EVS24 Stavanger, U.S. Department of Energy – Advanced Vehicle Testing Activity: Plug-in Hybrid Electric Vehicle Testing and Demonstration Activities

The U.S. Department of Energy’s Advanced Vehicle Testing Activity tests plug-in hybrid electric vehicles (PHEV) in closed track, dynamometer, and on-road testing environments. The purpose of this testing is to determine the potential of PHEV technology to reduce petroleum consumption. It also allows documentation of PHEV driving and charging profiles and electric charging infrastructure requirements. As of March 2009, the Advanced Vehicle Testing Activity has initiated testing on 12 PHEV models from aftermarket conversion companies and original equipment manufacturers. In addition to performing controlled dynamometer and on-road testing, AVTA has collected in-use data from 155 PHEVs operating in 23 U.S. states and Canadian provinces. This fleet has demonstrated an average increase in cumulative fuel economy of 22 to 55% when in charge depleting mode, as compared to charge sustaining mode. Charge depleting range has varied from 32 to 64 miles, depending on the vehicle and battery pack. In ideal conditions, some vehicles have achieved monthly fuel economy results of 80 to 120 miles per gallon through frequent charging and less aggressive driving styles.


Introduction
The U.S. Department of Energy's (DOE) Advanced Vehicle Testing Activity (AVTA) tests plug-in hybrid electric vehicles (PHEV) in closed track, dynamometer, and on-road environments. The purpose of this work is to document the petroleum reduction potential of the PHEV concept.
This work includes evaluating vehicle energy consumption and efficiency, as well as plug-in charging infrastructure requirements and the impact of plug-in vehicles on the electrical grid.
AVTA provides advanced technology vehicle performance benchmark data for technology modelers, research and development programs, and technology goal setters. AVTA testing results also assist fleet managers in making informed vehicle purchase, deployment, and operating decisions. AVTA is part of DOE's Vehicle Technologies Program. AVTA's light duty PHEV testing activities are conducted by the Idaho National Laboratory and the Electric Transportation Engineering Corporation, with.
Argonne National Laboratory performing dynamometer testing.
AVTA regularly reports PHEV test results to industry stakeholders [1,2] and the general public via its website [3]. The purpose of this paper is to summarize PHEV testing and evaluation to date.

PHEV Testing Overview
As of March 2009, AVTA has initiated dynamometer, track, and controlled on-road testing of 12 PHEV models from original equipment manufacturers (OEM) and conversion companies. Eight PHEV conversion models have been evaluated in fleet usage.
The PHEV models tested are listed below, with battery pack or cell manufacturer listed separately, where appropriate. Note that when different battery chemistries or battery manufacturers are used by the same conversion company, they are considered different models due to potentially different battery and vehicle performance.

Baseline track testing
The first step in the AVTA's PHEV testing program is baseline performance testing conducted on a closed track. Test metrics measured include the following: Acceleration 0 to 60 mph Acceleration 1/4 mile and 1 mile Braking from 60 mph Fuel economy with and without the air conditioning on at fixed speeds (usually 45 and 60 mph) Plug-in battery pack charging time Vehicle coast down coefficients.

Baseline dynamometer testing
Following track testing, PHEVs are tested on a chassis dynamometer. Each PHEV undergoes a 5-day testing regime, which includes (at least) 26 Urban Dynamometer Driving Schedule (UDDS), Highway Fuel Economy Driving Schedule (HWFEDS), and US06 test cycles. Testing includes charge depleting (CD) and charge sustaining (CS) test cycles, hot and cold starts, and repetition of cycles with and without air conditioning.
For more information on AVTA test procedures, a comprehensive PHEV test manual can be found at the AVTA website [4].

