Presenter: Janice Lin, StrateGen Consulting

1. Integra7ng Distributed Solar and Storage
A Compelling Value Proposi>on to Realizing a Smarter Grid
Janice Lin | Managing Partner of StrateGen Consul7ng, Director CA Energy Storage Alliance
September 29, 2009
San Diego Solar Energy Conference

2. Agenda
1. StrateGen Consul>ng and CESA Overview
2. Energy Storage and California’s Smart Grid
3. Ra>onale for Early Focus on Distributed Applica>ons
4. The Value Proposi>on for Integra>ng Distributed PV and Storage
5. Energy Storage Market Barriers and Policy Recommenda>ons
6. Summary
1

3. StrateGen Overview
StrateGen helps businesses create sustainable value
through clean energy solu7ons
Smart Grids
Enernex
Structured Finance Energy Controls
HMH Energy Resources StrateGen Core Team Ron Hofmann
» Deep industry knowledge in clean
Regulatory energy; core focus on solar & storage
» Strong analy>cal and financial Solar Advisory
Barkovich & Yap, Inc
Andy Skumanich
Douglass & Liddell capabili>es Lori Mitchell
Roger Levy
» Strategic management exper>se
» Product development and project
Energy construc>on experience Market Research
Efficiency » Project leadership and management American LIVES, Inc.
ConSol New Heights Research
Corporate
Sustainability
Tyler Elm
2

4. StrateGen Overview: Our Clients
Our clients represent a wide range of organiza7ons, including those
central to the clean energy market, and those trying to enter
U.S. Department
of Energy
3

5. The California Energy Storage Alliance (CESA)
Our Goal: Expand the role of storage technology to promote the growth of
renewable energy and create a more stable, secure electric system
» CA‐focused advocacy group represen>ng
energy storage stakeholders
» Focus on storage coupled with renewable
energy integrated into the smart grid
» Current priori>es/ac>vi>es include:
– CPUC
• SGIP AES implementa>on
• DG (DER) cost benefit methodology
• Smart Grid OIR
– Storage legisla>on: SB 412, SB 14, AB 44
– CEC Integrated Energy Policy Report planning
– CAISO Implementa>on of FERC order 719 and
890
– Vision for Storage in CA!
4

6. Energy Storage is a key enabling technology
Smart Grid Renewables Integra7on
“Our expecta7on is that
this [smart grid] network
will be 100 or 1,000 7mes
larger than the Internet”
‐ Cisco, May 2009
Peak Load Growth Transmission Constraints
5

7. Storage is a necessary component of the smart grid
Transmission Distribution Load Serving
Operator Operator Entity Industrial
Customer
Distributed
Resources Commercial
Customer
Energy Substation
Storage
Microgrid / Residential
Customer Multi-
Other sustainable communities Dwelling Unit
Substations
6
Diagram courtesy of PG&E

8. Storage is a necessary component of the smart grid
Commercial & Industrial
Transmission Distribution Load Serving
Operator Operator Entity Industrial
Customer
Distributed
Resources Commercial
Customer
Energy Substation
Storage
Substation
Microgrid / Residential
Customer Multi-
Other sustainable communities Dwelling Unit
Transmission Scale Substations
Microgrid Residential
7
Diagram courtesy of PG&E

9. Energy Storage Reduces GHG Emissions
Percent CO2 / MWh
Reduc7on Shi_ing Peak vs. Off-peak CO2 Emission Rate* (Tons/MWh)
from Peak to Off‐Peak:
Afternoon - Heavy A/C Use
Peaker Plants
SCE: 33% reduc7on
PG&E: 26% reduc7on
SDG&E: 32% reduc7on
Other times - Very Little A/C Use
Also ~56% lower NOx Better Use of Utility Assets
emissions
E3 Calculator Tons CO2 / MWh
Summer Summer Summer
On-Peak Mid-Peak Off-Peak *Southern California Edison Data
Utility
PG&E 0.6709 0.6068 0.4949
SCE 8 0.7247 0.6322 0.4862
SDG&E 0.6872 0.5807 0.4705

