BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA

Transcription

1 BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA Application of San Diego Gas & Electric Company (U902M) for Review of its Model Assessment Proceeding Pursuant to Decision Application (Filed May 1, 2015) And Related Matters. Application Application Application JOINT UTILITIES REPORT ON JUA MULTI-ATTRIBUTE FUNCTION TEST DRIVE Laura Earl San Diego Gas & Electric Company 8330 Century Park Court, CP32D San Diego, CA Telephone: (858) Facsimile: (619) Attorneys for SAN DIEGO GAS & ELECTRIC COMPANY Frank McNulty Southern California Edison Company 2244 Walnut Grove Avenue Rosemead, CA Phone: (626) Fax: (626) Attorney for SOUTHERN CALIFORNIA EDISON COMPANY J. Michael Reidenbach Pacific Gas and Electric Company 77 Beale Street, B30A San Francisco, CA Telephone: (415) Facsimile: (415) Attorney for PACIFIC GAS AND ELECTRIC COMPANY Nancy Whang San Diego Gas & Electric Company 8330 Century Park Ct., CP32 San Diego, CA Phone: (213) Fax: (213) Attorney for SAN DIEGO GAS & ELECTRIC COMPANY Dated: October 13, 2017

2 BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA Application of San Diego Gas & Electric Company (U902M) for Review of its Model Assessment Proceeding Pursuant to Decision Application (Filed May 1, 2015) And Related Matters. Application Application Application JOINT UTILITIES REPORT ON JUA MULTI-ATTRIBUTE FUNCTION TEST DRIVE In accordance with The Joint Status Report dated August 11, 2017, of Pacific Gas and Electric Company, Southern California Edison Company, San Diego Gas and Electric Company and Southern California Gas Company (Joint Utilities), together submit their Joint Utilities Report as Attachment A. The undersigned is authorized to sign this pleading on behalf of the Joint Utilities. // // 1

3 Respectfully Submitted, J. MICHAEL REIDENBACH By: /s/ J. Michael Reidenbach J. MICHAEL REIDENBACH Pacific Gas and Electric Company 77 Beale Street, B30A San Francisco, CA Telephone: (415) Facsimile: (415) Dated: October 13, 2017 Attorney for PACIFIC GAS AND ELECTRIC COMPANY 2

4 Attachment A

5 October 13, 2017 Joint Utility Report on Test Drive Results for Multi-Attribute Analysis Pursuant to the updated schedule issued by Administrative Law Judge Colette Kersten on October 5, , the Joint Utilities 2 provide the following narrative report and PowerPoint presentation (see Appendix A). This narrative report and presentation, which evaluate several test drive risks from a financial, safety, and reliability perspective, is intended as a follow-up to the safety-focused analysis that was submitted by the Joint Utilities on September 8, 2017, and presented by the Joint Utilities on September 28, Please note that this narrative portion of the report is intended as a complement to the slide presentation in Appendix A; it is not intended to be a standalone document. For a more detailed review of the regulatory and legislative process leading up to this point, please see the summary in the September 8, 2017, report. The Joint Utility Approach (JUA) Guiding Principles 3 The Joint Utilities have sought to meet or exceed guidance provided by the California Public Utilities Commission (CPUC) in their effort to develop a common risk management methodology (called the Joint Utility Approach, or JUA). JUA was developed to address several criteria that resulted from an analysis of CPUC decisions and materials: Risk Focused: The starting point for JUA is risk identification and analysis, which provides an understanding of the magnitude of the top risks (safety and otherwise) for the utility. Focused: The JUA would like to use the safety attribute to identify the top safety risks for the Risk Assessment Mitigation Phase (RAMP). The JUA analyzes risk mitigations in a multi-attribute context of safety, along with other risk impacts, e.g. financial and reliability impacts. Probabilistic: The JUA allows risks to be quantified as a distribution of probability and impact. 1 See Administrative Law Judge Ruling Updating Schedule In Response To Comments And Entering Phase Two S-MAP Workshop #2 Staff Summary Into The Record, A (filed October 5, 2017). 2 Pacific Gas and Electric Company (PG&E), San Diego Gas & Electric Company (SDG&E), Southern California Edison Company (SCE), and Southern California Gas Company (SoCalGas). 3 Please see Attachment A for additional details and information. 1

6 Simple, Clear, and Transparent: The JUA is based on simple and intuitive concepts that can be understood and evaluated by experts and non-experts alike. Uniform: The JUA was developed by the four large California utilities to measure top safety risks and mitigations in a consistent fashion. Comparable: The JUA Risk Spend Efficiency scores are a significant first step that provides insight into cross-utility comparability. Cost-effective: The JUA does not require substantial effort to implement across all four large California utilities, and allows the utilities to grow and innovate their respective approaches to risk management. Accurate: The JUA provides risk and mitigation analysis that can provide meaningful input into decision making, alternative analysis, and General Rate Case (GRC) resource requests. The Joint Utilities have incorporated these criteria into the JUA. The JUA is intended to provide a consistent, repeatable, and quantitative approach to risk management for both asset and non-asset risks. The JUA Process: High-Level The JUA process has three major phases: Risk Identification & Analysis Mitigation Identification & Analysis Resource Decision Input Risk Identification & Analysis The risk identification process is informed by a combination of existing risk registers, CPUC input, senior management input, industry practices and conferences, brainstorming sessions and workshops, cross-utility discussions, and the non-cpuc regulatory or statutory requirements the Joint Utilities face as publicly traded enterprises (e.g. SEC risk reporting). Risk analysis builds on risk identification by quantifying the identified risks as a function of probability and impact. The building block for risk analysis is the bowtie statement, which provides a structure to disaggregate a risk into three primary components: drivers that cause a risk event, the risk event itself, and outcomes. The bowtie approach allows a risk to be structured and analyzed as a distribution of potential outcomes and the associated probability impact level of those outcomes; this can be performed in terms of a single attribute (e.g. safety) or in terms of multiple attributes (e.g. safety, 2

