Modeling Challenges:

Dynamic Interactions Between Inverters, Grid Forming Inverters

V. Gevorgian, S. Shah, P. Koralewicz, NREL

Challenges for Distribution Planning, Operational and Real-time Planning Analytics Workshop

Washington, DC

May 17, 2019

NREL | 2

Services by Multi-Technology (Hybrid) Plants

– Dispatchable renewable plant operation

• Long-term and short-term production forecasts

• Capability to bid into day-ahead and real-time energy markets like conventional generation

– Ramp limiting, variability smoothing, cloud-impact mitigation

– Provision of spinning reserve

– AGC functionality

– Primary frequency response (programmable droop control)

– Fast frequency response (FFR)

– Inertial response:

• programmable synthetic inertia for a wide range of H constants emulated by BESS

• Selective inertial response strategies by wind turbines

– Reactive power/voltage control

– Advanced controls: power system oscillations damping, wide-area stability services

– Resiliency services: black-start, islanded operation

– Stacked services

– Plant electric loss reduction, AEP increase

– Selective plant configuration for BESS: ability to serve a whole wind power plant, or selected rows/turbines

– Battery SOC management

– Optimization model-predictive control strategies – work in progress

– Revenue optimization

NREL-First Solar Project

NREL-PG&E project

NREL | 3

Control Interactions

Governor

Controls

Field Control

(AVR)

Pitch Control

Q & V Control

DC Bus

Control

PLL

Current

Control

Sy

nch

ronou

s G

ener

ato

rs

Win

d T

urb

ines

, P

V &

Sto

rage

Inver

ters

,

HV

DC

AGC

Transmission Network

Dynamics

Filters

Torsional Modes

Grid Support Functions

PWMHVDC Controls

• Power electronics-based generation and T&D systems

have increased control interaction problems

NREL | 4

Impedance-Based Analysis

• Impedance responses of inverter/plant and grid are compared– Impedance intersection points give frequencies of resonance modes

– Phase difference at intersection points gives damping

Grid( )gZ s

Loop gain:

Zg(s)/Zp(s), Zg(s)/Zn(s)

( )

( )

p

n

Z s

Z s

NREL | 5

4 kHz300 Hz 2 kHz100 Hz 1 kHz150

100

50

0

50

100

4 kHz300 Hz 2 kHz100 Hz 1 kHz20

30

40

50

60

( )pZ s

Mag

nit

ud

e (d

B)

Ph

ase

(deg

.)

( )gZ s

( )gZ s

( )pZ s

SCR: 5

SCR: 2

172o

182o

195o

205o

SCRGrid

Inductance, Lg

Resonance

Frequency

Phase

Margin

5 4.6 mH 641 Hz +8o

4 5.7 mH 584 Hz -2o

3 7.6 mH 512 Hz -15o

2 11.5 mH 441 Hz -25o

• Unstable Resonance for Weak

Grids

– Unstable for SCR<5.0

– Resonance Frequency Decreases

with SCR and its “Severity”

Increases

-

-

-

Resonance: Frequency and Damping

NREL | 6

SCR=4

SCR=3

SCR=2

Resonance-Generated Distortions

NREL | 7

NREL Flatirons Campus

Siemens

2.3 MW

GE/Alstom

3 MW

GE 1.5 MW

Gamesa

2 MW

Research Turbines

2 x 600 kW

PV Array

1 MW

2.5 MW dynamometer

5 MW dynamometer,

7MVA CGI1MW / 1MWh BESS

FS PV array

• Total of 12+ MW variable renewable generation currently • 7 MVA Controllable Grid Interface (CGI) • Multi-MW energy storage test facility• 2.5MW and 5 MW dynamometers (industrial motor drives)• 13.2 kV medium voltage grid• 1.5 MW total PV capacity

NREL | 8

NWTC Controllable Grid Platform

NWTC Wind Turbines

SunEdison1 MW PV Array

Controllable Grid Interface (CGI)for Grid and Fault Simulation

Switchgear Building

XcelSubstation

1 MW / 1 MWh BESS

Controlled grid,CGI Bus

Regular grid,Xcel Bus

(7 MVA continuous / 40 MVA s.c.)

