Optimization of Resilient, Reliable, and Renewable Energy Infrastructure for Large-Scale Essential Services in Urban Environments

Optimization of Resilient, Reliable, and Renewable Energy Infrastructure for Large-Scale Essential Services in Urban Environments Final Report Submitted to: Joe Leiva Research, Development & Demonstration Southern California Gas Company 555 West 5th Street, Los Angeles, CA 90013 Submitted by: Professor Jacob (Jack) Brouwer, Director Victor Klumper Dr. Robert Flores, Senior Scientist Clean Energy Institute University of California, Irvine Irvine, California 92697-3550 Tel: (949) 824-1999 x221 Fax: (949) 824-7423 jbrou[email protected]ci.edu October 03, 2025 Final Report University of California, Irvine iClean Energy Institute Table of Contents Table of Contents ........................................................................................................................... i List of Figures ............................................................................................................................... iv List of Tables ............................................................................................................................... vii Executive Summary ................................................................................................................... viii 1 Introduction ........................................................................................................................... 1 1.1 Purpose ............................................................................................................................. 1 1.2 Literature Review ............................................................................................................. 2 1.3 Current Work and Contributions ...................................................................................... 5 2 Key Assumptions ................................................................................................................... 6 3 Scenarios ................................................................................................................................. 6 4 Resiliency Analysis ................................................................................................................ 8 4.1 Wildfire/PSPS Event Zones ............................................................................................. 8 4.2 Resiliency of the Existing Electrical Grid/Model Verification ...................................... 12 4.2.1 Zone 1 – Lower Palm Springs Corridor .................................................................. 13 4.2.2 Zone 2 – Upper Palm Springs Corridor .................................................................. 14 4.2.3 Zone 3 – Victorville Corridor ................................................................................. 15 4.2.4 Zone 4 – Santa Clarita Corridor .............................................................................. 16 4.2.5 Zone 5 – Central Corridor ....................................................................................... 19 4.2.6 Conclusions on Existing Grid Resiliency Under Full Loads .................................. 20 4.3 Residential & Commercial Building Stock Critical Loads ............................................ 20 4.3.1 Residential Load Profile ......................................................................................... 21 4.3.2 Commercial Load Profile ........................................................................................ 21 4.3.3 Adjusting Load Profiles .......................................................................................... 21 4.3.4 Zone 3 ..................................................................................................................... 22 4.3.5 Zone 4 ..................................................................................................................... 23 4.3.6 Zone 5 ..................................................................................................................... 24 4.3.7 Conclusions on Critical Load Resiliency ................................................................ 24 5 Pipeline Hydrogen as a Resiliency Augmentation ............................................................ 25 5.1 Combining Models ......................................................................................................... 25 Final Report University of California, Irvine iiClean Energy Institute 5.2 System Response to Removal of Non-Renewables ....................................................... 26 5.3 Leveraging Pipeline Hydrogen....................................................................................... 27 5.3.1 Segment 1 – Palm Springs to Compressor Station ................................................. 30 5.3.2 Segment 2 – Compressor Station to Basin .............................................................. 32 5.4 Simulating the Inclusion of Hydrogen Within OpenDSS .............................................. 32 5.5 Including Natural Gas Demand ...................................................................................... 37 5.6 Fuel Cell Sizing Problem ............................................................................................... 39 5.7 Resiliency of Resulting System...................................................................................... 41 5.7.1 Zone 1 – Pipeline Included ..................................................................................... 41 5.7.2 Zone 2 – Pipeline Included ..................................................................................... 47 5.7.3 Zones 1 AND 2 – Pipeline Included ....................................................................... 49 5.7.4 Zone 3 – Pipeline Included ..................................................................................... 52 5.7.5 Zone 4 – Pipeline Included ..................................................................................... 54 5.7.6 Zone 5 – Pipeline Included ..................................................................................... 57 5.7.7 Conclusions on Hydrogen-Reinforced System Resiliency ..................................... 60 6 Performance Under Complete Electrification of End-Uses ............................................. 60 6.1 Resiliency Analysis ........................................................................................................ 62 6.2 Comments on Transportation Electrification ................................................................. 65 7 Technoeconomic Analysis ................................................................................................... 65 7.1 Base System Resiliency Analysis Results ...................................................................... 65 7.1.1 Base System - Zone 3 Disabled Upgrade Cost ....................................................... 66 7.1.2 Base System – Zone 4 Disabled Upgrade Cost ...................................................... 67 7.1.3 Base System – Zone 5 Disabled Upgrade Cost ...................................................... 68 7.1.4 Base System – Total Upgrade Cost ........................................................................ 68 7.2 Hydrogen-Augmented System Resiliency Analysis Results ......................................... 69 7.2.1 Hydrogen-Augmented System – Zone 3 ................................................................. 69 7.2.2 Hydrogen-Augmented System – Zone 4 ................................................................. 69 7.3 Hydrogen-Augmented System Additional Pipeline Cost............................................... 69 8 Conclusion ............................................................................................................................ 69 9 References ............................................................................................................................. 71 Appendix A – Data Collection for Electric Grid Model .......................................................... 72 Geospatial Data ......................................................................................................................... 72 California Electrical Transmission Lines .............................................................................. 72 Final Report University of California, Irvine iiiClean Energy Institute Southern California Edison Distributed Energy Resource Interconnection Map .................. 73 California Power Plants ......................................................................................................... 74 Load & Generation Profile Data ............................................................................................... 76 California Independent System Operator .............................................................................. 76 Residential & Commercial Building Stock ........................................................................... 77 Appendix B - Model Development & Simulation Methodology ............................................. 79 OpenDSS Electrical Grid Model ............................................................................................... 79 Language Format & Requirements for Model ...................................................................... 79 Python Algorithm .................................................................................................................. 79 Public Use Microdata Areas & Substation Selection ............................................................ 80 Transmission Infrastructure Model ....................................................................................... 82 Establishment & Verification of a Base Case ....................................................................... 83 Load Profile Assignment ....................................................................................................... 83 Final Report University of California, Irvine ivClean Energy Institute List of Figures Figure 1: Average reliability and outage rate metrics for the electrical and natural gas networks across the United States .................................................................................................................. 3 Figure 2:Map of the Southern California electrical transmission grid with major substations. Substations ...................................................................................................................................... 7 Figure 16: Southern California gas network ................................................................................... 7 Figure 3: PSPS Events in 2023 ....................................................................................................... 9 Figure 4: Southern California Fire Hazard Areas ......................................................................... 10 Figure 5: Partitioned Fire Zones ................................................................................................... 10 Figure 6: Base Loads & Generation Overload Plot (Zone 1 Disabled) ........................................ 13 Figure 7: Base Loads & Generation Overload Plot (Zone 2 Disabled) ........................................ 14 Figure 8: Base Loads & Generation Overload Plot (Zone 3 Disabled) ........................................ 15 Figure 9: Base Loads & Generation Overload Plot (Zone 4 Disabled) ........................................ 16 Figure 10: Base Loads & Generation Overload Plot (Zones 3 and 4 Disabled) ........................... 18 Figure 11: Base Loads & Generation Overload Plot (Zone 5 Disabled) ...................................... 19 Figure 12: Base & Critical Load Profile Comparison .................................................................. 22 Figure 13: Zone 3 Performance Under Critical Loads, Overload Plot ......................................... 22 Figure 14: Zone 4 Performance Under Critical Loads, Overload Plot ......................................... 23 Figure 15: Zone 5 Performance Under Critical Loads, Overload Report ..................................... 24 Figure 17: Southern California Gas Pipeline & Electrical Infrastructure ..................................... 26 Figure 18: Removal of Natural Gas Generation Overload Plot (Base Loads) .............................. 27 Figure 19: Line 3568 Ampacity Plot ............................................................................................ 28 Figure 20: Overloaded Line Ampacities ....................................................................................... 29 Figure 21: Hydrogen Mass Flow Rate .......................................................................................... 30 Figure 22: Pipeline Segment 1 Performance, Full Mass Flow Rate ............................................. 31 Figure 23: Pipeline Segment 1 Performance, Divided Mass Flow Rate ....................................... 