Your 2019 Battery Performance Scorecard FAQs answered
In early November we released our 2019 Battery Performance Scorecard, which provides independent ranking and evaluation of battery vendors based on testing performed in DNV GL’s laboratories. Since the release, we’ve had some great feedback from those who have downloaded the report. We’ve also received a lot of the same questions, which prompted me to create this blog to answer the top FAQs.
How many cells have you tested?
There are nearly 500 individual test curves a that contribute to the findings of the 2019 Battery Performance Scorecard. And the database continues to grow.
How does Battery XT work?
Battery XT uses 20 years of published literature on battery degradation curve equations to predict whether your warranty will be upheld. The semi-empirical formulae we use are similar to what you’d find in the USABC, INL, and NREL models for battery degradation. We do one thing different than the literature: we use recent, real-world datasets from the Scorecard to calibrate the model to your application.
The Scorecard data allows DNV GL to perform infinitely possible interpolations with Battery XT. What makes DNV GL’s Battery XT unique is that the data comes from today’s batteries. Your batteries. The resolution in the SOC conditions for the data exceeds the accuracy of most model assumptions, and with each passing month more data is added to DNV GL’s extensive database. We’ve advanced the field from a generic analysis to an analysis that can be performed on your batteries directly. If you haven’t tested with us, we can at least narrow the field to a chemistry category with multiple batteries to use for model calibrations. This is less valuable that having your own data directly, but it’s a step.
What is the value of testing?
Every automotive manufacturer with a PHEV or EV (which is now pretty much all of them) uses testing to verify how their batteries will perform. GM’s test lab in Warren, Michigan, is one of the most advanced in the industry. It houses thousands of battery channels and GM can even test battery packs with active variation of temperature, humidity, and even vibration while the pack is cycling in a simulated drive cycle. Few stationary energy storage EPC or development entities have the balance sheet or impetus to fund such a lab.
The good news is the stationary storage industry doesn’t have to bootstrap a bunch of its own labs. It has one. The BEST Test and Commercialization Center, funded by NY-BEST and operated by DNV GL, has done it for you. Luckily for you, our customers, you don’t each have to invest $25M in your own battery facility like GM did. Instead, NY-BEST and the state of New York invested about $6M so that you could have a piece of it. This investment lets all of you benefit from an industry pool of expertise and data.
A Scorecard testing program costs less than 1% of a multi-MW energy storage project. If it saves you from $1M+ in warranty dispute and arbitration costs later, wouldn’t you say that’s worth it? The return on safety testing can be substantially higher.
Give us a call and we’ll quote a PQP for you. We can usually reduce the overall cost for testing programs when we are testing more than one battery, as we can combine the setup costs and better optimize channel utilization around your program.
What is your test method?
There are a few things unique about DNV GL’s Scorecard Testing Protocol. The most unique feature is that we test at intermediate states of charge. We’ve done away with the concept of “depth of discharge” (DOD) and focused more on the problem of determining what state of charge (SOC) range is best for the battery. Most energy storage (and even automotive) power profiles don’t operate like a continuous discharge from the top. They spend a lot of time at intermediate states of charge fluctuating around an average with varying SOC swings. And that average is not always 50%.
It took us about 10 years to perfect this protocol and the past 3 years have shown its payoff. While our battery laboratory is state of the art and capable of any test that any automotive, commercial, public, or research laboratory can perform, we focused the Scorecard testing programs on real-world simulations of battery operations. We don’t have rest periods, and the batteries are temperature controlled in chambers but we aren’t counteracting the effects of thermal lag, just like the batteries would see in the field. We have 40 test conditions and we operate within a design of experiments matrix varying these four variables: average SOC and SOC swing, temperature, and c-rate. Those four variables cover the entire field of battery abuse, and we map it for every battery we test.
Have you tested (brand) batteries?
Most of the time we can’t answer this question because the battery industry is so protective of its data. We know that’s frustrating. It’s frustrating for us too. That’s why we’re publishing this Scorecard, to advance the industry despite this data sharing problem the manufacturers have created.
DNV GL would love for some of the manufacturers to volunteer to be named. It would demonstrate their confidence in their product.
What is the name of each manufacturer in the Scorecard?
Sorry, folks. The battery manufacturers won’t let us tell you. The battery industry is highly protective of its data. They place strict confidentiality on disclosure of their name and data specific to their product. So DNV GL is restricted to presenting this data in an anonymized and aggregated form.
Perhaps our readers find this frustrating. So do we. So we did the next best thing: we showed you the whole field so you can judge whether your battery fits in any of these categories. We did that because we knew that’s the best we can do right now, and people need to know. Still not sure? Give us a call, ask us to test your battery, and let us name it in the next Scorecard. It will take all of us to change this protectionist culture from the battery suppliers.
What is the chemistry of each of the anonymous names?
In 2018 we told you the chemistry of each of the anonymized letters for each battery. In 2019, we have a few chemistries that are distinctive and can be identified by simply naming their chemistry, which would again violate battery manufacturer confidentiality concerns. So this year, to accommodate this concern, we did not distinguish them.
Until the battery manufacturers test with us directly and give us permission to be named, we will have this problem. We have two suggestions:
- If you are a battery manufacturer, contact us, test with us, and volunteer to be named.
- If you are an EPC, developer, or lender, require your battery manufacturers to test with us.
We’re actually showing everyone that any battery can do lots of things if you size it right, integrate it appropriately, and do your due diligence. That should be good for everyone. Our testing should help move everyone towards easier project financing.
What do I need to know about battery safety?
All batteries can burn. We haven’t yet seen one that is inflammable if subjected to flame or high temperature from a cartridge heater. So, our first message on battery safety is that most batteries are made of plastics and use alcohol-based solvents in their electrolytes; as they burn, they emit flammable gases and these need to be ventilated. Second, the gases can also be toxic, similar to a plastics fire. And third, preventing batteries from cascading and keeping them cool with water is the best way to fight a fire. Ask your battery manufacturer what they are doing about cell-to-cell cascading. If their answer is quantitative and addresses failure at the cell, they will likely have an interesting approach on safety. If their answer sounds like a recited line or prepared response with no detail about the rate at which cells can cascade, then, well… you’ll have your answer.
What’s the best battery chemistry?
All of the products tested have their strong suits. The story the Scorecard tells is that a battery can do many things, and how you use is the important consideration. We hope the Scorecard shows you that there is a battery out there that will work for you; you just have to make sure it doesn’t have properties that are not well matched to your application. That’s what our testing and analysis tells you.
What’s the safest cell?
They can all burn. All have safety considerations. All have an appropriate safety management plan.
How many years of testing is in the Scorecard?
Some of the test data in the scorecard goes back to 2009 when DNV GL started battery testing, but the majority of the data is from active commercial products either fully commissioned or in late stage development between 2018-2020.
It is annoying the manufacturers are anonymized.
Yep, it is. Anyone on the buy-side of the market needs to tell the battery manufacturers, contractually, that they must disclose their degradation and safety data. PVSyst wouldn’t have worked unless solar module manufacturers agreed to test, and they didn’t do that until their buyers made them do it. It’s about time the battery manufacturer supply chain did the same. Until they do, we’re all slowed down because of them.
If you have an additional question, or you’d like to set up a meeting to discuss what the Scorecard means for you, please contact me. To learn more, join our webinar on January 30th at 2 PM EST: Leveraging the 2019 Battery Performance Scorecard in North America.