from adapting to transforming in a ch i tilithanging...

From adapting to From adapting to transforming in a transforming in a

h i tilith i tilitchanging utility changing utility industryindustryEelco de Jong

industryindustry

Georgia Tech Energy SeriesNovember 12th, 2014

Eelco de Jong

CONFIDENTIAL AND PROPRIETARYAny use of this material without specific permission of McKinsey & Company is strictly prohibited

Key themes for todayMCKINSEY PROPRIETARY AND CONFIDENTIAL

Transforming, not adapting

Disruption: glass half empty

Actions that define winners

McKinsey & Company | 1

Are these headlines alarmist or “in-the-money”?MCKINSEY PROPRIETARY AND CONFIDENTIAL


What are the facts?

Demand growth forecasts are being adjusted… down

A f d i EIA AEO P

Historical growth rateLong-term electric consumptionannual growth rate1

2.62.4

As forecasted in EIA AEO, Percent

1.51.8

0.7

1.1 -1.1%

McKinsey & Company | 31 Normalized to 2005-25 CAGR for all AEOs

1960-1990 1990-2010 082004 06 2010

Five technology-based drivers are transforming the industryMCKINSEY PROPRIETARY AND CONFIDENTIAL

Unconventional gas and oil

New “economic pillar” that rebases the value of generation

Centralized renewables

Policy and technology curve leading to commercially competitive build

Distributed generation

Fast decreasing costs improve competitiveness causing relocation of generation to lower voltage

Energy efficiency

Innovations, policies and big data driving next wave of adoption; integrating customer into supply curve

Customer awareness

g g pp y

Customer expectations, experience and activity rising to unparalleled levels


levels

Views on transformation change

When you see a So, despite the risks that

MCKINSEY PROPRIETARY AND CONFIDENTIAL

When you see a disruptive technology

come into your space if you don’t

So, despite the risks that a rapidly growing level of

distributed energy resources penetration and space, if you don t

embrace it… the people who try and cling to the past get

resources penetration and other disruptive challenges

may impose, they are not currently being discussed cling to the past get

rolled overcurrently being discussed

by the investment community and factored

i t th l tiDavid Crane,

CEO NRG

into the valuation

Edison Electric Institute


Cutting through the hype--utilities need to develop a perspective on these questions


Growing consensus▪ What is the threat?

▪ How real is it?

How concer- Growing consensus

“yes” but expected timing differs

▪ How real is it?

▪ How impactful?

▪ How immediate?

nedshould utilities b ? How immediate?be?

H i ti▪ Is there a viable defensive

l ? Huge variation spanning “disaster” to new ideas for growth to limited mindshare

play?

▪ Is there an offensive opportunity?

What should regulated to limited mindsharey

▪ How do options get framed within a 4-6% EPS growth aspiration?

gplayers do? Glass is half

empty…or half


aspiration? p yfull?

If past is a guide: Utilities have adapted to significant changes in last 15 years, but “traditional” core competencies have not changed


Success in “traditional” areasSuccess in “traditional” areas of strength Mixed record in new areas

▪ Retail and customer-facing ▪ New generationbusinesses

▪ New technology businesses▪ T&D value creation

So far utilities have had to adapt – not transform – butadapt not transform but future is going require transformation

McKinsey & Company | 7SOURCE: McKinsey Energy Practice




Actions that define winners


Trends are having a meaningful impact on load growthDEMAND DESTRUCTION MCKINSEY PROPRIETARY AND CONFIDENTIAL

100% = baseline load forecast for 2023 Kwh load

Calculated demand across selected US States

2%100%

5%

2%

18%

2%2%

-22%7%

83%

5%

Efficiency standards

Competitive efficiency b i

New energy efficient t h l

Utility EE programs /d d

Total demand

Distributed Solar PV

Co-generation

2023 demand

10 year growth

2012 demand

McKinsey & Company | 9SOURCE: McKinsey Electric Power Practice

business models

technology/demand elasticity

Price erosion the key enabler for residential LEDs

Compact fluorescent lights took off

ENERGY EFFICIENCY

… and LED bulbs are rapidly

CFL share1 and ASP

p garound $5-10 …

Retail price of LED retrofit bulb, Dollars

p yapproaching that price point

60

(USD)

60

(Percent)

40

50

40

50

20

30 30

20

0

10

’951990 ’05’00 ’10

10

0

McKinsey & Company | 10SOURCE: U.S. DOE, U.S. EPA, GE Web site

951990 0500 101 U.S. market shares on annual replacement sales.

