Autoresearch: Independent confirmation of US IOU distribution capex (~$51B / ~43% / +160%) and T&D equipment inflation (wires +150%, poles +50%) plus PBR-vs-rate-base status as of 2026

Generated by /autoresearch on 2026-09-19. Synthesized across 3 rounds from government HTML/PDF (EIA, BLS, OSTI/LBNL, NCUC) plus EEI HTML, NARUC HTML, and one secondary price-driver paper (Rhodium). Anchored by Grokipedia Electric_power_transmission (technical primer; no capex or PPI figures). See Provenance. Treat as raw material — review before promoting. No buy/sell/size. Do not re-rate PWR / ETN / GEV / FIX. Context: vault/projects/stock-market (mechanism aging-grid-replacement-to-td-capex-supercycle; steps 2/3/4 tagged partial on a single Columbia Energy Exchange podcast)

Summary

EIA primary does independently confirm the level and the 20-year growth: major-utility distribution capital investment reached $50.9 billion in 2023, up $31.4 billion / 160% from 2003, with the 2022–23 jump described as replacement, upgrade, and weather-hardening rather than load-driven expansion (EIA Today in Energy, 2024-11-18). EEI’s HTML industry page prints a higher 2023–24 pair — $56.7 billion (2023) and $60.2 billion (2024) nominal distribution investment, $635 billion cumulative since 2001 — and does not state a 43% share (EEI Industry Data).

The podcast’s “43% of IOU capital expenditures” is not independently confirmed on that denominator. The nearest fetched primary-adjacent figure is LBNL’s FERC Form 1 reconstruction: “Distribution is currently the largest source of CapEx (44% of the total in 2023)” among utilities that report all CapEx categories, with the 2023 distribution print $47.4 billion on 112 reporters (LBNL / OSTI 2483675, Dec 2024). LBNL’s May 2026 distribution-cost scoping study cites EEI (2025) as 31% of all electric IOU capital spending since 2015 being distribution equipment — a multi-year EEI share, not 43% (LBNL, May 2026). Do not collapse 31 / 44 / 43.

On equipment inflation: wood poles +50% since 2019 is independently confirmed by BLS PPI WPU08710101 (~+52% vs 2019 average as of August 2026). Wires/cables +150% is not a fetched official point estimate. BLS PCU3359293359291 (power wire and cable, nonferrous) is +116% vs 2019 average as of August 2026 and peaked near +167% in June 2025. Rhodium’s April 2026 paper states a 50–152% range across four categories (wire and cable, transformers, switchgear, wood poles) — that is a secondary range, not an EIA/EEI/FERC/10-K confirmation of the podcast’s two point estimates (Rhodium, 2026-04-21; BLS API WPU08710101; BLS API PCU3359293359291).

Performance-based ratemaking has not materially decoupled utility earnings from rate-base capex as of 2026. NARUC still writes the revenue requirement as RR = ra(RB) + O&M + D + T. LBNL still treats PBR as a medium-term overlay that augments cost-of-service, and still finds a CapEx-over-OpEx bias. North Carolina’s operating PBR statute is residential sales decoupling + small PIMs + a multi-year rate plan that still books billions of capital (NARUC Desk Reference §3; LBNL, May 2026; NCUC PBR report, 2025-03-31).

Findings

1. Distribution capex level and 20-year growth: $50.9B and +160% are EIA-primary

EIA’s 18 November 2024 Today in Energy note, built from FERC financial reports (via Ventyx Velocity Suite), is the independent source that matches the podcast’s level and growth almost exactly (EIA, 2024-11-18):

Capital investment in distribution infrastructure increased by $31.4 billion, or 160%, from 2003 to 2023.

More than one-fifth of this increase occurred between 2022 and 2023 when spending increased by $6.5 billion to a total of $50.9 billion as utilities replaced and upgraded aging equipment and installed new lines, transformers, and other equipment to help neighborhood electricity grids withstand extreme weather events and to manage the intermittency of renewable resources.

Arithmetic check: $50.9 − $31.4 = $19.5 billion in 2003; $31.4 / $19.5 ≈ 161%. The “nearly $51 billion” and “+160% over ~20 years” claims survive as EIA primary. The same note’s component lines are also load-bearing for the replacement/hardening reading: overhead lines/poles/towers $17.4 billion in 2023 (+11% y/y, +220% vs 2003); underground lines $11.8 billion (more than doubled over 20 years); line transformers $7.5 billion (+23% y/y, “supply chain and manufacturing issues”); distribution substations $6.1 billion (+184% vs 2003); customer-sited $5.1 billion; distribution storage $97 million → $723 million (same).

