Hollis Select Board Transparency Archive

PH7PR19MB589572803B64856395072A44B3EC2

RE: memo on Brookfield

From

• Renee Lewis <[email protected]>

To

• Jim Boutin ([email protected]) <[email protected]>
• Terry Walters <[email protected]>
• Paul Mattor <[email protected]>
• Richard Morin <[email protected]>
• Dan Yarumian <[email protected]>
• Mark Woodruff <[email protected]>

Date

Thu, Jun 25, 2026 · 1:09 p.m. ET

• Brookfield_Renewable_US_Disposals_with_Locations_2015-2025 (6).docxOpen original

  Read from the file

  Brookfield Renewable US Asset Disposals — Ledger with Locations, 2015–2025 US-located disposals, including orphaned / non-core sales. Locations are as named in filings and contemporaneous reporting; “not disclosed” means the company did not specify states or sites. Compiled June 2026. ◆ HYDRO SPOTLIGHT — IDENTIFIED The two large hydro portfolios both resolve to the same asset: Smoky Mountain (formerly Tapoco) — four dams (Cheoah, Santeetlah, Calderwood, Chilhowee) on the Little Tennessee and Cheoah Rivers across east Tennessee and western North Carolina, FERC Project No. 2169. Brookfield bought it from Alcoa for $600M in 2012, sold 50% to Argo Infrastructure Partners in 2023, and sold the remaining 50% across 2025–early 2026 to fully exit. “Non-core” reflects portfolio fit, not asset quality: Smoky signed a 10-year TVA offtake (2024) and supports a $435M private placement (2025). The 30 MW plant (2024) remains unidentified by state. None was sold at a loss.