100 kW Rooftop c-Si multi-crystalline PV solar system

The costs of solar panels continues to go down Wafer

BOS

Cell

ModuleSOLAR PV COST OUTLOOK Installation type

5 KW Rooftop (residential)100 KW Rooftop (commercial)

Stand-alone Cost Decrease10 MW Ground-mounted (wholesale)


Best-in-class installed system cost (no margins)USD/Wp (2011 dollars)

100 kW Rooftop, c Si multi crystalline PV solar systemPolysilicon

Levelized Cost of Energy (LCOE)for 100 kW Commercial Rooftop System1

USD/kWh (2011 dollars)USD/Wp (2011 dollars)

0 280.300.320.340.36

3.6

4.02011-2015 2016-2020

Polysilicon price

USD/kWh (2011 dollars) 2020 Approx. Commercial Retail Prices2

V. Good SunGood SunModerate Sun

0.200.220.240.260.28

2.4

2.8

3.2

8%2%

6%8%

10%

Productivity

Procurement

Optimized s stem design

ScaleIncremental techimprovements

Polysilicon pricedecline

Italy

New York

Japan

0.100.120.140.160.18

1 2

1.6

2.01% 6%

6%4%

1%5%

system designProductivity

ProcurementScale

Incremental techimprovements Germany

California

FranceAustralia

New York

Spain

0.060.08

0.8

1.2Optimized system design India

1000 GW

McKinsey & Company | 11SOURCE: Industry experts, Photon, GTM, NREL, EIA, Enerdata, press search, company websites, McKinsey analysis

1 Assumed 7% WACC, annual O&M equivalent to 1% of system cost 0.9% degradation per year, constant 2011 dollars, 15% margin at module level (EPC margin included in BOS costs). 2 Very good sun conditions = 19% capacity factor, good sun conditions = 16% capacity factor, moderate sun conditions = 11% capacity factor.

Cost reduction is bringing more states “in the money”SOLAR PV – ADOPTION RATES

20122013201420152016201920202021202220232024202520262027202820292030

States at “grid parity” for residential distributed solar

McKinsey & Company | 12SOURCE: McKinsey analysis

Actual observed installation rates: rapid solar uptake at “grid parity”SOLAR PV – ADOPTION RATES

5.0

Solar adoption rate% of total demand erosion in that year

HIAZ

3.0

2.5 NMNJ3

HI

1.5

2.0

NJ-Com3

GermanyNV

1.0

0.5

NJ Com

060400-20-40-60-140 20-30-50 10-10 30 50 70

McKinsey & Company | 13SOURCE: BMU, BSW, GTM Research, Ventyx Energy Velocity, press search; team analysis

Price discount %Solar LCOE (incl. incentives) vs. residential rate

Conservative view: distributed solar PV hits a tipping point in 2020 and will grow to ~180 GW installed base by 2030

SOLAR PV DEMAND OUTLOOK MCKINSEY PROPRIETARY AND CONFIDENTIAL

Expected installed capacityGW, distributed solar, residential + commercial

20%

180 GW

+20% p.a.All other

GANJNY

83

29

10 GW TXCAFL


2015 203020252020SOURCE: McKinsey Electric Power Practice

Winning players are aggressively working down “soft costs” NORTH AMERICA SOLAR COST FOCUS MCKINSEY PROPRIETARY AND CONFIDENTIAL

3.32

Expected reduction in installed PV solar system costs – residential

USD/Wp

1.421.63

3.32

0.520.67

0 43

0.380.320.59 0.68

0.43

Installer margin

2011 Others1Installationlabor

Customer acquisition

2020

McKinsey & Company | 15SOURCE: McKinsey analysis, Expert interviews, NREL, LBNL

1 Includes Sales Tax, Permit Fees and PII (permitting inspection and interconnection)2 Variability based on sales tax cut realization of 50% in pessimistic scenario




A ti d fi i iActions defining winners


ImplicationsMCKINSEY PROPRIETARY AND CONFIDENTIAL

Players who continue to rely only on value from centralized generation and current utility model will struggle

Winners are beginning to stake out plans anchored in long-term growth perspective

Winners are redefining their risk profile – no silver bullet, need to make multiple bets


a e u p e be s

McKinsey’s view is that companies need to simultaneously manage “three horizons” to take the glass half-full approach


“Find ways to grow”

“Find new ways to change the industry”

“Sustain earnings growth to invest in the future”

“Find ways to grow”

Find new growth

Adopt new business models

Optimize the core

Find new growth opportunities

▪ Master new technologies and products

▪ Get close to the▪ Protect the core

b i ▪ Get close to the customer

▪ Innovate the business model – including rate

business▪ Invest in growth

opportunities: transmission,

▪ Optimize operational performance

▪ Sustain EPS in a tough structuresrenewables

▪ M&A?

genvironment

▪ Create headroom to fund future growth


To win, utilities must develop three characteristics

Mastering technology / product development

Getting closer to customers

Developing new business models and services

▪ Ability to execute ▪ Low cost customer ▪ Ability to monetize quickly – like a start-up

▪ Procurement of technology @ lowest TCO

acquisition▪ Ability to truly segment

customer based on behaviour

information / customer combination

▪ What partnerships can help utilities


paccelerate?

Opportunities needs to be understood from both regulated and unregulated view

NOT EXHAUSTIVE


Customer apps Distributed Gen Storage PHEV/EVEnergy efficiency Data collection IT integration

Rate based solar

3rd party data monetization

Rate based storage


Structured programs

Customer Sat. caustion

Analytics for Auto OEM

Regulated

CHP mgmt Critical storageTime-of-Use driven voluntary

Awareness-driven prog.

