Chapter 15 Energy Storage Management Systems

CHAPTER 15

ENERGY STORAGE MANAGEMENT SYSTEMS

Tu Nguyen, Ray Byrne, David Rosewater, Rodrigo Trevizan

Sandia National Laboratories

Abstract

Over the last decade, the number of large-scale energy storage deployments has been increasing

dramatically. This growth has been driven by improvements in the cost and performance of energy

storage technologies, the need to accommodate renewable energy generation, as well as incentives

and government mandates. Energy management systems (EMSs) are required to utilize energy

storage effectively and safely as a flexible grid asset that can provide multiple grid services. An

EMS needs to be able to accommodate a variety of use cases and regulatory environments.

Key Terms

Arbitrage, battery management system (BMS), customer demand charge reduction, device

management system (DMS), distribution deferral, energy management system (EMS), energy

storage, energy time shift, frequency regulation, optimal operation, power conversion system

(PCS), renewable, renewable smoothing, safety, small signal stability, state-of-charge (SOC),

state-of-health (SOH), transmission deferral, voltage support

1. Introduction

Energy storage applications can typically be divided into short- and long-duration. In short-

duration (or power) applications, large amounts of power are often charged or discharged from an

energy storage system on a very fast time scale to support the real-time control of the grid. In long-

duration (or energy) applications, large amounts of energy are supplied to and pulled from the grid

on much slower time scale. Some examples of power applications include frequency regulation,

voltage support, small signal stability, and renewable smoothing. Energy applications include

energy arbitrage, renewable energy time shift, customer demand charge reduction and transmission

and distribution deferral. More details on energy storage applications are discussed in Chapter 23:

Applications and Grid Services.

There are two main requirements for the efficient operation of grid storage systems providing the

above applications and services:

1. Optimal control of grid energy storage to guarantee safe operation while delivering the

maximum benefit

2. Coordination of multiple grid energy storage systems that vary in size and technology while

interfacing with markets, utilities, and customers (see Figure 1)

Therefore, energy management systems (EMSs) are often used to monitor and optimally control

each energy storage system, as well as to interoperate multiple energy storage systems. This

chapter provides an overview of EMS architecture and EMS functionalities. While it is a high-

level review of EMS, it can be the starting point for any further reading on this topic.

Chapter 15 Energy Storage Management Systems

Figure 1. Energy Management System Overview

1.1. Energy Management System Architecture Overview

Figure 1 shows a typical energy management architecture where the global/central EMS manages

multiple energy storage systems (ESSs), while interfacing with the markets, utilities, and

customers [1]. Under the global EMS, there are local EMSs that are responsible for maintaining

safe and high-performance operation of each ESS. Just as an ESS includes many subsystems such

as a storage device and a power conversion system (PCS), so too a local EMS has multiple

components: a device management system (DMS), PCS control, and a communication system (see

Figure 2). In this hierarchical architecture, operating data go from the bottom to the top while

commands go top to bottom. For example, in the case of a battery energy storage system, the

battery storage modules are managed by a battery management system (BMS) that provides

operating data such as the state of charge, state of health, battery cell temperature [2]. These data,

together with the operating data of the PCS, are given to the local EMS for calculating the charge

or discharge power that are sent to the PCS as power commands. While delivering these required

powers, the PCS also interfaces with the BMS to ensure that none of the battery limits are violated.

In a highly centralized architecture, the optimal dispatches (i.e., power commands) are calculated

at the control center and sent to each local EMS. In a highly decentralized architecture, the central

EMS may not exist, therefore, EMS functions are only performed at the local EMSs.

Chapter 15 Energy Storage Management Systems

Figure 2. Energy Management System Hierarchy Architecture

1.2. Storage Device Management

The DMS includes a set of functions (software) that are responsible for: 1) safe operation, 2)

monitoring and state estimation, and 3) technology specific functions (such as conditioning cycles

to prolong life in some battery technologies) (see Figure 3). These DMS functions are designed to

maintain safe operation and high performance of the storage device as well as to provide operating

data to the EMS and PCS. They are often implemented on a DMS device (hardware) that is capable

of sensing, monitoring, control, and communication.

Figure 3. Device Management System Functions

1.2.1. Ensuring safe operation of energy storage device

Grid-scale ESSs can store a significant amount of energy. Therefore, safety mechanisms, either

passive or active, are required to prevent that energy from causing a hazard.

