WEBINAR: Learn how innovations in insurance unlock the full potential of battery storage systems
REGISTER TODAY
.svg){kind=logo}
.svg){kind=logo}
Knowing the inner state of a battery is essential for accurately predicting its lifetime and state of health. One useful method is analyzing the open-circuit voltage curve. In this article, we will explore how the open-circuit voltage curve is obtained through field data and how further analysis techniques such as incremental capacity analysis and differential voltage analysis can be used to better manage batteries and ensure optimal performance and safety.
Batteries are one of, if not THE most valuable component in many applications, including electric vehicles (EV), residential solar energy systems, and battery energy storage solutions (BESS). To get maximum value, it is critical to have a complete understanding of the battery.
This understanding gives battery owners and operators useful intelligence, including:
However, it is impossible to look inside a running battery, so how can we access the state of a battery in operation?
There are several diagnosis techniques that can be applied to evaluate the state of batteries in operation. However, some of these techniques require specific hardware and measurement methods, such as electrochemical impedance spectroscopy (EIS) or differential thermal voltammetry, which are not always available.
An alternative option, which does not require specific hardware, is analyzing the open-circuit voltage (OCV) curve of batteries. To calculate the OCV, sensors measuring the voltage, current, and temperature of each battery cell are sufficient. These values are already tracked by the battery’s inbuilt battery management system (BMS). Therefore, extracting and analyzing the OCV of a battery is an accessible and preferred way to investigate the state of a battery in operation.
The open-circuit voltage (OCV) curve is the voltage of a battery as a function of the state of charge when no external current is flowing and all chemical reactions inside of the battery are relaxed. Each battery chemistry and cell type have an individual OCV curve based on its inner state, which is why the OCV curve can be compared to a fingerprint. The OCV curve is analyzed as it evolves throughout the battery’s life because these changes are correlated with changes in the inner state of the battery. Figure 1 shows the OCV curve of a Nickel Manganese Cobalt (NMC) battery.
Changes in the OCV curve visualize capacity loss, and different degradation modes of the battery can be distinguished with them. Degradation modes group aging mechanisms into three groups:
Batteries can be used more efficiently, prolonging their lifetime, by understanding these modes. Additionally, rapid changes of the inner state can be used as a precursor for safety-relevant events.
Incremental capacity analysis (ICA) and differential voltage analysis (DVA) are used to analyze the changes in the OCV curves. These methods make the changes of the OCV curves visible and quantifiable. Additionally, the changes detected through ICA and DVA analysis can be linked to the presented degradation modes. This is why these methods can be used to quantify the degradation modes and provide a better understanding of the inner state of the battery. The ICA differentiates the capacity by the voltage (dQ/dV), whereas the DVA differentiates the voltage by the capacity (dV/dQ).
Get battery analytics trends and industry knowledge delivered directly to your inbox. Because we batteries.
Extraction of the OCV curve during operation is difficult. The challenge is that many different reactions take place inside a battery, which all influence the measured voltage of a battery, and these reactions cannot be easily distinguished. Further, only the voltage on the poles of a battery can be measured.
The measured voltage of a battery is always a combination of, but not limited to, the following four different effects. The effects are depicted in Figure 2.
* Equilibrium voltage and open-circuit voltage are not the same, however for simplification reasons these terms are used as synonyms.
Let’s look at what these four different effects represent.
The OCV represents the potential difference between the active material of the battery and the electrolyte with the lithium-ions. This is why the OCV curve depends on the material used in a battery and is especially influenced by the anode and cathode material. Therefore, the OCV curve of a lithium iron-phosphate (LFP) battery looks different compared to a battery with a nickel-manganese-cobalt cathode. And even within the family of NMC batteries, the OCV curves look different when comparing NMC nickel-rich batteries to traditional NMC batteries. Besides the anode and cathode material, the OCV curve is influenced by the electrolyte concentration and the temperature.