7 reliability, and financial attributes). The bowtie approach allows for flexibility in the specific quantitative approach that is used to measure the frequency of event occurrence and the associated consequence distributions. 4 Although different risks require different assumptions, data sources and statistical methods, in all cases the JUA model yields a bottom-line expected value 5 figure and tail average values. The expected value and tail average values reflect both probability and impact. The expected value expresses an average outcome in terms of impact per year; the tail average can help utilities analyze high impact, low probability events. Utilities may use one or both of these values for risk prioritization and mitigation selection. The JUA risk analysis approach allows risks to be understood in terms of relative magnitude. For example, a utility would use the results of the safety risk analysis phase of the JUA to inform the scope of risks for a RAMP filing (not to the exclusion of other risks as informed by the CPUC and other key internal or external stakeholders). 6 Mitigation Identification & Analysis As with the case in risk analysis, the JUA approach then informs the Mitigation Identification & Analysis phase. The JUA allows mitigations to be understood in terms of their impact to safety and all relevant risk attributes. Mitigations are defined and quantified relative to their ability to change the probability of a risk event and/or its potential consequences. Each mitigation can be quantified in terms of its forecasted risk reduction on an expected value basis and/or on a tail average basis; once this is combined with the forecasted cost of the mitigation, the JUA allows for potential mitigations to be evaluated and compared in terms of forecasted risk reduction per dollar. 7 Resource Decision Input The risk and mitigation analysis phases of the JUA help quantify our understanding of the magnitude of a risk, the potential impact of mitigations, and the relative efficiency of various mitigation options. These data points provide a starting point for the resource decisions, which includes a broader discussion that considers factors such as risk tolerance, policy goals, current 4 Different risks warrant different levels of modeling sophistication based on factors such as the significance of the risk, the cost effectiveness of modeling, data availability, and feasibility. 5 Expected value is a statistical term that reflects an average outcome based on the probability of an event and the associated outcome. For example, if you have two chances to flip a coin, the expected value that at least one flip will be heads is 75%. 6 See Attachment B for RAMP requirements. The JUA supports each utility s ability to comply with the RAMP requirements. 7 This ratio of risk reduction per dollar has been described with several terms, such as Comparator and Risk Spend Efficiency. 3

8 practices and standards, emerging industry best practices, cross-risk interactions and mitigations, mandated requirements, resource constraints, obstacles to mitigations, and other considerations. This evaluation and discussion informs resource decisions for funding requests to be presented in a GRC. 8 The JUA Process: Attributes 9 The Joint Utilities are proposing to supplement the safety-only attribute from the prior report with two additional attributes: The reliability impact would be an index composed of SAIDI and SAIFI (which measure electric service reliability) and Number of Customers Affected and Customer Minutes of Interruption (which measure gas service reliability). The financial impact, which does not have sub-attributes, would include economic costs to the public, including recoverable costs for the utility. The safety impact, which was included in the prior Joint Utilities report, includes fatalities and injuries to the public, employees, and contractors. The Joint Utilities have focused on these three potential attributes as a current reflection of their core objective and commitment to deliver safe, reliable, and affordable energy. Each attribute and sub-attribute can be defined in terms of three parameters: 1. The natural unit measures the real-world impact of the event. For example, safety is measured in injuries and fatalities. Financial impact is measured in dollars. 2. The top-end value (also called the scaler or limit or range ) defines an upper limit to the impact level measured by natural units. 3. The weighting puts a relative value on the attribute relative to the other attributes. The combined weight of all attributes equals 100%. The three top-level attributes would be common across the Joint Utilities, but the JUA would also allow a utility to choose unique sub-attributes, weights, and top-end values (See Appendix A for an illustrative generic example, as well as the utility-specific illustrative examples, of how these values can be defined). 8 Note that the scope of the risk discussion in a GRC would be defined by the Rate Case Plan (RCP) and other appropriate guidance for that respective utility. 9 The specific attributes and weightings are illustrative and subject to senior management approval as well as applicable regulatory approvals and stakeholder input. 4

9 A simple process allows the attributes to be converted from their natural unit into a risk score: attribute natural unit value attribute top-end range value * attribute weight = weighted attribute risk score For example, assume that a large California utility has identified a risk that could have a financial impact of $2 billion. 10 Also assume that the top-end value for financial impact is $5 billion, and that the financial attribute is weighted at 25%. One would do the following: $2B $5B = * 25% = other weighted attribute scores = total weighted multi-attribute risk score Regardless of whether a utility is analyzing a risk pre- or post-mitigation, this approach allows the utility to measure and convert the impact(s) from natural units into a weighted and cumulative risk score. Finally, because the mitigations are evaluated in terms of a numerical value of risk reduction, the JUA allows for a straightforward calculation of how much risk reduction a particular mitigation provides per dollar. This calculation can be performed solely on the safety attribute, or at a combined level for all attributes. As discussed above, the RSE scores at both the safety attribute level and the combined attribute level provide a starting point for a discussion that considers additional factors. Test Drive Results Please see Appendix A for the full results for the test drive problems for each utility. The data includes pre-mitigation EV scores, EV mitigation reductions, and Risk Spend Efficiency calculations. The JUA in the GRC and Company Decision Process As described in Appendix A, each utility is developing and evolving its respective approach to risk management as well as how the JUA feeds into RAMP filings and GRC resource requests. 10 A large whole number has been selected for this example in the interest of mathematic simplicity. 5

10 These differences reflect regulatory cycles for GRC and RAMP filings, customer base, utility size, systems and infrastructure, geographic and climate region, and other factors. Please see Appendix A for further detail as presented by each utility. 6

