Is the streetcar any more energy-efficient than the bus?

Many claim that the streetcar, running on steel wheels against smooth rails, is inherently more energy efficient than the bus, running on squishy pneumatic tires along rough pavement.  Both have to run mixed with traffic, expending more energy stopping and starting at intersections and with traffic. 
Electricity, considered a clean panacea by many, does all of its polluting at the plant.  Right now we get 86% of our electricity from boiling water to make steam to turn a turbine, a 30% efficient reaction.  While much of this is nuclear, we still get 65% of our electricity from burning good old carbon.

To compare energy for buses and streetcars, I had to compare a mostly diesel- and gas-fueled fleet of buses with an entirely electricity-fueled fleet of streetcars.  This is easy enough to compare in terms of “Gallon-Gasoline Equivalent” (GGE) or British Thermal Units (BTU).  These conversion factors allow us to look at different kinds of fuels and Kilowatt-hours in a straight comparison.  The trick is that electricity is not just about the energy used by the streetcar, but by the power plant.  As the FIVE major cities with streetcars have installed bases of large coal power plants, I had to go ahead and multiply the Kilowatt Hours (kWH) reported for streetcars by 3.33 to approximate the real energy use of steam generated electricity.

Average, Maximum and Minimum Energy Efficiency of 802 Bus and 5 Streetcar transit systems, 1997-2014

The above graph shows Passenger Miles per Gallon (PMPG), rather than “Gallons Gasoline equivalent per trip, or British Thermal Units per Trip, because PMPG is more relatable to the driving American that is likely to be reading this.  Notice that the average energy efficiency of either of these modes, about 40 PMPG, is very comparable to a 25 MPG fleet average car with 1.6 average occupancy in traffic.  I have to look into whether the high performing bus and streetcar systems are because of high passenger occupancy or highly efficient vehicles.  I suspect the former.

I have found data on the state-by-state mix of “renewable” versus “consumable” energy, but only for 2009.  I would want to find this data for the 1997-2014 period to make a reasonable comparison of energy use.  Again, there is nothing magic about electricity.  It is just burning fuel somewhere else.

Screen Shot 2016-03-15 at 7.52.24 AM
^ Click to Play

The reason this took two weeks to write was that I had to get all the yearly reports for energy use together, convert their units for different fuels into a single until of energy , and then put all that together into a time-series spreadsheet*.  Mostly working out the formula and filling the column, but every years data and presentation format was different.  This is what happens when you give a staff of highly educated bureaucrats a full year of dull accounting of over two thousand different transit systems.  They monkey with the presentation, just to keep themselves awake. I noticed, much to our chagrin, that the NTD** energy reports for 2009, 2010, and 2011 are 1,000 times less than they report.  There is a header that reports (in 1,000s) on top of each.  I made the executive decision to correct this in my data, even though it wasn’t what was reported from NTD.

* All while working a job during the day and my PhD research at night.
** National Transit Database