浑水做空 Edison International：称 SCE 应为 Eaton 山火承担约200亿美元无上限赔偿

Edison International Should be
Liable for ~$20 Billion
Newly surfaced filings lead us to believe that EIX subsidiary Southern California Edison
(“Edison”) acted with conscious and willful disregard for the rights and safety of others,
which caused the catastrophic Eaton Fire. If the state ultimately agrees, Edison could be
liable for in excess of ~$20 billion. We believe it is virtually certain that, at the least, Edison
will be found imprudent and be liable for at least $4.3 billion.

Altadena, California, January 8, 2025. Photo: Matt Gush/Shutterstock.
October 13, 2026


Page 2 of 54
“Left Disclosure”: As of the time and date of this report, Muddy Waters (defined below) is short the securities of, or derivatives
linked to Edison International (Edison) (the “Covered Issuer”). Upon publication, we intend to begin covering a substantial
majority – possibly all – of our short positions. As we elaborate below, our risk reduction is not a reflection of a lack of
conviction in our opinions or the facts presented; rather, it has to do with managing risk in a manner that is prudent for a
fiduciary of our investors’ money.

This report has been published by Muddy Waters, LLC (“Muddy Waters Research” or “we” or “us”). Muddy Waters Research is under
common control and affiliated with Muddy Waters Capital LLC (“Muddy Waters Capital”). Muddy Waters Research is an online
research publication that produces due diligence-based reports on publicly traded securities, and Muddy Waters Capital LLC is an
investment adviser registered with the U.S. Securities and Exchange Commission. This presentation is the property of Muddy Waters
Research. Muddy Waters Research and Muddy Waters Capital, collectively their respective affiliates and related parties, including, but
not limited to any principals, officers, directors, employees, members, clients, investors, consultants and agents, are referred herein to as
“Muddy Waters”.
Muddy Waters will continue transacting in the securities of Covered Issuer for an indefinite period after this report on the Covered
Issuer, and we may be net short, net long or flat positions in the Covered Issuer’s securities after the initial publication of this report,
regardless of our initial position and views herein.
We are a “for profit” journalistic organization with a non-traditional revenue model – rather than accepting advertising money or
subscriptions, we finance our journalism through taking positions in the securities of companies on which we report. This revenue model
enables us to report in great depth on a limited number of investigations and also entails our taking significant financial risk on behalf of
ourselves and our clients. In order to manage risk, we must close open positions as we deem prudent. We do not provide “price targets”,
although we may express our opinion of what the security is worth. An opinion of the value of a security differs from a price target in
that we do not purport to have any insight as to how the market as a whole might value a security – we can only speak for how we,
ourselves, view its value. We therefore do not hold a position until it reaches a certain price target, nor do we hold positions until they
reach the price at which we have expressed a valuation opinion. There are numerous factors that enter into investment decisions aside
from opinions of the value of the security, including without limitation, the borrow cost of a shorted security, the potential for a “short
squeeze”, prudent risk sizing relative to capital and volatility, reduced information asymmetry, the opportunity cost of capital, client
expectations, the ability to hedge market risk, our perception of the efficacy of market regulators and gatekeepers, our perception of the
resource imbalance between us and Covered Issuers, and our moods and gut feelings. Therefore, you should assume that upon
publication of this report, we will, or have begun to, close a substantial portion – possibly the entirety – of our positions in the Covered
Issuer’s securities.
We are not providing you with a recommendation to buy or sell securities of the Covered Issuer. We are articulating our reasons at the
time of publication we have positions in the securities of the Covered Issuer.
We have no duty or obligation to update this report or update you on the size or direction of any position we hold in a Covered Issuer.
We do not provide investment advice to any person, unless our affiliate has entered into an investment adviser-client relationship with
that person in writing.
All information and opinions set forth herein are for informational purposes only. Under no circumstances should any information or
opinions herein be construed as investment advice, as an offer to sell, or the solicitation of an offer to buy any securities or other
financial instruments.
This report is opinion journalism. We are providing our journalistic opinions about issues of concern to the general public. The
opinions, information, and reports set forth herein are solely attributable to Muddy Waters Research. Before making any investment
decision, you should do your own research and due diligence before making any investment decision with respect to securities of or
derivatives linked to the Covered Issuer.
This report represents the views of Muddy Waters Research only and is based on publicly available information. To the best of our
knowledge, all information contained herein is accurate and reliable and has been obtained from publicly available sources that we
believe to be accurate and reliable. The information presented herein is “as is,” without warranty of any kind, whether express or
implied. This report contains a large measure of analysis and opinion. All expressions of opinion are subject to change without notice.
By viewing and accessing this report, you further agree to the following terms of use:
Muddy Waters shall not be liable for any claims, losses, costs, or damages of any kind, including direct, indirect, punitive, exemplary,
incidental, special or consequential damages, arising out of or in any way connected with this report. This limitation of liability applies
regardless of any negligence or gross negligence of Muddy Waters. You accept all risks in relying on the information and opinions in
this report.
You agree that any dispute between you and Muddy Waters arising from or related to this presentation shall be governed by the laws of
the State of Texas, without regard to any conflict of law provisions. You knowingly and independently agree to submit to the personal
and exclusive jurisdiction of the state and federal courts located in Austin, Texas and waive your right to any other jurisdiction or
applicable law, given that Muddy Waters are based in Austin, Texas.
The failure of Muddy Waters to exercise or enforce any right or provision herein shall not constitute a waiver of such right or provision.
If any provision of these terms of use is found by a court of competent jurisdiction to be invalid, the parties nevertheless agree that the
court should endeavor to give effect to the parties’ intentions as reflected in the other provisions set forth herein, in particular as to this
governing law and jurisdiction provision.
You agree and understand that, by the time you read this report, we may be covering or have already covered (i.e., bought back) our
short position, and we are unlikely to increase our short positions unless it is in our financial interest to do so. You should not make any
investment decision based your interpreted view of our positioning in the Covered Issuer’s securities.
You agree that regardless of any statute or law to the contrary, any claim or cause of action arising out of or related to this presentation
must be filed within one (1) year after the occurrence of the alleged harm that gave rise to such claim or cause of action, or such claim or
cause of action be forever barred.


Page 3 of 54
Summary
Newly assembled evidence drawn entirely from Southern California Edison’s own regulatory filings
indicates the utility consciously disregarded public safety when it failed to consider de-energizing the four
220-kilovolt Transmission Circuits running through Eaton Canyon on the evening of January 7, 20251 —
the transmission corridor that Los Angeles County fire investigators have now formally concluded started
the Eaton Fire.2 If the California Wildfire Fund Administrator finds Edison acted with “conscious or
willful disregard of the rights and safety of others”, it would mean the utility is liable to reimburse the
fund approximately $20 billion, which would be ~$20 per share discounted, post-tax.3 Regardless,
Edison’s failures over years and in the days and hours leading up to the ignition of the Eaton Fire almost
certainly amount to “imprudence”, which would mean a $4.3 billion liability, or approximately $6.50 per
share discounted, post-tax.4 The facts we have uncovered show that the Transmission Circuits’ ignition
of the fire wasn’t only foreseeable – it was preventable; and, Edison seemingly failed at every turn to
avert the disaster.
Edison repeatedly failed to fix its primary fire risk model to account
for extreme wind events
Since 2019, Edison was aware that its fire science program, and its principal risk assessment tool in
particular, were deficient – particularly with respect to high wind events. As a result of these
inadequacies, Edison appears to have not even considered de-energizing the Transmission Circuits in
Eaton Canyon that caused the ignition of the Eaton Fire. The model that determined which circuits were
considered for a shutoff, called the “Fire Potential Index” (or “FPI”), assigned its maximum wind speed
value to sustained winds of 29 mph. In other words, Edison knew that FPI could not distinguish between
a 29 mph breeze and the 50 mph winds forecast for January 7, 2025.

1 The four 220 kV transmission circuits running through Eaton Canyon are Eagle Rock–Mesa, Goodrich–Gould, Mesa–Vincent No. 1 and Mesa–
Vincent No. 2 (together, the “Transmission Circuits”). See S. Cal. Edison Co., Public Safety Power Shutoff (PSPS) Post-Event Report: January
2–17, 2025 Event, at 7 tbl. 1 n.3 (Mar. 3, 2025), https://www.cpuc.ca.gov/-/media/cpuc-website/divisions/safety-and-enforcement-
division/reports/psps-post-event-reports/2025/20250104sce-psps-post-event-reportdeenergizedfinal.pdf [hereinafter Jan. 2025 Post-Event
Report].
2 Press Release, Cnty. of L.A. Fire Dep’t, County of Los Angeles Fire Department Releases the Investigative Reports on the Cause of the Eaton
Fire (Aug. 4, 2026), https://lacounty.gov/2026/08/04/county-of-los-angeles-fire-department-releases-the-investigative-reports-on-the-cause-of-
the-eaton-fire/ [hereinafter LACoFD Eaton Fire Release]; CAL FIRE & Cnty. of L.A. Fire Dep’t, Investigation Report, Case No.
25CALAC009087 (Eaton Fire) (released Aug. 4, 2026),
https://file.lacounty.gov/SDSInter/lac/1214310_EatonArsonInvestigationReportFinal8.3.26redacted.pdf [hereinafter Eaton Fire Investigation
Report].
3 Cal. Pub. Util. Code § 3292(h)(3)(A) (West 2026) (the reimbursement cap does not apply if the Wildfire Fund administrator determines that the
utility’s “actions or inactions that resulted in the covered wildfire constituted conscious or willful disregard of the rights and safety of others”).
The ~$20 billion liability is a Muddy Waters estimate. Reference points, all of which measure insured loss rather than utility liability, include
Milliman’s estimate of $13.7–$22.8 billion of Eaton Fire insured loss and loss adjustment expense. Cal. Earthquake Auth., Wildfire Fund
Administrator’s Annual Report to the California Catastrophe Response Council and the California Legislature, at 9–10 (Aug. 15, 2026),
https://www.cawildfirefund.com/sites/wildfire/files/documents/2026/2026-cwf-annual-report.pdf [hereinafter Wildfire Fund 2026 Annual
Report]. Per-share figures are Muddy Waters estimates using 384,813,979 EIX common shares outstanding as of July 23, 2026, Edison Int’l & S.
Cal. Edison Co., Quarterly Report (Form 10-Q), cover page (quarter ended June 30, 2026),
https://www.sec.gov/Archives/edgar/data/0000827052/000082705226000059/eix-20260630.htm [hereinafter Q2 2026 Form 10-Q], and a $20
billion nominal payment stream ($5 billion in year five, $1 billion a year in years six through fifteen and a further $5 billion in year fifteen),
discounted at 8–9% from 2026.
4 Prudence / imprudence is not a binary decision. A utility found imprudent reimburses the Wildfire Fund for the lesser of the disallowed costs
and the cap; given the size of the Eaton Fire claims, even a partial disallowance would likely reach the ~$4.3 billion cap. So even if Edison is
deemed to be mostly prudent, which we think is unlikely, it will likely still incur a ~$4.3 billion liability. See Cal. Pub. Util. Code §
3292(h)(2)(A)–(C); Edison Int’l, April 2026 Business Update, at 14 (Apr. 28, 2026) (Exhibit 99.1 to Form 8-K),
https://download.edison.com/406/files/20263/3364725cb1f85c4c0b8293fc9b5a77e5/eix-april-2026-business-update.pdf [hereinafter April 2026
Business Update] (“Liability cap of 20% of T&D Equity Rate Base (~$4.3Bn for SCE for 2025)”).


Page 4 of 54
We learned from Edison’s 2022 Wildfire Mitigation Plan that when it began using FPI for PSPS in 2019,
it observed deficiencies in the model, including the 29 mph wind speed cap and an underweighting of
wind speeds. (Edison had adopted FPI in 2018, basing it on an index San Diego Gas & Electric
developed.5) Edison concluded that FPI’s relative weighting of factors “does not reflect the realities” of
how wind “can dominate the fire environment.”6,7 Below is the passage from that plan, which Edison
filed with its regulators in February 2022, nearly three years before the Eaton Fire.

Edison’s 2022 Wildfire Mitigation Plan Update, § 7.3.2.4.1, filed February 2022: the limitations Edison observed in
FPI during the 2019 PSPS events, including that wind speed “should have a higher weighting.” Red boxes added
for emphasis.

5 S. Cal. Edison Co., 2026 Public Safety Power Shutoff Action Plan, § II.D, at 25 (May 28, 2026), https://www.cpuc.ca.gov/-/media/cpuc-
website/divisions/safety-and-enforcement-division/psps/sce-2026-psps-action-plan_28-may-2026.pdf [hereinafter 2026 PSPS Action Plan]
(“SCE’s original FPI was based on SDG&E’s FPI. SCE adopted it in 2018 and began using it for PSPS in 2019.”); see also S. Cal. Edison Co.,
2022 Wildfire Mitigation Plan Update, at 275 (Feb. 18, 2022), https://www.sce.com/sites/default/files/custom-
files/SCE%202022%20WMP%20Update.pdf [hereinafter 2022 WMP Update] (“SCE’s current FPI is based on SDG&E’s index, which was
adopted in 2018 and used for PSPS in 2019.”).
6 2022 WMP Update, supra note 5, at 275–77 (“SCE’s current FPI is based on SDG&E’s index, which was adopted in 2018 and used for PSPS in
2019. During the 2019 PSPS events, SCE observed limitations in its current FPI. SCE added a fuel-loading modifier in 2019 to account for areas
where fuels are sparse and unlikely to support a significant fire. In 2021, SCE calibrated the index and was able to raise FPI thresholds across
much of its HFRA as a result. SCE is looking to improve upon its current FPI in subsequent iterations. For example, with the current FPI, fire
potential is capped at 17 which limits its ability to differentiate high end events. Also, weather, fuel moisture, and green-up are essentially
weighted the same in the FPI, which does not reflect the realities of how these factors each contribute to fire potential. For example, wind speed
should have a higher weighting since wind can dominate the fire environment. Finally, the fuels portion of the index is heavily dependent on live
fuel moisture, but there are other fuel moisture variables to consider that are equally important. . . . In 2022, SCE is running FPI 2.0 in parallel
with the current FPI to demonstrate the difference and improvements over the current index, and will make refinements to FPI 2.0 as needed,
based on its evaluation of the outputs. If FPI 2.0 demonstrates a significant improvement over the current FPI, SCE expects that FPI 2.0 will
replace the current FPI before the start of the 2023 fire season.”).
7 Id. at 97 tbl. SCE 4-18 (“The uncertainty associated with the current FPI is due to the inherent error in the various weather and fuel inputs and
also due to limitations of the model. These limitations include: 1) all three components (weather, fuels, and green-up) are essentially weighted the
same, and 2) the wind speed input is capped at 29 mph which limits the index’s ability to account for wind events stronger than that.”).


Page 5 of 54
Edison was supposedly testing a “FPI 2.0” to adjust for these deficiencies in 2022. But for 2023 through
2028 Edison budgeted only $75,000 for FPI 2.0 “enhancements”. It is obvious from the record that
Edison had not been utilizing FPI 2.0 the night of the fire.8 The original model, which Edison knew to be
deficient, particularly with forecast high winds, was still the gate for Edison’s PSPS process. Below is an
excerpt from the Rate Case Workpapers that Edison filed in 2023, showing Edison’s budget for FPI 2.0
enhancements.

Edison’s 2025 General Rate Case workpapers, filed May 2023: $75,000 for FPI 2.0 in 2023, then zero in every year
through 2028. Red boxes added for emphasis.
Since the fire, Edison has supposedly addressed the wind speed deficiencies in its fire science program.
Edison stated in August 2026 that it now places greater weight on winds. Seemingly the scrutiny that
Edison was under caused it to complete the task in approximately only a year and a half. It stands to
reason that had Edison been similarly motivated at some point in the five years preceding the fire, there
would have been no catastrophe.
Edison did not follow override procedures for the Transmission
Circuits to account for winds forecast to exceed FPI’s capabilities
FPI is one-half of the first step in deciding whether to initiate a Public Safety Power Shutoff (PSPS) on a
circuit. If both the FPI score and forecast wind speeds for a given circuit meet or exceed thresholds that
Edison has established, then Edison places the circuit in scope. When Edison places a circuit in scope, it
means that Edison will initiate a process to consider de-energizing the circuit. Edison’s procedures also
anticipate situations when only one – but not both – of the FPI or forecast wind speed thresholds are
breached. It calls this situation a “rogue circuit”. Edison has procedures for evaluating rogue circuits to
determine whether they should manually be placed in scope. The rogue circuit scenario seems
particularly important, given the known limitations of FPI.

8 S. Cal. Edison Co., 2023–2025 Wildfire Mitigation Plan, Revision 3.1, at 514–16 (Nov. 14, 2024),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2023-2025/SCE%202023-
2025%20WMP%20R3.1%20(November%2014%2C%202024).pdf [hereinafter 2023–2025 WMP R3.1] (stating that SCE “is developing” FPI
2.0); S. Cal. Edison Co., 2025 General Rate Case Workpapers: Wildfire Management, PSPS, and Other Wildfire Activities, SCE-04, Vol. 5, pt. 4,
A.23-05-010 (Cal. P.U.C. filed May 12, 2023) [hereinafter 2025 GRC Workpapers] (budgeting $75,000 for “FPI 2.0 Enhancements” in 2023 and
$0 in each year from 2024 through 2028); see also Jan. 2025 Post-Event Report, supra note 1, at 9 (Visual 1, the FPI equation applied in the
January 2025 event); S. Cal. Edison Co., 2025 Wildfire Mitigation Plan Annual Implementation Report, Docket No. 2025-EC_AIR, Plan
Objective 6, at 5 (Cal. Off. of Energy Infrastructure Safety Apr. 1, 2026),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2023-
2025/SCE%202024%20WMP%20Annual%20Implementation%20Report.pdf [hereinafter 2025 Annual Implementation Report] (FPI 2.0 “was
not fully implemented in 2025”).


Page 6 of 54
The forecast wind speeds for the San Gabriel valley foothills, where Altadena is, were up to 50 mph
sustained and gusts of 80 to 100 mph.9 These each exceeded the scoping thresholds of 40 mph / 58 mph,
respectively.10 An excerpt of an Urgent Fire Weather Message the National Weather Service issued at
2:37 PST on January 7, 2025, for the following ~24 hours is below.

Moreover, Edison has established “activation thresholds” for when either forecast sustained wind speeds
or gusts should automatically place a circuit in scope, if its FPI score is elevated, but below threshold.11
Based on activation thresholds published in 2021 for the Eaton Canyon Transmission Circuits, the
forecast wind speeds met the activation thresholds of 40 mph sustained and 58 mph gusts.12 The
Transmission Circuits’ actual FPI was likely between 12 and 13. It is possible that Edison raised the
Transmission Circuits’ thresholds subsequently. As of 2023, two of the circuits’ thresholds were
unchanged – we have been unable to find more recent thresholds for the Transmission Circuits.

