Hybrid Distributed Wind and Battery

Energy Storage Systems

Jim Reilly,1 Ram Poudel,2 Venkat Krishnan,3
Ben Anderson,1 Jayaraj Rane,1 Ian Baring-Gould,1
and Caitlyn Clark1
1 National Renewable Energy Laboratory
2 Appalachian State University
3 PA Knowledge

NREL is a national laboratory of the U.S. Department of Energy Technical Report

Office of Energy Efficiency & Renewable Energy NREL/TP-5000-77662
Operated by the Alliance for Sustainable Energy, LLC June 2022
This report is available at no cost from the National Renewable Energy
Laboratory (NREL) at www.nrel.gov/publications.

Contract No. DE-AC36-08GO28308

 Hybrid Distributed Wind and Battery
Energy Storage Systems
Jim Reilly,1 Ram Poudel,2 Venkat Krishnan,3
Ben Anderson,1 Jayaraj Rane,1 Ian Baring-Gould,1
and Caitlyn Clark1
1 National Renewable Energy Laboratory
2 Appalachian State University
3 PA Knowledge

Suggested Citation
Reilly, Jim, Ram Poudel, Venkat Krishnan, Ben Anderson, Jayaraj Rane, Ian Baring-
Gould, and Caitlyn Clark. 2022. Hybrid Distributed Wind and Batter Energy Storage
Systems. Golden, CO: National Renewable Energy Laboratory. NREL/TP-5000-77662.
https://www.nrel.gov/docs/fy22osti/77662.pdf.

NREL is a national laboratory of the U.S. Department of Energy Technical Report

Office of Energy Efficiency & Renewable Energy NREL/TP-5000-77662
Operated by the Alliance for Sustainable Energy, LLC June 2022
This report is available at no cost from the National Renewable Energy National Renewable Energy Laboratory
Laboratory (NREL) at www.nrel.gov/publications. 15013 Denver West Parkway
Golden, CO 80401
Contract No. DE-AC36-08GO28308 303-275-3000 • www.nrel.gov
 NOTICE

This work was authored in part by the National Renewable Energy Laboratory, operated by Alliance for Sustainable
Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. Funding
provided by U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Wind Energy
Technologies Office. The views expressed herein do not necessarily represent the views of the DOE or the U.S.
Government.

This report is available at no cost from the National Renewable

Energy Laboratory (NREL) at www.nrel.gov/publications.

U.S. Department of Energy (DOE) reports produced after 1991

and a growing number of pre-1991 documents are available
free via www.OSTI.gov.

Cover Photos by Dennis Schroeder: (clockwise, left to right) NREL 51934, NREL 45897, NREL 42160, NREL 45891, NREL 48097,
NREL 46526.

NREL prints on paper that contains recycled content.

Acknowledgments
We are thankful to all project team members from partnering laboratories on the Microgrids,
Infrastructure Resilience, and Advanced Controls Launchpad project:

• Idaho National Laboratory

• Pacific Northwest National Laboratory
• Sandia National Laboratories.
We also express our sincere gratitude to our industry advisory board members for their valuable
insights and real-world test system recommendations during the March 2020 advisory board
meeting: Venkat Banunarayanan (National Rural Electric Cooperative Association), Chris Rose
(Renewable Energy Alaska), Rob Wills (Intergrid), Paul Dockrill (Natural Resource Canada),
Jeff Pack (POWER Engineers), Arvind Tiwari (GE Global Research), Kristin Swenson
(Midcontinent Independent System Operator), Jonathon Monken (PJM), and Scott Fouts (QED
Wind Power).

The authors would also like to thank the peer reviewers Jennifer King (National Renewable
Energy Laboratory) and Jack Flicker (Sandia National Laboratories) for their thorough review.

iii
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
List of Acronyms
AC alternating current
BESS battery energy storage system
DC direct current
DER distributed energy resource
DFIG doubly-fed induction generator
HVS high voltage side
Li-ion lithium-ion
LVS low voltage side
MIRACL Microgrids, Infrastructure Resilience, and Advanced Controls Launchpad
MW megawatt
NREL National Renewable Energy Laboratory
PV photovoltaic(s)
SM synchronous motor
SOC state of charge
WTG wind turbine generator

iv
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Executive Summary
For individuals, businesses, and communities seeking to improve system resilience, power
quality, reliability, and flexibility, distributed wind can provide an affordable, accessible, and
compatible renewable energy resource. Distributed wind assets are often installed to offset retail
power costs or secure long term power cost certainty, support grid operations and local loads,
and electrify remote locations not connected to a centralized grid. However, there are technical
barriers to fully realizing these benefits with wind alone. Many of these technical barriers can be
overcome by the hybridization of distributed wind assets, particularly with storage technologies.
Electricity storage can shift wind energy from periods of low demand to peak times, to smooth
fluctuations in output, and to provide resilience services during periods of low resource
adequacy.

Although interconnecting and coordinating wind energy and energy storage is not a new concept,
the strategy has many benefits and integration considerations that have not been well-
documented in distribution applications. Thus, the goal of this report is to promote understanding
of the technologies involved in wind-storage hybrid systems and to determine the optimal
strategies for integrating these technologies into a distributed system that provides primary
energy as well as grid support services. This document achieves this goal by providing a
comprehensive overview of the state-of-the-art for wind-storage hybrid systems, particularly in
distributed wind applications, to enable distributed wind system stakeholders to realize the
maximum benefits of their system. As battery costs continue to decrease and efficiency continues
to increase, an enhanced understanding of distributed-wind-storage hybrid systems in the context
of evolving technology, regulations, and market structure can help accelerate these trends.

v
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Table of Contents
1 Introduction ........................................................................................................................................... 1
1.1 Advantages of Hybrid Wind Systems ........................................................................................... 1
1.2 Considerations and Challenges of Hybrid Wind Systems ............................................................. 4
2 Wind-Storage Hybrids: Possible Configurations .............................................................................. 6
2.1 AC-Coupled Wind-Storage Hybrid Systems ................................................................................ 8
2.2 DC-Coupled Wind-Storage Hybrid System .................................................................................. 8
2.3 Comparison of AC and DC Configurations ................................................................................ 10
3 Hybrid System Controls: Stable Integration and Maximum Utilization ........................................ 13
3.1 Distributed Hybrid System Controls ........................................................................................... 14
3.1.1 Essential Reliability Services and Stability .................................................................... 14
3.1.2 Frequency Response....................................................................................................... 14
3.1.3 Voltage and Reactive Power Support............................................................................. 15
3.1.4 Flexibility and Economic Grid Services ........................................................................ 15
3.1.5 Enabling Fast and Accurate Response ........................................................................... 16
3.2 Modeling Controls and Time Scales ........................................................................................... 16
4 Operation and Dispatch of Wind-Storage Hybrids.......................................................................... 18
4.1 Wind-Storage Hybrids Optimal Dispatch ................................................................................... 18
4.2 Wind-Storage Hybrids Supporting Black Start ........................................................................... 19
5 Techno-Economic Sizing of Wind-Storage Hybrids ....................................................................... 22
5.1 Storage Cost Models ................................................................................................................... 22
5.2 Wind-Hybrid Models .................................................................................................................. 23
6 Conclusion .......................................................................................................................................... 25
References ................................................................................................................................................. 27

vi
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
List of Figures
Figure 1. Possible wind-storage hybrid configurations................................................................................. 7
Figure 2. Dominant wind turbine technologies. ............................................................................................ 7
Figure 3. Common topology of an AC-coupled wind-storage hybrid system. ............................................. 8
Figure 4. Schematic of DC-coupled photovoltaic-plus-storage systems. ..................................................... 9
Figure 5. Four-port DC/DC converter for an isolated system. .................................................................... 10
Figure 6. Hierarchy of hybrid system control ............................................................................................. 13
Figure 7. Dispatch of photovoltaics-plus-storage system on a typical day ................................................. 19
Figure 8. Distributed black start of wind turbines in an island mode. ........................................................ 20
Figure 9. Battery cost projections for 4-hour Li-ion systems ..................................................................... 23

vii
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
1 Introduction
A distributed hybrid energy system comprises energy generation sources and energy storage
devices co-located at a point of interconnection to support local loads. Such a hybrid energy
system can have economic and operational advantages that exceed the sum of the services
provided by its individual components because of synergies that can exist between the
subsystems. The coordination between its subsystems at the component level is a defining
feature of a hybrid energy system. Recently, wind-storage hybrid energy systems have been
attracting commercial interest because of their ability to provide dispatchable energy and grid
services, even though the wind resource is variable. Building on the past report “Microgrids,
Infrastructure Resilience, and Advanced Controls Launchpad (MIRACL) Controls Research
Road Map,” which highlights the challenges and opportunities for distributed wind grid
integration and control mechanisms, this report initiates and establishes a baseline for future
research on wind-storage hybrids in distribution applications (Reilly et al. 2020).

The objective of this report is to identify research opportunities to address some of the challenges
of wind-storage hybrid systems. We achieve this aim by:

• Identifying technical benefits, considerations, and challenges for wind-storage hybrid

systems
• Proposing common configurations and definitions for distributed-wind-storage hybrids
• Summarizing hybrid energy research relevant to distributed wind systems, particularly
their control, operation, and dispatch
• Suggesting strategies for sizing wind-storage hybrids
• Identifying opportunities for future research on distributed-wind-hybrid systems.

