Why do electric powertrains not perform like ICE (internal combustion engine) powered ones in track situations? (however for drag racing, high performance electric vehicles do just fine). What’s this “throttling” all about w.r.t. BEV range, performance, and charging capability? Fielding questions like this, there seems to be continual “puzzlement” in comparing power obtainable from Electric (all-electric or BEV – battery electric vehicle) vs ICE powertrains, especially when one wants to drive the vehicle aggressively or in track applications. This article will attempt to explain some of the overview level issues in comparing the two types of vehicle powertrains.
In general, ICE powertrains, if properly cooled, can produce peak power and torque for longer durations than electric powertrains. Simply put, the power (horsepower – hp or kilowatts – kW) production chain for an electric powertrain looks like this -> battery system (ESS – energy storage subsystem) -> the controller(s)/inverter(s) -> electric motor(s). In general, these 3 main elements produce the motor shaft hp/kW (different than and usually greater than hp available at wheels). Only looking at a high power electric motor and its inverter/controller vs. a high power ICE, observe the power/torque vs. RPM curves below. For the electric powertrain, the Tesla Model S and Borg-Warner IPM (internal permanent) motor figure below also shows typical power vs. RPM trends.
High Power IPM Electric Motor based drive subsystem (top figure below) and Tesla Model S drive subsystem (dotted lines are power in kW) vs. high Power ICE (lower figure below)
(For Model S depending on specific tires and its 9.73:1 fixed gear ratio = ~113~118 x mph = motor RPM)
The weakest link in this chain determines the maximum motor shaft horsepower that can be obtained for peak short duration and continuous operation of the electric powertrain. The demand for peak hp vs. time requested by driver may reveal one of these 3 components, even if cooled properly, can not sustain the power being requested. When performing the systems engineering and design of an electric powertrain, the top requirements driving the design is what the worst-case operational scenarios will be for the powertrain, so components and subsystems can be sized accordingly for these conditions. Quantifying these vehicle load profiles can sometimes be done using actual track data if the car is instrumented to record data that can be translated into power demand vs. time.
The “battery hp”, or “battery electrical “horsepower” or “battery chemical horsepower” in most cases is the limiting factor in producing the maximum “motor shaft horsepower” over time of driver request and its power changes as the State of Charge (SOC) of the ESS changes over time of use. The ESS voltage x current in amps x “ESS cell type factor” = power in watts x 1.34 = power in hp. However, the controller-inverter and motor power output decreases as the ESS SOC and voltage decrease, known as “voltage sag”, (unless regen braking is able to maintain SOC by putting energy back into the ESS), as shown in example figure below. For Lithium (LiFePO4) Batteries, some designers simply use a conservative “ESS cell type factor” of 80% to summarize the maximum discharge characteristics of the ESS cells being used, such that the Voltage x Amps x .80% = Usable Available Power.
Not all cells produce their “peak or maximum” current for same amount of time and have same voltage sag while producing this current plus due to a cell’s internal resistance, may heat up faster, hence requiring more effective cooling to allow it to deliver the requested power for longer time periods. For example, let’s think of individual cells within an ESS as the whole ESS as an illustration, and say we consider a cell which is a compromise between energy and power, such as the SDI INR18650-30q, a max 5C (15 amps or 5 x its 3ah rating) for max discharge of 10-20 minutes (“full throttle”), while we keep the cell temperature is below its permanent damage temp (70-75°C).
Continuing with our illustrative example, see figure below showing discharge vs. time and notice how the voltage drops vs. time at max recommended discharge of 15 amps and the total time the cell produces its max continuous current of 15 amps – again, imagine that this set of curves is for the whole ESS voltage (not just one cell) as it is dropping this fast when “full throttle” is applied and its effect on the power output from the controller-inverter and motor. It must be mentioned that depth of discharge (or lowest level of state of charge – SOC, and not going above 15 amps – think of this level of discharge as “full throttle” until cell gets to its lowest recommended level of charge state) and cooling is critical to avoid shortened cycle life of ESS cells. This specific cell could lose 20% of its charge/discharge cycles if these operational parameters are not strictly controlled. Going above 15 amps, the Samsung 30Q datasheet (Section 7.9) mentions operation at 20A (short “power bursts”), however operation at that level can be done as long as we don’t let the battery get above 75°C, its maximum rated operating temperature. https://imgur.com/a/VOxh5 (great resource on SDI INR18650-30q test results and other cells testing is Mooch)
As you can see from these curves, very light hp demand (not much fun on the track) from the powertrain results in long time duration for the ESS energy delivery (~3 hours at 1 amp to ~14.5 hours duration at .2 amp to minimum acceptable voltage for this cell) before ESS voltage starts dropping which will reduce the potential for peak hp when demanded. Larger ESSs (higher kWh capacity) would then demand less current per cell (see lower amp discharge curves below, hence increasing time ESS can produce more hp. There are more technical issues involved in proper high performance ESS design that go beyond the objective of this article, like cell internal resistance (IR) creating cell heating at high discharge currents (we always seek lowest IR cells for performance ESS designs), ESS/cell cooling system effectiveness, etc. Refer to Tesla Model S on the track below. (figure src: https://lygte-info.dk/review/batteries2012/Samsung%20INR18650-30Q%203000mAh%20%28Pink%29%20UK.html )