9 Nat’l Weather Serv., L.A./Oxnard, Urgent—Fire Weather Message, RFWLOX (WWUS86 KLOX 072237) (Jan. 7, 2025, 2:37 PM PST),
archived at Iowa Env’t Mesonet, https://mesonet.agron.iastate.edu/wx/afos/p.php?pil=RFWLOX&e=202501072237.
10 S. Cal. Edison Co., Quantitative and Qualitative Factors for PSPS Decision-Making app. (May 2021) [hereinafter Q&Q Factors (2021)]
(activation thresholds of 40 mph sustained / 58 mph gust for the Eagle Rock–Mesa, Goodrich–Gould, Mesa–Vincent No. 1 and Mesa–Vincent
No. 2 circuits); S. Cal. Edison Co., PSPS Event Data Workbook: October 26 – November 2, 2023 PSPS Event, tbl. T03 (Dec. 1, 2023)
[hereinafter Oct. 2023 Event Workbook] (same activation thresholds for Mesa–Vincent No. 1 and No. 2); see also S. Cal. Edison Co.,
Amendments to 2025 PSPS Post-Event Reports, R.18-12-005, at 7 (PDF pagination) (Cal. P.U.C. Feb. 4, 2026),
https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M597/K705/597705050.PDF [hereinafter Feb. 2026 Post-Event Report Amendments] (“Sub-
transmission and transmission circuits also use the NWS High Wind Warning level as their baseline,” defined as 40 mph sustained and 58 mph
gusts).
11 Oct. 2023 Event Workbook, supra note 10, tbl. T03 (listing separate sustained and gust activation thresholds for each circuit). But see S. Cal.
Edison Co., 2020 Wildfire Mitigation Plan Remedial Compliance Plan (July 27, 2020) [hereinafter 2020 Remedial Compliance Plan] (“If the
wind speed forecasts exceed these activation thresholds in a location with elevated FPI, SCE initiates PSPS activation and notifications.”). Note:
“elevated” FPI broadly implies actual FPI that meets or breaches threshold FPI.
12 Q&Q Factors (2021), supra note 10; see also Feb. 2026 Post-Event Report Amendments, supra note 10, at 7 (PDF pagination).


Page 7 of 54
On the day of ignition, winds in Eaton Canyon gusted to 63 mph by 2:20 p.m. and sustained 75 mph at
their peak.13 And yet, even though Edison knew FPI scored a 29-mph breeze and a 100-mph Santa Ana
identically, it seemingly did not place the Eaton Canyon Transmission Circuits “in scope”.
The CPUC had designated the portion of Eaton Canyon with the Transmission Circuits a Tier 3 district
(extreme risk of ignition and rapid, catastrophic fire spread) in its High Fire Threat District framework.
This is the highest potential threat level.
Because the forecast and actual wind speeds for Eaton Canyon met or exceeded the 2021/3 thresholds
while the broken FPI did not, the Transmission Circuits, in our view, should have been deemed rogue
circuits and placed in scope. Yet Edison apparently did not do so. However, on January 7th, there were
three distribution circuits adjacent to the ignition point, which Edison did put in scope despite two of their
FPIs not meeting thresholds. In doing so, Edison appears to have treated them as rogue circuits. After
seemingly manually placing them in scope, Edison ran its second step model, called FireCast, on them.
Edison then de-energized these distribution circuits approximately two to three hours before the fire
ignited.
FireCast screamed out for circuits adjacent to the ignition to be placed
in scope yet Edison still did not do so
After a circuit is in scope, Edison applies a second fire risk model, called FireCast, to it. FireCast weighs
simulated wildfire consequences against the harm of cutting power and produces a ratio. If the ratio is at
or above 1.0, the fire risk exceeds the outage risk, which means that Edison should de-energize the line.
When Edison applied FireCast to the three adjacent distribution circuits (Kinneloa, Lamanda, and Video),
FireCast produced ratios many times above threshold, ranging from 39 to 469. Given those distribution
circuits were adjacent to the ignition point and that their ratios were well above threshold, it seems
unconscionable that Edison did not then place the Transmission Circuits in scope, particularly given that
those three distribution circuits and the Transmission Circuits had the same FPI threshold of 13, which
was Edison’s highest threshold, network-wide.
Edison de-energized Kinneloa, Lamanda and Video two to three hours before the fire ignited, further
calling into question why Edison did not put the transmission lines in scope. The Eaton Canyon map
below shows the three de-energized circuits in red, the Transmission Circuits in blue, and the ignition
point.14

13 Nat’l Weather Serv., L.A./Oxnard, Public Information Statement, PNSLOX (NOUS46 KLOX 071848) (Jan. 7, 2025, 10:48 AM PST), archived
at Iowa Env’t Mesonet, https://mesonet.agron.iastate.edu/wx/afos/p.php?pil=PNSLOX&e=202501071848 (“Eaton Cyn (SCE) 62 MPH 0840
AM”; “Henniger Flats RAWS 79 MPH 0758 AM”); Nat’l Weather Serv., L.A./Oxnard, Public Information Statement, PNSLOX (NOUS46
KLOX 080002) (Jan. 7, 2025, 3:53 PM PST), archived at Iowa Env’t Mesonet,
https://mesonet.agron.iastate.edu/wx/afos/p.php?pil=PNSLOX&e=202501080002 (“Eaton Cyn (SCE) 63 MPH 0220 PM”); Jan. 2025 Post-Event
Report, supra note 1, § 2, Q. 2, at 12 (“Peak observed wind speeds in areas of concern were 75 mph sustained and 100 mph gusts during this
event.”). The 75 mph figure is SCE’s peak across all areas of concern for the January 2–17 event, not an Eaton Canyon reading. The Eaton
Canyon weather station (944 feet) sits below the Transmission Circuits. The Henninger Flats RAWS (2,799 feet), upslope of the Transmission
Circuits running through Eaton Canyon, recorded gusts of 79 mph at 7:58 a.m. and 74 mph at 8:58 a.m. on January 7. Iowa Env’t Mesonet, Iowa
State Univ., HNGC1 (Henninger Flats) Observation History: January 7, 2025,
https://mesonet.agron.iastate.edu/sites/obhistory.php?station=HNGC1&network=CA_DCP&date=2025-01-07.
14 Map by Muddy Waters Research. Circuit routes traced from the map attached to Letter from S. Cal. Edison Co. to Wildfire Safety Enf’t
Branch, Cal. Pub. Utils. Comm’n (Feb. 6, 2025) (Pub. Util. Code § 315 report on the Eaton Fire), https://www.cpuc.ca.gov/-/media/cpuc-
website/industries-and-topics/documents/wildfire/wildfire-2025/eaton-fire---315-letter.pdf [hereinafter Feb. 6, 2025 § 315 Letter]; de-energized
circuits from Jan. 2025 Post-Event Report, supra note 1, att. C, tbls. 2–3; ignition point from Letter from Paul Pimentel, Claims Principal
Manager, S. Cal. Edison Co., to Ed Pike, Wildfire Safety Enf’t Branch, Cal. Pub. Utils. Comm’n (Jan. 27, 2025) (interim update on the Eaton


Page 8 of 54

De-Energizing Eaton Canyon would likely have not led to additional
power outages to Edison customers
The purpose of FireCast is to weigh the cost of a PSPS in terms of loss of electricity to customers against
the simulated consequences of a wildfire. Because of redundancy built into the grid, it appears that if
Edison had de-energized the transmission lines on January 7th, no additional customers would have lost
power. Edison’s load data and the state grid operator’s records suggest the system had room to absorb the
loss of the Eaton Canyon circuits on the evening of January 7.
At 6 p.m., Edison’s demand was approximately 12,000 MW against Edison’s estimated winter capacity of
approximately 20,000 MW. Across CAISO, the state grid operator, demand was approximately 27,000
MW against a 2025 peak of approximately 44,500 MW.15 Neither Edison nor the state’s system appeared
constrained.16 Due to fire damage, the Transmission Circuits locked out (de-energized) anyway that night

Fire), https://www.cpuc.ca.gov/-/media/cpuc-website/divisions/safety-and-enforcement-division/electric-safety-incidents-reported-by-ious/2025-
incidents/eaton-fire-update-1-27-25.pdf [hereinafter Jan. 27, 2025 Interim Letter].
15 Cal. Indep. Sys. Operator, Hourly System Demand Data for the CAISO Balancing Area and S. Cal. Edison Co. (Jan. 7, 2025 and 2025 peak-
demand days), https://www.caiso.com/outlook/history/20250107/demand.csv (CAISO demand of 26,819 MW at 6:00 p.m. on Jan. 7, 2025)
[hereinafter CAISO Demand Data]; Cal. Indep. Sys. Operator, 2025 Statistics (CommPR/02/2026) (2026),
https://www.caiso.com/documents/2025-statistics.pdf [hereinafter CAISO 2025 Statistics] (2025 system-wide peak of 44,506 MW on August 21,
2025). Muddy Waters estimates Edison’s available winter capacity at roughly 18,750–22,500 MW, after a 25% reduction for winter maintenance
outages, based on an interview with a former SCE senior executive. Interview by Muddy Waters Research with Former Senior Executive #1, S.
Cal. Edison Co. [hereinafter SCE Executive #1 Interview].
16 See CAISO Demand Data, supra note 15; SCE Executive #1 Interview, supra note 15.


Page 9 of 54
between approximately 7 pm and 12:30 am.17 A former Edison operations executive told us that each of
the four Eaton Canyon circuits has at least two alternative sources of supply — an N-2 design standard,
meaning the system is built to keep running after losing two transmission lines — and that de-energizing
a circuit “without impacting the function of the grid” is routine.18
Moreover, on January 12th, and on January 20th, Edison de-energized all four Transmission Circuits,
which “did not result in any customer outages”, further showing that de-energizing Eaton Canyon on
January 7th likely would have had no impact on customers.19
Edison was aware it could not disable a recloser, making it even more
dangerous to leave the canyon energized during an extreme wind
storm
Edison’s failure to place the Transmission Circuits in scope appears even more egregious because it
seemingly knew that it was unable to disable automatic reclosers on at least one circuit connected to a
Transmission Circuit that runs through the canyon. Automatic reclosers re-energize a faulted line to test
whether the fault has cleared — routine practice in the case of an ordinary fault, such as one caused by a
bird strike. However, because wind events present the risk of pushing a second surge into a line that may
be arcing, it is standard practice to preemptively disable reclosers during forecast high wind events, and
especially before Red Flag Warnings from the NWS. Edison’s own plan says it has done so “for at least
30 years,” and its internal operating bulletin has included guidance on blocking transmission reclosers
since at least 2019.20
Edison’s February 6, 2025, incident report to the CPUC states that on the adjacent Eagle Rock–Gould
circuit at approximately 6:11 p.m. a fault was detected, and “an earlier attempt to disable the automatic
reclosing function was unsuccessful; therefore, it automatically tested the line to determine if the fault had
cleared.”21 Surveillance video from the Gerrish Swim and Tennis Club in Pasadena shows two flashes at
the ignition point roughly three seconds apart, matching the two faults in Edison’s own records.22
A former San Diego Gas & Electric (SDG&E) operator, shown the sequence, said a line under a
Particularly Dangerous Situation Red Flag Warning “should be on zero shot,” that is it should be

17 Feb. 6, 2025 § 315 Letter, supra note 14, at 3 (Mesa–Vincent No. 2 relayed at approximately 7:17 p.m. on January 7; Mesa–Vincent No. 1 and
Goodrich–Gould at approximately 10:35 p.m.; Eagle Rock–Mesa at approximately 12:35 a.m. on January 8).
18 SCE Executive #1 Interview, supra note 15. An N-2 contingency standard means the transmission system is designed to remain operational
after the loss of two transmission lines.
19 Jan. 2025 Post-Event Report, supra note 1, at 7 & tbl. 1 n.3.
20 2023–2025 WMP R3.1, supra note 8, at 333 (“for at least 30 years”); S. Cal. Edison Co., 2020–2022 Wildfire Mitigation Plan, Revision 3, at
162 (Mar. 18, 2020), https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/SCE's%202020-
2022%20Wildfire%20Mitigation%20Plan%20-%20Revision%2003.pdf [hereinafter 2020–2022 WMP Rev. 3]; S. Cal. Edison Co., 2019 Wildfire
Mitigation Plan Progress Update (as of Dec. 31, 2019), R.18-10-007, at 3 (OP-1) (Cal. P.U.C. filed Jan. 15, 2020), https://www.cpuc.ca.gov/-
/media/cpuc-website/industries-and-topics/documents/wildfire/wildfire-2019/r1810007-sce-2019-wmp-progress-update.pdf [hereinafter 2019
WMP Progress Update] (revising SOB 322 to include “blocking transmission reclosers”).
21 Feb. 6, 2025 § 315 Letter, supra note 14, at 2.
22 Kevin Ozebek, New Surveillance Video Shows What Could Be the Start of the Eaton Fire, Attorneys Say, ABC7 L.A. (June 10, 2026, 11:34
AM PDT), https://abc7.com/post/new-surveillance-video-shows-flashes-origin-eaton-fire-attorneys-say/19265902/; Melody Petersen, New
Evidence Confirms Edison’s Idle Line Ignited Eaton Fire, Lawyers Say, L.A. Times (June 9, 2026).


Page 10 of 54
disabled. A second former SDG&E operator said that opening the breaker, via digital communications,
can generally be done “quickly and remotely.” 23
The fact that Edison had been unable to disable the reclosers on a connected circuit and still did not place
the Transmission Circuits in scope adds to the weight of our belief that Edison acted with conscious and
willful disregard of the rights and safety of others leading up to the fire.
The likely direct cause of the fire, induction, was a risk of which
Edison was well aware
The leading theory is that the direct cause of the ignition was induction from one or more of the energized
Transmission Circuits to an adjacent idle transmission line that had been out of service since 1971; and,
that the current then arced to the base of the idle line’s tower, igniting the brush.24,25,26 Edison “considers
an induction hazard to exist wherever transmission lines run parallel for over 600 feet or when the de-
energized conductors are located within transmission/sub-transmission corridors.” 27
Eaton Canyon met both prongs of the induction risk test. The idle Mesa–Sylmar conductor sat inside a
transmission corridor, which alone meets the second prong. And the parallel spans running to the origin
tower measure roughly 1,150, 1,800 and 2,600 feet — two, three and four times the 600-foot length at
which Edison’s own standard says the hazard exists.28

23 Interview by Muddy Waters Research with Former Operator #1, San Diego Gas & Elec. Co. [hereinafter SDG&E Operator #1 Interview];
Interview by Muddy Waters Research with Former Operator #2, San Diego Gas & Elec. Co. [hereinafter SDG&E Operator #2 Interview].
24 LACoFD Eaton Fire Release, supra note 2 (investigators found the fire was caused by “electrical arcing events that took place on the out-of-
service Southern California Edison (SCE) tower”).
25 Steven Lawrence, Eaton Fire Investigation Concludes: Electrical Arcing on Idle Edison Tower Caused the Catastrophic Blaze That Killed 19
& Wiped Out Altadena, S. Pasadenan (Aug. 4, 2026), https://southpasadenan.com/eaton-fire-investigation-concludes-electrical-arcing-on-idle-
edison-tower-caused-the-blaze-that-killed-19/; Petersen, supra note 22; Melody Petersen, Key Takeaways from Investigation into Edison’s Role
in the Deadly Eaton Fire, L.A. Times (Aug. 20, 2026) (reporting that the idle line “hadn’t carried power since 1971”).
26 Jan. 27, 2025 Interim Letter, supra note 14 (reporting the County’s “preliminary origin area . . . around coordinates N34.1860422292
W118.09357612511549 in Eaton Canyon” and identifying towers M6T1, M24T3 and M16T1, the last carrying the idle Mesa–Sylmar line);
Eaton Fire Investigation Report, supra note 2.
27 Cal. Pub. Utils. Comm’n, Safety & Enf’t Div., Incident Investigation Report: Incident No. E20180118-01, at 7 (Nov. 9, 2021) (Jan. 18, 2018
Kramer Junction contractor fatality), https://www.cpuc.ca.gov/-/media/cpuc-website/divisions/safety-and-enforcement-
division/documents/citations/sce/kramer-junction/enclosure-1--sed-incident-investigation-report-dated-november-9-2021_redacted2.pdf
[hereinafter Kramer Junction Investigation Report].
28 Muddy Waters measurement, using Google Earth (image date Dec. 2023), of the parallel spans to towers T1, T2 and T3; the measurements are
straight-line and therefore conservative. See Kramer Junction Investigation Report, supra note 27, at 7 (600-foot standard).


Page 11 of 54

Measured spans of the parallel transmission lines running to the origin tower — roughly two to four times the 600-
foot length at which Edison’s own standard deems an induction hazard to exist.
In January 2018, at Kramer Junction, an Edison contract worker died when an energized 220-kV circuit
built parallel to a de-energized conductor induced a current into it during a transfer operation. The
CPUC’s Safety and Enforcement Division investigated for three years and cited Edison $4.5 million.29
Edison is attempting to evade accountability
The company’s formal regulatory position remains that it “will be able to make a good faith showing that
[SCE’s] conduct with respect to its transmission facilities in the preliminary area of origin was consistent
with the actions of a reasonable utility.”30 But on top of the foregoing inactions leading up to the tragedy,
on January 7th, Edison also knew the Palisades Fire was already burning out of control and that regional
suppression resources were committed — a factor its own plan lists as relevant to shutoff decisions.31 It
further knew, an hour before ignition, that the City of Los Angeles had declared a state of emergency.32
Twenty-two months before the fire, Edison had told a CPUC workshop that shutoffs “could have reduced
the number of structures damaged by past fires by more than 90%,” measured across its own wildfires
larger than 1,000 acres from 2015 to 2020.33 Yet, Edison did not place the transmission lines in scope.

