Architecture of distribution system operators (DSOs) and transmission-distribution coordination in a decentralized, layered electricity network based on renewable energy. Presentation for Stanford University Bits & Watts, June 2017.

1. Electric Distribution Grids
in a 21st Century Energy System
Lorenzo Kristov
Principal, Market & Infrastructure Policy
Bits and Watts Initiative Meeting
Stanford University – June 1, 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. Decarbonization of electricity and the greater economy
requires whole-system rethinking of the electric system.
v Three broad forces are reshaping the electric system
§ Policies – and increasingly economics – driving a shift to renewable
energy and away from fossil generation at all scales
§ Decisions of customers, communities and cities to determine their
energy sources and uses for efficiency, resilience, local economic
and system benefits, and lower environmental impacts
§ Proliferation of powerful, cost-effective technologies to enable the
desired changes
v These forces upset all aspects of the traditional paradigm of
central-station generators and one-way power flows
§ The new behind-the-meter (BTM) “market” for energy services is
eroding the grid-based commodity energy market
§ Electric distribution systems must re-conceptualize their mission and
value, and develop new functional capabilities
§ Policy makers, regulators and planners need to learn to navigate
bottom-up-driven change
Page 3

4. Change is so all-encompassing it’s hard to identify and
stay focused on the essentials.
v The duck curve must be solved to sustain renewable expansion
§ Accelerate multiple solutions: storage to time-shift supply, flexible
demand-side resources, regional diversity, decentralization
§ Specify and compensate performance attributes that address the
problem but don’t yet have clear revenue opportunities
v Recognize where targeted policies may miss or even work
against the broader systemic objectives
§ Net energy metering (NEM) and zero net energy (ZNE) mandates
accelerate rooftop PV adoption but worsen the duck curve
§ For the objective to decarbonize transportation, personal EVs
constitute only one of several strategies
v Assess potential changes within a whole-system grid architecture
§ Consider a future “integrated decentralized” electric system, where
multi-user and community microgrids provide local energy needs and
resilience while using the bulk system for residuals.
Page 4

5. Recent events surpass previous forecasts of net load
and afternoon ramp with high solar PV on the system.
Typical Spring Day
Net Load 10,386
MW on April 9, 2017
Page 5
Actual 3-hour ramp
12,960 MW on Dec.
18, 2016

6. Distributed solar PV in CAISO, for system peak < 50 GW
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 6

7. DER business models seek to provide services and earn
revenues at multiple levels of the system.
“DER” = all energy resources connected at distribution level, on
customer side or utility side of the customer meter
– Plus communications & controls to aggregate & optimize DER
v Behind the end-use customer meter, on-site energy facilities
– Time of day load shifting, demand charge management
– Customized reliability & service resilience – smart buildings,
microgrids, critical loads
v Distribution system services
– Increase hosting capacity, defer new infrastructure
– Operational services – voltage, power quality
v Transmission system and wholesale market
– ISO spot markets for energy, reserves, regulation
v Bilateral energy contracts with customers, DOs & LSEs
– Renewable energy, resource adequacy
v Peer-to-peer transactions, via distribution-level markets
Page 7

8. Benefit-cost assessment of DERs and microgrids suffers
from benefits not yet defined, valued, compensated.
v Smoothing the ducklings
– System-level duck is comprised of 1000s of “ducklings” –
similar load shape, with “feathers” (real-Cme volaClity)
– Manage local variability locally
– Smoothing the ducklings can be carbon free, rather than
using gas to manage system-level duck
v Resilience
– Widely recognized, but not yet well defined or valued
– “Ability to maintain essen/al services for extended periods
of /me (months) under major stress or disrup/ve event”
– Resilience is essenCally a local aKribute
Page 8

9. Increasing DER volume and diversity requires a new
T-D interface coordination framework.
Area of Activity Challenges of High DER What’s Needed
System operations • Diverse DER behaviors & energy flows,
esp. with aggregated virtual resources
• Hard to forecast impacts at T-D
interfaces
• DO is not informed of DER wholesale
market transactions
• Multi-use DER may receive conflicting
dispatches/signals (from DO and ISO)
• Distribution grid real-time visibility
• Real-time forecasting of DER
impact at each T-D substation
• Coordination procedures between
ISO, DO and DER re wholesale
DER schedules & dispatches
• Dispatch priority re multiple use
applications (MUA)
T & D infrastructure
planning
• Long-term forecasting of DER growth &
impacts on load (energy, peak, profile)
• Assess DER to offset T&D upgrades
• Distribution grid modernization needs
• Align processes for T&D planning
and long-term forecasting
• Specify required performance for
DER to function as grid assets
• Modernize grid in logical stages for
“no regrets” investment
System reliability &
resilience
• High DER may make traditional top-down
paradigm obsolete
• Develop new models to “layer”
responsibilities for reliability
Market & regulatory
issues:
• Wholesale v retail
• Monopoly v
competition
• Complexity of central dispatch of high
DER volume in wholesale market
• Unclear boundary between competitive v.
utility services, while utilities & insurgents
pursue new business models
• Explore “Total DSO” aggregation
of DER for wholesale market
transactions
• Consider optimal scope of
regulated distribution monopoly
with high DER
Page 9

