Truck Aerodynamics Observation Study

A Small Study of Trucking Companies Running Aerodynamic Tractor Trailers

Study Design

 

Over the course of one year – from May 2012 through May 2013 – I methodically watched and collected data on trucks during my morning and evening daily weekday commute through New Jersey.  My 50 mile one way commute was was primarily in the truck lanes of the New Jersey Turnpike (Exit 7 to Exit 10 and back – 35 miles for each time I drove that segment) and to a lesser extent the western portion of Interstate 195 through the center of the state (seven miles each leg), and shorter segments of a few other roads (eight miles more each way).

 

My study was not conducted by just randomly recording the names of trucking companies whose trucks contained aerodynamic devices, but by applying a fairly stringent set of rules to my data collection.  These rules were:

 

• I only counted tractor trailers pulling standard dry vans or reefers between 40 feet and 53 feet long. I avoided counting open flat beds, car carriers, bulk carriers, moving trailers, box trucks, 20 foot cargo containers, doubles in any configuration (road trains) and anything else that didn’t fit my narrow definition.
  • The primary reason for this was that the overwhelming number of vehicles with aerodynamic devices on them were tractors paired with either 40 or 53 foot trailers.
  • This eliminated the need to count hundreds of additional vehicles a day; everything from open top municipal waste vehicles, heavy duty dump trucks, large box trucks, numerous moving vans, tractors pulling 20 foot cargo boxes, curtainsides, and other specialty trucks. Counting probably wouldn’t have been possible given that I also needed to be operating my own vehicle at the same time.
• Counting started from the moment I pulled out of the driveway to the moment I pulled into my work parking space, and vice versa.
• The same two radio stations (WNYC FM 93.9 – Greatest Radio Station On Earth and WBBR AM 1130 Bloomberg Radio – Best Financial Radio Ever) were on every day so my level of concentration, while not total, was fairly consistent.
• No eating, shaving, or looking at women on the ride, only staring at trucks.
• Trucks in my field of view were counted, and those with aerodynamic devices (primarily trailer side skirts) were counted as a subset of the total.
  • Counting was only initiated in my field of view, but turning around and rearview mirror counting was done if a truck’s ID or aero status was uncertain or a cluster count was not completed.
  • While skirted and boat tailed trucks were easily spotted, the brands of those trucks were often not identifiable.
  • At first, trailer side skirts and boat tails were the only devices identified. Over time, I began to see under trailer fairings, which were probably there in lower concentration the whole time.
• Trucks were counted if they were on the active roadway, in both the north- and south- bound lanes, moving or stopped. Trucks spotted entering or exiting toll exchanges (Turnpike exits), were counted. Finally, trucks not yet off the roadway or almost on the roadway entering or exiting rest stops were counted. Trucks within rest stops were off limits, no matter how few or many, skirts or not.
• Names were spoken into a voice transcription system; first my HTC EVO 4G and later my Samsung Galaxy S4 (running either Google Voice Recognition, Swype Voice Recognition, Samsung S-Voice, and/or ???) or Nuance Dragon Naturally Speaking on a laptop during the ride and corrected at home in the evening.
• While initial specific company sightings were recorded, company prevalence totals were estimated by memory and not by counts; full counts proved too heavy a lift given the process I set up.
  • One definite side effect of this choice is that for the less common companies in my ranking system (rank three out of five and below), skirt use ranking is unlikely to be accurate. For example, while I may have affirmatively identified that “Joe Rigger Trucking” is running skirts on the first truck I see, when I finally sat down to estimate what percent of Joe Rigger’s trucks were running skirts during the course of the study, I not only won’t remember that I saw a dozen Joe Rigger Trucks in total (or the same Joe Rigger Truck 12 times, or something in between), but I won’t remember specifically if one, two, or three of those trucks were running skirts. I might remember if EVERY truck had skirts, but otherwise, my number is an estimate based on sparse mental cataloging. As a result, many of the less common companies have trailer skirt rankings in the middle (3), which represents a decent guess, “n/a”, “I don’t know” or “I don’t remember because I don’t have a photographic memory”.
• Additional information collected on companies was primarily done via Internet searches (Google), along with some company phone calls. Hank’s Truck Forum and Hank’s Truck Pictures were two extremely useful websites for identifying and correcting company names . And I expect that visitors to this site will continue to improve company identities, as well.