29 Press Release, Cal. Pub. Utils. Comm’n, CPUC Staff Propose $4.5 Million SCE Penalty for Safety Failure (Dec. 16, 2022),
https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-staff-propose-4-million-sce-penalty-for-safety-failure. The citation was issued under
General Order 95, Rule 31.1, for allowing the contractor to proceed without meeting three requirements of SCE’s Contractor Safety Management
Standard; the induction finding appears in the underlying investigation report rather than as the charged violation. See Kramer Junction
Investigation Report, supra note 27, at 7.
30 Q2 2026 Form 10-Q, supra note 3, MD&A (“Eaton Fire”) (“Based on the information it has reviewed, SCE believes that it will be able to make
a good faith showing that its conduct with respect to its transmission facilities in the preliminary area of origin was consistent with the actions of
a reasonable utility.”); Edison Int’l & S. Cal. Edison Co., Quarterly Report (Form 10-Q), MD&A (“Eaton Fire”) (quarter ended Mar. 31, 2026),
https://www.sec.gov/Archives/edgar/data/827052/000082705226000042/eix-20260331.htm (same); see also Edison Int’l, Second Quarter 2026
Financial Results, slide 15 (July 30, 2026) (Exhibit 99.3 to Form 8-K),
https://www.sec.gov/Archives/edgar/data/827052/000082705226000058/eixq22026earningstelecon.htm.
31 2023–2025 WMP R3.1, supra note 8, at 514, 624, 627; Jan. 2025 Post-Event Report, supra note 1, at 13 (“Multiple fires and lack of aerial
suppression resources factored into adjustment of FPI and de-energization prior to reaching wind thresholds.”).
32 Press Release, Off. of Mayor Karen Bass, City Declares Local Emergency in Response to Palisades Fire and Ongoing Wind Event, President
Biden Pledges Support for Impacted Areas (Jan. 7, 2025), https://mayor.lacity.gov/news/city-declares-local-emergency-response-palisades-fire-
and-ongoing-wind-event-president-biden; Off. of the Mayor, City of L.A., Declaration of Local Emergency, Council File No. 25-0030 (Jan. 7,
2025, 5:04 PM) (signed by Acting Mayor Marqueece Harris-Dawson), https://cityclerk.lacity.org/onlinedocs/2025/25-0030_rpt_mayor_01-07-
24.pdf.
33 S. Cal. Edison Co., Presentation at the CPUC Fast Trip, Unplanned Outages, and Distribution Reliability Workshop (Mar. 17, 2023)
(presentation by R. Fugere, Principal Manager, Wildfire Mitigation Strategy).


Page 12 of 54
Edison instead has pointed the finger of blame in several directions. Edison has countersued Los Angeles
County, Pasadena Water and Power, five other water agencies and Southern California Gas, alleging
delayed evacuation warnings, inadequate firefighting water and late gas shutoffs.34
Despite the Eaton Fire killing 19 people, Pedro Pizarro’s compensation rose 19.8% in 2025 to
$16,544,559.35 The board suggested that Edison’s senior executives were also victims of the fire when it
noted in the proxy that their equity had “shed nearly 40% of their value” since the fire.36 In particular, the
board saw no reason to dock executives’ safety-related compensation. Both Edison’s 2025 and 2026
Proxy Statements establish that “Safety and compliance are therefore foundational and events such as
fatalities or significant non-compliance issues can result in the reduction or elimination of annual
incentive awards for all or some plan participants, depending upon the Committee’s assessment of the
circumstances.”37 The 2025 scoring matrix shows that there were “no qualifying events” and therefore
“no deduct” for serious injuries to the public from system failure or significant non-compliance events.38

Edison’s 2025 corporate performance scoring matrix. Red boxes added for emphasis.
Meanwhile, certain safety targets themselves moved in management’s favor. The covered-conductor
installation goal — miles of insulated wire, the company’s flagship hardening measure — was set at
1,050 circuit miles for 2024, against which Edison installed 813 and scored zero out of six. For 2025, the
target was cut to 650 miles, a figure below what the company had actually built the year before. Edison
installed 690 and scored eight out of six — above maximum.39 The 650-mile target was approved by the
committee in February 2025, the month after the fire.40
We believe the facts establish that in failing to update FPI for years, and in failing to manually place the
Eaton Canyon Transmission Circuits in scope on January 7, 2025, despite all that Edison knew, it acted

34 SoCal Edison Sues LA County, Several Agencies over Eaton Fire Responsibility, NBC L.A. (Jan. 17, 2026, 11:28 PM),
https://www.nbclosangeles.com/news/local/socal-edison-sues-la-county-eaton-fire/3833083/ (describing SCE’s complaint against Los Angeles
County agencies, the alert vendor Genasys, Pasadena Water & Power and five other water agencies, and a separate complaint against Southern
California Gas Co.).
35 Edison Int’l, Proxy Statement (Schedule 14A), at 61 (Mar. 13, 2026),
https://www.sec.gov/Archives/edgar/data/827052/000110465926027655/eix-20260423xdef14a.htm [hereinafter 2026 Proxy] (Summary
Compensation Table). Pizarro’s 2024 total compensation was $13,809,571.
36 Id. at 42.
37 Id. at 43; Edison Int’l, Proxy Statement (Schedule 14A), at 37 (Mar. 14, 2025),
https://www.sec.gov/Archives/edgar/data/827052/000155837025003002/eix-20250424xdef14a.htm [hereinafter 2025 Proxy].
38 2026 Proxy, supra note 35, at 45, 47 (2025 EIX and SCE Corporate Performance Scoring Matrices — Goal Details). The “Actual Score” cell is
merged across all four foundational goals and reads “No Deduct.”
39 Id.; 2025 Proxy, supra note 37, at 38, 40.
40 2026 Proxy, supra note 35, at 43.


Page 13 of 54
with willful and conscious disregard for the rights and safety of others. When viewing the callous lack of
acknowledgement by Edison’s board of directors of the company’s safety failures, it seems that
significant financial penalties are needed to change the culture of Edison and reduce the risk of it starting
another catastrophic fire.


Page 14 of 54
Table of Contents
Part 1: Edison Knew for Years That the Models Determining Whether to Keep the Eaton Canyon
Transmission Circuits Energized Were Faulty ........................................................................ 15
1.1 FPI was central to the PSPS process Edison says it follows, including for the Transmission Circuits ......... 15
1.2 FPI could not measure winds above 29 mph; Edison knew that in 2019, claimed it was developing a
replacement, and apparently defunded it after 2023 ............................................................................................ 16
1.3 After the Fire, Edison admitted its model could not “capture extreme wind events” and recast the 2019
defect as a 2025 “learning” ................................................................................................................................... 19
1.4 The Transmission Circuits should have been in-scope before January 7th .................................................... 20
1.5 Same threshold, different treatment: Edison overrode FPI 13 for the circuits it was watching and seemingly
never scored the Transmission Circuits at all ....................................................................................................... 24
1.6 Edison failed to consider FireCast, a second decision tool that was signaling to consider the Transmission
Circuits for PSPS, because FireCast requires FPI at threshold ............................................................................ 25
1.7 Another risk Edison knew, but failed to act upon by placing the Transmission Circuits in-scope: A recloser
could not be disabled ............................................................................................................................................ 27
1.8 Other factors that evening Edison knew that called for manually placing the Transmission Circuits in-scope
and de-energizing them: the Palisades Fire, no air support, a lightly loaded grid, and a redundant network ...... 29
Part 2: Edison Shareholders Could Be Liable for ~$20 Billion. Imprudence Seems Near Certain; Willful
Disregard Is Likely in Our Opinion ...................................................................................... 31
2.1 The framework: AB 1054, the Wildfire Fund, and the prudence inquiry ...................................................... 31
2.2 Imprudence and willful disregard: what Edison’s plan required, and what Edison knew ............................. 32
2.3 Damages – Imprudence Versus Willful Disregard: a ~$15 billion question .................................................. 34
2.4 Scenario A – mere imprudence (or partially prudent as prudent / imprudent decision is not binary): ~$4.3
billion, ~$6 to $6.50 per share, post-tax, discounted ............................................................................................ 35
2.5 Scenario B – willful disregard: ~$20 billion, ~$20 to $21 per share, post-tax, discounted ........................... 35
APPENDIX ................................................................................................................... 36
A. Edison’s statements on FPI 2.0, 2019–2026 .................................................................................................... 36
B. Exhibits 7–19, referenced in Part 1 and Appendix A ...................................................................................... 40




Page 15 of 54
Part 1: Edison Knew for Years That the Models
Determining Whether to Keep the Eaton Canyon
Transmission Circuits Energized Were Faulty
Edison’s public narrative is that it ran its process, and that its models did not call for shutting off the four
220 kV Transmission Circuits running through Eaton Canyon because an unspecified “threshold” was not
met. The four circuits are Eagle Rock–Mesa, Goodrich–Gould, Mesa–Vincent No. 1 and Mesa–Vincent
No. 2 (together, the “Transmission Circuits”). In Edison’s telling, an idle line was likely induced.41 The
fine print in Edison’s regulatory record repeatedly demonstrates a persistent and knowing failure to
maintain a fire-science modeling program that could adequately prepare for the risk of extreme wind
events.
Edison’s PSPS process broadly has two parts. First, forecast wind speed and the FPI are the two central
inputs that decide which circuits come “in scope” for potential de-energization. In Edison’s words,
transmission lines come into scope when forecasted windspeeds and Fire Potential Index (FPI) exceed
PSPS thresholds.42 Second, real-time readings against de-energization thresholds decide whether an in-
scope circuit is shut off.43 Had the Transmission Circuits been in-scope, we believe Edison would have
de-energized them and avoided the tragedy.
For at least five years, Edison knew its FPI was inadequate for extreme wind events. This inadequacy
appears to be the reason for why Eaton Canyon was not in-scope for a PSPS on January 7th, 2025.
However, knowing the inadequacy, and several other factors, in our opinion, Edison acted with willful
disregard by not placing the Canyon Transmission Circuits in-scope, let alone de-energizing them.
We believe the forecast winds cleared the activation thresholds (i.e., the thresholds that determine what is
in-scope) that should have put them in-scope. Every indication is that Edison kept the circuits out of
scope.
1.1 FPI was central to the PSPS process Edison says it follows,
including for the Transmission Circuits
Edison’s Wildfire Mitigation Plan (“WMP”), which it files with its regulators, describes a sequential
PSPS decision process.44 Transmission Circuits are scoped in for PSPS when forecast wind speeds and
FPI exceed PSPS thresholds.45 Edison’s published PSPS timeline begins planning four to seven days
before a potential PSPS event. It activates an Incident Management Team (“IMT”), which is a group that

41 Cameron Kiszla, Power Company CEO Indicates Fault for Eaton Fire, KTLA 5 (Dec. 17, 2025, 10:21 AM PST), https://ktla.com/news/local-
news/power-company-ceo-indicates-fault-for-eaton-fire/ (quoting Pizarro that SCE equipment was “likely the cause” of the fire); see also Q2
2026 Form 10-Q, supra note 3, MD&A (“Eaton Fire”) (describing “a de-energized idle SCE transmission facility in the preliminary area of
origin” and stating that “it is likely that its equipment was associated with the ignition”).
42 S. Cal. Edison Co., Amendment to 2025 PSPS Post-Season Report, R.18-12-005, at 134 (Cal. P.U.C. July 24, 2026), https://www.cpuc.ca.gov/-
/media/cpuc-website/divisions/safety-and-enforcement-division/reports/2025-psps-post-season-reports/r1812005-sce-amended-2025-psps-post-
season-report.pdf [hereinafter 2025 Post-Season Report Amendment]. The same protocol appears in prior-year post-season reports. See, e.g., S.
Cal. Edison Co., 2023 PSPS Post-Season Report, R.18-12-005, § III.E.2, at 23 (Cal. P.U.C. Mar. 1, 2024), https://www.cpuc.ca.gov/-
/media/cpuc-website/divisions/safety-and-enforcement-division/reports/2023-psps-post-season-reports/sce_postsr1_3-1-2024.pdf [hereinafter
2023 Post-Season Report] (stating that transmission lines are in scope if forecast wind speeds and FPI exceed PSPS thresholds).
43 SCE Executive #1 Interview, supra note 15; Jan. 2025 Post-Event Report, supra note 1, at 8 & n.5.
44 2023–2025 WMP R3.1, supra note 8, at 625 fig. SCE 9-02 (“PSPS Decision Making Flowchart / Diagram”); see also Jan. 2025 Post-Event
Report, supra note 1, § 2, at 8–16.
45 2025 Post-Season Report Amendment, supra note 42, at 134; see also 2023 Post-Season Report, supra note 42, § III.E.2, at 23.


Page 16 of 54
makes de-energization decisions.46 Given the forecast as of January 3, 2025, Edison had four days to
bring the circuits into scope for shut-off.47
1.2 FPI could not measure winds above 29 mph; Edison knew that in
2019, claimed it was developing a replacement, and apparently
defunded it after 2023
FPI is a relatively simple formula.48 It adds a fuels term (dryness, live fuel moisture, green-up and a fuel-
loading modifier) to a weather score, Wx. Wx is read from a lookup table of sustained wind speed
against dew point depression. Problematically, the table’s highest wind band is “≥29 mph.”49 The
National Weather Service (“NWS”) calls a 29-mph wind a “strong breeze.”50 The forecast issued at 2:37
PST on January 7th repeated prior Red Flag Warnings calling for the most destructive windstorm since
2011, an increased risk for large fires with very rapid spread, and extreme fire behavior in the foothills of
the San Gabriel valley:51
“The strongest winds with this event are expected to be this evening [January 7] into
early Wednesday afternoon when widespread damaging wind gusts of 50 to 80 mph are
likely. The San Gabriel mountains, Santa Susana mountains, and foothills of the San
Gabriel/San Fernando Valleys will likely see areas of destructive wind gusts between 80
and 100 mph!...
THIS WILL LIKELY BE THE MOST DESTRUCTIVE WINDSTORM SEEN SINCE 2011
WINDSTORM THAT DID EXTENSIVE DAMAGE TO PASADENA AND NEARBY
FOOTHILLS OF THE SAN GABRIEL VALLEY. ANY COMMUNITIES ALONG

46 2023–2025 WMP R3.1, supra note 8, at 532 fig. SCE 8-52a (PSPS flowchart).
47 Id. at 630 (“At up to 72 hours before a forecast potential PSPS event (event), SCE meteorologists and fire scientists continue to review weather
conditions, using both internal and external weather models and National Weather Service forecasts, alerts, and warnings. The PSPS Incident
Management Team (IMT) is activated and develops a list of circuits that could potentially be impacted by a PSPS event based on the forecast and
predictive modeling.”). SCE has full authority over its transmission network: in 2019, Edison told the CPUC that its own incident teams would
decide on transmission shutoffs and that CAISO had agreed that “if there is a required deenergization, CAISO would call for an outage in
response to an SCE request.” S. Cal. Edison Co., Progress Report on the Implementation of De-energization Guidelines Set Forth in Appendix
“A” of Decision 19-05-042, R.18-12-005, § VIII.B, at 24 (Cal. P.U.C. Sept. 4, 2019), https://www.sce.com/sites/default/files/inline-
files/SCE_PSPSProgressReport.pdf [hereinafter Sept. 2019 De-energization Progress Report]. Edison’s 2022 exercise script and April 2024
tabletop each addressed CAISO coordination. S. Cal. Edison Co., 2022 Functional Exercise Ground Truth, at 4; S. Cal. Edison Co., 2024
Tabletop Breakout, slide 5 (Apr. 2024); see also S. Cal. Edison Co., 2022 Tabletop Situation Manual, at 29.
48 S. Cal. Edison Co., Response to Data Request CalAdvocates-SCE-2020WMP-03 (Mar. 18, 2020). Question: “The following questions relate to
SCE’s Fire Potential Index (FPI) model. a. Is the FPI model based on physical modeling, statistical analysis, machine learning, or a different
modeling technique? b. Has SCE verified or validated the accuracy and precision of its FPI model? If so, please provide a validation analysis,
with plots or graphs, showing the accuracy and error of SCE’s FPI model.” SCE’s response: “a. The FPI is a linear equation developed by
SDG&E that incorporates dead and live fuel moisture, the state of green-up of the annual grass, and weather conditions which consists of wind
speed and dew point depression. This equation is loosely based on a physical model in the sense that all inputs influence the potential for a large
fire to occur. The FPI also includes a fuel loading modifier to adjust the index downward (towards lower fire potential) in areas where vegetation
is sparse. This modifier incorporates a geospatial fuel loading layer to determine the amount of vegetation across each SCE circuit within its
HFRA which helped to determine whether the load would be light, moderate, or heavy. b. SCE has not been able to validate the accuracy of the
FPI due to a lack of historical data. This historical data is necessary for backcasting the index and comparing to historical fire data. For the time
being, SCE has relied on SDG&E’s index calibration until sufficient data can be acquired to perform a thorough calibration and analysis of the
index across SCE’s service territory.”
49 Jan. 2025 Post-Event Report, supra note 1, at 9 (Visual 1, “Fire Potential Index Equation”); see also 2022 WMP Update, supra note 5, at 97
tbl. SCE 4-18 (wind speed input “capped at 29 mph”).
50 Nat’l Weather Serv., Beaufort Wind Scale, https://www.weather.gov/mfl/beaufort (last visited Oct. 10, 2026) (Force 6, “Strong Breeze,” 25–31
mph).
51 Urgent—Fire Weather Message (Jan. 7, 2025), supra note 9; see also Nat’l Weather Serv., L.A./Oxnard, Area Forecast Discussion, AFDLOX
(FXUS66 KLOX 070117) (Jan. 6, 2025, 5:17 PM PST) (“Fire Weather” section) (Gomberg), archived at Iowa Env’t Mesonet,
https://mesonet.agron.iastate.edu/p.php?pid=202501070117-KLOX-FXUS66-AFDLOX (similar language).