Monitoring

and State

Estimation

Diagnosis

and Safety

Technology

Specific

Functions

Measure (voltage, current,

temp) and estimate the

device states (SOC, SOH)

Ensure safety of the device

through active and passive

protection

Flow battery electrolyte

rebalancing or Li-ion cell

balancing

Chapter 15 Energy Storage Management Systems

1.2.1.1. Passive Safety

Passive safety includes the protection measures that do not do any work until they are triggered to

action. They are used to minimize the damage to the storage device and to the environment in

worst-case scenarios including short-circuits, thermal runaway, and hazardous chemical leakage.

Energy storage devices are typically protected against short-circuit currents using fuses and circuit

breakers. Thermal isolation or directed channeling within electrochemical packs is often employed

to prevent or slow the propagation of thermal runaway in Lithium-ion (Li-ion) batteries. Vanadium

redox flow batteries (VRFB) are designed to prevent the leakage of the electrolyte into the

environment through secondary and tertiary containment. Enclosures for flywheels are often

reinforced and located underground to contain the destructive kinetic energy released by a

catastrophic failure.

1.2.1.2. Active Safety

Active safety includes the protection measures that have control and monitoring capabilities. They

are used to protect storage device against undesirable working conditions such as over-charge,

over-discharge, and over-temperature that significantly reduce the life of the device [3]. The

fundamental unit of an active protection mechanism is the feedback control process where:

1. data is collected from the process being controlled,

2. the controller decides what protection actions are needed,

3. actions are implemented though actuators such as circuit interrupts, the power conversion

system, or even fire suppression systems.

For example, a BMS active protection mechanism will disconnect a battery module if its

voltage/current/temperature limits are violated or a DMS will take a flywheel off-line if its rotor

speed exceeds a threshold.

1.2.1.3. Fault Diagnosis

Some faults are easy to diagnose, such as when a smoke detector activates. Others are more

difficult, such as an internal short circuit in a Li-ion battery. Early fault identification or fault

prediction can enable a DMS to anticipate when a device failure may occur in the future by

identifying the precursors to such events. Fault diagnosis methods can be classified into

knowledge-based methods, model-based methods, and data-driven methods [4]. This is an active

area of ongoing research [5].

1.2.2. Monitoring and State Estimation

The DMS measures quantities such as voltages, currents and temperatures, and estimates the

quantities or device’s states that cannot be measured directly. The following sections describe the

three principal states of an energy storage device.

Chapter 15 Energy Storage Management Systems

1.2.2.1. State-of-Charge Model

The state-of-charge (SOC) is the ratio between the remaining energy and the maximum energy

capacity of an ESS while cycling [6]. In a small number of energy storage technologies, the SOC

can be measured directly, but in general the SOC can only be estimated through other measurable

parameters. For instance, the SOC of a pumped hydro plant can be determined directly from the

reservoir level while the SOC of electrochemical batteries such as Li-ion or lead-acid batteries,

can only be estimated through voltage and current measurements.

The parameters needed for estimation of the SOC differ for various energy storage technologies.

Table 1 summarizes the required parameters for estimating SOC of several common storage

technologies. In some cases, the SOC can be estimated using a simple model. In other cases, a

more sophisticated model may be required, but there is a trade-off between accuracy and

computational cost. A simple model is easy to develop and implement but might lead to large

errors, whereas a more sophisticated model would provide better accuracy at the expense of higher

computational requirements and more difficulty in development and implementation. A

comprehensive overview of SOC modeling can be found in “Battery Energy Storage Models for

Optimal Control” [6].

Table 1. Measurable Quantities for SOC Calculation

Technology Measurable parameters for SOC calculation

Compressed air energy storage

(CAES)

Pressure, volume, temperature, discharge profile

Electrochemical batteries Voltage, current, temperature, age

Flywheels Rotor speed, moment of inertia

Pumped hydro Reservoir level

Superconducting magnetic energy

storage (SMES)

Current, inductance

Ultra-capacitors Voltage, capacitance/impedance, temperature

Vanadium redox flow battery Voltage, temperature, electrolyte concentration

1.2.2.2. State-of-Heath Model

The state-of-health (SOH) is the present health divided by the initial health of an energy storage

device [6]. Health is measured differently in different technologies, but energy capacity is the most

commonly used proxy parameter. At some critical SOH, the battery becomes unusable or

unreliable for given applications and should be replaced. The SOH of an electrochemical cell can

be estimated from the ratio of its current capacity to its rated capacity. In lead-acid batteries for

example, at approximately 80% of initial capacity the batteries start to become increasingly

unreliable for backup power applications and should be replaced to continue to supply this service.