11 Attachment A Proposed Acceptance Criteria Criteria Explanation Source Comment JUA Meets Criteria? Risk Focused Focused Simple The Commission has indicated in multiple documents that the GRC process needs to shift from a project focus to a risk focus. While the Commission needs to address all risk attributes (e.g. Reliability) the Model Assessment Proceeding is focused on safety. The process for identifying and evaluating risks and the related mitigations is relatively simple and straightforward. R , Order Instituting Rulemaking to Develop a Risk Based Decision-Making Plan for Energy Utilities It will be beneficial to hold an S-MAP in order to fulfill the objective of providing Commission staff and other parties the opportunity to analyze and understand the various models and methodologies the energy utilities plan to use to prioritize safety in their GRC proceedings, and to manage, mitigate, and minimize such risks. See D at 51. D Historically, the GRC project was focused on projects and costs. With the letter of March 5, 2012, the Commission indicated a need to address both costs and utility risks. The Commission has made it clear on numerous occasions that the focus for this effort is. Criteria to determine any priorities should be the fulfillment of stated Commission goals, ability to impact short-term change, transparency, reasonableness and accuracy of results, ease of preparation The JUA is risk focused, allows identification of the top safety risks for the utility, and connects projects (i.e., mitigations) to risks. The JUA looks at mitigations from the perspective of their impact to safety, as well as to other important risk attributes that utilities must consider to ensure safe, reliable and affordable service. The JUA results and overall methodologies are simple to explain, accessible for non-experts, and may be implemented across the enterprise for all enterprise

12 Proposed Acceptance Criteria Criteria Explanation Source Comment JUA Meets Criteria? Transparent Accurate The inputs and outputs from the process must be rigorous, but the sources of data and the results of the analysis need to be transparent to all stakeholders. The process will result in accurate assessments. It is our intent that the adoption of these additional procedures will result in additional transparency and participation on how safety risks for energy utilities are prioritized by the Commission and the energy utilities See D at 3. D and implementation, among other things. See D at 173. Criteria to determine any priorities should be the fulfillment of stated Commission goals, ability to impact short-term change, transparency, reasonableness and accuracy of results, ease of preparation and implementation, among other things. See D at 173. risks. The JUA process is transparent, in part because it is simple to explain and understand (see above) and all data and results used for decision making are provided. The output of the JUA methodologies will result in assessments as accurate as reasonably possible, and will allow the utilities and stakeholders to rely on those assessments, along with other data points, in making decisions. Uniform While the specific risk weightings will be unique to each utility, the overall risk methodologies used by the large California IOUs are [T]here may be modeling, methodologies, or approaches, that are common to certain energy utilities, and such components It may not be feasible to use models and methodologies that employ a one size fits all approach for all of the energy The JUA methodology applied with the large California IOUs will be uniform. Some of the 8

13 Proposed Acceptance Criteria Criteria Explanation Source Comment JUA Meets Criteria? Probabilistic Cost Effective the same. The large California IOUs should strive to be more quantitative, and implement more probabilistic modeling into their risk methods. The methods applied by each of the utilities will not require new investments in models or extensive O&M expenses to meet the RAMP requirements. may be able to be used interchangeably among these utilities. See D at 26. D at 13. Pub. Util. Code 963(b)(3): The commission shall take all reasonable and appropriate actions necessary to carry out the safety priority policy of this paragraph consistent with the principle of just and reasonable cost-based rates. See also Senate Bill (SB) No. 705, adding 963 to the Public Utilities Code (subsequently expanded to electric in SB 900). utilities. Some flexibility in how each utility assesses its risk, and manages, mitigates, and minimizes its risk will likely be needed. See D at impact weightings and range values may be different depending on utility-specific judgments. The JUA methodology incorporates probabilistic modeling, and the most significant risks presented in the RAMP will use probabilistic modeling. The JUA methodology can be implemented today, without significant incremental costs. The process for many risks will be an improvement based on what utilities are currently developing and will not require significant additional investment. Comparable While the utilities will use unique The JUA methodology 9

14 Proposed Acceptance Criteria Criteria Explanation Source Comment JUA Meets Criteria? User Friendly Useful approaches, the resulting risk mitigation evaluations will be able to be compared for purposes of explaining variations in approaches. The methods and practices used by the large California IOUs will be user friendly for those required to understand, implement, and provide input/data on risks and mitigations. The results of the RAMP filings and related analysis will be useful to all stakeholders. It will be beneficial to hold an S-MAP in order to fulfill the objective of providing Commission staff and other parties the opportunity to analyze and understand the various models and methodologies the energy utilities plan to use to prioritize safety in their GRC proceedings, and to allows for comparability of mitigations, especially when looking at the JUA -only lens. This allows the utilities to explore mitigation strategies by discussing differences between their mitigation approaches. The JUA methodology is user friendly, and in many cases is simply an expansion of the concepts utilities already employ today for risk and mitigation assessments. The results of the JUA methodology, from the risk scores to the mitigations effectiveness values, are helpful for understanding mitigation funding and effectiveness. 10

15 Proposed Acceptance Criteria Criteria Explanation Source Comment JUA Meets Criteria? manage, mitigate, and minimize such risks. See D at 51. Used The results will be used by all stakeholders including the utilities. R , Order Instituting Rulemaking to Develop a Risk Based Decision-Making Plan for Energy Utilities The OII was launched in order to produce a solution that will be used. The results of the JUA methodology can be used today for riskinformed decision making. Not a Significant Effort to Implement Does Not Mask The proposed approach can be implemented using existing resources and does not require numerous changes in business processes. The approach does not have the potential of masking significant safety risks or mitigations. Criteria to determine any priorities should be the fulfillment of stated Commission goals, ability to impact short-term change, transparency, reasonableness and accuracy of results, ease of preparation and implementation, among other things. See D at 173. R , Order Instituting Rulemaking to Develop a Risk Based Decision-Making Plan for Energy Utilities The Commission has consistently stated that it wants to ensure significant safety risks are addressed. The JUA methodology can be implemented without many changes to existing business processes, and allows for growth, innovation, and maturity improvements over time. The JUA looks at mitigations using many lenses, including a safety-only lens and a multiattribute lens. [Same as Focused above] 11

16 Attachment B The JUA Compliance with RAMP The JUA satisfies the requirements of the S-MAP proceeding, as discussed in the body of this Narrative Report, while also supporting each large California utility as it fulfills its RAMP obligations in a consistent manner. Currently, the S-MAP Interim Decision requires large utilities RAMP filings to: 1. Identify its top safety risks 2. Describe the controls and mitigations currently in place and a plan for improving mitigation of each risk 3. Present at least two alternative mitigation plans for each risk 4. Present an early stage risk mitigated to cost ration or related risk reduction per dollar spent (also referred to as Risk-Spend Efficiency [RSE]) 5. Move toward probabilistic calculations as much as possible 6. Describe the company s safety culture, executive engagement and compensation policies 7. Respond to immediate or short-term crises outside of the RAMP and GRC processes; identify lessons learned 8. Complete annual accountability reports 12