Page 17 of 54
HIGHWAY 118 AND 210 CORRIDORS WILL BE AT HIGHEST RISK FOR
COMPARABLE WIND DAMAGE...
AREAS IN A RED FLAG WARNING (ESPECIALLY DURING THE PDS TIME
FRAME) WILL HAVE INCREASED RISK FOR LARGE FIRES WITH VERY RAPID
FIRE SPREAD, EXTREME FIRE BEHAVIOR...”
The NWS began forecasting an imminent high wind and high fire risk event for Los Angeles and Ventura
Counties beginning on January 3rd. The Red Flag Warning was officially made on January 5th. The high
wind / high fire forecasts for the San Gabriel foothills became stronger each of the preceding days.52
On January 7th sustained winds were far above 30 mph. The Cal Fire (California Department of Forestry
and Fire Protection) investigator reported “the approximate wind speeds at the time of the fire in the area
of the transmission towers were between 35 and 65 mph, with gusts up to 80 mph.”53 Edison’s own post-
event report records peaks of 75 mph sustained and 100 mph gusts during the period of the event
(covering January 2nd to January 17th).54 Those winds could contribute, at most, 6 out of a possible
maximum of 17 points to FPI. In practice, the cap meant that FPI understates risks from high wind
events. As a result, scoping decisions that should have been automatic often needed human override.
For more than five years prior to the Eaton Fire, Edison repeatedly told its regulator the fix to FPI was
coming, moved the goal posts, and then apparently stopped funding it (Appendix A sets out the detailed
chronology with full footnotes). During 2019 PSPS events, Edison noted the wind speed cap was
problematic and that “wind speed should have a higher weighting” in the model (Exhibit 16). Also in
2019, Edison told regulators it was working on improving FPI.55 Likely the principal reason for seeking
to improve it was to correct for its inability to fully account for high winds (Exhibit 15). By the end of
2020, Edison said development of the new version of FPI was complete. In 2021, it stated the new
version would replace the old one by the end of 2022. But at the same time, in September 2021, it
somehow justified increasing FPI scoping thresholds.56

52 Nat’l Weather Serv., L.A./Oxnard, Urgent—Fire Weather Messages, RFWLOX (Jan. 3–7, 2025), archived at Iowa Env’t Mesonet,
https://mesonet.agron.iastate.edu/cgi-bin/afos/retrieve.py?pil=RFWLOX&sdate=2025-01-03&edate=2025-01-
07&fmt=text&limit=20&order=asc; see especially Nat’l Weather Serv., L.A./Oxnard, Urgent—Fire Weather Message (Red Flag Warning),
RFWLOX (WWUS86 KLOX 052240) (Jan. 5, 2025, 2:40 PM PST), archived at Iowa Env’t Mesonet,
https://mesonet.agron.iastate.edu/wx/afos/p.php?pil=RFWLOX&e=202501052240 (first Red Flag Warning, “IN EFFECT FROM 10 AM
TUESDAY TO 6 PM PST THURSDAY”).
53 Eaton Fire Investigation Report, supra note 2, at 53–54.
54 Jan. 2025 Post-Event Report, supra note 1, § 2, Q. 2, at 12 (“Forecasted versus actual weather parameters”) (forecasting 25–50 mph sustained
winds, isolated areas to 60 mph, and gusts of 45–80 mph, isolated to 100 mph; “Peak observed wind speeds in areas of concern were 75 mph
sustained and 100 mph gusts during this event.”).
55 S. Cal. Edison Co., 2019 Wildfire Mitigation Plan, R.18-10-007, at 63 (Cal. P.U.C. filed Feb. 6, 2019),
https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M263/K645/263645320.PDF [hereinafter 2019 WMP] (“In 2019 SCE will begin Phase II of
the FPI project intended to increase capability by adding more granular weather data, expanding the coverage to all of SCE’s service territory,
and integrating historical weather data.”); see also Decision on Southern California Edison Company’s 2019 Wildfire Mitigation Plan Pursuant
to Senate Bill 901, D.19-05-038, R.18-10-007, § 8.1, at 30 (Cal. P.U.C. May 30, 2019), https://energysafety.ca.gov/wp-
content/uploads/docs/misc/docket/298487636.pdf (same sentence, with “its FPI project”); S. Cal. Edison Co., Advice Letter 4153-E, exh. A
(“SCE’s 2019 Wildfire Mitigation Plan (WMP) Progress Update – Dec 31, 2019”), at 6 (Jan. 31, 2020), https://energysafety.ca.gov/wp-
content/uploads/docs/wmp-2020/news/2020_01_31_sce-al-4153-e.pdf (listing “Fire Potential Index Phase 2: Enhance Capabilities of FPI (SA-2)”
as complete); 2019 WMP Progress Update, supra note 20, at 6 (“FPI Phase 2 consisted of the FPI being calculated at the circuit level across its
HFRA below 6,000 feet.”).
56 See 2019 WMP, supra note 55, at 63 (FPI Phase II); S. Cal. Edison Co., 2020 Wildfire Mitigation Plan Annual Report on Compliance (EC
ARC) Pursuant to PUC Section 8386.3(c)(1), att. A, at 7 (Mar. 31, 2021), https://energysafety.ca.gov/wp-
content/uploads/docs/misc/wsd/sce_2020-arc_20210331.pdf [hereinafter 2020 Annual Report on Compliance] (“Completed development of FPI
2.0”); S. Cal. Edison Co., 2021 Wildfire Mitigation Plan Update (Revision), at 107 (June 3, 2021),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2021/SCE%202021%20WMP%20Update%20Revision_Redline.p
df [hereinafter 2021 WMP Update (Rev.)] (“FPI 2.0 to replace current FPI” by the end of 2022); 2022 WMP Update, supra note 5, at 548 (PSPS


Page 18 of 54
The replacement, supposedly five years in the making, did not arrive in time for Eaton. Instead, Edison
appears to have defunded the project in 2023. It omitted the launch date, and its 2025 General Rate Case
(“GRC”), filed that May, budgeted $75,000 for FPI 2.0 in 2023 and nothing in any year from 2024
through 2028.57 In the same rate case, in which Edison budgeted $5 million for two supercomputers,
Edison called its 2021 fire-science spending, which ran 39% over budget, “necessary and prudent.”58 The
machines were to be, at least in part, used to compute FPI (Exhibit 1).
Exhibit 1: Edison’s 2025 GRC workpapers, filed May 2023. The O&M forecast (top) budgets $75,000 for “FPI 2.0
Enhancements” in 2023 and $0 in each year from 2024 through 2028. The capital forecast (bottom) budgets
$2,500,000 for a replacement supercomputer in 2025 and another $2,500,000 in 2027, the machines used to
compute FPI (see Appendix A). Boxes added by MW.


Source: S. Cal. Edison Co., 2025 General Rate Case Workpapers: Wildfire Management, PSPS, and Other Wildfire Activities, SCE-04, Vol. 5, pt.
4, A.23-05-010 (Cal. P.U.C. filed May 12, 2023).
In August 2026, Edison told a CPUC safety briefing that the latest FPI version could capture extreme
events, with “no cap on the weather component [i.e., wind score combined with dew point depression].”
Around the same time, it told a community meeting that, “Beginning in early September [2026], wind

activation threshold raised from FPI 12 to 13 “for most areas and most events”); Letter from Kevin M. Payne, S. Cal. Edison Co., to Marybel
Batjer, President, Cal. Pub. Utils. Comm’n, at 3 (Sept. 16, 2021), https://www.cpuc.ca.gov/-/media/cpuc-website/consumer-
support/documents/psps/sce-response-to-cpuc-aug-31-2021-letter.pdf [hereinafter Payne Letter] (“Starting September 1, 2021, SCE has set the
FPI at 13 for most areas and most events.”); infra Appendix A.
57 2025 GRC Workpapers, supra note 8.
58 S. Cal. Edison Co., Amended Testimony: Wildfire Management, Part 4: PSPS and Other Wildfire Activities, SCE-04, Vol. 5, pt. 4A, A.23-05-
010, § I.A.5.c(1), at 98, lines 1–5 (Cal. P.U.C. filed Nov. 20, 2023) (amending pt. 4, filed May 12, 2023),
https://docs.cpuc.ca.gov/PublishedDocs/SupDoc/A2305010/6905/520913742.pdf (“SCE was authorized $4.342 million for 2021 O&M expense
in connection with Fire Science and Advanced Modeling. SCE incurred recorded costs of $6.049 million due to necessary and prudent spending
on Advanced Modeling Computer Hardware, Fire Science Enhancements, the Fuel Sampling program, Remote Sensing, Surface and Canopy
Fuels Mapping, Fire Spread Modeling, and Weather and Fuels Modeling activities.”); see also id. at 98 fig. I-15. The 39% overspend is Muddy
Waters’ calculation from those two figures ($6.049 million recorded vs. $4.342 million authorized).


Page 19 of 54
carries more weight in the PSPS formula to align more closely with wildfire risk.”59 (emphasis added by
MW; see Exhibits 18 and 19)
1.3 After the Fire, Edison admitted its model could not “capture
extreme wind events” and recast the 2019 defect as a 2025 “learning”
Edison’s own records show that it was still using the flawed FPI with the 29mph wind speed cap on
January 7th 2025.60 On November 14, 2024, 54 days before the Eaton Fire, Edison refiled the nearly
identical WMP passage it had filed in 2021, 2022 and 2023, stating it “is developing” FPI 2.0 to put
“more emphasis on wind speed” and “highlight extreme events.” (Emphasis added.)
Edison’s event data are consistent with that model. Twelve circuits were shut-off, including distribution
and sub-transmission with sustained winds of 40 mph or more, scored between 10.94 and 12.98 (with 12
or 13 thresholds) on FPI’s 1–17 scale. That is not materially higher than the three Altadena circuits
beside the Transmission Circuits, which scored 12.08 to 13.08 with their stations reading about 29 mph
sustained (§ 1.5). Sand Canyon, which was de-energized at 43.5 mph sustained and 74.75 mph gusts,
scored an FPI of only 11.19 (vs a 13 FPI threshold).61
After the Eaton Fire, Edison again conceded that the FPI it used for de-energization decisions in January
2025 was flawed. It told the California Public Utilities Commission (“CPUC”) that it had developed a
new FPI to address limitations in the existing model. Those limitations included its assessment of fuel
conditions and its “inability to capture extreme wind events.”62 (Emphasis added by MW; see Exhibit
7.)
The pre-fire filing, the event data and the post-fire fixes clearly indicate that the model in force during the
fire was the flawed legacy version.
Edison seems to have retroactively created “FPI 1.5” when it wrote that “While FPI 2.0 was not fully
implemented in 2025, key components were operationalized as part of FPI 1.5”.63 It is unclear what
changed with the supposed FPI 1.5 to us. The Office of Energy Infrastructure Safety (OEIS) is asking
Edison to explain its recent changes to FPI seemingly because it is also unclear to it.64

59 S. Cal. Edison Co., Interagency Annual Public Briefing on Safety Culture and Public Safety Power Shutoffs (PSPS), slide 18 (Aug. 20, 2026)
(presentation by T. O’Toole & J. C. Anderson to the CPUC and Energy Safety), https://www.cpuc.ca.gov/-/media/cpuc-website/divisions/safety-
policy-division/meeting-documents/sce-2026-cpuc-safety-culture-briefing_082026_final.pdf [hereinafter 2026 Safety Culture Briefing] (“Ability
to capture extreme events with no cap on the weather component”); S. Cal. Edison Co., Wildfire Safety Community Meeting: Santa Barbara
County, slide 20 (Aug. 27, 2026), https://www.sce.com/sites/default/files/custom-files/PDF_Files/WSCM_Presentation_8_27_Santa_Barbara.pdf
[hereinafter Santa Barbara Community Meeting] (“wind carries more weight”).
60 2023–2025 WMP R3.1, supra note 8, at 514–16; see also Jan. 2025 Post-Event Report, supra note 1, at 9 (Visual 1); 2025 Annual
Implementation Report, supra note 8, Plan Objective 6, at 5 (FPI 2.0 “was not fully implemented in 2025”).
61 Jan. 2025 Post-Event Report, supra note 1, att. C, tbl. 2 (“Factors Considered in Decision to De-Energize”); Muddy Waters analysis of the
workbook. Wind and FPI values are those recorded at the time of the de-energization decision.
62 2025 Post-Season Report Amendment, supra note 42, at 135.
63 2025 Annual Implementation Report, supra note 8, Plan Objective 6, at 5.
64 Cal. Off. of Energy Infrastructure Safety, Decision for the Southern California Edison Company 2026–2028 Base Wildfire Mitigation Plan,
Docket No. 2026-2028-Base-WMPs, § 7.1.3, at 40–41 (Feb. 23, 2026),
https://efiling.energysafety.ca.gov/eFiling/Getfile.aspx?fileid=60317&docket=2026-2028-Base-WMPs&shareable=true (“Further, SCE stated in
its WMP that it last updated its FPI in 2021. This statement appears to contradict SCE’s representation to the Acton community that it recently
altered its ‘FPI criteria.’ However, SCE did not describe any recent changes to the FPI criteria or calculations in its WMP. Therefore, it remains
unclear what changes SCE made to its FPI criteria that would increase PSPS initiations by 20 to 40 percent, as SCE stated in the community
meeting.”); see also id. § 7.3.1, at 44–45 (SCE-26B-10, requiring “[a]ny changes to SCE’s FPI criteria related to or affecting its PSPS program”).


Page 20 of 54
Given Edison’s disclosures to the CPUC, FPI 1.5 did not appear to correct the wind speed flaws.65 In
May 2026, Edison told the CPUC that it made its FPI refinements “primarily based on learnings from the
January 2025 fires.”66 As § 1.2 shows, the supposed effort to fix FPI was six years old and had seemingly
since been defunded. In our opinion, presenting it as a post-fire lesson is revisionism aimed squarely at
minimizing Edison’s culpability.
1.4 The Transmission Circuits should have been in-scope before
January 7th
As far as we can tell from Edison’s filings, Edison did not scope in the Transmission Circuits for de-
energization. We conclude as much from the CPUC post-event inquiries (see § 1.6).
Public commentary from management has not addressed the failure to bring the Transmission Circuits
into scope. Instead, they have redirected the discussion towards the wind thresholds, arguing that the
thresholds had not been met – but they have also yet to produce what those threshold numbers were.
In a December 2025 interview, Edison’s CEO, Pedro Pizarro, was asked why a forecast of “life-
threatening winds” and a 63 mph gust by 2:20 PM were not enough to shut off the transmission lines. He
said the winds “had not hit the criteria — that Threshold criteria — for de-energizing those transmission
lines.” He could not say what those criteria were. He offered a follow-up interview that, to our
knowledge, never took place.67 In a different interview, he said that he didn’t think 100 mph winds were
“fully forecast” at the time of ignition. We think that this is misleading because the forecast clearly called
for the circuits to be in-scope. 68
While Pizarro would not provide the thresholds, we found relevant thresholds in Edison’s regulatory
records. In a July 2020 filing, Edison told regulators that two factors drive its PSPS activations (i.e.,
putting circuits in-scope for shut-off):
“SCE considers the National Weather Service Wind Advisory levels (defined as 31 mph
sustained wind speed and 46 mph gust wind speed) and the 99th percentile of historical
windspeeds in the area to set activation thresholds. If the wind speed forecasts exceed
these activation thresholds in a location with elevated FPI, SCE initiates PSPS activation
and notifications.”69

65 In its 2025 Annual Implementation Report, Edison appears to have invented FPI 1.5 for the first time: “While FPI 2.0 was not fully
implemented in 2025, key components were operationalized as part of FPI 1.5.” 2025 Annual Implementation Report, supra note 8, Plan
Objective 6, at 5. In 2026, Edison released a PSPS Action Plan in which it states: “Analysis was performed when FPI 1.5 was in effect, though
that FPI version does not directly affect the estimated windspeed reduction calculation.” 2026 PSPS Action Plan, supra note 5, at 9 tbl. 1 n.8.
66 2026 PSPS Action Plan, supra note 5, § II.D, at 25 (“During 2025, primarily based on learnings from the January 2025 fires, SCE made
additional refinements to its FPI”).
67 Interview with Pedro J. Pizarro, President & Chief Executive Officer, Edison International (KTLA 5 television broadcast Dec. 17, 2025). The
exchange: Reporter: “There was a major Red Flag event, you knew days in advance. National Weather Service calling it life-threatening winds.
Why wasn’t that enough to shut off the [transmission]?” Pizarro: “So we did shut off the power on some of the distribution lines in the area —
not on the larger transmission line.” Reporter: “You did have . . . the higher wind gusts even before the fire — 2:20 PM, 63 miles per hour.”
Pizarro: “Our [SCE] team was monitoring all of that. And they had not hit the criteria — that Threshold criteria — for de-energizing those
transmission lines.” Reporter: “So what would be? . . . Is there a 70 mile per hour? An 80?” Pizarro: “I don’t have those numbers in my head.
Happy to give you a full-on interview with our team.” To the best of our knowledge, the threshold criteria have not been provided.
68 L.A. Times Studios, Edison’s CEO Pitches His Plan for Eaton Fire Survivors, YouTube, at 57:17 (2025),
https://www.youtube.com/watch?v=ZFPa4FbtgVs [hereinafter L.A. Times Pizarro Interview].
69 2020 Remedial Compliance Plan, supra note 11 (“Overview of SCE’s Remedial Compliance Plans for Class A Conditions”).


Page 21 of 54
In 2021, Edison disclosed circuit-level activation thresholds for its entire network; specifically the
Transmission Circuits had activation (i.e., in-scope) thresholds of 40 mph sustained and 58 mph gust. 70
Then in 2023, in a PSPS Post-Event workbook, which is part of the public record, Edison specifically
disclosed the activation thresholds for two of the four Transmission Circuits. Based on these thresholds,
Mesa–Vincent No. 1 and No. 2, two of the four Transmission Circuits, should have come into PSPS scope
at an activation threshold of 40 mph sustained or 58 mph gust.71 Therefore, when forecasted wind speeds
(sustained or gust) hit either of those thresholds and forecasted FPI meets criteria, those two circuits
should be in-scope for PSPS. The NWS forecast Edison for January 7th called for 35–50 mph sustained
winds with gusts of 80 to 100 mph in the San Gabriel foothills.72
The Transmission Circuits traverse the Extreme (Tier 3) High Fire Threat District of these foothills for
over 15 miles. That forecast cleared the gust threshold outright. Already, half of the key criteria for
scoping in the circuits had been met prior to January 7th. The other half of the key criteria, the FPI
threshold, was not met because, as Edison knew, it capped the impact of wind at 29 mph and
underweighted wind in its calculations.
Exhibit 2: Edison’s PSPS activation thresholds for the four Transmission Circuits, 2021–2025. All four carried 40
mph sustained / 58 mph gust when Edison last published them in 2021. The two it scoped into its October 2023
event carried the same figures then. We suspect they were unchanged on January 7, 2025, partly because Edison
did not disclose them. The NWS forecast for January 7 was 35–50 mph sustained with 80–100 mph gusts in the San
Gabriel foothills, far past the 58 mph gust activation threshold.
Circuit (sustained / gust,
mph)
2021 activation
threshold (published)
2023 activation
threshold (PSPS Post
Event workbook)
2025 activation
threshold (Eaton Fire)
Mesa–Vincent No. 1
40 / 58
40 / 58
40 / 58 (MW estimate; not
disclosed)
Mesa–Vincent No. 2
40 / 58
40 / 58
40 / 58 (MW estimate; not
disclosed)
Eagle Rock–Mesa
40 / 58
Not disclosed
40 / 58 (MW estimate; not
disclosed)
Goodrich–Gould
40 / 58
Not disclosed
40 / 58 (MW estimate; not
disclosed)
Sources: 2021 — S. Cal. Edison Co., Quantitative and Qualitative Factors for PSPS Decision-Making app. (May 2021); 2023 — S. Cal. Edison
Co., PSPS Event Data Workbook: October 26 – November 2, 2023 PSPS Event, tbl. T03 (Circuit Thresholds) (Dec. 1, 2023).
Notes: Not disclosed = not in the October 2023 workbook or in Edison’s January 2025 post-event report. Edison’s circuit-level activation
thresholds were largely unchanged between 2021 and January 2025. In January 2025, Edison seemingly set the Altadena distribution circuits’

70 Q&Q Factors (2021), supra note 10. The Mesa–Vincent thresholds were unchanged in 2023. Oct. 2023 Event Workbook, supra note 10, tbl.
T03; see infra Exhibit 2. We think the activation thresholds for all four Transmission Circuits, including both Mesa–Vincent lines, were again
unchanged in 2025. The activation thresholds of the ~300 circuits Edison de-energized in January 2025 were largely unchanged from 2021
activation thresholds. See also Feb. 2026 Post-Event Report Amendments, supra note 10, at 7 (PDF pagination) (transmission circuits “use the
NWS High Wind Warning level as their baseline”).
71 Oct. 2023 Event Workbook, supra note 10, tbl. T03 (“Circuit Thresholds”); see also S. Cal. Edison Co., Amendment to PSPS Post-Event Report
for 10/29/23 De-Energization Event, R.18-12-005, app. A, at A-9 to A-10 (Cal. P.U.C. Feb. 16, 2024), https://www.cpuc.ca.gov/-/media/cpuc-
website/divisions/safety-and-enforcement-division/reports/psps-post-event-reports/2023/r1812005sce-amendment-to-psps-post-event-report-for-
102923-deenergization-event-1.pdf (listing Mesa–Vincent No. 1 and No. 2 among the circuits scoped for the October 2023 event, with a FireCast
output ratio of “INF”).
72 Nat’l Weather Serv., L.A./Oxnard, Urgent—Fire Weather Message (Fire Weather Watch), RFWLOX (WWUS86 KLOX 032219) (Jan. 3,
2025, 2:19 PM PST), archived at Iowa Env’t Mesonet, https://mesonet.agron.iastate.edu/wx/afos/p.php?e=202501032219&pil=RFWLOX; Nat’l
Weather Serv., L.A./Oxnard, Urgent—Fire Weather Message (Particularly Dangerous Situation Red Flag Warning), RFWLOX (WWUS86
KLOX 062324) (Jan. 6, 2025, 3:24 PM PST), archived at Iowa Env’t Mesonet,
https://mesonet.agron.iastate.edu/wx/afos/p.php?e=202501062324&pil=RFWLOX.