10. Energy Storage Framework for California
Phase I Phase II Phase III
Start with smaller customer sited Couple with distributed Demonstrate large scale
storage, linked with AMI wholesale (higher value FiT or transmission level
U>lity owned storage PPA) storage (50‐100MW+/
demonstra>ons – capital deferral
Demonstrate storage as key site)
20 MW ancillary services storage enabler of sustainable Accelerate distributed
demonstra>ons with CA ISO microgrids storage deployment for
mul>ple applica>ons
2009‐2011 2012‐2015 2015‐2020
Goal: 25-100 MW deployed 100-500 MW deployed 1,500 MW+ deployed
9

11. Ra7onale for Early Focus on Distributed Applica7ons
Distributed applica7ons offer the poten7al to capture and bundle the
greatest number of value streams
Customer + Utility + System Operator + Society
• Reduced energy • Integ of renewables • Ancillary services • More
and demand costs • Load leveling • Grid integra>on renewables
• Emerg. back up • T&D relief / deferral • Improved grid • Fewer emissions
• Demand response • Improved power reliability & • Lower power
• Improved quality security costs
reliability • Reduced peak & • More jobs
spinning reserve
requirements
10

12. Conceptual Overview of Storage + PV (1+1=3)
Distributed Solar + Storage Poten7al Value Streams
Base load
» Charge during off‐peak and discharge during
peak to reduce demand charges
» Firm up addi>onal demand savings from
Load & Solar
Genera>on renewables
» Can share inverter / power condi>oning
equipment with solar or other renewables
» Poten>al to act as dispatchable power for
Net Load is S>ll u>lity demand‐response
Coincident with
Peak Demand » Can shik peak demand
Charges
» Poten>al to leverage SGIP and FITC for both
technologies
» Poten>al to provide emergency back up
Storage to Shik
capabili>es
Net Peak Load to
Off Peak Periods
11
11

13. Distributed Applica7ons Are U7lity‐Scale
Small distributed systems can have a grid‐scale impact
Substation Sited Storage Demand Curve after
Typical Summer Implementation of
Daily Demand for 3,000 MW solar
CA-ISO Region
44
42 Peak - Shaving Impact
of 13 GWh storage
40 Equivalent to 5 kWh
Storage for each kW
CA Generation (GW)
Customer Sited 38 Installed Solar
36
34
32
30
28
26
24
0 6 12 18 24
Hour of Day
Source: EPRI
12

14. Value Proposi7on: Our Approach
We have extensively analyzed the drivers of value for storage coupled
with PV. We’ll present results of the base case and sensi7vi7es.
» Select a specific applica>on/customer site
» Model the value proposi>on of standalone PV system for the site
» Design a hypothe>cal energy storage system to be paired with the
specific site and PV system
» Model the value proposi>on of integrated PV and storage
» Run key sensi>vi>es to incen>ves and other project specifica>ons
The base case for PV Only and PV + Storage are presented on the
following pages
13

15. Value Proposi7on: PV Only Base Case
PV Only Base Case: A 350 kW PV system has a 5.9% IRR from avoided
energy costs
Key Assump7ons Load Shape Impacts
» Use: customer‐sited 350kW rookop PV
system
» All‐equity turnkey purchase
» CAPEX: $5,000/kW DC
» O&M: $15/kW/yr
» Incen>ves:
 30% FITC
 $0.22 CSI PBI
 5 Year MACRS Deprecia>on
Results Load & PV Output
» Aker Tax IRR: 5.9% » High school load profile in California
» Simple Payback: 12 years » PVWaps 20deg fixed >lt in So. Cal. (1)
1) 0.5% annual degrada>on factor & 0.77 derate factor
14

16. Value Proposi7on: PV + Storage Base Case
PV + Storage Base Case: A 100kW, 4h baner + 350 kW PV system has a
7.6% IRR from peak load shi_ing and PV avoided energy costs
Key Assump7ons(1) Load Shape Impacts
» Use: customer‐sited storage system coupled
with a 350kW PV system
» All‐equity turnkey purchase
» Bapery Size: 100kW, 4h dura>on
» Bapery Round Trip AC Efficiency: 65%
» CAPEX: $475kWh ($2,850/kW)
» Variable O&M: $0.075/kWh discharged (2)
» Fixed O&M: $6/kW/yr
» Incen7ves: None
Results Storage Op7mal Sizing
» Aker Tax IRR: 7.6% » Analyze net peak load with PV system
» Simple Payback: 11 years » SCE TOU8 Peak Period 12PM – 6PM
» Use storage to shave akernoon peak
» 100kW, 4h bapery
(1) SCE TOU8 tariff, CA high school load profile
(2) This cost accounts for replacement costs of battery cells, parasitic loads, and consumables
15