A wide range of energy storage technologies are available, but we will focus on lithium-ion (Li-
ion)-based battery energy storage systems (BESS), although other storage mechanisms follow
many of the same principles. The Li-ion technology has been at the forefront of commercial-
scale storage because of its high energy density, good round-trip efficiency, fast response time,
and downward cost trends.

1.1 Advantages of Hybrid Wind Systems

Co-locating energy storage with a wind power plant allows the uncertain, time-varying electric
power output from wind turbines to be smoothed out, enabling reliable, dispatchable energy for
local loads to the local microgrid or the larger grid. In addition, adding storage to a wind plant
can enable grid-forming or related ancillary grid services such as inertial support and frequency
responses during transitions between grid-connected and islanded modes. A hybrid system can
also increase revenue by storing rather than wasting energy that cannot be used because of
system rating limits or the absence of loads.

Additional benefits of hybrid energy systems can come from sharing components between other
generation sources such as inverters and optimizing electrical system ratings and interconnection
transformers. It is worth noting, however, that limiting the full system rating can result in a
decrease in revenue. For example, the use of storage during periods of high wind energy output

1
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
might be limited restricted because of a limit on the total power output of the combined system.
For this reason, rigorous assessments—including hybrid system modeling, planning, and sizing
of the components—are critical to maximize system benefits based on the application, expected
load, and desired grid services. An assessment should also consider the specific grid and local
weather conditions.

The following are some high-level benefits of wind-storage hybrid systems:

• Dispatchability of variable renewable resources. A storage system, such as a Li-ion

battery, can help maintain balance of variable wind power output within system
constraints, delivering firm power that is easy to integrate with other generators or the
grid. The size and use of storage depend on the intended application and the
configuration of the wind devices. Storage can be used to provide ramping services, as
has been done with wind installations in Kodiak and along the Alaskan Railbelt with
wind facilities in Anchorage or Fairbanks; for time-of-day shifting, as was deployed in
Kotzebue, Alaska; or to allow for transitions between sources, as has been deployed in
Tuntutuliak and other remote Alaskan communities. In larger grid-connected systems,
photovoltaics (PV) has a diurnal cycle that fits well with a 4-hour storage cycle, charging
the storage device during the day to expand energy supply to, typically, evening peak
load hours. Depending on a site’s wind profile and the driver for energy services, a wind-
storage hybrid system will require different considerations for storage size. These
requirements have prompted storage asset developers and owners to look to new battery
technologies beyond the short-duration Li-ion systems deployed so far (Energy Storage
Systems, Inc. 2016). Various technologies are evolving to provide long-duration storage.

• Economic impact. The demand for electricity varies with time, changing with time of
day, weather, and various socioeconomic factors. Similarly, the price of electricity also
varies with system conditions, congestion, and time of day. A storage system can
leverage this varying pricing to schedule its charging and discharging to increase the
effectiveness of energy arbitrage. Research has also shown that arbitrage can be achieved
across energy and ancillary markets to improve the economics of wind-storage hybrids
(Das, Krishnan, and McCalley 2015). This economic value proposition further improves
for a hybrid resource, which can rely on low-cost renewable energy (or no-cost renewable
energy at times when curtailment requires shutting down wind turbines) to charge and
sell in the larger grid’s energy and ancillary markets. The benefits of a hybrid system
depend on the resource configuration and specific context of the project, and research is
needed to tailor hybrid solutions to specific locations and grid scenarios.

• System flexibility. Modern energy systems require electricity to maintain constant

frequency and voltage. However, wind energy is a variable resource that, when combined
with a variable load, increases the overall power variability of the energy system. Hence,
maintaining a balance of supply and demand requires balancing engineering and
economics. To achieve this balance, balancing authorities have look-ahead generation
scheduling and operational planning, starting from day-ahead unit commitment and
dispatch and continuing to real-time dispatch at 5 minutes. This scheduling ensures that
generation resources have sufficient flexibility (e.g., headroom capacity and ramping
capabilities) to meet the energy, load following, ramping, and ancillary service

2
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 (regulation and spinning reserves) requirements for reliability. As system size decreases,
there are fewer devices on the grid and less need to stabilize frequency and voltage,
requiring faster system response even below 1 hertz (Hz). Regarding flexibility, hybrid
wind systems can provide:
o Load leveling or energy shifting to avoid steep ramps and negative prices caused
by excess renewable generation
o Complementarity with solar, thereby mitigating issues such as the duck curve
(California ISO 2016), with its mismatch between generation and load, leading to
severe morning and evening net-load ramps
o Ramping up or down to support the increase in the frequency and severity of
ramping events in the grid related to increasing variable renewable contributions.
With improved wind forecasting and adequate energy storage, hybrid systems can
provide ramping capability, thereby avoiding generation scarcity events and real-
time price spikes that would otherwise necessitate expensive gas generation starts.
• Enhanced grid stability. In a power system, especially localized grids, generation and
demand must remain balanced to maintain stability. This balance ensures that voltage,
frequency, and small-signal oscillations remain within acceptable North American
Electric Reliability Corporation and American National Standards Institute levels. A
storage system can function as a source as well as a consumer of electrical power. This
dual nature of storage combined with variable renewable wind power can result in a
hybrid system that improves grid stability by injecting or absorbing real and reactive
power to support frequency and voltage stability.

• Grid reliability and resilience. A distribution hybrid system with local loads can also
function as a microgrid, and the microgrid, with appropriate controls, can operate in both
grid-tied and islanded modes. A microgrid with on-site renewable generation and storage
can enhance grid resilience and ensure power supply to critical loads during major
physical or cyber disruptions. Additionally, a distributed wind system can support a
stable and reliable grid when hybridized with storage as well as dispatchable generation
as appropriate. Further reliability improvements can be made by adding redundancy to
the system (by physically distributing assets with parallel capabilities) or using advanced
controls to provide services (such as black start capabilities).

• Economics with common and standardized components. Most modern utility-scale

wind turbines have power converters to allow for variable-speed operation of the wind
generator for maximum efficiency and to convert the power to grid-standard voltage and
frequency. The power converter may include AC/DC and DC/AC conversion. A battery
storage system also requires such power converters to regulate charging/discharging.
Other relevant services that these power converters can provide include ramp rate,
decoupled control of real and reactive power for frequency and voltage support, and DC-
to-AC power conversion in an AC grid-tied scenario. These services are also relevant to
many other distributed energy resources (DERs). Battery systems can utilize the existing
power converter and inverter hardware infrastructure in a wind turbine, and the
components can be optimally sized for their intended uses. The incremental cost of the

3
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 hardware, even when the component size is increased, can be an economic option for
some deployments, especially in an isolated environment or use case.

• Other benefits from the circular economy and recycling. Small-scale wind energy
developers are looking at the economics of employing used batteries from the
transportation industry. Bergey Windpower Co. is planning to use secondhand battery
systems from a nearby Nissan electric car factory to create a home microgrid system. The
Bergey Excel 15 Home Microgrid System uses 18-kilowatt-hour (kWh) recycled electric
vehicle battery packs (Bergey 2020). The batteries used in electric vehicles can be
evaluated for a range of options for reuse and recycling. The research at National
Renewable Energy Laboratory has revealed that the second use of electric vehicle
batteries is both viable and valuable (NREL 2020). NREL’s battery second-use calculator
can be used to explore the effects of different repurposing strategies and assumptions on
economics. Before batteries are recycled to recover critical energy materials, reusing
batteries in secondary applications, like the Excel 15 Home Microgrid System, is a
promising strategy (Ambrose 2020). The value propositions from the circular economy
can make wind-hybrid systems a cost-effective as well as an environmentally friendly
option for a reliable and resilient energy system.

1.2 Considerations and Challenges of Hybrid Wind Systems

Although a hybrid wind system has many benefits, it can pose operational challenges as well.
The following are some high-level considerations and challenges when considering the
deployment of a wind-storage hybrid system or upgrade of a standalone wind power plant to
include storage:

• Complicated dispatch and valuation of combined resources. A variable wind resource

can cause cycling of the battery, which can affect its life cycle (Wenzl et al. 2005; Corbus
et al. 2002). How daily cycling compares with random charge/discharge is an economic
question that may be specific to the context. The hybrid system may have challenges
associated with co-location, such as transmission constraints and inverter capacity limits.
Some of these challenges can be managed with a better forecast and control/dispatch
logic. However, a detailed assessment for specific grid scenarios and weather situations is
needed to size the hybrid systems appropriately and optimize resource utilization.
Further, the economic assessment must maximize storage utilization while reducing
curtailments and battery cycling, especially for isolated power systems (Baring-Gould et
al. 2001).

• Feasibility studies are not as defined and generic as they are for conventional
generators. For systems on a central grid, governing market rules and policy incentives
can make or break the finances of a wind-hybrid project. A co-located wind-storage
system can share infrastructure to provide reliable power at a low cost. Such a system
may also qualify for incentives such as the investment tax credit, provided it complies
with terms and conditions specific to the state, region, or country. In some states, a
battery system must get 75% of its energy from renewable energy sources such as solar
and wind to qualify for the investment tax credit. Depending on policy, the hybrid system
may or may not make sense technically and/or financially.

4
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 • The current production tax credit for wind does not consider the addition of energy
storage. There are also operational limits to hybridization. These will depend on the
available resource in a region and the ability to forecast and develop appropriate resource
bids or self-schedules (if participating in markets or central dispatch and compensation
mechanisms) to enhance the value of a hybrid system. The investment tax credit for PV
was expanded to include investments in battery storage (NREL 2018b), but the
production tax credit for wind does not include such considerations.