Low income acceleration

Analytics for Auto OEM

C&I aggregation

ISO

Microgrid (rate basing emphasis)

U l t d

ISO energy bidding

Installation services

Load aggreg. for DR

Home energy retail

On-site mgmtCHP end-to-end

Bundle with Cust. Apps

Charging infrastructure

C tISO CHP


IT integration services

Program consulting

End-to-end delivery

Data architecture

Un-regulated Customer segmentation

ISO energy bidding

CHP comm. mgmt

Charging software

Microgrid (scope of service spans)

McKinsey & Company | 20MCKINSEY PROPRIETARY AND CONFIDENTIALSOURCE: McKinsey Electric Power Practice

Microgrid (scope of service spans)

NOT EXHAUSTIVEUtilities actively pursuing “active response”MCKINSEY PROPRIETARY AND CONFIDENTIAL

Utilities using unregulated arms

Utilities pushing regulated boundaries

Utilities using traditional channelsProducts

▪ Smart meter i / l ti

Description

Efficiency & home solutions

services/solutions▪ Home solutions

(e.g. HVAC repairs, home improvement, smart devices)

+smart devices)

▪ C&I solutions

Distributed

▪ Solar PV▪ Mini CHPDistributed

resources

▪ Demand side

M bilit▪ EV charging

Flexibility management


Mobility stations

The challenge – utilities need to think “beyond commodities”

Energy retailers bundled free energy

Old model“The pure-play energy manager”

“

New model: make money on the added services

bundled free energy management and services products

▪ “Energy management package”: $8.99/month plus one-time $50 activation fee (on top of standard home security package)home security package)

▪ Includes 1 smart thermostat, 1 appliance control, 12 EE light bulbs and home energy monitoring/advice

“The digital home provider” ▪ Energy management controls and

monitoring bundled with home automation / security package from incumbent telco or cableco

McKinsey & Company | 22SOURCE: McKinsey Energy Practice

€/MWh €/customer

Companies are exploring alternatives to our current rate model

Challenge Examples

No accurate price▪ Can we sustain a volumetric rate vs. largely

fixed cost base?No accurate price signals ▪ Net metering -- who is paying for back-up and

grid?

No cost recovery of core services

▪ Services that are “free”: universal access, back-up power

▪ Are customers willing to pay the full costs of g p ythese?

▪ What if utilities were allowed to build distributed solar?Limited ability to play in

competitive sectorssolar?

▪ What would happen if utilities would be in charge of your energy efficiency?

McKinsey & Company | 23SOURCE: McKinsey

What does the airline industry teach us? Average adjusted yield on US domestic airline routes, cents per seat mile


Organized

What does the airline industry teach us? ROIC for representative groups

Organized

Jet providers

Booking platforms

Organized labour Direct competitorsAirports

Jet providers

Booking platforms

Organized labor Direct competitorsAirports

207 4providers

Fuel suppliers Substitutes

providers

Fuel suppliers Substitutes

16 164

18 2

Lessors

Travelers

Lessors

Travelers

107 5

127

Airframe makersTravel Agents

Web travel agents

Low cost carriersAirframe makers

Travel Agents

Web travel agents

Low cost carriers

7 5 7


gg

Glass half full or half empty?


Integrating Energy Efficiency into the Distributed g g gy yEnergy Resource Mix

M a r i l y n B r o w nB r o o k B y e r s P r o f e s s o r o f S u s t a i n a b i l i t y

S c h o o l o f P u b l i c P o l i c yyG e o r g i a I n s t i t u t e o f Te c h n o l o g y

C o l l a b o r a t o r s :B e n S t a v e r & A l e x S m i t h ( G e o r g i a Te c h )

J o h n S i b l e y ( S o u t h f a c e E n e r g y I n s t i t u t e )

T H E A G I L E U T I L I T Y: A L I G N I N G D I S T R I B U T E DT H E A G I L E U T I L I T Y: A L I G N I N G D I S T R I B U T E D G E N E R AT I O N W I T H C O N S U M E R D E M A N D

N o v e m b e r 1 2 , 2 0 1 4

Background: Challenges to the Traditional Utility Business ModelTraditional Utility Business Model

Recent trends are challenging the traditional cost-of-servicetilit b i d lutility business model:o Technologieso Economicso Economicso Policies

Vicious or Virtuous Cycle?

These trends (“disruptive threats”) are placing upward tilit tpressure on utility rates:

Are alternative business models needed?

Source: Peter Kind (2013). Disruptive Challenges: Financial Implications and Strategic Responses to a Changing Retail Electric Business. Edison Electric Institute.

Origins of our Research Project

Project goal: develop a tool to illuminate the impacts of Project goal: develop a tool to illuminate the impacts of ratepayer-funded EE programs and advance the debate on best utility business practices

What are the pros and cons of different approaches to allocating the costs and benefits of ratepayer-funded energy-efficiency (EE) programs. gy y ( ) p g

For more information on the project, see: Marilyn A. Brown, Benjamin Staver, Alexander M. Smith, and John Sibley. 2014. "Business Models for Utilities of the Future: Emerging Trends in the Southeast," School of Public Policy, Georgia Institute of Technology, Working Paper #84, http://cepl.gatech.edu/drupal/node/69.

Thanks to the Energy Foundation for their support.

Methodology

Review business models in the Southeast and define a “prototypical” approach that uses three features:a prototypical approach that uses three features: the recovery of program costs, the treatment of lost contributions to fixed costs, and , the provision of utility incentives.