Thus, 80% of rated capacity is considered to be 0% SOH for many applications and battery types.

The most straightforward method for estimating SOH of batteries is to measure the impedance, as

it is generally proportional to the capacity loss [7] [8] [9]. In many cases, this method has high

error when the capacity loss and impedance are not well correlated. Model-based methods [10]

[11] and data-driven methods [12] [13] can be employed to provide more accurate estimation.

Chapter 15 Energy Storage Management Systems

1.2.2.3. Thermal Models

In many energy storage systems designs the limiting factor for the ability to supply power is

temperature rather than energy capacity [6]. This is clearly the case in thermal storage

technologies, where temperature can be used as a direct measurement of SOC, but this is also the

case in many battery systems. Batteries can reach a high temperature limit long before they reach

a low voltage limit on discharge, meaning that the EMS needs a thermal model of the batteries to

correctly predict battery operational limitations.

1.2.3. Technology Specific Functions

1.2.3.1. Equalization of Battery Cells

For multi-cell battery packs, unequal cell voltages increase the risk of over-charge/over-discharge.

To prevent these problems, the DMS must monitor and periodically equalize the battery cell

voltages. There are two classes of cell equalization (i.e., cell balancing) methods:

Passive balancing – includes two steps: 1) fully charge the battery pack, and 2) remove the

excess charge from the cells with highest voltages through passive resistors until their voltages

reduce to a reference value. This method is relatively inexpensive, but it is potentially

inefficient and does nothing for abnormally low voltage cells. Also, if the cooling system

cannot effectively dissipate the heat from the passive resistors, it increases the cell temperature

thereby reducing the battery life.

Active balancing – the charge is transferred from higher voltage cells to lower voltage cells.

Switched capacitors, inductors/transformers, and power electronics (converters) can be utilized

to redistribute the charge between cells. The advantages of this method are that it is more

efficient, it reduces heat generation, and it addresses both high and low voltage cells. However,

it involves more expensive components, more complicated controls, and increases the

instrumentation cost. It also introduces additional failure mechanisms into the module design.

1.2.3.2. Electrolyte Rebalancing of Vanadium Redox Flow Batteries

For VRFBs, electrolyte rebalancing between the two half-cells is important to prevent differential

ion transfer across the membrane and side reactions lowering the efficiency of the battery [14].

The electrolyte imbalance is often detected by comparing the SOC of both half-cells. Once the

electrolyte imbalance is quantified, a chemical reductant can be added to the positive electrolyte

to balance the oxidation states [14].

2. Power Conversion Control

The majority of energy storage devices employ a direct current (DC) interface. Therefore, a PCS

is required to integrate with the alternating current (AC) power grid. The purpose of the PCS is to

provide bi-directional conversion and electrical isolation.

Power conversion system architectures are described in more details in Chapter 13: Power

Conversion Systems. The PCS management system often includes at least two levels of control:

primary and secondary. The primary control (low level) includes the module level controllers that

generate the drive and gate signals for the power converters’ semiconductor switches given the

operating mode from the secondary control and the states of PCS and energy storage device. The

secondary control (high level) specifies the operating mode of the system given the power

Chapter 15 Energy Storage Management Systems

commands (e.g., charge and discharge rate) from the EMS and the energy storage states (e.g., SOC

and temperature) from the DMS. Because the primary control of PCS is covered in Chapter 13,

this chapter only describes the operation of the secondary control.

2.1. Secondary Control

The secondary control performs the high-level management that determines the operating mode

for each of the power converters. The three most common modes are: charging, discharging, and

standby.