32 Figure 24: Pipeline Segment 2 Performance, Divided Mass Flow Rate ....................................... 32 Figure 25: Electrical Grid, Natural Gas Pipelines, Natural Gas Plants ........................................ 33 Figure 26: Remaining Overloads .................................................................................................. 34 Figure 27: Additional Hydrogen to Address Overloads ............................................................... 34 Figure 28: Previously Overloaded Lines After Hydrogen Implementation.................................. 35 Final Report University of California, Irvine vClean Energy Institute Figure 29: Pipeline System Performance With Final Hydrogen Mass Flow Rate, Divided ......... 36 Figure 30: Hydrogen Profiles........................................................................................................ 37 Figure 31: System Performance Under Full Hydrogen Load ....................................................... 38 Figure 32: Base System Hydrogen Profile.................................................................................... 40 Figure 33: Zone 1 Disabled – System Overload Plot ................................................................... 42 Figure 34: Hydrogen Mass Flow Rate for Full Load Zone 1 ....................................................... 44 Figure 35: Pipeline Performance Report, Full Loads, Zone 1 ...................................................... 45 Figure 36: Pipeline Performance Report, Critical Loads, Zone 1 ................................................. 46 Figure 37: Zone 2 Disabled – System Overload Plot ................................................................... 47 Figure 38: Pipeline Performance Report, Full Loads, Zone 2 ...................................................... 48 Figure 39: Pipeline Performance Report, Full Loads, Zones 1 and 2 ........................................... 49 Figure 40: Complete Pipeline Performance Report, Full Loads, Zones 1 and 2 .......................... 50 Figure 41: Pipeline Performance Report with One Additional Pipe, Base Loads, Zones 1 & 2 .. 51 Figure 42: Zone 3 Disabled – System Overload Plot ................................................................... 52 Figure 43: Pipeline Performance Report, Full Loads, Zone 3 ...................................................... 53 Figure 44: Zone 4 Disabled – System Overload Plot ................................................................... 54 Figure 45: Zone 4 Disabled – Pipeline Hydrogen Applied ........................................................... 55 Figure 46: Pipeline Performance Report, Full Loads, Zone 4 ...................................................... 56 Figure 47: Zone 5 Disabled – Pipeline-Included System Plot ...................................................... 57 Figure 48: Zone 5 Disabled – Pipeline Hydrogen Applied ........................................................... 58 Figure 49: Pipeline Performance Report, Full Loads, Zone 5 ...................................................... 59 Figure 50: Whole-Home Electrification Package, Min Efficiency – Residential ......................... 61 Figure 51: Base & Whole-Home Electrification Load Comparison ............................................. 61 Figure 52: Hydrogen Profile Comparisons ................................................................................... 62 Figure 53: Pipeline Performance Report, Electrification Package, Base System ......................... 63 Figure 54: Pipeline Performance Report, Electrification Package, Zones 1 and 2 Disabled ........ 64 Figure 55: California Electrical Transmission Lines geospatial map ........................................... 73 Figure 56: DERiM substation geospatial map .............................................................................. 74 Figure 57: Power Plants ................................................................................................................ 75 Figure 58: CAISO Overall Generation Profiles ............................................................................ 76 Figure 59: CAISO Renewable Generation Profiles ...................................................................... 76 Final Report University of California, Irvine viClean Energy Institute Figure 60: ResStock aggregate load profile .................................................................................. 77 Figure 61:ComStock aggregate load profile ................................................................................. 78 Figure 62: Key Substations (Buses) .............................................................................................. 81 Figure 63: Map of All Substations in PUMA delineations ........................................................... 81 Figure 64: Grid Infrastructure ....................................................................................................... 82 Figure 65: CAISO load profile compared with DERiM aggregated load profile ......................... 84 Figure 66: Final Load Profiles for PUMA region G0606503 ....................................................... 84 Figure 67: Generation Plant Map .................................................................................................. 85 Figure 68: August 8th Load & Generation Profiles ....................................................................... 86 Figure 69: Adjusted Load & Generation Profiles ......................................................................... 87 Figure 70: Base Case Voltage Profile ........................................................................................... 90 Figure 71: Base Case Overload Plot ............................................................................................. 90 Figure 72: Adjusted Reactance Voltage Profile ............................................................................ 95 Figure 73: Adjusted Reactance Overload Plot .............................................................................. 95 Figure 74: Initial OpenDSS Net Load .......................................................................................... 97 Figure 75: Iterations of net load profiles....................................................................................... 98 Figure 76: Iterations of Natural Gas Duty Cycle .......................................................................... 98 Figure 77: Final Natural Gas Profiles ........................................................................................... 99 Figure 78: Zero Net Load Overload Plot ...................................................................................... 99 Figure 79: Final Base Case Overload Plot .................................................................................. 102 Final Report University of California, Irvine vii Clean Energy Institute List of Tables Table 1: Reliability vs. Resiliency [2] ............................................................................................ 2 Table 2: Base Loads & Generation Voltage Report (Zone 3 Disabled) ....................................... 15 Table 3: Base Loads & Generation Overload Report (Zone 3 Disabled) ..................................... 15 Table 4: Base Loads & Generation Overload Report (Zone 4 Disabled) ..................................... 17 Table 5: Base Loads & Generation Voltage Report (Zone 4 Disabled) ....................................... 17 Table 6: Base Loads & Generation Overload Report (Zone 5 Disabled) ..................................... 19 Table 7: Base Loads & Generation Voltage Report (Zone 5 Disabled) ....................................... 19 Table 8: Residential End Uses ...................................................................................................... 21 Table 9: Zone 3 Performance Under Critical Loads, Overload Report ........................................ 22 Table 10: Zone 4 Performance Under Critical Loads, Overload Report ...................................... 23 Table 11: Removal of Natural Gas Generation Overload Report (Base Loads) ........................... 27 Table 12: Zone 1 Disabled – Pipeline-Included Electrical Performance Report .......................... 43 Table 13: Zone 2 Disabled – Pipeline-Included Electrical Performance Report .......................... 47 Table 14: Zone 3 Disabled – Pipeline-Included Electrical Performance Report .......................... 52 Table 15: Zone 4 Disabled – Pipeline-Included Electrical Performance Report .......................... 54 Table 16: Zone 4 Disabled – Pipeline Hydrogen Applied, Overload Report ............................... 55 Table 17: Zone 5 Disabled – Pipeline-Included Electrical Performance Report .......................... 57 Table 18: Zone 5 Disabled – Pipeline-Included Electrical Performance Report .......................... 58 Table 19: Base System - Zone 3 Disabled Overload Report ........................................................ 66 Table 20: Base System – Zone 3 Upgrade Cost ........................................................................... 66 Table 21: Base System – Zone 4 Upgrade Cost ........................................................................... 67 Table 22: Base System – Zone 5 Upgrade Cost ........................................................................... 68 Table 23: Zone 4 system details.................................................................................................... 69 Table 24: Base Case Voltage Report ............................................................................................ 91 Table 25: Base Case Overload Report .......................................................................................... 91 Table 26: Base Case Voltage Report ............................................................................................ 93 Table 27: Adjusted Reactance Voltage Report ............................................................................. 95 Table 28: Adjusted Reactance Overload Report ........................................................................... 96 Table 29: Zero Net Load Overload Report ................................................................................. 100 Final Report University of California, Irvine viii Clean Energy Institute Executive Summary Resilient, reliable, and renewable energy infrastructure is essential for critical applications such as data centers, hospitals, and military facilities, which demand strict energy service uptime despite the challenges of urban space constraints and the intermittent nature of renewable energy sources like solar and wind. Moreso, outages extending beyond a few hours reduce quality of life in ways that spur affected residents to pursue back-up generation. Solutions for increasing resiliency include 1) hardening energy infrastructure, 2) developing distributed generation, and 3) increasing demand response capabilities. Hardening energy infrastructure, in many cases, includes expanding and undergrounding energy transmission systems, both of which are expensive processes. Likewise, developing distributed generation based on solar and battery energy storage is difficult due to solar capacity size constraints based on building rooftop area coupled with the high expense backup battery energy systems that are rarely utilized in order to ensure reliability. An alternate solution could be to utilize existing the gas transmission system coupled with distributed generation to increase the capacity of locally sited fuel cells. This approach has potential because the existing natural gas transmission infrastructure is substantially more reliable than the electric transmission system. Additionally, introducing renewable hydrogen into the gas transmission system would reduce and eliminate carbon emissions associated with gas pipeline operation, resulting in carbon neutral backup generation. This work aims to demonstrate that renewable hydrogen, integrated with affordable, renewable grid power, offers a robust solution to achieve 100% renewable energy goals while meeting stringent reliability and resiliency standards. Through comprehensive modeling of Southern California's electrical and gas networks, this effort analyzes the dynamic interaction of hydrogen systems—including fuel cells, electrolyzers, and batteries—with local and remote renewable energy sources, the electric grid, and gas networks to address grid outages and improve energy system resilience. This report focuses on how the existing gas transmission system can firm up electricity supply in the face of a less resilient and reliable electric transmission system. Resiliency is examined from two perspectives: 1) when meeting the entire Southern California electrical load, and 2) when meeting a reduced Southern California electrical load based on critical loads from a societal and individual resident