17 APPENDIX A

18 Appendix A Joint Utilities Approach (JUA) to Risk Assessment JUA Multi-Attribute Workshop November 6, 2017 All data and materials in this document are illustrative and not meant to represent actual risk assessments. 1

19 Overview 1. Background of Model Assessment Proceeding (S-MAP) 2. The Joint Utilities Approach (JUA) Methodology and its Application 3. Detailed Walk-through 1. SDG&E / SoCalGas 2. PG&E 3. SCE 4. Conclusion 5. Next Steps All data and materials in this document are illustrative and not meant to represent actual risk assessments. 2

20 Overview In 2014, Senate Bill 900 states that the Commission shall develop formal procedures to consider safety in a rate case application by electric or gas corporations Develop the models and methodologies all utilities use to evaluate risks and mitigations S-MAP: Model Assessment Proceeding RAMP: Risk Assessment Mitigation Phase Identify top safety risks and mitigation proposals developed using the models and methodologies adopted in S-MAP Incorporate the RAMP submission into utility rate cases Rate Cases All data and materials in this document are illustrative and not meant to represent actual risk assessments. 3

21 Guiding Principles for the JUA Methodology Development Risk Focused Focused CPUC Docs Policy Statement (July 10, 2014), Action Plan and Regulatory Strategy (February 12, 2015), Policy and Planning Division and Enforcement Division Quantifying Risk: Building Resiliency into Utility Planning (January 23, 2014), Cycla Report (May 16, 2013), Liberty Consulting Report (May 6, 2013), and Enforcement Division Risk Assessment section Staff Report on SoCalGas and SDG&E s GRC (March 27, 2015), and Enforcement Division Risk Assessment section Staff Report on PG&E s GRC (March 7, 2016), Enforcement Division Evaluation Report on the Risk Evaluation Models and Risked-based Decision Frameworks in A , et al (March 21, 2016), S- MAP Decision D (August 18, 2016), S-MAP Scoping Memo and Ruling of Assigned Commissioner in A (December 13, 2016), and SED Report on SCE s GRC A (January 31, 2017) Probabilistic Simple / transparent (understandable by nonexperts) Uniform Comparable Cost-effective modeling / Not a significant effort to implement Accurate Acceptance Criteria See Appendix for text. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 4

22 Application of the JUA Methodology Identify and Analyze Top Risks Evaluate Mitigations & Alternatives Discussion: Mitigations, Risk Tolerance, Leading Practices Input to Resource Decisions Step 1 Utilities identify and categorize risks Step 2 Utilities perform JUA Attribute assessment to identify top safety risks and calculate safety comparators Step 3 Utilities perform JUA Multi- Attribute evaluation of proposed mitigations and alternatives for the top safety risks Step 4 Calculate Multi-Attribute Risk Spend Efficiency (RSE) Utility Decision-Making and GRC Process Cycla Evaluation Model Step 1 Identify Threats Step 2 Characterize Sources of Risk Step 3 Identify Candidate Risk Control Measures (RCMs) Step 4 Evaluate the Anticipated Risk Reduction for Identified RCMs Step 5 Determine Resource Requirements for Identified RCMs Step 6 Select RCMs Considering Resource Requirements and Anticipated Risk Reduction Step 7 Determine Total Resource Requirements for Selected RCMs Step 8 Adjust the Set of RCMs to be Presented in GRC Considering Resource Constraints Step 9 Adjust RCMs for Implementation following CPUC Decision on Allowed Resources Step 10 Monitor the Effectiveness of RCMs All data and materials in this document are illustrative and not meant to represent actual risk assessments. 5

23 Quantitative Risk Assessment Different risks warrant different levels of modeling sophistication based on various factors such as the significance of the risk, the cost effectiveness of modeling, data availability and feasibility. The JUA methodology can be adapted to different levels of sophistication and can be implemented quickly. Model Sophistication Data Driven SME Risk Assessment Mitigation Effectiveness Risk Plan RAMP Proposal All data and materials in this document are illustrative and not meant to represent actual risk assessments. 6

24 JUA: Multi-Attribute Test Drive Overview Overall Risk Reliability Financial The safety impact of a risk event includes fatalities and injuries of the public, employees and contractors. # of Fatalities # of Injuries Reliability Index Score Electric Reliability Score* Financial Impacts ($ -- e.g., environment, compliance, claims) The reliability attribute top measurement is the reliability index which is a composite of the gas reliability index and electric reliability index. Gas Reliability Score** The financial impact of a risk event may includes economic costs to the public, including recoverable costs for the utility. * Electric Reliability Score is composed of SAIDI and SAIFI ** Gas Reliability Score is composed of Customers Affected and Customer Minutes All data and materials in this document are illustrative and not meant to represent actual risk assessments. 7

25 JUA: Multi-Attribute Illustration The attribute framework allows for utility-specific weights, sub-attributes, and top-end range boundaries. The natural units of each attribute are turned into risk units with the following process: attribute natural unit value attribute top-end range value * attribute weight For example, assume a financial impact is identified as having financial EV of $2B: $2B $5B = * 25% = other attribute scores = total multi-attribute risk score Multi-Attribute Methodology Attribute Unit Top End (Scaler) Weighting SU 10 50% Reliability RU 1 25% Financial $ $5 Billion 25% Reliability Unit Breakdown Gas/Electric Unit Top End (Scaler) Weighting Gas # of Customers 1.5 Million 50% Electric SAIDI Index 1,000 25% Electric SAIFI Index 5 25% All data and materials in this document are illustrative and not meant to represent actual risk assessments. 8