EIA is explicit that this is capital investment on the distribution system, not total distribution spending (capex + O&M + customer accounts). An older EIA note for 2019 had already separated those: utilities spent $57.4 billion on electric distribution in 2019, of which $31.4 billion was capital, $14.6 billion O&M, $11.5 billion customer expenses — and said distribution spending had outpaced customer and sales growth because the incremental dollars were “replacing aging equipment, modernizing… and fortifying distribution structures against weather-related damage” (EIA, 2021-07-20). The 2019 capex print ($31.4B) is the same dollar increment EIA later uses as the 2003–2023 increase, not a 2019 level — do not confuse the two $31.4 billion figures.

EEI’s HTML industry page, fetched 2026-09-19, is a second official industry series and is not identical to EIA (EEI Industry Data):

• Distribution investment: $56.7 billion (2023), $60.2 billion (2024) nominal.
• Transmission: $30.0 billion (2023), $32.6 billion (2024); $39.9 billion projected 2025; ~$178 billion planned 2025–2028.
• Cumulative distribution since 2001: $635 billion nominal.

EEI > EIA on the 2023 distribution print ($56.7B vs $50.9B). Different universe (EEI member IOUs / holding-company construction vs EIA “major utilities” FERC filers) and possibly different capital-vs-construction definitions. Both are above $50 billion. Neither HTML page states +160%.

LBNL’s December 2024 FERC Form 1 reconstruction is a third series: distribution CapEx $31.5B (2019) → $47.4B (2023) on the utilities that report that category (112 reporters in 2023) — +50% over 2019–2023, “more than double the rate of inflation” (LBNL / OSTI 2483675). $47.4B is below both EIA $50.9B and EEI $56.7B because Form 1 coverage is a subset (LBNL: 172 of 207 IOUs filed Form 1 in 2023; category counts differ). Use EIA for the podcast’s $51B / +160% pair; use LBNL for the 2019–23 growth rate and the 44% share (next section).

Step-2 implication: the podcast’s ~$51B and +160% (2003–2023) are independently confirmed by EIA. The “mostly replacing and hardening, not expanding capacity” clause is supported in direction by EIA’s 2023 and 2019 notes and by LBNL May 2026 (15 of 22 sampled utilities put managing the existing system as the largest planned distribution CapEx category; capacity expansion is largest for only 2). It is not a 100% replacement print: LBNL cites EEI (2025) that 2025 investment drivers were 33% adaptation/hardening/resilience, 28% replacement, 31% expansion (LBNL, May 2026).

2. The 43% share: not confirmed as “43% of IOU capex”; nearest fetched figure is 44% of a FERC Form 1 complete-reporter subset

No fetched EIA, EEI HTML, FERC, or utility 10-K page states “43% of investor-owned-utility capital expenditures.” EEI’s Industry Data HTML does not print a functional share at all (EEI Industry Data). EEI PDFs that search indexed as “Distribution $59.8B 32%” were not fetched (off the *.gov PDF whitelist) and are not treated as confirmation.

What was fetched:

• LBNL / DOE December 2024 slides, FERC Form 1: “Distribution is currently the largest source of CapEx (44% of the total in 2023)” among the stacked-bar subset of utilities reporting all CapEx categories. Absolute 2023 prints on (non-overlapping) reporter sets: generation $19.0B (75 utilities), transmission $20.0B (99), distribution $47.4B (112), other $10.8B (112). LBNL warns those absolutes cannot be summed across categories because the reporter sets differ (LBNL / OSTI 2483675, pp. 21, 43–44).
• LBNL May 2026: “EEI estimates that 31% of all electric IOU capital spending since 2015 has been for distribution equipment” (footnote: EEI 2025) (LBNL, May 2026, p. 13).
• Implied shares using unfetched EEI total-capex headlines that appeared only in search snippets ($168.9–178.2B) are not used here. Arithmetic on fetched EEI HTML alone cannot produce 43%: distribution $56.7B (2023) / transmission $30.0B is 65% of T&D, not of total capex.

Honesty rule: 43% is not independently confirmed. A nearby 44% exists, but it is LBNL’s complete-reporter electric functional CapEx share for 2023, not EEI holding-company total (which includes gas-related and other). A 31% multi-year EEI figure, cited by LBNL, cuts the other way. Do not launder the podcast’s 43% through the 44% as if they were the same statistic.