Fleet demonstrations
In addition to controlled testing, AVTA monitors vehicle performance in uncontrolled fleet usage.   Vehicles are driven by fleet participants to perform a variety of missions. The majority of the vehicles in the fleet are driven for commercial use by more than one driver. Approximately 10% of the miles logged in the fleet were logged by vehicles driven for private use.
Vehicles are equipped with onboard controller area network (CAN) data loggers. These loggers capture time history data to monitor gasoline and electricity consumption and numerous other vehicle, charging, and environmental parameters. Automated database routines store and analyze these data. Summary metrics are generated to quantify energy consumption and characterize vehicle operation and charging behavior. Reports are sent to fleet participants monthly, and overall fleet summary reports are posted to the AVTA website [5]. Basic results from fleet data analysis are presented in this paper. More detailed analyses will be presented in future works.  Results shown in this section and Section 4 are from a Hymotion Prius with a production prototype, or "v1," battery pack. Figure 2 shows gasoline fuel consumption and electrical energy consumption of the Hymotion Prius in repeated UDDS cycles. The first three cycles show CD operation, starting with a cold start in the first cycle. The transition to CS mode takes place during the fourth cycle as the Hymotion battery pack depletes fully. Remaining cycles represent CS mode performance and demonstrate how cumulative gasoline fuel economy decreases as the distance driven in CS mode increases.

Hymotion Prius UDDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

Hymotion Prius HWFEDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

EnergyCS Prius Conversion with Valence Battery
EnergyCS converted a Toyota Prius to a PHEV by replacing the stock Prius NiMH pack with a 9 kWh Li-ion battery pack. The EnergyCS pack is charged by plugging into a standard 110 V outlet.
It also accepts regenerative braking and engine charging. The Toyota Prius with EnergyCS conversion pack (hereafter referred to as the EnergyCS Prius) is a blended-mode PHEV.
As previously mentioned, EnergyCS has produced packs with cells from multiple manufacturers. Test data shown in this section and Section 4 come from an EnergyCS Prius with Valence cells.

EnergyCS Prius UDSS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

EnergyCS Prius HWFEDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

Hymotion Escape Conversion with A123Systems Battery
Similar to the Hymotion Prius, the A123Systems Hymotion conversion pack for the Ford Escape Hybrid is a supplemental battery pack installed adjacent to the stock Ford 1.8 kWh NiMH pack.
The Hymotion pack for the Escape is a 12 kWh Liion pack, which is charged by plugging into a standard 110 V outlet. It does not accept regenerative braking or engine charging. The Ford Escape with Hymotion conversion (hereafter called the Hymotion Escape) operates in blended mode. Figure 6 shows gasoline and electricity consumption of the Hymotion Escape in repeated UDDS cycles.

Hymotion Escape UDDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

HybridsPlus Escape UDDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh

HybridsPlus Escape HWFEDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh Figure 9: Hybrids Plus Escape HWFEDS testing.

Electrovaya Escape Conversion with Electrovaya Battery
The Electrovaya PHEV conversion of a Ford Escape Hybrid involves replacement of the stock Ford Escape HEV battery. Electrovaya uses a 12 kWh Li-ion battery pack, which is charged by plugging into a standard 110 V outlet. The pack also accepts regenerative braking and engine charging. The Ford Escape Hybrid with Electrovaya conversion pack (hereafter referred to as the Electrovaya Escape) operates in blended mode. Figure 10 shows gasoline and electricity consumption of the Electrovaya Escape in repeated UDDS cycles. Figure 11 shows gasoline and electricity consumption of the Electrovaya Escape in repeated HWFEDS cycles.

Electrovaya Escape HWFEDS Testing
Cycle MPG Cumulative MPG Cumulative AC kWh Figure 11: Electrovaya Escape HWFEDS testing.

Renault Kangoo
The 2003 Renault Kangoo is a series PHEV, with true electric-only operation during CD mode. It has a 25 kWh Saft nickel-cadmium battery pack. When the pack is fully depleted, a 0.5 L, two-cylinder ICE runs to generate electrical energy for tractive power.     Table 5 shows on-road accelerated testing results for the EnergyCS Prius. The three 40-mile cycles exhibited low electrical energy consumption and fuel economy due to a pack failure.  Table 6 shows on-road accelerated testing results for the Hymotion Escape.  Table 7 shows on-road accelerated testing results for the Electrovaya Escape.  Table 8 shows on-road accelerated testing results for the Renault Kangoo. The two longest cycles were omitted due to problems with the vehicle's ICE and inverter.  Table 9 contains basic statistics for fleet data collected in 2008.