17. CA’s Energy Storage Regulatory Market
California’s regulatory framework is rapidly evolving to accelerate
deployment of storage, especially storage coupled with solar
» Cri>cal Peak Pricing (CPP)
» Self Genera>on Incen>ve Program (SGIP)
» Permanent Load Shiking (PLS)
» Federal Investment Tax Credit (FITC)
» CEC Integrated Energy Policy Report 2009 update will include storage
» Poten>al medium term opportuni>es
• Emergence of Feed in Tariff with differen>al rates for renewables coupled with storage
• Standard offer for PLS
The sensi>vity to the first four factors is explored on the following pages
Source: StrateGen’s current regulatory knowledge
16

18. CPP: Cri7cal Peak Pricing Sensi7vity
The SCE TOU8 tariff is scheduled to convert to default cri7cal peak
pricing (CPP) in October 2009
Non‐CPP vs. CPP Impacts to IRR1 CPP Overview
» CPP will be the default tariff for SCE
customers with peak demand >500kW in
October 2009
» SCE will role out CPP to smaller customer
classes within 1 to 2 years
» Average of 9 events/yr (12/yr max) during the
summer months (with 24h advanced
no>fica>on)
» CPP Structure:
 On‐peak energy charge : $1.37/kWh
 Peak demand charge reduced to $7.26
($19.73 w/o CPP)
Modeled Scenario: Es>mated October
» Assume storage device can react to and
2009 SCE TOU 8 CPP Tariff
prepare for 24h no>fica>on of CPP event
1. See Appendix for detailed modeling assump>ons
17

19. SGIP: Evolu7on of the SGIP
D. 01‐03‐073 AB 2778 Nov 2008
SGIP established to implement AB 970. Creates financial Only wind and fuel CPUC concludes that AES systems
incen>ves for distributed genera>on technologies that provide cell DG cannot be added to the SGIP as
“energy conserva>on demand‐side management and other technologies stand‐alone technology, but do
ini>a>ves in order to reduce demand for electricity and reduce qualify for SGIP qualify when coupled with
load during peak demand periods.” eligible wind or fuel cells
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010+
D.02‐04‐004 AB 2778 SB 412 Q1‐Q2
CPUC Opinion reveals that Removes all PV Restores CPUC An>cipated
“Legislature expressed no incen>ves from SGIP authority to add new implementa>on of
guidance on the extent or
technologies to SGIP storage incen>ves
scope of incen>ves for (awai>ng Gov. for solar
distributed genera>on.”
signature) applica>ons
SGIP admin implement
2008 decision: storage
applica>ons accepeted
May 8, 2009
NOTE: Details about current SGIP structure in appendix
18

20. SGIP Energy Storage Incen7ve Structure
Incen7ve Renewable Non‐renewable Energy
System Size Structure Fuel Cell Fuel Cell Wind Storage
0‐1 MW 100% $4.50 $2.50 $1.50 $2.00
1‐2 MW 50% $2.25 $1.25 $0.75 $1.00
2‐3 MW 25% $1.125 $0.625 $0.375 $0.50
Minimum technical opera7ng parameters:
» Ability to be used daily in concert with an on‐site genera>on resource, and s>ll meet its 20‐
year life>me requirement.
» Ability to handle hundreds of par>al discharge cycles each day.
» Ability to be discharged for at least four hours of its rated capacity to fully capture peak
load reduc>ons in most u>lity service territories.
» Ability to meet Ins>tute of Electrical and Electronics Engineers, Inc. interconnec>on
standards.
» Must comply with all local environmental and air quality requirements.
19