Compile public data on a “stereotypical” southeastern utility and EE program.

Use GT-DSM to examine the prototypical happroach

GT-DSM and its manual can be downloaded at:http://cepl gatech edu/drupal/node/69http://cepl.gatech.edu/drupal/node/69

The Prototypical Approach Used in the SoutheastSoutheast

The prototypical approach is highlighted below forThe prototypical approach is highlighted below for each “leg” of the three-legged stool:

Examples in the Southeastp

Business Model Feature Extent of Usage in the Southeast of

sts Amortized with a Carrying Cost Not used by southeastern electric utilities

Recovery

oProg

ram

Cos

Expensed and Recovered Contemporaneously

General practice across the Southeast

P

g Utility

from

y Sales

Straight Fixed Variable Rate Used by some gas utilities in the Southeast but not used by southeastern electric utilities

Lost Revenue Adjustment Mechanism Arkansas Kentucky Louisiana Mississippi

Decou

pling

Profits

fElectricity Lost Revenue Adjustment Mechanism Arkansas, Kentucky, Louisiana, Mississippi,

North Carolina, South Carolina, and Virginia Per Customer Decoupling A number of states in the U.S., but none in

the Southeast

ision of

orman

ce

entiv

es

Shared Savings based on net benefits from the Program Administrator Cost (PAC) test

Georgia, North Carolina, and South Carolina

Shared Savings based on net benefits Arkansas and Kentucky

Prov

iPe

rfo

Ince Shared Savings based on net benefits

from the Total Resource Cost (TRC) test Arkansas and Kentucky

Return on Program Costs Virginia

GT-DSM The GT-DSM model is laid out in three Sectors: The Customer Sector focuses on the electricity rate and utility

bill and how an EE program affects them. Residential and commercial programs can be modeled, either as bundled programs or as individual programs.

The Utility Sector focuses on the revenues and costs to the utility and how an EE program affects those revenues and costs. Three modules: the Performance Incentive Module, thecosts. Three modules: the Performance Incentive Module, the Deferred Capital Investment Module, and the Rate Case Module.

The Cost Benefit Analysis (CBA) Sector produces estimates of The Cost-Benefit Analysis (CBA) Sector produces estimates of four of the standard cost-effectiveness tests for utility-operated EE programs that account for different stakeholder perspecti es to energ efficiencperspectives to energy efficiency.

The GT-DSM Model is Laid Out in Th S tThree Sectors

Customer Sector:• Impact of EE program on electricity rate and utility bill.• Two types of rate classes (Residential & C/I)

o Bundled programs or individuallyo Bundled programs or individually

Utility Sector:I t f EE d tilit t• Impact of EE program on revenues and utility costs.

o Performance incentiveo Deferred capital investmento Rate case

Cost-Benefit Analysis Sector:• Estimate for four standard cost-effectiveness testsEstimate for four standard cost effectiveness tests

o Utility-operated EE programso Alternate stakeholder perspectives for EE

The “Stereotypical” Southeast Utility

Based on public filings describing the Georgia Power Company in 2012 and the energy-efficiency programsCompany in 2012 and the energy efficiency programs proposed by the company in its 2013 IRP filing. The Georgia Power Company is the largest utility in Georgia. We do

not purport to replicate it in GT DSMnot purport to replicate it in GT-DSM. Serves 2.4 million customers, with annual sales of 81.1 TWh and a

peak demand of 15.4 GW. The number of customers is expected to grow by 1 0% per year and sales and demand are expected to growgrow by 1.0% per year, and sales and demand are expected to grow 1.24% annually. Annual earnings are $1.2 billion based on an 11.25% return on equity from a rate base of $19.5 billion.

Fuel and purchased power costs are assumed to increase by 6 5% Fuel and purchased power costs are assumed to increase by 6.5% per year. Major capital investments are programmed over the next several years to build out new baseload capacity, make environmental retrofits and improve transmission and distributionenvironmental retrofits, and improve transmission and distribution facilities.

The “Stereotypical” Southeast Utility (cont.)

Average rates are 12 ¢/kWh for residential customers and 8 ¢/kWh f i l d i d t i l t¢/kWh for commercial and industrial customers. Residential rates are collected through volumetric charges. The commercial and industrial rate includes a volumetric charge of 6

¢/kWh, plus a demand charge, equal to $10/kW in the first year.

The utility has a peak cost period of 2-7pm on weekdays from June to September. This represents roughly 3.7% offrom June to September. This represents roughly 3.7% of the year. Rate cases are filed every three years.

The capital structure is 54% equity and 46% debt, with a f f % fcost of debt of 4.2%. The weighted average cost of capital

is 8%.

The Portfolio of Residential EE ProgramsThe Portfolio of Residential EE Programs

A collection of programs:p g The end-use specific programs include lighting, air

conditioning, and other large home appliances. The whole home programs cover both existing and new The whole home programs cover both existing and new

homes and generally include insulation and select large appliances.

$ $ Annual costs of $8.3 million for incentives and $9.8 million for administrative costs.

Set to save 57.8 GWh and 10.2 MW annually for each year of y ythe measure and program lifetimes.

Average measure life is assumed to be 10 years. 8% of the residential energy efficiency program savings occur 8% of the residential energy-efficiency program savings occur

during the utility’s peak period, much more than the roughly 3.7% of the year that occurs during the peak.

