Search Results

Field experience with electric vehicles has shown in a significant number of cases, the performance of the batteries starts to degrade in a few months or a few thousand miles resulting in unhappy vehicle owners. This has occurred even for batteries for which the manufacturer has claimed a cycle life of several hundred deep discharge cycles. In this paper, the reasons are explored for this large difference between the expected and experienced battery cycle life and what life cycle testing should be done to greatly reduce the uncertainty in battery pack life. Test procedures for battery life testing are discussed and it is shown that there is a large difference in the cycle life that would be inferred from test results for one or two modules compared to that from testing a pack of many modules (at least ten).

Track and dynamometer testing of the Eaton Dual Shaft Electric Propulsion (DSEP) minivan has been performed by the Idaho National Engineering Laboratory (INEL). The dynamometer testing included constant speed tests up to 88 km/h and driving cycle tests on the SAE J227a C and D cycles and the Federal Urban Driving Schedule (FUDS) as well as maximum effort acceleration tests. The dynamometer data were analyzed to determine the energy consumption (Wh/km) of the DSEP vehicle for the various driving modes and to project the range of the vehicle if the NiF170 nickel-iron battery had been at its rated capacity. Ranges of 90-125 miles at constant speeds and about 70 miles on the driving cycles were projected. Comparisons were made of the performance of the DSEP vehicle and the ETX-II and the TEVan minivans, which have been developed on other DOE and EPRI programs using lead-acid, nickel-iron, nickel-cadmium, and sodium-sulfur batteries.

The use of electric vehicle road test data to characterize various missions in terms of velocity time-histories and battery power profiles is discussed. The data utilized were taken on the Griffon Van at Detroit Edison using the VDAS on-board data system, which stores the data in several frequency-of-occurrence matrices. Time-based data are not available from VDAS. Construction of velocity-time cycles from the mechanical power-velocity matrices has proven to be difficult. A new VDAS, which stores the vehicle/battery data in time-based files, has been designed and built at the INEL. Bench testing of the unit is near completion and in-vehicle testing should begin in June 1989. The new VDAS utilizes the same sensors as in the original VDAS and will yield essentially the same data variables. The frequency-of-occurrence matrices can be constructed from the time-based data using a Vax 2000 workstation and commercially available software.

Series of designs of compact and full-size passenger cars and minvans were formulated using state-of-the-art electric driveline components and battery modules/cells. The performance of each of the designs was simulated using the ELVEC. Computer runs were made for constant speeds between 40 and 88.5 km/h and the J227D and FUDS driving cycles as well as maximum effort accelerations. The simulations indicated for the three vehicle types the target ranges and acceleration times could be met for both the near-production and advanced batteries. The targets were consistent with those established for the various batteries by the DOE Battery Goals Task Force in 1988. The energy consumption values calculated for vehicles utilizing DC drivelines were consistently lower by 15-25% than those for vehicles using AC drivelines primarily due to the higher efficiency of the transmission in the DC power-train system.

Studies of battery management systems (BMS), including state-of-charge (SOC) algorithms, were performed in the Battery Test and Vehicle Dynamometer Laboratories at the INEL. The test and evaluation of the SOC algorithms indicated that the capacity rate dependent and adaptive approaches to determining battery SOC yielded RMS errors between the indicated and actual states-of-charge of less than 10% if battery temperature and aging effects were not dominate. The BMS tested were not sufficiently accurate and reliable to permit their use as the primary means of charge control during the various INEL test programs. Considerable progress has been made in the development of BMS, but further work is needed before the systems can be used with confidence in either the charge or discharge modes without careful attention from the user.

Comparisons have been made between data obtained from dynamometer tests of various electric vehicles and computer simulations of the same vehicle-battery combinations for several driving cycles. The vehicles included in the study were the ETV-1, Bedford Van, Unique Mobility, ETX-1, and DSEP(TB-1). The batteries studied were the ALCO 2200, Gel/cel 3, EV5T, ETX-100(CHL12), and the NIF-170. The comparisons indicated that the energy consumption values obtained using ELVEC agree within 10% with test data for both constant speed and variable power driving schedules. The range comparisons were less consistent, but the predictions agreed with the data to within 10% if the vehicle battery was in good condition and the controller did not limit battery power at low states-of-charge.