 

Study Caveats (in addition to the above)

 

In spite of the relatively organized system set up to collect this data, during the course of the study it became clear that a number of systemic issues affected the data collection.  Because these issues persisted consistently throughout the study, and the study was conducted consistently through the entire year, we believe that errors did not affect the overall trends identified in the study, nor the specific companies or technologies identified.  Some of these study design issues follow:

 

• While I my did best to count just tractor trailer combos within my size range, I know that I missed some and occasionally counted wrong sized ones.
• Depending on the section of road, trucks may have been harder to spot in some sections. This resulted in artificially low aerodynamic modification counts; however, given they were consistently counted in this manner, relative changes in skirt ratios remain valid.
  • For example, Jersey barriers (those concrete dividers between opposing traffic) separate north- and south-bound lanes on the New Jersey Turnpike, resulting in a small but consistent missed side skirts and under trailer devices on trailers.
• Although my normal time on the road was 8:15 am to 9:45 am and 5:45 pm to 6:45 pm, daily variations could easily be a half an hour either earlier or later at least once a week , and perhaps an hour or more once or twice a month. Most cataloging and counts were within the stated window, but not all.
  • Time effects are multitudinous. Drivers and/or companies with particular types of equipment may decided to driver earlier or later, resulting in a perhaps biased view of the real mix of trucks on the road.
  • Some tolls in the region have peak and off peak rates (e.g. E-ZPass) which have changed at times over the past five years, potentially changing the types and/or volumes of trucks moving during particular periods of time, and perhaps in the middle of the study period.
  • More generally, events known and unknown to me can easily alter the volume, time and mix of traffic on the road at a given time. For example, accidents upstream, say on the New York Thruway, might change when a group of trucks arrives on the New Jersey Turnpike.  Also, if a port area road is closed, port container trucks travel times may be delayed, resulting in an increase or decrease in the port truck to regular truck ratio during my count.  This will impact not just the count but the ratio of skirted to non-skirted trucks since port trucks did not have trailer side skirts during my study.
  • When drivers break can affect road truck assortment. Breaks can be for rest, meals, legally mandated rest requirements, meeting friends, bathroom pitstops, health, boredom, and refueling.
  • There was always plenty of light to count trucks in the morning, but evenings were impaired or lost from between October 1st through March 1st when sunset times are 6:30pm or earlier.
  • Time effects can also apply to the day of week, since certain types of deliveries are done on certain days e.g. stocking up for weekends on Fridays or running replenishment runs on Mondays.
  • These same phenomenon could easily extend to multi-week or monthly phenomenon, such as monthly reorders or new product releases.
  • Finally, normal seasonal variations would certainly occur such as outdoor furniture deliveries being more common in the spring, Christmas increasing volumes in the fall, etc.
  • I only cataloged and counted during weekdays, and did NOT make the run on on state holidays, federal holidays, weekends, when I was sick, and during vacations. Not counting during weekends leaves out a crucial driving period which may have altered what I observed and counted.
• Epic New Jersey Turnpike lane widening began in 2012 and continued through 2015. The ever changing lane configurations, closures, equipment mobilizations, and new overpass occlusions certainly changed my counts in direct ways as well as ways unbeknownst to me e.g. a lane realignment unknowingly changing my viewing angle.
• More traffic in my travel lanes during my commute didn’t just slow me down but lead to a higher residence time on the road, increasing the number of potential truck sightings versus an average commute day. Higher than average traffic in the opposite lane may also result in additional aero outfitted truck sightings. Traffic in either direction may alter the visibility of trucks for better or worse depending on speed, lighting, weather, segment of commute, etc.
• Trucks two abreast may mask correct totals, and certainly masked the identification of aerodynamic devices. More than two abreast would have been unlikely because the separate truck lanes along my commute were only two lanes wide during the study, and when trucks were forced onto the “cars only” lanes,  trucks are prohibited from the left lane.
• Regarding lane closures, either the car or truck lanes were closed on occasion during commuting hours, forcing all traffic onto the remaining lanes. While I am nearly certain I never conducted a count study during these periods of time, it bears mentioning as a potential factor for projects involving traffic counts.
• I know that I occasionally double counted trucks within a day because many trucks return during the same day, allowing me to pass them a second time. In addition, even during the same trip double counts might have occurred because of occasional side-of-road or rest-stop stops either on the part of the trucks or myself.