26 Risk XYZ, with 1 mitigation, with cost of $20 million Attribute EV (Current) EV (Post- Mitigation) JUA: Multi-Attribute Illustration Weighting Top-End (Scaler) % 10 Reliability % 1 Financial $40M $20M 25% $5 B Reliability Sub-Attribute Gas (# of customers) EV (Current) EV (Post- Mitigation) Weighting Top-End (Scaler) % 1.5 Million Electric (SAIDI) % 1,000 Electric (SAIFI) % 5 Reliability Unit Current: (0) * 50% + (300/1000) * 25% + (0.8/5) * 25% = Post-Mitigation: (0) * 50% + (200/1000) * 25% + (0.6/5) * 25% = Risk Score RSE Current: (1/10) * 50% + (0.115/1) * 25% + (40/5000) * 25% = 0.08 Post-Mitigation: (0.8/10) * 50% + (0.08/1) * 25% + (20/5000) * 25% = 0.06 ( ) / $20M = 0.1 per $100M All data and materials in this document are illustrative and not meant to represent actual risk assessments. 9

27 JUA Test Drive SDG&E OH Conductor (Wire Down) Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Wires down on Distribution system from any cause (except wildfire) SDG&E considered the number of wire down events as the main trigger. The value of 80 per year is the number of wire down events. The mitigations shown only impact the reduction of wire down events, as opposed to any reduction in consequence given a wire down Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation 15% reduction is estimated based upon replacing 100 miles 15% (from 80 SDGE Targeted of system, and using subject matter expertise to determine down to Reconductor - $50M/yr how many fewer wires down there would be with that much 68/year) OH replacement. SDGE Targeted Reconductor - $150M/yr 30% (Down to 56/year) 30% reduction is estimated based upon replacing 300 miles of system, and using subject matter expertise to determine how many fewer wires down there would be with that much OH replacement. $150M Project Cost $450M Project Cost 40 (0.038) (0.0031) (0.085) (0.0065) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. 10 Illustrative Data Developed for the Purposes of JUA Test Drive

28 JUA Test Drive SDG&E High Pressure Gas Pipeline Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Pipeline incident on a high pressure pipeline, excluding dig-ins National PHMSA data from 2010 to present was used to identify the frequency and consequences of high pressure gas incidents. SDG&E separated out third-party dig-ins from this analysis, due to that cause being included in a separate risk Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Targeted Replacements 30 mile 14.4% Estimated Reduction in Likelihood SDG&E Pipe Replacement (30 miles of worst performing). SME input for illustrative study. $150M Project Cost 80 (0.0006) (0.0002) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test Drive 11

29 JUA Test Drive SoCalGas High Pressure Gas Pipeline Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Pipeline incident on a high pressure pipeline, excluding dig-ins National PHMSA data from 2010 to present was used to identify the frequency and consequences of high pressure gas incidents. SCG separated out third-party dig-ins from this analysis, due to that cause being included in a separate risk Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Targeted Replacements 30 mile Targeted Replacements 300 miles 1.9% Estimated Reduction in Likelihood 10.1% Estimated Reduction in Likelihood SCG Pipe Replacement (30 miles of worst performing). SME input for illustrative study. SCG Pipe Replacement (300 miles of worst performing). SME input for illustrative study. $150M Project Cost $1,500M Project Cost 80 (0.0012) (0.0003) (0.0063) (0.0044) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test Drive 12

30 JUA Test Drive SDG&E Workplace Violence Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk A violent incident occurs at a workplace site Data used: Fatalities from OSHA; non-fatal injuries from BLS; number of private industry workers from FRED; SME (riskier than national average) Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Mitigation 1 8% Estimated Reduction in Likelihood Implementation of all mitigations and new programs (Risk Analyst, Incident Management Database improvements, social media monitoring) $15.18M Project Cost 7 (0.0026) (0.0001) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. 13 Illustrative Data Developed for the Purposes of JUA Test Drive

31 JUA Test Drive SoCalGas Workplace Violence Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk A violent incident occurs at a workplace site Data used: Fatalities from OSHA; non-fatal injuries from BLS; number of private industry workers from FRED; SME (riskier than national average) Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Mitigation 1 9% Estimated Reduction in Likelihood Implementation of all mitigations and new programs (Risk Analyst, Incident Management Database improvements, social media monitoring) $4.62M Project Cost 5 (0.0054) (0.0003) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. 14 Illustrative Data Developed for the Purposes of JUA Test Drive

32 JUA Test Drive SDG&E Workforce Planning Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk A safety incident occurs due to inadequate knowledge transfer to new utility employees Data used: SME data (for safety impact and frequency). This is the risk of not having an appropriate workforce with the right skills to meet business needs due to the acceleration of workforce attrition and changing business needs Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Mitigation 1 Improve Job proficiency 48.5% Estimated Reduction in Likelihood Data used: SME provides the percentage reduction of incident rate, which is then converted into a reduction in likelihood. $3.93M Project Cost 2 yr O&M (0.0479) (0.0029) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. 15 Illustrative Data Developed for the Purposes of JUA Test Drive

33 JUA Test Drive SoCalGas Workforce Planning Risk Description A safety incident occurs due to inadequate knowledge transfer to new utility employees Comments/Notes Data used: SME data (for safety impact and frequency). This is the risk of not having an appropriate workforce with the right skills to meet business needs due to the acceleration of workforce attrition and changing business needs. Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator: Only RSE*: All Attributes Combined Mitigation Mitigation 1 Improve Job proficiency 20.2% Estimated Reduction in Likelihood Data used: SME provides the percentage reduction of incident rate, which is then converted into a reduction in likelihood. $2.34M Project Cost 2 (0.0201) (0.0012) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. 16 Illustrative Data Developed for the Purposes of JUA Test Drive