level. Subsequent work will reduce loads further, focusing on societal critical loads only (hospitals, communication, transportation, health and safety, water, and other critical infrastructure/loads). The resiliency events considered in this work focus on examining the disabling of major electrical transmission lines that power Southern California. The main motivation for considering these transmission lines are public safety power shutoff (PSPS) events, and wild fires. A far less likely cause is equipment failure unrelated to PSPS and wildfire events. Gas transmission outages are not considered in this work because the gas transmission system has historically achieved resiliency and reliability metrics that are orders of magnitude better than electric transmission systems. This work makes a key assumption that the gas transmission has been converted from carrying natural gas to hydrogen. Prior work has shown that the current natural gas transmission has sufficient capacity to power gas power plants in and around Los Angeles, Orange, San Bernadino, and Riverside counties that make up the Greater Los Angeles Area, boosting system resiliency and reliability. However, the main interest of this work is in examining the electric and gas Final Report University of California, Irvine 5 Clean Energy Institute 1.3 Current Work and Contributions This work focuses on the electrical and natural gas networks in Los Angeles, Orange County, and parts of Riverside and Orange County. Geographical demarcation is based on the location of fire risk zones surrounding the major population centers in Southern California. Given the facts summarized in the prior section, the topic of interest for this report is recruiting natural gas infrastructure to cooperate with electrical infrastructure to leverage its superior reliability. Specific to the issue of wildfires, it would be difficult to come up with a situation where natural gas pipeline infrastructure could be responsible for igniting a fire. Thus, gas infrastructure could be used as a backup distribution system when part of the electrical infrastructure either fails or needs to be disabled. To achieve this, we can consider powering fuel cells using H2 that is transported from points of production across California to points of use inside the Greater Los Angeles Area using existing gas transmission lines. In essence, the methodology of this study will be to analyze the behavior of the Southern California electrical grid under various operating conditions, and then evaluate the ability of the natural gas pipeline infrastructure to help in delivering adequate energy to all demand nodes using hydrogen. A major contribution of this work is the development of an electric transmission model for Southern California. Publicly available data sources were use to develop a highly detailed alternating current power flow representation of this system, including the magnitude and location of residential, commercial, and industrial electrical loads throughout Southern California. To perform this study, a model previously developed to simulate transmitting H2 using natural gas pipelines running on hydrogen was used [7]. This reference fully details the physics and execution of this gas pipeline model. The contents of this report focus on the different resiliency analyses performed during this work. A complete description of the data collection methods and model development is provided in the Appendix. Final Report University of California, Irvine 6 Clean Energy Institute 2 Key Assumptions The key assumptions made in this work are: • Existing natural gas transmission and distribution pipes have been repurposed to carry 100% H2. This implicitly assumes that: o California regulations have been amended to allow the injection and transmission of H2 into existing gas pipelines o Gas transmission pipeline components can safely transmit H2 o Gas transmission components that cannot safely or reliabily transmit H2 have been replaced o Utility customers can safely use H2 at their facilities/residences o There is a consistent and steady supply of renewable H2 that can be injected into pipelines that serve the Southern California Gas service territory o H2 can be produced and delivered to pipelines at a relatively low cost that would enable consumer use • Utility customers have the ability to shed load during a resiliency event. We do not resolve how load shedding occurs – if there is a technology (i.e., smart panel or smart appliance/load) that enables automatic load shedding, or if individual utility customers actuate their loads to reduce demand during a resiliency event (i.e., unplug appliances, turn off air conditioner, turn of lights). • Fuel cells are used as the H2 distributed generation installed across the Greater Los Angeles Area to improve system resiliency and reliability. • Natural gas power plants inside the Greater Los Angeles Area have been shut down. Specific scenarios assume that these power plants have been retrofitted to use H2 or have been replaced with H2 fuel cells. • Transportation electrification is not considered 3 Electric and Natural Gas Transmission Infrastructure This study is focused on infrastructure in the Southern California area – both electrical and natural gas. Maps showing the infrastructure that is considered are shown in Figure 2 and Figure 3 for the electrical and natural gas transmission systems respectively. Figure 2 also indicates major substations captured in the electrical transmission systems. These substations are located at points where the high voltage electrical transmission system interfaces with lower voltage transmission and distribution circuits. Electrical loads across the Southern California Edison territory are aggregated around these major substations indicated in Figure 2. The substations in Figure 2 are split between substations that are also connected to utility scale natural gas electricity generation, and substations that are not. Final Report University of California, Irvine 7 Clean Energy Institute Figure 2: Map of the Southern California electrical transmission grid with major substations. All substations serve electrical utility cusomters either directly or through other substations that are connected through lower voltage transimisison and distribution circuits not captured in this map. Substations are differentiated between substations that are connected to utility scale natural gas electricity generation, and substations that are not. Figure 3: Southern California gas transmission network Final Report University of California, Irvine 8 Clean Energy Institute 4 Resiliency Analysis To evaluate the resiliency of the electrical grid as it stands, we refer to Table 1, which states that resiliency can be partially quantified by a metric to describe the fragility/survivability of the electrical grid, specifically in unusual conditions. It has been demonstrated in practice that the grid is capable of operating when all lines are active, and the model reflects this. However, when subjected to a ‘disruptive event’ as referred to by IEEE in [1] in which normal operation of the grid is affected in ways that can lead to power outages, voltage instabilities, or system failures, the grid may be unable to respond depending on the magnitude of the event. Common types of disruptive events include natural disasters and extreme weather, cybersecurity and physical attacks, equipment failures, unexpected or excessive electric demand, and operational/human errors. The key result presented in this section is that the loss of major electric transmission lines has the potential to reduce the reliability of the Southern California electric grid, potentially leading to a resiliency event. Under current load conditions, additional demand response is absolutely necessary if other measures are not taken to ensure grid resiliency and reliability against wildfires and other PSPS causing events. 4.1 Wildfire/PSPS Event Zones In 2023, there were a total of 13 Public Safety Power Shutoff (PSPS) events. At least one PSPS event occurred in August of 2023. In Los Angeles, August saw the highest average temperature of all months in 2023 ([8]), meaning the cooling loads were likely higher than normal. Therefore, analysis of PSPS events coupled with the elevated August load profiles is most pertinent to the evaluation of the grid’s resiliency, as it provides results for a ‘worst-case’ scenario. Final Report University of California, Irvine 9 Clean Energy Institute Figure 4: PSPS Events in 2023 The DERiM resource as discussed in Appendix B provides a map of areas that have high fire risk. Transmission lines running through these areas are therefore candidates for shutoffs. Figure 5 shows a map of the fire hazard areas surrounding the LA basin, as well as the electrical infrastructure running through these areas. The fire areas were partitioned into 5 zones. Figure 6 shows these partitioned fire risk areas. All transmission lines in the model that run through these areas were grouped together based on zone. To simulate PSPS events, the lines running through each zone could be disabled depending on which zone was flagged to be a wildfire threat. While PSPS events are outages scheduled to prevent the ignition of a wildfire by transmission infrastructure, this zone analysis doubles as a potential scenario where a wildfire does occur in these zones, leading to the destruction of the corresponding transmission infrastructure. In analyzing the resiliency of the system to PSPS events, the model will be run with both individually disabled as well as combinations of disabled zones. Final Report University of California, Irvine 10 Clean Energy Institute Figure 5: Southern California Fire Hazard Areas Figure 6: Partitioned Fire Zones A variety of outage scenarios are considered in this work. These scenarios also consider different generation scenarios where the natural gas electrical generators located inside Los Angeles, Orange, Riverside, and San Bernadino county are kept operational using natural gas or converted to hydrogen. These different scenarios are listed in Table 2. This table also provides a high level summary on simulation results, indicating if electric transmission overloads occur in the simulation Final Report University of California, Irvine 11 Clean Energy Institute and if existing gas pipelines could carry the necessary quantity of hydrogen. Additional information in this table include: • Electric Load Definition: This describes the electric load scenario across Southern California. “Base” indicates electric loads based on current electricity use in Southern California. “Critical” indicates that a resiliency event is occuring and electric loads have been curtailed to “critical loads only”. “100% Electrified” indicates that most residential, commercial, and industrial natural gas loads have been electrified and have been added to the normal electric load defined in the “Base” scenario. • Generation Fuel: This indicates the type of fuel used to power the current natural gas generation located at substations as indicated with red dots in Figure 2. In the “H2” scenario, we assume that these generators can be repowered with hydrogen. • Disabled Zones: Zones refer to fire zones shown in Figure 6. If a zone is disabled, we assume that all major transmission lines running through this fire zone are deenergized. • Overloads: This indicates results from our simulations for the electric transmission system. A “Yes” indicates that the electrical transmission system experiences an overload somewhere in the system within this scenario. • Pipeline Success: This indicates if the existing natural gas infrastructure can safely handle the gas that is required to meet customer and power plant fuel demand. A “N/A” indicates that the scenario assumes natural gas is flowing through the pipelines. Since the current natural gas system can handle customer and power plant demand, these scenarios are assumed to be viable and are not simulated. A “Yes” indicates that natural gas in all pipelines has been replaced with hydrogen, and the pipelines can handle the hydrogen demand. A “No” indicates that natural gas in all pipelines has been replaced with hydrogen, and the current system does not have sufficient gas carrying capacity to supply the desired fuel energy flow rate. A “Partially” indicates that natural gas in all pipelines has been replaced with hydrogen, and that the gas pipeline system can handle the fuel transition through most hours of the day, but must be augmented with fuel storage near or at the customer to successfully meet demand. In other words, the gas pipeline system in the “Partially” case can handle the average gas demand, but cannot handle peak gas demand. Final Report University of California, Irvine 12 Clean Energy Institute Table 2: Definition of different generation and outage scenarios. The table also indicates if electric transmission system overloads occur, if the pipeline is capable