21. SGIP: SGIP Incen7ve Sensi7vity
SGIP incen7ves for solar + storage projects will be very anrac7ve,
even more anrac7ve than standalone storage
SGIP Impacts to IRR1 SGIP Current Status
» SB 412 (Kehoe, Blakeslee, Skinner)
restores CPUC authority to add
technologies into the SGIP
» Storage will be explicitly added, for
standalone applica>ons as well as
coupled with PV
» SB 412 awaits the Governor’s signature,
will be effec>ve January 2010, subject to
CPUC and SGIP working group
Modeled Scenario: SGIP @ $2,000/kW for implementa>on >ming
Storage
1. See Appendix for detailed modeling assump>ons
20

22. PLS: Program Overview
» Permanent Load Shiking (PLS) is shiking energy from one >me period to
another on a permanent basis. It provides several system benefits:
 Reduce need for capacity investments
 Reduce likelihood of shortages (was done in response to 2006 heat wave)
 Lowered system costs (fewer peaking plants)
» Main applica>on has been to offset peak demand from A/C
» The CPUC created a program to encourage PLS because there was no
incen>ve to do this (other than TOU rates) and PLS was caught in between
DR and EE programs, thus gexng few incen>ves
» In 2007 PG&E and SCE implemented “Shik and Save” programs to
implement this, with budgets of $10M each
 These programs only allow for Thermal Energy Storage systems
 PG&E uses ICE Energy exclusively
21

23. PLS: PLS Incen7ve Sensi7vity
PLS doubles storage’s IRR, but the California programs are closed to
future RFPs and only thermal storage is currently eligible
PLS Impacts to IRR1 PLS Current Status
» The CPUC issued a Final Decision for the 2009‐2011
Demand Response Program that ended further
RFPs for PLS
» The Decision did, however, order the u>li>es and
Energy Division Staff to provide construc>ve input
to future PLS programs, including a possible
‘standard offer’ (e.g.. Fixed $/kW shiked)
» Benefit for PV – this incen>ve could be combined
with the CSI and possibly SGIP incen>ves as well
» SCE’s Nightshik Program is currently only available
for thermal storage and is closed to new
Modeled Scenario: SCE’s Nightshik technologies:
Program @ $1,110/kW2  $1,110/kW
 Based on a nominal ra>ng, not peak ra>ng similar to
a PTC ra>ng for PV2
1. See Appendix for detailed modeling assump>ons
2. 100% nominal ra>ng used for the purposes of this analysis
22

24. FITC: FITC Incen7ve Sensi7vity
Storage may be eligible for the FITC when coupled with PV. If so, it will
raise the IRR’s approximately 50% higher than they are today
FITC Impacts to IRR1 FITC Current Status
» FITC is 30%
» Chadbourne & Parke has indicated that storage is
currently eligible for the FITC when paired with
renewables, but only if charged by renewables and
not grid power
» This par>cular applica>on works well in
conjunc>on with CPP tariffs because the bapery
can be charged on the weekend off‐peak with PV
to prepare for CPP events (out of scope for this
project)
» In the future, the FITC may be applicable to storage
Modeled Scenario: Unrestricted Grid without these current restric>ons
Charging @ 30% FITC » StrateGen’s model assumes storage can charge
unrestricted from grid and not from PV
1. See Appendix for detailed modeling assump>ons
23

25. Incen7ves: Combining CPP, SGIP, & FITC
Combining all incen7ve value streams has favorable financial
outcomes, making Storage + PV a very anrac7ve value proposi7on
Combined Impacts to IRR1
Note: The ability to combine and capture all of these value streams is
s>ll specula>ve pending clarifica>on of programs’ rules
1. See Appendix for detailed modeling assump>ons
24

26. Market Barriers to Full Deployment of Energy Storage
» Cost / Economics
– Many storage technologies have not demonstrated scale economies
– Analyzing impact of storage, especially coupled with renewables, is very challenging
» Technology
– Many solu>ons, all with tradeoffs
– First demonstra>ons of new applica>ons are difficult to implement
» Regulatory / Policy
– Difficult to aggregate complete value streams provided by storage
– Percep>on that u>lity‐scale storage = large equipment only
– Tariff design does not reflect true cost of producing and delivering power on peak
– Incomplete CA ISO implementa>on of FERC Order 890/719 energy storage tariff for
regula>on
– Unclear net metering rules for storage + renewables projects
– Storage is new enough that it is not yet considered in all relevant aspects of
regula>on and policy making
25

