Greenhouse Gas Emissions Reductions from Traffic Improvements Are at Best Negligible 4th Order Effects

Checking some citations within a draft report that I have been reading led me to pervasive alternative facts about the relationship between greenhouse gas emissions and traffic improvement strategies.  It seems that not only are we daily subjected to incorrect interpretations about the sources of global warming, but we are additionally subjected to incorrect ideas about what to do about GHG emissions.  People are asserting, without proper analysis, that traffic engineering improvements designed to reduce delays can result in significant GHG savings.  I was able to find many examples on the Internet, but let me quote just one from the authoritative organization, AASHTO.

“Greenhouse gas reductions may be achieved through maximizing the efficiency of the transportation system, through effective management of available road capacity and reducing delays. Intelligent transportation systems that cut down on congestion and idling in traffic include ramp metering, message signs warning of disruptions, real-time traveler information, variable speed limits to help cars flow smoothly onto a highway, and advanced traffic signal controls that change timing based on traffic load.”  AASHTO, “Real Transportation Solutions for Greenhouse Gas Emissions Reductions”

Statements like this, although unsupported by evidence, are fundamentally harmless.  There are huge safety and mobility benefits from most traffic engineering improvements, so can this white lie be justified?  I am OK with it, just as long as everyone (wink, wink) in our profession is in on the joke.  Unfortunately, there appears to be many true believers, and these true believers are distracting us from what we really need to accomplish.

Order of effects are thus.

First order:  Reduction in vehicle miles traveled (VMT).

Second order:  Technological changes to the vehicle fleet (while driving those VMT).

Third order:  Improvements in how vehicles are operated (when traveling those VMT).

Fourth order:  The net difference between vehicle operational savings and the losses from VMT increases attributable to induced demand, if any.

I think we can all agree than any traffic engineering improvement, such as raising speed limits, that allows free-flowing traffic to go faster than about 50 mph is counterproductive, both for safety and GHG reductions.

Third order effects include such items as voluntary speed reductions, better braking and acceleration habits, minimizing unnecessary idling, and chaining multiple trips to avoid cold operation.  Third order effects are wonderfully personal, reminding everyone about the seriousness of global warming, without actually accomplishing very much directly.

Fourth order effects are negligible at best and negative at worst, since they are the difference between two, fairly equal, counterbalancing factors.  As discussed in my earlier blog (“Answer to the Fuel Savings Riddle”), increased speeds cause more travel.  I argued for a paradigm of constant person-hours-traveled when evaluating fuel savings and GHG emissions reductions.  Undoubtedly, there are many places where it is unreasonable to expect more travel when there are faster speeds, but those places don’t have a lot of congestion, anyway.  The big potential GHG reductions are in large urban areas with lots of traffic, where people self-limit their travel because of time-budget constraints.

Asserting positive GHG benefits from any traffic improvement program would involve a very difficult analysis process.  An essential requirement is a sufficiently elastic travel demand model, which properly considers induced travel from increased average speeds on trips.  Since global warming is a long-range problem, the analysis must also be long-range and must include a full slate of behavioral and land-use changes.  Comparatively few agencies have this capability, although it is considered to be practical technology.  Anything less is dishonest.

Before and after changes in traffic would need to be precisely assessed, likely with a well-regarded microsimulation package.  The speed outputs of most (but not all) travel models are not good enough for this task, without extensive post-processing.  A good DTA might be sufficient, however.

The most logical source of emissions estimates would be MOVES, but I am personally OK with direct application of the Davis equation, as illustrated by my previous blog.  The application of the Davis equation would require second-by-second speed and acceleration data for every vehicle, which would be cumbersome, but not impossible.

And, after doing all this work, you will likely not see any benefits whatsoever.

Of course, the proof of the pudding is ground truth through a before-and-after implementation study over a very long period of time.  Unfortunately, such a study would be ridiculously expensive for the amount of insights gained.  And I doubt that we would have the patience for it, as well.

As always, your opinions are welcome.

Alan Horowitz, Whitefish Bay, September 10, 2017