3. Equipment inflation since 2019: poles ~+50% confirmed; wires +150% not a fetched official point estimate

Wood poles — confirmed, BLS primary. Commodity PPI WPU08710101 (“Wood Poles, Piles, and Posts Owned and Treated by the Same Establishment”), monthly, June 1985=100, fetched via the BLS public API for 2019–2026 (BLS API WPU08710101):

The podcast’s “wood poles … 50% increase since 2019” is inside the BLS band (latest +52%; peak +61%). This is independent of the Columbia podcast.

Wires and cables — not confirmed as +150% official. Industry PPI PCU3359293359291 (“Power Wire and Cable, Made from Nonferrous Metals (Purchased Wire)”), December 1982=100 (BLS API PCU3359293359291):

A +150% print is not the August 2026 index. It is near the 2023–25 plateau (+141% to +167%). Stating “wires and cables +150% since 2019” as a current fact overstates the latest BLS print and understates the mid-2025 peak. The honest BLS statement is: power wire and cable roughly doubled-plus since 2019 and briefly approached +150–170% before easing.

Sibling BLS series (same API method) show other T&D kit also up, but not +150%:

• Power and distribution transformers except parts (WPU11740999): 2019 avg 195.39 → July 2026 381.783 = +95% (still rising; July is the peak in this pull) (BLS API WPU11740999).
• Switchgear, switchboard, industrial controls (WPU1175): 2019 avg 231.16 → August 2026 430.247 = +86% (BLS API WPU1175).

Rhodium (21 April 2026) is independent of the June 2026 podcast but is not EIA/EEI/FERC/10-K. It writes: “Wire and cable, transformers, switchgear, and wood poles have increased 50-152% in cost since 2019, compared to a 29% increase in the overall CPI” (Rhodium). That is a four-category range. It is the obvious sibling of the podcast’s “poles 50% / wires 150%” pair (low end ≈ poles, high end ≈ wire). Do not treat Rhodium as confirmation of those two point estimates. It is useful as a contemporaneous secondary that used the same PPI family; the confirmation work is the BLS series above.

LBNL May 2026 independently notes the same mechanism without the podcast numbers: “One of the reasons for recent growth in distribution CapEx is that equipment prices have grown well-above the pace of inflation”; “post-pandemic supply-chain constraints and elevated equipment prices contributed”; “prices since 2021 increased much faster than CPI and remain elevated” (LBNL, May 2026, p. 19). EIA’s transformer line (“+$7.5 billion … supply chain and manufacturing issues”) is the government spending-side corroboration, not a price index (EIA, 2024-11-18).

Step-4 implication: poles +50% — confirmed (BLS). Wires/cables +150% — not independently confirmed as a current official figure; BLS supports large, above-CPI inflation that peaked near that neighborhood in 2025 and has since eased to ~+116%. Transformer (+95%) and switchgear (+86%) are the rest of Rhodium’s range, still well above CPI.

4. PBR has not materially decoupled earnings from rate-base capex as of 2026

The structural incentive in the mechanism (earn the authorized return on capital built) is still the default US formula. NARUC’s commissioners’ desk reference states it without hedging (NARUC Desk Reference §3.1):

The basic formula for Revenue Requirement Determination can be expressed as follows: RR = ra (RB) + O&M + D + T where: … RB = rate base (original cost of invested utility plant in service net of accumulated depreciation & adjustments).

PBR appears in the same manual as an Alternative Form of Regulation that may be chosen over traditional cost-of-service if it protects consumers and is in the public interest. NARUC’s definition of PBR is “tying growth in revenues or rates to a metric other than costs” via multi-year rate plans, revenue decoupling, and/or PIMs — not “earnings no longer depend on rate base” (NARUC Desk Reference §3.5). Decoupling, as NARUC defines it, separates profits from total sales so revenue tracks the authorized requirement; it does not remove the return-on-rate-base term (§3.6). Trackers/riders (including Pennsylvania-style DSIC) reduce regulatory lag on capital and, NARUC notes, “can reduce the utility’s incentives to control costs and encourage utilities to spend more on infrastructure” (same).

NARUC’s PBR State Tracking Map was last updated 3 September 2026 and shows 51 “participating” states (NARUC PBR map). “Participating” on a developments tracker is not “51 states have replaced cost-of-service.” The map page fetched here has no state-level table in the extracted HTML (it is a map widget). Do not read 51 as a count of fully decoupled jurisdictions.

