The Obstacle Course on the Path to Repurposing Used Electric Vehicle Batteries (EVB). Part II: Battery Grading

This is the second blog in our series, this time we’re going to dive into our bread and butter: battery grading.  ReJoule has now tested over 900 kWh worth of batteries - the equivalent of 9 fully electric delivery trucks, and we wanted to share progress. We gave you a preview of this in a prior social media post (LinkedIn, Twitter, Instagram), but here’s a deeper dive!  

What we’ll cover:

1) Status quo: how batteries are tested today

2) What is UL 1974 and some of its limitations?

3) ReJoule’s recommendations for a faster battery grading process

Our analysis is based on LFP (lithium iron phosphate) battery modules from an electric truck (one battery pack is pictured above). Our project goal is to achieve a 30% lower installed cost for used batteries versus new batteries.  The grading process is an added step, so any improvements here will help us achieve our cost target.

Status quo: how batteries are tested today

After you disassemble from packs to modules, you have to sort the modules.  You may be able to buy them pre-sorted (ReJoule can help with that!)  If not, you start by looking for signs of physical damage or leakage. 

After a physical inspection it's time to figure out the battery’s available energy. It’s known that batteries experience degradation differently, even if it’s in the same electric vehicle (EV). We have found a state-of-health (SOH) can vary as much as 15% between modules! This meansthat, with existing methods and technologies, capacity testing needs to be done on every battery module to know the health.


Batteries are not allowed to be shipped fully charged (usually 30% state-of-charge), so you need to fully charge and then fully discharge it.  Depending on the battery size, this process can take 6 to 10 hours per battery and drive up your electricity bill.  Furthermore, the equipment required to do this testing can be very expensive and you’ll need additional equipment if you want to be able test batteries concurrently.  For those reasons, we’ve learned not everyone checks every battery.  There is no requisite standard in the industry at the moment.

What about UL1974?  

UL 1974 is a standard to evaluate the performance of used batteries, but is not yet widely adopted.  It includes a procedure to examine and sort batteries and various tests to assign a grade to a used battery, as well as recommendations for how to repackage the used batteries.  The test procedures include physical inspection, battery management system (BMS) operation checks, historical data gathering, open circuit voltage (OCV) testing, capacity testing, direct current internal resistance (DCIR) testing, and self-discharge testing. Assuming zero down time, we estimate it would take 40+ hours to complete a UL1974 test per used battery.  

We estimate this process adds an incremental 30% to the cost (breakdown of the estimate at the end of the blog).  As an example: if you bought the batteries for $100 per battery module, it means it will cost you another $30 per battery just to test it.  If you want to repackage and sell the batteries, that will add more cost.  


While we strongly encourage repurposing and adopting a common test protocol, we understand it is difficult to add so much additional overhead. However, the UL1974 test procedure can be modified to reduce test time to reduce cost without sacrificing accuracy and safety. How so? Well I’m glad you asked!  


ReJoule’s recommendations for a rapid battery grading process

ReJoule’s fast grading technology expands on a proven fast battery testing technique, electrochemical impedance spectroscopy (EIS). These tests measure a battery’s Impedance, a crucial battery health metric that can be used to model a battery’s true health and remaining useful life (RUL). 

We created and compared our process to UL 1974 as a case study, here are our initial findings:  

We added an EIS test and were able to dramatically reduce the time it takes to get the battery’s capacity without sacrificing accuracy.  Based on the testing we’ve done with the LFP modules we’ve achieved:

  • 2% accuracy in 5 minutes test (> 70x faster than status quo)

  • 4% accuracy in 30 seconds test

Below you can find some of our results. On the left are our early results on a capacity prediction model based on EIS.  The diagonal line shows where 100% accuracy would be.  On the right, the ACI of a cell as its capacity degrades during accelerated aging. The data shown is for NMC 18650 cells. This data was generated using funds from NSF Phase I award.

ReJoule has also modified UL1974 to find and disqualify batteries earlier in the process. Time to rejection, as measured in # tests before rejection, is a key optimization metric.  If you will ultimately reject N% of the batteries, identifying that earlier in the process increases the efficiency of the testing process as less overall tests are performed. 

For example, we found that our results with the UL 1974 defined DCIR procedure were inconclusive, as a result we modified the test to save 7 hours of test time per battery. In the case study, we aimed to stay true to the UL 1974 guidelines.  However, when we found a test was not useful as a grading criteria, we removed it for practicality.

This is an ongoing case study, so follow us on social media to stay tuned on future improvements.  There’s still a lot of room for improvement, so please engage with us in the comments so we can continue to make improvements.  You’re welcome to share your own challenges and recommendations too, so we can learn from each other!

Next steps 

We will expand the screening process to different battery types (chemistries and form factors) so we can continue to refine this process.  Form factors include modules and packs as well as cylindrical and pouch cell types.  Our cofounder and CEO, Steven is a UL 1974 standards technical panel member and will work with UL on advancing the standard to make it more accessible and viable.   

Reference: How did we arrive at a 30% added cost from UL 1974?

We start by assuming $75 per kWh, this is a 50% discount from the 2019 estimate of $150 per kWh for new batteries.  Using our LFP modules results in a cost of $105 per module. At almost 500 modules tested, this results in ~$50,000 as the cost of the used batteries.  

We estimate the added cost to process these batteries would be an additional $15K, thus arriving at 30%.  We estimated and aggregated the costs to grade the batteries across four categories to get to the $15K:

  1. Direct labour: this is the time an operator spends disassembling, inspecting, and testing the batteries.

  2. Non-recurring engineering (NRE): there is no one-size-fits-all for batteries so each new battery design requires a bit of engineering to set it up.  The more batteries are tested, the lower this cost becomes on a cost per unit basis. 

  3. Equipment cost: for high voltage EV batteries, this equipment is easy $50K or more.  We allocate a small percentage of that cost to this case study. Same as with the NRE, the cost per unit goes down the more units you test.

  4. Overhead costs: this includes rent (the space for the batteries as a per cent of your total facility), utilities, and any custom cables or parts. For instance, you may need to buy specific software to be able to perform a BMS check. 

Contact us at info@rejouleenergy.com or leave a comment if you want to leverage our technology to test your batteries.

Cheers, 

The ReJouligans


Legal notice:

This document was prepared as a result of work sponsored by the California Energy Commission. It does not necessarily represent the views of the Energy Commission, its employees, or the State of California. Neither the Commission, the State of California, nor the Commission’s employees, contractors, or subcontractors makes any warranty, express or implied, or assumes any legal liability for the information in this document; nor does any party represent that the use of this information will not infringe upon privately owned rights. This document has not been approved or disapproved by the Commission, nor has the Commission passed upon the accuracy of the information in this document.©2021 ReJoule Incorporated.  All Rights Reserved.

~~~~~

Previous
Previous

The Obstacle Course on the Path to Repurposing Used Electric Vehicle Batteries (EVB). Part III: Lack of Critical Battery Information

Next
Next

The Obstacle Course on the Path to Repurposing Used Electric Vehicle Batteries (EVB). Part I: EVBs aren’t designed to be taken apart!