorite podcast which features weekly readings of excellent science fiction short stories. If you listen, you may notice some similarities in format between Escape Pod and Talking Traffic. If you’re going to steal ideas, steal good ones. Happy Birthday Steve!
If you like what you hear, or if you hate what you hear, or if you think I’m a callous goon for my comments about the crash warrant, please send an email to Bill@talking traffic.org. Or you can leave a comment on the show notes at www.talkingtraffic.org. I appreciate all the feedback I get, even if people tell me I talk too fast. I can’t help it, I’m from New England. Have a great week!