LBNL’s May 2026 DOE-funded scoping study is the most direct 2026 official-adjacent verdict on whether the build-to-earn flywheel has broken (LBNL, May 2026):

Cost-of-service regulation for IOUs can create utility incentives that are misaligned with least-cost electric service.

There is evidence that utility ROEs are higher than financial theory suggests are necessary, resulting in a bias towards CapEx over potentially cheaper OpEx solutions, which could lead to higher electricity rates.

Cost recovery tracker mechanisms have become more common, but risk further weakening utility incentives for cost containment if not carefully designed.

PBR is listed there as a medium-term tool to augment COSR, not a completed national shift: “Augment cost-of-service regulation to strengthen utility performance incentives — Use multi-year rate plans to contain utility costs by reducing frequency of rate cases, and apply targeted performance incentives” (same, p. 9). IOU rate-increase requests hit $18 billion in 2025; PUC approval of requested amounts averaged 64% over 2021–2025, both “multi-decadal highs” (same, p. 7). That is the opposite of a system that has stopped paying for rate-base growth.

North Carolina is the most fully specified 2025–26 PBR implementation fetched. N.C.G.S. § 62-133.16 defines PBR as the bundle of (1) a residential decoupling mechanism (RDM), (2) one or more PIMs, and (3) a multi-year rate plan (MYRP) including an earnings-sharing mechanism (ESM) (NCUC report, 2025-03-31). The annual review still tests whether earnings exceeded authorized ROE (ESM refund) and then trues up PIM and RDM riders. A Duke Energy Progress order cited in that report approved, for April 2025–March 2026: RDM recovery of $39.4 million (0.2324 ¢/kWh), ESM $0.00, and a PIM rider of 0.002 ¢/kWh ($0.71 million) (same). Sales decoupling and a 0.002 ¢/kWh PIM do not take the return off rate base. SEC 8-K snippets that surfaced in search (Duke Carolinas / Progress MYRP capital on the order of $3.8–4.4 billion, T&D ~55–60% of the MYRP) were not fetched (SEC 403 this pass) and stay unverified.

Rhodium, secondary, is directionally the same: authorized ROEs “held on average around 9–10%” from 2019 through 2025; “Building more infrastructure at the same approved RoE would, all else equal, increase utility profits. This highlights the underlying economic incentive for IOUs to spend more capital” (Rhodium, 2026-04-21). California ROE cuts of 30 bp are noted as politically salient and not nationally decisive (same).

Step-3 implication: the mechanism’s falsifier — “widespread adoption of PBR that decouples earnings from capital deployed” — has not fired as of 2026-09-19. PBR is spreading as MYRP + sales decoupling + small PIMs on top of rate base, and in NC it still schedules capital into the MYRP. Arent’s podcast hedge (“evolving landscape… no state really stands out”) is more consistent with the fetched official record than a “PBR has arrived and broken the flywheel” reading.

5. What this pass does not do

• Does not re-rate or size PWR, ETN, GEV, or FIX.
• Does not treat the Columbia Energy Exchange podcast as a source, and does not treat a secondary recap of that podcast as confirmation.
• Does not fetch EEI PDFs (off-whitelist). EEI HTML and LBNL’s citation of “EEI (2025)” are the EEI-adjacent record.
• Did not obtain a 2025/2026 utility 10-K body (SEC 403 on the Duke filing this pass). Replacement-vs-expansion color is EIA + LBNL, not issuer IR.

Add-evidence quotes (for mechanism steps 2–4)

Use these as independent, non-podcast citations. Do not paste the podcast numbers back in as if EIA/BLS had uttered them in the podcast’s exact wording.

Step 2 (capex wave / replacement-hardening):

1. EIA (18 Nov 2024): “Capital investment in distribution infrastructure increased by $31.4 billion, or 160%, from 2003 to 2023.” “spending increased by $6.5 billion to a total of $50.9 billion as utilities replaced and upgraded aging equipment and installed new lines, transformers, and other equipment to help neighborhood electricity grids withstand extreme weather events.” (https://www.eia.gov/todayinenergy/detail.php?id=63724)
2. LBNL May 2026: “Among a sample of 22 utilities, 15 identify managing the existing distribution system as the largest category of planned distribution capital expenditure; capacity expansion is the largest CapEx category for just 2 utilities.” “Asset replacement, safety and reliability, and resilience are all important drivers of CapEx for managing the existing system.” (https://eta-publications.lbl.gov/sites/default/files/2026-04/utility_distribution_costs_final.pdf)
3. Share caveat, do not file as 43% confirmation: LBNL/OSTI Dec 2024: “Distribution is currently the largest source of CapEx (44% of the total in 2023).” LBNL May 2026: “EEI estimates that 31% of all electric IOU capital spending since 2015 has been for distribution equipment.”