of carrying the desired hydrogen, and the section that describes the scenario Scenario Index Electric Load Definition Generation Fuel Disabled Zones Overloads Pipeline Success Report Section 1 Base Base None No N/A N/A 2 Base Base 1 No N/A 4.2.1 3 Base Base 2 No N/A 4.2.2 4 Base Base 3 Yes N/A 4.2.3 5 Base Base 4 Yes N/A 4.2.4 6 Base Base 3,4 Yes N/A 4.2.5 7 Base Base 5 Yes N/A 4.2.6 8 Critical Base 3 Yes N/A 4.3.4 9 Critical Base 4 Yes N/A 4.3.5 10 Critical Base 5 No N/A 4.3.6 11 Base H2 None Yes N/A 5.2 12 Base H2 1 Yes Yes 5.7.1 13 Base H2 2 Yes Yes 5.7.2 14 Base H 2 1,2 Yes No 5.7.3 15 Base H2 3 Yes Yes 5.7.4 16 Base H 2 4 Yes Partially 5.7.5 17 Base H2 5 Yes Partially 5.7.6 18 100% Electrified H2 1,2 Yes No 6.1 4.2 Resiliency of the Existing Electrical Grid/Model Verification The base system, which includes August load and generation profiles, must first be assessed for resiliency to these PSPS events. First, each of the five previously specified zones will be disabled, leading to five simulations where a major transmission pathway is offline. After this, combinations of two disabled zones will be simulated. OpenDSS requires the specification of a “source bus”, which compensates for any net load in the system by providing importation or exportation as if the system is a microgrid connected to a larger grid. In all simulations, this source bus has been selected to be the easternmost bus in Palm Springs. This bus was chosen because of the implications of potentially placing renewable generation in the adjacent desert and evaluating the ability of the transmission corridors there to transport this generation to the basin. Each of the five established fire risk zones will be disabled individually and the resulting circuit will be evaluated based on performance. Final Report University of California, Irvine 13 Clean Energy Institute 4.2.1 Zone 1 – Lower Palm Springs Corridor Zone 1 is the lower Palm Springs corridor. Figure 7 shows that disabling it has no significant impact on the Greater Los Angeles area circuits, as there are no reported current overloads, nor voltage anomalies. This is a reasonable result, as the transmission lines in the Palm Springs corridor are robust, but not responsible for much power transmission within this configuration, as there is less than 2GW of generation in the area. The key driver of this result is that other electric transmission lines have sufficient current-carrying capacities to supply the electricity that would have otherwise been supplied through the lower Palm Springs corridor without inducing a line overload or experiencing unacceptable voltage drop throughout the electric transmission system. Figure 7: Base Loads & Generation Overload Plot (Zone 1 Disabled) Final Report University of California, Irvine 14 Clean Energy Institute 4.2.2 Zone 2 – Upper Palm Springs Corridor Figure 8 shows that disabling Zone 2, the upper transmission corridor in Palm Springs, leads to no overload or voltage anomaly reports. Figure 8: Base Loads & Generation Overload Plot (Zone 2 Disabled) Final Report University of California, Irvine 21 Clean Energy Institute 4.3.1 Residential Load Profile The ResStock residential load profile can be broken down into a set of end-uses, of which several could be considered critical. Below is a table of each end use, whether it is or is not considered critical in the analysis, and a rationale. Table 9: Residential End Uses END USE CRITICAL? REASONING CEILING FAN Y Some buildings may not be equipped with A/C, fan is last resort for cooling CLOTHES DRYER N Clothes can be hang-dried CLOTHES WASHER N Clothes can be washed in bathtub COOLING Y Livable indoor temperature is critical DISHWASHER N Dishes can be washed in sink FREEZER Y Storage of food is essential HEATING Y Temperature control is essential HOT TUB N Recreational item HOT WATER Y Necessary for washing EXTERIOR LIGHTING Y Exterior visibility at night promotes safety INTERIOR LIGHTING Y Visibility in home promotes safety GARAGE LIGHTING Y Same as interior lighting MECHANICAL VENTS Y May be important to be able to close vents in the case of wildfires affecting air quality PLUG LOADS 50% Internet connection, as well as some critical end uses, may be tied to this POOL HEATING N Recreational item RANGE OVEN Y Ability to cook food is essential REFRIGERATOR Y Storage of food is essential WELL PUMP Y Ability to pump water is essential 4.3.2 Commercial Load Profile As with ResStock’s breakdown, ComStock can be broken down into end uses. All end uses are assumed to be critical because their naming indicates they are all tied to either lighting, heating, cooling, or some other system whose importance of operability is ambiguous. 4.3.3 Adjusting Load Profiles To generate a critical load profile for each substation, the existing base load profile for that substation is multiplied by a factor at each hour. For each hour, the percentage of the total load deemed critical, based on the assertions made in sections 4.3.1 and 4.3.2, is close to 80%. This Final Report University of California, Irvine 22 Clean Energy Institute 20% reduction may be enough to alleviate some of the overloads experienced when disabling the zones. Figure 13 shows the resulting critical load profile for one of the substations. There is a consistent decrease in the load for each hour. Zones 3, 4 and 5 showed to cause problems when disabled in the full load case examined. They will again be tested under the reduced critical load. Figure 13: Base & Critical Load Profile Comparison 4.3.4 Zone 3 Figure 14: Zone 3 Performance Under Critical Loads, Overload Plot Table 10: Zone 3 Performance Under Critical Loads, Overload Report Final Report University of California, Irvine 23 Clean Energy Institute Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 5743 500 0 TRUE 3125 114 4 385 5741 500 0 FALSE 3125 114 4 385 5742 500 0 FALSE 3125 114 4 385 3622 220 8 FALSE 3894 112 3 414 3654 500 7 TRUE 3108 113 3 368 5744 220 10 TRUE 2497 108 2 177 Despite the reduction in load, the same lines are overloaded, although in comparison with the results for Zone 3 in 4.2.3, the overloads are reduced slightly in percentage and significantly in duration. 4.3.5 Zone 4 Figure 15: Zone 4 Performance Under Critical Loads, Overload Plot Table 11: Zone 4 Performance Under Critical Loads, Overload Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3597 500 76 FALSE 3112 114 6 372 Final Report University of California, Irvine 24 Clean Energy Institute 3598 500 76 FALSE 3110 114 6 370 3892 500 20 TRUE 4678 114 4 568 3893 500 2 TRUE 3134 114 4 394 3894 500 1 FALSE 6270 114 4 790 3889 500 26 FALSE 3106 113 4 366 3890 500 31 FALSE 3062 112 4 322 3891 500 26 FALSE 3120 114 4 380 As with the performance of Zone 3, there is a marginal reduction in the overload percentage and a significant decrease in the overload duration. 4.3.6 Zone 5 Figure 16: Zone 5 Performance Under Critical Loads, Overload Report Under critical loads, disabling Zone 5 no longer leads to overloads. 4.3.7 Conclusions on Critical Load Resiliency Despite the 20% reduction in loads, there are still significant overloads that occur when Zones 3 and 4 are disabled. To fully avoid overloads in such a scenario, the reduced electrical load would have to be decreased further beyond the critical residential and commercial loads outlined in Section 4.3.1, which could be a realistic strategy as 80% is a generous estimate of what portion of loads should be considered critical. Final Report University of California, Irvine 25 Clean Energy Institute 5 Pipeline Hydrogen as a Resiliency Augmentation California aims to transition to 100% renewable energy by 2045. A scenario of interest is therefore one where the non-renewable generation in the model is disabled entirely. In the model, this implies the removal of all natural gas generation. Can this lost generation be replaced by renewable generation sited in locations outside of the Greater Los Angeles Area? More specifically, will the electrical transmission infrastructure be able to support the added strain resulting from the removal of in-basin generation? If not, how can the natural gas infrastructure be leveraged to accomplish this? The analysis in this section differs from the prior because the prior section only considered if the current electrical transmission system can successfully handle transmission line outages. In many instances, the electric transmission system experiences overloads and voltage issues. This section introduces a remedy to transmission issues during a resiliency event where the gas transmission system is used to delver hydrogen to fuel cells and zero emission generators located across the Greater Los Angeles Area. 5.1 Combining Models To introduce hydrogen as a renewable energy transmission medium, the existing natural gas infrastructure could be leveraged to deliver hydrogen to key points in the electrical transmission grid. Since the state of California is pushing its constituents to transition from non-renewable to renewable energy, a most logical step is to replace all natural gas generation with hydrogen fuel cells, supplied with hydrogen through the pipeline network. The California State Geoportal, mentioned in 4, provides a geospatial dataset for the natural gas pipeline infrastructure in Southern California. Figure 17 shows the natural gas pipeline network overlaid on the electrical transmission network of the model. The yellow substations host all natural gas generation in the model. As can be seen, the natural gas pipelines run close to many of these substations. Therefore, to conduct an analysis on the resiliency of the system when hydrogen transmission is included, all natural gas plants will be removed from the model, and a subset of the yellow substations that run closest to natural gas pipelines will be equipped with fuel cells to meet the duty cycle of the now inactive natural gas plants. The pipeline model will be used to evaluate whether the pipelines can transmit the necessary volume of hydrogen to power these fuel cells. Only pipes 1, 2, 3, 4, and 5 in will be simulated as they cover the path from Palm Springs to a set of junctions in the basin that branch off into many other pipelines. Once the hydrogen is transported to these junctions, it is assumed that the pipeline infrastructure within the basin can distribute it. Final Report University of California, Irvine 26 Clean Energy Institute Figure 17: Southern California Gas Pipeline & Electrical Infrastructure 5.2 System Response to Removal of Non-Renewables To evaluate the ability to phase out non-renewable generation and replace it with hydrogen, it is important to determine the system response to the removal of natural gas generation. Natural gas is responsible for a significant portion of generation, thus its removal will have a major impact on the delicately balanced system. Final Report University of California, Irvine 27 Clean Energy Institute Figure 18: Removal of Natural Gas Generation Overload Plot (Base Loads) Table 12: Removal of Natural Gas Generation Overload Report (Base Loads) Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3568 500 67 FALSE 4235 155 7 1495 5698 500 66 FALSE 4288 156 7 1548 3653 500 65 FALSE 6011 146 6 1901 3569 220 69 FALSE 2464 106 2 144 3553 220 68 TRUE 2515 108 2 195 With the duress of the full loads as derived from CAISO historical data, removing in-basin natural gas plants puts significant stress on the Palm Springs transmission corridors when this missing generation is supplied by the source bus. Before running a more complicated approach, we can determine simply whether the pipelines that run parallel to these transmission lines are able to alleviate these overloads. 5.3 Leveraging Pipeline Hydrogen The overloaded lines are not constantly overloaded. Below is an example of the ampacity profile on one of the overloaded lines. Figure 19 shows that line 3568 is overloaded from hour 13.5 onwards, peaking at hour 20. Final Report University of California, Irvine 28 Clean Energy Institute Figure 19: Line 3568 Ampacity Plot To alleviate this overload from the line, we convert this amperage first to power, then to hydrogen. 