34 Risk Description Rupture of transmission pipeline may result in loss of containment and/or uncontrolled gas flow leading to potential public safety issues, prolonged outages, property damages and/or significant environmental damage. JUA Test Drive PG&E Gas Transmission Pipeline Failure Comments/Notes Historical company data and PHMSA data on number of incidents occurred with and without ignition by the nine ASME B31.8S threat categories (hereafter known as risk drivers); subject matter expert judgement regarding the effectiveness of the mitigations. Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Description Effectiveness and Data Sources Total Project Cost and Useful Life Risk Reductions Δ EV: Δ EV: All Attributes Combined Comparator RSE* Mitigation Valve Automation Vintage Pipe Company ~8.2% (Exposure of ~540 approximate miles of automated Data and SME pipe); ~14.1% approximation of onshore ignited incidents; Judgement 600% effectiveness Company ~0.3%(Exposure of 22.5 miles, 3 years worth); ~14.1% Data and SME approximation of onshore ignited incidents; 600% Judgement effectiveness $89.2M. ~180 miles per year at $0.2M per miles $181.69M. ~7.5 miles per year at $8.1M per miles 65 (0.0011) ( ) (0.0008) ( ) Strength Testing Company ~5%(Exposure of 325 miles, 3 years worth); ~14.1% Data and SME approximation of onshore ignited incidents; 600% Judgement effectiveness $358.79M. ~108 miles per year at $1.1M per miles 7 (0.0015) ( ) ILI Company ~13.9%(Exposure of 906 miles, 3 years worth); ~14.1% Data and SME approximation of onshore ignited incidents; 600% Judgement effectiveness $984.54M. ~302 miles per year at $1.09M per mile 7 (0.0023) ( ) Illustrative Data Developed for the Purposes of JUA Test Drive * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project.

35 JUA Test Drive PG&E Electric Distribution Overhead Conductor Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Failure of or contact with, energized electric distribution primary conductor results in public safety issues, significant environmental damage, prolonged outages, or significant property damage. Company data on wire down events by cause; Failure rates by conductor size and type in corrosion zones; Count of PUC reportable 3rd party events related to Distribution OH Primary conductor; assumption based on historical Distribution engineer investigations on Wire Down events that 30% of Wire Down events may remain energized Mitigation Description Targeted Conductor Replacement (4 ACSR) in Corrosion zone ($110.8M/yr) Focus on highest risk circuits based on historical vegetation caused wire down events for underground conversion Clear vegetation directly above OH Primary conductor Company Data Company Data Company Data and SME Judgement Effectiveness and Data Sources 0.8% (Exposure 630 miles (3 years) / system) * 25.5% (Equipment failure conductor connector / total wires down) * 1060% (effectiveness in reducing Equipment failure caused wire down events related to Conductor or Connector assets: 10.6x multiplier applied since WD/100 mile rate in corrosion zones are 5.3 compared to.5 in non-corrosion zone areas) 0.2% (Exposure 150 miles 3 years worth/ system) * 42.4% (Vegetation caused / total wires down) * 791% (effectiveness in reducing Vegetation caused wire down events per mile: 13 worse performing circuits make up 11.31% of Vegetation wire down events and only 1.43% of total miles % / 1.43% = 791%) 6% (Exposure 4950 miles 3 years worth / system) * 42.4% (Vegetation caused / total wires down) * 16.9% (effectiveness in reducing Vegetation caused Wire Down events: 70% reduction on branch outage on circuit miles worked (per historical PS&R analysis), 24.18% of Vegetation wires down events are from branch caused outages (other categories included full tree failures, trunk failures, etc.), 70%*24.18% = 16.9%) Total Project Cost and Useful Life $332.64M. Replace 210 Miles a year at $100/ft ($528k/mile) $450.00M. 50 Miles a year at $3M/mile $17.82M miles in top 40% REAX = 1650 miles a year at $3,600/mile * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Risk Reductions Δ EV: Δ EV: All Attributes Combined Comparator RSE* 40 (0.0235) ( ) (0.0073) ( ) (0.0048) ( ) Illustrative Data Developed for the Purposes of JUA Test Drive

36 JUA Test Drive PG&E HR Skilled and Qualified Workforce Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk The risk of Employees performing work The baseline likelihood and consequence for this risk is a composite of the data attributable to for which they are not skilled or qualified employee error from PG&E's asset based RAMP risks. Those risks are Gas Storage Wells, Gas presents to PG&E and the public related Maintaining system capacity, Gas Compression & Processing facility, Gas Measurement & Control to safety, reliability, affordability and facility, Gas Measurement & Control downstream, Gas Distribution - Non-cross bore, Gas security. Transmission pipeline, Electric Distribution OH conductor, and Electric Transmission OH conductor Mitigation Description Portable Technology-- Technical Access-- Provide to All Field Employees Portable Technology-- Technical Access-- Provide to Supervisors Crew Leads SME Judgement SME Judgement Effectiveness and Data Sources Total Project Cost and Useful Life Risk Reductions Δ EV: Δ EV: All Attributes Combined Comparator Based on expertise and assessment data 1.43M 2 (0.0379) ( ) Based on expertise and assessment data 1.43M 2 (0.0288) ( ) RSE* Portable Technology-- Qualification Status SME Judgement Based on expertise and assessment data 0.50M 2 (0.0196) ( ) /7 Technical Support Desk SME Judgement Based on expertise and assessment data 2.00M 2 (0.0296) ( ) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test 19 Drive

37 JUA Test Drive SCE Distribution OH Conductor Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Overhead conductor down in service leading to public contact with the conductor, a wildfire, or property damage. 860 Annualized wire down events has varied over the years that it has been recorded Risk Reductions Description Reconductoring the highest-rse circuits at a $100M per year pace. SMEcalibrated model Effectiveness and Data Sources Based on the expected reduction in wire down events by reconductoring small conductor on the highest RSE circuits. Total Project Cost and Useful Life $300M 40 years 669 $447K per mile Δ EV: Δ EV: All Attributes Combined Comparator RSE*: All Attributes Combined (0.131) (0.007) Mitigation Reconductoring the highest-rse circuits at a $150M per year pace. SMEcalibrated model Based on the expected reduction in wire down events by reconductoring small conductor on the highest RSE circuits. $450M 40 years 998 $447K per mile (0.161) (0.008) Reconductoring the highest-rse circuits at a $200M per year pace. SMEcalibrated model Based on the expected reduction in wire down events by reconductoring small conductor on the highest RSE circuits. $600M 40 years 1,340 $447K per mile (0.213) (0.011) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test 20 Drive

38 JUA Test Drive SCE Active Shooter Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk Deliberate violent actions of current or former worker leading to serious injuries and/or fatalities to self or others which can potentially have safety and financial impacts Active Shooter events over a 16-year period matching the defined risk statement occurring at a Fortune 500 or similar organization Mitigation Description Effectiveness and Data Sources Total Project Cost and Useful Life Security personnel assigned to populated facilities Active Shooter training SMEestimate Based on the expected deterrence level provided by security personnel presence Based on the expected response to an Active Shooter event $7M $0.05M Estimated annual cost. One-time cost for training development. Risk Reductions Δ EV: Δ EV: All Attributes Combined Comparator RSE*: All Attributes Combined (0.012) (0.001) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test Drive