• Integrating multiple technologies is complex, and plug-and-play solutions are

needed to simplify design. The literature review conducted as part of this report is
intended to inform the development of control solutions to maximize the benefits of
wind-hybrid system configuration and sizing. A “plug-and-play” distributed wind turbine
system is needed to enhance the market share and realize the full potential of wind to
serve the global demand for clean energy. A defining aspect of “plug and play” is
continued innovation on par with evolving grid codes and other technology solutions.
Considering the possible range of benefits, challenges, and opportunities, this paper will explore
how wind-hybrid systems, with a current focus on wind-storage hybrid systems, can be
efficiently configured to operate within different environments. A detailed quantitative study will
be undertaken later, and results will be reported. Taking lessons learned from other hybrid
technologies (e.g., hybrid-solar or hybrid-hydro [Poudel, Manwell, and McGowan 2020]) in the
energy industry, this literature review aims to identify the opportunities and challenges of wind-
hybrid systems in various operational use cases. These use cases include isolated grids or
microgrids in island mode, grid-connected resources providing energy and ancillary services to
the grid, and the ability to transition from grid-connected to island mode.

5
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
2 Wind-Storage Hybrids: Possible Configurations
Increasingly, wind turbines are being coupled with batteries to mitigate variability and
uncertainty in wind energy generation at a second-by-second resolution. Storage may be
integrated with wind turbines in three ways:

1. Virtually, if the hardware is not co-located but is controlled as a single source

2. Physically co-located yet separately metered and dispatched as a separate source
3. Co-located behind the same meter, in which case the two components act as a singular
source with respect to the grid.
Within this context, wind-storage hybrids can also be coupled in two ways:

1. AC-coupled, in which wind and storage share a point of common coupling on an AC-bus
2. DC-coupled, in which wind and storage share a point of common coupling on a DC-bus.
As we discuss in Section 2.3, AC coupling can be done in all three storage integration cases, but
to date DC-coupled systems are exclusively behind-the-same-meter systems.

In a wind power plant, which may contain two or more wind turbines, the storage can be sited
either at the power plant level (i.e., central storage, as shown in Figure 1a) or at the individual
wind turbine level (i.e., integrated storage, as shown in Figure 1b). Individual turbine-level
storage can either be deployed as a unit behind the dedicated turbine interconnect, typically with
a lower-voltage AC connection, or integrated behind the turbine power converter, which will
take place at a DC voltage. For example, each of the 100 GE 1.6-megawatt (MW) wind turbines
at Tehachapi has 200 kWh of integrated storage (Miller 2014) in the DC link. Unlike turbines
with integrated storage that use the turbines’ existing power conversion equipment, a wind power
plant with AC-connected individual or central storage requires additional equipment such as a
dedicated power converter, switchgear, and transformer. This is one of the trade-offs that need to
be considered when choosing a storage topology and location. A study of the GE turbines at
Tehachapi builds on a precursor study (Fingersh 2003) that explored using the turbine’s
controller and power electronics system to operate an electrolyzer to generate hydrogen from
water, thereby using a component-level strategy for a hybrid system. The GE study (Miller 2014)
does not provide many details about the sizing of integrated storage and the associated power
electronics architecture; we believe this is an opportunity for future research.

6
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 a) Central storage at the plant level b) Integrated storage at each turbine

Figure 1. Possible wind-storage hybrid configurations

A hybrid system can be coupled on a common DC bus, AC bus, or both, depending on the type
of wind turbine. The four main types of wind turbines are summarized in Figure 2 (Singh and
Santoso 2011). Some of these configurations are more amenable to sharing DC-to-AC-
conversion equipment. A review paper (Badwawi, Abusara, and Mallick 2015) presents power
electronics topologies and control for hybrid systems. A good description of AC versus DC solar
coupling, including their pros and cons with reference to the solar energy industry, is
documented in (Marsh 2019).

Figure 2. Dominant wind turbine technologies.

Source: Singh and Santoso (2011)
Key: DFIG – doubly-fed induction generator; IM – induction motor; SM – synchronous motor.

7
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
2.1 AC-Coupled Wind-Storage Hybrid Systems
In an AC-coupled wind-storage system, the distributed wind and battery connect on an AC bus
(shown in Figure 3). Such a system normally uses an industry-standard, phase-locked loop
feedback control system to adjust the phase of generated power to match the phase of the grid
(i.e., synchronization and control). To integrate electrical power generated by DERs efficiently
and safely into the grid, grid-side inverters accurately match the voltage and phase of the
sinusoidal AC waveform of the grid (Denholm, Eichman, and Margolis 2017).

An AC-coupled wind-storage system has some advantages over DC-coupled systems. AC-
coupled systems use legacy hardware and standardized equipment commonly available in the
market, making them relatively easy to install. In an AC-coupled system, energy stored by the
battery can be independent of the output of the wind turbine, allowing the combined system to be
sized and operated based on the energy and grid services that the project will provide. Two
independent units will also have a high total capacity because both units can provide full output
simultaneously. In this scenario, the battery storage can have fewer charging/discharging cycles
than it would in the DC-coupled system. However, this may not always be the case if the hybrid
system is in an isolated mode of operation.

For Type 3 and Type 4 wind turbines (see Figure 2), an AC-coupled wind-storage system would
require two inverters: one DC/AC one-way inverter for the wind (after the DC/AC converter)
and a bidirectional DC/AC inverter for the battery system for charging/discharging, as depicted
in an example system shown in Figure 3. The power conversion equipment is costly but allows
the full capacity of both generation sources to be used.

Figure 3. Common topology of an AC-coupled wind-storage hybrid system.

Source: Adapted from Denholm, Eichman, and Margolis (2017)

2.2 DC-Coupled Wind-Storage Hybrid System

In a DC-coupled wind-storage system, the wind turbine and BESS are integrated at the DC link
behind a common inverter, as detailed for PV by Denholm, Eichman, and Margolis (2017) and
adapted for wind-plus-storage systems in Figure 4. The electricity generated by the wind turbine
is rectified and coupled with the BESS, and the battery is maintained through the DC-DC
converter. The grid-side inverter can be one-directional (i.e., DC/AC) or bidirectional, and the

8
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
battery can store energy from just the turbine or from both the turbine and the grid. This is shown
in Figure 4 and discussed in further detail for PV by Denholm, Eichman, and Margolis (2017).

Figure 4. Schematics of DC-coupled wind-storage systems.

Source: Adapted from Denholm, Eichman, and Margolis (2017)

In a DC-coupled system using a one-directional DC/AC inverter, the battery can only be charged
using the wind turbine. Some states and federal programs offer tax credits for such systems
(NREL 2018b). With a bidirectional inverter, the stacked value streams for the BESS may
increase because it can serve energy-shifting functions and participate in energy arbitrage. In
addition, such a system may qualify for tax credits and other incentives available to one-
directional inverters.

Type 3 and Type 4 wind turbines share many of the same components as energy storage systems
and can often share a significant portion of AC/DC and DC/AC infrastructure, with a DC link
capacitor in between (Miller 2013, 2014). In this case, a battery with a DC output can be
connected directly or via its own bidirectional DC-DC converter for power regulation. This type
of storage is known as an integrated storage in the DC link of the wind turbine. A recent master’s
degree thesis at the Norwegian University of Science and Technology evaluated he modular
multilevel converter for medium-voltage integration of a battery in the DC link (Rekdal 2018). A
multilevel converter is a method of generating high-voltage waveforms from lower-voltage
components. Modular multilevel converters are considered a promising battery interface as they
have very high efficiency; excellent AC waveforms; and a scalable, modular structure, while also
allowing for the use of semiconductors with low ratings. However, there is not much research
available in the public domain about how to optimize the size of integrated storage for given
wind power plant sizes and energy resources.

9
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
For hybrid systems, there has been recent interest in revisiting multiport DC/DC converters to
share power electronics components, simplify operational logics, and develop compact/efficient
architectures. For an isolated application, Zeng et al. (2019) present a four-port DC/DC converter
that can handle wind, PV, battery storage, and loads (see Figure 5). The authors claim that their
multiport converter has the advantage of using a simple topology to interface with sources of
different voltage/current characteristics.

Figure 5. Four-port DC/DC converter for an isolated system.

Source: Zeng et al. (2019)
Key: WTG = wind turbine generators; LVS = low voltage side; HVS = high voltage side; BAT = battery storage; PV =
solar photovoltaic.

2.3 Comparison of AC and DC Configurations

Both AC and DC wind-storage hybrids have advantages and disadvantages that depend on the
details of the specific installation. For example, direct-drive, Type 4, full-conversion wind
turbines (e.g., EWT, Enercon) are suitable for integrated AC and DC coupling, as both an AC
and DC bus exist in their typical configuration. However, conventional Type 1 turbines are more
suited for AC coupling because of the lack of a DC bus in their typical configuration. The
configuration also depends on the specifics of the project and economic factors such as market
price for energy and grid services, and tax credit policies for hybrid plants. The following is a
high-level comparison of characteristics of AC and DC hybrid configurations:

• AC system maturity and battery independence. AC-coupled systems use standard AC

interconnection equipment available in the market that is easy to install. This allows more
flexibility in the sizing of the wind turbine and battery, both in terms of power and
capacity. In an AC-coupled system, energy stored by the BESS can be independent of the
output of the individual wind turbines.