The Portfolio of Commercial EE ProgramsPrograms

Targets both small and large commercial buildings. The small commercial program includes appliances, lighting, and

insulation. The other commercial programs are from either a long list of prescriptive

facility improvements or from a custom built incentives program. Annual costs of $13.7 million for incentives and $5.5 million for

administrative costs. Designed to save 241 GWh and 55.3 MW annually for each year of the

measure and program lifetimes. The average measure life is assumed to be 15 years. Since the

f 10programs are proposed to deploy measures for 10 years and the measures are assumed to operate for 15 years, our analysis of the impacts of these programs extends for 25 years.

10% of program savings are during the utilities peak period which is 10% of program savings are during the utilities peak period, which is more than for the residential program and also much more than the roughly 3.7% of the year that constitutes the peak.

Res ltsResults

The Impact of Commercial EE Programsp gUtility Economics Customer Economics

R tAverage

C N ACumulative Earnings in

$Billions

Return on Equity (%)(25-Year Average)

Commercial Energy

Bill ($/year)

Participant Energy Bill

($/year)

Non-participant Energy Bill

($/year)

Average Commercial Energy

Rate (¢/kWh)

Utility Without EEUtility Without EE Programs 47.02 11.46 28,107 NA NA 12.37

+ Commercial EE Programs 45.22 11.04 26,747 22,293 28,070 12.35g

+ Program Cost Recovery & Shared Savings Incentives

45.51 11.10 26,782 22,322 28,106 12.37

• Utility economics can be hurt by EE programs, but all customers can benefit.Th t t i l b i d l t 99 7% f tilit i b t t i

+ Prototypical Business Model 46.79 11.41 27,015 22,516 28,351 12.50

• The prototypical business model restores 99.7% of utility earnings, but rates rise by 1.0%. ROE exceeds authorized level of 11.25%.

• Rates still lowered after recovery of program costs and incentives.

The Impact of Residential EE Programsp gUtility Economics Customer Economics

Average

Cumulative Earnings in

$Billions

Return on Equity (%)(25-Year Average)

gResidential Energy

Bill ($/year)

Participant Energy Bill

($/year)

Non-participant Energy Bill

($/year)

Average Residential

Energy Rate (¢/kWh)

Utility Without EEUtility Without EE Programs 47.02 11.46 2,533 NA NA 19.23

+ Residential EE Programs 45.84 11.18 2,484 2,343 2,533 19.22

+ Program Cost Recovery & Shared Savings Incentives

45.98 11.22 2,488 2,346 2,537 19.25

• Utility economics can be hurt by EE programs, but all customers can benefit.

+ Prototypical Business Model 46.88 11.43 2,511 2,367 2,560 19.42

y y p g• The prototypical business model restores 99.7% of utility earnings, but rates rise

by 1.0%. ROE exceeds authorized level of 11.25%.

The Prototypical Business Model’s Impact on RatesImpact on Rates

Rates decline with EE Programs, but increase when lost utility revenues are recovered.

1 8%

2.0%

tes

Residential Customers Commercial/Industrial Customers

1.2%

1.4%

1.6%

1.8%

entia

l and

C/I

Ra

0 4%

0.6%

0.8%

1.0%

crea

se in

Res

ide

0.0%

0.2%

0.4%

% In

c

Note: Compared to operating an EE program without any business model features

Year

Average Change in Energy Bills

SFVR

Com

mer

cial

Prototype

C

SFVR

Res

iden

tial

0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5%

Prototype

A E C t P ti i t E C t N P ti i t E C t

Note: Compared to operating an EE program without any business model features.SFVR = straight fixed variable rate for lost revenue recovery

Avg Energy Cost Participant Energy Cost Non-Participant Energy Cost

Findings: Impact on Utility Earnings

SFVR

cial

Base Case

Prototype

Com

mer

c

Authorized Earnings Earnings Without EE

Prototype

SFVR

entia

l

Base Case

Prototype

Res

ide

Earnings Without EE Authorized

Note: Compared to operating an EE program without any business model

$- $0.2 $0.4 $0.6 $0.8 $1.0 $1.2 $1.4 $1.6 $1.8 $2.0 Change in Earnings ($ Billions)

Program Cost Decoupling Incentive Authorized Base Case

p p g p g yfeatures.SFVR = straight fixed variable rate for lost revenue recovery

The “DRIPE” Effect – Demand Reduction Ind ced Price EffectInduced Price Effect

EE programs reduce rates by eliminating a greater proportion of more expensive on-peak than off-peak fuel expenditures. D f i “ b ild ” i t l t fit d Deferring “new builds,” environmental retrofits, and T&D upgrades would be additional benefits, but these are not specified for the stereotypical utilitythese are not specified for the stereotypical utility.

Even if the utility recovers program costs and is paid incentives, there can be downward pressure on rates , pbecause of the “DRIPE” effect.

But with this combination, the utility is still left short of the earnings and ROE it would receive without the EE programs.

Conclusions

Utility earnings are reduced by EE programs, but they can be restored by alternative business modelsthey can be restored by alternative business models.

With these alternative models, EE programs: cause modest increases in electricity rates,y , reduce average bills for all customers, significantly cut the electricity bills of participants.