115 kV

13.2 kV tie-line

GE 1.5 MW

Siemens 2.3 MW

Alstom 3 MW

Gamesa 2 MW

DC

AC

DC

DC

DC

DC

AC

13.2 kV 13.2 kV

First Solar430 kW PV array

GE 1.25 MW / 1.25 MWh BESS

13.2 kV

Aerial view of the siteAC

AC

AC

AES

Grid forming

Flexible testbed for many black start schemes

Image source: NREL

NREL | 9

Impedance Measurement Using CGI

• 7-MVA, 13.2-kV grid simulator (CGI)

+

• Perturbed voltages

• Response currents

NREL | 10

Impedance of 1-MW/13.2-kV BESS System

• Inverter-interfacing 1-MW/1-MWh battery energy storage system

– Voltage perturbation magnitude

• Impedance response

• Different control

elements dominate at

different frequencies.

NREL | 11

Impedance of First Solar PV Plant

Shahil Shah, 04/19/2019

Mag

nit

ud

e (d

BW

)P

has

e (D

EG

.)

Vendor 1, 4x40 kW Inverters: 160 kW, 0 VAR

Vendor 2, 2x125kW Inverters: 250 kW, 0VAR • Phase response goes

below -90 degrees

between 60 Hz and

200 Hz: Can

potentially interact

with grid and create

undamped resonance

NREL | 12

Resonance of Sungrow Inverters for Weak Grid

Shahil Shah, 04/19/2019

Mag

nit

ud

e (d

BW

)P

has

e (D

EG

.)

2 x125kW Inverters: 250 kW, 0VAR

Grid Impedance: Lg = 1.2 H (SCR: 1.5) • 125 kW inverters will form undamped

resonance between 100 and 200 Hz if the

grid is inductive with SCR less than 1.5

on base of 250 KW, 13.2 kV

– Corresponding grid inductance is 1.2 H

NREL | 13

Power-Domain Impedance

• Transfer function from Active Power to Frequency at Point of Interconnection

NREL | 14

Frequency Response Characterization

Loss of generation

Applications:– Real-time Estimation of

Inertia, Primary Frequency Response (PFR), Nadir, etc.

– Frequency Support Design by Renewable Generation

20%

droop 10%

droop

5%

droop

5%

droop10%

droop

20%

droop

NREL | 15

Grid-Forming Inverter from Outside

• BESS Inverter

• Impedance

measurements can

quantify different

aspects of grid-

forming ability

Forming

Following

Forming

Following

NREL | 16

Black-start of Wind Power Plant (13.2 kV system)

Grid forming inverter – 7 MVA

Main problems:• Harmonics / resonances• Grounding / protection

• Inrush currents

Significant overvoltage

Solutions:• Additional filtering• Transformer neutral point grounding resistor• Grounding transformers• Oversized inverter (40 MVA SC capability)

Real event: Resonance observed between power converters during black start at NWTC

Image source: NREL

NREL | 17

Challenges and Future Research

• Adoption of impedance characterization by industry

– Root-cause finding, grid codes, control design, impedance specs

• Impedance measurement using grid simulators

– High-fidelity model validation

– Control design; Testing for grid codes

• New impedance-based tools

– Design of grid-support functions

– Testing of grid-forming ability of inverters

NREL | 18

CGI RTDS

DC

AC

AC

DC

DC

AC

DC

AC

DC

AC

7 MVA Controllable Grid Interface

Regular grid

RjX

Programmable line impedance

40 MVA for 2 sec

Real-time model of a power system

Distribution System Testbed for Islanded Testing

=~

NREL BESS 1MW/1MWh230kV

Z1Z2

70/230kV12.47/70kV

T1T2T3

0.4/12.47kV

CGI POI

12.47kV, Old River

CB1CB4

CB2

SEL-487

Z3

Z4

F1

650 kVAPF=0.6893

Z5Z6

F2Z7

PV

924 kWZ8

SW1 800 kVAR

Z9

R

Z10

SW2

800 kVAR

Z11

F3

560 kVAPF=0.6825

F4

Z12

PV

150 kW 150 kW 150 kW

Z16

CB3

T4

M1000HPPF=0.85

12.47kV/480V

CHIL

PHIL

SEL-487 SEL-487

CHIL

12.47/0.4kVT5

ISLAND

ISLAND FEEDER BREAKER

630631632

640

711712

642

652

643

662

644

645

646

647

691

692

671

672

682

651661

633

641

Grid Forming Control NREL – PG&E project

Image source: NREL

NREL | 19

Example: Bus 630 L-to-L Fault

Low Z fault Hi Z fault

Time (sec)

