We use physical models to estimate what is going on inside Li-ion battery cells. The model estimates internal states like temperature, State of Charge (SoC), and State of Health (SoH). This web page first presents an overview of the technology and then discusses applications for the electric grid and EVs.

The underlying IP has been developed at University of Washington (UW) for the last few years with funding from ARPA-E and other government agencies. These IPs are protected by pending and issued patents. Bii has exclusive rights to the commercial deployment of current and future research IP from UW related to battery management.

BACKGROUND

We change the batteries from being black boxes to managed assets; i.e., the batteries are operated based on a physical understanding of their behavior rather than simplistic constraints based on limited knowledge about the internal state. This approach enables us to relax conservative operating constraints to improve the value of the batteries.

Capabilities

We focus on two essential capabilities:

  • State estimation: We use a physical model of the internal battery behavior to estimate what is going on in terms of temperature, SoC, degradation, etc. These estimates are improved by a Kalman Filter to integrate the few measured data that are available. The result is higher accuracy (e.g., in SoC) than classical methods which justifies using a wider operating range.
  • Degradation: By knowing the battery capabilities and the benefits to be optimized, an optimization algorithm performs a trade-off between battery degradation versus the benefits generated by the batteries. For example, it may sometimes be worthwhile to fully discharge a battery even though that may double the battery degradation in the process.

Benefits

Our solution is used to modify the operating constraints, improve performance, and improve safety. For grid-connected applications, we expect 20% improvement; for example: 5% longer life, 7% lower capital cost, and 8% more revenue from grid services.

Current Approach

The traditional method of charging Li-ion batteries is constant current followed by constant voltage (CC-CV). It is easy to implement and has wide applicability; but, it is overly conservative. The desire to charge EVs in less than 30 minutes, and the need for frequent charge and discharge in grid applications, have led to modified CC charging protocols. However, the lack of real-time models has led to one-size-fits-all solutions that are kept constant for the life of the batteries. The results are easy to implement and have wide applicability; but, they are overly conservative.

The lack of real-time information about the internal state of the system has led to rigid operating windows specified by the battery manufacturers; see the figure below. To be useful, the window shown in the figure needs more dimensions: Temperature, current, and number of cycles.

battery-bounds

Example of Possible versus Allowed Operating Region for Batteries

Bii is replacing these conservative operating windows with a model-based, application-aware, management system that provides predictive capabilities and is less conservative because we have better physical understanding of what is going on inside the batteries. These results are deployed in a BMS with an increase of a few dollars in computer hardware cost.

Model-Based Approach

Our solution is based on fast-running P2D battery models applicable for all Li-ion batteries. Our integrated battery + economic models enable the batteries to produce cost effective solutions. Our solution a) operates across a wider SoC range, b) keeps the battery safe, c) maintains the battery internal temperatures within desired limits, and d) balances the cost of battery degradation against the opportunity value.

Our improvement claims are consistent with NREL reports that battery models are suitable for system design and control with life extensions of 20% to 50% being possible. The main factors reducing battery lifetime are time at high temperature, high C-rate, cycling at high DoD, and over-voltage. NREL has generated specific results for our models. Their validation testing is continuing.

GRID SCALE APPLICATIONS

Capabilities

We focus on these essential capabilities:

  • We use a battery degradation model to find the optimum charge and discharge profiles. Avoiding over-voltage, controlling the estimated internal battery temperature, and limiting deep discharge, help extend the lifetime of the batteries.
  • By knowing the battery capabilities and the benefits to be optimized, an optimization solution performs trade-off between battery degradation versus benefits generated by the batteries.

Benefits

Our solution modifies the operating constraints, improves performance, and improves safety. For grid-connected applications, we expect 20% improvement; for example: 5% longer life, 7% lower capital cost, and 8% more revenue from grid services.

How Is BII Unique?

We have tight integration between battery degradation models and economic benefits. Thus, our optimization algorithm is able to make the most economical tradeoff between cost of battery degradation versus economic opportunity. Our product can be a replacement or an enhancement to existing BMS products.

For more information, click on this link.

ELECTRIC VEHICLE APPLICATIONS

Capabilities

We focus on these essential capabilities:

  • We use a battery degradation model to find the optimum charge and discharge profiles. Avoiding over-voltage, controlling the estimated internal battery temperature, and limiting deep discharge, our model helps extend the lifetime of the batteries.
  • By knowing the battery capabilities and the benefits to be optimized, an optimization solution performs trade-off between battery degradation versus benefits generated by the batteries.

Deployment

Our solution is part of the battery and it interfaces with fast chargers to make optimum trade-off between rapid charging versus degradation of the batteries. Our solution can also control the discharge rate to reflect the driver’s choice in terms of how hard s/he wants to “push” the battery. Integration of our BMS with the drive-train in electric buses is attractive.

Benefits

For EV battery charging, the preferred benefit is fast charging. Charging from SoC=30% to 80% in 30 minutes is typically done by modified CC; i.e., 2 or more step changes in amount of current. There are three major standards to implement this type of fast charging: Tesla, SAE, and CHAdeMO. Battery Informatics' model-based BMS employs a more refined charging protocol, based on knowing the internal state, chemistry, and age of the cells, thereby charging faster with less degradation.

How Is Bii Unique?

Our battery model provides higher accuracy than the existing heuristic rules. That allows us to operate the batteries beyond the normal conservative operating limits. For example, we operate the batteries with higher DoD values with no appreciable increase in battery degradation, and we operate with optimized charge and discharge rates without increase in internal battery temperature.

For more information, click on this link.