Depending on the choice of business model, non-participant utility bills may also decline. S l ti th i ht b i d l i i t t t th Selecting the right business model is important to the future of EE programs.

Tying reward to performance is an important Tying reward to performance is an important principle for regulatory design.

Conclusions

The “utility of the future” discussion has largely y g yfocused on the rush to DG

Yet EE exerts similar stresses to utility economics and is likely to “scale up” significantly

With DER expanding and climate policy likely, we d t d fi t t i th t th tilit i d tneed to define strategies so that the utility industry

and consumers can continue to prosper as the grid evolvesgrid evolves

For More Information

Dr. Marilyn A. BrownBrook Byers Professor

28

Brook Byers ProfessorGeorgia Institute of TechnologySchool of Public PolicyAtlanta, GA [email protected] and Energy Policy Lab: http://www.cepl.gatech.edu

My WordMy Word Cloud

GT Understanding of CBA Tests Ratepayer Impact Measure (RIM)

Benefits: Avoided Supply Costs (Production and T&D)pp y ( ) Costs: Lost Revenues Caused by Reduced Sales, Program

Administration Costs, Program Incentives to Participants Total Resource Cost Test (TRC) Total Resource Cost Test (TRC)

Benefits: Avoided Supply Costs (Production and T&D) Costs: Program Administration Costs, Participant Measure CostsP Ad i i t t C t T t (PAC) Program Administrator Cost Test (PAC) Benefits: Avoided Supply Costs (Production and T&D) Costs: Program Administration Costs, Program Incentives to g g

Participants Participant Cost Test (PCT)

Benefits: Bill Savings Program Incentives to Participants Benefits: Bill Savings, Program Incentives to Participants Costs: Participant Measure Costs

GT Clean Energy Series The Agile Utility: Aligning Distributed Generation

with Consumer Demand

John Rossi – SVP Corporate Strategy Comverge [email protected]

Aligning Distributed

Generation with Consumer

Demand

Aligning Consumer Demand

with Generation

©2012 Comverge – Confidential and Proprietary

2

Outline

Comverge background

Evolving drivers for Demand Response

Playing field

Technology discussion


3

Solutions

Provider of Demand Management Solutions

Utility Clients

Comverge by the Numbers

6,000,000+ energy management devices deployed

1,600,000+ residential participants enrolled into DR programs

500+ Utility Customers

4

Solutions Demand Response Energy Efficiency Customer Engagement

Products IntelliSOURCE software platform IntelliTEMP thermostats IntelliPEAK load control switches

Services Program Design and Marketing Field Service, Installation, Maintenance & Support Measurement and Verification

Peak Price – Traditional Driver for DR


5

Cost of Electricity $300

$250

$200

$150

$100

$50

$0

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Cumulative Hours of Operation

$ / M

Wh

Hourly Wholesale Cost to Utility

Evolution of DR Requirements

Former Model: Controlled Supply Variable Demand New Model: Variable Supply Variable Demand


6

Ramping – Emerging Driver for DR


7

Ca “Duck Curve” caused by variation in solar over the course of a day

US Markets


8

Recent Regulatory Developments

FERC 745 overturned DR for energy Push to extend the ban to Capacity PJM published paper pushing shift to allow DR only through load serving

entities DR saved $9.3B in capacity market1

Disposition could force states to re-enable DR as a market force in the state

NY State proposes Reforming the Energy Vision (REV)2

Enable customers to better manage their energy costs System efficiency, bills, carbon, innovations, resiliency and competitive

markets 1http://www.rtoinsider.com/no-pjm-demand-response-no-prob/ 2.http://www3.dps.ny.gov/W/PSCWeb.nsf/a8333dcc1f8dfec0852579bf005600b1/26be8a93967e604785257cc40066b91a/$FILE/REV%20factsheet%208%2020%2014%20(2).pdf


9

Technology Enablers

Advance Metering Infrastructure (AMI) Allows pay-for performance DR Enables price-responsive rates

Two-way solutions enable analytic insights

Third Party Connected Devices WiFi connected thermostats Potentially a customer-supplied DR resource Tool for behavioral-based Energy Efficiency


10

Pay for Performance DR Three Characteristics Drive the Value of a DR Asset

Predictability – must know what quantity is available at any time

Reliability – if scheduled, resource must deliver

Timeliness - rapid start, long persistence


11

IntelliSOURCE – Bring Your Own Device

Assign fair value to load drop

Dynamically dispatch assets to achieve desired outcomes

Mix and match device constraints for desired load shape

Analytics from T-Stat data

A Tale of Two Houses


13

Cool Slope 0.4⁰/hr

Heat Slope 1.5⁰/hr

House 1 - Better Insulation, Undersized AC

14

Uses for Data: Remote Energy Audit Candidates for other programs EE and DR optimization



15

House 2- Poor Insulation, Oversized AC



House 2- Poor Insulation, Oversized AC

16

Custom Tip :

When away, set temp. up 3⁰, return temp. an hour before arriving home.