10. The design of TSO-DSO coordination for high DER is
inseparable from design of the future DSO.
v Bookend A: Current Path or “Minimal DSO”
– DSO maintains current distribution utility role, with enhancements only
as needed to ensure reliability with high DER volumes
– Large numbers of DER centrally optimized in ISO market
– DER engage in “multiple-use applications” (MUA) providing services to
end-use customers, DSO and wholesale market
v Bookend B: “Total DSO”
– DSO expands its role to include
• DER aggregation for wholesale market participation
• Optimizing local DER to provide wholesale market services
• Balancing supply-demand locally
• Managing DER variability to minimize impacts at T-D interface
• Potential to operate distribution-level transactive markets
– DSO provides a single aggregated bid to ISO at each T-D interface
– MUA are simplified, dispatch conflicts eliminated, because DSO is
responsible for response to ISO dispatches
Page 10

11. “Minimal DSO” retains traditional primary DO mission –
reliable operation & planning – with high DER.
ISO directly integrates all DER for both transmission and distribution system
operations. Requires ISO to incorporate distribution grid network model and have
complete real-time distribution grid state information.
This approach is not advised due to complexity & scaling risks

12. “Total DSO” – similar to an ISO at distribution level – is
the most robust & scalable model for high DER.
DSO directly integrates all DER for Local Distribution Area for each T-D Interface (e.g.,
LMP pricing node) and coordinates T-D interchange with TSO, so that ISO sees only a
single resource at each T-D interface and does not need visibility to DER. DSO manages
all intra-distribution area transactions, schedules and energy flows.
DSO coordinates a single aggregation of all DER at each T-D interface

13. The choice of DSO model implicates several key
power system design elements.
Design Element Minimal DSO Total DSO
Market structure Central market optimization by ISO with
large numbers of participating DER
DSO optimizes local markets at each T-D
substation; ISO market sees a single virtual
resource at each T-D interface
Distribution-level energy
prices
Locational energy prices based on LMP +
distribution component (e.g., LMP+D)
Based on value of DER services in local
market, including LMP for imports/exports
Resource/capacity
adequacy
As today, based on system coincident
peak plus load pocket & flexibility needs;
opt-out allowed for micro-grids
Layered RA framework: DSO responsible
for each T-D interface area; ISO responsible
to meet net interchange at each interface
Grid reliability paradigm Similar to today Layered responsibilities; e.g., DSO takes
load-based share of primary frequency
response
Multiple-use applications
of DER (MUA)
DER subject to both ISO and DSO
instructions
Rules must resolve dispatch priority,
multiple payment, telemetry/metering
DER subject only to DSO instructions, as
DSO manages DER response to ISO
dispatches & ancillary services provision
Regulatory framework Federal-state jurisdictional roles similar to
today
Explore framework to enable states to
regulate distribution-level markets
Comparison to existing
model
Current distribution utility roles &
responsibilities, enhanced for high DER
Total DSO is similar to a neighboring
balancing authority or TSO
Page 13

14. Simplicity reveals an architectural issue: “tier bypassing”
limits scalability, adds complexity and security risks.
TransCo
Merchant
Gen
Cust
Sites
DistCo
Microgrid
s
TSO/BA TSO/BA
TransCo
Merchant
Gen
Cust
Sites
Merchant
DER
DSO
Microgrid
s
Merchant
DER