𝑃𝑃=𝐼𝐼𝐿𝐿∙𝑉𝑉𝐿𝐿−𝐿𝐿 ∙√3 (1) 𝑚𝑚󰇗𝐻𝐻2= 𝑃𝑃∙𝑆𝑆𝑆𝑆 𝐿𝐿𝐿𝐿𝑉𝑉 𝐻𝐻2 ∙𝜂𝜂 𝐺𝐺𝐺𝐺𝐺𝐺 (2) Equation (1) converts single-phase current to three-phase power using the line-line (base) voltage. To obtain hydrogen mass flow rate, Equation (2) then takes this three-phase power and divides it by the lower heating value of hydrogen gas, as well as a conversion efficiency constant. It is then multiplied by a safety factor, to ensure that the line is not operating exactly at its operating capacity for those hours. These equations are applied to the profiles for each of the overloaded lines that run in parallel. For example, the lines from 339 to 347 in Figure 18 (3568, 5698, 3569, 3553 and 3554 in Figure 20) are included in this calculation, but the line from 347 to 367 is not, because this would be double counting the overload running through that branch. The lower heating value used is 120MJ/kg. A safety factor of 1.1 is assigned. A gas-to-power efficiency of 50% is assumed, based on known electrical efficiency data on PEMFCs ([9]). Therefore, the necessary power is converted to a theoretical hydrogen mass flow rate that would produce an equal power output when sent through a fuel cell. Final Report University of California, Irvine 29 Clean Energy Institute Figure 20: Overloaded Line Ampacities Figure 21 shows the resulting necessary hydrogen profile to alleviate the overloaded lines. This profile serves as a direct input into the pipeline model. Based on Figure 17, a simple network of a handful of pipes will be able to describe the pipeline system running in parallel with the overloaded lines. suggests the presence of pipes 1, 2, and 3 that meet at a compressor station, from where pipes 4 and 5 continue into the basin. It can be assumed that once the hydrogen has successfully passed through this set of pipes, the pipeline infrastructure in the basin, which hosts significantly more pipes as seen in Figure 17, can transmit the hydrogen to the different substations equipped with fuel cells. Assuming the ability to inject hydrogen into pipes 1, 2, and 3 in Palm Springs, we can use the MATLAB model discussed in section 1.3 coupled with corresponding pipeline parameters taken from to determine whether the pipeline network can transmit the hydrogen mass flow rate profile. Final Report University of California, Irvine 30 Clean Energy Institute Figure 21: Hydrogen Mass Flow Rate 5.3.1 Segment 1 – Palm Springs to Compressor Station The length of the pipeline segment between the Palm Springs injection point and the compressor station was assumed to be 100km, around 40% of the total length of the three pipelines according to . This assumption was made conservatively based on an observation of the distance between Palm Springs and the compressor station. Each of the three pipes can individually transmit the entire hydrogen mass flow rate, although the peak does approach the operational pipe pressure limit in pipes 1 and 3. These figures indicate the maximum pressure allowable in the pipe, the outlet pressure required at point of delivery, and the inlet pressure, or the pressure at which the gas in injected into the pipe. Final Report University of California, Irvine 37 Clean Energy Institute 5.5 Including Natural Gas Demand It cannot be assumed that the pipelines are completely empty and only need to transmit the hydrogen profile obtained in 5.4. While the need for natural gas in utility generation has been eliminated with the removal of natural gas plants in the basin, residential and commercial demand for natural gas still exists. ResStock & ComStock, as discussed in 4, have simulated natural gas demand profiles for both sectors on an average August day. These profiles are provided in equivalent kWh of energy, making their conversion to hydrogen straightforward. Since this data represents the entire state of California, it is divided by three to represent an estimate of the demand of the area of the study. These residential and commercial natural gas kWh profiles are summed up and converted to hydrogen, assuming a gas-to-power efficiency of 30%. This low efficiency is a conservative estimate of how the end uses normally serviced by natural gas would perform on a 100% hydrogen input. The resulting hydrogen profile is added on top of the previous profile, and the system is reassessed. Figure 30 shows that the residential and commercial natural gas equivalent hydrogen demand are relatively small compared to the required hydrogen output for the overloaded lines. Figure 30: Hydrogen Profiles After combining these profiles, they are fed through the pipeline system. Final Report University of California, Irvine 38 Clean Energy Institute Figure 31: System Performance Under Full Hydrogen Load The system can easily support the complete hydrogen mass flow rate profile that includes residential and commercial natural gas demand. In conclusion, the natural gas pipeline infrastructure would be able to support the necessary hydrogen transmission to power fuel cells to fully replace the natural gas generation in the Greater Los Angeles Ara. If Segment 2 can successfully transmit the hydrogen, this means that Segment 1 automatically can as well. Since Pipes 4 and 5 are assumed to have identical physical properties, we can assess the entire system performance from just one of these pipes alone. Thus, from here onwards, only Pipe 4 will be plotted for analysis purposes. 0 5 10 15 20 25 Time (h) 500 1000 Pressure (psi-g) Pipe 1 (Segment 1) Inlet pressure Outlet pressure Maximum pipe pressure 0 5 10 15 20 25 Time (h) 500 1000 Pressure (psi-g) Pipe 2 (Segment 1) 0 5 10 15 20 25 Time (h) 500 1000 Pressure (psi-g) Pipe 3 (Segment 1) 0 5 10 15 20 25 Time (h) 500 1000 Pressure (psi-g) Pipe 4 (Segment 2) 0 5 10 15 20 25 Time (h) 500 1000 Pressure (psi-g) Pipe 5 (Segment 2) Final Report University of California, Irvine 39 Clean Energy Institute 5.6 Fuel Cell Sizing Problem We have established that the pipeline infrastructure can carry the necessary hydrogen to the substations. However, to convert this hydrogen back into electricity to be injected into the grid, adequate fuel cell capacity is necessary. Based on the simulated hydrogen generation profile established for the base system in section 5.5, each of the substations requires 10s t0 1000s of MW fuel cell capacity. According to the Fuel Cell & Hydrogen Energy Association (FCHE, [10]), stationary fuel cell technology can produce around 10MW per acre of land. Assuming a linear relationship between power output and the area required, the substations would each require tens to hundreds of acres of space, land which is not readily available in the city. Furthermore, this is only considering the base circuit, not the added strain of disabling Zones, which would increase the required fuel cell power output even more. Instead of assuming the use of fuel cells, a near term solution would be to switch to H2 powered combined cycles using conventional gas turbine and steam turbine systems. Conventional combined cycles can operate at peak fuel to electricity conversion efficiencies exceeding 60%. Long term solutions could include higher efficiency fuel cell – gas turbine cycles where a high temperature fuel cell is thermally integrated with a gas turbine, using heat from the fuel cell process to replace combustion in a gas turbine. Theoretical analysis of this next generation power plant predict fuel to electricity conversion efficiencies in excess of 70% on a lower heating value basis ([11]). Several experimental fuel cell-gas turbine hybrids have successfully demonstrated efficiencies approaching and exceeding these theoretical values ([12]). For this work, we assume that power plants with a fuel to electricity conversion efficiency of 55% are used. This lower value is selected based on a) existing generators in the Greater Los Angeles Area that could be repowered using hydrogen, and b) to reflect that power plants operating on a electric grid with high penetration of intermittent renewable resources are regularly operated at part load where actual efficiency drops below rated efficiency. Final Report University of California, Irvine 40 Clean Energy Institute Based on this new efficiency assumption, the final hydrogen profile needed to supply the system after natural gas is completely removed is plotted below. Figure 32: Base System Hydrogen Profile As Pipe 4 can transmit its share of the hydrogen mass flow profile, we can conclude that the entire pipeline system does as well. 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 41 Clean Energy Institute 5.7 Resiliency of Resulting System Now that the pipeline has an active role in the energy transmission system, we can conduct the same analysis done in Section 4. Each of the five fire risk zones is disabled, the resulting overloads are plotted, and the question of whether the pipeline can support enough hydrogen to alleviate these overloads by supplying combined cycle hydrogen plants is answered, for both the full electrical loads as well as the critical loads established in 4.3. The results in this section consider the addition of H2 in the gas transmission system as a fuel that can be used in backup generation and other systems that can be used to send electricity back to the grid during an adverse electric transmission event. 5.7.1 Zone 1 – Pipeline Included Figure 33 shows overloads are experienced in the upper Palm Springs corridor under full electrical loads when zone 1 is disabled. In this situation, the pipelines running in parallel to the transmission lines are ideal for addressing the overloads. Taking the same approach as before, the amperage overloads on the transmission lines tabulated in Final Report University of California, Irvine 42 Clean Energy Institute Table 12 are converted to an equivalent hydrogen mass flow rate and then sent through the pipelines, on top of the mass flow rate established in section 5.5. By doing this, ideally the electrical overloads are addressed by the additional hydrogen mass flow into the basin. Figure 33 shows the circuit response to the removal of the lower transmission corridor, when there is no natural gas generation in the basin. Naturally, since that natural gas generation is replaced by renewable generation originating from bus 339, an electrical overload occurs in the other transmission corridor. Sections 5.7.1.1 and 5.7.1.2 show the ability of the pipeline to address this issue for full electrical loads and critical loads respectively. Figure 33: Zone 1 Disabled – System Overload Plot Final Report University of California, Irvine 43 Clean Energy Institute Table 13: Zone 1 Disabled – Pipeline-Included Electrical Performance Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3896 220 8 TRUE 2546 110 1 226 3876 220 7 FALSE 2705 117 1 385 3555 220 25 TRUE 2949 127 2 629 3556 220 25 TRUE 2913 126 2 593 3873 220 24 FALSE 3010 130 2 690 3569 220 69 FALSE 3494 151 2 1174 3553 220 68 TRUE 3565 154 2 1245 3554 220 72 TRUE 3353 145 2 1033 Final Report University of California, Irvine 44 Clean Energy Institute 5.7.1.1 Disabling Zone 1 Under Full Electrical Loads First, the system is evaluated based on its ability to transmit adequate hydrogen to meet full electrical loads. Figure 34 shows the resulting necessary hydrogen mass flow rate that eliminates the overloads in the circuit that arise when Zone 1 is disabled, under full electrical loads. Logically, the removal of transmission infrastructure running parallel to the pipelines has increased the involvement of the pipelines in energy transmission. This total profile will be sent through the pipeline system. Figure 34: Hydrogen Mass Flow Rate for Full Load Zone 1 Figure 35 shows the performance of the pipeline system on the necessary hydrogen mass flow rate under full electrical loads. Since Pipe 4 can transmit its portion of the hydrogen mass flow, Pipe 5 is also capable, and therefore Pipes 1, 2 and 3 are as well. Final Report University of California, Irvine 45 Clean Energy Institute Figure 35: Pipeline Performance Report, Full Loads, Zone 1 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 46 Clean Energy Institute 5.7.1.2 Disabling Zone 1 Under Critical Loads Since the pipelines can support hydrogen under full electrical loads, the ability to service critical loads is implied. As Figure 36 shows, under critical loads the pipeline is even more capable of transmitting the necessary hydrogen to the basin. Figure 36: Pipeline Performance Report, Critical Loads, Zone 1 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 53 Clean Energy Institute Figure 43: Pipeline Performance Report, Full Loads, Zone 3 Again, as the pipeline system can support the overloads created by the full electrical loads in this case, it is inferred that critical loads can also easily be met using this approach. 