IntelliSOURCE - DR Optimization Data to Inform Control Events

17

AMI Data Real Events Test Events Device Status Parcel Data Demographic Data SCADA Weather Device Telemetry Energy Price

More Accurate Predictions

Cost Optimized Dispatch (including 3rd party devices)

Precision Load Shape

Reduce Free Riders

Real Time Data Stream Analysis

Continuous Machine Learning


Utility Distribution Mgmt.

System

Head End Device–Specific APIs etc.

Settle Report Register Monitor Dispatch Demand Response Optimization

Comverge Direct Install

Comverge IntelliSOURCE®

Utility Back-Office

Distributed Energy

Resources RDP RDP RDP RDP: Retail Device

Provider (Google) RDP RDP

Demand Response Optimization

Summary

Changing supply, regulatory and technology landscape requires new requirements for demand response resources

Bring Your Own Device (BYOD) programs change cost structure of DR programs but add increased complexity for program management

Utilities must develop strategy to manage complexity while ensuring program optimization


19

John Rossi [email protected]


20

1 1

Georgia Tech Energy Speakers SeriesThe Agile Utility: Aligning Distributed Generation with Consumer Demand

November, 2014

Georgia Tech Energy Speakers SeriesThe Agile Utility: Aligning Distributed Generation with Consumer Demand

November, 2014

Prosumer-Based Decentralized ControlProsumer-Based Decentralized Control

Santiago GrijalvaGeorgia Institute of Technology

Santiago GrijalvaGeorgia Institute of Technology

2 2

Smart Grid functionality restores the balance

Hydro power plants

Nuclear Power Plants

Natural Gas Generators

Transmission System

Distribution Substations

Customers

Distributed storage

Solar Farms

Wind Farms

The Emerging GridThe Emerging Grid

Home EnergyStorage

EnergyEfficiency

PHEV

Rooftop Solar

Distributed wind

Commercial Customers

2© 2014 Georgia Institute of Technology

3 3

Unprecedented EvolutionUnprecedented Evolution

Domain Change Future System

Objectives • Reliability++, Economy++, Sustainability Sustainable

Sources • Fossil fuel to renewable• Bulk centralized to distributed• Highly Variable

RenewableDistributed, Two way

StochasticICCT • Can control entire system through SW

• Interdependency of physical and cyber• Privacy and cyber-security issues

Cyber-ControlledCyber-PhysicalSecure, Private

Actors • Consumers can also produce and store• Consumers seek their own objectives• Massive number of actors and devices

ProsumersSmart

Massive

33

© 2014 Georgia Institute of Technology

• Much more difficult to model and simulate grid complexity• Challenges in Control, Management, and Industry Architecture

4 4

ARPA-E Prosumer-Based Distributed Control Project

ARPA-E Prosumer-Based Distributed Control Project

• ARPA-E Green Energy Network Integration (GENI)• Jan. 2012 – Dec 2014 • $2.7 Million• Collaboration:

– Santiago Grijalva (power systems), – Magnus Egerstedt (networked control), – Shabbir Ahmed (stochastic optimization), – Marilyn Wolf (cyber-physical systems)– About 15 graduate students.

• Project Objective: Demonstrate a massively scalable decentralized control architecture that can support the requirements of Future Intelligent, Sustainable Electricity Grids.


5 5

Project SummaryProject Summary

• Project Elements– Reference Architecture– Theory: Decentralized Agent-Based Control and Decomposition-

based Optimization

– Technologies: • Power-Communications Co-Simulator• Electricity Operating System, • Distributed Controllers

– Large-Scale Simulation [TRL6]• IAB including MISO, PJM, NRECA, FERC, Brattle Group.• Vision inputs from about 100 stakeholders.• About 30 papers produced.


6 6

Concept 1: ProsumersConcept 1: Prosumers

• A generic model that captures basic functions (produce, consume, store) can be applied to power sub-systems at any scale.

• The fundamental task is power balancing:

• Energy services can be virtualized.

ExternalSupply Energy

Storage

LocalEnergy

Wires

Load 1 Load 2 . . . . Load n

INT G D Loss STO STOP P P P P P


7 7

Concept 2: Decentralized Electricity Industry

Concept 2: Decentralized Electricity Industry

Interconnection

ISO

Utility

Grid, Building, Home

7

• Interactions occur among entities of the same type (prosumers)• Can achieve massive decentralization


Hierarchical

Flat

8 8

Concept 3: Prosumer ServicesConcept 3: Prosumer Services


• Prosumer handles internal optimization and external coordination.

• Exposes standardized services

– Energy balancing– Frequency regulation– Reserve– Sensing and Information– Forecasting– Security– Self-identification– Voltage control– Black Start– Etc.

