It seems like every day we hear about another incremental breakthrough in battery technology. By tweaking existing battery chemistries or inventing new chemistries altogether,university researchers and startup companies have managed to increase the energy density, cycle life, energy efficiency, and safety of numerous potential grid battery technologies. There’s no doubt that these advances in energy storage hardware are important. However, I think sometimes the focus on energy storage hardware overshadows another major challenge facing grid batteries: software to decide what batteries should do and when they should do it.

Unlike renewable energy technologies, grid battery storage does not have the unique advantage of turning freely available wind and sunlight into useful electricity. Thus, capturing value from a grid battery is a lot more complicated than simply pointing it at the sun or facing it into the wind.

A battery’s value comes from the fact that it can do something few other devices can: store large amounts of electric energy in a relatively small form factor. However, this property alone does not make a grid battery valuable. Without the right instructions on when to charge and when to discharge, even the most advanced grid battery is still just an overpriced box of chemicals.

With the right software, grid battery storage could be programmed to operate in a way that not only maximizes its value across multiple applications, but also extends its lifetime. This sort of software would devise a charge-discharge schedule for the battery that maximizes its value without overcharging the battery, overdischarging the battery, or causing the battery’s capacity to fade too quickly.

Unfortunately, it seems this sort of software is still a long way away. Developing software of this kind is difficult because batteries use complex electrochemical reactions that are difficult to model and impossible to observe once the battery cell has been sealed. Furthermore, the potential value of a battery is often determined by external factors like the wholesale price of electricity or the level of electric demand inside a building, so it’s not always easy to decide exactly when to charge and when to discharge, even if you know everything that’s happening inside the battery.

The battery operational management problem is something I’ve studied extensively as a Ph.D. student in the Department of Mechanical Engineering at the University of Texas at Austin. The problem is interesting to me not only because I think it is an important one to solve, but also because I think it is possible to cost-effectively improve the performance and lifetime of battery storage from the software side.

In a paper recently published in the journal Energy, I developed an optimization program that takes a first crack at the problem of grid battery control. The optimization program contains a battery model that predicts how the charge-discharge schedule applied to a battery will affect its voltage and state of charge, both of which must be constrained within certain limits during operation. Using the information provided by the battery model, the optimization program intelligently tests thousands of different potential charge-discharge schedules to determine which one will maximize the value of the battery system over the day without overcharging or overdischarging the battery cells.

Optimization software can be used to decide when a battery participating in the electricity market should charge and discharge to maximize its value. This figure from a paper I published in the journal Energy shows how much charging (negative) or discharging (positive) power should be applied to a Li-ion battery cell to maximize its value over the day in the Electric Reliability Council of Texas (ERCOT) energy market.

Using the software, I was able to show that lithium-ion batteries used to buy and sell energy in Texas’s wholesale electricity market could capture far more value during a handful of operating days when electricity prices are extremely volatile. This result suggests that the life cycle value of a grid battery could be increased if its software were smart enough to know when to perform a grid duty cycle and when to transition to a standby operating state that maximizes the battery’s lifetime (e.g. by cooling or fully discharging the battery). Building this sort of smarts into the software is my next step.

I am hardly the only one working on grid battery control software for economic operational management. One industry notable is Growing Energy Labs Inc. (GELI), which has developed an entire operating system around grid-connected batteries. Another notable is Younicos, which bought up the remnants of Texas battery startup Xtreme Power. Younicos’s grid management software controls batteries as a bridge between multiple distributed energy sources connected to one or more loads.

I’m excited to see how grid storage software will develop as battery hardware drops in price over the coming years. The exciting part about the software world is that innovations move much more quickly than on the hardware side. Perhaps this time next year I’ll be writing about the next big open-source battery energy management software. We’ll have to wait and see.

(This news story is from Scientific American)

Cron Job Starts