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 54 Clean Energy Institute 5.7.5 Zone 4 – Pipeline Included Figure 44: Zone 4 Disabled – System Overload Plot Table 16: Zone 4 Disabled – Pipeline-Included Electrical Performance Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3622 220 8 FALSE 5081 146 9 1601 5744 220 10 TRUE 3256 140 9 936 304 138 3 TRUE 2515 108 4 195 3597 500 76 FALSE 3252 119 6 512 3598 500 76 FALSE 3250 119 6 510 5698 500 66 FALSE 2877 105 2 137 Final Report University of California, Irvine 55 Clean Energy Institute The circuit experiences some electrical overloads downstream of the main pipeline system we are modelling. These overloads are not easily addressed by adding more hydrogen, as this additional hydrogen has no effect on transmission infrastructure that does not run in parallel to the pipeline system. After applying the algorithm to determine the necessary hydrogen flow rate, the circuit is reassessed. Figure 45: Zone 4 Disabled – Pipeline Hydrogen Applied Table 17: Zone 4 Disabled – Pipeline Hydrogen Applied, Overload Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3597 500 76 FALSE 3186 116 7 446 3598 500 76 FALSE 3184 116 7 444 3892 500 20 TRUE 4408 107 3 298 3893 500 2 TRUE 2953 108 3 213 3894 500 1 FALSE 5909 108 3 429 3889 500 26 FALSE 2930 107 3 190 3890 500 31 FALSE 2889 105 3 149 3891 500 26 FALSE 2944 107 3 204 Final Report University of California, Irvine 56 Clean Energy Institute Figure 46: Pipeline Performance Report, Full Loads, Zone 4 With the inclusion of hydrogen, the overloads have shifted from where they were occurring in Figure 44 to a single transmission branch. These overloads occur because there are loads in the north-western part of the area where there are no generators. The hydrogen compensation algorithm is unable to converge on a solution where there are no overloads in the system, meaning the result of the analysis of this particular case should be considered inconclusive. We suspect that the pipeline system can address these overloads since the performance report (Figure 46) shows the system is below capacity. Most likely, a limitation in the model is preventing convergence. 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 57 Clean Energy Institute 5.7.6 Zone 5 – Pipeline Included Figure 47: Zone 5 Disabled – Pipeline-Included System Plot Table 18: Zone 5 Disabled – Pipeline-Included Electrical Performance Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3880 220 8 TRUE 4143 179 6 1823 3941 220 8 FALSE 4135 178 6 1815 304 138 3 TRUE 2936 127 6 616 3569 220 69 FALSE 2594 112 2 274 3553 220 68 TRUE 2647 114 2 327 3554 220 72 TRUE 2490 107 2 170 3599 220 11 TRUE 2713 117 3 393 3600 220 11 TRUE 2768 119 3 448 Disabling Zone 5 leads to major overloads both in the upper Palm Springs corridor as well as downstream within the basin. By including pipeline hydrogen, all these overloads are addressed, except a new overload is created in the transmission infrastructure in Irvine: Final Report University of California, Irvine 58 Clean Energy Institute Figure 48: Zone 5 Disabled – Pipeline Hydrogen Applied Table 19: Zone 5 Disabled – Pipeline-Included Electrical Performance Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3691 220 13 TRUE 2549 110 3 229 Final Report University of California, Irvine 59 Clean Energy Institute Figure 49: Pipeline Performance Report, Full Loads, Zone 5 Since the transmission lines between station 2133 and 381 do not run parallel to the pipeline, and since the only other electrical path to station 381 is cut off when Zone 5 is disabled, these lines become overloaded and cannot be helped by the pipeline infrastructure. The model does not account for generation south of the area, so in the real system it is likely that these overloads would not occur because power can still be transmitted to station 381 from the South. 0 5 10 15 20 25 Time (h) 200 300 400 500 600 700 800 900 1000 Pressure (psi-g) Pipe 4 (Segment 2) Inlet pressure Outlet pressure Maximum pipe pressure Final Report University of California, Irvine 60 Clean Energy Institute 5.7.7 Conclusions on Hydrogen-Reinforced System Resiliency When comparing the performance of the existing base system (section 4.2) to this section, the inclusion of the pipeline network significantly increases system resiliency. Most importantly, it achieves this while simultaneously eliminating the need for natural gas based electricity generation in the basin. Without disabling any zones, the pipeline network can easily support the base system on full electrical loads. Disabling Zones 1 or 2 leads to overloads in the upper and lower Palm Springs corridors respectively, and we have demonstrated that these overloads can be fully alleviated by the pipeline network in sections 5.7.1 and 5.7.2. All loads can be met by the pipeline system if Zone 3 is disabled. Disabling Zones 1 and 2 simultaneously, which is a realistic scenario as they are physically connected, would require the pipeline network to transport hydrogen exceeding its capacity. Specifically, Pipes 4 and 5 would not be able to transport adequate hydrogen to address the system’s electrical needs. An additional pipeline identical to Pipes 4 and 5 would need to be built. Disabling Zones 4 and 5 both lead to a system configuration where the hydrogen compensation algorithm does not converge on a solution where the grid has no electrical overloads. An AC power flow model is highly dynamic, and certain results are difficult to properly diagnose, so it is likely that some limitation in the way the model is set up is preventing this algorithm from converging. However, we are confident that the existing pipeline infrastructure would be capable of addressing the overloads seen when Zones 4 and 5 are disabled, based on the inspection of the performance reports. 6 Performance Under Complete Electrification of End-Uses The study performed in section 5 was conducted on the assumption that the electrical load profile taken from CAISO historical data accurately represents the complete energy demand in the area. However, these loads are purely electrical and do not account for alternative energy sources, primarily natural gas. To account for natural gas, section 5.5 converted its demand to a hydrogen demand profile, but the flaw with this approach is that it assumes that all natural gas end uses can be retrofitted to use hydrogen. If we instead assume that all natural gas end uses are converted to electrical loads, we can again run an analysis determining whether the pipeline network can support the additional loads. ResStock and ComStock provide complete electrification load profiles, ranging from lowto high-efficiency, whole-home electrification packages. This is only available at the state level, so we must leverage our other data to increase the resolution and assign new, complete electrification load profiles to the substations. To achieve this, we can determine the hourly ratio between the baseline electrical load and the complete electrification package load, then apply this hourly ratio to each of the substations’ current electrical load profiles, like how we determined critical load profiles in section 4.3. Final Report University of California, Irvine 61 Clean Energy Institute Figure 50: Whole-Home Electrification Package, Min Efficiency – Residential The upgrade package we are using assumes whole-home electrification at minimum efficiency. This was chosen to determine the system performance in the most conservative efficiency case. As an equivalent package is not provided for the commercial end uses in ComStock, this residential package will be used exclusively to adjust all loads. Figure 51: Base & Whole-Home Electrification Load Comparison Figure 51 demonstrates the increased electrical loads because of the whole-home electrification package. However, because we are now accounting for the residential and commercial natural gas end uses in the electrical profiles, we can eliminate the natural gas demand from the pipeline system. Final Report University of California, Irvine 62 Clean Energy Institute 6.1 Resiliency Analysis Figure 52 shows the increase in hydrogen transmission necessary due to the conversion to wholehome electrification. While natural gas end uses were converted to electrical demand and this decreased the hydrogen profile between hours 2 and 8 (when the electrical grid experiences no overloads), for the rest of the day the resulting electrical overloads demand more hydrogen transmission overall. Figure 52: Hydrogen Profile Comparisons Final Report University of California, Irvine 69 Clean Energy Institute 7.2 Hydrogen-Augmented System Resiliency Analysis Results For the few cases where the hydrogen compensation algorithm was unable to converge on a solution where none of the electrical transmission lines were overloaded, it is beneficial to determine the potential costs of upgrading these specific transmission lines as an alternative. To reiterate, in this case, it is assumed that all natural gas power plants in the LA basin are repurposed to use hydrogen, making it a highly renewable case. 7.2.1 Hydrogen-Augmented System – Zone 3 In the hydrogen-augmented system, the pipeline completely addresses all electrical overloads. Thus, no upgrade is necessary. 7.2.2 Hydrogen-Augmented System – Zone 4 Table 24: Zone 4 system details Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) Additional Required Capacity Per Phase (Amps) 3597 500 76 FALSE 3186 116 7 446 3598 500 76 FALSE 3184 116 7 444 3892 500 20 TRUE 4408 107 3 298 3893 500 2 TRUE 2953 108 3 213 3894 500 1 FALSE 5909 108 3 429 3889 500 26 FALSE 2930 107 3 190 3890 500 31 FALSE 2889 105 3 149 3891 500 26 FALSE 2944 107 3 204 7.3 Hydrogen-Augmented System Additional Pipeline Cost In one extreme case, part of the pipeline system was unable to handle the hydrogen transmission load. Introducing a new pipeline addressed this problem. Thus, it is important to determine the possible costs of building such a new pipeline, and to compare these costs to that of electrical transmission reconductoring discussed in section 7.2. 8 Conclusion This report analyzed the resiliency and reliability of the Southern California energy infrastructure, focusing on the potential of hydrogen augmentation through natural gas pipelines to enhance Final Report University of California, Irvine 70 Clean Energy Institute system performance under stress. Below are the key findings and quantified results from the analysis: • Baseline System Resiliency: o The Southern California grid experienced overloads exceeding 120% of operational limits during simulated outages in major transmission zones. These occurred for durations ranging from 2 to 9 hours. o Upgrading the electrical infrastructure to handle these overloads would cost an estimated $66 million, assuming installation costs are ten times the conductor costs. • Hydrogen-Augmented System Performance: o Inclusion of hydrogen transmission via pipelines significantly improved system resiliency. o Hydrogen pipelines effectively mitigated overloads in Zones 1, 2, and 3 during full load scenarios, without exceeding operational pipeline limits. o When Zones 1 and 2 were simultaneously disabled, the existing pipelines were insufficient. Adding one new 36-inch pipeline would enable hydrogen transmission to meet system demand at an estimated capacity increase of 45%. • Critical Load Management: o With loads reduced by 20%, most overloads were alleviated, demonstrating the system's ability to prioritize essential services during outages. o Under full electrification scenarios, pipeline limits were approached but not exceeded for single-zone outages. However, simultaneous outages of Zones 1 and 2 required additional pipeline infrastructure. • Future Outlook: o Transitioning natural gas pipelines to hydrogen transport offers a dual benefit: reducing dependency on fossil fuels and enhancing grid resiliency. o However, the complete electrification of end-uses and transportation would increase electrical loads by up to 70%, necessitating significant infrastructure investments in either hydrogen or electrical systems. This analysis supports the feasibility and economic advantage of using hydrogen infrastructure as a resiliency measure for critical urban energy systems. Further integration of renewable hydrogen and strategic pipeline expansion is recommended to future-proof the Southern California energy grid against increasing demand and extreme events. Final Report University of California, Irvine 71 Clean Energy Institute 9 References [1] Stankovic A, Tomsovic K, Mili ( L, Panteli M, Kavicky J, Thomas K, et al. Comments on the Definition and Quantification of Resilience IEEE Task Force on Definition and Quantification of Resilience. 2018. [2] Chi Y, Xu Y, Hu C, Feng S. A State-of-the-Art Literature Survey of Power Distribution System Resilience Assessment. 