Page 22 of 54
activation thresholds at or near their 2021 levels.73 By 2023, Edison appears to have stopped using the 99th percentile to set activation
thresholds for Transmission Circuits.
We think the Transmission Circuits breached Edison’s wind scoping test multiple times on January 7th
(see above) but not its FPI scoping test, which kept it out of scope for PSPS. Edison discloses FPI scores
only for circuits it de-energized.
Edison has since blamed FPI for not de-energizing the Transmission Circuits while failing to
acknowledge that for years, it knew of FPI’s high wind event deficiencies. In its May 2026 PSPS Action
Plan, it told the CPUC (Exhibit 3) that high winds alone do not warrant a shut-off:74
Exhibit 3: Edison tells the CPUC that high winds alone do not warrant a shut-off, and calls FPI “a critical forecast
metric.”

Source: S. Cal. Edison Co., 2026 Public Safety Power Shutoff Action Plan § II.D, at 25 (May 28, 2026).
Under Edison’s protocol in effect on the day of the fire, the primary route to a shut-off ran through FPI
(§§ 1.1, 1.6). When the CEO said the Transmission Circuits’ “threshold criteria” were not met (see
above), the threshold he meant should have been FPI. However, the FPI threshold was not the hard gate
Edison’s narrative implies: its incident team overrode it dozens of times that day for the distribution
circuits it was watching. On January 7, 2025, Edison de-energized 33 distribution circuits whose reported
FPI sat below the circuit’s own threshold — typically by 0.8 to 1.0 points, or 6–8% — including seven
before the 6:11 PM fault on the Transmission Circuits line, three of them (Kinneloa, Lamanda and
Broadcast) in the Altadena foothills beside it.75 Yet, Edison never extended the same judgment to the
Transmission Circuits. Because the index stops counting wind at 29 mph sustained (§ 1.2), nothing the

73 Jan. 2025 Post-Event Report, supra note 1, att. C, tbl. 3 (“Circuit Thresholds”) (Kinneloa, Lamanda and Video); Oct. 2023 Event Workbook,
supra note 10, tbl. T03.
74 2026 PSPS Action Plan, supra note 5, § II.D, at 25.
75 Jan. 2025 Post-Event Report, supra note 1, att. C, tbl. 2 (“Factors Considered in Decision to De-Energize”).


Page 23 of 54
wind did above that level on January 7th could have increased the Altadena circuits’ scores of 12.08–
13.08.76 The Transmission Circuits’ published threshold was 13 (§ 1.5).77
Edison’s recent revision of what it means to be in scope supports our view that Edison is now casting a
wider net because its prior narrow view of what it means to be in-scope likely overlooked the
Transmission Circuits during the Eaton Fire. On January 7, a circuit was in-scope when:
“Circuits at risk are deemed to be in scope when they are at risk for reaching event risk
thresholds.”78 (emphasis added – presumably FPI and wind).
By 2026:
“Circuits at risk are deemed to be in scope when they have a reasonable chance of de-
energization.”79
Nonetheless, failing to scope-in the Transmission Circuits is not a good excuse: Edison’s IMT was trained
to bring in the “rogue circuits” its weather forecast failed to include.
Real-time weather readings drove Edison’s shutoffs, but only for circuits already in scope. Because the
FPI was flawed, the Transmission Circuits were never scoped-in on January 7th. Edison’s own training
exercises treat that gap as part of the IMT’s job. A circuit the forecast left out but which the weather
brings in is a “rogue circuit.” Edison defined the term in its April 2022 after-action report, which called
for a written de-energization procedure.80 Edison rehearsed the process in its 2022 and 2024 exercise
scripts. In them, it seems the incident management team (IMT) promotes such a circuit into scope when
winds run “higher than originally forecast,” and requests its shut-off before winds reach or exceed its
shut-off thresholds (100% of threshold is not a necessity – see Exhibits 11 and 12).81
Edison followed the rogue-circuit process in actual PSPS events in November 2021 and November
2022.82 It did not follow it on January 7th for the Transmission Circuits.

76 Id. att. C, tbls. 2–3 (Table 2: Kinneloa and Lamanda FPI 12.08, Video 13.08, with sustained wind at the assigned station of 29.18, 29.18 and
29.42 mph; Table 3: FPI threshold 13 for all three). Wind and FPI values in Table 2 are those recorded at the time of the de-energization decision.
Id. at 8 n.5.
77 Oct. 2023 Event Workbook, supra note 10, tbl. T03 (FPI threshold 13; wind activation thresholds of 40 mph sustained / 58 mph gust; de-
energization thresholds of 62 / 74 mph (Mesa–Vincent No. 1) and 68 / 82 mph (Mesa–Vincent No. 2)). The workbook lists only the two Mesa–
Vincent circuits; Eagle Rock–Mesa and Goodrich–Gould do not appear in it, and SCE has not published FPI thresholds for them. We extrapolate
the Mesa–Vincent thresholds of 13 to all four Corridor circuits.
78 S. Cal. Edison Co., Quantitative and Qualitative Factors in PSPS Decision-Making Technical Paper, glossary, at 2 (Aug. 2021, rev. Nov. 6,
2023) (Attachment B to the January 2–17, 2025 Post-Event Report).
79 S. Cal. Edison Co., Quantitative and Qualitative Factors for PSPS Decision-Making, at 10 (rev. Oct. 2026),
https://www.sce.com/pspstechnicalpaper [hereinafter Q&Q Factors (Oct. 2026)] (“PSPS Terminology”).
80 S. Cal. Edison Co., After-Action Report: 2022 Public Safety Power Shutoff Tabletop Exercise, at 2 (Apr. 15, 2022) (Best Practices, “IMT Role
Familiarity”; Areas for Improvement, “Documentation for Rogue Circuit Processes”).
81 S. Cal. Edison Co., 2022 PSPS Functional Exercise Master Scenario Events List (Excel workbook) (“SCE 2022 PSPS Functional Exercise
MSEL FINAL”) (sheet “MainMSEL”). The 2022 functional exercise series ran May 16–19, 2022; its after-action report (June 10, 2022) was also
produced with SCE’s response to Energy Safety data request OEIS-P-WMP_2023-SCE-006, Question 2. Both the MSEL and that after-action
report were provided as Appendix D to S. Cal. Edison Co., 2022 PSPS Pre-Season Report, R.18-12-005, § IV (Cal. P.U.C. July 1, 2022),
https://on.sce.com/PSPSPreSeasonReporting. The MSEL is reproduced infra as Exhibit 12, panel B. See also S. Cal. Edison Co., 2024 PSPS
Functional Exercise Master Scenario Events List (“SCE 2024PSPSFunctionalExercise MSEL FINAL”).
82 S. Cal. Edison Co., Business Resiliency After Action Report: Event – 11.18.2022 PSPS IMT Event, at 1 (activation Nov. 16–20, 2022) (“During
the Period of Concern for this event, due to rapidly escalating wind conditions and elevated FPI levels, SCE had to de-energize customers on two
circuits in Ventura County that were not originally in scope for this event”), in S. Cal. Edison Co., Response to Energy Safety Data Request
OEIS-P-WMP_2023-SCE-006, Question 2, Docket No. 2023-2025-WMP-DRs, at 25 (PDF pagination) (June 8, 2023) [hereinafter OEIS-P-
WMP_2023-SCE-006 Response]; accord S. Cal. Edison Co., Public Safety Power Shutoff (PSPS) Post-Event Report: November 19, 2022 De-
Energization Event, at 4 (Dec. 6, 2022), https://www.cpuc.ca.gov/-/media/cpuc-website/divisions/safety-and-enforcement-division/reports/psps-


Page 24 of 54
1.5 Same threshold, different treatment: Edison overrode FPI 13 for
the circuits it was watching and seemingly never scored the
Transmission Circuits at all
Edison manually placed three nearby distribution circuits in-scope despite their FPI threshold not being
met. This further illustrates why, in our opinion, Edison acted with conscious willful disregard of the
rights and safety of others in the lead up to the ignition. Edison’s public explanation for leaving the
Transmission Circuits energized is that transmission lines have higher shut-off criteria than distribution
lines.83 On FPI, the index Edison calls the critical metric, that is not so.
The three Altadena distribution circuits Edison shut off on January 7th (Kinneloa, Lamanda and Video)
carried an FPI threshold of 13,84 identical to the two Mesa–Vincent Transmission Circuits it left
energized.85 Edison calculates FPI the same way for both with no distinction for whether a line is
transmission or distribution.86
Edison shut off Kinneloa and Lamanda at FPI of 12.08 and Video at 13.08. Two of the three were below
the 13 on their threshold list (Exhibit 4). It used the in-event discretion described in § 1.4, for the reasons
its own report gives: multiple fires and the lack of aerial suppression resources “factored into adjustment
of FPI and de-energization prior to reaching wind thresholds.”87
Exhibit 4: Same FPI threshold, different treatment. The Transmission Circuits thresholds are from Edison’s
October 26 – November 2, 2023 PSPS event workbook, the last time Edison scoped two of these circuits and
published their figures. We assume the 13 that Edison published for Mesa–Vincent Nos. 1 and 2 also applied to the
two Transmission Circuits it did not disclose. Network-wide, 13 is the upper band of Edison’s FPI thresholds,
Circuit
Type
FPI
threshold
FPI at decision,
Jan. 7
FireCast ratio
(shut off if >
1.0)
Shut off for
PSPS
Kinneloa
Distribution
13
12.08
39.4
3:33 PM, Jan. 7
Lamanda
Distribution
13
12.08
46.5
3:34 PM, Jan. 7

post-event-reports/2022/r1812005-sce-psps-post-event-report-for-11-19-22-de-energization-event.pdf (same sentence). See also S. Cal. Edison
Co., After-Action Report on the November 24, 2021 PSPS Event, in OEIS-P-WMP_2023-SCE-006 Response, at 70 (PDF pagination) (“the
number of circuits in scope increased from about 77 circuits on the 22nd to about 154 circuits (including downstream circuits) on the 24th. In the
end, winds ended up being much higher than originally forecasted”).
83 See Interview with Pedro J. Pizarro, President & Chief Executive Officer, Edison International (Bloomberg Television broadcast Jan. 13,
2025), https://www.bloomberg.com/news/videos/2025-01-13/edison-ceo-on-california-wildfires-lawsuit-video; L.A. Times Pizarro Interview,
supra note 68, at 39:16; see also Altadena Winds Weren’t Strong Enough to Warrant Edison Shutting Off Transmission Lines, CEO Asserts, L.A.
Times (Jan. 15, 2025) (quoting SCE CEO Steven Powell: “Usually winds would have to be 60 to 80 mph for the company to consider the de-
energization of transmission lines”).
84 Jan. 2025 Post-Event Report, supra note 1, att. C, tbls. 2–3 (Table 3: FPI threshold 13 for Kinneloa, Lamanda and Video; Table 2: FPI at the
time of decision of 12.08, 12.08 and 13.08, de-energized at 3:33, 3:34 and 3:52 p.m.). See infra Exhibit 4.
85 Oct. 2023 Event Workbook, supra note 10, tbl. T03 (FPI threshold 13 for Mesa–Vincent No. 1 and No. 2). SCE’s October 2023 workbook
assigns the Mesa–Vincent Transmission Circuits a threshold on the same 1–17 FPI scale. Eagle Rock–Mesa and Goodrich–Gould are not listed in
that workbook. See infra Exhibit 4.
86 Communication from Edison Int’l Investor Relations to Muddy Waters Research (stating that FPI reflects the fire potential of a geographic area
and, because it is determined by local weather and fuel conditions, is not inherently different for distribution versus transmission facilities located
in the same area).
87 Jan. 2025 Post-Event Report, supra note 1, at 13 (“Multiple fires and lack of aerial suppression resources factored into adjustment of FPI and
de-energization prior to reaching wind thresholds.”); see also id. at 42 (“SCE lowered the FPI threshold to 11 at times during this event”).


Page 25 of 54
Circuit
Type
FPI
threshold
FPI at decision,
Jan. 7
FireCast ratio
(shut off if >
1.0)
Shut off for
PSPS
Video
Distribution
13
13.08
469.5
3:52 PM, Jan. 7
Mesa–Vincent No. 1
Transmission
(Transmission Circuits)
13 (per 2023
workbook)
Not disclosed
Not run
Not until Jan. 12
Mesa–Vincent No. 2
Transmission
(Transmission Circuits)
13 (per 2023
workbook)
Not disclosed
Not run
Not until Jan. 12
Sources: Distribution circuits — S. Cal. Edison Co., Public Safety Power Shutoff (PSPS) Post-Event Report: January 2–17, 2025 Event, att. C,
tbls. 2–3 (Mar. 3, 2025). Transmission Circuits — S. Cal. Edison Co., PSPS Event Data Workbook: October 26 – November 2, 2023 PSPS
Event, tbl. T03 (Dec. 1, 2023) (see Exhibit 2); Eagle Rock–Mesa and Goodrich–Gould do not appear in the October 2023 workbook. The 13
threshold is the upper band Edison applies network-wide: S. Cal. Edison Co., 2023–2025 Wildfire Mitigation Plan, Revision 3.1, at 514, 714
app. B tbl. SCE B-01 (Nov. 14, 2024). FireCast output ratios — S. Cal. Edison Co., Public Safety Power Shutoff (PSPS) Post-Event Report:
January 2–17, 2025 Event, att. C, tbl. 4 (Mar. 3, 2025); a ratio above 1.0 calls for shut-off. No FireCast ratio exists for the Transmission
Circuits because Edison seems to have not run the model on them (§ 1.6).
1.6 Edison failed to consider FireCast, a second decision tool that was
signaling to consider the Transmission Circuits for PSPS, because
FireCast requires FPI at threshold
Under the WMP in force on January 7, once FPI put a circuit in-scope, the IMT was required to review
FireCast, Edison’s In-Event Risk Comparison Tool. FireCast compares simulated wildfire consequences
against the social costs of a power shut-off and produces a ratio.88 A ratio above 1.0 means the wildfire
risk exceeds the PSPS risk and suggests de-energization is warranted. The WMP is explicit that this
review is for all circuits in-scope and is considered on a circuit-by-circuit basis.89 FPI was the gate. The
same WMP says:
“At 24 hours before an event, the IMT looks at the twice-daily weather report to see if the
weather pattern has shifted and updates the list of circuits that might be impacted.
Additionally, the In-Event Risk Comparison Tool is run once to compare the risk of de-
energization against the risk of wildfire. The Firecast does not come into play if the
established FPI (discussed in section 9.2.1) does not warrant de-energization.”90
For circuits in-scope, or manually placed in-scope, FireCast seemingly worked accurately that night. Our
map of the Eaton Fire ignition area, built from Edison’s post-event report, shows the ignition point in
Eaton Canyon (red triangle), the four Transmission Circuits, and the distribution circuits Edison de-
energized under its PSPS protocol (Exhibit 5). The three distribution circuits adjacent to the ignition

88 Edison’s CEO appears to describe the FireCast process in an interview. L.A. Times Pizarro Interview, supra note 68, at 39:16. For the technical
detail, see 2023–2025 WMP R3.1, supra note 8, § 9.2, at 628 (“Wildfire simulations (using Technosylva FireCast modeling) for potential
ignitions based on dynamic, in-event weather and wind conditions in proximity to the circuits in scope for de-energization. These conditions are
used to determine the extent of an estimated fire footprint (or fire shed). The risk of a wildfire is calculated based on the number of structures,
population, and acres potentially threatened within the simulated fire shed.”); id. at 628 n.286 (“Technosylva is a suite of wildfire simulation
models or tools. While relying on a similar underlying fire propagation engine, each model is designed to support a unique use case. FireCast is
specifically designed to forecast ignition risk associated with electric utility assets over a 3-day horizon based on expected short-term weather
conditions.”).
89 2023–2025 WMP R3.1, supra note 8, at 625 fig. SCE 9-02 (“PSPS Decision Making Flowchart / Diagram”).
90 Id. at 630.


Page 26 of 54
point were all well above the 1.0 de-energization threshold, reading 39 (Kinneloa), 47 (Lamanda) and 469
(Video).91 None appear to have been a marginal call.
Edison acted on those readings, but only at the distribution level, on the lower-voltage lines with the
higher customer impact. It de-energized Kinneloa, Lamanda and Video between 3:33 and 3:52 PM on
January 7, 2025, roughly two and a half hours before the fire was reported at about 6:18 PM.
Yet, Edison left energized the Transmission Circuits running through the San Gabriel Mountain foothills,
an area designated as a Tier 3 High Fire Threat District (“HFTD”), where 80–100 mph winds were
forecast, and observed near Eaton Canyon on the morning of January 7.
Exhibit 5: Edison’s FireCast readings on the three distribution circuits around the ignition point ran 39 to 469
times its own shut-off threshold. The red triangle is the ignition point Edison reported. The four transmission lines
through Eaton Canyon are the Transmission Circuits. The numbers are FireCast output ratios for the circuits
Edison scoped and de-energized; a ratio above 1.0 calls for shut-off. No ratio exists for the Transmission Circuits,
because Edison never ran the model on it.

