Energy storage at various scales can be the key to integrating large amounts of renewable generating resources into the electric power system. Growth of storage is advanced by a combination of policies and economics. Presentation for the Portuguese National Committee of CIGRE, 2017.

1. The Role of Energy Storage in the
Future Electricity System
Lorenzo Kristov, Ph.D.
Principal, Market & Infrastructure Policy
Presentation for Portuguese National Committee of CIGRE
March 17, 2017

2. Ideas in this presenta-on are offered for discussion purposes only,
and do not reflect the views or policies of the California ISO.
2

3. Topics of this presentation
Ø Some facts about California ISO
Ø California’s energy policies and policy objectives
Ø Challenges to achieving the policy objectives
Ø A diverse toolkit for addressing the challenges
Ø Storage as the potential game changer
Ø Challenges to bringing storage on at scale
Ø How California is addressing these challenges
Ø Big questions and works in progress
Ø A vision of a possible future
Page 3

4. California Independent System Operator
§ 71,800 MW of power plant capacity
(installed capacity)
§ 50,270 MW record peak demand
(July 24, 2006)
§ 27,488 market transactions
per day (2015)
§ 25,685 circuit-miles of transmission
lines
§ 30 million people served
§ 240 million megawatt-hours of
electricity delivered annually (2015)
§ Not-for-profit public benefit
corporation
Page 4

5. Western Electricity Coordinating
Council (WECC):
• 38 of 76 balancing authorities are
in WECC
• Serves a population of
approximately 82.2 million
• Spans more than 1.8 million
square miles in all or part of 14
states
• In 2014, the balancing authorities
reported a total nameplate
capacity of 275,400 MW
WECC
BC
WA
OR
MT (p)
ID SD (p)
NV
UT
AZ NM(p)
CO (p)
NE (p)
TX (p)
WY
CFE
CA
AB
Two-thirds of the United States is served by
independent system operators (ISO/RTOs)
Quebec
Interconnection
Interconnection
Eastern
Interconnection
Page 5

6. CAISO resource mix
Page 6

7. California ISO is fully committed to achieving the
state’s energy and environmental policy objectives.
California’s policy objectives:
– Broad de-carbonization of the California economy
– Low-carbon, reliable electricity system
– Growth of distributed generation
– Customer choice for adoption of distributed resources
and devices of all types
– Environmental justice
– Greater resilience through community resources and
micro-grids
Page 7

8. Energy and environmental policy targets
2020 Policy Goals
• Greenhouse gas reductions to 1990 levels
• 33% of load served by renewable generation
• 12,000 MW of distributed generation
• End use of once-through cooling in coastal power plants
• 1,325 MW of energy storage
2025 Policy Goal
• 1.5 million electric vehicles
2030 Policy Goals
• 50% of load served by renewable generation
• Double energy efficiency of existing buildings
• Greenhouse gas reductions to 40% below 1990 levels
New legislation introduced
• 100% renewable generation by 2045
Page 8

9. 5,000 MW of additional transmission-connected
renewables by 2020 (mostly Solar PV)
(IOU data through 2017 and RPS Calculator data 2018 – 2020)
Page 9

10. Distributed Solar PV in California
2015 2016 2017 2018 2019 2020 2021
BTM Solar PV 3,695 4,903 5,976 7,054 8,146 9,309 10,385
0
2,000
4,000
6,000
8,000
10,000
12,000
MW
Estimated Behind the Meter Solar PV Build-outthrough 2021
Page 10

11. Achieving the objectives entails some challenges.
– Severe system load shape effects – the “duck” curve
– Real-time variability of renewable energy production
– Zero marginal cost energy depresses prices in the
wholesale spot market
– Distributed energy resources (DER) in the wholesale
market
– Hard-to-forecast DER adoption and behavior
– Regulatory frameworks, business models, industry
culture are all based on the traditional paradigm:
centralized generation and one-way power flows
Page 11

12. Recent events surpass previous expectations of net load
and afternoon ramp with higher solar generation.
Typical Spring Day
Net Load 11,663
MW on May 15,
2016
Slide 12
Actual 3-hour ramp
12,960 MW on Dec.
18, 2016

13. Solar production varies from one day to the next – first
week of March 2014
-500
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
6 7 8 9 10 11 12 13 14 15 16 17 18
Day_1 Day_2 Day_3 Day_4 Day_5 Day_6 Day_7 Average
Average
MW
Page 13

14. Wind production varies from one day to the next – first
week of March 2014
0
500
1,000
1,500
2,000
2,500
3,000
3,500
0 1 2 3 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Day_1 Day_2 Day_3 Day_4 Day_5 Day_6 Day_7 Average
Average
MW
Page 14

