New SAE J2908 Revision
The Journey to Settle the Question: What is “System Power” for xEVs and Can We Compare It to Existing Engine Ratings?
Guest Post by Michael Duoba, Research Engineer, Argonne National Laboratory
Posted: March 22, 2023
Buying a new vehicle, you are presented with a myriad of options. One such option is the engine. Do you choose the base 4-cylinder with 180 HP or the optional turbocharged version with 240 HP? Engine ratings are a familiar language for consumers going back many decades. Performance car bragging rights are derived from these specifications and SAE has been instrumental in providing rating standards so that bragging rights are well deserved and not a product of loosely defined terms or inflated assumptions.
It’s a new day and along with engine choices, we now have powertrain choices: HEVs, BEVs, PHEVs… sometimes referred to as “xEVs” or “electrified vehicles.” The standard engine power test (SAE J1349) does not apply to electrified powertrain systems and so comparing xEVs to conventional vehicles has been difficult. Electrified powertrains are complex and diverse with multiple power-producing components all operating differently but working together to provide total system power. This is a perfect problem for SAE standards to address by providing uniform approaches and comparable results for vehicle specifications so new technology can thrive.
System Power in an Engine Power World
The J2908 journey started in 2015. Automotive OEMs were providing power ratings in good faith, but these definitions were typically suited to their own powertrain configuration. For example, because the power-split HEV configuration is so complex (power recirculation among engine, generator, and motor) a popular definition was: System Power = engine rating + peak battery power. However, if we use the power-split definition for all possible HEVs, unacceptable anomalies can occur. For example, the BMW i3 has a BEV version, and a “range-extender” version which by definition is a PHEV. If we apply the “engine + battery” rating approach, we get different power ratings for two vehicles that provide the same power at the wheel.
To confront these inevitable anomalies the J2908 task force addressed the problem at the core. If we want system power, and we want to compare all powertrain types equally, we need to measure at a common point for all powertrains – we need to measure power at the wheel. This agnostic approach could be applied to all powertrains, including conventional engine-only vehicles. The J2908-2017 version specifies a wheel power test by defining thermal and battery SOC pre-conditioning, the dynamometer equipment, test procedures, and post-processing procedures to arrive at a system power measured at the wheel. The downside of this approach is that moving the measurement point from the engine shaft to the wheel hub lowers the power rating for a given vehicle because driveline losses from the engine (and/or motor) shaft to the wheel hub. At the time, no clear method emerged to determine power for all types of xEV powertrains directly comparable to existing engine power ratings. We tabled the much harder problem of finding the “upstream power” comparable to legacy engine power ratings for the future.
J2908 Revision to Determine the “Upstream” System Power
They say, “life comes at you fast.” In 2020, manufacturers' demand to revise the J2908 document led to its reopening. Despite our hopes that J2908-2017 solutions would prompt a migration to wheel power to resolve rating comparison issues, this never occurred (although there was anecdotal support for J2908 wheel-power from a car-savvy Motor Trend audience in an on-line forum). Our task was to define "upstream power" that compares to legacy engine power ratings, so we had to go back to the drawing board.
If wheel power isn’t the system power determination point, what should this point be? This critical question needed to be answered before any progress could be made. Electrified powertrains can have two or more components working together to provide total power. Adding up the peak power rating of each individual component does not provide the right answer, quite often we saw each component arrive at its respective peak output at different moments in time during wheel power testing in support of J2908-2017. It was then we realized we needed to keep the wheel power test to demonstrate peak system power condition but find a way to define and determine the upstream system power during the test.
Again, the challenge of applicability across different powertrain types and configurations had to be addressed. Our solution was to present a clearly defined definition for system power but offer flexibility in how that information is determined. The upstream version of J2908 system power is defined as: “The sum of mechanical shaft power outputs of all components contributing to the propulsion of the vehicle at the condition of rated wheel power (maximum power).” How do we determine each component’s shaft power? This depends on the application. Each automotive OEM has extensive component-level data for the system and can monitor all the powertrain components during the wheel power test to calculate what occurred at the shaft during the test. For some systems, accurate torque and speed are estimated by the controller in real time for each component. For some parallel HEVs, the easiest approach is using the wheel power measurement and calculating the upstream power using detailed efficiency data of the driveline shafts and gears. In any vehicle, if the determination result is accurate, any 3rd party tester could potentially install invasive torque and speed sensors on all the relevant components and directly measure the same system power result.
The hope is that automotive OEMs will begin to adopt the new method. Feedback from industry with their experience using J2908 will be instrumental in making refinements and improvements for even wider use. With greater transparency and consistency, customers can have greater confidence in the performance of these vehicles, which will be essential as the market for electrified vehicles continues to grow.
I want to thank everyone involved in this collaborative, international effort. It was a pleasure working with so many engineers from such diverse backgrounds including those from: the SAE J2908 task force, the UNECE EVE informal working group, the ISO/TC22/SC37/WG2 committee, and SAE staff.