27. Energy storage is deserving of its own asset class
category and immediate energy policy focus
Energy Storage
Goal:
Supply System Demand
Optimization
Transmission/
Distribution
26

28. Summary
 Storage is a key enabler of the smart grid
 Integra>ng storage with PV can boost the value proposi>on of PV
significantly
 Many new storage technologies are commercially available today
 Significant incen>ve programs for storage + PV will be in place soon
 With these incen>ves, the value proposi>on will be very aprac>ve
 Now is the >me to start developing projects
 CESA needs the PV community’s help to design effec>ve policy for
integrated PV + storage projects
 Storage is cri>cal to many key energy policy goals – it is deserving of its
own asset class in CA
27

29. For Further Informa7on, Contact:
28

30. CESA Policy Recommenda7ons (1 of 2)
Incen7ves
 “Fully implement” SGIP for storage (need incen>ves for standalone
and solar applica>ons)
 Establish increased rate of return for u>lity‐owned storage, similar to
renewables treatment
 Long term discounted financing for storage
 Tax incen>ves comparable to solar and wind (ITC and MACRS)
RD&D Funding
 Accelerate deployment of “integrated” demonstra>on projects under
various business models
 Leverage PIER matching funds for federal ARRA storage‐related
proposals in California
 Create California‐based Energy Storage Center of Excellence to
provide technical and policy leadership
29

31. CESA Policy Recommenda7ons (2 of 2)
Goal: Leverage storage under mul7ple ownership models to help
enable the Smart Grid, GHG reduc7on, EE, DR and the RPS
 Include storage in DG, DR, EE cost benefit methodologies
 Increase Feed‐in‐Tariff cap and price for renewables firmed with
storage
 Require storage (customer, 3rd party and/or u>lity owned) as part of
long term procurement process, including pursuing Standard Offers for
Distributed Energy Storage and pursuing storage eligibility for next
Permanent Load Shiking RFP
 Explore retail tariff design that encourages load shiking
 Implement energy storage tariff for regula>on (FERC Orders 890 and
719)
 Consider a peak reduc>on standard for state agency power purchases
 Clarify net metering rules for renewable energy projects with storage
30

32. Benefits of Storage – Renewable Integra7on
“Enabling technologies such as fuel switching in ‘smart’ appliances, dispatch‐able
load from plug‐in hybrid or other electric vehicles, or sta7onary energy storage
would be required to enable very high levels of PV contribu7on (>20%) to the
electric power system”.
‐ NREL Denholm & Margolis, April 2006
“When PV penetra7on reaches sufficiently high levels (e.g., 5 to 20% of total
genera7on), the interminent nature of PV can begin to have no7ceable, nega7ve
effects on the en7re grid” [requiring storage]
‐ US DOE, SEGIS‐ES, July 2008
“Storage will need to be part of our porwolio if going to 15 to 20 percent wind at a
na7onal level, otherwise it won’t be efficient at a lower level and it won’t get us
where we want to go environmentally”
‐ Electric Power Research InsNtute, March 2009
31

33. CA’s RPS implementa7on will increase the need for
regula7on and ramping
• Increased wind
penetration creates
need for greater
regulation capacity
and faster regulation
ramping capability
• Nov ‘07 CAISO report
identifies significant
additional regulation
requirements with
20% renewables
(about 10% wind
penetration)
Ancillary services can be provided today at 20 MW scale, and from systems as
small as 1 MW on the customer side of the meter
32

34. Value Proposi7on – Framework
We will now examine the other storage‐specific drivers to customer‐
sited energy storage systems paired with PV
1. StrateGen’s Approach to Energy Storage Modeling
2. Primary Drivers of Value
3. Base Case Results
4. Key Sensi>vi>es
5. Value Proposi>on Summary Findings
33