2.1.1. Charging Mode

Charging mode occurs when the EMS commands the energy storage device to charge. This mode

can include a power level, in which case the charge current is controlled to deliver the commanded

power. Depending on the SOC of the device, a different charging stage is selected. Figure 4 shows

a three-stage charging scheme for batteries. The three stages are [15]:

• Bulk charge (current control) – used for fast charging when the SOC is low

• Absorb charge (voltage control) – used to prevent overcharging the battery when the

SOC is higher than a certain level

• Float charge (voltage control) – used when the battery is close to fully charged

Figure 4. Three-stage Charging Scheme for Batteries

Chapter 15 Energy Storage Management Systems

This charging scheme can also be applied to ultra-capacitors. The charging time for an ultra-

capacitor is much shorter compared to a battery because a much higher current can be tolerated

during the bulk charge stage (see Figure 4). Similarly, the charging scheme for flywheels includes:

• Torque control mode (bulk charge)

• Speed control mode (absorb charge)

• Power control mode

2.1.2. Discharging mode

This mode occurs when the EMS commands the energy storage device to discharge at a power

level to provide certain grid services. Two critical factors that must be considered for an

electrochemical battery are: (1) a higher discharge current reduces the energy capacity, and (2)

SOC lower/upper limits are often required to increase cycle life.

2.1.3. Standby mode

Standby mode occurs at the end of charging period when the maximum SOC limit has been

reached. While in this mode, the energy storage device is not significantly charging or discharging.

For an electrochemical battery, float charge might be used to compensate for the self-discharge.

For a flywheel, a small charge current is used to maintain its nominal speed.

2.1.4. Advanced inverter functions

For most applications, the above basic operating modes are sufficient to fulfill the EMS

commands. However, there has been a push to incorporate advanced inverter functions within the

PCS’s operating modes, primarily to enable increased penetrations of distributed solar generation.

Examples of these functions include:

• Volt-var (voltage support)–regulate voltage by controlling reactive power output of the

PCS

• Volt-watt–regulate voltage by controlling real power output of the PCS

• Constant power factor–maintain a constant power factor at the PCS’s output

Depending on the evolution of these capabilities, some grid control functions may reside in the

PCS secondary control going forward. After the operating mode is specified by the secondary

control, control references are calculated and passed to the primary controller.

3. Communication Interface

To coordinate operations between different subsystems of an ESS, each subsystem must be

equipped with a communication interface. Fundamental requirements for a communication

interface of an ESS can be found in existing standards such as IEC 61850-7-420 and Modular

Energy System Architecture (MESA) (see Figure 5). Commercial systems often follow

standardized communication protocols. Modbus TCP is commonly used in a large number of

devices in today’s market because of its simplicity and flexibility. DNP3 plays an important part

in modern SCADA systems, which are widely used in power systems. Subsystems/devices often

use different communication protocols and, therefore, must be capable of integrating with a wide

range of devices/systems with different protocols. An example of such a platform is VOLTTRON

developed by PNNL [16].

Chapter 15 Energy Storage Management Systems

Figure 5. MESA Communication Basic Structure [17]

4. Optimal Operation of Energy Storage Systems for Various

Applications

4.1. Market Applications

The United States has seven regions that operate electricity and grid ancillary services markets. In

these regions the potential revenue of ESSs is dependent on the market products they provide.

Generally, the EMS tries to operate the ESS to maximize the services provided to the grid, while

considering the optimal operation of the energy storage device. In market areas, maximizing grid

services is typically aligned with maximizing revenue. The operation of an ESS can be viewed as

an optimization problem where maximizing revenue is the primary objective and the constraints

are the market requirements and physical limitation of the ESS (e.g., SOC limits).

4.2. Behind-the-Meter Applications

Behind-the-meter ESS optimal operations can be enabled through the EMS. The objective of the

EMS is to shift and shave the electricity usage of consumers by charging and discharging the ESS

to minimize their bills [18]. The savings often come from demand charge reduction, time-of-use

(TOU) energy charge reduction, and utilization of net-metering energy. Optimization techniques

used for market applications can also be used to achieve this objective while ensuring that storage

device constraints, such as maximum capacity, storage efficiency, running cost,

charging/discharging rate, etc., are satisfied. Most behind-the-meter optimization algorithms are

distributed or decentralized because centralized approaches are invasive to users and do not scale

well [19].

