June 2015

MEMORANDUM

SUBJECT:	Possible Tractor, Trailer, and Vocational Vehicle Standards Derived from Alternative Road Grade Profiles

FROM:	Lauren Steele, Houshun Zhang, OTAQ/ASD

TO:	Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles - Phase 2 - Docket EPA-HQ-OAR-2014-0827

Introduction

The purpose of this memorandum is to illustrate some possible effects on the calculated fuel consumption and CO2 emissions of vehicles caused by variations in the test procedure for simulating road grade in the Greenhouse gas Emissions Model (GEM).  This memorandum presents possible road grade profiles other than the one presented in the proposal, as well as an analysis of exploratory work  using the Phase 2 GEM with alternate road grades.  The road grade profiles described below include two that were developed by the Department of Energy's National Renewable Energy Laboratory (NREL) and one that was developed by the agencies.  For completeness the proposed profile that is in Phase 2 GEM is described as well. 

It is a goal of the agencies to select the most appropriate road grade profile(s) on the basis of being nationally representative as well as reasonably similar to real-world driving conditions.  It is not the agencies' intent to cause the standards to become more or less stringent as a result of selection of road grade profile.  Any effect on stringency is minimized by the fact that both the baseline case and the stringency case will be simulated in GEM with the same road grade profile.  However, it is expected that various engine and driveline technologies will perform differently over different road grade profiles, thus changing the resulting numerical results in all cases.

The agencies acknowledge that if a road grade profile is very aggressive, some vehicles may not be able to follow the full trace at speed.  In the preamble to this rulemaking we request comment on the merits of developing variable speed highway cycles.  Our goal is to adopt a cycle that is challenging enough to ensure that vehicles continue to be built for real world performance, while maintaining the simulated vehicle operation within a range for which GEM has been validated. 

Proposed Road Grade Profile

The agencies are proposing an interim road grade profile for development of the proposed standards, as shown below in Figure 1, in both the 55 mph and 65 mph cycles.  The grade profile was developed by Southwest Research Institute (SwRI) on a 12.5 mile stretch of restricted-access highway during on-road tests conducted for EPA's validation of the Phase 2 version of GEM.  The minimum grade in the interim profile is -2.1 percent and the maximum grade is 2.4 percent.  The profile includes 30 percent of the distance at grades of +/- 0.5 percent.  Overall, the profile includes approximately 50 percent of the time in relatively flat terrain with road gradients of less than 1 percent.

                                       
   Figure 1:  Proposed Road Grade Profile for 55 mph and 65 mph Drive Cycles

The following charts illustrate the degree to which the SwRI road grade profile in GEM matches the nationally representative profile, as characterized by NREL.  In Figure 2, the GEM profile has generally lower magnitude grades than the national data, with a lack of grades between 2 and 4%.  In Figure 3 the majority of half hills in the GEM profile are of length 0.5 to 1.5 miles, with a lack of half hills less than 0.5 mi and longer than 1.5 miles.

                                       
Figure 2:  Comparison of National Cumulative Absolute Road Grade with SwRI Route (Source: NREL)

                                       
Figure 3:  Comparison of National Cumulative Half Hill Distance with SwRI Route (Source: NREL)

NREL Road Grade Profiles 

A complete description of the methodology for generating the NREL road grade profiles, as well as several other possible alternate profiles, is found in a May 2015 report titled, "Development of Road Grade Profiles Representative of US Controlled Access Highways," available in the docket.  Numerical representations of these distance-based profiles are also available in the docket, in a format that is ready for modeling.

The minimum grade in the example 55 mph NREL profile shown in Figure 4 is -3.1 percent and the maximum grade is 3.8 percent.  The majority of the profile is within a range of grades of +/- 2 percent.  Of the five nationally representative 55 mph profiles generated by NREL in the final report, this one was selected as the example because the shorter length of the synthetic profile (compared to the relatively long local routes) has many advantages.  Also, the road grade is steeper than in the other synthetic profile.