Step 3 (incentives still on rate base; PBR has not decoupled):

1. NARUC: “RR = ra (RB) + O&M + D + T” where RB is rate base. (https://www.naruc.org/commissioners-desk-reference-manual/3-ratemaking-fundamentals-and-principles/)
2. LBNL May 2026: “There is evidence that utility ROEs are higher than financial theory suggests are necessary, resulting in a bias towards CapEx over potentially cheaper OpEx solutions.” PBR is listed as a tool to “Augment cost-of-service regulation.” (https://eta-publications.lbl.gov/sites/default/files/2026-04/utility_distribution_costs_final.pdf)
3. NCUC: statutory PBR is RDM + PIMs + MYRP/ESM; DEP’s 2025–26 PIM rider was 0.002 ¢/kWh. That is not earnings-decoupled-from-capex. (https://www.ncuc.gov/reports/2025performancereg.pdf)

Step 4 (equipment inflation):

1. BLS WPU08710101 wood poles: 2019 average 271.06 → August 2026 411.076 (+52%). (https://api.bls.gov/publicAPI/v2/timeseries/data/WPU08710101?startyear=2019&endyear=2026)
2. BLS PCU3359293359291 power wire and cable: 2019 average 248.82 → August 2026 536.495 (+116%); June 2025 peak 665.616 (+167%). Not a fetched +150% official current print. (https://api.bls.gov/publicAPI/v2/timeseries/data/PCU3359293359291?startyear=2019&endyear=2026)
3. Optional secondary (do not file as confirmation of the podcast split): Rhodium 21 Apr 2026, “Wire and cable, transformers, switchgear, and wood poles have increased 50-152% in cost since 2019, compared to a 29% increase in the overall CPI.” (https://rhg.com/research/us-electricity-prices-drivers/)

Contradictions and open questions

• $50.9B (EIA) vs $56.7B (EEI) vs $47.4B (LBNL/FERC Form 1) for 2023 distribution capex. Three official-adjacent series, three universes. The podcast’s “nearly $51 billion” matches EIA, not EEI or LBNL.
• 43% vs 44% vs 31%. 43% remains unfetched on the “share of IOU capex” denominator. 44% is LBNL’s 2023 complete-reporter functional share. 31% is LBNL citing EEI (2025) since 2015. Do not upgrade step 2’s 43% clause to confirmed.
• Replacement-not-expansion is directional, not exclusive. EIA and LBNL support aging + hardening as the main driver; EEI-via-LBNL still books 31% of 2025 investment as expansion.
• Wires +150% vs BLS +116% latest / +167% peak. The podcast number is a peak-era fit, not the August 2026 print. Transformer (+95%) and switchgear (+86%) were inside Rhodium’s range and were not named in the podcast.
• PBR “51 participating states” vs “no state really stands out.” NARUC’s 2026-09-03 map counts participation in a tracker, not completed decoupling. NC’s implemented PBR still pays on MYRP capital. Hawaii’s 2025 rebasing request (traditional rate case inside a PBR state) appeared only in a January 2026 secondary and was not fetched.
• No 2025/2026 utility 10-K body this pass (SEC 403). Issuer-level replacement vs new-business capex mix remains a gap.
• Grokipedia Electric_power_transmission is a physics/history primer (US transmission losses ~2–3%, AC vs HVDC, 2003 blackout). It does not carry the $51B / inflation / PBR claims and was not used as evidence for them.

Provenance

Rounds run: 3 of 3 (full)

Sub-questions by round:

Round 1 (broad survey):

1. What do EIA / FERC primary tables say about IOU distribution capex level, ~20-year growth, and share of total IOU capex?
2. What does EEI Industry Data (HTML) report for distribution vs transmission investment?
3. What do BLS PPI series say about T&D equipment inflation (wires/cables, wood poles) since 2019?
4. What is the current US status of performance-based ratemaking versus cost-of-service / rate-base ROE?