Chapter 15 Energy Storage Management Systems

5. Summary

This chapter provided an overview of EMSs needed for operating ESSs. Since ESSs as grid assets

are quite new, most system operators are still learning to efficiently operating them. The grid

operators need robust EMSs that can manage multiple technologies, and grid services in evolving

market structures. As the regulatory environment for energy storage is evolving quickly, there are

also challenges in developing generic models that work across market structures and technologies.

Even with recent progress, storage valuation/optimization continues to be challenging. Many

related areas require additional research. Examples of these areas include: 1) storage models that

fully reflect the performance and cycle life characteristics of ESSs, 2) optimization approaches for

stacked benefits, 3) energy management systems that enable the integration of massive deployment

of distributed energy resources.

Chapter 15 Energy Storage Management Systems

Tu Nguyen is a Principal Member of the Technical Staff at Sandia

National Laboratories. He is also a Senior Member of the Institute of

Electrical and Electronics Engineers (IEEE) and a former editor of

IEEE Transactions on Sustainable Energy. He received his B.S degree

in Power Systems from Hanoi University of Science and Technology,

Vietnam in 2007 and his Ph.D. degree in Electrical Engineering from

Missouri University of Science and Technology in 2014. Before

joining Sandia National Laboratories in September 2016, he worked

as a Postdoctoral Research Associate at University of Washington. In

2019, he received Outstanding Young Engineer Award from IEEE -

Albuquerque Section. His current research interests include energy storage analytics, microgrid

modeling and analysis, and the grid integration of distributed resources.

Ray Byrne is manager of the Electric Power Systems Research

Department at Sandia National Laboratories where he has been

employed since 1989. Prior serving as manager, he was a

Distinguished Member of the Technical Staff. He completed a B.S. in

electrical engineering at the University of Virginia, an M.S. in

electrical engineering at the University of Colorado, and a Ph.D. in

electrical engineering at the University of New Mexico. He also

completed an M.S. in financial mathematics (financial engineering) at

the University of Chicago. He currently serves as team lead of the

Equitable Regulatory Environment thrust area of the Sandia energy storage program. Awards

include Time Magazine invention of the year in robotics in 2001, as well as the Prize paper award

at the 2016 IEEE Power and Energy Society General Meeting for a paper on maximizing revenue

from energy storage in grid applications. He is a member of Tau Beta Pi, Eta Kappa Nu, and Sigma

Xi. He was elevated to IEEE Fellow in 2017 for contributions to miniature robotics and grid

integration of energy storage.

David Rosewater is a Principal Member of the Technical Staff at the

Sandia National Laboratories. He received a B.S. and an M.S. in

electrical engineering from Montana Tech of the University of

Montana, as well as a PhD in electrical and computer engineering from

the University of Texas at Austin. His research interests include

modeling and simulation, performance testing, safety, and

standardization of battery energy storage systems. From 2009 to 2011,

he worked at the Idaho National Laboratory developing advanced

spectral impedance measurement techniques for hybrid vehicle battery

cells. He is a Senior Member in the IEEE and currently chairs the IEEE

P2686 working group developing a recommended practice for design

and configuration of battery management systems in energy storage

applications. David holds a professional engineering license in the state of New Mexico.

Chapter 15 Energy Storage Management Systems

Rodrigo D. Trevizan is a Senior Member of Technical Staff at Sandia

National Laboratories. Rodrigo authored research papers on the

subjects of control of energy storage systems and demand response for

power grid stabilization, power system state estimation, and detection

of nontechnical losses in distribution systems. Rodrigo received a B.S.

and M.Sc. degree in Electrical Engineering from the Federal University

of Rio Grande do Sul, Brazil, in 2012 and 2014, respectively, a M.Sc.

in Power Systems Engineering from the Grenoble Institute of

Technology (ENSE3) in 2011 and a Ph.D. in Electrical and Computer

Engineering from the University of Florida in 2018.

Chapter 15 Energy Storage Management Systems

References

[1]

R. H. Byrne, T. A. Nguyen, D. A. Copp, B. Chalamala and I. Gyuk, "Energy Management and

Optimization Methods for Grid Energy Storage Systems," IEEE Access, vol. 6, no. 1, pp. 13231 -

13260, 2017.

[2]

M. T. Lawder, B. Suthar, P. W. C. Northrop, S. De, C. M. Hoff, O. Leitermann, M. L. Crow, S.

Santhanagopalan and V. R. Subramanian, "Battery Energy Storage System (BESS) and Battery

Management System (BMS) for Grid-Scale Applications," Proceedings of the IEEE,

vol. 102, no. 6,

pp. 1014-1030, 2014.