• DC systems for smaller and distributed hybrids. As the size of the DER project
increases, a clear demarcation begins to emerge between the AC and DC coupling based
on the economics of the project and other nontechnical constraints. A DC-based system is
known to interface better with other DC-based distributed generation on the system, but
currently is limited to rather small sizes. Such a system can communicate and supply

10
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 power over a single distribution line, and interconnection with other on-site DC
generation sources such as PV is simplified. Experts on the future of direct current in
buildings (Glasgo, Azevedo, and Hendrickson 2018) suggest that the two biggest barriers
for DC coupling are industry professionals unfamiliar with DC and comparatively small
markets for DC devices and components.

• Trends in power-electronic-interfaced sources and loads favoring DC coupling.

Recent advances achieved in power electronics—which made DC voltage regulation a
simple task—have increased the penetration of DC loads and sources and encouraged
researchers to reconsider DC distribution for portions of today’s power system to increase
overall efficiency (Elsayed, Mohamed, and Mohammed 2015). Although the
conventional rotating-electric machine-based power system predominantly operates via
AC transmission, microgrids intrinsically support DC power. Many distributed energy
systems are driven by static electronic converters (Gu, Li, and He 2014). Compared to its
AC counterpart, a DC microgrid has the potential to achieve higher efficiency, power
capacity, and controllability. Because of these advantages, a DC-based power system
with DC-coupled wind and storage is an enabling technology for microgrids, especially
in small-scale residential applications such as green buildings, sustainable homes, and
energy access applications in areas inaccessible by the national grid.

• System efficiency and cost. An AC-coupled system will have lower roundtrip efficiency
for battery charging than a DC-coupled system, which charges the battery directly and
does not have power flow through two inverters (one wind turbine inverter and one BESS
inverter). However, only a portion of the wind turbine power produced goes into the
storage and is thus subject to the losses. An NREL study based on a utility-scale PV
project suggests that using DC coupling rather than AC coupling results in a 1% lower
total cost (Fu, Remo, and Margolis 2018), which is the net result of cost differences
between solar inverters, the structural and electrical balance of system, labor, developer
overhead, sales tax, contingency, and profit. For an actual project, however, cost savings
may also need to account for additional factors such as retrofit considerations, system
performance, design flexibility, and operations and maintenance.
Further design considerations for different hybrid configurations to promote reliability and
flexibility include:

• DC systems. A DC-coupled wind-storage system requires one less inverter than an AC-
coupled system (see Figure 3), which reduces wiring and housing costs as well as
conversion losses. Type 3 and Type 4 wind turbines also have hardware components that
can be used for DC coupling at the DC link. Because the BESS is connected directly to
the distributed wind turbine system, excess generation that might otherwise be clipped by
an AC-coupled system at the inverter level can be sent directly to the BESS, which could
improve system economics (DiOrio and Hobbs 2018).

• AC systems. AC systems use off-the-shelf components, and they do not require

technology-specific modification or engineering. In addition, AC system components are
modular, which reduces retrofit costs, and they stack well with each other compared to a
DC-coupled system. They require less maintenance time because, unlike a DC-coupled

11
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 system, batteries do not need to be installed next to the bidirectional inverter. AC-coupled
systems can also use larger battery racks per megawatt-hour of battery capacity and thus
reduce the number of heating, ventilating, and air-conditioning and fire-suppression
systems in the battery containers (Fu, Remo, and Margolis 2018). These systems allow
manageable battery health monitoring and state-of-charge (SOC) planning with an
independent battery management system that has its own bidirectional DC-AC inverter
and can use redundant inverters that provide increased reliability and available capacity.

• Retrofit to add storage to existing generation. For a retrofit scenario with individual
wind turbines (i.e., adding battery storage to existing wind turbine generators), an AC-
coupled BESS may be the only practical option because of the extensive turbine-specific
modifications that would need to be implemented for a DC-coupled system.

• Synchronization. A hybrid system coupling in a DC common bus does not require the
synchronism an AC bus configuration requires. The voltage is fixed for all subsystems in
the hybrid system, and the current from each subsystem is controlled independently. A
battery bank connected directly or through a DC/DC link can regulate the DC bus
voltage. The subsystem can independently perform maximum power point tracking by
using an AC/DC converter for the wind turbine and DC/DC converter for the PV case. A
common DC/AC inverter maintains the voltage across the load.
The wind and solar industries have many similarities for AC- and DC-coupled systems.
Badwawi, Abusara, and Mallick (2015) present a summary of research regarding power
electronic topologies and control. Marsh (2019) also provides a good description of AC versus
DC solar coupling, including pros and cons related to the solar energy industry. A co-located
wind-storage system can share some components and leverage some transmission-level
constraints.

To expand on the grid support capabilities of wind-storage hybrids, GE conducted a study on

wind power plants with integrated storage on each turbine rather than central storage, along with
an extra inverter and transformer for redundancy (Miller 2014). There are always some trade-offs
involved in choosing a storage topology. The GE study does not present details about sizing
integrated storage but rather demonstrates the benefits of the technology. As part of the
MIRACL project, NREL plans to explore integrated storage sizing and configurations using
theoretical and computational approaches through desktop simulations and power-hardware-in-
the-loop validation.

12
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
3 Hybrid System Controls: Stable Integration and
Maximum Utilization
A defining feature of hybridization is the ability to coordinate generation to effectively balance
varying load or net load (load minus variable renewables), resulting in an economic dispatch of
the generation and storage assets. This is possible by controlling individual devices (e.g.,
generators, storage, load) within the hybrid system, or by controlling the hybrid system as a
single unit, providing a precise power output to benefit the overall power system. A system-level
controller utilizes algorithms to issue commands to each device within the hybrid system based
on load and variable renewable forecasts.

Tertiary (supervisory):
∆E(grid), set points from
grid dispatcher

Secondary (supervisory):
∆V, ∆f at hybrid plant or
microgrid level

Primary: ∆Load and

dynamics at individual
asset level

Figure 6. Hierarchy of hybrid system control

Typically, controls use a hierarchical architecture as well as two-way communication from

individual subsystems or devices in a hybrid system to achieve the best hybridization outcomes.
A typical hierarchical control can be classified into three levels, as shown in Figure 6. The
primary control manages the load/current sharing through droop control. The secondary control
responds to the steady-state error on the voltage and frequency, and the tertiary control maintains
coordination based on the status at the point of common coupling.

The objective of control is to maintain the electrical system parameters within acceptable limits
by balancing generation with demand at the hybrid system level, taking system constraints and
the health trajectory of subsystems and individual components into account. It should also be
recognized that these control functions are made at different time steps, with electrical system
parameter adjustments needing to happen very quickly whereas others, such as decisions based
on balancing load or varying renewable energy production, can typically be made over minutes
or hours.

13
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
3.1 Distributed Hybrid System Controls
Well-designed controls can enable several capabilities that improve hybrid system economics.

3.1.1 Essential Reliability Services and Stability

Hybrid controls should be flexible (or customizable) to accommodate various essential reliability
services that the hybrid system may provide. Control and coordination between the hybrid
technologies becomes more challenging as the contribution of variable renewables increases on
the grid and more is expected of hybrid systems to support grid stability. For example, toward
the end of rural extension lines or long transmission lines where voltage and frequency are more
sensitive to the dynamic load/generation (i.e., weakly interconnected systems), the phase-locked
loop measurement system, on which frequency and phase estimation and subsequent controls
rely, is known to have issues with frequency and phasor measurement, adversely affecting
stability. Therefore, the controls in a hybrid system should be able to ease and enhance the
stability of the services provided. The use of a battery to provide services such as inertial
response will also decrease mechanical load in the wind turbine (extending its life). Similarly,
wind turbines can provide damping control to offset oscillations (e.g., local, forced, or interarea),
which will be further enhanced with interconnected battery storage.

3.1.2 Frequency Response

In addition to the (natural or synthetic) inertial response to any generator outages causing
frequency drop, the grid typically uses three additional levels of frequency response: 1) primary
or governor response subject to frequency deviation beyond a “dead band”; 2) secondary
response that uses 4- to 6-second-level automatic generation control signals coming from a
central dispatcher (taking into account both frequency deviation as well as tie-line transactions)
into a variable called an area control error; and 3) tertiary response, typically coming from
additional reserves through market dispatch. For each of these services, there must be headroom
reserved from the maximum available power for both a wind turbine and PV plant. In a hybrid
plant, a battery can complement the variable renewable power and provide these frequency
response services, removing the need to curtail and reserve headroom in the wind turbine, unless
it becomes necessary for reliability reasons.

Droop control is a common way to control and coordinate multiple distributed resources in a
hybrid plant, allowing them to share power and support multiple grid services. A droop for a
resource with a rated power P(rated) in a power system with frequency f = 60 Hz is defined as:

1 ∆𝑃𝑃/𝑃𝑃(𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟) (1)
=
𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∆𝑓𝑓/60(𝐻𝐻𝐻𝐻)

For example, Xcel Energy has used wind turbine droop control for years (Porter, Starr, and Mills
2015). The most common droop setting used in many power systems is 5%, but in some cases
more aggressive 3% droop is used as well (NREL 2018b). A 5% droop means that a 5% change
in frequency would result in a 100% change in power. For a BESS system operating at 5% droop
control at a nominal frequency of 60 Hz, a decrease or increase in frequency of 3 Hz (i.e., 5/100
× 60 Hz) should deliver/absorb the rated power of the battery. However, the deliverability of the
power for a BESS and any source of generators will depend on the available headroom or the

14
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
current state of the resource (i.e., maximum generation, current generation set point, or battery
SOC).