39 JUA Test Drive SCE Unqualified Workforce Description Comments/Notes Pre-Mitigation EV Pre-Mitigation EV All Attributes Combined Risk SCE Employee and/or Contractor actions that led to serious injuries and/or fatalities to self, other employees and/or contractors and the public. The data needed for this test drive has been kept in manual documents before May Due to lack of time and resources, we chose to only include Employee and/or Contractor serious injuries to self or other employee and/or contractor that have been transferred to a database from 5/ /2016. The data for Employee and/or Contractor incidents that led to injury and/or fatality to the public included all incidents over a 2 year period (Years 2015 and 2016) and does not assume or infer that an SCE Employee and/or Contractor was found at fault for the incident Risk Reductions Description Effectiveness and Data Sources Total Project Cost and Useful Life Δ EV: Δ EV: All Attributes Combined Comparator RSE*: All Attributes Combined Mitigation The mitigation for this risk is training for T&D employees. SMEcalibrated model Based on SCE s year over year Days Away, Restricted, Transfer (DART) rate trend $37M Cost per year (0.692) (0.035) * RSE is Risk Spend Efficiency. This is the measure of risk reduction per $1 million cost for the mitigation that also reflects the useful life of the project. Illustrative Data Developed for the Purposes of JUA Test Drive

40 Questions? All data and materials in this document are illustrative and not meant to represent actual risk assessments. 23

41 SDG&E and SoCal Gas A Walk Through The RAMP All data and materials in this document are illustrative and not meant to represent actual risk assessments. 24

42 Outline Selecting Risks for RAMP RAMP Comprehensive discussion of risk Numerical representations Recommendations Overview and Detail View Overview methodology, give illustrative example for each step (indicated with blue type) All data and materials in this document are illustrative and not meant to represent actual risk assessments. 25

43 Overview Selecting Risks for RAMP Perform JUA safety assessment for each risk. Risks above certain threshold to go to RAMP Many techniques available when JUA safety assessment data present Additional risks at CPUC discretion to be included Climate Change Gas Storage All data and materials in this document are illustrative and not meant to represent actual risk assessments. 26

44 Detail View Selecting Risks for RAMP Risk Name Wildfires Caused by SDG&E Equipment (including Third Party Pole Attachments) Employee (previously combined as Employee, Contractor & Public risk in 2015) Catastrophic (10+) Extreme (3-10) High (1-3) Moderate (0.1-1) Low ( ) Electric Infrastructure Integrity Inadequate Knowledge Transfer Catastrophic Damage Involving Third Party Dig-Ins Customer (previously combined as Employee, Contractor & Public risk in 2015) Contractor (previously combined as Employee, Contractor & Public risk in 2015) Physical Security of Critical Electric Infrastructure Climate Change Adaptation Catastrophic Damage Involving a Medium Pressure Gas Pipeline Failure EV Quantification of all risks SDG&E and SoCalGas will model all risks with stochastic approaches Can present outputs in various forms JUA safety assessment buckets are an approach Unmanned Aircraft System (UAS) Incident Distributed Energy Resources (DERs) Aviation Incident Catastrophic Damage Related to Inadequacy of Operational Asset Records (previously combined as Records Management risk in 2015) Electric Grid Failure and Restoration (Blackout/Failure to Black Start) Catastrophic Damage Involving a High Pressure Gas Pipeline Failure Insufficient Supply to the Natural Gas Transmission System Workplace Violence Management of Emergency Spares for Major Electric Equipment Violation of Environmental Policies/Procedures IT System Compliance Insurance Coverage Issue PSEP Execution and Reasonableness Review Outcome Negative Customer Impacts Caused by Outdated Systems (new risk for 2016) Customer Privacy Regulatory Compliance Departing Load including Direct Access (DA) and Community Choice Aggregation (CCA) Flawed Electric Rate Design Capacity Restrictions or Disruptions to the Natural Gas Transmission System Massive Smart Meter Outage Cyber Security Major Project Delays (e.g. CNF, SOCRE, PSRP) IT Critical Infrastructure Risk (Natural Disasters) Gas Pipeline Regulatory Compliance Untimely and Unfavorable Regulatory Decisions Access to Capital Markets Don t attempt to read numbers. We are not responsible for blurred vision. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 27

45 Detail View Selecting Risks for RAMP JUA safety assessment primer Risk Name Wildfires Caused by SDG&E Equipment (including Third Party Pole Attachments) Employee (previously combined as Employee, Contractor & Public risk in 2015) Catastrophic (10+) Extreme (3-10) High (1-3) Moderate (0.1-1) Low ( ) EV Values in box represent likelihood that each consequence will occur. For example, there is a 15% likelihood that 1-3 Units will occur in a given year, for the Employee risk All data and materials in this document are illustrative and not meant to represent actual risk assessments. 28

46 Detail View Selecting Risks for RAMP Risk Name Wildfires Caused by SDG&E Equipment (including Third Party Pole Attachments) Employee (previously combined as Employee, Contractor & Public risk in 2015) Catastrop hic (10+) Extreme (3-10) High (1-3) Moderate (0.1-1) Low ( ) EV RAMP? Y Y Electric Infrastructure Integrity Y Inadequate Knowledge Transfer Y Catastrophic Damage Involving Third Party Dig-Ins Y Customer (previously combined as Employee, Contractor & Public risk in 2015) Contractor (previously combined as Employee, Contractor & Public risk in 2015) Physical Security of Critical Electric Infrastructure Climate Change Adaptation SED Catastrophic Damage Involving a Medium Pressure Gas Pipeline Failure Unmanned Aircraft System (UAS) Incident Distributed Energy Resources (DERs) Aviation Incident Catastrophic Damage Related to Inadequacy of Operational Asset Records (previously combined as Records Management risk in 2015) Electric Grid Failure and Restoration (Blackout/Failure to Black Start) Catastrophic Damage Involving a High Pressure Gas Pipeline Failure Y Insufficient Supply to the Natural Gas Transmission System Potential Thresholds. Above which the Risk would be included into RAMP Based on feedback, can adjust method of identification, or level of thresholds. Thresholds need not be identical across all utilities. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 29