2018 IEEE Power & Energy Society General Meeting (PESGM), 2018, p. 1–5. https://doi.org/10.1109/PESGM.2018.8586495. [3] Liss W, Rowley P. Assessment of natural gas and electric distribution service reliability. Gas Technology Institute 2018. [4] Williams AP, Abatzoglou JT, Gershunov A, Guzman-Morales J, Bishop DA, Balch JK, et al. Observed Impacts of Anthropogenic Climate Change on Wildfire in California. Earths Future 2019;7:892–910. https://doi.org/https://doi.org/10.1029/2019EF001210. [5] CPUC. Public Safety Power Shutoffs. Https://WwwCpuc.caGov/Psps/ n.d. [6] US Department of Energy. California Energy Sector Risk Profile 2015. [7] Heydarzadeh Z, McVay D, Flores R, Thai C, Brouwer J. Dynamic modeling of california grid-scale hydrogen energy storage. ECS Trans 2018;86:245. [8] Los Angeles Almanac. High/Low & Average Temperatures By Month & Year Downtown Los Angeles. Https://WwwLaalmanacCom/Weather/We04aPhp n.d. [9] Lohse-Busch H, Stutenberg K, Duoba M, Iliev S. Technology assessment of a fuel cell vehicle: 2017 Toyota Mirai. No ANL/ESD-18/12 Argonne National Lab 2017. [10] Fuel Cell & Hydrogen Energy Association. Stationary Power. Https://WwwFcheaOrg/Stationary n.d. [11] D.F. Chuahy F, Kokjohn SL. Solid oxide fuel cell and advanced combustion engine combined cycle: A pathway to 70% electrical efficiency. Appl Energy 2019;235:391–408. https://doi.org/https://doi.org/10.1016/j.apenergy.2018.10.132. [12] van Biert L, Woudstra T, Godjevac M, Visser K, Aravind P V. A thermodynamic comparison of solid oxide fuel cell-combined cycles. J Power Sources 2018;397:382–96. https://doi.org/https://doi.org/10.1016/j.jpowsour.2018.07.035. [13] U.S. Energy Information Administration. California State Energy Profile. Https://WwwEiaGov/State/PrintPhp?Sid=CA 2023. [14] Pro Wire & Cable. ACSR Conductor Pricing. Https://ProwireandcableCom/BuildingWire/Aluminum/Acsr/?Srsltid=AfmBOorG4GUtKmR6qVwCLqCeSBz6EFSdKOb9_vJHQiBe85U9Uu5uPoL n n.d. [15] California Energy Commission. California Energy Storage Systems Survey. Https://WwwEnergy.caGov/Data-Reports/Energy-Almanac/California-Electricity-Data/CaliforniaEnergy-Storage-System-Survey 2024. [16] LaForest JJ, General Electric Co. PMALTDiv; GECo, SNYEUS. Transmission Line Reference Book - 345kV and above. EPRI, NTRL; 1981. [17] Indulkar CS, Ramalingam K. Estimation of transmission line parameters from measurements. International Journal of Electrical Power & Energy Systems 2008;30:337–42. https://doi.org/https://doi.org/10.1016/j.ijepes.2007.08.003. [18] Dixon J, Moran L, Rodriguez J, Domke R. Reactive Power Compensation Technologies: State-of-the-Art Review. Proceedings of the IEEE 2005;93:2144–64. https://doi.org/10.1109/JPROC.2005.859937. [19] Aspen Environmental Group, CPUC. Transmission Structures. Https://IaCpuc.caGov/Environment/Info/Aspen/Cltp/Archive/Files_8_26_14/_4TransmissionStructures FactSheetPdf n.d. Final Report University of California, Irvine 72 Clean Energy Institute [20] Dwight HB, Farmer EB. Double Conductors for Transmission Lines. Transactions of the American Institute of Electrical Engineers 1932;51:803–8. https://doi.org/10.1109/T-AIEE.1932.5056167. Appendix A – Data Collection for Electric Grid Model Geospatial Data Data that describes the location and specifications of important infrastructure was necessary for the development of a model of the electrical and natural gas grids. California Electrical Transmission Lines The “California Electric Transmission Lines” dataset provided by the California Energy Commission contains all electrical transmission lines in the state of California and tabulates a set of features for each line. Among these features, those listed below are most significant to our model: • Line ID number • Line coordinates (all segments) • Line voltage rating (in kV) • Utility that owns the line • Operational status • Whether the line is a single or double circuit • Line length (in miles & feet) Figure 55 shows a subsection of the dataset in our area of interest: the LA basin and its surrounding regions. Final Report University of California, Irvine 73 Clean Energy Institute Figure 55: California Electrical Transmission Lines geospatial map Southern California Edison Distributed Energy Resource Interconnection Map Substations are key components of the electrical grid. From a modelling standpoint, they are nodes, interconnected by transmission lines. An area’s power generation, as well as load profiles pertaining to the area’s population, can be assigned to these nodes. Southern California Edison has a publicly available geospatial dataset, known as the Distributed Energy Resource Interconnection Map (DERiM). This dataset contains the details describing all substations in the Southern California area, including those owned by other utilities. Like the transmission line data set, this data set provides a list of features for each substation: • Substation ID • Substation coordinates • “System” to which the substation belongs • DER existing generation (in MW) • Projected load (in MW) • Maximum and minimum loads experienced during every hour of every day in a year Figure 56 shows the plotted substation data taken from DERiM. Final Report University of California, Irvine 74 Clean Energy Institute Figure 56: DERiM substation geospatial map Accompanying this geospatial data is a set of comma-separated values describing minimum and maximum experienced loads (in MW) for every substation. Every month is described by a minimum and maximum 24-hour load profile, where the values are based on the minimum and maximum loads experienced at each hour in the entire month. For example, if January 3rd experienced the lowest load at 13:00, while January 20th experienced the highest load at 13:00, they would be selected to describe January’s minimum and maximum loads at 13:00 respectively. While this data is not physical as it is an aggregation of data collected over month-long periods rather than describing real individual days, it will be useful in scaling and assigning real, physical data to the nodes. By combining the transmission line dataset with this substation dataset, it will be possible to build an A/C power flow model that treats the substations as nodes, and the transmission lines as connections between these nodes. California Power Plants Alongside geospatial data for the physical infrastructure that constitutes the electrical grid, our model necessitates geospatial data describing the magnitude, location, and type of power generation in the system. The California State Geoportal provides a dataset of all power plants in California, with the following set of features describing each: • California Energy Commission Plant ID • Plant Name • State of retirement (0 or 1) Final Report University of California, Irvine 75 Clean Energy Institute • Operator Company • County • Capacity (in MW) • Units • Primary Energy Source (SUN, NG, WIND…) While this dataset does not contain information regarding generation profiles for any of these plants, the capacity and location data will be useful in scaling and assigning such profiles to substations in their vicinities. Figure 57Error! Reference source not found. shows a plot of this dataset, again in our area of interest. Figure 57: Power Plants Note that this dataset describes utility-scale generation and does not account for DER resources such as rooftop solar, meaning that the DER data collected from SCE’s DERiM is independent from this dataset. Therefore, we can be confident that no generation will be double counted in the model. Final Report University of California, Irvine 76 Clean Energy Institute Load & Generation Profile Data California Independent System Operator To establish a working model that accurately describes the real transmission system in Southern California, real demand and generation data is required. The California Independent System Operator (CAISO) is a powerful resource that provides information on the operation of California’s electric power transmission. In its library, it contains electrical demand and generation data for every day since April 10th, 2018. The generation data is broken down by resource (renewables, natural gas, hydroelectric, etc.) in 5 minute increments, and the renewable generation is further broken down by renewable category (solar, wind, geothermal, etc.). Figure 58 and Figure 59 show this generation data as it is presented on the CAISO website, for August 8th, 2023. Figure 58: CAISO Overall Generation Profiles Figure 59: CAISO Renewable Generation Profiles Final Report University of California, Irvine 77 Clean Energy Institute This data represents the total statewide generation but cannot be separated based on utilities. It is therefore not possible to know how much of the total generation is provided by SCE and LADWP, the utilities of interest in this study. Load profile data is provided by the California ISO via its Open Access Same-time Information System (OASIS) on an hourly basis, broken down based on utilities. As a result, we can obtain a load profile that adequately represents our area of interest (served by most of SCE and LADWP). The limitation of this CAISO data is the fact that it is aggregated data (statewide for generation, utility-based for loads). This limitation can be compensated for using certain properties of the geospatial data. Residential & Commercial Building Stock The National Renewable Energy Laboratory provides building simulation data in two categories of building stock – residential (ResStock) and commercial (ComStock). Both categories contain data at the state level, which are partitioned into end uses. ‘Upgrade packages’, which simulate the predicted electrical loads of building stock that is partially or entirely electrified, are also available. Monthly data can be downloaded, where the user can obtain a 24-hr profile in units of kWh. Figure 60: ResStock aggregate load profile Final Report University of California, Irvine 78 Clean Energy Institute Figure 61:ComStock aggregate load profile Final Report University of California, Irvine 85 Clean Energy Institute the same approach as taken in assigning load profiles to the substations in 0, each substation is assigned a set of power plants based on their proximity. Figure 67 shows all power generation plants that are included in the model and assigned to their nearest substations. With a generation capacity value for every type of generation attached to each substation, the generation profiles can be divided amongst the substations accordingly. For example, if there is a total of 20GW of natural gas generation in the region, and a substation has a natural gas generation capacity value of 2GW, it will be responsible for 10% of the total natural gas generation profile. The total generation capacity for each generation type was found and tabulated. The generation profiles provided by CAISO in 0 are not partitioned by utility, but the load profiles provided by CAISO OASIS are. By taking the ratio between the SCE + LADWP loads and the statewide loads at each hour, we arrive at an array of ratios at each hour that can be multiplied elementwise by the generation profiles to obtain an approximation of the local generation profiles. Figure 67: Generation Plant Map Final Report University of California, Irvine 86 Clean Energy Institute 𝛼𝛼 = (𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 � � � � � � � � � �  𝑆𝑆𝑆𝑆𝑆𝑆 +𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 � � � � � � � � � �  𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐺𝐺)⊘𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 � � � � � � � � � �  𝑆𝑆𝐿𝐿 (4) 𝐺𝐺𝐺𝐺𝐺𝐺 � � � � � � �  𝑆𝑆𝑆𝑆𝑆𝑆+𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐺𝐺 = 𝛼𝛼⊙𝐺𝐺𝐺𝐺𝐺𝐺 � � � � � � �  𝑆𝑆𝐿𝐿 (7) In observation of Figure 68, notice that utility solar, the solar generation provided by the CAISO profile, has been scaled such that its maximum output does not exceed the solar generation capacity. Wind, hydroelectric, geothermal and imported profiles are untouched. Rooftop solar generation is the generation provided by the DERiM resource, which provides existing customerside solar generation capacity for each substation. This profile was constructed using these capacity numbers coupled with a solar generation profile for the Los Angeles area in August, downloaded from the NREL PVWatts tool. Two High-Voltage DC transmission lines known as Path 27 and Path 65 are responsible for a significant amount of imported power into the basin. With capacities of 2.4GW and 3.1GW respectively, they are more than capable of transmitting the importation profile seen in Figure 68, and thus will be inserted into the model for the purpose of supplying all importation to the system. Note that they are single-phase HVDC lines that will be treated as three-phase AC lines with the same capacity to avoid adding complexity to the model. The imports profile was adjusted such that there is constant importation of power at 50% of the maximum capacity of these two HVDC lines, a conservative estimate for simplicity and to avoid needing to ramp up the natural gas duty cycle to unreasonable levels. Figure 69 displays this adjusted import generation profile. Figure 68: August 8th Load & Generation Profiles Final Report University of California, Irvine 87 Clean Energy Institute Figure 69: Adjusted Load & Generation Profiles With a total generation profile for each type of generation established for the area covered by the model, this generation can be divided among the substations using the generation capacity values assigned to each substation. On Energy Storage This model of a base case will not include battery energy storage due to the sizing and placement of the existing storage in the state. According to the California Energy Storage System Survey ([15]), the state has a total energy storage capacity of 10.383GW. Most of this capacity is owned by utility companies. Because the generation profiles were taken directly from CAISO, the dynamics of storage and dispatch are already integrated. As such, for the base case, which aims only to simulate a real day, formally including battery energy storage as a generation