Together, the route data and power train models determine the energy consumption in each simulation. Our methodology affords access to established data and extends prior vehicle propulsion energy and emissions analyses. In Phase I of the EVALUATE project, the team utilized a 5- cycle EPA dyno schedule and fuel economy label weighting protocol and has continued to do so for the two light-duty vehicles investigated herein. In addition, owing to the larger classes investigated in the present study, Phase II has also called for independently derived travel demands and data from the literature for representative use cases. For more details on the theory, model development, source data, and applications, please refer to [1, 14].As discussed, we developed physics-based power train models developed in [1, 14] to accommodate target vehicle technologies of interest. In each vehicle category, we established characteristics for the baseline vehicles . We then proceeded to develop an electrified power train model for each vehicle category. To facilitate a direct comparison among vehicles using dissimilar energy sources, we identify vehicle specifications for a given vehicle classification and hold constant key parameters such as vehicle power output, capacity to sustain the required torque and speed, vehicle footprint, passenger and cargo capacity, auxiliary power requirements, and so forth. Table 2-1 below depicts some of these operative specs. In developing our model and conducting simulations, growing tray we have made every effort to represent real-world vehicle characteristics across the categories of interest.
For example, when an electrified variant has a greater gross vehicle weight than a comparable internal combustion vehicle within its class, we reflect that vehicle weight difference in the analysis. This is especially important at larger vehicle classes because EVs in these classes have proportionately heavier batteries, which then incur additional energy consumption. In this way, we provide simulated comparisons based on actual vehicles in the marketplace.Five distinct driving cycles comprise EPA test and labeling protocols for light-duty passenger cars and pickup trucks. The three 23°C tests include a derivative of the Urban Dynamometer Driving Schedule known as the Federal Test Protocol , the high acceleration aggressive driving schedule identified as the Supplemental FTP , and the Highway Fuel Economy Driving Schedule . The 35°C drive cycle is the Air Conditioning Supplemental FTP driving schedule referred to as SC03. The -7°C cold weather test schedule repeats the original FTP at the reduced temperature. For the light-duty pickup and cargo van investigated in the study, we have adopted the EPA “5- cycle” protocol and created an approach whereby a weighted mix of driving schedules is obtained to approximate major modes . With the original development of the vehicle architecture models, and assumptions around the weighted driving cycle protocols, the team’s next step was to develop a MATLAB/Simulink code that generated a series of energy consumption values based on inputs of vehicle type and driving cycle. These intermediate outputs were then combined to generate effective fuel economy values, analogous to the EPA 5-cycle approach, for the stipulated categories . This was done and a set of energy consumption outputs were generated.
Using the example of the light-duty pickup truck and van, representative outputs are depicted in Table 2-2 and Table 2-3.The next step was to define representative driving schedules for a range of small business and fleet needs serving a large urban area, such as metro Atlanta. For the light-duty pickup and van, we follow the approach used in Phase I, where we select two urban commutes of 80.5 km and 32.2 and a suburban vehicle use case of 48.2 km . For urban commutes, presumably into and out of a city like Atlanta, it is reasonable to employ the EPA “combined” rating and protocol to determine energy consumption for these trips. For the suburban errand use case, it is reasonable to employ the EPA “city” rating and protocol. This is summarized in Table 2-6 below. Shown in Figure 2-2 is a notional depiction of a baseline vehicle’s instantaneous power and cumulative energy for an example drive cycle. Figure 2-3 depicts a few of the standardized EPA dyno schedules that are fed into a 5-cycle weighting determination.While EPA protocols have been applied for LDV cars, pickups, and vans, we have relied upon a combination of physical models and the literature to define driving schedules for moving trucks and refuse trucks. Section 2.4 provides a more comprehensive discussion of the driving schedules assumed in this study for the moving trucks and refuse trucks. In short, we have relied upon experimental field data, as reported in peer-reviewed literature. We have also corroborated these observations against our physics-based models, in areas such as weight, aerodynamic drag coefficient, frontal area, and rolling resistance.Constructing representative charging profiles for commercial fleet EVs is a distinct and, in many ways, a simpler exercise than doing the same for personal EVs.
