…Exceptional service in the national interest

Renewable Systems Integration

And Microgrids

Operational Energy Security

Sandia National Laboratories is a multi-program laboratory operated and managed

by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation,

for the U.S. Department of Energy's National Nuclear Security Administration under

contract DE-AC04-94AL85000

SAND2010-7644P

Sandia National Laboratories

Energy Security Program

Energy Security Focus

Operational Energy Systems • Electric Power Assurance

• Microgrid, renewables, nuclear, storage,

control systems, cyber

• Transportation Energy Assurance • Combustion research, renewable fuels

Climate Change Science • Operational Impacts

• Assessments

Energy Security Roles

$250M DOE Energy Research Program

Support DoD on energy system, physical, and

cyber security

System integrator for the DOE/NNSA

Distributed Energy Technology Laboratory

Combustion Research Facility

DoD Installation Security Projects

Nuclear Design & Fuel Cycle

3

Energy Challenge - Harvest, Transform, and Control

Delivery of Available Energy

*EXERGY = AVAILABLE ENERGY = useful portion of energy that

allows one to do work and perform energy services

Electricity

Fuel

Heat

Cooling

Chemicals (such as

lubricants)

Clean Water

Fossil (coal, oil, gas)

Solar (including wind and hydro)

Geothermal

Nuclear

Plant, animal, and

human waste

CO2 & other energy

conversion

byproducts

Energy & Material Resources

Harvest, transform,

and deliver exergy* at

the necessary

amount and rate.

Energy Needs

or Services Energy

Processing

Generator

Transmission

Substation

Distribution

Load

Oil and Gas

Crude

Oil

Coal

Gas

Refinery

•Extensive storage in fuels

•Fixed infrastructure is inflexible

•Significant human interaction

Controlled Supply Fixed Infrastructure Random Load

Today’s Power Grid is Designed for Dispatchable

Centralized Generation

4

Loads are Predictable, Allowing

Essentially Open-loop Grid Control

5

AGC

Gen Load Power

Prediction State

Estimator Fixed

Infrastructure

California ISO

Feedback loop controls minor

frequency variations and output

voltage and power

Available resource forecast

Actual load

Day ahead demand forecast

0 4 8 12 16 20 24

Time of day (hours)

Pow

er

(GW

)

30

28

26

24

22

20

Capacity less largest unit

System Load

Hawaiian Electric Co. daily load vs capacity Forecasting is used to set generation

0 4 8 12 16 20 24

Time of day (hours) P

ow

er

(GW

)

1.2

1.0

0.8

0.6

Feedback is

often through

people

An Emerging Market: Preparing for Large-Scale Renewable Energy Integration

New Market Scenario: Climate change concerns, renewable portfolio

standards, incentives, and accelerated cost reduction driving steep

growth in U.S. renewable energy system installations.

6

1200

1000

800

600

400

200

Stochastic Sources (Negative Loads) Complicate Load

Forecasting

Wind power forecasting examples

7

This is weather forecasting!

Forecast

Average

6:00 9:00 12:00 15:00 18:00 21:00

Time (hrs)

0 2 4 6 8 10 12 14 16 18 20 22 24

Hours

400

300

200

100

0

MW

W/m

2

Solar Insolation, May 4, 2004 CST

Actual Wind Power

Forecasts

PV Output Can Vary Considerably on an Average Day

• Irradiance and PV system ac output for a typical partly cloudy day in July

• PV system rating: 1,200 kW ac, presently limited to 400 kW ac

(intentionally)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 104

0

200

400

600

800

1000

1200

1400

Time (1-second samples spanning 5:30 am to 9:30 PM)

Irra

dia

nce (

W/m

2)

and P

lant

outp

ut

(kW

)

La Ola PV System, Lanai, HI

Plant output

Irradiance

Source: SunPower

Castle & Cooke

Generator

Transmission

Substation

Distribution

Load

AGC

Gen Load Power

Prediction State

Estimator Fixed

Infrastructure

PV

Wind

Today, Stochastic Renewable Sources are Treated as

Negative Loads

9

State Estimator

Power Prediction

PV

Wind

Gen

PV

Wind

AGC

To Achieve Maximum Benefit Renewable Energy Needs

to be Treated as a Source

Fixed Infrastructure

Load

System efficiency can increase with reduction in excess generation capacity.