15. Source: J. Taft, P. De Martini & L. Kristov
Simple diagrams can obscure the true complexity.

16. DER growth and distribution-level markets trigger other
new policy and design questions.
v Open-access structure for distribution system operators (DSOs)
– Non-discrimination in distribution services, resource interconnection,
infrastructure investment, real-time re-dispatch as needed
– Is an independent DSO needed? Or can today’s utility DO be effectively
firewalled into a regulated “wires & markets” operator and competitive
affiliate offering retail services?
v Possible new boundary clarification for federal-state jurisdiction
– Could states regulate “sales for resale” that occur within a local
distribution-level market that doesn’t use transmission?
v Customized reliability, layered resource adequacy and reliability
– “Total DSO” aggregates all DER & customers below a T-D substation
and submits a single virtual resource to ISO at the interface
– ISO is responsible for system reliability only to the T-D interface
– DSO is responsible from T-D interface to the customer meter
– A micro-grid takes responsibility for its own reliability, and will island if
grid supply is limited or interrupted
Page 16

17. DSO = Local Distribution Area or
Community Microgrid
Regional Interconnection
The future grid may be an “integrated decentralized”
system, a layered hierarchy of optimizing sub-systems.
Page 17
Smart
building
Micro-
grid
Micro-grid
D
S
O
D
S
O
ISO
Balancing
Authority
Area
BAABAA
Smart
building
Smart
building
• Storage, DG and controls on
premises form a building-level
microgrid
• Each layer only needs to see
its interchange with next layer
above & below, not details
inside other layers
• ISO optimizes regional bulk
system only up to the T-D
interfaces
• Layered control structure
reduces complexity, allows
scalability, and increases
resilience & security
• Fractal structure mimics
nature’s design of complex
organisms & ecosystems.

18. Putting the pieces together: Architecture of a clean,
sustainable, and resilient 21st century electric system
The building blocks: “resilient communities”
v Local power systems based on renewable energy and storage,
designed to meet local needs and support local economy
v Integrated municipal systems with electricity at the core: water,
wastewater, solid waste, transport, telecoms, emergency services
v Eliminate fossil fuel combustion by electrifying everything and
closing the loops (e.g., extract fuel from the waste stream)
v Ramp up renewable energy and efficiency in buildings by integrating
with electric and thermal storage, dynamic demand management
The whole system: “decentralized and integrated”
v Layered control structure comprised of concentric, self-optimizing
buildings, microgrids, local distribution areas (DSOs)
v Able to island at each level if needed, but connected 99.9% of the
time for transacting energy and services
v Smart devices and IoT function within levels, not across levels
Page 18

19. And … what’s needed to get there ...
v Resolve policy ambivalence toward community choice aggregation
(CCA) to recognize essential role of communities in achieving state
objectives while aligning energy decisions to local needs
– Implement state initiatives to build human resource capacity in all
jurisdictions throughout the state
v Adopt policies to incentivize local smoothing of load profiles and
variability – e.g., distribution charges to reflect impact on the grid
– Eliminate NEM entirely: subsidize batteries for existing NEM customers;
incentivize PV + storage systems
– Update building standards to include thermal & battery storage, rooftop
PV, EV charging, “microgrid ready” control systems
– Redefine ZNE at the community level, not individual buildings
v Value resilience as a local attribute; create community microgrids
v Develop open-access framework for DSOs
v Redesign national reliability and resource adequacy frameworks to
align with layered grid architecture – each layer responsible for its
own reliability and adequacy, capable of islanding
v Consider high-DER scenarios in infrastructure planning
Page 19

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

21. References
v L. Kristov, P. De Martini, J. Taft (2016) “Two Visions of a Transactive Electric
System” (IEEE Power & Energy Magazine, May-June 2016):
http://resnick.caltech.edu/docs/Two_Visions.pdf
v P. De Martini & L. Kristov (2015) “Distribution Systems in a High Distributed Energy
Resources Future: Planning, Market Design, Operation and Oversight” (LBNL series
on Future Electric Utility Regulation):
https://emp.lbl.gov/sites/default/files/lbnl-1003797.pdf
v DOE (2017) Next Generation Distribution Platform Project: http://doe-dspx.org/
v L. Kristov (2015) “The future history of tomorrow’s energy network” (Public Utilities
Fortnightly, May 2015):
http://www.fortnightly.com/fortnightly/2015/05/future-history-tomorrows-energy-
network?
page=0%2C0&authkey=afacbc896edc40f5dd20b28daf63936dd95e38713e904992a6
0a99e937e19028
v L. Kristov & P. De Martini (2014) “21st Century Electric Distribution System
Operations”:
http://smart.caltech.edu/papers/21stCElectricSystemOperations050714.pdf
v J. Taft et al, Pacific Northwest National Laboratory, Grid Architecture site:
http://gridarchitecture.pnnl.gov