Computing Advances for the Smart Grid Vision

Steve WidergrenIEEE Computer SocietySmart Grid Vision Project

IEEE STC 20143 April 2014

Forces Shaping the Future of Energy

Worldwide, demand for energy -- and costs -- are rising

Climate change policiesset environment as a parameter in the cost equation

Technology is enabling new ways to get things done

Customers want morevalue for their energy dollar

2

Smart Grid Overview

3Source: www.smartgrid.gov

The Challenge Ahead is Complex

4

Electrify transportation sector to reduce dependence on imported oil

Deliver significant amounts of renewable generation quickly

Maximize benefits of end-use efficiency and storage

Meet future carbon and emissions constraints

Affordable, Power

Reliable Power

Secure Power

Historical Expectations

Emerging Expectations

What will it take to get us there?

Policy

Science &Technology

Knowledge andTransparency

Capital

• government legislation

• R&D Labs• Universities• Private Sector

• Venture capital• Institutional investments• New business models

A New Relationship with Electricity

Understand how we use electricityConsumption by appliance and activityMajor contributors of electricity useEfficiency and waste

Appreciate its time-varying valueLike freeways, system loads vary over timeExpense and limits to capacity - congestion

Realize we can make a differenceAlternative operating behaviorWhere investments in efficiency can pay-off

Important enablersInformation on usage and costsAutomation to represent our desires

5

Electricity Ecosystem Transformation

6

Smart grid opportunity• Paradigm shift in

electricity• Fresh business

models• Innovative

government and institutional roles

Drives change, restructure• Cylinders of

excellence crumbling

• Rapid evolution

Creates work and jobs• Affects utility,

industrial, commercial, residential sectors

• Technology deployment

• New services

Necessitates education• Greater

educated workforce needed

• Curriculum reorganization

• Holistic, systemic methods

Contents

Purpose and ObjectivesProject TeamProject ApproachVision Summary– Architectural Concepts (Tier 1)– Functional Concepts (Tier 2)– Technological Concepts (Tier 3)

Future Work

7

Smart Grid Vision Project (SGVP)Objective– Develop a “Smart Grid Vision” Report

Role of Computing, 2030 and beyond Incorporate futuristic concepts

Purpose– Stimulate research and development, education, standards

Project Groundrules and Assumptions– There are no wrong visions for the future– Not bounded by current understanding of technology– Not constrained by today’s policies and practices– Not driving toward a common end vision – this was not an

engineering exercise– Visions may be complimentary or contradictory

8

Smart Grid Vision Project TeamManagement Team– Dan McCaugherty (Athena Sciences Corp)– Steve Widergren (US Dept of Energy, Pacific Northwest National Lab)– Dr. Joe Chow (Rensselaer Polytechnic Institute)– Bill Ash, Sponsor IEEE SA– Dr. Roger Fujii, IEEE CS Executive, Division VIII

Topic Area Leadership– Cyber Security – Dr. William Sanders (Univ of Illinois – Urbana)– Software Systems Engineering – Dr. Andreas Tolk (Old Dominion Univ)– Operations/Monitoring/Control – Dr. Dave Cartes (Florida State Univ)– Planning/Analysis/Simulation – Dr. Joe Chow (RPI)– Communications Networks – Dan McCaugherty (Athena Sciences Corp)– Computing – Steve Widergren (DOE PNNL)– Visualization/Data Management – Dr. Joe Chow (RPI)

9

Project ApproachElaborate Visions in three Concept Tiers – Architectural (AC), Functional (FC), and Technological (TC)Visions along lower Tiers often stimulated higher Tiers

10

Capabilities

Architectural Concepts (11 Total)

Supply Side (4)– Renewable resources– Energy storage & balancing– Integrated islands– Isolated islands

11

Demand Side (4)– Utility demand response– Aggregated local energy– Self-owned base energy– Electric transportation

System Concepts (3)– Coherent system operations– Complex autonomous adaptive systems– System and cyber security

Complex Autonomous Adaptive Systems

12

Level 1: Self-awareness, Context-awareness

Level 2: Self-configuring, optimizing,

healing, protecting

Level 3: Self-adaptive

Sensors and intelligence to monitor states and behaviorAwareness of location and ops environment

Defend against attacks & mitigate effectsRespond to changesManage performance & resources to satisfy different users

Self-adaptation services & information flow between distributed grid services - balancing global properties and grid health managementEmergent behavior emblematic

Hierarchical aspects of autonomy [1]

[1] Salehie, M., Tahvildari, L. 2009. "Self-adaptive software: Landscape and research challenges." ACM Transactions on Autonomous and Adaptive Systems (TAAS) 4, no. 2.