Sources: FireCast output ratios — S. Cal. Edison Co., Public Safety Power Shutoff (PSPS) Post-Event Report: January 2–17, 2025 Event, att. C,
tbl. 4 (Mar. 3, 2025). Ignition point — coordinates reported in Letter from Paul Pimentel, Claims Principal Manager, S. Cal. Edison Co., to Ed
Pike, Wildfire Safety Enf’t Branch, Cal. Pub. Utils. Comm’n (Jan. 27, 2025). Circuit routes — traced by Muddy Waters from the map attached to
Letter from S. Cal. Edison Co. to Wildfire Safety Enf’t Branch, Cal. Pub. Utils. Comm’n (Feb. 6, 2025). Map by Muddy Waters.

91 Jan. 2025 Post-Event Report, supra note 1, att. C, tbl. 4 (FireCast output ratios); see also id. at 8 tbl. 2, 16 tbl. 4 (tables “continued in
Attachment C – PSPS Event Data Workbook”).


Page 27 of 54
As far as we can tell from Edison’s filings, Edison never generated a FireCast score for the Transmission
Circuits. The CPUC asked Edison to name “challenges in quantifying risks and benefits in terms of
determining the scope (size and duration) of the PSPS you conducted?” Edison answered (see also
Exhibits 9 and 10):
“The main challenge SCE experienced in 2025 was the timely availability of FireRisk
[AKA FireCast] model information for circuits that were not originally forecast in
scope for potential de-energization. These circuits were brought into scope based on
emergent weather conditions… While FireRisk modeling inputs are normally based on
information provided as part of SCE’s initial PSPS forecast, SCE does retain capability to
perform a risk benefit analysis for most emergent conditions based on FireRisk model
info for circuits within High Fire Threat Districts (HFTD). However, circuits that are
not originally identified in scope for a PSPS event and not included in the FireRisk
Model inputs cannot be assessed. SCE continues to refine and update its forecast
modeling, which continues to further improve forecasting for PSPS event scope to
address this challenge.” 92
Again, it is clear, Edison never manually placed the Transmission Circuits in-scope.
1.7 Another risk Edison knew, but failed to act upon by placing the
Transmission Circuits in-scope: A recloser could not be disabled
An automatic recloser re-energizes a line after it faults and trips, to test whether the fault has cleared.
“Test” means re-energizing the circuit. In a Red Flag windstorm, that means pushing a second surge of
power into a line that may be down, arcing or, as likely here, inducing current in an adjacent conductor.93
Disabling reclosers in those conditions is standard practice. Edison’s WMP says it has done so “for at
least 30 years” on all high fire-risk miles, triggered by exactly the Red Flag Warning in effect on January
7.94 Since at least 2019, Edison’s internal operating procedure, SOB 322, has required transmission

92 2025 Post-Season Report Amendment, supra note 42, at 137–38.
93 D.19-05-038, § 7.3, at 28 (summarizing SCE’s description of its practice: “SCE states that it utilizes remote automatic reclosers (RAR) that can
be remotely configured to prevent reclosing during times of elevated fire risk. According to SCE, this configuration provides the benefit of
avoiding the repeated energization of potential faulted conditions. Additionally, if a line or line section relays during an RFW or other elevated
fire weather condition, the line will not be re-energized until a visual check of all overhead conductors and equipment is completed to further
ensure re-energization does not create an ignition risk.”).
94 2023–2025 WMP R3.1, supra note 8, at 333.


Page 28 of 54
reclosers to be made non-automatic during PSPS proximity threats and fire-weather alerts.95 The
independent evaluator’s 2021 compliance report on Edison’s 2020 WMP spells out the rule.96, 97
A recloser’s “shot” setting is the number of times it re-energizes a faulted line before locking out. On
zero shot, the line stays dead after the first trip until an operator restores it. A former San Diego Gas &
Electric (“SDG&E”) operator noted that a line under a Particularly Dangerous Situation warning should
be “on zero shot.”98 A second former SDG&E operator said that opening the breaker, via digital
communications, can be done “quickly and remotely.”99 Because the breaker automatically reclosed, we
believe it was not set to zero shot.
Prior to the fire, Edison tried, and failed, to disable the reclosers on a connected transmission circuit.
Despite knowing the recloser was set to automatically reclose, Edison did not put the circuit in-scope.
Edison’s § 315 incident report to the CPUC, dated February 6, 2025, records what happened on Eagle
Rock–Gould.
The report states that at approximately 6:11 pm, a fault was detected and that “an earlier attempt to
disable the automatic reclosing function was unsuccessful; therefore, it automatically tested the line to
determine if the fault had cleared.”100 The fault had not cleared, so the line locked out (i.e. de-energized).
Surveillance video from a Pasadena Swim & Tennis club shows two flashes at the ignition point roughly
three seconds apart.101 We suspect the second is the recloser re-energizing a faulted line into a
windstorm.
Although Eagle Rock–Gould does not itself run through Eaton Canyon, it is connected to it.102 On a
networked grid, however, the breaker that sees a fault and re-tests the line can be miles from the fault, and
its recloser re-energizes the faulted path through the network. In a 2019 data response filed with the
CPUC, Edison said that because its transmission circuits can be networked, de-energizing one requires

95 2020–2022 WMP Rev. 3, supra note 20, at 162. Similar guidance appears in 2019 WMP Progress Update, supra note 20, at 3 (OP-1)
(“blocking transmission reclosers”). See also 2022 WMP Update, supra note 5, § 7.3.6.1, at 437–38 (SOB 322 contains “operational protocols for
making reclosers non-automatic . . . for designated overhead transmission, sub-transmission and distribution circuits,” and “[a] PSPS Proximity
Threat puts operating and reclosing restrictions on transmission and sub-transmission lines within one mile of a distribution line that is on the
period of concern (POC),” declared by the Grid Control Center Liaison “during PSPS activation and throughout a POC”).
96 NV5 & Guidehouse, Final Independent Evaluator Annual Report on Compliance (S. Cal. Edison Co. 2020 Wildfire Mitigation Plan), at 44
(June 30, 2021) (“Under Recloser Restrictions, transmission reclosers will be made non-automatic during FWT, TT, or PSPS proximity threats.
The PSPS IMT IC, in consultation with the GCC Liaison, will identify specific lines for reclosers to be made non-automatic. Sub-transmission
reclosers will be made non-automatic during RFWs (by county) and FWT, TT or PSPS Proximity Threat (by Switching Center and county).
Distribution reclosers will be made non-automatic by county or Switching center during TT, RFW and FWT. During FCZ, they will be made
non-automatic via Auto-322 program or manually.”). For comparison, LADWP in the same windstorm preemptively blocked reclosing on the
line nearest the Palisades ignition on January 6, 2025. L.A. Dep’t of Water & Power, Remarks by LADWP Executives About Wind and Wildfire
Response, January 2025 – Presented at Board of Water and Power Commission Meeting, Jan. 28, LADWP News (Jan. 30, 2025),
https://ladwpnews.com/remarks-by-ladwp-executives-about-wind-and-wildfire-response-january-2025-presented-at-board-of-water-and-power-
commission-meeting-jan-28/; L.A. Dep’t of Water & Power, LADWP Statement Regarding Palisades Fire Litigation, LADWP News (Mar. 25,
2025), https://ladwpnews.com/ladwp-statement-regarding-palisades-fire-litigation/ (reclosing on the 34.5 kV Royal–Monte Grande 1 sub-
transmission line, approximately a quarter mile from the fire’s origin, was blocked on January 6, 2025).
97 In August 2026, Edison disclosed, regarding a different incident, that “[t]he Moorpark-Oakpark 66 kV Circuit relayed at 2:08 p.m. but did not
automatically reclose as System Operating Bulletin (SOB) 322 was in effect.” Letter from Jose Moran, S. Cal. Edison Co., to Vanessa Aguila,
Wildfire Safety Enf’t Branch, Cal. Pub. Utils. Comm’n, at 1 (Aug. 28, 2026) (Pub. Util. Code § 315 report on the Kanan Fire),
https://download.edison.com/406/files/202608/711fc64ab7abd1e22c649a1393caad13/section-315-20260831-kanan-fire.pdf.
98 SDG&E Operator #1 Interview, supra note 23.
99 SDG&E Operator #2 Interview, supra note 23.
100 Feb. 6, 2025 § 315 Letter, supra note 14, at 2.
101 Ozebek, supra note 22.
102 Feb. 6, 2025 § 315 Letter, supra note 14 (attached circuit map of the Eaton Canyon corridor, showing the Eagle Rock–Gould line entering at
the lower left and running north to Gould); see also Jan. 27, 2025 Interim Letter, supra note 14 (Eagle Rock–Gould “connects the Gould
substation in La Cañada Flintridge to the Eagle Rock substation,” and “because SCE’s transmission system is networked, the fault on this
geographically distant line caused a momentary and expected increase in current”).


Page 29 of 54
opening the breakers at every source.103 Regardless, Edison knew before ignition that the disable attempt
had failed. If the fix was trivial, it should have been made; if not, the Connected Transmission Circuit
should have been placed in-scope.
Edison acted far more carefully with an analogous situation that same night. That evening, Edison lost
communication with its grid monitoring system (SCADA) across Ventura County and parts of northern
and eastern Los Angeles County. Edison’s incident commander reacted by de-energizing many LA and
Ventura county circuits directly from substations “out of an abundance of caution,” even with most below
their wind or FPI triggers.104
1.8 Other factors that evening Edison knew that called for manually
placing the Transmission Circuits in-scope and de-energizing them:
the Palisades Fire, no air support, a lightly loaded grid, and a
redundant network
Induction was the likely cause of the Eaton Fire and the Transmission Circuits were a known induction
hazard by Edison’s own definition. In 2018, a contractor working on a de-energized Edison line was
killed when an adjacent energized line induced current in it. The CPUC’s 2021 investigation report states
Edison’s own standard:105
“Electromagnetic induction is a phenomenon whereby an alternating electric current in
one conductor can cause an electric current to also flow in another nearby parallel
conductor, even if that parallel conductor is not otherwise energized. The degree to
which induction occurs, i.e., the magnitude of the current induced in the second
conductor, is proportional to the distance the two conductors run parallel to each other.
SCE considers an induction hazard to exist wherever transmission lines run parallel
for over 600 feet or when the de-energized conductors are located within
transmission/sub-transmission corridors.” (emphasis added)
Edison’s letter to the CPUC identified the location of the fire’s origin (Tower Location 1, or T1).106
Using Google Earth’s measuring tools, we measured the parallel transmission spans running to it. Each
span significantly exceeded 600 feet (Exhibit 6). The span leading to T1 (the origin) is approximately
1,150 feet, the span leading to T2 approximately 1,800 feet, and the span leading to T3 approximately
2,600 feet. These are roughly 2x, 3x and 4x the minimum length required to generate an induction

103 S. Cal. Edison Co., Response to Data Request Set MGRA-SCE-001, R.18-10-007, Question 07 (Cal. P.U.C. filed Mar. 13, 2019) (responses
dated Mar. 6–7, 2019), https://energysafety.ca.gov/wp-content/uploads/docs/misc/docket/272342921.pdf (Question: “How many of the circuits
specified in MGRA-1 can currently be remotely deenergized or re-energized via SCADA? Please specify answers for distribution and
transmission circuits separately.” Response: “. . . Transmission circuits may have more than one source of electricity and operate networked as
opposed to radially. As such, all of the associated circuit breakers for each source would need to be opened to de-energize the circuit. Some
transmission lines have sources or terminals which are not operated by SCE, where the source or terminal may be a neighboring utility or a
generating facility.”).
104 Jan. 2025 Post-Event Report, supra note 1, § 5, at 29–30 (“out of an abundance of caution (for circuits that were expected to approach or meet
criteria), the incident commander authorized the de-energization of some circuits directly from substations”). The passage appears in the report’s
discussion of missed customer notifications, which it attributes in part to “system and process issues”; the SCADA loss is the one example given.
105 Kramer Junction Investigation Report, supra note 27, at 7.
106 Jan. 27, 2025 Interim Letter, supra note 14 (the County’s “preliminary origin area . . . around coordinates N34.1860422292
W118.09357612511549 in Eaton Canyon”).


Page 30 of 54
hazard. Our measurements are straight lines (i.e., as the crow flies) and are therefore conservative,
because they do not account for the curvature of the lines as hung.
Exhibit 6: The parallel transmission spans leading to the fire’s origin at Tower Location T1, measured by MW in
straight lines: ~1,150 feet to T1, ~1,800 feet to T2 and ~2,600 feet to T3. Each exceeds the 600 feet at which Edison
considers an induction hazard to exist.

Source: Google Earth (image date Dec. 2023); spans measured by Muddy Waters. Fire origin per Letter from S. Cal. Edison Co. to Wildfire
Safety Enf’t Branch, Cal. Pub. Utils. Comm’n (Feb. 6, 2025).
The idle line in Eaton Canyon, on tower M16T1, runs next to the four energized 220 kV Transmission
Circuits, which parallel it for miles. A former Edison senior executive, shown Edison’s own § 315 map,
put this configuration “definitely on the higher category of induction risk.”107
The lack of fire suppression resources due to the Palisades fire was clear. The Palisades Fire started
burning in the morning. By the afternoon of January 7th, it was out of control on the other side of Los
Angeles County. It demonstrated in real time both what the day’s winds could do to an ignition and that
the region’s fire suppression resources were already committed. Edison’s WMP, and the Eaton Fire Post
Event report, lists fire suppression resource availability as an event-specific factor in PSPS decisions.108
According to Edison’s own data, de-energizing the Transmission Circuits likely would have cost
essentially nothing, because it is unlikely customers would have lost power as a result. Edison’s system
demand at 6 PM on January 7th was ~12,000 MW, against estimated Edison winter capacity of ~20,000
MW. That is roughly still just 60% of capacity after adjusting for estimated winter maintenance outages.
On January 7th, CAISO system-wide demand was ~27,000 MW, against a 2025 peak of ~44,500 MW.109
A former Edison senior executive also confirmed to us that each of the four Transmission Circuits has at

107 SCE Executive #1 Interview, supra note 15.
108 2023–2025 WMP R3.1, supra note 8, at 514, 624, 627; Jan. 2025 Post-Event Report, supra note 1, at 13.
109 CAISO Demand Data, supra note 15; CAISO 2025 Statistics, supra note 15 (2025 system-wide peak of 44,506 MW on August 21, 2025, at
6:07 p.m., based on 1-minute averages and including dynamic transfers). System-wide capacity generally runs about 20% above peak demand,
because CAISO maintains a reserve buffer of roughly that size. SCE Executive #1 Interview, supra note 15. The assumed 25% reduction in SCE
capacity for winter maintenance outages, yielding approximately 18,750–22,500 MW of available capacity, is a Muddy Waters estimate based on
the same interview.


Page 31 of 54
least two alternative sources of supply,110 and that de-energizing transmission lines “without impacting
the function of the grid” is, in general, routine.111 On the night of January 7th, all four de-energized
anyway, taken down by their own breakers: Mesa–Vincent No. 2 at 7:17 PM, Mesa–Vincent No. 1 and
Goodrich–Gould at 10:35 PM, and Eagle Rock–Mesa at 12:35 AM on January 8.112 Edison incurred
whatever cost a shut-off carried, minus the benefit. Edison subsequently seemingly proved the point
itself. On January 12th, and January 20th, it proactively de-energized all four Transmission Circuits for
PSPS with no customer outages.
Part 2: Edison Shareholders Could Be Liable for ~$20
Billion. Imprudence Seems Near Certain; Willful
Disregard Is Likely in Our Opinion
2.1 The framework: AB 1054, the Wildfire Fund, and the prudence
inquiry
AB 1054 lets the Wildfire Fund pay claims from utility-caused fires. If the Commission finds Edison
acted prudently, Edison repays nothing. If the Commission disallows the costs as imprudent (§ 1701.8),
Edison reimburses the Fund, but § 3292(h)(2) caps the reimbursement at 20% of its transmission and
distribution (T&D) equity rate base, about $4.3 billion.113 The cap falls away under § 3292(h)(3) if
Edison lacked a valid safety certification on the date of ignition ((h)(3)(B)), or if the Fund administrator
determines that its “actions or inactions that resulted in the covered wildfire constituted conscious or
willful disregard of the rights and safety of others” ((h)(3)(A)). Edison’s 2024 certification was in force
on January 7th,114 so the cap turns on the administrator’s finding under (h)(3)(A). Certification presumes
prudence only until a party raises “serious doubt” (§ 451.1(c)). Edison itself proposed the measure:
substantial compliance with its WMP “will demonstrate that SCE prudently operated its system.”115 On
that measure, we think it failed multiple times.