35. Modeling the value proposi7on of storage is
challenging. StrateGen’s approach:
34

36. Value Proposi7on – Key Drivers
StrateGen has extensively analyzed the drivers of value for storage.
We’ll present results of the base case and sensi7vi7es.
Customer Specific Technology Specific
Baseline Load CAPEX
» Requires peaky loads coincident with » Fully loaded cost of storage system including
u>lity peaks to maximize storage value baperies, controls, inverters, BOS, building,
etc.
Tariff
» Requires high peak demand charges or OPEX
>me‐of‐use tariffs with large spreads » Fixed and Variable O&M
between off‐peak and on‐peak prices » Consider running the bapery every weekday
» Cri>cal Peak Pricing is an improvement on peak
for storage » Includes replacement costs and consumables
Round Trip Efficiency
» AC to AC efficiency losses of charging and
discharging
35

37. Value Proposi7on – Base Case
Base Case: A 100kW, 4h banery has a 7.6% IRR from peak load shi_ing
when paired with a 350 kW PV system
Key Assump7ons(1) Load Shape Impacts
» Use: customer‐sited storage system coupled
with a 350kW PV system
» All‐equity turnkey purchase
» Bapery Size: 100kW, 4h dura>on
» Bapery Round Trip AC Efficiency: 65%
» CAPEX: $475kWh ($2,850/kW)
» Variable O&M: $0.075/kWh discharged (2)
» Fixed O&M: $6/kW/yr
» Incen>ves: None
Results Storage Op7mal Sizing
» Aker Tax IRR: 7.6% » Analyze net peak load with PV system
» Simple Payback: 11 years » SCE TOU8 Peak Period 12PM – 6PM
» Use storage to shave akernoon peak
» 100kW, 4h bapery
(1) SCE TOU8 tariff, CA high school load profile
(2) This cost accounts for replacement costs of battery cells, parasitic loads, and consumables
36

38. Customer Specific Sensi7vi7es
Three hypothe7cal load shapes and tariffs were selected to determine
sensi7vity of customer specific factors
Customer Load Applicable Tariff
» SCE TOU8 (Secondary):
August Daily Load Profile – $15.41/kW Summer Peak Demand Charge
– $0.1077/kWh Summer Peak Energy Charge
– 4 Month Summer
– 12 to 6 PM Summer Peak
» SCE TOU8 – CPP (Secondary):
– Same structure as TOU8, but with 12 Cri>cal
Peak Days during summer peak hours at
$1.36229/kWh
– Peak Demand Charge credit of $12.47/kW
to offset CPP events
» PG&E E‐19 (Secondary):
– $13.51/kW Summer Peak Demand Charge
– $0.1555/kWh Summer Peak Energy Charge
– 6 Month Summer
– 12 to 6 PM Summer Peak
37

39. Value Proposi7on – Customer Specific Drivers
A project’s load shape and tariff will have a significant impact on end‐
customer returns
Customer Load Applicable Tariff
38

40. Value Proposi7on – Storage Technology Drivers
CAPEX has a significant impact on end‐customer returns.
Round Trip Efficiency is typically less cri7cal
CAPEX AC Round Trip Efficiency
39

41. Value Proposi7on – Storage Technology Drivers
A project’s OPEX will have an impact on end‐customer returns and is
the least transparent variable for distributed applica7ons
OPEX Methodology
» OPEX has two components:
 Fixed costs that are incurred whether the
bapery is cycled or not
 Variable costs from cycling the bapery
» This graph focuses only on the variable costs
associated with kWh hours discharged
» The variable costs include consumables such
as water, cell stack replacement, parasi>c
loads, etc.
 Variable costs do not include energy
consumed or lost during charge/discharge
(accounted for in avoided energy costs)
40

42. Summary Value Proposi7on Analysis Findings
Retail applica7ons can provide anrac7ve value proposi7ons
Economic Drivers Realis7c Scenario
» Top economic drivers » Key Assump7ons
– CAPEX – 85% RTE
– Load shape – $475/kWh CAPEX
– Tariff – $2/W SGIP
– Incen>ves – $0.0500/kWh discharged OPEX
– SCE TOU8 CPP
» Addi7onal economic drivers (not – High school load profile
quan7fied) – Includes PV integra>on
– Ability to provide emergency backup/
UPS capability » Results of Storage + PV System
– Ability to comply with u>lity demand – 11.9% IRR
response ini>a>ves – 6 Yr Simple Payback
– Ability to par>cipate in more tariff
opportuni>es
41