Hymotion Prius
As with on-road accelerated testing, the primary metrics used to summarize PHEV fleet performance are fuel consumption and electrical energy consumption. Results are broken down by operating mode (CD or CS) relative to trips taken. Table 10 shows 2008 results for the fleet of 102 Hymotion Priuses. Note that a CD trip is one that begins and ends in CD mode. A mixed trip begins in CD mode and ends in CS mode. A CS trip begins and ends in CS mode. The majority of data in this section comes from production (or "v2") packs, although some vehicles in the fleet had "v1" packs in early 2008.
Sixty-three of these vehicles have the V2Green data logger, which measures alternating current (AC) energy from the grid directly.
The remainder of Hymotion Priuses in the fleet have a data logger that measures direct current (DC) energy at the battery. For the latter group, AC energy was calculated by using an AC-to-DC efficiency of 78%, based on observed average charger efficiency during dynamometer.
The average CD range from Hymotion Prius fleet data was observed to be 31.6 mi. This is the distance driven in one or more trips, starting with a full state of charge in the Hymotion battery pack and ending when the pack is fully depleted. It is important to note that CD range is not synonymous with all-electric range for blended-mode PHEVs. The ICE may cycle on and off during CD mode to meet tractive power and accessory load demands, heat the catalytic converter, and for other reasons.

Driving and charging behavior as they affect vehicle performance
The Hymotion Prius and other conversion PHEVs tested to date exhibit a wide range of energy efficiency as driving, charging, and external conditions vary. Figure 12 expands on cumulative fuel economy metrics shown in Table 10 to demonstrate this variation in the Hymotion Prius. Similar trends are seen with the other blendedmode conversion PHEVs evaluated by AVTA. Figure 12 contains distributions of monthly vehicle fuel economy by operating mode. All vehicle months in 2008 when a vehicle drove more than 300 miles are included in these distributions. The tops and bottoms of the boxes represent the 75 th and 25 th percentiles, respectively. Center bars represent median values. Individual data points shown are statistically considered outliers. The single largest factor determining a PHEV's fuel efficiency is its operating mode, which is a function of distance driven relative to charging frequency and duration. Hymotion Prius fleet driving is broken down by operating mode as shown in Table 11. Note that nearly half of the trips taken began in CS mode, meaning the Hymotion pack had been fully depleted prior to the start of the trip. This high proportion of driving in CS mode greatly reduces overall vehicle fuel economy.
Additional metrics describing plug-in charging are given in Table 12. Although the vehicles are being charged, on average, more than once for every day they are driven (Table 12, line 2), they are being driven further between charging events (Table 12, line 3) than their average CD range of 31.5 miles. This suggests the need for increased day-time "opportunity charging" to increase the proportion of miles travelled in CD mode.
There are several factors which may explain the lack of charging in the fleet. Because this is a largely commercial fleet where vehicles are most often used for accomplishing daily work tasks, vehicle operators may not be focused on looking for charging opportunities during the day. Also, work responsibilities may limit operators' access to charging infrastructure or time for charging. Furthermore, it is unclear how many operators are aware of the need for increased charging. Finally, commercial fleet vehicle operators may not be motivated to plug in their vehicles, because they likely do not personally pay for the more expensive gasoline fuel they consume instead.
AVTA has begun to study the impact of charging infrastructure availability on charging patterns [6].
Additional studies are being planned to examine charging behavior of private vehicle owners, and to explore the impact on driver education on charging patterns. Studies involving PHEVs operating in "rich" charging environments are also being planned.
Looking beyond operating mode and charging frequency, the energy efficiency of the Hymotion Prius and other blended-mode PHEV conversions is highly sensitive to usage conditions, due to vehicle system design limitations. Aftermarket conversion companies have limited ability to change the stock vehicle's control strategy, which was designed for a hybrid electric vehicle (HEV) system with much less electrical energy and power capacity. Hardware limitations, such as HEVoptimized electric drive component power limits and transaxle kinematics, also constrain fuel economy improvements as a PHEV.
One area where this sensitivity is manifested is driving aggressiveness. As a driver's demand for power increases beyond the capacity of the electric drive subsystem, the ICE must come on to meet power demand, even while in CD mode. Figure 13 shows how fuel economy decreases as driving aggressiveness increases in individual trips. In this case, aggressiveness is defined as the percent of time spent with the accelerator pedal position beyond 40% of full pedal travel during individual trips. Results come from 700 CD trips driven in ideal conditions (warm ambient and vehicle operating temperatures and no climate control usage).
Trips meeting these criteria were randomly selected from 2008 data. Other sources of performance variation include route type (e.g., city, rural, and highway.), ambient and vehicle temperature, use of climate control, and so on. These factors are explored in other works [7].

