Transcript slide deck - The California IDER and DRP Working Groups

Integrated Capacity Analysis
Working Group
January 6, 2017
In-person meeting
drpwg.org
1
Agenda
Time
Topic
A. Introduction, Overview of January Calendar, Purpose of WG Report
9:00-9:15
B. IOU presentations on final demo results
9:15 - 10:00
10:00 – 11:00 C. Q&A session with IOUs and WG members on results
11:00 – 12:00
12:00
D. WG recommendations and report drafting process
E. Next Steps and Lunch
2
ICA and LNBA Working Group Background
ICA and LNBA WG Purpose - Pursuant to the May 2, 2016, Assigned Commissioner’s Ruling (ACR) in
DRP proceeding (R.14-08-013), the Joint Utilities are required to convene the ICA and LNBA WG to:
1. Refine ICA and LNBA Methodologies and Requirements
2. Authorize Demonstration Project A and Project B
CPUC Energy Division role
•
Oversight to ensure balance and achievement of State objective (ensure adequate stakeholder representation in
consensus statements, keeping WG activities on track with Commission expectations/needs, demonstration project
results review, quality control on deliverables)
•
Coordination with both related CPUC activities and activities in other agencies (IDER CSF WG, CEC and CAISO
interagency matters, interconnection/Rule 21/SIWG, other proceedings that may impact or be impacted by
locational value calculation such as AB 350/IRP and LTPP/TPP/RPS)
•
Steward WG agreements into CPUC decisions when necessary
More Than Smart role
•
Engaged by Joint Utilities to facilitate both the ICA & LBNA working groups. This leverages the previous work of MTS
facilitating stakeholder discussions on ICA and LBNA topics.
3
CPUC Decision Points for
ICA/LNBA Working Group
Short-Term Final Reports
ICA/LNBA Working Group
January 6, 2017
Oakland, CA
4
Proposed Decision on Demo A&B
will need to address:
• Compliance with May 2, 2016 ACR (and
February 6, 2015 Guidance Document)
• ICA/LNBA Use Cases
• WG-recommended methodological adjustments
needed to achieve Use Cases
– Based on prescribed Demo methodologies and ShortTerm refinements
• Implementation requirements/schedule
5
Consensus Process Proposal for Discussion
For each topic, the WG will characterize its recommendation as:
1. General or majority consensus (by a show of hands or absence of
objecting voices)
2. Non-consensus, with alternative proposals
For non-consensus positions, MTS will identify the number of parties and list
party’s attribution and proposed alternative in the report
A comment period after the final report is filed is considered for parties to submit
final comments.
MTS will work with WG members to develop text for the report. For certain
sections, the WG may assign a chair and co-chair lead to coordinate development
of text.
6
Update on Timelines
Date
Short Term
Jan. 6
In-person meeting
Jan. 11
First draft circulated
Jan. 17
First round of edits due
Jan. 18-19
Second draft circulated
Jan. 20
Webinar
Jan. 27
Second and final round of edits due
Jan. 31
Report submitted
Long Term
• Determine schedule for discussions up to Q2
2017
• Begin LT refinement discussions
For consideration: Extend first round of writing to Jan. 12/13?
7
Integration Capacity Analysis (ICA) – Working Group
IOU Demo A Findings Overview
January 6, 2017
DRAFT
Purpose and Overview
Provide the ICA Working Group an overview of the Demo A finding with overview of the ICA
Roadmap
• Demonstration Project A Summary (15 Minutes)
–
–
–
–
–
–
Distribution Planning Areas (DPAs)
Integration Capacity Analysis (ICA) Methodologies
Overview of the Results
Comparative Assessment between ICA Methodologies
Mapping and Publication Overview
Additional Studies
• Challenges and Lessons Learned (15 Minutes)
– Modeling and Simulation of Power Systems
– Software, Script Development, and Computational Performance
– Mapping and Publication
• ICA Roadmap (15 Minutes)
– SCE System
– Short- and Long-Term Enhancements
DRAFT
9
Distribution Planning Areas (DPAs)
Southern California Edison
Pacific Gas & Electric
San Diego Gas & Electric
Distribution
Planning Area
Rector DPA
(Rural)
Johanna DPA
(Urban, Suburban)
Demo A DPA
Chowchilla DPA
(Rural)
Chico DPA
(Urban, Suburban)
Demo A DPA
Ramona DPA
(Rural)
Northeast DPA
(Urban, Suburban)
Demo A DPA
Substations
5
4
9
4
10
14
5
19
24
Feeders
44
38
82
20
41
61
19
137
156
Loading (MW)
314
217
531
155
235
370
65
815
880
Number of
Customers
49,700
25,100
74,800
13,000
125,000
138,000
20,538
204,985
225,523
Nodes
12,442
4,687
17,129
12,914
22,898
35,812*
40,459
228,049
268,643
In the selection of the DPAs, the IOUs considered characteristics such as average feeder
length, average electrical resistance, feeder loading levels, customer diversity, and installed
distribution equipment
* PG&E’s new GIS models in development have 20% more nodes for the demo A area due to enhanced geospatial accuracy using new GIS system.