Both our generation and our loads are now random!

Storage: Impedance and Capacity Matching Solution

• Specific applications justify electrical energy storage solutions

Autonomous Mars Rover

Solar

Solar

Autonomous Man on Earth

Super Capacitor

(Storage)

Robot

Mobility

Available

Power In 12 Volts

1 Amp

12 Volts

3 Amp

Required

Power Out

Earth

PV

Solar Cells

Available

Power In

Fossil Fuel

(Storage)

End

Uses

Required

Power Out

Lifestyle

Lifestyle

*NOTE: Required Power Out > Available Power In; Impedance/Capacity Mismatch

Impedance/Capacity Mismatch (Duty Cycle)

1

10

10

10

10

10

10

10 1 10 10 10 0.1

2

3

4

5

6

2 3 4

Specific Energy (Watt-hrs/kg)

Elastic Element

Pneumatic

Flywheel

Photovoltaic Batteries

Internal Combustion

Fuel Cell

Microgrids Can Provide a Pathway for More Effective

Use of Renewable Energy: Optimized Storage and Info

Flow

Generator

Transmission

Substation

Distribution

Loads

With distributed generation

and storage, electric

power can be provided

when the grid is down X

Storage and generation on load side,

sized to match energy performance

needs

PV

Wind

State Estimator

Power Prediction

PV

Wind

Gen

UPFC

UPFC

UPFC Load

PV

Wind

AGC

PV

Wind

UPFC - Unified power flow controllers

Low-Level Distributed Nonlinear Control Enables

Stability and Transient Performance

14

Gen

Gen

Nonlinear control

maintains stability

and performance

)()()()( xVxVxTxTH cc

xVxVxTxTH cc

dtHdtVVTTJI cc 88

i

N

i icc

Hm

HdtHdtJI cc

1

00

1 where,0][

8][

1

SNL’s Hamiltonian-Based Nonlinear Control Theory

Addresses Stability and Performance

Equations from a microgrid can be used to construct a Hamiltonian.

Asymptotic stability

is achieved by satisfying the following constraints

Fisher Information Equivalency provides link to and minimization

of information flow and energy storage.

Hdtc

Individual microgrid

Hamiltonian

Addition of cost functions allow for optimization to a particular solution.

Chosen to minimize storage, conventional generation, etc.

Kinetic Energy Potential Energy

0][00

dtLGdtH cc

0)()(where ,0 *** xVxVxxH c

R.D. Robinett and D.G. Wilson, “Hamiltonian Surface Shaping with Information Theory and Exergy/Entropy Control for Collective Plume

Tracing”, Int. J. Systems, Control and Communications, Vol. 2, Nos. 1/2/3, 2010

Assumptions Limit Most Common Models

Reduced Network Model (RNM)

00

cossin1

2

kk

n

kll

klklklklkkkkkmkkk

kk

MDwhere

FCDGEPM

• Each Synchronous Machine represented as voltage phase with constant magnitude E’ behind transient reactance

• Various network components assumed to be insensitive to changes in frequency

• Mechanical angle of Synch-Mach. rotor assumed to coincide w/ electrical phase angle of voltage phase behind transient reactance

• Loads represented as constant impedances thereby eliminating DAEs

• Mechanical power input to generators assumed constant (Pmk)

• Saliency is neglected

• Stator resistance is neglected

Modify to include

variable definition

Instantaneous power balance

*M Ghandhari, “Control Lyapunov Functions: A Control Strategy for Damping of Power Oscillations in Large Power Systems,” KTH,

2000

Simplest Network Model:

One-Machine Infinite Bus (OMIB)

BPPJ

BPPJ

BPPJ

m

m

em

sin

sin1

][1

max

max

RM

e

m

e

m

B

J

P

P

T

T

Turbine-Generator Rotor System

Mechanical turbine torque N-m

Electromagnetic counter torque N-m

Mechanical turbine power W (constant)