The minimum grade in the example 65 mph NREL profile shown in Figure 5 is -2.2 percent and the maximum grade is 3.3 percent.  The majority of the profile is within a range of grades of +/- 2 percent.  Of the four nationally representative 65 mph profiles generated by NREL in the final report, this one was selected as the example because the shorter length of the synthetic profile (compared to the relatively long local routes) has many advantages.  Also, the length of distance spent in a steep down-slope is less than in the other synthetic profile.
                                       
          Figure 4: Example NREL road grade profile for 55 mph cycle

                                       
          Figure 5:  Example NREL road grade profile for 65 mph cycle


EPA Road Grade Profile 

The profile depicted in Figure 6 is 11.48 miles (18.44 km) long, its longest half hill spans 2.87 miles (4.47 km) and its highest grade equals 2.5%.  It was developed to align with the national road grade data using a methodology differing from NREL's and may offer some advantages. Of several profiles developed by EPA, this one was selected for analysis in part due to its aggressive uphill. 

A complete description of the methodology for generating the EPA road grade profile is found in a May 2015 memorandum titled, "Development of an Alternative, Nationally Representative, Activity Weighted Road Grade Profile for Use in EPA GHG Certification of Medium- and Heavy-Duty Vehicles," available in the docket.

                                       
       Figure 6:  Example EPA road grade profile for both cruise cycles


Exploratory Work Analyzing the Effect in GEM of Alternative Road Grade Profiles

The agency has conducted exploratory work to evaluate the effects on vehicle performance as simulated in GEM of applying different road grade profiles. The objective of this analysis is to understand how the default transmission auto-shift responds to demands to move heavy vehicles over different road grades.  This auto-shift is being used for the stringency development as well as the future compliance tool under the Phase 2 rulemaking.[,]  The results are inconclusive, but do offer insights as to how GEM could be further improved to represent more realistic real-world driving conditions.  

For this analysis, three profiles were examined: (1) the EPA profile described above and presented in Figure 6 (Profile 1); an amplified version of Profile 1 referred to as Profile 2, and (3) the proposed profile described above and presented in Figure 1 ("Baseline").  Figures 7 and 8 show the road grades and elevations, respectively, of these three profiles explored in this study.  


                                       
                Figure 7 Alternate Road Grade Profiles Explored


                                       
                Figure 8 Alternate Elevation Profiles Explored
                                       
It can be seen that Profiles 1 and 2 are similar in term of shape, except that Profile 2 has a higher road grade up to 4%.  In fact, Profile 2 was derived from Profile 1 by multiplying a constant to Profile 1 in order to achieve a maximum road grade of 4%.  This hypothetical road grade is mainly used to check how the auto-shifting algorithm responds with a road grade up to 4%, and was not designed to align with the national data.  The Baseline and Profile 1 have similar road grade in terms of maximum road grade, although elevation is quite different.  Compared to the Baseline, both Profiles 1 and 2 have a long hill, which can test how the default transmission auto-shift responds to this condition. 

In order to evaluate the impact of the road grade on the vehicle performance as well as check the fidelity of the default transmission auto-shift, a vehicle with the key parameters defined in Table 1 is selected for this study, which represents a class 8 tractor with specifications aligning with some technologies that could be adopted in the 2027 timeframe.

      Table 1 Class 8 Combination Tractor Sleeper Cab Modeling Parameters
                               Regulatory Class
                              Class 8 Combination

                            Sleeper Cab - High Roof
                                    Engine
                           455hp rating map for 2027
                            Transmission gear ratio
              [0 14.8 10.95 8.09 5.97 4.46 3.32 2.45 1.81 1.35 1]
                              Gearbox Efficiency
      98% for the gear with a 1:1 gear ratio, and 96% for all other gears
                                     CdA 
                                       4
                               Steer CRR (kg/t)
                                      4.3
                               Drive CRR (kg/t)
                                      4.5
                              Trailer CRR (kg/t)
                                      4.7
                          Axle Mechanical Efficiency
                                      96%
                               Total weight (kg)
                                     31978
                                  Axle ratio
                                     2.26
                               Number of Axles 
                                       5
                          Default Axle Configuration
                                      6x4
                        Electrical Accessory Power (W)
                                      300
                        Mechanical Accessory Power (W)
                                     1000
                      Environmental Air Temperature (°C)
                                      25
                                Payload (tons)
                                      19