• My direction of travel, south in morning and north in the evening, meant that the vast majority of trucks I counted were only on one side – northbound trucks in the morning and southbound trucks in the evening.  Thus my commute, combined with variable direction of travel by trucks based on time of day travel, haul type, etc., may have affected the count totals and aerodynamic
• Where the country was in the economic cycle will have additional effects and as you’ll see below, I perhaps found it reflected in my truck counts.
• Consumer preferences for products of different sizes, shapes, weights, composition and countries of origin sets the mix and number of trucks on the road independent of the inning of an economic recovery.  Evolution of consumer preferences here; and country of origin based on labor cost, quality, and resource availability; will also change truck mix over time as well.
• Regional geographic variation exists everywhere, but regional consumption evolution can change what is shipped within a region over time. For example, millennials are drinking more wine and liquor than Gen X’ers, changing what kind of alcohol is shipped, and where.
• My ability to count trucks, and spot ones with aero modifications improved over the course of the study
  • By the end of the study, not only was I able to ID truck brands out of the corner of my eye, I got better at identifying less obvious aerodynamic modifications like under trailer fairings.
• Goods movement business models were almost entirely ignored, resulting in some guaranteed errors in the data collection.
  • For example, both companies and individual tractor owners tow trailers for other companies. As such, the tractor label will often not match the trailer brand. For reasons mentioned elsewhere, I could seldom document the mismatch, and mention it only occasionally in my spreadsheet database of sightings.
  • Trailers are often leased, and since many of my identifications are based on branding printed on trailers, I often ended up identifying trailer leasing companies and occasionally even trailer manufacturers and not the company actually towing the box around.
  • Older tractors and trailers are sold out of fleets and often poorly debranded, allowing the old brand to still be identified.
• Some companies and drivers choose different roads for different reasons.
  • Load contents may determine on which roads certain goods are even legal to transport with such dangerous loads being confined to certain roads or transportation times.
  • Tolls avoidance may be an official corporate business decision, or a personal decision of a contracted or independent driver; as toll prices rise, these avoidance decisions will increase.
  • Routes may also be chosen for scenic reasons, family proximity, access to a favorite gentlemen’s club or male revue and a multitude of other reasons
• Branding, logos and identification is severely lacking on smaller company tractors and trailers, resulting in a bias toward the identification of larger company rolling stock. Swift and Roadway are impossible NOT to see because of the brand emblazoned across the trailer, while a six inch cursive identification on the door of a tractor is often difficult to see even when right next to it and thus impossible to ID when passing in the opposite lane, while wearing sunglasses, peering over Jersey barriers, and commuting on a road undergoing construction for the study duration.
  • This bias results in the hundreds of smaller companies and independent operators being undercounted relative to the larger companies. As you will see in my conclusion below, I was only able to document a third of the trucks outfitted with aerodynamic devices, with the remaining two thirds of trucks generally having small brand lettering and unlabeled trailers. This may also lead to a country bias with the less well branded pairs more likely to be local (Pennsylvania, New Jersey, New York, etc.) companies than Canadian companies.
• Since the study route is along THE largest and most important trucking corridor in my state and is directly associated with one of the countries largest ports, I see a much larger share of container towing rigs than on an average road in the United States. Because the highly integrated goods movement system for ports – material aggregation, loading shipping, unloading and distribution systems – are slow to change due to complexity, tradition, inertia, and the high capital costs for changing those systems, fuel efficiency improvements to the surface transportation (trucking and rail) portion of that system has not been a focus over the last couple decades. As a result, port  organizations and their affiliated companies appear to have made little outward change to their systems, whether those changes be aerodynamic improvements to chassis or trailers, or the purchase or contracting of high efficiency tractors.
• While each of the data tables I compiled will have additional background, caveats, legends and the like, the study should probably not be held up to strict statistical standards. One man collecting data with a few rules does not a rigorous scientific study make.  Don’t expect perfect adherence to the rules of observation, level of measurement (i.e. nominal / discrete, ordinal, interval, ratio, continuous) consistency, and other rules of formal science.

