The systems-analysis model is an important tool for examining the behavior of various overall vehicle system configurations. System models integrate models of individual vehicle components and power train components to predict component and overall system performance. Their primary use is in the development of performance specifications for individual system components to provide optimal overall system performance and in the study of system performance over a range of driving patterns. They can also provide information on future components and total vehicle systems for cost and reliability analysis and for trade-off studies. Over the past few years, the PNGV systems analysis team has developed the PSAT model for these purposes. PSAT is a forward, or “driver-driven,” model (i.e., component and vehicle performance are calculated from driver inputs).
During the past year, the systems-analysis team has made encouraging progress in the development of the PSAT model and in providing analysis support to the other PNGV technical teams. A new structure has been set in place for overall development of the PSAT model. Through the national laboratories, DOE has assumed the overall responsibility for funding the development of the PSAT model, in parallel with the development of another vehicle systems model, called ADVISOR, which has been developed by the National Renewable Energy Laboratory. The ADVISOR model is a backward, or drive-cycle-driven, model (i.e., it calculates the power train performance required to drive the vehicle through a specified drive pattern). ADVISOR, which continues to evolve, is available to all users via the Internet; PSAT is available through a secure web site
and now has more than 100 registered users. The ADVISOR model runs quickly, requires modest computer resources, and is useful for concept development and evaluation. The PSAT computer code incorporates more sophisticated dynamic models of component behavior and readily allows analysis of transient vehicle performance and control system development. PSAT has been upgraded and improved during the past year in accordance with the PNGV systems-analysis team’s work plan, and that process is continuing. An advanced training class for PSAT held in January 1999 was attended by more than 50 people. PSAT has been used to examine the benefits of improved control algorithms (Oakland University) and optimization techniques (University of Michigan) on HEV performance.
The new structure for the coupled development of these two models through joint management and funding by DOE, with industry in a consulting/advisory role, makes excellent sense. Based on different logics, these models are suited to different objectives, and the similarities in many of the subsystem component models will allow developments that improve these subsystem models to be used in both PSAT and ADVISOR.
ASSESSMENT OF THE PROGRAM
The HEV propulsion system, which combines an engine, such as the four-stroke CIDI engine, with energy-storage devices, such as batteries, in a lightweight vehicle, is essential to meeting the 80 mpg fuel economy goal within the time frame of the PNGV program. There are many ways to configure an HEV power train system, and developing and then optimizing such systems with their many interactive components is a challenging task.
The recent announcement by EPA of its proposed Tier 2 emissions standards starting in 2004 has made the modeling of the emissions of HEVs an urgent task for the PNGV systems-analysis team so that extensive emissions/fuel economy/ performance/cost trade-off studies can be carried out. Part of this task is to ensure that the emissions models in PSAT and ADVISOR are sufficiently complete and accurate for this purpose. A key issue here is modeling the performance of CIDI engine exhaust catalysts for NOx and traps for PM and validating these models against experimental data. Because the HEV uses a smaller engine and can control transients through the battery/electric motor propulsion system component, the models should explore whether the HEV configuration provides significant additional opportunities for emissions reduction beyond the lowest emissions levels provided by a stand-alone CIDI engine power train.
A second important task for the PNGV systems analysis team will be to develop a more complete model for fuel-cell power train systems. The fuel-cell-based propulsion system, with its inherently low-emissions characteristics and potential for high efficiency, is a promising longer term technology being pursued by the PNGV program and automobile companies. An unresolved question for
the fuel-cell system is how the choice of fuel—hydrogen, methanol, or gasoline/ hydrocarbon—will affect the overall system configuration and performance. With hydrogen as the fuel stored on the vehicle, a hybrid system with electrical energy storage will not be required (although it may improve vehicle performance and fuel economy, as well as lower costs). With liquid fuels, the dynamic response of the onboard methanol or gasoline-to-hydrogen reformer is not likely to be adequate to follow vehicle start-up and driving transients, and an HEV system with energy storage will be required. The performance, efficiency, emissions, and costs of these different fuel-cell systems will have to be systematically analyzed. To date, only limited systems-analysis studies of fuel-cell HEV systems have been carried out, and the most promising configurations have yet to be defined. This is an important area for the PNGV systems-analysis team to focus on.
The validation and review of systems models is continuing. The Toyota Prius HEV has been used as a source of data for model assessment. PNGV teams have reviewed critical component models and plans for model improvement. The 4SDI technical team’s interaction with the systems model (engine efficiency and emissions predictions, engine mass and warm-up, after-treatment-device models) is especially important. The committee is concerned that PNGV has not devoted enough resources to validation of the PSAT system model. Although validation will be challenging, largely because the data are limited, the PNGV systems analysis team should put more emphasis on this task.
A third important task that should be addressed by the systems analysis team is the development of a generic system/subsystem cost model, which is vital at this phase of the program. The committee recognizes that specific vehicle-level analyses have been performed by the individual automotive companies on a proprietary basis, but a more general model indicating the relative importance of various subsystems and components in the cost structure for different configurations would be of great benefit to sponsors and reviewers of the program. If subsystem cost targets have not been met, trade-offs may be necessary, perhaps even in the goals of the program.
Other sections of this report highlight the need for systems analysis of trade-offs among performance, fuel economy, emissions, and cost for alternative primary power plants, fuel composition, degree of hybridization (relative magnitude of engine and battery/motor power, use of regenerative braking for recharging), and impact of accessory loads. For fuel cells, important issues are the trade-offs between efficiency of the fuel cell, overall system energy conversion efficiency, component sizes, and cost. For batteries, an important question is how failure to attain cost goals will influence the optimum degree of hybridization and the overall system energy conversion efficiency.
Recommendation. Given the potential of fuel-cell technology for meeting the efficiency and emissions objectives of the PNGV program, the systems-analysis team should increase its efforts to develop more complete and accurate fuel-cell system and component models to support the development and assessment of fuel-cell technology.