Functional Concepts (27 Total)

Systemic (8)Cyber Security– Information security– Control security– Privacy– Supply chain resilience– Intrusion tolerance

Software/Systems Engineering– Unsupervised autonomy– Social nodes– Autonomous validation

13

Enabling (7)Comm. & Networks– Intelligent devices/nodes– Converged communications– Hardware/Software refresh

Visualization and Data Mgt– State awareness– Failure awareness, restoration

Markets and Economics– Wholesale power market– Dynamic demand side markets

More Functional Concepts – Performance (12)

Operations, monitoring, and control subtopic (8)– Bulk system transmission dynamic operations– Operations congestion detection– Power flow forecasting in distribution networks– Direct load control events– Island-to-island stable power flow control– Automated grid load flow coordination– Process coordination of industrial manufacturing– Commercial and industrial building coordination

Planning, analysis and simulation subtopic (4)– Bulk system transmission planning– Asset management and maintenance– Resilient systems– Command, control, and automated functions

14

Automated Intrusion Tolerance ConceptHighly integrated cyber-physical components– Detect attacks automatically– Diagnose root cause accurately– Real-time, adaptive response to malicious adversaries

Regular manual intervention unnecessaryEach time instant accurately determines security stateMonitor and detect exploitation of known vulnerabilitiesIntrusion tolerance strategies’ algorithms compare criticality of assetsMathematical decision making framework selects optimal tolerance strategy

15

Distributed System Architecture (4)– Self-integrating systems and standards – Distributed multi-agent architecture – Virtual computing architecture– Messaging-oriented middleware

Computer Applications (7)– Market-Inspired (transactive) control– Monitoring and control/modeling and simulation tools– Signal processing for control, protection and performance

qualification/performance monitoring– State estimation analysis algorithms– Contingency, preventive, and corrective control analysis– Stochastic analysis for system operations, planning, forecasting – Prognostics and asset management

Technological Concepts (21 Total)

16

Technological Concepts

Information Science (4)– Visualization– Artificial Intelligence, data analytics, fast mathematics and high-

performance computing– Internet and real-time systems– Software verification and validation

Cyber Security (6)– Trusted component validation– Portable identity – bidirectional authentication support– Hierarchical sense making (HSM) and collaborative HSM agent

networks – Massive parallel pattern detection– Patterns for agile self-organizing security– Information security technology

17

18

Distributed Multi-Agent Architecture

Bottoms-Up organic control- Individual needs negotiated- Regional imbalances

smoothed- Power flow checked against

stability constraints

Local optimization coordinated step-wise up, forming ever-larger networks

Monitoring & Control/Modeling & Simulation (M&S) Tools

Problem:Proprietary control system M&S tools do not interactDifficult to leverage integrated M&S component capability for dynamic SCADA system simulation

19

Future Vision:Standard data exchange formatsScalable fidelity (size, performance, level of detail)Accommodate third party M&S tools (e.g., state estimation)Grid device vendors supply compatible modelsUses – grid design, planning, operations