  Year Asset &amp; capacity Location (as disclosed / identified) Buyer Gain / loss &amp; notes 2015 Wind facility — 102 MW California Third party (not named in filing) Gain $53M ($12M net). BEP held 22%. 2016 — none identified — — Acquisition year (Isagen; Holtwood/Pennsylvania hydro). 2017 — none identified — — Acquired 51% of TerraForm Power. 2018 — none identified — — Raised TerraForm stake to 65%. 2019 — none identified — — Disposals were Europe / Asia / South Africa. 2020 Wind portfolio — 852 MW (40% interest) United States (states not disclosed; TerraForm-era assets) Not disclosed in filing FFO impact $15M, 760 GWh. Blended w/ Ireland. 2021 Wind portfolio — 391 MW (4 farms) California (Alta Wind VIII 150 MW, Windstar 120 MW, Coram 22 MW) + New Hampshire (Granite 99 MW) NextEra Energy Partners Mature onshore wind; ~2x invested capital. 2022 — none in US — (US held-for-sale building up) — Year&apos;s disposals were Asia solar (19 MW). 2023 Q1 ◆ HYDRO — 378 MW Smoky Mountain (78% interest) TN &amp; NC — Cheoah, Santeetlah, Calderwood, Chilhowee dams (Little Tennessee / Cheoah Rivers; FERC P-2169) Argo Infrastructure Partners (50%) + Brookfield Corp affiliates (28%) Deconsolidated. BEP kept 22%, no proceeds. 2023 Q2 Wind assets (acquired 2022) United States (not further specified) Not disclosed Quick resale of recently acquired wind. 2024 Q2 ◆ HYDRO — 30 MW asset United States (single plant; state not disclosed) 3rd party + institutional partners Gain $29M ($6M net). Sold above book. 2024 Q2 Battery storage — 60 MW United States (state not disclosed) 3rd party + institutional partners Gain $5M ($2M net). 2024 Wind — 845 MW (partial; Shepherds Flat) Oregon (repowered onshore wind) Institutional buyer (50% across 2024–25) One of largest US wind repowerings. 2025–26 ◆ HYDRO — 403 MW Smoky Mountain, remaining 50% TN &amp; NC — same four dams as 2023 (full exit of the system) 3rd party (25%) + BAM-managed fund (25%) “Non-core” = portfolio fit, not quality. TVA PPA + $435M financing in place. 2025 Distributed generation — 700 MW United States (multi-site; Luminace-related) 3rd party (47%) + BAM-managed fund (53%) Part of majority Luminace exit. 2025 Solar portfolio — 833 MWdc (613 MWac) United States (sites not disclosed) Not disclosed Agreed subsequent to Q3 2025. 2025 Solar — Deriva platform United States (sites not disclosed) Not disclosed ~$400M ($70M net). Closed Q1 2026. 2025–26 Wind &amp; solar — ~2,300 MW (two-thirds) United States (dev-platform-built; sites not disclosed) Buyer consortium (framework w/ recurring sales) Closes H1 2026; +$1.5B future framework. Amber rows (◆) = hydro disposals. “BAM-managed fund” = a private fund managed by Brookfield Asset Management (an affiliate); such partial sales are stated to be at the same value agreed with the unaffiliated third party. Ranges reflect figures refined between quarterly filings. Notes on location disclosure A key finding for your geographic focus: Brookfield discloses asset location only sparingly in its own filings. The disposal notes almost always read simply “in the U.S.” without naming states, rivers, or plants. Where precise locations appear in this ledger, they generally come from the buyer’s disclosure or trade-press coverage, not Brookfield’s. The standout example is the 2021 sale to NextEra, which named all four wind farms and their states. Where locations are firmly established 2021 — 391 MW wind to NextEra Energy Partners ($733M): Alta Wind VIII (150 MW), Windstar (120 MW) and Coram (22 MW) in California, plus Granite (99 MW) in New Hampshire. This is the most geographically precise US disposal on record. 2024–25 — the 845 MW Shepherds Flat wind sits in Oregon. 2015 — the 102 MW wind facility is in California. Where location was recovered through triangulation The two large hydro portfolios (2023 and 2025) were disclosed by Brookfield only as “in the U.S.,” but both were identified as Smoky Mountain by cross-referencing the buyer (Argo Infrastructure Partners), the offtaker (a 2024 TVA power-purchase agreement that named all four dams), the lenders (a 2025 $435M private placement), and a sister Brookfield fund’s holdings table that lists “U.S. Hydro (Smoky Mountain).” The capacity figure (377–403 MW) was the thread tying the vague filing to the named asset. The 30 MW plant (2024), the 2020 wind, and the 2025 solar/DG portfolios remain unspecified as to site — those would need the same buyer-or-FERC reconstruction. Geographic pattern that emerges US wind disposals cluster on the West Coast / mountain-West (California, Oregon) with one Northeast outlier (New Hampshire). The identified hydro disposals both sit in the Tennessee Valley / western Carolinas (the Smoky Mountain four-dam system) — a cluster somewhat apart from Brookfield’s larger Northeastern, New York and Maine hydro concentrations, which is the most likely reason it carried the “non-core” label. FERC license surrenders &amp; decommissioning by Brookfield entities Separate from sales, you asked which Brookfield Renewable entities have not renewed a FERC license or moved to decommission a plant. These are concentrated in Maine and are matters of relicensing conflict rather than asset sales. The clearest cases: ▸ Bar Mills (FERC No. 2194), Saco River, Hollis, Maine — Brookfield White Pine Hydro filed a notice of intent to surrender the license on Nov 30, 2020, the first step