On the flip side of the max cell discharge scenario is the very crucial max charge scenario which occurs during regen braking while driving the electric powertrain aggressively. If one can optimally use regenerative braking and maintain higher ESS SOC and hence voltages, than higher power can be obtained. Usually an ESS cell max discharge and charge are closely related in terms of performance. An important side issue is that battery cells have an estimated number of full charge and discharge cycles tied to their cell temperatures, so depending on conditions controlling fast charge rates from either external charger or a car’s regen system is quite often designed in by major OEMs. (Tesla for example “throttles” down the max charging rate to avoid damaging the car’s batteries)
This is a good article to dive deeper into the obtainable “ESS or battery hp” for an electric powertrains, even though the discussion is for ESSs at a smaller scale – the math is the same. https://insideevs.com/ev-nerdgasm-batteries-cbre-dream-build-next-chapter/
State-of-the-art high power inverter/controllers tend to have short peak power time periods, e.g. 30 seconds, so one would use its maximum continuous current rating of these devices to size a powertrain for a specific desired “peak hp”. This also applies to the motor itself, even if properly cooled. 30 second burst peak power capability may be adequate for a 0-62mph runs, but as has been shown with road versions of Tesla Model S on the track, which was not designed for optimal track performance, constant requests for peak power produces over-heating and forces the system to degrade the allowable power or “throttling down” to what the system can produce continuously, keeping all components in their proper temperature range and not over-stressing the ESS cells. (see linked articles below for supporting evidence)
JB Straubel, CTO, Tesla explains the challenges in communicating “maximum horsepower” of an electric powertrain compared to an ICE powertrain. “There is some confusion about our methodology for specifying ‘equivalent’ horsepower ratings for our all-wheel drive, dual motor vehicles,” Straubel said. “Attempting to directly correlate horsepower ratings in petroleum burning vehicles to horsepower in an electric vehicle is a difficult challenge.” https://www.tesla.com/blog/tesla-all-wheel-drive-dual-motor-power-and-torque-specifications
As impressive as the new racing series “Model S for racing”, called the “Electric GT Model S P100D” was on the track, however, the Model S powertrain was not designed for track use. British racing driver and former Top Gear and Fifth Gear television presenter Tiff Needell recently got behind the wheel of the Electric GT’s Tesla Model S P100DL race car. https://www.teslarati.com/tesla-model-s-p100dl-electric-gt-track-tested/ Tiff Needell ultimately noted that the vehicle’s battery (ESS) still showed some heating issues, as revealed by the Model S’ control system throttling (reducing) performance to safegauard the powertrain, after only 1.5 hot laps around the Barcelona track and fans were required to be directed over the vehicle’s battery pack after runs in an attempt to lower the ESS temps – the ESS cooling system needs an upgrade. The veteran racing driver stated, however, that if Electric GT can find a workaround with the Model S’ battery heating issues, the racing series could very well be quite successful. However, the Model S P100DL powertrain does do quite fine in drag racing where it is able to accelerate harder than it can brake.
The Tesla Model 3 powertrain design has considered track or aggressive driving as one of its “worst-case” scenarios and recent tests show good results when compared to the Model S track performance as mentioned above and in this linked article. “Model 3 was able to handle nine laps on a racetrack without the car’s software limiting or throttling of its performance, so this is noteworthy.” https://www.teslarati.com/tesla-model-3-laguna-seca-raceway-track-racing-video/ Adding to this success, just recently, a non-Performance Long Range Model RWD Model 3 with aftermarket suspension and brakes won the Canadian Sport Compact Series Time Attack series, defeating a Porsche Boxer and a Mazda RX-8 to claim the title. During the entire event, Sasha Anis, the driver of the vehicle, noted that the Model 3 did not experience any heating problems at all. The Model 3 Performance, which Elon Musk stated would be 15% faster than a BMW M3 around the track, is expected to perform even better during spirited driving.. The upcoming dual motor Tesla Model 3 performance version may even have a “track setting” for improved longer duration high hp demands and regen capabilities.
TAKE-AWAY POINTS: Electric powertrains vs. ICE powertrains
1 – Peak torque at zero electric motor RPM is a great advantage over ICE powertrains. Current top performing stock BEV cars best their ICE powered brethren.
2 – Weakest link in electric powertrains is usually the ESS when longer time duration max hp is demanded. In addition, optimized ESS cooling systems are required. (in our powertrain designs for performance, a/c boosted chillers are used in the ESS cooling loop to assist)
3 – Both max discharge and charge performance of ESS cells are crucial for power production and effectiveness of regenerative braking
4 – Large capacity ESSs can greatly help in producing more power for longer durations, where the voltage decrease or “sag” and current demanded per cell are both less, however a properly cooled ICE powertrain can produce peak power for longer durations.
5 – Regenerative (regen) braking of electric powertrains allows recovery and transfer of kinetic energy into additional electrical energy or range stored in the ESS (battery system) for the vehicle, minimizing lost energy due to heat, as is the case with mechanical brakes – a regen powertrain will decrease wear and maintenance on mechanical brakes.
Stay tuned to this post … updates to be added as more questions arise …