Round 2 (drill-down):

1. Can the 43% share be found on an EIA / FERC Form 1 / LBNL primary, or computed from a fetched total-capex denominator? — targeting the unconfirmed 43%.
2. What do BLS transformer and switchgear PPIs show since 2019, and does any official series print +150% for wire/cable? — targeting the equipment-inflation gap.
3. Has PBR, in states that actually implemented it (NC and the NARUC tracker), removed the return-on-rate-base term? — targeting step 3.

Round 3 (resolve remaining uncertainty):

1. Confirm LBNL’s 44% / $47.4B FERC Form 1 table from the OSTI PDF, and LBNL May 2026’s EEI 31% plus replacement-vs-expansion and COSR-bias language. — targeting the share contradiction and the PBR overlay.

Anchor source (Grokipedia, fetched before round 1):

• Electric power transmission — 25,017 chars extracted (cap), 259 citations, source: api via _lib/grokipedia.py — technical and historical primer (AC vs DC, overhead vs underground, US losses ~2–3%). No IOU capex totals, no PPI, no PBR. Search also surfaced investor_owned_utilities_in_the_united_states, utility_ratemaking, and performance_based_regulation; those slugs were not fetched (budget).

URLs fetched (14 successful, 3 failed / skipped):

Round 1:

• EEI Industry Data — official industry HTML — $56.7B / $60.2B distribution, $30.0B / $32.6B transmission, $635B since 2001. No 43%.
• EIA Today in Energy, 18 Nov 2024 — official (FERC-based) — $50.9B, +160% / +$31.4B, 2003–2023; replacement/hardening language. WebFetch timed out; recovered with browser-UA curl.
• EIA Today in Energy, 20 Jul 2021 — official — 2019 distribution $57.4B total / $31.4B capex; spending outpaced sales because of replacement and fortifying.
• Rhodium, Caught in the Current, 21 Apr 2026 — secondary research shop — 50–152% equipment range; IOU distribution capex +35% / +$14B since 2019; ROE still 9–10%; CapEx incentive. Not treated as confirmation of the podcast’s two point estimates.
• BLS PPI WPU08710101 — official — wood poles +52% vs 2019 avg as of Aug 2026.
• BLS PPI PCU3359293359291 — official — power wire and cable +116% latest / +167% Jun 2025 peak.
• [Failed: FRED CSV endpoints for the same series] — HTTP/2 INTERNAL_ERROR then timeout; BLS API used instead.

Round 2:

• NARUC PBR State Tracking Map — official association HTML — updated 2026-09-03, 51 participating states. Map widget; no per-state table in the extract.
• NARUC Desk Reference §3 — official — RR = ra(RB)+…; PBR as AFOR overlay; decoupling = sales, not capex.
• LBNL Electric Utility Distribution Costs, May 2026 — DOE-lab PDF (lbl.gov, whitelist) — EEI 31% since 2015; 15/22 existing-system vs 2/22 expansion; COSR CapEx bias; PBR as medium-term overlay; EEI 2025 mix 33/28/31 hardening/replacement/expansion.
• NCUC PBR implementation report, 31 Mar 2025 — state-government PDF — NC PBR = RDM + PIMs + MYRP/ESM; DEP PIM 0.002 ¢/kWh.
• BLS PPI WPU11740999 — official — transformers +95% vs 2019 avg (Jul 2026).
• BLS PPI WPU1175 — official — switchgear +86% vs 2019 avg (Aug 2026).

Round 3:

• OSTI biblio 2483675 — DOE landing — identifies LBNL/NREL Dec 2024 slides.
• OSTI servlet / LBNL Retail Electricity Price and Cost Trends 2024 Update — DOE-lab PDF — 44% of 2023 CapEx; $47.4B distribution CapEx; 2019–23 distribution CapEx +50%; remaining under COSR.
• [Failed: Duke Energy 10-K HTML on sec.gov] — HTTP 403. Replacement-vs-expansion issuer mix left to EIA + LBNL.

Not fetched (and not treated as confirmed): EEI Industry Capital Expenditures PDF; EEI Financial Review PDFs; ScottMadden 2023 industry update PDF; Oregon PUC / Virginia SCC / Indiana IURC PBR PDFs (search-indexed; budget). FRED HTML pages timed out; numbers taken from BLS API instead.

Tools used: WebSearch, WebFetch, _lib/grokipedia.py fetch, curl (browser UA) for EIA / BLS API / *.gov PDFs after WebFetch timeouts. No X pass (--include-x not set). Priors skipped per request. Generated: 2026-09-19 21:00 UTC