110 SCE Executive #1 Interview, supra note 15 (“An N-2 transmission contingency means designing the transmission system so it remains
operational even after losing two transmission lines. If one line fails, the system must still have enough capacity and resilience to withstand the
loss of a second line.”).
111 Id.
112 Feb. 6, 2025 § 315 Letter, supra note 14, at 3 (Mesa–Vincent No. 2 relayed at approximately 7:17 p.m. on January 7; Mesa–Vincent No. 1 and
Goodrich–Gould at approximately 10:35 p.m.; Eagle Rock–Mesa at approximately 12:35 a.m. on January 8).
113 Cal. Pub. Util. Code § 3292(h)(2)(C)(i)(ia)(IA) (formerly § 3292(h)(2)(C)(i)(I)) (cap of “20 percent of the electrical corporation’s total
transmission and distribution equity rate base”); Edison Int’l & S. Cal. Edison Co., Annual Report (Form 10-K) (Feb. 2025) (fiscal year ended
Dec. 31, 2024), https://www.sec.gov/Archives/edgar/data/827052/000082705225000022/eix-20241231.htm [hereinafter 2024 Form 10-K]
(Wildfire Insurance Fund discussion: the AB 1054 liability cap equals 20% of the equity portion of the utility’s transmission and distribution rate
base, excluding general plant and intangibles, in the year of the prudency determination); April 2026 Business Update, supra note 4, at 14. Prior
Form 10-Ks put the cap at approximately $3.4 billion (2022), $3.8 billion (2023) and $3.9 billion (2024). But see Cal. Wildfire Fund Adm’r,
Appendix 1 (AB 1054 liability cap calculations) (updated July 21, 2026),
https://www.cawildfirefund.com/sites/wildfire/files/documents/2024/appendix_1.pdf [hereinafter Wildfire Fund Appendix 1] (SCE cap of
approximately $3.63 billion (2022), $3.87 billion (2023) and $4.07 billion (2024)).
114 Cal. Off. of Energy Infrastructure Safety, Issuance of Southern California Edison’s Safety Certification, Docket No. 2024-SCs, TN15094
(Dec. 11, 2024),
https://efiling.energysafety.ca.gov/EFiling/GetFile.aspx?filePath=D%3a%5cFileThat%5cFileServer%5cPublic%5cPublicDocuments%5c2024-
SCs%2fTN15094_20241211T145722_Issuance_of_Southern_California_Edison%27s_Safety_Certification.pdf (finding that SCE’s October 8,
2024 request met the requirements of Cal. Pub. Util. Code § 8389(e) (West 2024) (now § 8389(a)) and that the certification is valid for 12 months
from issuance); see also Press Release, Cal. Off. of Energy Infrastructure Safety, Energy Safety Issues Safety Certifications to PG&E, SDG&E,
SCE, and BVES (Dec. 2024). Edison also posts the issuance letter at download.edison.com (“2024-sce-safety-certification.pdf”).
115 2019 WMP, supra note 55, § 1.1, at 7. But see Guidance Decision on 2019 Wildfire Mitigation Plans Submitted Pursuant to Senate Bill 901,
D.19-05-036, at 4 (Cal. P.U.C. May 30, 2019), https://energysafety.ca.gov/wp-content/uploads/docs/misc/docket/296577466.pdf (declining to
find that substantial compliance with a WMP establishes that a utility acted prudently when it later seeks cost recovery: “Senate Bill 901 did not


Page 32 of 54
2.2 Imprudence and willful disregard: what Edison’s plan required,
and what Edison knew
Imprudence asks whether Edison acted as a reasonable operator would have, given what it knew or should
have known.116 Departures from its own approved plan seemingly should at least constitute imprudence,
answering that question.
On conscious or willful disregard of the rights and safety of others, the cap exception presumably borrows
the language of malice from California’s punitive-damages statute, Civil Code § 3294: “despicable
conduct which is carried on by the defendant with a willful and conscious disregard of the rights or safety
of others.”117 Courts read this as a knowledge test. It asks whether the defendant knew, and proceeded
anyway; “should have known” is not enough.
Below are the actions and inactions that, in our opinion, constitute willful disregard for the safety of
others. We note four of the below are the express violations of WMP. We believe any one of the five
below could constitute willful disregard, exposing Edison to a potential $20 billion+ nominal, pre-tax
financial liability.
•
Transmission Circuits never brought into scope for de-energization when all indications
suggested they should have been (§§ 1.4, 1.5)
o
WMP rule broken: if wind thresholds and FPI required the Transmission Circuits to be
in-scope, the FireCast review was mandatory. But Edison knew FPI understated risks
from the wind event forecasted days in advance.
o
Edison trains its teams to pull in the “rogue circuits” the weather forecast misses, a term
it defined, and called for a procedure on, in 2022.
•
For five years, Edison seemingly knew FPI could not distinguish between high winds and
extreme wind:118 (§§ 1.2–1.4)
o
WMP rule broken: the WMP made FPI a critical gate for PSPS decisions, transmission
included, and in 2021 set FPI 2.0 to replace it by the end of 2022 (Appendix A, FPI 2.0
timeline).

redefine the ‘prudent manager’ test”); id. at 15 (describing SCE’s proposal), 38 (“Any provision in a WMP that represents that approval of the
Plan constitutes a determination on cost, reasonableness, or prudency is disapproved.”).
116 Cal. Pub. Util. Code § 451.1(b) (West 2026) (wildfire costs are just and reasonable where “the conduct of the electrical corporation related to
the ignition was consistent with actions that a reasonable utility would have undertaken in good faith under similar circumstances, at the relevant
point in time, and based on the information available to the electrical corporation at the relevant point in time,” taking into account factors within
and beyond the utility’s control); id. § 451.1(c) (the conduct of a utility holding a valid safety certification is deemed reasonable “unless a party to
the proceeding creates a serious doubt as to the reasonableness of the electrical corporation’s conduct,” after which the utility “has the burden of
dispelling that doubt and proving the conduct to have been reasonable”). Section 451.1 was amended by Assembly Bill 1054 and again by
Assembly Bill 1513. 2019 Cal. Stat. ch. 79, § 6 (Assemb. B. 1054); 2019 Cal. Stat. ch. 396, § 23 (Assemb. B. 1513). The Commission’s general
prudence test is the same “known or should have known” inquiry: “reasonable and prudent means that at a particular time any of the practices,
methods, and acts engaged in by a utility follows the exercise of reasonable judgment in light of facts known or which should have been known at
the time the decision was made,” and a reasonable and prudent act “is not limited to the optimum practice, method, or act to the exclusion of all
others.” Re S. Cal. Edison Co., D.87-06-021, 24 Cal. P.U.C.2d 476, 486 (1987), quoted in D.02-08-064, I.01-06-047 (Cal. P.U.C. Aug. 22, 2002)
(Southwest Gas Corp.), https://docs.cpuc.ca.gov/published/Final_decision/18738-03.htm; accord D.90-09-088, 37 Cal. P.U.C.2d 488, 499 (1990);
Decision Denying Application, D.17-11-033, A.15-09-010, at 10 (Cal. P.U.C. Nov. 30, 2017) (San Diego Gas & Electric Co. 2007 wildfire costs),
https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M200/K045/200045020.PDF (applying the standard to utility-ignited wildfire costs).
117 Cal. Civ. Code § 3294(c)(1) (West 2026) (defining “malice”).
118 2025 Post-Season Report Amendment, supra note 42, at 135; see also 2022 WMP Update, supra note 5, at 97 tbl. SCE 4-18 (wind speed input
“capped at 29 mph”).


Page 33 of 54
o
Edison told its regulator in 2020, 2021, 2022, 2023 and 54 days before the fire that it built
the fix, budgeted $0 for FPI 2.0 enhancements in 2024 and it was not launched in time for
the Eaton Fire.
•
Reclosing on Eagle Rock–Gould (§ 1.7)
o
WMP rule broken: the WMP bars “automated or manual re-energization” of a transmis-
sion line that relays within a mile of a monitored distribution line,119 and the WMP and
SOB 322 required recloser blocking under the Red Flag Warning in effect.120
o
Edison knew that its attempt to disable reclosing had failed, and it seemingly knew so be-
fore ignition; the line tested into an uncleared fault seconds after tripping. We think this
may have caused the fire and / or exacerbated it.
•
By the afternoon of January 7th, the Palisades fire was spreading rapidly, straining public
safety resources;
•
Rule broken: the WMP lists fire-suppression resource availability as an event-specific
factor in PSPS decisions, and Edison’s transmission protocol says it takes “event-specific
conditions into account” in deciding whether to de-energize. On January 7th, Edison ap-
plied that factor to the distribution circuits beside the Corridor, including shutting them
two off well below their FPI thresholds — and never applied it to the Transmission Cir-
cuits.
o
Edison knew that the Palisades fire was spreading rapidly. The city of Los Angeles had
declared a State of Emergency. Local fire suppression and public safety resources were
rapidly being depleted. 121,122
•
Edison seemingly could have de-energized the Transmission Circuits without significant
power loss to customers
o
Edison had the sole authority from CAISO to de-energize transmission and its outlined
2019 protocol indicated it would coordinate with CAISO days, model load issues, ahead
of a potential transmission de-energization.123 According to Edison: “SCE’s grid is oper-
ated to withstand next-contingency (N-1) and some multiple-contingency (N-2) condi-
tions on an on-going basis.”124 We believe grid redundancy and excess power capacity

119 2023–2025 WMP R3.1, supra note 8, § 9.2, at 626–27 (“When a transmission line is within the one-mile boundary of a monitored distribution
line, the transmission line has operating restrictions placed into effect to prevent an automated or manual re-energization if the transmission line
was to relay.”); see also 2022 WMP Update, supra note 5, § 8.2.3, at 551 (same protocol).
120 2020–2022 WMP Rev. 3, supra note 20, at 162; 2019 WMP Progress Update, supra note 20; 2022 WMP Update, supra note 5, § 7.3.6.1, at
437–38 & n.186 (describing SOB 322: a PSPS Proximity Threat “puts operating and reclosing restrictions on transmission and sub-transmission
lines within one mile of a distribution line that is on the period of concern,” and “blocking reclosers means that no attempted re-energization can
take place automatically, possibly preventing a second relay operation”).
121 City Declares Local Emergency, supra note 32; Declaration of Local Emergency, supra note 32.
122 2023–2025 WMP R3.1, supra note 8, at 514, 624, 627; Jan. 2025 Post-Event Report, supra note 1, at 13 (“Multiple fires and lack of aerial
suppression resources factored into adjustment of FPI and de-energization prior to reaching wind thresholds.”).
123 Sept. 2019 De-energization Progress Report, supra note 47, § VIII.B, at 24 (“SCE intends to alert the CAISO as early as three days ahead of a
potential PSPS event”; “[t]his would allow SCE and CAISO to perform the necessary load flow studies”); see also 2023 Post-Season Report,
supra note 42, § III.E.2–4, at 23–24 (contingency analysis “based on forecasted load”; potential transmission outages submitted to CAISO
“typically 3 days in advance”); Decision Adopting De-Energization (Public Safety Power Shut-Off) Guidelines (Phase 1 Guidelines), D.19-05-
042, § 4.6.2, at 60 (Cal. P.U.C. May 30, 2019) (recording CAISO’s position that “de-energization decisions rely entirely on the transmission
owners”).
124 2025 Post-Season Report Amendment, supra note 42, at 133.


Page 34 of 54
across CAISO on January 7th would likely have led either no outages or manageable out-
ages.125
In our view, defunding a remedy for a known safety defect is the paradigm of conscious disregard.
Edison knew its gate was inadequate for January 7th, the Transmission Circuits were an induction hazard,
a safeguard had failed, and a shut-off would likely have had minimal impact. In our opinion, that is
conscious disregard of safety , and the cap should be voided.
2.3 Damages – Imprudence Versus Willful Disregard: a ~$15 billion
question
AB 1054 leaves Edison shareholders three outcomes. The gap between the last two is about $15 billion,
or ~$40 a share undiscounted.
Finding
Consequence
Pre-tax
Shareholder
liability
Per share
(undiscounted,pre-
tax)
Wholly prudent
operator
Wildfire Fund pays; no reimbursement
~$0
~$0
Wholly / partially
imprudent (cap applies)
Reimburse Fund, capped at 20% of
T&D equity rate base, as disclosed by
Edison126
~$4.3 billion127
~$11
Conscious / willful
disregard (cap voided)
Reimburse Fund without cap
~$20 billion+128
~$52
Per-share figures use 384,813,979 EIX common shares outstanding as of July 23, 2026 (EIX Form 10-Q,
Q2 2026, cover). The ~$4.3 billion is the statutory cap as applied to SCE’s T&D equity rate base; the
~$20 billion is our estimate of the credible tail of Eaton Fire liability if the cap is voided. No probability
weighting is applied.

125 See CAISO Demand Data, supra note 15; SCE Executive #1 Interview, supra note 15.
126 April 2026 Business Update, supra note 4, at 14 (“Liability cap of 20% of T&D Equity Rate Base (~$4.3Bn for SCE for 2025)”); see also id.
at 13 (“SCE 2025 cap: ~$4.3 Bn”).
127 Cal. Pub. Util. Code § 3292(h)(2)(C)(i)(ia)(IA) (formerly § 3292(h)(2)(C)(i)(I)) (cap of “20 percent of the electrical corporation’s total
transmission and distribution equity rate base”); 2024 Form 10-K, supra note 113 (Wildfire Insurance Fund discussion: the AB 1054 liability cap
equals 20% of the equity portion of the utility’s transmission and distribution rate base, excluding general plant and intangibles, in the year of the
prudency determination); April 2026 Business Update, supra note 4, at 14. Prior Form 10-Ks put the cap at approximately $3.4 billion (2022),
$3.8 billion (2023) and $3.9 billion (2024). But see Wildfire Fund Appendix 1, supra note 113 (SCE cap of approximately $3.63 billion (2022),
$3.87 billion (2023) and $4.07 billion (2024)).
128 Muddy Waters estimate. Reference points, all of which measure insured loss rather than utility liability: Milliman (Sheri Scott, Principal &
Consulting Actuary), presentation to the California Catastrophe Response Council (July 24, 2025), estimating Eaton Fire insured loss and loss
adjustment expense at $13.7–$22.8 billion, as summarized in Wildfire Fund 2026 Annual Report, supra note 3, at 9–10, which also records PCS’s
reported $15.2 billion Eaton estimate, Verisk’s $8–$10 billion, Moody’s $7–$12 billion, approximately $8 billion of insurance payments on close
to 28,000 Eaton claims reported by the California Department of Insurance as of March 31, 2026, and the Administrator’s view that eligible
Eaton claims “may well exceed the total claim-paying capacity of the AB 1054 Account”; Gallagher Re, Natural Catastrophe and Climate
Report: First Half 2025, at 5 tbl. 1, 9 tbl. 2 (July 16, 2025), https://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/news-and-
insights/2025/july/h1-2025-natural-catastrophe-and-climate-report.pdf (Eaton: $28 billion economic loss, $17 billion insured); Cal. Earthquake
Auth., Enhancing California’s Resiliency to Natural Catastrophes (Apr. 7, 2026),
https://www.cawildfirefund.com/sites/wildfire/files/documents/2026/sb254-natcatresiliencyreport-locked-4-7-26.pdf (“Early estimates of insured
losses exceed $15 billion; total losses are estimated to be well above that figure”; “The Eaton Fire may well deplete the $21 billion Wildfire
Fund”). Utility liability additionally includes uninsured and underinsured property, wrongful-death and injury claims, business interruption and
public-entity costs. SCE has agreed to settle subrogation claims at 52 cents on the dollar. Edison Int’l & S. Cal. Edison Co., Current Report (Form
8-K) (Sept. 15, 2025), https://www.sec.gov/Archives/edgar/data/827052/000082705225000081/eix-20250915x8k.htm (SCE to pay “$0.52 for
each dollar in claims paid or to be paid . . . up to an agreed upon cap”).


Page 35 of 54
2.4 Scenario A – mere imprudence (or partially prudent as prudent /
imprudent decision is not binary): ~$4.3 billion, ~$6 to $6.50 per
share, post-tax, discounted
The Wildfire Fund administrator’s determination follows the resolution of claims, so reimbursement is
unlikely before roughly year three. Discounting a single ~$4.3 billion payment, approximately year three,
at 8–9% gives ~$6 - $6.50 per share, post tax (28% tax rate).
2.5 Scenario B – willful disregard: ~$20 billion, ~$20 to $21 per
share, post-tax, discounted
An uncapped, nominal ~$20 billion would not be paid at once. We model a $5 billion first installment in
the year the liability is assessed, approximately year three, $1 billion annually in years six through fifteen,
and an additional $5 billion in year fifteen. That is a $20 billion nominal stream. Discounted at 8–9%
from 2026, the stream is worth ~$20 per share, post tax (28% tax rate).


Page 36 of 54
APPENDIX
This appendix has two parts. Appendix A sets out, in Edison’s own words and in date order, what it told
its regulators about FPI 2.0 between 2019 and 2026. Appendix B collects Exhibits 7 through 19, which
are referenced in Part 1 and Appendix A.
A. Edison’s statements on FPI 2.0, 2019–2026
2019 — Edison announces FPI Phase Two and buys supercomputers to
compute it
•
Edison begins Phase II of the FPI project, intended to “increase capability by adding more
granular weather data.”129
•
Supercomputers compute FPI:
o
SCE’s plan, as filed (Feb. 6, 2019): “In 2019, SCE will procure and install a High
Performance Computing Cluster (HPCC) that will generate forecasts of weather and fuel
conditions at high resolution. Greenness of the vegetation, moisture content of the dead
and live fuels, relative humidity, and wind data from the HPCC will be used to
comprehensively assess wildfire risk across the area (e.g., HPCC will compute FPI).”
130(emphasis added)
o
The CPUC’s decision approving the plan (May 30, 2019), recording SCE’s commitment
in the Commission’s words: “In 2019, SCE will install a HPCC that will generate
forecasts of weather and fuel conditions at high resolution. The HPCC will
comprehensively assess wildfire risk across the area. For example, the HPCC can help
compute FPI and can run fire spread models.”131
•
Machine learning is being incorporated too:
o
“Rolinski [SCE Fire Scientist] and SCE’s meteorologists are already using the
supercomputer for help in forecasting elevated fire conditions, but he said the data will
really pay off over the next year as they build their database and apply machine learning
to the historical data.” 132
2020 — FPI 2.0 development declared “completed”
•
Q4 2020: “Completed development of FPI 2.0 … formulas were completed in Q4.” 133

129 2019 WMP, supra note 55, at 62. A nearly identical sentence appears in D.19-05-038, § 8.1, at 30.
130 2019 WMP, supra note 55, at 63. SCE also states: “Having this information will enable SCE to more accurately understand the fuel’s
receptivity to fire. In addition, this dataset will be used to run fire spread models which will determine potential risks of past, current, and future
event scenarios. Furthermore, the HPCC will be used to generate and store weather and fuel conditions over a 30 plus year period which will
provide valuable insight into the nature of wildfire behavior and allow SCE to relate weather and fuel parameters to historical fire occurrences.”
131 D.19-05-038, § 8.1, at 31 (“SCE Proposal – Situational Awareness and Alternative Technologies”). The passage is the Commission’s
summary of SCE’s plan, not SCE’s own text.
132 Mary Ann Milbourn, SCE Harnesses Technology to Hone Wildfire Forecasts, Energized by Edison (Sept. 6, 2019),
https://energized.edison.com/stories/sce-harnesses-technology-to-hone-wildfire-forecasts.
133 2020 Annual Report on Compliance, supra note 56, att. A, at 7.


Page 37 of 54
2021 — FPI 2.0 to run in parallel; a 2022 launch date is set and FPI
thresholds are raised
•
SCE told regulators (Exhibit 14): “it created a fuels index and weather component for FPI 2.0,
and then backcasted the FPI 2.0 calculations 40 years. SCE also had FPI 2.0 calculated for each
Fire Climate Zone back to 1980 and operationalized to produce daily circuit‐level output.”134
•
2020 performance (Exhibit 13): Backcast FPI for calibration. Developed FPI 2.0 equations
emphasizing wind speed and diversity of fuel conditions. 135
•
Projected target by end of 2021: “1) Backcast 20 years of FPI using FPI 2.0 before typical height
of fire season (Q3) to determine historical performance compared to current FPI 2) Run FPI 2.0
in parallel with the current FPI and compare outputs for the 2021 fire season.” 136
•
Projected target by end of 2022 (Exhibit 13): “FPI 2.0 to replace current FPI” by end of 2022.137
•
September 1, 2021: SCE’s PSPS activation threshold raised from FPI 12 to 13 “for most areas
and most events.”138
o
“In 2021, SCE calibrated the index and was able to raise FPI thresholds across much of
its HFRA as a result.”139
2022 — Edison admits the legacy FPI caps wind speed, then misses its fire-
science targets
•
SCE, in writing (Exhibit 16): “The uncertainty associated with the current FPI is due to the
inherent error in the various weather and fuel inputs and also due to limitations of the model.
These limitations include: 1) all three components (weather, fuels, and green-up) are essentially
weighted the same, and 2) the wind speed input is capped at 29 mph which limits the index’s
ability to account for wind events stronger than that.”140
o
“For example, with the current FPI, fire potential is capped at 17 which limits its ability
to differentiate high end events. Also, weather, fuel moisture, and green-up are
essentially weighted the same in the FPI, which does not reflect the realities of how
these factors each contribute to fire potential. For example, wind speed should have a
higher weighting since wind can dominate the fire environment. Finally, the fuels
portion of the index is heavily dependent on live fuel moisture, but there are other fuel
moisture variables to consider that are equally important. SCE formulated a new FPI
(2.0) to address the limitations stated previously, by placing more emphasis on wind
speeds and adding a new fuels component to account for the diversity of fuel conditions
across the SCE’s service area.”141 (emphasis added by MW)

134 2022 WMP Update, supra note 5, at 276.
135 2021 WMP Update (Rev.), supra note 56, at 107 (PDF Pagination).
136 Id.
137 Id. (Fire Science, SA-2, Sec. 7.3.2, Projected Target: “FPI 2.0 to replace current FPI” by end of 2022).
138 2022 WMP Update, supra note 5, at 548; see also Payne Letter, supra note 56, at 3 (activation threshold set at FPI 13 “for most areas and most
events” starting September 1, 2021).
139 2023–2025 WMP R3.1, supra note 8, at 516. The same sentence appears verbatim in 2026 PSPS Action Plan, supra note 5, § II.D, at 25.
140 2022 WMP Update, supra note 5, at 97 tbl. Col. 18.
141 Id. at 275.