15. These challenges are all manageable with a toolkit
of complementary strategies.
– Integrate renewables with controllable/dispatchable renewables
– Wholesale market models to enable DER and storage participation
– New coordination framework between ISO and distribution utilities
for T-D interface operations and integrated planning
– Use DER to flatten load profiles and mitigate variability locally
– Price signals to incentivize work-place EV charging and other uses
of plentiful daytime solar energy
– Aggregate DER for dynamic demand management
– Regional coordination to optimize diversity benefits
– Micro-grids for local resilience
– Whole-system grid architecture for a more decentralized system
Page 15

16. Storage could be the game changer.
• Energy storage of various types and scales could meet
the challenges of high-volume solar PV
– At grid-scale: charge at low/negative over-gen prices &
make good use of mid-day excess supply
– Co-locate with utility-scale PV and wind to smooth output
to the grid
– Discharge to mitigate steep ramp & late afternoon peak
– Provide 2-way demand response, frequency response
and regulation
– With rooftop PV: minimize back-flow on distribution, to
increase circuit “hosting capacity”
– Manage local variability locally & flatten load profiles

17. Bringing storage on at scale has some challenges, and
requires paradigm shift to focus on energy uses.
• Value storage services, not just kWh or kW
– Thermal storage in buildings
– De-carbonize transportation
• Key values storage could offer are not yet valued financially
or defined as revenue-producing services
– Resilience benefits
– Local mitigation of volatility and extreme load profiles
• Storage can decentralize reliability
– Improve physical and cyber security
– Support self-optimizing local systems, micro-grids
• Change in the electric system is being driven as much by
bottom-up adoption as top-down policy
– Regulators have less control, must become facilitators,
enablers of change

18. CAISO’s wholesale market model for storage.
v “Non-Generator Resource” (NGR)
– Designed for a resource that can move seamlessly between
consuming and injecting energy at different times
– Market optimization manages storage characteristics including
state of charge (SOC), maximum charge amount (limit on total
energy it can provide), maximum charge and discharge rates
– Resource can choose to have ISO manage SOC, or manage
itself and bid SOC
– Can be used by transmission-connected and distribution-
connected resources
– Can interconnect to ISO grid under our normal interconnection
procedures – as a generator with negative output
– Can provide regulation and other ancillary services
– Visible to ISO and settled 24x7 (comparable to a generator)
– Does not use baselines to measure performance
Page 18

19. Examples of NGRs operating today
v Utility-scale batteries
– 2014 market operational – Vaca Dixon 2 MW / 14 MWh NAS battery,
100% dedicated to ISO wholesale market participation
– 2016 market operational – Yerba Buena 4 MW / 28 MWh NAS battery,
customer R&D facility, San Jose.
Half energy reserved for islanding/backup for adjacent customer
facility
– 2016 market operational – SCE Tehachapi wind farm, 8 MW/32MWh
v Aggregation of electric vehicles: Los Angeles Air Force Base
– Bi-directional power flow (V2G)
– 500-600 kW capacity
– V2G capable sedans, trucks, vans
– Began ISO market participation December 2015
– Partnership effort between DOD, SCE, CPUC, CEC, ISO, Lawrence
Berkeley National Labs, Kisensum, and others.
Page 19

20. Storage behind the meter can participate as demand
response.
v Proxy Demand Resource (PDR)
– Accommodates traditional load curtailment and behind-the-
meter (BTM) devices such as energy storage
– May be metered at load meter as traditional DR or via sub-
metering of BTM devices to measure performance
– In all cases DR performance is calculated against a baseline
– PDR can be an aggregation of service accounts in a sub-LAP
– DR cannot inject energy into the grid (only load modification)
– Currently can only be dispatched to reduce demand
– Not a 24x7 resource; only available when bid per program or
resource availability parameters
v Future PDR enhancements:
– Dispatch to increase consumption; provision of regulation
– New performance evaluation baselines to fit diverse end-uses
Page 20

21. Examples of PDR
v AMS model: Battery behind commercial building meter manages
demand charges
– Aggregation of multiple sites forms PDR in ISO market
– DR response is measured at the batteries, not the premises
– DR response is measured against “normal” battery operation
based on recent similar non-dispatch hours
– Energy consumed to charge battery is simply retail load
v Future PDR enhancement example
– PDR bids to “consume energy” at negative wholesale prices
• If the negative bid is economic, the ISO will dispatch and pay the
PDR for energy consumed at the LMP
• ISO payment partially offsets retail rate for load
– PDR can bid to provide regulation down by increasing
consumption
Page 21