In our Phase I report, we consulted four primary data sources to explore observed charging behavior for personal EVs from which we manufactured representative charging profiles: a synthetic dataset generated by a separate Georgia Tech research team evaluating the benefits and challenges of smart charging algorithms; the Georgia Power Electric Vehicle Rate scheme; a ChargePoint dashboard portal and database that has been aggregated for workplace charging on the Georgia Tech campus since 2015; and a verbal consultation with Escalent, a third party research firm. These independent data sources were corroborated and used to develop four representative charging profiles for personal EV charging events. The obvious difference between charging behavior for personal EVs and commercial EVs is the shift in emphasis from convenience charging to charging schedules constrained more severely in space and time by the demands of business. Fleet vehicles have operational obligations that must be fulfilled punctually and reliably. Commercial EV charging behavior is therefore primarily a function of business characteristics. To develop charging profiles for commercial EVs, we envisioned conceivable business use cases for each vehicle type included in Phase II . In order to improve our understanding of daily vehicle usage for these vehicle-application combinations, we reviewed U.S. national vehicle miles traveled data from the Vehicle Inventory and Use Survey. This revealed that daily VMT for LDV are less than 50 miles about 82% of the time, and less than 100 miles about 93% of the time. Similarly, daily VMT for MD in the class 3-6 range are less than 50 miles 68% of the time, and less than 100 miles about 84% of the time. Such empirical data was useful in developing realistic daily driving demands that are explored in this study. Light trucks and vans may most typically be deployed by residential service businesses such as lawn care, and pest control or by electricians and plumbers. It is assumed that the on-road driving cycles for these vocations are somewhat similar. As such, these business use cases were clustered into a representative category labeled “Residential Home Services.” Representative driving cycles were then used to calculate the combined energy consumption rates of light trucks and vans belonging to residential service fleets, as described in the previous section. Similarly, we reasoned local moving trucks would have a fairly consistent mixture of on-roaddriving cycles and produced a combined energy consumption rate for a composite medium duty truck, drying and curing bud as described in the previous section. We then used a distribution of vehicle miles traveled to calculate a series of cumulative energy consumption totals comparable to a diverse variety of business operations and scales. Modeled VMT figures were determined to be reasonable assumptions based on real-world operational data for service vans from NREL’s FLEET DNA, which publishes 29 days of dynamometer driving data from four service vans in operation in the United States as shown in Figure 2-4.A similar approach was employed for refuse trucks. Driving cycles for refuse trucks are less variable than for other vehicles, but distances traveled might vary more depending on service area and distance to the landfill and are generally greater.
We used higher vehicle miles traveled numbers for the calculation of total cumulative energy consumption for refuse trucks. It is in the nature of the businesses described by each use case that the vast majority of on-road activity occurs at predictable, well-defined times. From experience, “business hours” for residential home service businesses are very likely always during the day, between around 8 AM to 6 PM. This is similarly true for moving businesses, although with perhaps slightly less consistency to accommodate the occasional client requiring moving services outside of typical hours due to scheduling conflicts or other miscellaneous logistical reasons. Refuse truck service is even more regular, with well-defined service schedules and routes. Regardless, if business activity always or most often occurs during certain segments of the day, it leaves fewer segments of the day available for charging events. In terms of charging location, we reasoned further that most businesses with EV fleets would have two options for where to charge their vehicles: at the business home base or at a public EV charging station in the field.It was under these assumptions that five commercial EV charging profiles were developed. Three charging profiles are representative of a business that charges their EVs at their business’ “home base,” some central garage, parking lot, or depot where vehicles return at the end of each day and domicile overnight, with each of the three profiles having a different “after hours” start time for the charging event . Two charging profiles are representative of a business that charges vehicles during the workday using charging equipment in the field . To facilitate comparisons, we designed the charging profiles and only reported simulated outputs if the EV battery was recharged to 100%. We assumed businesses would tend to opt for the lowest charging level that fulfills charging demand within the time constraints of the business’ activity. For the main batch of simulations, Level 2 charging systems were used for light trucks and vans, and Level 3 charging systems were used for moving and refuse trucks. Level 2 was sufficient to completely recharge the batteries of light trucks and vans using every charging profile at all VMT levels, except for field charging starting at 3 PM at 50 and 100 miles traveled per day. Level 3 was sufficient to completely recharge the batteries in every scenario. The Level 2 charging profiles are shown in Figure 2-5 as examples, with on-times of 5PM, 8PM, 1AM for the charging that occurs at the location of the “Business,” and on-times of 12PM and 3PM for charging that occurs at locations in the “Field,” respectively. Please note that the time axis has been shifted with a start at 8AM to accommodate charging cycles that extend past midnight on a given day. The five charging profiles using Level 2 and Level 3 charging systems, four or three VMT levels depending on the use case, and four vehicle types enable the very broad applicability of our simulations and findings to many different business operations and scales.This second phase of research leveraged the extended grid modeling and optimization work described in our Phase I report. We used the same merit-order dispatch estimation framework based on actual data reported by the Southern Company Balancing Authority . These data are high-resolution and provide high detail of individual plants and generating units for all technologies. The methods used to develop the dispatch model and generate the marginal grid emissions assumptions are described in detail in [29]. With similar motives to Phase I, demonstrating the nuanced implications of emissions assumptions for quantifying abatement and comparing EVs to other vehicle technologies requires the development of a series of grid emissions assumptions.Monthly average and hourly emissions were assembled for August, October, and December. Hourly and marginal hourly emissions were collected for representative days from each of those months. Grid characteristics, including total load, demand curves, and to some extent available generation resources, evolve seasonally. In warm climates , demand for electrical power and subsequently grid emissions intensity are highest during the summer months and particularly in the evenings as people return home, turn on their lights, stoves, and televisions, and crank up their air conditioning units. The SERC grid experiences similar fluctuations in the winter months, and less extreme fluctuations in the shoulder months of spring and autumn. Using example emissions from summer, winter, and shoulder months affords the exploration of emissions variations within a 24-hour day at different times of the year and their effects on total cumulative emissions from EV charging events.