As the contribution level of variable renewable energy grows in a microgrid, additional design
challenges emerge for the integration of BESS and appropriate levels of droop settings. A 2015
study (Weaver et al. 2015) looked at the energy storage requirements of DC microgrids with
high-penetration renewables under droop control. This study suggested that decentralized control
architecture is possible with a distributed or adaptive droop control that is subject to evolving
net-load disturbance or area control error fluctuations, and, consequently, the energy storage
requirements in a microgrid may be minimized with the optimal choice of droop settings.
Another study related to DC microgrids (Zhao and Dörfler 2015) demonstrated that the droop
control strategy can achieve fair and stable load sharing (even in the presence of actuation
constraints) or follow set points provided by the economic dispatch.

3.1.3 Voltage and Reactive Power Support

In addition to frequency support, voltage and reactive power control are another criterion for
hybrid plants. The control may be to maintain a specific voltage set point or power factor at the
point of interconnection or to maintain the voltage within American National Standards Institute
limits of 0.95 to 1.05 per unit (or 0.90 to 1.1 per unit) for certain locations or contingency
situations. Type 1 and Type 2 wind turbines will typically need external reactive power
resources, such as capacitor banks and static synchronous compensators (STATCOMS) to
provide reactive support and voltage control. Type 3 wind turbines come with a limited range
(+/- 30%) of reactive power control, given the size of the rotor-side converter. Type 4 turbines
come with a full range of reactive power capabilities, just like a PV inverter, and could be
operated like a STATCOM even when the turbine is not producing real power. However, given
the need to curtail real power to produce reactive power, the storage in the hybrid plant can
alleviate the issue by providing reactive power support. With the help of energy storage, the
hybrid plant’s range of reactive power control can be increased and maximized to support the
required power factor or voltage performance.

3.1.4 Flexibility and Economic Grid Services

Although all the services mentioned are needed to ensure that the hybrid power plant can be
integrated into the grid and support grid reliability and stability, the most important factors from
a project developer’s perspective are the nest utilization of all assets and maximizing profit. To
do that, the modelers will have to understand every possible combination of individual devices
suited for a particular location and develop optimization and management algorithms that can
harness the synergies among various components. For example, among various objective
functions of the hybrid resource optimization and control, reducing energy from a diesel genset
is a desirable outcome, especially in a remote, isolated grid scenario. Efficient use of fuel, or
hedges against winter fuel shortages, should also be accounted for when designing and operating
hybrid plants. For example, this is the case in Alaskan microgrid designs.

In a grid-tied scenario, maximizing the revenue from energy and ancillary markets will be key.
The increased use of variable renewable energy resources has also increased the necessary
reserve, regulation, and ramping capability needed in the grid. A wind-storage hybrid plant is
well-suited to provide these flexibility and ancillary services in addition to firm dispatchable
energy.

15
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
3.1.5 Enabling Fast and Accurate Response
Although energy storage can make wind turbines more versatile when hybridized, appropriate
controls and tests must be done to ensure that coordination and response times are good enough
to provide the necessary services. For example, fault ride-through and black-start capability will
need prompt response and even near-instantaneous synchronization with the grid. NREL
researchers have achieved Li-ion battery response times of less than 30‒40 milliseconds (ms)
(NREL 2018a). The response time also depends on which mode the Li-ion battery is operating
in. In a grid-following mode, the response time is about 25 ms, whereas it is about 50 ms in a
grid-forming mode.

In grid-forming mode, a hybrid resource is the primary source of the voltage and frequency
regulation. The underlying inverter of the hybrid resource consists of voltage and current
regulators working together to maintain the nominal state of the grid. The grid-forming inverter
may work as the master or work in parallel with other inverters in the microgrid. The main
challenges during grid-forming mode are to maintain the stability of operation during changing
set points and ensure black start of the microgrid (Fusero et al. 2019). During transitions, such as
connecting and disconnecting from the utility grid or energizing and de-energizing other DERs
in islanded mode, the grid-forming inverter should be able to resynchronize the system with
minimum transients. The mode requires correcting active and reactive power sharing in tandem
with other DERs. To summarize, the inverter in grid-forming mode should be able to mimic the
dynamic behavior of synchronous generators. A precise control of the virtual inertia of the
inverter is important for system stability in both grid-following and grid-forming modes.

3.2 Modeling Controls and Time Scales

The wind-hybrid models used for a simulation could be discrete or continuous. Depending on the
time scales of a discrete simulation, we can capture various dynamics of the hybrid system using
simulation models ranging from electromagnetic transient to the phasor solution at a given
frequency (e.g., 60 Hz). Different resolutions and fidelities of model physics are essential to
capture events ranging from dynamics to minute and hourly deviations. The ability to visualize
and generate data demonstrating the interactions of inverters and batteries at various scales will
aid in an expanded understanding of stability. In general, the power system simulation models
for wind-hybrid systems may be classified as:

• Detail electromagnetic transient simulation (about 1 nanosecond-microsecond, including

modeling power electronics switching).
• Average simulation (about 100 microseconds-milliseconds; good enough to capture the
electrical transients, phase imbalances, faults, and dynamics).
• Phasor solution (at 60 Hz, typically balanced modeling). Sometimes, we may be
interested in a solution at a frequency, such as 60 Hz. A phasor solution solves a much
simpler set of algebraic equations relating to the voltage and current phasors. This
method computes voltages and currents as phasors. Phasors are complex numbers
representing sinusoidal voltages and currents at a particular frequency (Mathworks 2020).
They can be expressed either in Cartesian coordinates (real and imaginary) or in polar
coordinates (amplitude and phase). As the electrical states are ignored in the phasor
solution, the simulation is therefore much faster to execute.

16
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 • Hybrid simulation and co-simulation (in which certain spatiotemporal characteristics
could be modeled with higher fidelity whereas others could use simpler models for faster
computation). Such modeling can also be done using co-simulation of several existing
tools of varying modeling fidelity to ensure scalability to larger systems and faster
computation. Hybrid simulations may combine simulations at various time scales and
model topologies. One example is combining the electromagnetic transient and transient
stability simulations (Athaide 2018). Another example is the co-simulation of bulk
transmission systems, along with market dispatch, and the individual distribution system
feeders that may connect to a hybrid distributed wind system. The bulk system and
market representation may have to be modeled at 5-minute time scales, whereas the
distribution network may have to be simulated at a higher temporal resolution to respect
voltage bounds (quasi-static steady state).

17
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
4 Operation and Dispatch of Wind-Storage Hybrids
Operation and dispatch of wind-storage hybrids depend on the intended function as well as the
configuration of the hybrid in relation to the external power grid. For example, a hybrid system
operating in an isolated grid may differ significantly than the same hybrid system in grid-
connected mode. In an isolated grid, the wind-storage hybrid system may need to operate as a
grid-forming asset, whereas in the grid-connected mode it could normally operate in a grid-
following mode. This is a common challenge for generation employed in microgrids, and the
complexity increases slightly for a hybrid system in a microgrid.

4.1 Wind-Storage Hybrids Optimal Dispatch

Operating a wind-storage hybrid system involves uncertain intrinsic and extrinsic factors. One of
the major flaws of energy storage dispatch algorithms is that they are often based on forecasts
relying on perfect foresight and/or historical trends. These forecasts are used to optimize net
benefits of the operation and dispatch for a set of geospatial and temporal constraints.
Optimizing operation is governed by technical and economic requirements and can include
multiple time scales or multiperiod formulation of the operation and dispatch of a wind-storage
hybrid system. A margin for error must be included for a real-world system to ensure that its
technical and economic goals are met.

A hybrid system model can have different objectives than the individual subsystem models. The
model may include objective functions, such as optimizing revenue from co-optimized markets,
not just from energy, which is a departure from how energy storage and distributed wind turbines
have been traditionally modeled and dispatched. A wind-storage hybrid system mitigates
variability by injecting more firm generation into the grid. This is particularly helpful in high-
contribution systems, weak grids, and behind-the-meter systems that have different market
drivers. A battery combined with a wind generator can provide a wider range of services than
either the battery or the wind generator alone.

A study conducted for an isolated system (Barley and Winn 1996) examined three dispatch
strategies. The results illustrate the nature of the optimal strategy for two simple dispatch
strategies load following and cycle charging (HOMER Energy 2020) for a minimum run time.
The study found that the combination of a simple diesel dispatch strategy with the frugal use of
stored energy is virtually as cost-effective as the ideal predictive strategy.

An NREL study compared an independently coupled and uncoupled dispatch of PV and storage
for a day with a DC-coupled dispatch. As shown in Figure 7, in this case, the DC-coupled system
seems to lose revenue because the shared 50-MW inverter cannot fully utilize the storage system
(the total solar and storage power output is limited to a 50-MW inverter limit) (Denholm,
Eichman, and Margolis 2017). However, such a system (with inverter and load ratio > 1) at times
can avoid clipped energy by forcing the storage to charge with the excess power from PV.

18
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 Figure 7. Dispatch of photovoltaics-plus-storage system on a typical day

Several considerations remain regarding operating and dispatching hybrid plants in grid-tied
mode, including:

• If the hybrid plant is self-scheduled, it needs an algorithm to use forecasts of distributed

wind and prices to dispatch the hybrid wind and storage, considering the maximal
utilization of the storage SOC for multiple look-ahead periods.
• If the hybrid plant will be dispatched by a centralized scheduler and dispatcher, then new
challenges and opportunities arise for the construction of bids and offers that will be sent
from the hybrid plants. If the plant is wind only, then forecasts with their bounds are
typically sent, and in some rare utility-scale applications, the ability of the wind plant to
provide down-regulation (by curtailment) is communicated. If the plant has energy
storage, then communication of SOC and charging and discharging schedules will be key.
For a hybrid plant, the central dispatcher may only want to know 1) the maximum and
minimum generation capability (considering forecasts, available SOC, and price forecasts
for maximizing storage arbitrage); 2) up and down ramp rates for 5- and 10-minute
intervals relevant for regulation and spinning reserve services (from storage rates and
forecasted wind ramps); and 3) operational cost, which may be a function of nominal
wind turbine and storage operational costs, including the impact of cycling on battery
life.