Software Verification and Validation

Future Smart Grid is beyond the scale of existing high integrity systems

Expresses proper models of system element behaviors subject to formal methods

– Use of domain specific modeling languages

– Machine learning needed to construct the formal specifications

– Computational intelligence needed to guide the verification mechanism

Continuous run-time verification needed for adaptive elements

– High performance simulations constantly evaluating emergent behaviors

20

21

Technological Concepts

1: S

elf-

inte

grat

ing

syst

ems a

nd st

anda

rds

2: D

istr

ibut

ed m

ulti-

agen

t arc

hite

ctur

e

3: V

irtua

l com

putin

g ar

chite

ctur

e

4: M

essa

ging

-orie

nted

mid

dlew

are

5: M

arke

t-In

spire

d (t

rans

activ

e) co

ntro

l

6: M

onito

ring

and

cont

rol/

mod

elin

g an

d si

mul

atio

n to

ols

7: In

form

atio

n pr

oces

sing

for c

ontr

ol,

prot

ectio

n an

d pe

rfor

man

ce

qual

ifica

tion/

perf

orm

ance

mon

itorin

g

8:St

ate

estim

atio

n an

alys

is a

lgor

ithm

s

9: C

ontin

genc

y, p

reve

ntiv

e an

d co

rrec

tive

cont

rol a

naly

sis

10: S

toch

astic

ana

lysi

s for

syst

em o

pera

tions

, pl

anni

ng, f

orec

astin

g

11: P

rogn

ostic

s and

ass

et m

anag

emen

t

12: V

isua

lizat

ion

13: A

rtifi

cial

Inte

llige

nce,

dat

a an

alyt

ics,

fast

m

athe

mat

ics a

nd h

igh-

perf

orm

ance

co

mpu

ting

14: I

nter

net a

nd re

al-t

ime

syst

ems

15: S

oftw

are

verif

icat

ion

and

valid

atio

n

16: T

rust

ed co

mpo

nent

val

idat

ion

17: P

orta

ble

iden

tity

– bi

dire

ctio

nal

auth

entic

atio

n su

ppor

t

18: H

iera

rchi

cal s

ense

mak

ing

19: M

assi

ve p

aral

lel p

atte

rn d

etec

tion

20: P

atte

rns f

or im

plem

entin

g ag

ile se

lf-or

gani

zing

secu

rity

21: I

nfor

mat

ion

secu

rity

tech

nolo

gy

Functional Concepts1: Information security x x x x x x x2: Control security x x x x x x3: Privacy x x x4: Supply chain cyber resilience in software and hardware x x x x x x x x x x5: Automated intrusion tolerance x x x x x x6: More dependence on unsupervised autonomy x x x x x x x x x7: Social nodes x x x x x x x x x x8: Smart Grid autonomous validation x x x x x x x9: Proliferation of intelligent devices and nodes x x x x x x x x x x x x x x10: Secure converged communications x x x x x x x x11: Smart Grid hardware and software refresh x x12: State awareness x x x x x x x x x x x13: System failure awareness, emergency response and system restoration x x x x x x x x x x x x x x14: Wholesale electric power market policy, operation and design x x x x15: Emergent dynamic demand side markets x x x x x16: Bulk system transmission dynamic operations x x x x x x x x x17: Operations congestion detection x x x x x x x x18: Power flow forecasting in distribution networks x x x x x19: Direct load control events x x x x x20: Island-to-island stable power flow control x x x x x x x x21: Automated grid load flow coordination x x x x x x x x22: Advanced process coordination of industrial manufacturing x x x x x x x x x23: Commercial and industrial building coordination x x x x x x x x x24: Bulk system transmission planning x x x25: Asset management and maintenance x x x x x x26: Resilient systems x x x x x x x x x x x x27: Advanced command, control, and automated functions x x x x x x x x x

Future Work

Pursue opportunities for StandardsPromote education and training for relevant computer science disciplinesRe-address the visions in 2 to 5 years– Impact of emerging sciences– New smart grid concepts

Business/economic value propositions Lessons learned Promising technologies

22

Authors Acknowledgement

23

Al Valdes, University of IllinoisAndreas Tolk, Old Dominion UniversityAndrew Wright, N-Dimension SolutionsAnnabelle Lee, Electrical Power Research InstituteAranya Chakrabortty, North Carolina State UniversityDan McCaugherty, Athena Sciences CorporationDave Cartes, Florida State UniversityDavid P. Chassin, Pacific Northwest National LabDave Hardin, EnerNOCDiane Hooie, National Energy Technology LabEdmond Rogers, University of IllinoisEugene Litvinov, ISO New EnglandGlen Chason, Electrical Power Research InstituteJennifer Bayuk, Stevens Institute of TechnologyJianhui Wang, Argonne National LaboratoryJoe H. Chow, Rensselaer Polytechnic InstituteKeith Thal, Sypris SolutionsKlara Nardstedt, University of IllinoisKumar Venayagamoorthy, University of MissouriLeigh Tesfatsion, Iowa State UniversityMark Blackburn, Stevens Institute of Technology

Mark Smith, Sandia National LabMathias Uslar, OFFISNikos Hatziargyriou, National Tech University of AthensPaulo Ribeiro, Eindhoven University of TechnologyPrabir Barooah, University of FloridaRakesh Bobba, University of IllinoisRick Dove, Paradigm Shift InternationalSaman Zonouz, University of MiamiSean Meyn, University of FloridaSebastian Lehnhoff, OFFISSteve Ray, Carnegie Mellon UniversitySteve Widergren, Pacific Northwest National LabSteven Low, California Institute of TechnologySvetlana Pevnitskaya, Florida State UniversityTim Yardley, University of IllinoisTodd Montgomery, InformaticaWenxin Liu, New Mexico State UniversityWilliam Sanders, University of IllinoisZbigniew Kalbarczyk, University of IllinoisZhi Zhou, Argonne National Lab

Smart Grid Vision Report for Computing Link

http://www.techstreet.com/ieee/products/1857774

Contacts:Dan McCaughertydan@athenasciences.comSteve Widergrensteve.widergren@ieee.org