Figures 9 and 10 below present the simulated results of the vehicle described in Table 1 over Profile 1 and a 65 mph drive cycle.
                                       
Figure 9 Vehicle speed traces, gear number, engine speed and torque over 65 mph cycle for Profile 1
                                       
       Figure 10 Engine operating points over 65 mph cycle for Profile 1

From Figure 9, it can be seen that the engine is under-powered during the long hill operation because the vehicle is unable to follow the targeted vehicle speed trace.  The transmission stays in the top gear. The engine operating points as shown in Figure 10 seem to make sense, suggesting that the auto-shifting is able to handle this road grade profile up to 2.5%.  

The situation changes when higher road grade with a maximum of 4% is evaluated.  Figure 11 and 12 below present the simulated results of the vehicle described in Table 1 over Profile 2 and a 65 mph drive cycle.

                                       
Figure 11 Vehicle speed traces, gear number, engine speed and torque over 65 mph cycle for Profile 2
                                       
                                       
       Figure 12 Engine operating points over 65 mph cycle for Profile 2
                                       
Figure 11 shows that the vehicle cannot meet the targeted speed trace due to the under-powered engine. When the vehicle climbs the hill, there is a long delay for 50 seconds before shifting to lower gear. This results in the engine operating points moving to the lower speed end along the full torque curve as indicated in Figure 12.  In reality, the truck running on the road should also have a minimum engine speed limit.  In contrast, the default auto-shifting algorithm currently allows the engine speed to drop below 900 rpm without shifting.  These behaviors indicate an opportunity to improve GEM for the final rulemaking.  Another key observation and area for improvement is how the engine behaves during climbing once the transmission shifts to a lower gear. As can be seen in Figure 11, the engine first drops the speed a little bit after 1520 seconds, and then keeps rising for more than 80 seconds without shifting.  This causes the engine to run along the peak torque curve while the speed rises, which is not a very efficient zone.  

Table 2 shows the tractor performance in grams of CO2 per ton-mile over the individual GEM drive cycles as well as the composite cycle with weightings for a sleeper cab tractor, using the three different highway cruise road grade profiles explored in this study.  No road grade was applied to the ARB Transient cycle.  As expected, the vehicle emits more CO2 when driving the more aggressive road grade using Profile 2.

Table 2 GEM performance (g CO2/ton-mi) of a hypothetical 2027 class 8 sleeper cab tractor over 3 road grade profiles
                                       
                                   Baseline
                                   Profile 1
                                   Profile 2
                                     55mph
                                     62.24
                                     58.63
                                     68.5
                                     65mph
                                     71.73
                                     67.13
                                     74.28
                                      ARB
                                    128.17
                                    128.17
                                    128.17
                                   Weighted
                                     73.69
                                     69.42
                                     76.45
                                       
                                       
This exploratory work illustrates areas for improvement in the calibration of the agency's transmission auto-shift as indicated by Figures 11 and 12, when high road grade is used for highway cruise drive cycles.  To attain the agencies' goal of representing real world driving, we believe we would need to make corrections on the GEM auto-shifting.

In the preamble, the agencies have asked stakeholders for comments, hopefully supported by testing data, to help us to calibrate our auto-shift model.  More importantly, EPA has developed a comprehensive testing plan to obtain first-hand testing data.  This plan includes the powertrain tests for both medium heavy-duty and heavy heavy-duty powertrains conducted at SwRI, and heavy heavy-duty powertrains with three transmission setups at Oak Ridge National Laboratory.   Up to eight different road grade profiles will be tested, and a large data set will be collected to see how the transmissions are shifted, thus recalibrating some model constants used in the auto-shifting table.  It is expected that a well-calibrated GEM would be able to address a wider range of operations including high road grade routes.
