As automotive companies transition their product portfolio to fully electric, the development and validation of the electric powertrain and battery is a key enabler to success. The electric powertrain, which consists of the battery, traction inverters, and electric motors, or a combination of components like an e-axle, is a primary cost driver, has a tremendous effect on the driver’s user experience, and has the highest risk for expensive recalls and warranty issues. Because of the importance of the electric powertrain, automotive customers must transition their existing development and validation facilities to electric component and system testing which often requires a substantial increase in needed electrical power. However, intelligent installations can use green energy generation, second life energy storage, and high-efficiency test systems to reduce, or even eliminate this challenge.
The demand for power is clear with some basic, back of the napkin, math. If we have battery packs that average 100kWh, and the test profiles must include simulations of repeated DC fast charging (20 percent to 80 percent SOC) with the target recharge of 20 min, which seems to be the desired target time by many consumers, this fast charge event needs to be 80% - 20% = 60% * 100kWh for a DC fast charge capacity of 60kWh. Then, for a 20-min DC Fast Charge we would need (60 min/20 min)*60kWh) or 280kW of power. In addition, a typical validation program would have 20 packs (or more) in continuous test for multiple years, so just for these tests we would need a facility power of 280kW*20 = 5.1MW.
Of course, the matching E-Drive testbeds would have similar requirements confirming that tremendous power is needed to test these electric powertrain components and this power requirement in both directions as electrified powertrains have the enormous benefit of returning energy as well as consuming it. But this bi-directionality doesn’t help us (on the surface) with our needed power connections.
We can manage this power need with “lab management” software which can monitor test profiles to predict power consumption in the future and limit some testbeds in case power limits are being reached, but this can often delay testing or cause issues where faults occur and excess energy is needed, or doesn’t have any place to go. To take the solution a step further, we must outfit these test facilities with systems that use the available energy effectively in the first place.
There are several methods for this combining a mix of utilizing green energy like solar and wind to supplement the energy needs, a stationary energy source like second life batteries to store or provide excess energy when needed, and efficient test systems that have high electrical efficiency and can share the energy between testbeds for both AC and DC applications. If done correctly, the net energy effect on the facility can be zero, or in many cases, provide energy back to the facility to support the building.
If we start from the DUT (Device Under Test) and work our way back to the power source, we can start to see how this can easily play out.
Combining Test Systems
If we keep it simple and focus on EV battery and traction motor testing, starting with the battery, we need to provide energy to charge the battery which simulates the real-world condition of a DC fast charge event or regenerative braking during driving. We also need energy for electric motor testing to drive the electric motor which is replicating the electric motor accelerating the car. But we also need to consume energy from the battery to simulate the real-world situation when it is powering the electric motor to accelerate the car or absorb energy from the traction motor during the previously mentioned regenerative braking events.
If we think of this myopically, and only consider one test channel, we will have a building AC grid connection to provide the AC power connection to the Active Front End (AFE). This AFE then converts the AC power to DC and then a separate DC/AC or DC/DC converter is used to control the energy to/from the DUT. If isolation is needed for the test channel, then an isolation transformer is connected between the building grid and the AFE. Typical test systems today have a total efficiency of 85 percent to 90 percent, which is good, but we can do better, especially if we step back and view all the test system needs and look at multiple channels together.
If we have the capability of combining test systems together, on a so-called “Common DC Bus,” the possibilities for energy re-use with higher efficiencies is much greater. So, in addition to losing less energy due to losses, we can also save costs (and equipment footprint) by utilizing a single grid connection and a single AFE for multiple channels.
Terms like “back-to-back” testing have been in the industry for years, but now the energy can be circulated at a higher efficiency inside the test system decreasing the energy usage of the overall testfield. If channel-to-channel isolation is needed, new, high-performance Silicon Carbide devices can be used for high-frequency transformers inside the test equipment to provide isolation in a much smaller size. Then, with proper test coordination, back-to-back battery cycling could be like moving water from one cup to another, and back again, only feeding in the minimal losses from the grid.
Once we have test equipment with a common DC bus, we can take it even further and have combined AC and DC applications sharing energy. For example, if we take a circular energy system like an e-axle testbed, we can use a DC supply to supply DC power to the production traction inverter or a universal inverter which then drive the electric traction motor(s) connected to the axle. The axle is then connected to two output dynamometers which absorb this energy and brings it back into the test system onto the common DC bus, closing the energy loop. Only the loses are needed from the grid.
To achieve the next level of energy freedom, additional DC channels can be used to tie in one or more energy storage systems (like a second life battery BESS) to the common DC bus. This, tied into the automation system, could be used to intelligently store, and use, the DC energy as needed to fully support the testing requirements. For example, for facilities that have renewable energy during the day with solar could charge the BESS during daylight hours and use that energy to feed the test system losses at night utilizing the common DC bus to minimize energy losses. Or if charged batteries are brought into the test facility, the initial discharge energy is stored in the BESS for charging the test batteries at a later time.
Even with this high level of re-utilization of energy or green energy supported test systems, we also must try to keep pushing the efficiencies of the power electronics used in the test systems. Even a 1 percent improvement on efficiency in a 280 kW system is significant when you consider that validation programs running 24/7 take years to complete. Utilizing new wide bandgap semiconductors like SiC and GaN enable control techniques which can increase the efficiencies to 99 percent or even higher. In addition, the higher switching frequencies enabled by these new devices can also reduce the physical size of passive filter components reducing losses even further.
As the automotive sector continues to transition to a significant portion of their portfolio being electric, they must either build the needed test facilities to support this transition or update their current facilities. This typically requires a tremendous increase in needed power infrastructure. However, by utilizing the latest energy sharing test system technologies and supplemental energy supply and storage methods tied directly into these test systems, this burden is minimized.
Constant advances in technologies are also providing electrification test equipment with a significant efficiency savings over older, existing, test systems the OEM might already have where the increased operating cost savings can quickly pay for the upgrade, especially compared to constant repairs of aging systems. Here, the combination of increased energy efficiency and the ability to share and store energy could reduce existing operating costs of EV labs by 50 percent or more.
This article was written by Don Wright, VP of Engineering, Unico LLC (Franksville, WI). For more information, visit here .