Time (sec)

Bu

s #

Bra

nch

#

Time (sec)

Time (sec)

NREL | 20

Islanded Distribution Circuit Restoration with BESS

BESS Active and Reactive Powerin Grid Forming Mode

Motor start eventPHIL testing results using 1 MW/1MWh BESS

Image source: NREL

NREL | 21

BESS Inverter Fault-ride Through in Grid Forming Mode While Operating an Islanded System

PHIL testing results using 1 MW/1MWh BESS

Image source: NREL

NREL | 22

Definition of Black Start

NERC definition: “A generating unit(s) and its associated set of equipment which has the ability to be started without support from the System or is designed to remain energized without connection to the remainder of the System, with the ability to energize a bus, meeting the Transmission Operator’s restoration plan needs for Real and Reactive Power capability, frequency and voltage control, and that has been included in the Transmission Operator’s restoration plan”

Image source: NREL

NREL | 23

Black Start Stages

The black start process can be divided into three stages:

• Preparation stage

• Network reconfiguring

• Load restoration

A typical restoration plan for bulk power system includes the following

essential steps:

System status identification: blackout boundaries and location in

respect to critical loads, status of circuit breakers, capacity of available

black start units, etc.

Starting at least one black start unit to supply critical loads such as

nuclear or large thermal power plants

Progressive restoration: step-by-step supply of other loads avoiding

over and under voltage conditions

The restoration strategies:

• Serial – simpler strategy, slower but more stable

• Parallel – quicker but more complex

Conventional top down approach

Conventional centralized BS units to start the network

Passive loads, waiting to be energized

Changing black start paradigm

Microgrids

Bottom-up approach

Image source: NREL

NREL | 24

Configurations of Integrated PV/BESS Plants for Black Start

Co-located starter for a black start resource

Remote starter for a black start resource

PV + storage as fully functional black start resource

Collective black start resource

Image source: NREL

NREL | 25

PV-BESS Black starting a Gas Turbine Generator

Main challenge:• Energizing transformers and feeders• Mid-size gas turbines employ starting motors• Black start inverters need to be sized to provide necessary inrush current

Possible solutions:• Oversized inverters for inrush current • Equip all plant motor loads with soft starters of VFDs • Partial solution – energize transformers with tap positions at highest number of turns

Image source: NREL

NREL | 26

Consecutive Start of Aux Plant Motors

4x250 kW

1 MW

• Inverter current limit is exceeded during motor starting

• Limits capacity of motors that can be started with inverters

• Solutions:• Oversized inverters • Soft starters or VFDs for all

motors

Energizing transformers and feeder

Motor 1 start

Motor 2 start

Motor 3 start

Motor 4 start

MVA limit of inverters

Motor 1 start Motor 2 startMotor 4 startMotor 3 start

Energize transformer 1

Energize transformer 1

Energizefeeder

Image source: NREL

NREL | 27

Inverters in Current limiting Mode

4x250 kW

1 MW

• Inverter current limit is not violated but lack of inrush current causes voltage collapse

• Motors fail to start

MVA limit of inverters

Image source: NREL

NREL | 28

Constant V/Hz Soft Start

4x250 kW

1 MW

Inverters operate as VFDs with constant V/Hz ratio

• The whole circuit is started as a single VFD with constant V/Hz ratio

• Smooth start of all motors• No overcurrent conditions

during energizing/start up• No need to oversize

inverters

MVA limit of inverters

Image source: NREL

NREL | 29

HPP / Supercapacitor Energy Storage for Improved Restoration Process

Super cap storage response

• INL-NREL project• Industry partners:

• Idaho Falls Power• Siemens• Maxwell

• Black start of the feeder circuit with a fully black-start capable hydro power plant (slow governor)

• Supercapacitor energy storage is used to assist in frequency stabilization during system restoration

Image source: NREL

NREL | 30

Challenges, Conclusions

• Today and future restoration strategies should align with the changing network paradigm

• Modern grid forming inverters can contribute into black start / restoration with more superior reactive power capabilities compared to conventional synchronous generators

• Inherent inverter current limit is one most important factor for black start applications

Q [MVAR]

P [MW]

Pmin Pmax

(prime mover limit)