Electromagnetic counter torque W

Mass polar moment of inertia

Damping torque coefficient

Rotor shaft velocity in mechanical r/s

Angular velocity in electrical r/s

Power angle measured in electrical rad

RMRMem BJTT

BPPJH

PJH

VTH

m

sin

cos12

1

max

max2

Define Hamiltonian and Hamiltonian rate:

(Energy)

(Power flow)

RNM:

Design of the OMIB with UPFC Using Hamiltonian

Surface Shaping and Power Flow Control (HSSPFC)

)cos()sin(1 21max eee uuPP

UPFC general controller form:

Select nonlinear PID control laws:

dKKKu

dKKKu

t

sIDsPe

t

sIDsPe

eee

eee

02

01

coscoscos

sinsincos

Perform HSSPFC steps results in:

-Static stability condition:

-Dynamic stability condition (power flow):

))cos(1)(1(2

1max

2

sPeKPJH

dtdKPdtKPBt

sID ee

01max

2

max

- Region of stability increased

- Integrator: faster system response

max

1sinP

Pms

Hamiltonian Surface

H

UPFC

Variable Generation and UPFC Connected

to Infinite Bus

2

2

rrefTrefTm

mmm

emg

ra

refr

rr

garTrT

MKJu

uuu

TuT

MT

TTKJ

ref

ref

Wind

Turbine UPFC

Infinite B

us

emT

refTgarT

refrTT

TuKH

JTTKH

JJH

22

2

1

2

1

Select wind generation and UPFC nonlinear PID controllers from HSSPFC:

Model:

Hamiltonian and Power Flow:

Static stability condition and dynamic stability condition: similar to

previous example

cossinsin

)(

21max eeee

m

uuTT

PIDfnu

)(

)(

2

1

PIDfnu

PIDfnu

e

e

Variable (no longer constant)

Microgrid with UPFCs and Wind Turbine Connected to

Grid (Infinite Bus)

Wind

Turbine

22 E

Gen 2

12Xj

Infinite Bus

2222212322

112111130

11

2232112

12121131

2

1

2

1112111130

11

1121211311

1131212

2

1

2

1

121211211113

10

11

2

1

2

2222212321122222

cossin

cossin

sinsin

sinsin

0

0

cossin

sinsin

cos1cos12

1

2

1

sincossin1

;

,,

;

cossin1sin

2

2

uuCDP

uuCdKKK

CC

CCK

M

J

uuCdKKK

CCKJH

CCKJH

CuuCT

KdKKu

MKJu

uuu

TuTMT

TTKJ

uuCCDPM

m

t

IDTPT

t

IDT

PTT

PTT

e

D

t

IPm

rrefTrefTm

refrmmm

emgra

garTrT

m

mmm

mm

m

m

mmm

ref

ref

Gen 2:

Wind Turbine:

PV

Wind

State Estimator

Power Prediction

PV

Wind

Gen

UPFC

UPFC

UPFC Load

PV

Wind

AGC

PV

Wind

UPFC - Unified power flow controllers

High-Level Control Enables

Prioritization and System Adaptability

21

Gen

Gen

High-level control

optimizes system priorities

Nonlinear control

maintains stability

and performance

A Highly Interconnected Microgrid

Will Result from these Advancements

22

How do you connect

System components

in an efficient, cost

effective manner?

Generation

Sources or

other

Microgrids UPFC

Power Distribution

Connections Agent

Control

Agent

Control

Agent

Control

Agent

Control

Agent

Agent

Loads

Storage

Systems

Dispatchable

Generation

RE

Control

Agent

RE

Communication

network

These Microgrids will be Building

Blocks for Large Networks

23

UPFC

Power Distribution

Connections Agent

Control

Agent

Control

Agent

Control

Agent

Control

Agent

Agent

Loads

Storage

Systems

Dispatchable

Generation

RE

Control

Agent

RE

µG 2

µG 3

µG 1

Communication

network

24

DOE Has Identified Challenges That Must Be

Addressed to Increase PV Penetration

http://www1.eere.energy.gov/solar/solar_america/rsi.html

• 14 RSI Reports

• Advanced Grid Planning and Operations

• Utility Models, Analysis and Simulation Tools

• Advanced PV System Designs and Technology Requirements

• Development of Analysis Methodology for Evaluating the Impact of High Penetration PV