47 Overview RAMP Discussion RAMP Discussions for each risk: Current programs / mitigations in place Quantitatively describe risk Bow-tie; drivers, consequences Multi-Attribute scoring Constraints Interactions with other risks Goals Possible mitigations Effectiveness Multi-Attribute Efficiency Propose mitigations All data and materials in this document are illustrative and not meant to represent actual risk assessments. 30

48 Detail view RAMP (Electric Infrastructure) Discussion of current efforts Standards in place Backcountry Design Guide (Construction standards in SDG&E s Fire Threat Zone (FTZ)) CPUC rules and General Orders Wind loading (known local conditions) Practices in place Reliability reports and analysis High SAIDI outage analysis Reliability Director s Council Fire Director s Steering Committee Substation Reliability team with CBM WRRM modeling Wires down analysis LiDAR All data and materials in this document are illustrative and not meant to represent actual risk assessments. 31

49 Detail view RAMP (Electric Infrastructure) Quantitatively Describe Risk JUA safety assessment output Catastrophic Extreme High Moderate Low Risk Name EV (10+) (3-10) (1-3) (0.1-1) ( ) Electric Infrastructure Integrity Historical Events Bow-tie Drivers Aging infrastructure Over-utilized equipment (above capacity) Equipment failure System Protection issues Consequences Serious injury or fatality Environmental impacts Reliability Customer Satisfaction All data and materials in this document are illustrative and not meant to represent actual risk assessments. In future RAMP, all modeling will be stochastic, incorporating as much actual data as possible. Modeling techniques will be described in detail in RAMP. 32

50 Detail view RAMP (Electric Infrastructure) Quantitatively Describe Risk Demonstrate baseline MAUT scores Electric Infrastructure EV : 0.3 Reliability: SAIDI 20 minutes, SAIFI 0.3 outages, no gas concerns Finance: $80M Electricity Infrastructure score Reliability Unit = (0) * 50% + (20/1000) * 25% + (0.3/5) * 25% = 0.02 Overall Risk EV: (0.3/10) * 50% + (0.02/1) * 25% + ($80/$5000) * 25% = 0.024» Risk Score: 24,000 Risk Score = Risk EV * 1,000,000. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 33

51 Overview - RAMP Constraints Discuss things that can slowdown or prevent implementation of mitigating efforts Environmental Permits Labor/technology Interactions with other risks Does the risk and/or its mitigations have an impact on other risks Example: Converting from OH to UG electric construction can reduce fire risk and public safety risk from wires down All data and materials in this document are illustrative and not meant to represent actual risk assessments. 34

52 Detail view RAMP (Electric Infrastructure) Constraints Substation Transformers have significant lead time, and can require modification of substation layouts. Substation configuration changes can have environmental reviews. Distribution work can require traffic permits, environmental permits System Protection: Significant changes in technology can require testing and pilot studies prior to a full scale roll-out Interactions with other risks Overlap with several initiatives including wildfire risk, employee risk, electric grid failure, etc. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 35

53 Overview - RAMP Goals Is the emphasis: To improve safety, using the most effective safety mitigations possible To maintain safety at current levels, using the most efficient safety mitigations possible To focus on overall risk score by applying the most efficient multiattribute mitigations possible Each risk is different All data and materials in this document are illustrative and not meant to represent actual risk assessments. 36

54 Detail View RAMP (Electric Infrastructure) Possible mitigations Transformer replacements Transformer monitoring equipment OH Conductor and pole hardening System Protection upgrades or new installations All data and materials in this document are illustrative and not meant to represent actual risk assessments. 37

55 Detail View RAMP (Electric Infrastructure) Illustrative Mitigations Mitigation 1 (Transformer Replacements) Primarily improves Reliability Long term impacts Mitigation 2 (Large Scale OH hardening) Improves Reliability and Long term impacts Mitigation 3 (Transformer Monitors) Improves Reliability Shorter term impacts All data and materials in this document are illustrative and not meant to represent actual risk assessments. 38

56 Detail View RAMP (Electric Infrastructure) Current Mitigation 1 Mitigation 2 Mitigation 3 Units Reliability Units Finance $80M $60M $60M $70M Risk Score RSE per $1M Comparator per $1M ( Efficiency) Unit Improvement ( Effectiveness) Illustrative conclusions Mitigation discussion Mitigation 2 has more constraints due to permitting and resource issues. Mitigation 1 and 3 are relatively less complicated SDG&E recommends Mitigation 1 and Mitigation 3 Mitigation 1 has nearly the same amount of Improvement as Mitigation 2, but does so more efficiently. The RSE for Mitigation 3 fares well not only against other Electric Infrastructure mitigations but against other risk mitigations as well. Mitigations 1 and 3 also match company and Commission objectives of modernizing the grid with technology. All data and materials in this document are illustrative and not meant to represent actual risk assessments. 39

57 Overview GRC Unless the situation has changed since RAMP filing GRC to seek funding based on discussions and recommendations in RAMP Specific levels of funding to be viewed across enterprise with strongest recommendations most likely to remain in GRC All data and materials in this document are illustrative and not meant to represent actual risk assessments. 40

58 PG&E An End-to-End Process All data and materials in this document are illustrative and not meant to represent actual risk assessments. 41

59 Pre RAMP: Baseline Risk Assessment PG&E 1. [Cycla Step 1] Identify Threats Risk Event identification (Center of the bow tie) 2. [Cycla Step 2] Characterize Sources of Risk Risk Driver identification and quantification (Left-hand side of bow tie) Risk Consequence identification and quantification (Right-hand side of the bow tie) 3. List current controls that are embedded in the current state quantification # Current Controls 1 Control A 2 Control B 3 Control C 4 Control D 4. Repeat Steps 1-3 for another risk to assess the company risk register All data and materials in this document are illustrative and not meant to represent actual risk assessments. 42