DRAFT
10
Integration Capacity Analysis (ICA) Methodologies Alignment
Streamline Method
Iterative Method
Limits
Southern California
Edison
Pacific Gas & Electric
San Diego Gas & Electric
Thermal
Thermal
Equipment Thermal
Rating
X
X
X
Steady State
Voltage
Steady State
Voltage
Over or Under
Voltage
X
X
X
Voltage Fluctuation
Voltage Fluctuation
Change in Voltage of
3%
X
X
X
Protection
Protection
Relay reduction of
reach
X
X
X
Safety Reliability
Safety Reliability
No reverse on
SCADA devices
X
X
X
DRAFT
11
General Demo Result Trends
DRAFT
Demo Example Results by Distance
Typical Rural Results
DRAFT
Typical Urban Results
Typical Results (Mixed chart)
ICA Methodologies –
Comparative Assessment
• Iterative ICA methodology more
accurate in determining power
flow impacts, but more difficult
to determine hosting capacity;
issues often arise with with
simulation convergence.
• Streamlined ICA Methodology is
significantly faster, but may lack
ability to determine specific
complex operational issues
On average, the SCE Iterative and
Streamlined ICA results differed
between 18% to 30% on a node to
node basis.*
SCE Demo A Results Overview – Average hosting capacity ranges.
At a high level, Streamlined and Iterative ICA Methodologies align on results. The difference is apparent
when node-to-node comparisons are established though, with difference ranging between 18% to 30%
* PG&E’s iterative results created many non-convergent and/or zero results that were not reasonable to evaluate across entire demo area for report
DRAFT
14
Mapping and Data Sharing
Southern California Edison
Pacific Gas & Electric
San Diego Gas & Electric
http://on.sce.com/derim
Will provide on publication
https://sempra.maps.arcgis.com/
Circuit
Section ID
Voltage (kV)
Substation
System
Customer Breakdown % (Ag, Comm, Ind, Res, Other)
Existing Generation (MW)
Queued Generation (MW)
Total Generation (MW)
Integration Capacity, Uniform Generation (MW)
Integration Capacity, Uniform Load (MW)
Integration Capacity, Typical PV System (MW)
All map content can be downloaded via ESRI Open Data
same
same
Map
Attributes
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
DRAFT
15
Mapping and Data Sharing
Southern California Edison
Pacific Gas & Electric
San Diego Gas & Electric
same
same
Load Profiles
http://on.sce.com/derimwebapp
Attributes
1. Typical high-load day (288 data points)
2. Typical light-load day (288 data points)
Interactive. Hover over to obtain x/y coordinates.
DRAFT
16
Mapping and Data Sharing
Southern California Edison
Downloadable
Data
Pacific Gas & Electric
San Diego Gas & Electric
Will provide on publication
http://on.sce.com/drpdemos
Files Available
1.
2.
3.
4.
5.
Demo A Final Report
ICA Translator
Load Profiles
Customer Type Breakdown
Detailed ICA Results by Circuit
Additional
Resources
http://on.sce.com/derimguide
DRAFT
Will provide on publication
Will provide on publication
17
1.
2.
3.
4.
ICA Translator
Load Profiles
Customer Type Breakdown
Detailed ICA Results by Circuit
TBD
Computation Summary
Southern California
Edison*
Pacific Gas & Electric San Diego Gas & Electric
Average Run Time per Feeder
(Streamlined Method)
2 minutes
7 minutes
30 minutes
Average Run Time per Feeder
(Iterative Method)
23 minutes
83 minutes
1,620 minutes
Data Size
6.66GB
~100GB
4.79GB
* SCE’s run time statistics reflect limitation category and hourly reduction measures, but does not account for
data processing.