 

Study Data, Results & Conclusions

 

Along with a growing body of knowledge of the goods transportation industry, four data sets were collected during the course of the study.

 

First, in the first month and every few months subsequently, I chose a day to count every truck on the road and every aero device equipped truck, on one or both legs of my commute.  If I had counted just one leg, I doubled the number for a full day number.  At the end of the study, I put the five or so sets of numbers into table Table 1: Overall Truck Count and Skirted Truck Count seen below and via the link.  Noting that the numbers roughly fit a consistent rise in the total number of trucks on the road, and an even faster rising percent of trucks using aerodynamic devices, I created a percent rise in both series that fit well. Then I saved over my original data not realizing how valuable it would be later in this project. Darn >_<. I recounted in March of 2014 to check what had changed and found that the rise in the number of trucks on the road had slowed considerably (or had bounced around…I have no way of knowing for sure) but skirted trucks as a percent of all trucks had climbed.  Future updates will be available on the site at some point.

 

Month	Year	Trucks Daily	Aero Mod Trucks Daily	Aero Mod Trucks as a Percent of Daily Total	Fitted Data	Estimated trucks monthly	Estimated Aero Mod Trucks Monthly
May	2012	250	2.5	1.00%	Fitted	4500	45
June	2012	269	4	1.49%	Fitted	4842	72
July	2012	290	4	1.38%	Fitted	5220	72
August	2012	312	5	1.60%	Fitted	5616	90
September	2012	336	6	1.79%	Fitted	6048	108
October	2012	362	8	2.21%	Fitted	6516	144
November	2012	390	9	2.31%	Fitted	7020	162
December	2012	420	11	2.62%	Fitted	7560	198
January	2013	453	13	2.87%	Fitted	8154	234
February	2013	487	19	3.90%	Fitted	8766	342
March	2013	525	24	4.57%	Fitted	9450	432
April	2013	565	27	4.78%	Fitted	10170	486
May	2013	609	35	5.75%	Fitted	10962	630

 

 

I gleaned a few other nuts from the data:

• The number of trucks on the road continued to rise, with the the recovery from the recession one probable reason.
• Trailer side skirts and other aerodynamic devices are fitted onto an increasing percent of trailers, not just total number of trailers.
• Not only did I not maintain my original counts, but I did not track of the time of day, day of the week, and time of month my total and aero device count spot checks. Many of the caveats listed above apply to the data in Table 1: Overall Truck Count and Skirted Truck Count since only one count was done each month.

 

Companies using trucks with various streamlining devices are listed in Table 2: List of Trucking Companies Running Aerodynamic Tractor Trailers. Some of the findings from the data collected include:

 

• Out of the approximately 100,000 trucks that I laid my eyes on during the course of the study, 234 specific companies were identified using one or more aerodynamic device. If this represents only a third of the total trucks using aero modifications, which is approximately the number I felt I identified correctly, then up to 476 more companies may have been using these technologies during the course of the study; however, the number is certainly lower than this accounting for the number of repeat appearances by the same companies, or even the same rolling stock from those companies.
• Around six companies during the course of the study – Swift Transportation, Prime Inc., Schneider National, JB Hunt, Celadon Trucking, and Werner Enterprises – probably accounted for 50 percent of all trailer side skirts on the road, with the top ten probably accounting for two thirds of the skirts on the road.
• Thirty six (36) of the identified companies on the list are publicly traded.
• Thirty six (36) states are represented, with Texas (TX), Tennessee (TN) and Illinois (IL) the top three states for trucking companies with aerodynamic modifications.
• Canadian companies represent 37 of 234 (16 percent or so) of aerodynamically outfitted truck companies identified during the study which is remarkable enough given that our busy corridor is filled with trucks from over 70 percent of states in the United States of America.  As a percent of ALL trailer side skirts, gap reducers and boat tails on the road, the Canadian ratio is much, much higher than American companies, with possibly 50 percent of all Canadian trucks using trailer side skirts or other modifications.  This makes sense as Canada has been much more progressive in adopting energy efficiency measures across industries.
• Five foreign companies and three government companies were also identified, and thirty-one (31) companies remained unidentified.
• After the on the road observation ended, it was found that 32 company websites display a picture of a truck with one or more aerodynamic devices on their homepage, either as the primary image or in one or more slides.
• Twenty-eight (28) companies were not primarily “transportation companies”. That statement can be a bit hard to pin down due to the various business models that exist in trucking. For example, how does one determine whether an importer that specializes in transporting its own items but also transports other similar items under contract, and gets paid for transportation service, it a transporter or wholesaler?  What percent of its own items would a company need to transport before it was NOT considered a “transportation” company? Rhetorical question for now, but interesting.