The resistance voltage represents the ohmic voltage drop of a battery. The resistance of the battery depends on the resistance of the poles, current collectors (grid) and electrolyte. As soon as a current to the battery is applied, an ohmic voltage drop can be seen. If you charge the battery, the voltage will rise, whereas if you discharge the battery, the voltage will drop. As this change is linear with the applied current, the more current you apply, the higher the voltage drop is. Additionally, the resistance depends on the battery temperature. The higher the temperature, the lower the battery resistance. The reason for this is that the resistance is mainly influenced by the resistance of the electrolyte. With increasing temperatures, ion transportation speed is increased in the electrolyte, which is the dominant effect of the internal resistance of the battery. The resistance is also influenced by the age of the battery and increases as they age.
More information can be found in our Ultimate Guide to Battery Aging.
Reaction voltage describes the energy needed to change the molecules from the charge to the discharge state. These reactions take place on the surface of the active material. Therefore, the reaction voltage is proportional to the inner surface area. Additionally, it depends on the current since a higher current increases the reaction speed.
Diffusion voltage results from concentration gradients caused by electrochemical processes. Diffusion processes take time, so a diffusion voltage is present even after the current is disconnected. Time duration of the diffusion processes depend on the temperature. Higher temperatures accelerate the diffusion processes. The OCV voltage is only isolated when all reactions are no longer active.
Figure 2 depicts the voltage curve of a constant current discharge of a battery. Two voltage curves are shown. The measured voltage curve and the OCV curve.
To investigate OCV curve changes, the overpotentials (U_resistance, U_reaction, U_diffusion) need to be deduced from the measured voltage (U_measured) to extract the OCV curve (U_OCV) using the following equation:
U_measured = U_OCV - U_resistance - U_reaction - U_diffusion
Models of these overpotentials can be used to characterize an OCV, but doing so with field data is especially challenging because measurement errors provide an additional layer of complexity. Once OCV data is extracted, trends in behavior can be analyzed using ICA/DVA methods. Figure 3 shows an example from field data.
We have already seen that ICA analysis of the OCV Curve can be used to detect the loss of lithium inventory (LLI), which can lead to capacity fade. Predictive Battery Analytics uses the OCV curve to detect capacity anomalies for battery fault detection, calculate the state of health, and predict the battery lifetime. Figure 4 compares the average OCV curve for this cell type from our database to an OCV curve of a cell from one of our customers in BESS. A deviation in capacity indicates a faulty cell that could impair the performance and safety of the entire rack. A safety alert is triggered and enables the system operator to exchange the affected module before the error becomes critical.
Understanding the inner state of a battery is crucial to predicting batteries lifetime, estimating the state of health, and detecting capacity anomalies. This knowledge can be extracted from field data by analyzing the OCV curve, the fingerprint of each cell type, and evaluating its changes with ICA and DVA. A Predictive Battery Analytics Platform can automate this process and detect atypical patterns in the OCV curve by comparing them to a database of ideal OCV curves per cell chemistry. The insights provided will help operators optimize battery management to prolong lifetime and early exchange modules with safety-critical cells.
Dr. Georg Angenendt is a scientist and entrepreneur with expertise in mobility and utility-scale battery energy storage systems (BESS). His research on testing, modeling, commissioning, and optimization of battery storage systems has been published in international journals and at conferences. Since 2020, he is the Chief Technology Officer at ACCURE Battery Intelligence, developing advanced analytics software to help companies assess battery risk, ensure safety, and maximize asset value. His personal passion is Martial Arts: mixed martial arts, luta livre, grappling, boxing and Brazilian jiu-jitsu.
ACCURE helps companies reduce risk, improve performance, and maximize the business value of battery energy storage. Our predictive analytics solution simplifies the complexity of battery data to make batteries safer, more reliable, and more sustainable. By combining cutting-edge artificial intelligence with deep expert knowledge of batteries, we bring a new level of clarity to energy storage. Today, we support customers worldwide, helping optimize the performance and safety of their battery systems. Visit us at accure.net.