9 9

Concept 4: Layered Energy Cyber-Physical SystemConcept 4: Layered Energy Cyber-Physical System

10 10Device Layer

Local Control Layer

Cyber-Layer

System Control Layer

Market Layer

Flow Controller

Utility-scale PV

Hydropower

Geothermal

Gas Storage

Wind Turbine Synchronous Gen

Power Electronics

ControllersIEDs PMU

SensorsProtection Relays

Data Concentrators

SCADAWAN HPCFirewallBid Data

State Estimator

SCADAEMS AGCDSA

Security Assessment

SCOPFSCUCRegulatory Framework

Market Management System

Device Layer

Local Control Layer

Cyber-Layer


Market Layer

Utility-scale PV

Distribution Transformer

District Heating

Gas Storage

Mid-size WindUtility-scale Storage

Micro PMU

Smart MetersVar Regulatros Reclosers

SensorsProtection Relays

GIS

SCADADatabase HPCFirewallCustomer Data

Hosting Capacity

OMS RestorationCVR

Customer SystemBilling DMSDemand Response

Pricing Module

Forecasting

ISO/Transmission Distribution Utility

Feeder Reconfiguration


11 11Device Layer

Local Control Layer

Cyber-Layer


Market Layer

PV CellsStorage Devices

Building Loads

Gas Storage

Diesel GenHydrogen Fuel Cell

Micro PMU

Smart MetersActuators Sensors

ProtectionsControllers

Wireless System

SCADACampus LAN HPCFirewallDatabase

Islanding

BalancingSA Forecasting

Decentralized Controller

Virtual Power PlantPricing Module

OptimizationEMS

Device Layer

Local Control Layer

Cyber-Layer


Market Layer

Appliances

Air Conditioner

Water Heater

Pumps

Electric VehicleLighting

Flow Sensors

ActuatorsTemperature

Micro PMUOccupancy Sensors

Security

SCADADatabase PCFirewallLAN

MonitoringSystem Level ControlAlarm System

SchedulingBEMSDemand Response

Optimization Module

Forecasting

UPSRooftop PV Battery

Microgrid Building


12 12

Decentralized ControlDecentralized Control

• Self-Optimizing Regions in a Large-Scale RTO System.

• Tie-Line Bus LMP convergence using Decentralized Optimization.


13 13

Benefits of Decentralized ArchitectureBenefits of Decentralized Architecture

1. Scalable to infinite number of control points.2. Reduces need for massive communication infrastructure.3. Leverages sensing investment: smart meter, PMU, IED.4. Enables otherwise intractable optimization problems5. Supports integration of DERs6. Eliminates single point of failure7. Supports all forms of distributed intelligence8. Empowers customers9. Increases information privacy10.Enhances cyber-security11. Incrementally deployable12.Backward compatible


14 14

Grid OS

Simulation Testbed

Simulation Testbed API

Grid OS API

Decentralized ApplicationsARPA-E Energy

Scheduler

Distributed Power Control Protocols and Libraries

Application Framework

Technologies TRL [5-6]Technologies TRL [5-6]

• Decentralized Energy Scheduler– Much faster than state-of-the art for

large-scale ISO model.– Scales down to distribution/grid/home.

• Decentralized Frequency Regulation– Stabilization of large-scale ISO.– Scales down to distribution/grid/home

• Grid Operating System– Mathematically-proven protocols– Application framework– Distributed Control Library

• Co-Simulator– Power and communications

14

Communication Simulator

Power System Simulator

Middleware Models

Decentralized Frequency Regulation


15 15

March 2014, Team wins DOE ACC Business Model Competition

March 2014, Team wins DOE ACC Business Model Competition

• Regional DOE Competition focused on innovative business models for clean energy.

• Team proposed business models of an distributed control-based energy internet.

• Received first price at $100k.


16 16

July 2014: Incorporation of ProsumerGrid, Inc.

July 2014: Incorporation of ProsumerGrid, Inc.

• Decided to form start up company.• ProsumerGrid, Inc. to further develop and

commercialize ARPA-E project software that allows the effectively coordination and operation of emerging interacting energy systems.• Computational Simulation Software• Decentralized Real-time Control Systems


17 17

John Higley, M

Santiago Grijalva (Principal Investigator)• Associate Director for Electricity/Professor, Georgia Tech• Former Director of Power Systems Center at NREL

Marcelo Sandoval (Entrepreneurial lead)• Georgia Tech EE PhD Candidate, MBA• Certificates: Intl. Business, Entrepreneurship, Lean Six Sigma

John Highley (Mentor)• Owner of Energy and Environmental Enterprises• Retired Managing Partner for Deloitte’s Global Energy & Utilities

September, 2014 NSF I-CORPs ProgramSeptember, 2014 NSF I-CORPs Program

• To validate hypotheses, refine business model, and establish product market fit.

• I-CORPs Team


18 18

I-CORPs InterviewsI-CORPs Interviews

ISOs: regional Electric Utilities System Operators: Cities

Facilities Energy Managers: Buildings

7

18 10

Residential Home Owners

65

19 19

I-CORPs HighlightsI-CORPs Highlights• Utilities vary in their level of sensing

and automation, and have different regulations and renewable targets.

• Operational complexity growing fast.• Needs vary, but there is a common

theme around DER integration.• Wanted: a software system capable of

coordinating large-numbers of distributed energy subsystems.– Multi-layer, multi-scale simulation/analysis– Decentralized real-time control engines.


20 20

I-CORPs Decision: GO!I-CORPs Decision: GO!

Tested the problem

Identified customer problems and needs

Found Product Market Fit!

Tested our value propositions

Found Partners for Pilot Project


21 21

Thanks!Thanks!

• For information on decentralized control and the ARPA-E Project contact

• Santiago Grijalva: [email protected]

• For information on ProsumerGrid, Inc. contact Marcelo Sandoval: [email protected]

from adapting to transforming in a ch i tilithanging...

Documents