Supporting Information. For. Evaluating the Potential of Platooning in. Lowering the Required Performance Metrics of

Transcription

1 Supporting Information For Evaluating the Potential of Platooning in Lowering the Required Performance Metrics of Li-ion Batteries to Enable Practical Electric Semi-Trucks Matthew Guttenberg, Shashank Sripad, and Venkatasubramanian Viswanathan Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania

2 Platoon configuration The configuration of the vehicles involved in the platoon is a degree of freedom that impacts the overall efficiency. A computational fluid dynamics study 1 explored various configurations of medium-duty trucks in a platoon to determine the optimal configuration for overall platoon efficiency. Based on the analysis, the study showed that one truck behind another in a single lane yields the greatest reduction in the average drag coefficient of about 23%. 1 Validation of Platoon Drag Reduction Model The PATH project study was performed using models of vans rather than semi-trucks and this raises the question of whether these results will still hold for semi-trucks. In order to validate the data from the PATH project, we compare the predicted aerodynamic drag utilizing the data from the study 1 against other studies, 2,3 as summarized in Figure 1. The benchmarking cases include (i) a study by NREL based on wind tunnel experiments to determine the drag of a platoon of two trucks relative to a single truck, 2 and (ii) an experimental study exploring cab shapes comparing aerodynamics of a clean cab relative to an aerodynamically dirty cab. 3 For the cases considered, an average reduction in the coefficient of drag was determined meanwhile our model was used to predict the reduction in the average coefficient of drag using the same set up for each of the cases. The results are shown in Figure 1 and the dotted line represents a perfect agreement between model and data. This figure shows that, in general, the model matched the NREL data 2 much better than the data from Hammache et al. 3 with an MAE and RMSE of and respectively compared to an MAE of and an RMSE of These values approximate the uncertainty for the potential drag coefficient reductions and show that the model can be used to reasonably predict the average reduction in the drag coefficient for semi-truck platoons. 2

3 C d, reduction [Data from Ref (14,15)] Hammache (15) NREL (14) MAE Ref(14) = RMSE Ref(14) = MAE Ref(15) = RMSE Ref(15) = C d, reduction [Present Model] Figure 1: A comparison of the average drag coefficient reduction obtained from wind tunnel tests conducted by NREL 2 and Hammache et al., 3 (C d,reduction [Data]), and the average drag coefficient reduction predicted by our model, (C d,reduction [Present Model]), given the same initial parameters as each of the cases performed by the respective wind tunnel experiments. The dashed line represents the points where the predicted value exactly matches that of the corresponding experiment thus the farther away a value is from the line, the more inaccurate the prediction from the model is. Platoon Energy Figure 2 shows the energy required to overcome the drag force relative to the total energy is plotted against the root mean square velocity and the number of trucks in the platoon. Figure 2 shows that for a single vehicle, increasing the traveling speed from 40 to 60 mph causes an increase in the fraction of the drag energy with respect to the total energy from 0.22 to 0.4 while in a platoon of seven, the fraction would go from only 0.12 to Thus being in a platoon can significantly decrease the energy required to overcome the drag force. 3

4 Figure 2: Fraction of energy needed to overcome drag forces relative to the total energy required to overcome all of the forces acting on the truck as a function of the root mean square velocity (mph) and the number of trucks in the platoon. Since the drag force increases with an increase in velocity of the vehicle relative to the wind, we can see that at high velocities of about 60 mph, up to 40% of the total energy is spent on overcoming the drag force. These estimates provide insight into the reason behind the effectiveness of platooning in increasing the energy efficiency of a semi-truck. Uncertainty in Battery Pack Size Figure 3 shows the decrease in battery pack size represented in a different format where three ranges were specified to make it a function of only the number of trucks in the platoon. In addition, the error bars in Figure 3 provide additional bounds of uncertainty based upon the standard deviation resulting from the Monte Carlo simulations. 4

5 Figure 3: Average pack size required by a vehicle in the platoon measured in kwh as a function of the number of vehicles in the platoon for given ranges. The error bars represents one standard deviation from the mean value. The pack size also reflects the improvements in energy efficiency of trucks with platooning. Battery Swapping and Autonomous Driving While the overall benefits of platooning have been shown, we would like to point out that battery swapping and autonomous driving still have the potential to further improve upon these results. The cost of the battery pack significantly increases with the driving range of the truck, thus in order to make long-haul electric semi-trucks commercially feasible, it would have to be designed with a driving range that is well below the technical feasibility limit as well as the driving range of diesel trucks. One of the proposed approaches to solve the issue of commercial feasibility and limited range is to employ battery swapping. Battery swapping is an infrastructure strategy 4 in which charging stations for electric vehicles would also facilitate the swapping of depleted battery packs with fully charged packs thereby providing a way to combat the limited range issue, however, it requires massive infrastructure and tech- 5

6 nological development in order for the strategy to be effectively implemented. 4 Furthermore, the energy and power requirements of the vehicle strongly depend on the driving velocity, a parameter that autonomous driving has the potential to significantly alter. However, an analysis of changes in driving patterns is well beyond the scope of the present study and will be tackled in the future. Abbreviations PATH: Partners for Advanced Transit and Highway NREL: National Renewable Energy Laboratory MAE: Mean Absolute Error RMSE: Root Mean Square Error References (1) Vegendla, P.; Sofu, T.; Saha, R.; Kumar, M. M.; Hwang, L.-K. Investigation of Aerodynamic Influence on Truck Platooning. SAE Tech. Pap. 2015, (2) Salari, K. DOEs Effort to Improve Heavy Vehicle Fuel Efficiency through Improved Aerodynamics _o_web.pdf, 2016; Online Article, Accessed Date: 20-September (3) Hammache, M.; Michaelian, M.; Browand, F. Aerodynamic forces on truck models, including two trucks in tandem. SAE Tech. Pap. 2002, (4) Mak, H.-Y.; Rong, Y.; Shen, Z.-J. M. Infrastructure planning for electric vehicles with battery swapping. Manage Sci. 2013, 59,