 

As the study progressed, it became clear that the money saving and fuel saving devices I was observing were only the tip of the iceberg in terms of types and makers of those devices; Table 3: List of Tractor Trailer and other Heavy Vehicle Energy Efficiency Products and Companies is the result of not only the empirical observation of equipment in the real world, but extensive Internet research.  Some of the results of this data collecting includes:

 

• Any early version of this table contained 150 primarily aerodynamic products and devices; it currently contains 192 products (October 2015).  It has become clear that the variety of devices on the market is large and growing. It is expected that this list will grow and categories will be added.
• Approximately 93 devices fulfill EPA Smartway certification in Smartway’s Aerodynamics categories (as of October 2015), and by extension California Air Resources Board mandates.  Fourteen (14) additional devices may also fulfill Smartway mandates; clarification is pending communication with EPA.
• Perhaps because of size or the necessity of local installation, the majority of the products identified are made by North American companies in the United States or Canada.  In addition to the Canadian companies, six foreign companies do manufacture some of the identified products.
• Over twenty types of products were identified during the initial investigation (June 2014), but it is expected that this number will rise significantly.
• Eight publicly traded companies manufacture these products, on US or foreign stock exchanges.
• Many technologies identified in the list are not identifiable unless the stationary semi tractor trailer is examined closely e.g. low rolling resistance tires looks just like normal tires straight on.  Only at an angle can one see that one fat tire replaces two skinny ones.
• A few technologies being used overseas, such as teardrop trailers, are entirely absent from my New Jersey study corridor and probably from the United States.

 

Table 4: List of Common Trucking Companies NOT Running Aerodynamic Tractor Trailers was compiled to understand if there was any insight to be gleaned from observing trucks not using fuel saving tools. I found:

 

• A majority of the trucks not running aerodynamic trucks are pulling shipping containers associated for one or more of the Port of New York and New Jersey terminals – Port Newark-Elizabeth Marine Terminal, the Howland Hook Marine Terminal, Port Jersey Marine Terminal and Red Hook Marine Terminal, account for 60 percent of this group (by my rough estimate).
• Interestingly, a much higher percentage of the most common non-aerodynamic rig companies are publicly traded, with over 40 percent of this group being traded in foreign or North American exchanges.
• Almost fifty percent are foreign owned or run, which makes sense given the longer histories of foreign (particularly European) container shipping companies; often with roots in government entities, city states, and/or old, powerful families. These organizations are (or were) large and rich, allowing them to create globe girdling systems of ports, boats and other transportation infrastructure.  Turning these systems, which have become many times larger and more complex in the decades (or centuries) since their creation, requires massive amounts of capital and coordination to change – insert metaphor for turning a tanker ship here!

 

Since this study was conducted, aerodynamic device adoption has continued, increasing not only the percent of companies using aerodynamic equipment, but the list of aerodynamics users now includes some of the most commonly sighted companies previously not running aerodynamic equipment   Also since the study concluded, I have observed slowly changing ratios of companies on the road, perhaps due to the relative competitiveness of companies and a recovering and evolving economy. As I publish new documents and data, and make minor edits to the existing study, those changes and updates will be noted or linked to on the Fuel Efficient Truck Origin Story page.

Minor updates May 14, 2016; Published on September 20, 2015