Grid Resilience and Intelligence Platform (GRIP) Use-case CEC Advanced Simulation Program TAC Meeting Oakland, California 6 September 2019

Grid Resilience and Intelligence Platform (GRIP) Use-case Alyona Ivanova (SLAC) CEC Advanced Simulation Program 6 September 2019 This presentation was prepared with funding from the California Energy Commission under grant EPC-17-046. SLAC National Accelerator Laboratory is operated for the US Department of Energy by Stanford University under Contract No. DE-AC02-76SF00515

Import & Validate (OpenFIDO) Outline External datasets External models Shared datasets Shared models Import & Validate (OpenFIDO) Simulate (HiPAS & GLOW) Analyse (custom) CSV Files Plots MySQL Data Copy/link (Cloud Security) Introduction Objectives Resilience overview Anticipation Pole model Pole degradation User Interface Absorption Model Overview Planned Improvements Acknowledgements The GRIP project is funded by the US Department of Energy under the Grid Modernization Laboratory Consortium (GMLC).

Grid Resilience and Intelligence Platform (GRIP) Applies artificial intelligence and machine learning to anticipate, absorb and recover from extreme events that affect the Distribution Energy System

GRIP Innovation and Impact GRIP applies AI and ML for grid resilience. National impact with platforms and analytics Facilitates streamlining use of ML/AI applications for distribution resources. Final deliverable: open-source commercially available product. Year 1: Anticipation (Completed) Year 2: Absorption (In-progress) Year 3: Recovery (Future Work)

Anticipation Objectives Determine use-cases for resilience. Asset and protective device location and mapping Predicting vulnerabilities to extreme weather conditions Switch re-configuration Predicting ferroresonance occurrences Secondary voltage optimization with DERs Vegetation management Optimized work plans considering budget hardening options Develop a new platform based on pre-existing Google tools. Use previously DOE funded projects (VADER, OMF) as basis for GRIP Test and validate anticipation solution with data and models provided by National Rural Electric Cooperative Association (NRECA).

Core GRIP Simulation Analysis GridLAB-D implementation incorporates vulnerability analysis Analytical pole and line vulnerability model using weather data Calculation of pole vulnerability index, electrical fault propagation and restoration time Wind stress simulation represents worst case scenario Support for arbitrary vulnerability simulations User ability to specify the electrical system model Calculations of stresses loading to pole failures Cables tension Pole-mounted equipment Pole tilt angle Wind and ice loading

Internal pole degradation model Applicable to wood poles only Propagation of internal core degradation to outer edges Degradation defined by minimum shell thickness End-of-life thickness: 2” Characterized by the difference between the outer core and inner core moment calculations Accounts for pole base failures 2”

IEEE Standard test models Image obtained from DOI: 10.1109/TSG.2013.2288868 Image obtained from DOI: 10.1109/TDC.2010.5484381

GRIP Further Design Creating a dashboard that not only lists the simulations, but highlights status and results of each simulation.

Refining the layout data visualization

Absorption Dynamic reconfiguration of the network into virtual islands after distribution circuit damage due to weather disturbances Maximize amount of load that can be maintained during the fault Uses DG, batteries, flexible DERs (water heaters, distributed batteries, EV chargers) when supporting virtual islands Model development (SLAC/GridLAB-D) + Controls (Packetized Energy) + UI (Presence) Typical sequence: (1) Fault occurs (2) Fault isolation (3) Reconfiguration (Virtual Islanding) (4) Load balancing

VIRTUAL ISLANDING TEST CASE built in GridLAB-D 001 Node representing the Bulk Grid VIRTUAL ISLANDING TEST CASE built in GridLAB-D 1 NC NC 2 NC 002 4 Legend NC Switch or circuit breaker/recloser that is closed (hot) 3 101 201 301 Switch or circuit breaker/recloser that is open (not hot) 5 7 9 102 202 Solar1 Solar2 302 Solar3 xyz Distribution circuit node (or collection of nodes) with (eg) hundreds of customers. 6 10 8 11 12 103 203 303 Fault location NO NO Battery1 Battery2 Battery3

CASE 1: SINGLE FEEDER FAULT 001 Node representing the Bulk Grid CASE 1: SINGLE FEEDER FAULT 1 NC NC 2 NC 002 4 Legend NC Switch or circuit breaker/recloser that is closed (hot) 3 101 201 301 Switch or circuit breaker/recloser that is open (not hot) 5 7 9 102 202 Solar1 Solar2 302 Solar3 xyz Distribution circuit node (or collection of nodes) with (eg) hundreds of customers. 6 10 8 11 12 103 203 303 Fault location NO NO Battery1 Battery2 Battery3

CASE 1: STEP 1 BREAKER TRIPS 001 Node representing the Bulk Grid CASE 1: STEP 1 BREAKER TRIPS 1 NC NC 2 NC 002 4 Breaker opens Legend NC Switch or circuit breaker/recloser that is closed (hot) 3 101 201 301 Switch or circuit breaker/recloser that is open (not hot) 5 7 9 102 202 Solar1 Solar2 302 Solar3 xyz Distribution circuit node (or collection of nodes) with (eg) hundreds of customers. 6 10 8 11 12 103 203 303 Fault location NO NO Battery1 Battery2 Battery3

CASE 1: STEP 2 FAULT ISOLATION 001 Node representing the Bulk Grid CASE 1: STEP 2 FAULT ISOLATION 1 NC NC 2 NC 002 4 Legend NC Switch or circuit breaker/recloser that is closed (hot) 3 101 201 301 Switch or circuit breaker/recloser that is open (not hot) 5 7 Sw. opens 9 102 202 Solar2 302 Solar1 Solar3 xyz Distribution circuit node (or collection of nodes) with (eg) hundreds of customers. 6 10 Sw. opens 8 11 12 103 203 303 Fault location NO NO Battery1 Battery2 Battery3

CASE 1: STEP 3 RECONFIGURATION 001 Node representing the Bulk Grid CASE 1: STEP 3 RECONFIGURATION 1 NC NC 2 NC 002 4 Legend Switch or circuit breaker/recloser that is closed (hot) 3 101 201 301 Switch or circuit breaker/recloser that is open (not hot) 5 7 9 102 202 Solar1 Solar2 302 Solar3 xyz Distribution circuit node (or collection of nodes) with (eg) hundreds of customers. 6 10 8 11 12 103 203 303 Fault location Battery1 Battery2 Battery3

Planned Improvements Planned enhancements to the pole vulnerability model Add effect of changing wind direction Add ice build-up model and line loading effects Extend taxonomy of impacts of vegetation on lines, poles, and equipment Pole degradation Pole top failures due to equipment, weather and animal impacts

Vegetation Risk Assessment Risk modeling and quantification questions related to vegetation Effect of vegetation on the electrical system Other effects beside vegetation falling on lines (e.g., line-to-tree faults, etc.) Additionally the effect of the electrical system on the surrounding community Model the fire initiation and propagation risk through vegetation Develop a vulnerability index for the surrounding community Account for recent history, damage, impact zones, magnitudes Use US Forest Service models to assist model development for fire behavior

GRIP Technical Team SLAC National Accelerator Laboratory PresencePG Packetized Energy National Rural Electric Cooperative Association Lawrence Berkeley National Laboratory

Acknowledgements GRIP Technical Advisory Group Members

Questions?