22. DERP is a market participant that aggregates DER.
v Distributed Energy Resource Provider (DERP)
– Executes DERP agreement with ISO, may create multiple DER
aggregations (DERAs), each containing different sub-resources
– A DERA may mix DER types, as long as resource performance
characteristics can be modeled in ISO market system
– DERA and sub-resources are all directly metered: no baselines
– Sub-resources of a DERA must all be within a single sub-LAP
– Minimum DERA size is 0.5 MW; maximum size 20 MW if aggregating
across multiple p-nodes (LMP pricing nodes, T-D substations)
– Individual sub-resources must be less than 1.0 MW
– A DERA across multiple ISO p-nodes must specify distribution factors
and must follow same distribution factors in response to ISO dispatch
– Resource will utilize NGR model (allows injections, settled 24x7, does
not use baselines)
– Resource cannot contain sub-resources participating in NEM
– Resource cannot contain DR sub-resources using a baseline
Page 22

23. Storage resources seek to provide multiple services and
earn revenues at multiple levels of the system.
v Behind the end-use customer meter
– Time of day load shifting, demand charge management,
storage of excess solar generation
– Service resilience – smart buildings, microgrids, critical loads
v Distribution system services
– Deferral of new infrastructure
– Operational services – voltage, power quality
v Transmission system & wholesale market
– ISO spot markets for energy, reserves, regulation
– Non-wires alternatives to transmission upgrades
v Bilateral contracts with load-serving entities
– Resource adequacy capacity
– Renewable energy
Page 23

24. High DER penetration requires enhanced operations
coordination at T-D interfaces.
v Diverse end-use devices and resources with diverse owners/
operators will affect:
o Net end-use load shapes, peak demands, total energy
o Direction of energy flows, voltage variability, phase balance
o Variability and predictability of net loads and grid conditions
v ISO market systems see DER at T-D substations, have no visibility
to distribution grid conditions or impacts; distribution utility is not
aware of DER bids and dispatches
v DER providing services to customers and distribution system will
affect the T-D interfaces; need accurate real-time forecasting and
local management of DER variability
v CAISO has been working with distribution utilities to develop an
operations coordination framework for high penetration of DER
Page 24

25. Big questions and works in progress
v Future revenue opportunities for storage
– Synthetic inertia, primary frequency response (governor)
– Resilience – smart building, community micro-grid islanding
– Storage as transmission or distribution asset
v Multiple-use applications (MUA)
– Resource can provide services to and earn revenue from more than
one entity (customer, distribution operator, transmission/wholesale)
– CAISO and CPUC are developing MUA rules for storage (e.g.,
dispatch priority; double payment; wholesale/retail)
v DSOs, local electricity markets and micro-grid systems
– New distribution system operator (DSO) models may expand
opportunities for distribution-level storage
– We need an “open access” regulatory framework for DSOs analogous
to wholesale market framework for transmission
– We need a layered reliability framework to allow hand-off of reserve
and adequacy requirements from ISO to DSO at the T-D interface
– Revisit some federal-state regulatory structures (US)
Page 25

26. DSO operates Local Distribution
Area or Community Micro-grid
Regional Interconnection
A vision of a possible future: the electricity grid as a
layered hierarchy of optimizing sub-systems.
Page 26
Smart
building
Micro-
grid
Micro-grid
D
S
O
D
S
O
ISO
Balancing
Authority
Area
BAABAA
Smart
building
Smart
building
• Storage on premises can
support self-optimizing local
electricity systems
• Each tier only needs to see
interchange with next tier
above and below, not the
details inside other tiers
• CAISO focuses on regional
bulk system optimization
while DSO coordinates
DERs
• Layered control structure
reduces complexity, allows
scalability, and increases
resilience & security

27. Sources and resources
Using renewables to integrate renewables
• http://www.caiso.com/Documents/UsingRenewablesToOperateLow-
CarbonGrid.pdf
Future Distribution Systems, Platforms and DSOs
• http://doe-dspx.org
• https://emp.lbl.gov/sites/all/files/FEUR_2%20distribution
%20systems%2020151023.pdf
Grid Architecture
• http://gridarchitecture.pnnl.gov
Decentralization and Layered Optimization
• http://resnick.caltech.edu/docs/Two_Visions.pdf
• https://www.academia.edu/12419512/
A_Future_History_of_Tomorrows_Energy_Network
Page 27

28. Thank you.
Lorenzo Kristov
LKristov@caiso.com
Market & Infrastructure Policy
Page 28