A closer look at coulombic efficiency

Peter Ulrix and Stef Leemans of PEC explain how the very delicate process of measuring a cell’s coulombic efficiency could significantly speed up development times.

In May, the California Center for Sustainable Energy (CCSE) put out its results for America’s biggest survey of plug-in drivers yet. More than 2,000 California EV owners responded, and while 92 percent of them are satisfied overall with their EV purchase, almost all of them want a longer range. In fact, while the respondents drive only about 29 miles a day on average, only 10 percent of them are happy with a range of 100 miles or less, and 57 percent of them want a range of more than 150 miles. Since this was a survey only of drivers who already bought EVs, we can divine that ICE-driving customers would probably want even longer ranges before considering going electric.

The CCSE survey merely proves what we’ve all known anecdotally: that drivers by and large want ranges that are double or even triple what most EVs currently offer. We also know that developing battery technologies will be able to provide those ranges eventually, but that could be a decade or more down the line using current projections. What we could really use are some technological breakthroughs that speed up the battery development process.

Cycle testing has proved to be one of the great bottlenecks to battery development. However, a technology that measures the coulombic efficiency of a battery cell – the ratio of the energy delivered by the battery during discharge to the energy stored during recharge – could speed up the testing times significantly – that is, if or when the process can be made commercially viable.

Here to explain some of the positives and negatives of measuring coulombic efficiency are two men from PEC, an energy storage logistics, manufacturing, and testing specialist with offices around the world. Peter Ulrix, VP Sales and Marketing, and Stef Leemans, Product Manager, are working to integrate PEC’s standard battery cyclers  with coulombic efficiency ratings.

 

Charged: How can coulombic efficiency measurements shorten the development cycle for new batteries?

Peter Ulrix: In the automotive industry, people want cells that can cycle many times – for electric vehicles, we’re looking at 5,000-10,000 cycles – to reflect eight or nine years as a warranty period. If you want to test all 10,000 cycles every time you make potential improvement to cells, it can take years before you say, “okay, now we have an improvement.” You have to go faster. 

One way to speed up testing is to look at the coulombic efficiency of a cell. Basically, the coulombic efficiency is the equation between what you put into the cell in terms of energy, and what you get out of the cell. Typically, you have to put a little bit more energy in the cell than what you will get out after full discharge. This is because of the inefficiency of the cell, caused by internal resistance and aging. Measuring the coulombic efficiency of a cell shows you how efficient it is. If a cell’s coulombic efficiency has a value of 1, then it means that whatever you put in, you can send out, but this is not reality, of course. 

If you put a cell through a limited number of cycles, like 10 to 50, then coulombic efficiency starts stabilizing, and you will begin seeing a trend. If you extrapolate that, it can give you a significant idea of how the cell will behave in the longer term, before you start testing thousands of cycles. If you see something in those first cycles that is not showing a good coulombic efficiency, then you can immediately say, “let’s stop, because it will never be good enough in the long term.” 

It’s like an early detection system, so you can have a quick evaluation method for new chemistries, separator material, or whatever part of the cell you’re trying to improve.

 

Charged: If a perfect cell has an efficiency of 1, what is the average range of real cells? 

Stef Leemans: With cells having different lifespans from a few hundred to a few thousand cycles, you see coulombic efficiencies ranging anywhere up to 99.99 percent. Different lithium cells all have their own characteristics: e.g. if you increase power, you might lose on the cell’s cycle life. If you focus on safety, you might decrease performance or increase cost. Trade-offs have to be made.  

 

Charged: What kind of hardware do you need for measuring coulombic efficiency?

SL: There are four main factors that need to be controlled and measured extremely accurately. You need accurate current control with high-resolution readings – that’s one. You need an accurate cut-off voltage – that’s two. You need very fast sampling – that’s three. And the capacity depends on the temperature of the cell, so you need extremely high temperature stability.

PU: We use a combination of our standard multi-channel cyclers and auxiliary I/Os. They have really good time resolution and temperature stability – we do millisecond sampling and millisecond capacity calculations. Most other commercially available systems are far away from that, except for some custom-built single-channel systems. 

We also use cell-specific, customized holders that have built-in cooling, with very accurate temperature regulation. 

Then you just start cycling, charging and discharging based on a C/10 constant current. We’re able to achieve an accuracy on the coulomb efficiency measurement down to 10 ppm – 10 parts per million. 

 

Charged: So is this technique commonly used today?

PU: It’s not used on a big scale because of its technical challenges. It only makes sense to perform these tests when you can achieve an accurate measurement of coulombic efficiency. If it becomes imprecise, then it cannot give you any meaningful prediction of the impact on cycle life. 

SL: If you don’t measure accurately enough, and then extrapolate to thousands of cycles, you get a huge spread on the potential end result. 

PU: The big challenge is controlling the environment. The climatic conditions have to be very stable. When we started doing experiments with our equipment, we even noticed differences between the daytime and nighttime – the air conditioning system of the building was not stable enough to compensate for the temperature differences. 

If we talk about development of large-format cells that are used in hybrid cars and EVs, C/10 is not what you’re typically running on those cells. So if we can accurately control the cell temperature in high-current applications, then I think that coulombic efficiency could definitely be a game-changer in terms of cell development.

 

Charged: Why is it difficult to offer a commercial solution?

PU: It should be sold as a turn-key solution. Besides the equipment, you need infrastructure – for cooling the cell, and cooling the whole circuitry – which is cell-dependent.

There are standard cell formats like 18650, but if we talk about large-format cells, they’re always different. For each of these form factors, you have to build a cell holder with built-in temperature control. This requires customization and testing for each system being installed.

 

Charged: As these cooling techniques become more commercially available, could it speed up the development process for new materials industry-wide?

SL: Absolutely.

 

This article originally appeared in Charged Issue 8 – JUN 2013