profile was deemed unnecessary. Transmission Line Parameters The parameters governing transmission line dynamics in OpenDSS, namely operating ampacity, resistance, capacitance and resistance, are all defined in a ‘linecode’. A linecode needs to be defined for each type of transmission line included in the model – 500kV, 220kV and other high-voltage lines are built using differently sized conductors and thus will not have the same electrical properties. Southern California Edison declined to provide details on the conductors used in their transmission lines, and therefore some estimates must be made to proceed. To make an initial estimate on the physical parameters of the transmission lines in the circuit, transmission line conductor data was necessary. Aluminum Conductor, Steel Reinforced (ACSR) conductors are the standard conductors used in transmission line infrastructure. In the Transmission Line Reference Book [16] published by EPRI, a table containing critical information Final Report University of California, Irvine 88 Clean Energy Institute on ACSR conductors is provided. From this table, which contains dozens of different conductors, a set of conductors were selected to be assigned to the different transmission lines in the model. These conductors were chosen based on the assumption that higher-voltage lines will have higher current-carrying capacities, though this is not verifiable. The values shown in the table apply to individual conductors, of which there are three in a three-phase transmission line. Some transmission lines carry double circuits, meaning there are two separate three-phase circuits with a conductor each, a total of six conductors. The transmission line dataset discussed in 0 accounts for this, as it states for every line in the dataset whether that line is a single or double circuit. Double-circuit transmission lines will be treated as two identical single-circuit transmission lines in parallel. OpenDSS linecode requires capacitance specifications which are not included in the table taken from [16]. Instead, physical measurements of capacitance for a 220kV line were taken from [17], then applied to the other voltages simply using numerical ratios between the voltages to increase or decrease the capacitance. With this conductor data and the knowledge of the number of conductors in every transmission line, we have a first working model of the electrical transmission system and can begin to adjust parameters according to whether the model behaves as expected. Upon resolving a circuit, OpenDSS provides analytics that allow the user to determine whether the circuit is operating correctly. Among these analytics, the per-unit (pu) voltage system and the ampacity overload spreadsheet are the most indicative of issues. The p.u. voltage system shows whether the circuit components are operating at their specified voltages, without significant underor over-voltages. Ideally, the p.u. voltage is equal to 1 throughout the entire circuit, although in practice the agreed-upon acceptable range is from 0.95 to 1.05, a 5% error. Thus, if a transmission line is rated to operate at 500kV, but drops to below 475kV, this is considered an undervoltage. Overand under-voltages are caused by significant events. A lightning strike hitting a transformer would be one example of a cause for an over-voltage, while an under-voltage can occur when the demand for power in a circuit exceeds the rated power of a component in the circuit, leading to a voltage drop. In the base case circuit, it is assumed that such events do not occur, and that the circuit is fully capable of transmitting the power generated and demanded. Thus, overand under-voltages indicate that transmission line parameters are not properly specified. Ampacity overloads occur when lines in the circuit are required to transmit currents that exceed their operating capacities. Overloads lead to the degradation of the transmission infrastructure and should be avoided. Lines operating at over 105% of their capacity will be overloaded. OpenDSS provides an overload report on the circuit, showing the overloaded lines, their ampacities, and the hours in which these overloads occurred. Using these metrics to describe the system’s performance, a first run of the model can be conducted and evaluated. Final Report University of California, Irvine 89 Clean Energy Institute The chosen parameters for the transmission lines do not lead to what would be considered an acceptable voltage profile in practice. The two red lines in Figure 70 mark the acceptable p.u. voltage range (0.95-1.05), and the blue lines denote the transmission lines in the model, starting at Palm Springs on the left where the source bus is situated (note that the lines move from left to right on the graph, while on the map they are really moving from East to West). From inspection, Figure 70 shows a significant voltage drop across the power lines. Final Report University of California, Irvine 90 Clean Energy Institute Table 26 shows a minimum undervoltage of 0.63 pu, with a total of 19 undervoltage buses. This is far below the acceptable minimum. Figure 70: Base Case Voltage Profile Figure 71: Base Case Overload Plot Final Report University of California, Irvine 91 Clean Energy Institute Table 25: Base Case Voltage Report Maximum Overvoltage Maximum Hour Number of Overvoltages at Max Hour Minimum Undervoltage Minimum Hour Number of Undervoltages at Min Hour 1.006711 5 0 0.619941 19 249 Table 26: Base Case Overload Report Final Report University of California, Irvine 92 Clean Energy Institute Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) 3564 500 25 FALSE 1486 108 2 3694 220 10 TRUE 1740 150 3 3622 220 8 FALSE 2941 254 12 3685 220 16 TRUE 2116 182 3 3664 220 16 TRUE 2099 181 3 3892 500 20 TRUE 2671 195 1 3893 500 2 TRUE 1782 130 1 3894 500 1 FALSE 3564 260 1 3895 500 12 FALSE 2190 160 2 3703 220 9 TRUE 3112 268 3 3896 220 8 TRUE 1459 126 2 3876 220 7 FALSE 1550 134 2 297 500 10 FALSE 1510 110 1 298 500 11 FALSE 1500 109 1 3555 220 25 TRUE 1425 123 3 3556 220 25 TRUE 1408 121 3 3873 220 24 FALSE 1455 125 3 3903 220 6 TRUE 1530 132 3 3699 220 6 TRUE 1516 131 3 3569 220 69 FALSE 1420 122 4 3553 220 68 TRUE 1450 125 4 3889 500 26 FALSE 1775 130 1 3890 500 31 FALSE 1776 130 1 3891 500 26 FALSE 1783 130 1 3653 500 65 FALSE 3440 251 3 3916 220 37 TRUE 1414 122 1 3696 220 5 TRUE 1376 119 3 3568 500 67 FALSE 1785 130 6 5698 500 66 FALSE 1809 132 6 3554 220 72 TRUE 1362 117 3 3691 220 13 TRUE 1339 115 2 3697 220 5 TRUE 1335 115 2 3860 220 1 TRUE 1278 110 1 Final Report University of California, Irvine 93 Clean Energy Institute Table 27: Base Case Voltage Report Maximum Overvoltage Maximum Hour Number of Overvoltages at Max Hour Minimum Undervoltage Minimum Hour Number of Undervoltages at Min Hour 1.006711 5 0 0.619941 19 249 The parameter that has the most impact on this voltage drop was determined through trial and error to be the reactance in the linecodes. Despite the assumption that the conductor reactance specifications provided by the ACSR data table in [16] are accurate, it was found that the model could only produce an acceptable voltage profile when the reactance was reduced by an order of magnitude. While no source in the literature uses line reactance values below those in the ACSR table, there is no other simple way to produce acceptable voltage profiles in the model than by reducing the reactance. An explanation for this discrepancy between the real system and the model is the lack of reactive power compensation in the model. According to a review of Reactive Power Compensation published in IEEE [18], reactive power compensation is used to “[increase] the maximum active power that can be transmitted. It also helps to maintain a substantially flat voltage profile at all levels of power transmission” ([18], page 1). One method used to achieve reactive power compensation is known as ‘Series Compensation’, where capacitors are installed in series to “decrease the equivalent reactance of a power line at a rated frequency” ([18], page 2 section B). Thus, the direct manual reduction of reactance leads to the same result as Series Compensation (being reduced reactance along the power lines). Ultimately, the need for reactive power compensation in the real transmission circuit arises from the fact that the transmission line reactance is a fixed physical parameter. It is therefore essential as a physical way to reduce equivalent reactance and avoid voltage drops. However, in the simulated model of the system, the line reactance is not fixed, and can be manipulated directly, achieving the same result as, and eliminating the need for, reactive power compensation. Final Report University of California, Irvine 94 Clean Energy Institute Adjusting the transmission line reactance parameters across all linecodes by an order of magnitude, OpenDSS yields the following results. While Figure 72 shows that the order of magnitude reduction in the line reactance has produced an acceptable voltage profile, and Figure 73 shows fewer overloaded transmission lines than Figure 71, there are still undervoltages and many overloaded lines in the circuit. Table 28: Adjusted Reactance Overload Report Line Voltage (kV) Length (km) Double Circuit Highest Amperage (Per Phase) Percentage of Capacity Time Spent Overloaded (Hours) 3564 500 25 FALSE 2566 187 3 3694 220 10 TRUE 2341 202 6 3622 220 8 FALSE 4433 382 11 3685 220 16 TRUE 2346 202 4 3664 220 16 TRUE 2327 201 4 3892 500 20 TRUE 2836 207 3 3893 500 2 TRUE 1891 138 3 3894 500 1 FALSE 3783 276 3 3895 500 12 FALSE 2681 196 4 3703 220 9 TRUE 2936 253 3 297 500 10 FALSE 1649 120 4 298 500 11 FALSE 1637 119 4 3889 500 26 FALSE 1877 137 3 3890 500 31 FALSE 1878 137 3 3891 500 26 FALSE 1886 138 3 3653 500 65 FALSE 2168 158 4 3912 220 14 TRUE 1306 113 3 3568 500 67 FALSE 1873 137 3 5698 500 66 FALSE 1897 138 3 3691 220 13 TRUE 1676 144 4 3903 220 6 TRUE 1452 125 2 3699 220 6 TRUE 1439 124 2 3687 220 9 TRUE 1459 126 2 3688 220 9 TRUE 1454 125 2 3690 69 9 TRUE 1459 126 2 3860 220 1 TRUE 1495 129 2 3742 220 6 TRUE 1246 107 2 3916 220 37 TRUE 1269 109 2 shows the set of transmission lines in the circuit that experience overloads at least once in the 24 hour simulation. Lines for which the ‘Double Circuit’ entry is TRUE should be counted twice, as they have identical lines running in parallel that experience the same overload. Note that these lines are experiencing amperages between 100% and 500% of normal operating capacity. Before these overloads can be addressed, first the loads and generation must be completely balanced in the circuit. Final Report University of California, Irvine 101 Clean Energy Institute source citing the prevalence of bundled conductors could be found, the CPUC corroborates the statement that the phases of lines rated above 200kV “can consist of multiple (bundled) conductors” [19]. Furthermore, ‘double’ conductors (bundles containing two conductors) have been discussed in literature as early as 1932 [20]. Most importantly, the poor performance of the current model suggests the existence of bundled conductors in the real system. To address the current overloads, all overloaded lines will be assigned new linecodes based on how much they are transmitting over their operating capacity. Double, triple and quadruple conductor bundles will have double, triple and quadruple current carrying capacities respectively. Additionally, their resistances will be divided by two, three and four respectively based on the parallel resistor law: 1 𝑅𝑅𝐺𝐺𝑇𝑇𝐺𝐺 = � 1 𝑅𝑅𝑝𝑝 (5) Reactance and capacitance may not be affected as simply as resistance, as suggested by [20], which states double conductors lead to merely a 20% decrease in reactance and a 20% increase in capacitance. It will be assumed that double conductors abide by these changes, and that triple and quadruple conductors have the same reactance and capacitance as double conductors, for simplicity. As a first adjustment to the system, it will be assumed that all lines rated 220kV and over will be equipped with double conductors at minimum. Overloads that persist after this adjustment will be addressed as follows: • Lines equipped with single conductors that experience overloads between 100-200% will be assigned double conductors. • Lines equipped with double conductors that experience overloads between 100-150% will be assigned triple conductors. • Lines equipped with double conductors that experience overloads exceeding 150% will be assigned quadruple conductors. Final Report University of California, Irvine 102 Clean Energy Institute After making this adjustment, the system experiences no overloads and no voltage anomalies, confirmed by Figure 79. Thus, with the system fully meeting electrical loads without current overloads anywhere in the circuit, we have established a base case representative of the real electrical transmission system and can now proceed with analyzing its performance. Figure 79: Final Base Case Overload Plot