4.2 Wind-Storage Hybrids Supporting Black Start

Black start is the procedure used to restore power when it is lost. It requires a gradual ramping up
of wind turbine power in coordination with other subsystems, including controllable loads.
Wind-storage hybrids of the correct capacity can support black starts of microgrids in island
mode and in permanently isolated grids. In grid-connected mode, the grid normally provides the
required reference voltage to start a wind turbine. Black start is an advanced operation that
requires collaboration and coordination among many subsystems, including storage, using an
advanced control algorithm.

Wind turbines can provide black start in conjunction with an inverter (grid forming) and external
auxiliary power supplies such as battery storage to maintain a minimum DC voltage to initiate
the power ramp-up operation. In the case of the SMA Solar Technology inverter at NREL’s

19
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Flatirons Campus microgrid (SMA 2016), the black-start operation starts when, after closing the
DC load-break switch, the inverter checks for voltage at the AC terminals. If no AC voltage is
applied, the AC disconnection unit is closed, and the configured AC voltage set point is ramped
up. The AC voltage set point is usually specified via an external plant control using a Modbus
protocol. If an AC voltage already exists to the inverter terminal, the inverter can synchronize
with the external auxiliary power supply, close the AC disconnection, and support the power
grid. The start voltage must be at least 20% of the nominal AC voltage.

Wind turbines have demonstrated the ability to provide a black start in some special
circumstances. Figure 8 demonstrates a black-start operation utilizing three distributed wind
turbines in an isolated grid. This illustration (Majumder 2020) demonstrates how control systems
gradually adjust the DC voltage, AC voltage, and load to build up the voltage reference for the
second wind turbine to come online and aid the black-start process.

Figure 8. Distributed black start of wind turbines in an island mode.

Source: Majumder (2020)

In Figure 8, the black-start operation starts at time, t1, with wind turbine generator 1 (WTG1)
energized using an external auxiliary supply to bring the bus voltage up to 40% of the reference
voltage at t2. From t2 to t3, the wind turbine attains a steady operation at 0.5 MW. At t3, WTG3
is brought into the process and the load in the bus is increased accordingly to 1.2 MW to match
the generation. The voltage ramps up linearly following an external AC reference and reaches

20
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
the reference voltage at t5. The system remains at steady state until t6, at which point WTG2 is
energized fully to deliver the rated 0.5 MW of power.

Obviously, the black-start operation of the wind turbine is contingent upon the wind resource. An
integrated storage in the DC link of the wind turbine may function as an external auxiliary source
during the operation. For a microgrid with more than one inverter, a superordinate plant control
is required to coordinate various stages of the black start among the inverters.

In the United Kingdom, National Grid ESO has started an ambitious project called Distributed
ReStart (National Grid ESO 2020), which plans to demonstrate the black-start service through
the coordinated operation of DERs.

21
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
5 Techno-Economic Sizing of Wind-Storage Hybrids
Techno-economic evaluation of hybrid plants depends on both the benefits and costs (e.g.,
investment, installation, balance of system, soft, life cycle, and operational costs). Benefits could
include increased revenue by utilizing otherwise trimmed variable renewable energy. Some
components could also be shared for effective cost reduction. With the added flexibility of
energy storage, a hybrid wind power plant may be able to provide—in addition to firm energy—
flexibility and ancillary services with very high dependability. However, because of the shared
inverter, the system may generate less revenue under configurations of hybrid coupling that limit
storage operation during periods of high wind output. We will review some of these trade-offs in
this section, based on the state-of-the-art sizing methods proposed for wind-storage hybrids in
the open-source literature.

The sizing of storage in a wind-storage hybrid depends on various factors, such as resource
profile, load profile, desired storage functions, energy, and other essential reliability services
pricing signals, and the time scale of the analysis. Here, our focus will be on batteries that can
capture and store excess wind turbine energy and send it to the utility grid or a local microgrid as
necessary. The batteries can be integrated with each wind turbine or installed at the wind farm
level, as shown in Figure 1.

The techno-economic sizing of wind-storage systems depends largely on cost models of storage
and wind-hybrid systems. Such sizing tools go beyond conventional decision -making based on
levelized cost of energy-based decision-making. These computer-aided-engineering tools aim to
capture market structure more accurately, along with synergies and value streams from grid
services that may exist at different levels of the co-located subsystems. The market price signal
can make or break the viability of storage for an integrated wind hybrid project. Hence, it is very
important that the different value streams of a hybrid system be evaluated fairly. Some of the
value streams of a wind-hybrid system are not recognized (or are taken for granted) in the legacy
energy market structure that is dominant today.

5.1 Storage Cost Models

In this section, we summarize storage cost models of Li-ion batteries, using data from both the
energy and vehicle industries. We anticipate that the cost models will not deviate significantly
for a hybrid wind power plant compared to a hybrid PV plant, even if a typical wind turbine is
AC, whereas PV is DC. The analyses we include here are taken mainly from Denholm, Eichman,
and Margolis (2017); Fu, Remo, and Margolis (2018); and Cole and Frazier (2019).

An NREL study (Cole and Frazier 2019) looked at the cost projection for 4-hour Li-ion systems
in 2018 dollars. Figure 9 shows the overall capital cost for a 4-hour battery system. Regional
capital cost multipliers for battery systems range from 0.948 to 1.11, with Long Island having the
highest multiplier. This study uses a separate cost projection for the power and energy
components of Li-ion systems. Although the range is considerable, all projections show a decline
in capital costs, with cost reductions of 10%‒52% by 2025.

22
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 Figure 9. Battery cost projections for 4-hour Li-ion systems

Another study analyzed the total net present cost of the hybrid system and compared it with a
system without storage (Dufo-López and Bernal-Agustín 2015) to determine cost per kilowatt-
hour (cycled) of the Li-ion batteries for an economically feasible project. The techno-economic
evaluation of grid-connected storage under a time-of-day electricity tariff suggests that the Li-ion
battery cost would need to be reduced to about 0.085 $/kWhcycled.

5.2 Wind-Hybrid Models

There are a handful of first-generation tools to support techno-economic sizing of storage in
relation to wind-hybrid systems. The popular wind-hybrid models in the industry use
performance analysis at hourly or subhourly time scales. The performance-analysis-based tools
focus on energy balance at each time step of simulation for a typical year. The default time step
of many such models is an hour; hence, there will be 8,760 time steps in a typical year. The most
popular models are Hybrid Optimization of Multiple Energy Resources (Lilienthal 2005); the
Distributed Energy Resources Customer Adoption Model (Stadler et al. 2014, Stadler et al.
2016); and Hybrid2 (Manwell et al. 2006, Baring-Gould 1996), among others. There are in-
house NREL models such as Renewable Energy Integration and Optimization (Cutler et al.
2017) and the System Advisor Model (Blair et al. 2018) for sizing and analyzing hybrid systems,
all of which include the value of resilience (i.e., hours of support during complete grid outage).
These tools use exhaustive performance analysis and/or some optimization techniques like
mixed-integer linear programming to determine the optimal storage size. These models help
design and optimize hybrid systems generally based on the levelized cost of energy or other
relevant objective functions under a set of constraints. They also use market price signals on a
limited basis ($/kWh) but at times miss the value streams associated with hybridization, such as
enhanced essential reliability services; spatiotemporal values of energy and ancillary services
resulting from changing conditions and transmission congestions; associated value streams; and
sharing of infrastructure at component levels. The metrics based on levelized cost of energy-

23
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
based metrics do not consider the difference in value between various distributed-wind-plus-
storage configurations. There are not many studies that compare the cost of AC-coupled
distributed wind with DC-coupled distributed-wind-hybrid systems. However, there are some
solar studies that can be used to make an educated guess. Some extra components are needed for
AC-coupled systems, and corresponding labor and balance-of-system costs may range from 1%
to 5% depending on the size and geospatial coordinates of the hybrid project.

There are other tools, such as NREL’s Hybrid Optimization Performance Platform software
(National Renewable Energy Laboratory. Version 1.0. (2021). ), that further consider the synergy
of wind turbine and hybrid systems at the component level and optimize their use. In addition to
quantifying value streams associated with energy and capacity services, they also provide a value
methodology to evaluate the essential reliability services that a wind-hybrid system may provide.

A Joint Institute for Strategic Energy Analysis white paper (Ericson et al., “Hybrid Storage
Market Assessment,” 2017) gives an optimistic evaluation of hybrid storage markets. The paper
evaluates which markets are best suited for battery storage and storage hybrid systems and
reviews regulations and incentives that support or impede the implementation of stand-alone
storage and battery hybrids. California is found to be the most attractive geographic market for
U.S. battery storage because of its storage mandates, high renewables penetration, and regulatory
framework conducive to battery storage projects.

Recently, the scope for adding batteries to grid-connected wind projects is expanding around the
world (Parnel and Stromsta 2020), building on the considerable momentum that already exists
for hybrid solar-plus-storage plants. An earlier study (Ericson et al., “U.S. Energy Storage
Monitor,” 2017) forecasts a twenty-two-fold increase in battery storage and hybrid system
capacity in the United States by 2023 compared to the 2017 baseline.

