from adapting to transforming in a ch i tilithanging...
TRANSCRIPT
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
McKinsey & Company | 2
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
McKinsey & Company | 4
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
McKinsey & Company | 5
Cutting through the hype--utilities need to develop a perspective on these questions
MCKINSEY PROPRIETARY AND CONFIDENTIAL
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
McKinsey & Company | 6
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
MCKINSEY PROPRIETARY AND CONFIDENTIAL
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
Key themes for todayMCKINSEY PROPRIETARY AND CONFIDENTIAL
Transforming, not adapting
Disruption: glass half empty
Actions that define winners
McKinsey & Company | 8
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)
MCKINSEY PROPRIETARY AND CONFIDENTIAL
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
McKinsey & Company | 14
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
Key themes for todayMCKINSEY PROPRIETARY AND CONFIDENTIAL
Transforming, not adapting
Disruption: glass half empty
A ti d fi i iActions defining winners
McKinsey & Company | 16
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
McKinsey & Company | 17
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
MCKINSEY PROPRIETARY AND CONFIDENTIAL
“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
McKinsey & Company | 18
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
McKinsey & Company | 19
paccelerate?
Opportunities needs to be understood from both regulated and unregulated view
NOT EXHAUSTIVE
MCKINSEY PROPRIETARY AND CONFIDENTIAL
Customer apps Distributed Gen Storage PHEV/EVEnergy efficiency Data collection IT integration
Rate based solar
3rd party data monetization
Rate based storage
3rd party data monetization
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
3rd party data monetization
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
McKinsey & Company | 21
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
McKinsey & Company | 24
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
McKinsey & Company | 25
gg
Glass half full or half empty?
McKinsey & Company | 26
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
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
6
Ramping – Emerging Driver for DR
©2012 Comverge – Confidential and Proprietary
7
Ca “Duck Curve” caused by variation in solar over the course of a day
US Markets
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
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
Cool Slope 2.1⁰/hr
Heat Slope 1.7⁰/hr
15
House 2- Poor Insulation, Oversized AC
Cool Slope 2.1⁰/hr
Heat Slope 1.7⁰/hr
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
©2014 Comverge – Confidential and Proprietary
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
©2012 Comverge – Confidential and Proprietary
19
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.
© 2014 Georgia Institute of Technology
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.
© 2014 Georgia Institute of Technology
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
6© 2014 Georgia Institute of Technology
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
© 2014 Georgia Institute of Technology
Hierarchical
Flat
8 8
Concept 3: Prosumer ServicesConcept 3: Prosumer Services
8© 2014 Georgia Institute of Technology
• 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
System Control 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
© 2014 Georgia Institute of Technology
11 11Device Layer
Local Control Layer
Cyber-Layer
System Control 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
System Control 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
© 2014 Georgia Institute of Technology
12 12
Decentralized ControlDecentralized Control
• Self-Optimizing Regions in a Large-Scale RTO System.
• Tie-Line Bus LMP convergence using Decentralized Optimization.
12© 2014 Georgia Institute of Technology
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
13© 2014 Georgia Institute of Technology
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
© 2014 Georgia Institute of Technology
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.
© 2014 Georgia Institute of Technology
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
© 2014 Georgia Institute of Technology
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
© 2014 Georgia Institute of Technology
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.
© 2014 Georgia Institute of Technology
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
© 2014 Georgia Institute of Technology
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]