• Distribution System Performance Analysis for High Penetration PV

• Enhanced Reliability of PV Systems with Energy Storage and Controls

• Transmission System Performance Analysis for High Penetration PV

• Renewable System Interconnection Security Analysis

• Solar Resource Assessment: Characterization and Forecasting to Support High PV Penetration

• Test and Demonstration Program Definition to Support High PV Penetration

• Value Analysis

• PV Business Models

• Production Cost Modeling for High Levels of PV Penetration

• PV Market Penetration Scenarios

DOE is Addressing Wind Systems Integration

Challenges

Challenges:

• Transmission Interconnection & Congestion

• Lack of knowledge of operational impacts and integration costs of wind energy

• Shortage of power system professionals with knowledge of wind energy

• Policy treatment of wind energy as an electricity resource

DOE Action:

• Assess wind’s potential to serve our Nation’s electricity needs

• Develop tools to assist the electric utility industry analyze wind energy

• Perform operational and interconnection studies with industry stakeholders nationwide

• Provide education curriculum for the next generation of wind energy professionals

• Reach out to federal, state, and local stakeholders on the challenges and solutions to wind energy integration

Results:

• Set the path for wind industry to accelerate its penetration

• Increase body of knowledge on wind/grid interconnection

• Help grow the delivery of emission-free energy from roughly 1 percent to the AEI’s vision of 20 percent of our Nation’s electricity usage

25

Wind Radar Program

• Project Goals:

– Streamline existing federal requirements

– Establish coordinating mechanism

– Identify technical solutions (wind turbine & radar side)

• Long-Term Goal:

– Clear, timely, predictable Federal agency decision-making on wind siting processes

• Past Accomplishments

– Supported interagency working group

– Conducted radar baseline tests

– Created wind radar interaction fact sheet

– Performed case studies

– Performed industry outreach

– Supported IEA EU and NATO wind radar processes

– Participated in joint DOE-INL-FAA-DOD assessments

• Partners: SNL, INL, DOD, DOE, DOT, DHS, FAA, USDA, Interior, Commerce, others TBD

Next Steps, Summary, and Conclusions

• Next Steps

– Continue to refine models/controls

– Further numerical simulations

– Baseline microgrid system analysis

• Summary

– Renewables bring technical and operational challenges

– We cannot continue to treat stochastic renewables as negative loads

– Energy storage is not a silver bullet

• Conclusions

– Microgrids can effectively align renewable energy resources with mission needs

– Applied HSSPFC two-step process to analyze feedback controllers

– Optimize UPFC with variable generation requirements

Backup Charts

Control Theory Needs to be Expanded from Simple

to Complex Examples

Example system with control input v(t).

qkqdtkqktv dip )(

qRtvqC

qL )(1

qqkRqdtkqqkC

qLH dip

1

qVqVqTqkqC

qLH cp

)()(

2

1

2

1

2

1 222

Storage Load, L Generation, G

Control gains

PID controller:

Hamiltonian:

Derivative of the Hamiltonian:

Controller gains are chosen for specific

performance within the solution space defined by:

0)0()0(where0 ,0)()()( cc VVqqVqVqTH

0][00

dtLGdtH cc

R.D. Robinett and D.G. Wilson, “Nonlinear Power Flow Control Applied to Power Engineering”, Int’l Symposium on Power Electronics,

Electrical Drives, Automation, and Motion, SPEEDAM 2008, June 11-13, Ischia, Italy, pp. 1420-7.

• Devices have two modes of operation

– Operating in shunt with power line: injected currents are

controlled

Static Var Compensator (SVC)

– Operating in series with the power line: inserted voltages are

controlled

Unified Power Flow Controller (UPFC)

Controllable Series Capacitor (CSC)

Quadrature Boosting Transformer (QBT)

– AKA Controllable Series Devices

Flexible AC Transmission System (FACTS)

Power Electronic Controllers

• Offer greater control of

power flow

• Secure loading

• Damping of power system

oscillations

Injection Models of CSDs

•Valid for load flow and angle stability analysis

•Helpful to understand impact on power systems

•Model useful for purpose of developing control laws