24
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
6 Conclusion
In this report, we provide a comprehensive overview of the state-of-the-art for wind-storage
hybrid systems, particularly in distributed applications, to enable distributed wind system
stakeholders to realize the maximum benefits from their system. The goal of this report is to
promote understanding of the technologies involved in wind-storage hybrid systems and to
determine the optimal strategies for integrating these technologies into a distributed system that
provides primary energy as well as grid support services.

In our summary of technical benefits and modeling considerations, we identify the main benefit
from storage integration with wind to smooth power output and match energy production with
demand. In addition to smoothing output from the variable wind resource and supporting grid
stability, coupling wind energy generation with a storage system can provide quick-response
frequency and voltage support as well as active power control. Wind-storage hybrid systems can
also support black start of a power system, which can be very beneficial in bringing a power
system back online following a major grid disruption.

Our comparison of distributed-wind-storage hybrid system configurations highlights that turbine

technology, the size of the distributed system, as well as non-technical factors such as market
price for energy and grid services as well as tax credit policies determine which configuration is
best suited to meet generation and load demands while keeping the grid stable. Control strategies
to enable these configurations to meet energy and service demands include baseline reliability
and grid stability control, but also frequency response, as well as voltage and reactive power
support. Additional considerations for controls include enabling flexibility for optimal and
resilient control and achieving time scales for measurement and response that enable these assets
to provide advanced services and operation.

In our assessment of optimal operation and dispatch for distributed-wind-storage hybrid systems,
we highlight the dependence of this optimal operation on the distributed system configuration.
Namely, whether the distributed system is behind or in front of the meter, and whether it is grid
connected or not dictates the optimal operation to achieve both market and grid resilience
benefits.

Similarly, our review of techno-economic feasibility models for hybrid power plant design
indicates that the techno-economic sizing of wind-storage systems depends largely on the system
configuration (whether it is grid connected or not, behind the meter or not) as well as storage
system costs. The hybrid plant design models considered in this report aim to capture market
structure accurately, along with synergies and value streams from grid services. The market price
signal determines the viability of storage in hybrid project design. Hence, it is critical to
comprehensively evaluate hybrid plant value streams, some of which are not recognized by our
current energy market participation and compensation structures.

Based on our assessment of the state-of-the-art of wind-storage hybrid energy systems,

particularly for distributed system applications, opportunities for future work include:

• Developing well-documented, publicly available models for both AC and DC systems

25
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
 • Expanding on the opportunities that complementary wind and solar resources might
provide to a power system
• Evaluating systems in a simulated and power-hardware-in-the-loop environment to aid in
the development of useful case studies to support industry acceptance of distributed-
wind-storage hybrid systems
• Using wind-storage hybrid simulations to assess various configurations to support the
development of advanced sizing methods for AC- and DC-coupled wind-storage hybrid
systems
• Including other distributed energy resources (such as solar) into distributed hybrid
systems research.
The opportunities for future work outlined here have directly impacted the research to be
addressed through the remainder of the MIRACL project, under which this report was written.
With the remaining life of the project, we plan to conduct research and develop further publicly
available reports that address each of these opportunities.

26
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
References
Ambrose, Hanjiro. 2020. “The Second-Life of Used EV Batteries.” Union of Concerned
Scientists: The Equation (blog). Accessed May 27, 2021. https://blog.ucsusa.org/hanjiro-
ambrose/the-second-life-of-used-ev-batteries/.

Athaide, Denise. 2018. “Electromagnetic Transient—Transient Stability Hybrid Simulation for

Electric Power Systems with Converter Interfaced Generation.” Master’s thesis, Arizona State
University.
https://repository.asu.edu/attachments/211375/content/Athaide_asu_0010N_18449.pdf.

Badwawi, Rashid al, Mohammed Abusara, and Tapa Mallick. 2015. “A Review of Hybrid Solar
PV and Wind Energy System.” Smart Science 3(3): 127–138.
10.1080/23080477.2015.11665647.

Baring-Gould, E. Ian. 1996. Hybrid2: The Hybrid System Simulation Model, Version 1.0, User
Manual. Golden, CO: National Renewable Energy Laboratory. NREL/TP-440-21272.
https://www.nrel.gov/docs/legosti/old/21272.pdf.

Baring-Gould, E. Ian, Charles Newcomb, David Corbus, and Raja Kalidas. 2001. “Field
Performance of Hybrid Power Systems.” Presented at the American Wind Energy Association's
WINDPOWER 2001 Conference, Washington, DC, June 4–7. National Renewable Energy
Laboratory, Golden, CO, NREL/CP-500-30566. https://www.nrel.gov/docs/fy01osti/30566.pdf.

Barley, C. Dennis and C. Byron Winn. 1996. “Optimal Dispatch Strategy in Remote Hybrid
Power Systems.” Solar Energy 58(4–6): 165–179. https://doi.org/10.1016/S0038-
092X(96)00087-4.

Bergey, Michael. 2020. “Business Model for Rural Cooperative Distributed Wind Microgrids.”
Presented at Distributed Wind Energy Association Distributed Wind 2020, Arlington, VA,
February 26–27. Bergey Windpower, Norman, Oklahoma.
https://distributedwind.org/event/distributed-wind-2020-lobby-day-business-conference/.

Blair, Nate, Nick DiOrio, Janine Freeman, Paul Gilman, Steven Janzou, Ty Neises, and Michael
Wagner. 2018. System Advisor Model (SAM) General Description. Golden, CO: National
Renewable Energy Laboratory. NREL/TP-6A20-70414.
https://www.nrel.gov/docs/fy18osti/70414.pdf.

California ISO. 2016. “Fast Fact: What the Duck Curve Tells Us About Managing a Green
Grid.” https://www.caiso.com/Documents/FlexibleResourcesHelpRenewables_FastFacts.pdf.

Cole, Wesley and A. Will Frazier. 2019. Cost Projections for Utility-Scale Battery Storage.
Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-73222.
https://www.nrel.gov/docs/fy19osti/73222.pdf.

27
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Corbus, David, Charles Newcomb, E. Ian Baring-Gould, and Seth Friedly. 2002. “Battery
Voltage Stability Effects on Small Wind Turbine Energy Capture: Preprint.” Presented at the
American Wind Energy Association WINDPOWER 2002 Conference, Portland, OR, June 2–5.
National Renewable Energy Laboratory, Golden, CO, NREL/CP-500-32511.
https://www.nrel.gov/docs/fy02osti/32511.pdf.

Cutler, Dylan, Dan Olis, Emma Elgqvist, Xiangkun Li, Nick Laws, Nick DiOrio, Andy Walker,
and Kate Anderson. 2017. REopt: A Platform for Energy System Integration and Optimization.
Golden, CO: National Renewable Energy Laboratory. NREL/TP-7A40-70022.
https://www.nrel.gov/docs/fy17osti/70022.pdf.

Das, Trishna, Venkat Krishnan, and James D. McCalley. 2015. “Assessing the Benefits and
Economics of Bulk Energy Storage Technologies in the Power Grid.” Applied Energy 139: 104–
118. https://doi.org/10.1016/j.apenergy.2014.11.017.

Denholm, Paul, Josh Eichman, and Robert Margolis. 2017. Evaluating the Technical and
Economic Performance of PV Plus Storage Power Plants. Golden, CO: National Renewable
Energy Laboratory. NREL/TP-6A20-68737. https://www.nrel.gov/docs/fy17osti/68737.pdf.

DiOrio, Nicholas, and Will Hobbs. 2018. “Economic Dispatch for DC-Connected Battery
Systems on Large PV Plants.” Presented at 2018 PV Systems Symposium—Grid Integration
Track, Albuquerque, NM, May 3. https://pvpmc.sandia.gov/download/6559/.

Dufo-López, Rodolfo and José L. Bernal-Agustín. 2015. “Techno-Economic Analysis of Grid-

Connected Battery Storage.” Energy Conversion and Management 91: 394–404.
10.1016/j.enconman.2014.12.038.

Dykes, Katherine, Jennifer King, Nicholas DiOrio, Ryan King, Vahan Gevorgian, David Corbus,
Nate Blair, Kate Anderson, Greg Stark, Craig Turchi, et al. 2020. Opportunities for Research
and Development of Hybrid Power Plants. Golden, CO: National Renewable Energy Laboratory.
NREL/TP-5000-75026. https://www.nrel.gov/docs/fy20osti/75026.pdf.

Elsayed, Ahmed T., Ahmed A. Mohamed, and Osama A. Mohammed. 2015. “DC Microgrids
and Distribution Systems: An Overview.” Electric Power Systems Research 119: 407–417.
https://doi.org/10.1016/j.epsr.2014.10.017.

Energy Storage Systems, Inc. 2016. Beyond Four Hours. Portland, OR.
https://www.essinc.com/wp-content/uploads/2016/11/Beyond-Four-Hours_ESS-Inc-White-
Paper_12_2016_mr.pdf.

Ericson, Sean, Eric Rose, Harshit Jayaswal, Wesley Cole, Jill Engel-Cox, Jeffery Logan, Joyce
McLaren, Kate Anderson, and Doug Arent. 2017. U.S Energy Storage Monitor Q2 2017 Full
Report. GTM Research Group. Golden, CO: National Renewable Energy Laboratory.
NREL/MP-6A50-70237. https://www.nrel.gov/docs/fy18osti/70237.pdf.