DRAFT
18
Lessons Learned and Challenges – Electric System Modeling
• Nodal granularity requires precise models and possibly manual editing
– 1.Load, DER, connectivity, and equipment sometimes requires manual intervention
– 2.SCADA and AMI must be reconciled
– 3.Unknown phasing can limit model accuracy
• Nodal/hourly simulation requires rigorous processing
–
–
–
–
1.Feeder length & complexity increase number of iterations
2.VAr supply and/or equipment settings can lead to non-convergence of simulations
3.Error tolerance in load allocation and power flow solution may need to be increased
4.Common bus solution leads to longer processing and more non-convergence of simulations
• Such complex results are difficult to validate
– 1.Large volume of data is difficult to manually validate
– 2.Zero ICA may result from unexpected or indirect electrical conditions
– 3.Expected result validation requires tedious manual simulation (spot checks)
DRAFT
19
Lessons Learned – Computational Challenges
• Development of scripts
– Translating theory to practice is an iterative method requiring extensive knowledge of engineering and
computer science principles
– Must be coordinated with software vendors and relies on incremental software updates to ensure proper
functionality for complex analysis.
• Long Runtime Required
– Computational time required for each methodology is exacerbated by the amount of feeders, feeder
complexity, method used, and simulated hours.
• Scalability of Systems
– Scripts, databases, and processing capabilities need to be scaled up to handle periodic ICA updates of the
entire distribution system
• Data Management and Quality Assurance
– The ICA data structures need to progress into more robust enterprise friendly databases
– Results Validation
• Volume of results makes it increasingly difficult to validate data
• Expected vs obtained results based on manual simulation (spot checks)
• ICA Process
– Due to complexities in process, from data gathering and modeling to simulation and publication, automated
end-to-end process will be required to meet goal of analyzing the entirety of distribution system
DRAFT
20
Lessons Learned – Publication Challenges
• Granular nodal/hourly result set is too big for current map publication interface.
Reduction of data to be published and/or different data interface (i.e. Azure cloud) could
alleviate concerns.
• IOUs would like to limit future downloadable data to only the most actionable data going
forward that can practicably be applied towards current operational framework.
• Alternative solutions to publishing interactive load profiles should be explored as well.
The web map applications are not the ideal platform for publishing non-geospatial time
based load profiles.
• Current mapping platforms are expected to be sufficient for node results, but full testing
has not been conducted.
– SCE has identified an investment in the Distribution Resource Plan External Portal (DRPEP, 2018 GRC)
that would greatly enhance the user experience.
– PG&E currently has no specific funding for a new DRP data portal
– SDG&E currently has no specific funding for a new DRP data portal, but has made some modifications
to existing portal for Demo A
DRAFT
21
ICA Roadmap
• Proposal to perform final ICA implementation within 12 months following final resolution
from the PUC.
• IOUs to evaluate the resource requirements to further enhance the efficiency and
scalability of full ICA deployment.
• SCE has identified a number of investments in its 2018 GRC (SMT, DRPEP, LTPT, and
GAA) that will continue to enhance overall ICA efficiency and customer experience.
DRAFT
22
Recommendations
Activity from Ruling
3.1.a Update schedule for Demo A results
Southern California Edison
Pacific Gas & Electric
San Diego Gas & Electric
N/A
N/A
N/A
IOUs should continue to use tech agnostic approach for
2017 implementation. IOUs should not be required to
make assumptions about CAISO dispatch or other
specific resource types.
same
same
IOUs should continue to study smart inverters with the
WG, but should not be required to incorporate smart
inverters in the 2017 implementation.
same
Same
IOUs should provide uniform inverter gen and uniform
load, as was done with Demo A.
same
Same
3.1.c Recommend a format for the ICA maps
to be consistent and readable to all California
stakeholders across the utilities’ service
territories with similar data and visual
aspects (color coding, mapping tools, etc.)