Charging behavior as it affects the electrical grid
As there is much interest in how PHEVs will affect the electrical grid, metrics are provided that summarize plug-in charging time and energy consumption.  Note that the average time spent plugged in per charging event (Table 13,     The EnergyCS data logger records DC energy only. AC charging energy for these vehicles was calculated by multiplying DC energy discharged during driving by a round-trip efficiency of 77%. This efficiency value was observed during dynamometer testing. The average CD range for these vehicles is 31.5 miles.

EnergyCS Prius
Metrics describing plug-in charging profiles are given in Table 16.

Toyota Prius with Manzanita Micro Conversion
The Toyota Prius with Manzanita Micro conversion (or Manzanita Prius) uses a 5-kWh, Hawker valve-regulated lead acid (VRLA), supplemental battery pack. The pack does not broadcast any parameters to CAN; therefore, it is not possible to sense the operating mode of the vehicle. Table 17 shows 2008 overall energy consumption for two Manzanita Priuses with the Hawker VRLA pack.

Hymotion Escape
Tables 18 and 19 show 2008 energy consumption and trips/distance with respect to operating mode for two Hymotion Escapes. AC energy was calculated by using an AC-to-DC efficiency of 81%, based on observed average charger efficiency during dynamometer testing. The average CD range for these vehicles is 51.7 miles.
Metrics describing plug-in charging profiles are given in Table 20.

Hybrids Plus Escape
Tables 21 and 22 show 2008 energy consumption and trips/distance with respect to operating mode for seven Hybrids Plus Escapes with Hybrids Plus packs. AC energy was measured directly for these vehicles using the V2Green data logger. The average CD range for these vehicles is 63.8 miles.
Metrics describing plug-in charging profiles are given in Table 23.

Conclusion
AVTA testing of PHEVs in controlled conditions shows potential for significant petroleum displacement. AVTA fleet demonstrations also show that some conversion PHEVs operating in actual-use conditions have achieved impressively high fuel economy (i.e., 80 to 120 miles per gallon for Prius conversions) sustained over significant distances. On average, however, most conversion PHEVs evaluated in the field fall short of this potential. A major reason for this is that the vehicles are not plugged in often enough relative to the distance driven. This may be due to such factors as operator awareness, time availability, and availability of charging infrastructure. Future work is warranted to understand the impact of these factors on charging behavior and to identify enablers to overcoming obstacles to plug-in charging.
Fleet average CD fuel economy was 22 to 55% higher than CS fuel economy, depending on the vehicle. While these gains are significant, they again fall short of the vehicles' demonstrated petroleum displacement potential.
The difference is largely a result of pronounced vehicle sensitivities to varying driving conditions, which sensitivities are possibly due to the limits of aftermarket conversion vehicle design. Some of these sensitivities can be managed through modifying driving behavior, as has been shown with driving aggressiveness. Further improvement may require more highly integrated conversion PHEVs or purpose-built, optimized PHEVs.
AVTA will continue partnering with industry and government stakeholders to evaluate these vehicles as they become available.