Lagging power factor

Leading power factor

Field current limit

Armature heating constraints

Under excitation limit

Lagging Qmax

Leading Qmax

0

Synchronous Generator

Winding end region heating limit

Synchronous generator

PV Inverter and Type 4 wind turbine

BESS Inverter

Synchronous condenser

Type 3 wind turbine

Typ

e 3

WTG

STATCOM

Recommendations for future studies• Fault performance of grid forming inverters – needs to be robust and standardized• Robust seamless transition between grid forming and grid following• If grid forming is present, do we really need grid following anymore? If so, what shares of GM and GFL are optimal?• Impedance characterization of grid forming inverters• Grid stability impacts of grid forming• Validated grid forming inverter models are needed for various renewable and storage technologies for successful black start studies• At scale PHIL testing of black start-capable renewable resources is an important tool to discover potential issues, test mitigating solutions

and validate models

Image source: NREL

NREL | 31

Summary of CGI#2 SpecificationsPower rating• Continuous AC rating - 19.9 MVA at 13.2kV and 34.5 KV• Overcurrent capability (x5.7 for 3 sec, x7.3 for 0.5 sec)• 4-wire 13.2 kV or 35.4 kV taps• Continuous operational AC voltage range: 0 - 40 kVAC• Continuous DC rating – 10 MW at 5 kVDC

Possible test articles • Types 1, 2, 3 and 4 wind turbines• PV inverters, energy storage systems• Conventional generators• Combinations of technologies / hybrid systems• Responsive loads

Voltage control (no load THD <1%)• Balanced and unbalanced voltage fault conditions (ZVRT, LVRT and 140% HVRT) –

independent voltage control for each phase on 13.2 kV and 34.5 kV terminals• Response time – less than 1 millisecond (from full voltage to zero, or from zero

back to full voltage) • Programmable injection of positive, negative and zero sequence components• Long-term symmetrical voltage variations (+/- 10%) and voltage magnitude

modulations (0-10 Hz) – SSR conditions• Programmable impedance (strong and weak grids, wide SCR range corresponding

to a POI with up to 250 MVA of short circuit apparent power) • Injection of controlled voltage distortions • Wide-spectrum (0-2kHz) impedance characterization of inverter-coupled

generation and loads• All-quadrant reactive power capability characterization of any system

Frequency control• Fast output frequency control (3 Hz/sec) within 45-65 Hz range• 50/60 Hz operation• Can simulate frequency conditions for any type of power system • PHIL capable (can be coupled with RTDS, Opal-RT, Typhoon, etc.)• Coupled with PMU-based wide-area stability controls validation platform

New features • 5 kV MVDC grid simulator (PHIL capable)• Voltage or current source operation• Seamless transition between voltage and current source modes• Emulation of full set of resiliency services:

• Black start• Power system restoration schemes• Microgrids

• Flexible configurations are possible when combined with CGI#1:• Two independent experiments• Parallel operation• Back-to-back operation• Emulation of isolated, partially or fully grid-connected microgrids

20 MVA

NREL | 32

Variable Generation with Storage Ongoing PV, wind, hydro + storage projects (DOE and industry funded):• WETO – storage to enhance APC by wind power• GE – wind storage, hybrid wind/solar/storage• SETO - First Solar hybrid solar PV/storage systems• WPPTO – integrated storage with ROR hydro plants• PG&E EPIC project – synthetic inertia, grid forming, distributed

applications• AES – DC-coupled PV-storage peaker plant

Main current areas of research for storage integration: • Essential reliability services• Co-optimized controls development• Grid stability issues • Resiliency: Grid forming/black start/microgrids• Storage value streamsNew research enabled by Flatirons Campus:• Fundamental operational characteristics of other storage technologies:

• SMES, Supercaps, Flywheels, BESS, Flow batteries• Multi-POI PHIL experiments up to 20MW with CGI#1 and CGI#2 – 13.2kV, 34.5 kV• Virtually interconnected experiments for 100s of MWs – What are the limits of scalability?• PSH and CAES emulation using by-directional dynamometers• Fast EV charging• Thermal storage, hydrogen / CH4 storage• DC and MVDC microgrids

www.nrel.gov

Thank you

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 Solar Energy Technologies Office and Wind Energy Technology Office. The views expressed in the article do not necessarily represent the views of the DOE or the U.S. Government. The U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this work, or allow others to do so, for U.S. Government purposes.