Page 38 of 54
•
Replacement “before the start of the 2023 fire season.”142
•
December: calibration target reported “met”143
•
Tom Rolinski, SCE Fire Scientist (Exhibit 17): “FPI 2.0 has been in development and is now
being evaluated against the current FPI at both the Fire Climate Zone level and at the circuit level.
While FPI 2.0 still needs to be calibrated against historical fire data to develop breakpoints and
PSPS thresholds, SCE can still compare its output with the current FPI on a daily basis. Initial
results show FPI 2.0 is more sensitive than the current FPI as it is more responsive to
changes in windspeed.” (emphasis added by MW)144
•
At the same time, Edison said it had missed its 2021 “fire science” targets because other work
was higher priority. It specifically claimed this did not affect the pace of FPI 2.0 development.145
o
“SCE did not meet target. Vendor developed a climatology output containing a 40-year
history of wildfires for multiple variables but unable to complete because vendor work
was reprioritized to support other emergent work.”146
2023 — No launch date, FPI 2.0 is “slowly integrating,” and the GRC
budgets $75,000 then nothing
•
2023–25 WMP: FPI 2.0 “slowly integrating” — target launch date removed.147
•
2025 GRC workpapers: ~$75K for FPI 2.0 over three years and $0 in 2024–2028.148
•
By SCE’s own account, the model existed and was producing output.
2024 — Still “developing” FPI 2.0, 54 days before Eaton
•
November 14, 2024 — 54 days before Eaton: SCE refiles the FPI 2.0 passage: it “is developing”
FPI 2.0 to put “more emphasis on wind speed” and “highlight extreme events.”149
•
The capped index remains in service for Eaton.
2025 — Eaton burns under the legacy FPI, and the goal posts move again
•
January 7, 2025: shut-off decisions rely on the legacy FPI.
•
2025 GRC proposed decision – goal posts moved again: “In 2022, SCE formulated a new Fire
Potential Index (FPI 2.0) by placing more emphasis on wind speeds and adding a new fuels

142 Id. at 277.
143 S. Cal. Edison Co., Q4 2022 Quarterly Notification, Docket No. 2022-SCs, at 4 (Cal. Off. of Energy Infrastructure Safety Feb. 1, 2023),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2022/SCE%20Q4%202022%20Quarterly%20Notification%20wit
h%20Attachment.pdf.
144 S. Cal. Edison Co., Response to Energy Safety Data Request OEIS-SCE-22-003, Question 03 (Program Targets – Fire Science SA-8), at 1–2
(Mar. 25, 2022) (prepared by Tom Rolinski, Fire Scientist; request received Mar. 22, 2022) [hereinafter OEIS-SCE-22-003 Response].
145 Id.
146 2022 WMP Update, supra note 5, at 128; OEIS-SCE-22-003 Response, supra note 143; S. Cal. Edison Co., 2022 Wildfire Mitigation Plan
Update: First Supplement, at 32, tbl. 5.3-1, 2021 (perf.) (Apr. 8, 2022),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2022/2022%20WMP%20Update_First%20Supplement.pdf.
147S. Cal. Edison Co., Redlines to SCE’s 2023–2025 Wildfire Mitigation Plan, at 516 (Apr. 2, 2024) (containing the quoted language).
148 2025 GRC Workpapers, supra note 8.
149 2023–2025 WMP R3.1, supra note 8, at 516.


Page 39 of 54
component. SCE plans to continue to evaluate the performance of FPI 2.0 with the long-term
goal of integrating it into the PSPS decision-making process.”150
•
June 2025, five months after Eaton: SCE issues an RFP for consulting work to evaluate FPI
2.0.151
•
August 2025: SCE’s Q2 2025 Quarterly Notification to Energy Safety makes no reference to FPI
2.0.152
•
August 2025: SCE community presentation mentions updated FPI criteria driving 20-40% more
PSPS days.153
2026 — Edison’s “learnings” confirm FPI 2.0 never launched in time
•
January: 2026–28 WMP — “direct usage in PSPS decision-making is still under
consideration.”154
•
April: — “goal of implementing [FPI 2.0] in 2026.”155
•
May: PSPS Action Plan — refinements “primarily based on learnings from the January 2025
fires.”156
•
August 2026, 19 months after the Eaton Fire: SCE tells the CPUC (Exhibit 19) that the “latest
version” has “no cap on the weather component.”157 A week later, it tells a community
meeting (Exhibit 18) that, beginning in early September, “wind carries more weight in the PSPS
formula to align more closely with wildfire risk.”158 This fixes the error SCE described in 2022.

150 Cal. Pub. Utils. Comm’n, Decision on Test Year 2025 General Rate Case for Southern California Edison Company, A.23-05-010 (July 28,
2025), adopted as D.25-09-030 (Cal. P.U.C. Sept. 23, 2025), at 445:
https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M580/K788/580788967.PDF
151 S. Cal. Edison Co., Response to 2026–2028 Base WMP Revision Notice, R1, Docket No. 2026-2028-Base-WMPs, cover letter app. E, at 30–31
(Sept. 15, 2025) (“RN-SCE-26-03 Evidence”; “2025 FPI 2.0 Request for Proposal Summary”) (Doc5201065583).
152 S. Cal. Edison Co., Q2 2025 Quarterly Notification, att. A, at 3–4 (Aug. 1, 2025),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2026-
2028/SCE%20Q2%202025%20Quarterly%20Notification%20with%20Attachment.pdf (listing Fire Science activities without any mention of FPI
2.0).
153 S. Cal. Edison Co., Wildfire Safety Community Meeting: Acton, slide 5 (Aug. 27, 2025), https://www.sce.com/sites/default/files/custom-
files/PDF_Files/Acton_SCE_WSCM_PPT_8_27_25.pdf (“Updates to PSPS Program”); see also S. Cal. Edison Co., Community Safety Meeting
Archive, https://www.sce.com/outages-safety/wildfire-safety/community-safety-meetings/community-safety-meeting-archive (last visited Oct. 10,
2026); S. Cal. Edison Co., Wildfire Safety Community Meeting: Mono and Inyo Counties, slide 13 (Sept. 30, 2025) (presentation by J. Schneider,
PSPS Operations), https://www.sce.com/sites/default/files/custom-files/PDF_Files/Mono-Inyo-SCE_WSCM-Presentation_9_30.pdf (“Updates to
PSPS Program”: “When new FPI criteria applied, suggests 20-40% more PSPS activation days and event size up to twice as large in 2025
assuming similar weather to 2024”).
154 S. Cal. Edison Co., 2026–2028 Base Wildfire Mitigation Plan, Revision 3, at 431 (Jan. 22, 2026),
https://www.sce.com/sites/default/files/AEM/Wildfire%20Mitigation%20Plan/2026-2028/SCE%202026-
2028%20Base%20WMP%20R3%20(January%2022%2C%202026)a.pdf.
155 2025 Annual Implementation Report, supra note 8, § II.3, at 18, (Fire Science (SA-8) row: “While not fully implemented in 2025, SCE intends
to build on the progress with a goal of implementing in 2026”); see also id. Plan Objective 6, at 5 (FPI 2.0 “not fully implemented in 2025”; key
components operationalized as “FPI 1.5”).
156 2026 PSPS Action Plan, supra note 5, § II.D, at 25 (“During 2025, primarily based on learnings from the January 2025 fires, SCE made
additional refinements to its FPI”).
157 2026 Safety Culture Briefing, supra note 59, slide 18 (“PSPS Decision-Making Evolves as Risk Models Improve”); see also Press Release,
Cal. Off. of Energy Infrastructure Safety, Energy Safety and CPUC to Hold Public Briefings on Utility Safety Practices, Take Public Comment
(Aug. 27, 2024), https://energysafety.ca.gov/news/2024/08/27/energy-safety-and-cpuc-to-hold-public-briefings-on-utility-safety-practices-take-
public-comment/ (annual briefings “satisfy the requirement for board-of-director-level reporting under Public Utilities Code section 8389(e)(5) in
partial fulfillment of the requirements for a safety certification”).
158 Santa Barbara Community Meeting, supra note 59, slide 19.


Page 40 of 54
•
October 2026: SCE’s revised PSPS decision-making paper seems to supersede FPI with a Fire
Behavior Matrix whose weather component squares wind speed “because of its strong influence
on fire spread” (Exhibit 8).159
B. Exhibits 7–19, referenced in Part 1 and Appendix A
Exhibit 7: SCE’s own 2025 post-season filing concedes its current fire modeling had an “inability to capture ex-
treme wind events” and a new FPI will be implemented

Source: S. Cal. Edison Co., Amendment to 2025 PSPS Post-Season Report, R.18-12-005, § IV (Response to Q1), at 135 (Cal. P.U.C. July 24,
2026).
Exhibit 8: Edison’s post-Eaton model change is telling. The October 2026 Fire Behavior Matrix squares wind
speed; the model in service on January 7 capped it at 29 mph.
Edison’s new Fire Behavior Matrix seems to supersede FPI. Its weather component is wind speed squared times the
dryness of the air. Edison explains that wind is weighted twice because of its “strong influence on fire spread even

159 Q&Q Factors (Oct. 2026), supra note 80, at 3 (Fire Behavior Matrix; weather component formula and wind-speed weighting).


Page 41 of 54
when other conditions, such as less dry fuels, would otherwise reduce the overall fire threat.” Red box added by
MW.

Source: S. Cal. Edison Co., Quantitative and Qualitative Factors for PSPS Decision-Making, at 3 (rev. Oct. 2026). Red box added by Muddy
Waters.
Exhibit 9: Edison’s own after-action report on its October 22–24, 2022 PSPS event lists, as area for improvement
3, that “[m]anual intervention was necessary to produce in-event risk calculator results for sub-transmission


Page 42 of 54
circuits.” Its Corrective Actions Table rates the item High priority, calls for modifying the calculator “to account
for sub-transmission circuit impacts,” and dates it only “Q1.”

Source: S. Cal. Edison Co., Business Resiliency After Action Report: 2022.10.22 SCE PSPS Event, at 2–4, in S. Cal. Edison Co., Response to
Energy Safety Data Request OEIS-P-WMP_2023-SCE-006, Question 2, Docket No. 2023-2025-WMP-DRs, at 21–23 (PDF pagination) (June 8,
2023). Underline added by Muddy Waters.


Page 43 of 54
Exhibit 10: FireCast — what Edison told regulators, 2022–2026. In each post-season report from 2022 through
2025, Edison named the same “main challenge” with its in-event risk tool, in nearly the same words: circuits
outside its initial PSPS forecast could not be assessed. The year changed, a hedge was added in 2024, and the tool
was renamed in 2025. Maroon marks the words that changed from the prior filing.

Sources: S. Cal. Edison Co., 2022 PSPS Post-Season Report, at 27 (Mar. 1, 2023); S. Cal. Edison Co., 2023 PSPS Post-Season Report, at 25
(Mar. 1, 2024); S. Cal. Edison Co., 2024 PSPS Post-Season Report, at 56 (May 1, 2025); S. Cal. Edison Co., 2025 PSPS Post-Season Report
(Mar. 2, 2026), as amended by S. Cal. Edison Co., Amendment to 2025 PSPS Post-Season Report, at 137–38 (July 24, 2026). All filed in R.18-
12-005 (Cal. P.U.C.).
Exhibit 11: Edison’s training for circuits approaching their thresholds. Its 2024 exercise script moves circuits
through the same steps: imminent shut-off notices at about 81–88% of their wind thresholds, a shut-off request to
the Incident Commander at 94–99%, in every case “FPI meets criteria,” and promotion of circuits outside the plan
when winds run above forecast.
(A) 84% of thresholds, FPI met → imminent shut-off notices (2024 #13)

(B) 79% → “nearing imminent notification criteria. No action required yet” (2024 #154)

(C) A circuit that “isn’t in the POC or on the WATCH LIST,” at 82%, FPI met (2024 #153)



Page 44 of 54
(D) Promoted to the POC after winds run “higher than originally forecast” (2024 #155)

(E) 98% of thresholds, FPI met → “Request IC authorization to Open” (2024 #137)

Source: S. Cal. Edison Co., 2024 PSPS Functional Exercise Master Scenario Events List (“SCE 2024PSPSFunctionalExercise MSEL FINAL”;
page headers read “2023 . . . Full-Scale Exercise”). The script states no percentage rule and does not say whether “thresholds” means
activation or de-energization thresholds.




Page 45 of 54
Exhibit 12: “Rogue circuits”: Edison’s definition, and the term in use in its exercise scripts, 2022–2024.
(A) The April 2022 tabletop after-action report defines rogue circuits as circuits not forecast to meet Monitored Circuit
List criteria that meet or exceed PSPS thresholds in real time, and recommends a written de-energization procedure (pp.
2–3).




Page 46 of 54
(B) The May 2022 functional exercise script introduces Rim as a “Rogue Circuit” that “was not originally on POC,”
tracks it from 83% to 97% of its wind thresholds, and has the team request shut-off authorization rendered from the
Excel workbook with other rows and columns hidden).

(C) The 2024 functional exercise script promotes “rogue circuit BUTTERFIELD” into scope after Weather Services
reports winds “higher than originally forecast,” a circuit that earlier was “not in the POC or on the WATCH LIST”

Sources: S. Cal. Edison Co., After-Action Report: 2022 Public Safety Power Shutoff Tabletop Exercise, at 2–3 (Apr. 15, 2022); S. Cal. Edison
Co., 2022 PSPS Functional Exercise Master Scenario Events List (Excel workbook, sheet “MainMSEL”); S. Cal. Edison Co., 2024 PSPS
Functional Exercise Master Scenario Events List.


Page 47 of 54
Exhibit 13: February 2021: Edison’s sworn program target — “FPI 2.0 to replace current FPI” by the end of
2022.


Source: S. Cal. Edison Co., 2021 Wildfire Mitigation Plan Update § 5.3, at 104 tbl. 5.3-1 (List and Description of Program Targets, Last 5
Years) (Feb. 5, 2021) (“FPI (SA-2)” row). Red box added.


Page 48 of 54
Exhibit 14: February 2021: Edison’s FPI 2.0 was already in development and backcasting in 2020, and was
expected to “replace the current FPI before the start of the 2022 fire season” — the same sentence it would refile a
year later with “2023.”

Source: S. Cal. Edison Co., 2021 Wildfire Mitigation Plan Update § 7.3.2.4.1 (Fire Potential Index (FPI) (SA-2)), item 4 (“FPI 2.0 Development,
Testing (Backcasting) and Evaluation”), at 198 (Feb. 5, 2021). Red boxes added.
Exhibit 15: February 2021: Edison tells its regulator why FPI 2.0 exists — it “will put more emphasis on wind
speeds” — four years before the wind-blind index made the call on January 7, 2025.

Source: S. Cal. Edison Co., 2021 Wildfire Mitigation Plan Update § 7.3.2.4.1 (Fire Potential Index (FPI) (SA-2)), items 1–2, at 197 (Feb. 5,
2021). Red box added.


Page 49 of 54
Exhibit 16: FPI 2.0: Edison’s own 2022 account of the index it still runs — wind input “capped at 29 mph,” com-
ponents “weighted the same” — and the replacement it expected to deploy “before the start of the 2023 fire sea-
son.”



Page 50 of 54





Page 51 of 54

Source: S. Cal. Edison Co., 2022 Wildfire Mitigation Plan Update § 8.2.3, at 548 fig. SCE 8-7 (Fire Potential Index Equation) & accompanying
text (Feb. 18, 2022); id. § 7.3.2.4.1 (Fire Potential Index (FPI) (Fire Science SA-8)), items 1–2, 4, at 275–77; id. § 4.5.1, at 97 tbl. SCE 4-18
(Wildfire and PSPS Risk Model Inventory) (“Model Uncertainty” column of the “Fire Potential Index (FPI)” row, rendered with other columns
hidden).


Page 52 of 54
Exhibit 17: March 2022, in writing to its regulator: Edison’s fire scientist reports FPI 2.0 is “more responsive to
changes in windspeed” than the index it would leave in charge on January 7, 2025.

Source: S. Cal. Edison Co., Response to Energy Safety Data Request OEIS-SCE-22-003, Question 03 (2022 WMP Update, Program Targets –
Fire Science SA-8), response iii, at 1–2 (Mar. 25, 2022) (prepared by Tom Rolinski, Fire Scientist; request received Mar. 22, 2022) (rendered
continuously across the page break with the question and responses i–ii hidden). Red box added.


Page 53 of 54
Exhibit 18: September 2026, twenty months after Eaton: Edison tells customers wind now “carries more weight in
the PSPS formula” — the fix it described in February 2022.

Source: S. Cal. Edison Co., Wildfire Safety Community Meeting: Santa Barbara County, slide 20 (“Changes to PSPS Protocols”) (Aug. 27,
2026) (“Changes That Could Increase Impacts” panel, rendered with the “Changes Reducing Impacts” panel hidden),
https://www.sce.com/sites/default/files/custom-files/PDF_Files/WSCM_Presentation_8_27_Santa_Barbara.pdf.


Page 54 of 54
Exhibit 19: August 2026: Edison tells the CPUC the “latest version” of its FPI can “capture extreme events with
no cap on the weather component” — the limitation it had put in writing in February 2022.

Source: S. Cal. Edison Co., Interagency Annual Public Briefing on Safety Culture and Public Safety Power Shutoffs (PSPS), slide 18 (“PSPS
Decision-Making Evolves as Risk Models Improve”) (Aug. 20, 2026) (presentation by T. O’Toole & J. C. Anderson to the CPUC and Energy
Safety) (left text panel rendered with the backcast panel and graphic hidden). Red box added.