[3]

T. Reddy, Linden’s Handbook of Batteries, New York: McGraw-Hill, 2010.

[4]

X. Hu, K. Zhang, K. Liu, X. Lin, S. Dey and S. Onori, "Advanced Fault Dianosis for Lithium-ion

Battery Systems: A review of Fault Mechanism, Fault Features, and Diagnosis Procedures," IEEE

Industrial Electronics Magazine, vol. 14, no. 3, pp. 65-91, 2020.

[5]

M. K. Tran and M. Fowler, "A Review of Lithium-ion Battery Fault Diagnosis Algorithms: Current

Progress and Future Challenges," Algorithms, vol. 13, no. 3, pp. 1-18, 2020.

[6]

D. M. Rosewater, D. A. Copp, T. A. Nguyen, R. H. Byrne and S. Santoso, "Battery Energy Storage

Models for Optimal Control," IEEE Access, vol. 7, no. 1, pp. 178357 - 178391, 2019.

[7]

F. Huet, "A review of impedance measurements for determination of the state-of-charge or state-of-

health of secondary batteries," Journal of Power Sources, vol. 70, no. 1, pp. 59-69, 1998.

[8]

J. Vetter et al., "Ageing mechanisms in lithium-ion batteries," Journal of Power Sources, vol. 147,

no. 1, pp. 269-281, 2005.

[9]

H.-F. Yuan and L.-R. Dung, "Offline state-of-health estimation for high-power lithium-ion batteries

using three-point impedance extraction method," IEEE Transactions on Vehicle Technology, vol.

66, no. 3, pp. 2019-2032, 2017.

[10]

A. Bartlett, J. Marcicki, S. Onori, G. Rizzoni, X. G. Yang and T. Miller, "Electrochemical model-

based state of charge and capacity estimation for a composite electrode lithium-ion battery," IEEE

Transactions on Control System Technologies, vol. 24, no. 2, pp. 384-399, 2016.

[11]

M. Cacciato, G. Nobile, G. Scarcella and G. Scelba, "Real-time model-based estimation of SOC and

SOH for energy storage systems," IEEE Transactions on Power Electronics, vol. 32, no. 1, pp. 794-

803, 2017.

[12]

S. Sarkar, D. K. Jha, A. Ray and Y. Li, "Dynamic data-driven symbolic causal modeling for battery

performance & health monitoring," in 18th International Conference on Information Fusion, 2015.

[13]

G. You, S. Park and D. Oh, "Diagnosis of electric vehicle batteries using recurrent neural

networks," IEEE Transactions on Industrial Electronics, vol. 65, no. 6, pp. 4885-4893, 2017.

[14]

S. Corcuera and M. Skyllas-Kazacos, "State-Of-Charge Monitoring and Electrolyte Rebalancing

Methods for the Vanadium Redox Flow Battery," European Chemical Bulletin,

vol. 1, no. 12, 2012.

Chapter 15 Energy Storage Management Systems

[15]

T. A. Nguyen, X. Qiu, I. Guggenberger, M. L. Crow and A. C. Elmore, "Performance

Characterization for Photovoltaic-Vanadium Redox Battery Microgrid Systems," IEEE

Transactions on Sustainable Energy, vol. 5, no. 4, pp. 1379 - 1388, 2014.

[16]

S. Katipamula, J. Haack, G. Hernandez, B. Akyol and J. Hagerman, "VOLTTRON: An Open-

Source Software Platform of the Future," IEEE Electrification Magazine, vol. 4, no. 4, pp. 15-22,

2016.

[17]

MESA, "Open Standard for Energy Storage," 2017.

[18]

T. A. Nguyen and R. H. Byrne, "Maximizing the cost-savings for time-of-use and net-metering

customers using behind-the-meter energy storage systems," in 2017 North American Power

Symposium (NAPS), 2017.

[19]

A. Safdarian, M. Fotuhi-Firuzabad and M. Lehtonen, "A distributed algorithm for managing

residential demand response in smart grids," IEEE Transactions on Industrial Information, vol. 10,

no. 4, pp. 2385-2393, 2014.