28
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Ericson, Sean, Eric Rose, Harshit Jayaswal, Wesley Cole, Jill Engel-Cox, Jeffery Logan, Joyce
A. McLaren, Katherine H. Anderson, J. Douglas Arent, John Glassmire, et al. 2017. Hybrid
Storage Market Assessment. The Joint Institute for Strategic Energy Analysis (JISEA). Golden,
CO: National Renewable Energy Laboratory. NREL/MP-6A50-70237.
https://doi.org/10.2172/1399357.

Fingersh, L.J. 2003. Optimized Hydrogen and Electricity Generation from Wind. Golden, CO:
National Renewable Energy Laboratory. NREL/TP-500-34364.
https://www.energy.gov/sites/default/files/2014/03/f12/34364.pdf.

Fu, Ran, Timothy Remo, and Robert Margolis. 2018. 2018 U.S. Utility-Scale Photovoltaics-Plus-
Energy Storage System Costs Benchmark. Golden, CO: National Renewable Energy Laboratory.
NREL/TP-6A20-71714. https://www.nrel.gov/docs/fy19osti/71714.pdf.

Fusero, Michelle, Andrew Tuckey, Alessandro Rosini, Pietro Serra, Renato Procopio, and
Andrea Bonfiglio. 2019. “A Comprehensive Inverter-BESS Primary Control for AC
Microgrids.” Energies 12(20): 3810. https://doi.org/10.3390/en12203810.

Glasgo, Brock, Inês Lima Azevedo, and Chris Hendrickson. 2018. “Expert Assessments on the
Future of Direct Current in Buildings.” Environmental Research Letters 13(7): 074004.
https://doi.org/10.1088/1748-9326/aaca42.

Gu, Yunjie, Wuhua Li, and Xiangning He. 2014. “Frequency-Coordinating Virtual Impedance
for Autonomous Power Management of DC Microgrid.” IEEE Transactions on Power
Electronics 30(4): 2328–2337. https://doi.org/10.1109/TPEL.2014.2325856.

Hybrid Optimization of Multiple Energy Resources (HOMER) Energy. 2020. “Cycle Charging.”
https://www.homerenergy.com/products/pro/docs/latest/cycle_charging.html.

Lilienthal, Peter. 2005. “HOMER Micropower Optimization Model.” Presented at the 2004 DOE
Solar Energy Technologies Program Review Meeting, October 25–28, 2004, Denver, CO.
National Renewable Energy Laboratory, Golden, CO, NREL/CP-710-37606.
https://www.nrel.gov/docs/fy05osti/37606.pdf.

Majumder, Rajat. 2020. “Weak Area Network Control of Wind Turbine Generators.” Webinar,
April 20. https://www.esig.energy/event/webinar-weak-area-network-control-of-wind-turbine-
generators/.

Manwell, J.F., A. Rogers, G. Hayman, C.T. Avelar, J.G. McGowan, U. Abdulwahid, and K. Wu.
2006. Hybrid2–A Hybrid System Simulation Model:Theory Manual. Golden, CO: National
Renewable Energy Laboratory.

Marsh, Jacob. 2019. “AC vs. DC Solar Battery Coupling: What You Need To Know.”
EnergySage, Accessed June 18, 2020. https://news.energysage.com/ac-vs-dc-solar-battery-
coupling-what-you-need-to-know/.

29
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Mathworks. 2020. “Introducing the Phasor Simulation Method.” Accessed June 17, 2020.
https://www.mathworks.com/help/physmod/sps/powersys/ug/introducing-the-phasor-simulation-
method.html.

Miller, N.W. 2013. “GE Wind Plant Advanced Controls.” 1st International Workshop on Grid
Simulator Testing of Wind Turbine Drivetrains. Golden, CO: GE Energy Consulting.
https://www.nrel.gov/grid/assets/pdfs/turbine_sim_12_advanced_wind_plant_controls.pdf.

Miller, N.W. 2014. “GE Experience With Turbine Integrated Battery Energy Storage.” Presented
at the 2014 IEEE PES General Meeting, July 27–31, National Harbor, MD. https://www.ieee-
pes.org/presentations/gm2014/PESGM2014P-000717.pdf.

National Grid ESO. 2020. “Distributed ReStart: What Is the Distributed ReStart Project?”
Accessed August 19, 2020. https://www.nationalgrideso.com/future-energy/projects/distributed-
restart.

National Renewable Energy Laboratory (NREL). 2018a. “March Developments at NREL Shave
Response Time for Large-Scale Batteries Down to Milliseconds.” Energy Systems Integration
Newsletter, March 2018. https://www.nrel.gov/esif/esi-news-201803.html.

National Renewable Energy Laboratory. 2018b. Federal Tax Incentives for Energy Storage
Systems. Golden, CO: National Renewable Energy Laboratory. NREL/FS-7A40-70384.
https://www.nrel.gov/docs/fy18osti/70384.pdf.

National Renewable Energy Laboratory. 2020. “Battery Second-Use Repurposing Cost

Calculator.” Accessed July 17, 2020. https://www.nrel.gov/transportation/b2u-calculator.html.

National Renewable Energy Laboratory. Hybrid Optimization and Performance Platform

(HOPP). Version 1.0. (2021). https://github.com/NREL/HOPP

Parnel, John and Karl-Erik Stromsta. 2020. “Storage Hybrid Plants Becoming More Attractive in
Maturing Wind and Solar Markets.” Greentech Media, Accessed June 12, 2020.
https://www.greentechmedia.com/articles/read/storage-co-location-getting-more-attractive-for-
maturing-wind-and-solar-markets.

Porter, Kevin, Kevin Starr, and Andrew Mills. 2015. “Variable Generation in Electricity
Markets.” Presented at the Utility Variable Generation Integration Group Fall Technical
Workshop, October 15. Reston, VA: Utility Variable-Generation Integration Group.
https://www.esig.energy/download/variable-generation-electricity-markets-kevin-porter-kevin-
starr-andrew-mills/.

Poudel, R.C., J.F. Manwell, and J.G. McGowan. 2020. “Performance Analysis of Hybrid
Microhydro Power Systems.” Energy Conversion and Management 215: 112873.
https://doi.org/10.1016/j.enconman.2020.112873.

30
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.
Reilly, J., R. Poudel, V. Krishnan, R. Preus, I. Baring-Gould, B. Anderson, B. Naughton, F.
Wilches-Bernal, and R. Darbali. 2021. Distributed Wind Controls: A Research Roadmap for
Microgrids, Infrastructure Resilience, and Advanced Controls Launchpad (MIRACL). Golden,
CO: National Renewable Energy Laboratory. NREL/TP-7A40-76748.
https://www.nrel.gov/docs/fy21osti/76748.pdf.

Rekdal, Kristin. 2018. “Battery Energy Storage Integration via DC/AC Converter in Grid
Connected Wind Turbines.” Master’s thesis, Department of Electric Power Engineering,
Norwegian University of Science and Technology. https://ntnuopen.ntnu.no/ntnu-
xmlui/bitstream/handle/11250/2507658/18855_FULLTEXT.pdf

Singh, Mohit and Surya Santoso. 2011. Dynamic Models for Wind Turbines and Wind Power
Plants. Golden, CO: National Renewable Energy Laboratory. NREL/SR-5500-52780.
https://www.nrel.gov/docs/fy12osti/52780.pdf.

SMA. 2016. “Operating manual: SUNNY CENTRAL STORAGE.” SMA.

https://files.sma.de/downloads/SCS-BE-E7-en-12.pdf.

Stadler, Michael, Gonçalo Cardoso, Salman Mashayekh, Thibault Forget, Nicholas DeForest,
Ankit Agarwal, and Anna Schönbein. 2016. “Value Streams in Microgrids: A Literature
Review.” Applied Energy 162(4): 980–989. 10.1016/j.apenergy.2015.10.081.

Stadler, Michael, Markus Groissböck, Gonçalo Cardoso, and Chris Marnay. 2014. “Optimizing
Distributed Energy Resources and Building Retrofits With the Strategic DER-CAModel.”
Applied Energy 132: 557–567. 10.1016/j.apenergy.2014.07.041.

Weaver, Wayne W., Rush D. Robinett III, Gordon G. Parker, and David G. Wilson. 2015.
“Energy Storage Requirements of DC Microgrids With High Penetration Renewables Under
Droop Control.” International Journal of Electrical Power & Energy Systems 68(6): 203–209.
10.1016/j.ijepes.2014.12.070.

Wenzl, Heinz, E. Ian Baring-Gould, Rudi Kaiser, Bor Yann Liaw, Per Lundsager, Jim Manwell,
Alan Ruddell, and Vojtech Svoboda. 2005. “Life Prediction of Batteries for Selecting the
Technically Most Suitable and Cost Effective Battery.” Journal of Power Sources 144(2): 373–
384. https://dx.doi.org/10.1016/j.jpowsour.2004.11.045.

Zeng, Jianwu, Jiahong Ning, Xia Du, Taesic Kim, Zhaoxia Yang, and Vincent Winstead. 2019.
“A Four-Port DC-DC Converter for a Standalone Wind and Solar Energy System.” IEEE
Transactions on Industry Applications 56(1): 446–454. 10.1109/TIA.2019.2948125.

Zhao, Jinxin and Florian Dörfler. 2015. “Distributed Control and Optimization in DC
Microgrids.” Automatica 61: 18–26. https://doi.org/10.1016/j.automatica.2015.07.015.

31
This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.