3.1.d Evaluate and recommend new methods
that may improve the computational
efficiency of the ICA tools and process in
order to calculate and update ICA values
across all circuits in each utility’s service
territory in updated ICAs more frequently
and accurately
3.1.e Evaluate ORA's recommendation to
require establishment of reference circuits
and reference use cases for comparative
analyses of Demo Project A results
IOUs should work with the WG to implement minor
modifications to the maps and data sets.
same
same
IOUs should not be required or expected to implement
major revisions for the 2017 implementation.
same
same
IOUs should continue to improve efficiency where
possible. The approach in Demo is acceptable for
system-wide implementation. (see p. 76 of SCE report)
IOUs should continue to improve efficiency where
possible. PG&E supports streamlined for system-wide
analysis and finds iterative not to be feasible for full
system analysis
IOUs should continue to improve efficiency where possible.
The approach used by SDG&E in Demo is acceptable for
system-wide implementation, but is infeasible for frequent
updates.
Additional use of reference circuits is not necessary but
SCE is acceptable to explore additional reference
circuits.
A more complex reference circuit than the IEEE 123 test
feeder will be helpful, but reference set should be limited
to a single circuit due to complexities of comparison as
observed in Demo A
3.1.f Establish a method for use of Smart
Meter and other customer load data to
develop more localized load shapes to the
extent that is not currently being done
The approach used in the Demos is appropriate for
System wide implementation with increased utilization
of smart meter data based on technology advances to
remove to remove obstacles observed in Demo A.
The approach used in the Demos is appropriate for
System wide implementation
3.1.g Establish definite timelines for future
achievement of ICA milestones including
frequency and process of ICA updates
DRAFT
IOUs should update ICA on a monthly basis following
the initial system-wide implementation, if possible
given computing resources. Initial system wide
deployment should allow 12 months from PUC final
ruling
ICA should be updated no faster than on a monthly basis
following the initial system-wide implementation. Initial
system wide deployment should allow 12 months from
PUC final ruling
3.1.b Recommend methods for evaluation of
hosting capacity for the following resource
types: i) DER bundles or portfolios,
responding to CAISO dispatch; ii) facilities
using smart inverters
While reference circuits can be valuable, they will require
work that will take away from system wide implementation
The approach used in the Demos is appropriate for System
wide implementation
updates should be done on a semi-annual basis for
interconnection purposes. Scenario analysis can be
performed annually.
Other Recommendation
Southern California Edison
Pacific Gas & Electric
San Diego Gas & Electric
 recommend a blended approach with adequate levels of
iterative power flow method and streamline method
 Recommend a blended approach using iterative
and streamlined method where deemed appropriate
due to use case and practicability
 For Interconnection use case iterative method is
recommended. This information is what would be used for ICA
maps
 Iterative method is recommended for
Interconnection use case . PG&E did not find iterative
feasible for full distribution analysis and nodal mapping
and thus should be limited to interconnection process.
 For other use cases such as Planning or scenario analysis,
the streamline method is recommended
 Streamlined method is recommended for Planning
use case in which full nodal and system wide analysis
can be performed efficiently.
Timing of initial systemwide implementation
 IOUs should be given at least 12 months to perform initial
system-wide implementation, assuming above
recommendations are adopted. Additional requirements may
require additional time for completing ICA
same
same
Single phase feeders
 IOUS should continue to study potential expansion to
single phase feeders. IOUs should not be required to include
single phase feeders in the 2017 implementation
same
Same
Long-term refinements
 Long-term refinements being discussed in the working
group. The Q1 Decision should not address long-term
refinements.
same
same
Overall Methodology
DRAFT
24
• SDG&E recommends a blended approach with iterative
power flow method used for interconnection, and the
streamline method used for planning
• For Interconnection use case iterative method is
recommended. This information is what would be used
for ICA maps
• For other use cases such as Planning or scenario
analysis, the streamline method is recommended
WG Report Recommendations: Proposed Outline
1.
Executive Summary
1. Key Takeaways:
1.
2.
3.
2.
3.
Use cases
Methodology
Implementation schedule and
requirements
4.
5.
Evaluation of Results
1. ORA 12 Success Criteria
2. Other?
ICA Implementation Requirements and Time
Frame (timing/frequency of updates)
1. Initial implementation deployment
2. Further long term refinements
Use Cases
Methodology Recommendations
B. Evaluation for 1) DER bundles or portfolios
and 2) facilities using smart inverters
Others?
C. Map format
D. Computational efficiency
E. Comparative analyses/reference circuits
1. Response to recommendations by
IOUs of use of iterative/streamlined
approaches
F. Method for smart meters and localized
load shapes
25
www.drpwg.org
26