toward decommissioning. The units had not run for three years; a surrender application and decommissioning plan (partial dam breach) followed. This is the most clear-cut voluntary surrender — driven by an inoperable, uneconomic-to-repair plant. ▸ Ellsworth Project (Union River &amp; Graham Lake dams), Maine — operated by Black Bear Hydro Partners. Its last license expired in 2017; it has run on annual temporary licenses since. Maine DEP denied the required water-quality certification (2020, and a draft denial again in 2025), and the company has said removal of the Ellsworth and Graham dams “may be an unfortunate but necessary outcome.” Not yet surrendered, but relicensing is in serious jeopardy — a possible forced (non-voluntary) decommissioning. ▸ Lower Kennebec dams — Shawmut (FERC No. 2322), Lockwood, Hydro-Kennebec, Weston, Maine — operated by Brookfield White Pine Hydro, Merimil LP and Hydro Kennebec LLC. State agencies (Maine DMR, NMFS) and a 2020 Kennebec River Management Plan amendment have pushed to remove these four dams over endangered-fish concerns; Maine DEP denied water-quality certification for Shawmut’s relicensing. Brookfield has litigated to block removal (including a D.C. Circuit case and an injunction against Maine DMR). Contested and unresolved — relicensing opposed, removal not yet ordered. Focus risk: Bar Mills retained powerhouse &amp; the municipal abandonment problem The Bar Mills surrender carries a specific long-tail risk worth isolating. BWPH is not fully removing the project. Per its own plan, it will retain the powerhouse (1956 structure; original project 1919) “for operational purposes, such as materials and equipment storage and office space,” and will leave a portion of the spillway and the entire eastern half of the dam in place. The powerhouse will not be demolished; no soil testing is proposed around it for that reason. The mechanism that creates the risk: per BWPH’s own FAQ, “once the surrender application has been approved by FERC, there is no longer a FERC project boundary.” When surrender is granted, FERC’s safety and financial jurisdiction ends entirely. The retained powerhouse, spillway remnant and water-retaining structures become ordinary private real estate under state/local authority only. There is no FERC decommissioning bond, no federally-mandated maintenance trust, and no federal obligation that runs with the land to a future owner. BWPH’s maintenance duty binds BWPH under the FERC-approved decommissioning plan — it is not clearly a perpetual covenant enforceable against a downstream purchaser after federal jurisdiction terminates. So the scenario of concern is real: if BWPH (or a successor Brookfield entity) later sells or abandons the parcel to a thinly-capitalized buyer that cannot fund upkeep or removal of an aging, partially-breached dam and powerhouse, the residual liability could fall to the municipality. Enforceability would depend entirely on state dam-safety law and whatever deed covenants exist — not on FERC. The retained structures are not liability-free. The un-removed eastern dam sits adjacent to the Rogers Fibre Mill EPA Superfund site, and its concrete foundation is itself a water-retaining structure FERC required inside the project boundary. Decommissioning also affects municipal infrastructure (the Canal Road dry hydrant, which BWPH proposes to relocate in consultation with Hollis). Geographic nuance: the exposure is split between two towns. The retained powerhouse and BWPH-owned access road, parking and canal lands are on the Hollis side; the un-removed eastern dam portion and the adjacent Superfund parcel are owned by the Town of Buxton. Hollis’s exposure centers on the powerhouse and western structures BWPH retains; Buxton carries parallel exposure on the eastern remnant on its own land. A small town’s inability to absorb future maintenance or removal cost applies to both, and is not addressed by any federal financial-assurance mechanism in the surrender process. Potential mitigations a stakeholder could pursue (not legal advice): request, during the open FERC surrender docket, that the decommissioning order condition approval on (a) a funded maintenance/removal escrow or surety that survives transfer, (b) recorded deed covenants binding successors-in-interest to maintain or remove the retained structures, and (c) a state dam-safety registration that names a financially-responsible party. These are the kinds of conditions that, if not imposed before surrender is granted, become very difficult to attach afterward — because FERC jurisdiction will have ended. Is the Hollis fear precedented? Comparable cases The concern — an aging dam/powerhouse passing to an owner without the balance sheet to maintain or remove it, ultimately landing on the public — is a documented failure mode, not speculation. The enabling structure is the single-purpose LLC holding one hazardous structure with no parent guarantee. Key precedents: ▸ Cheboygan, Michigan (closest analog). A near-perfect realization of the Hollis fear. The Cheboygan Lock &amp; Dam complex is split: the Michigan DNR owns the 1922 dam, while a private LLC owns the adjoining hydropower plant that once served a now-shuttered tissue factory. The FERC license exemption traces back to Procter &amp; Gamble (1983) and passed through a chain of owners (Great Lakes Tissue, then Hom Paper XI LLC, controlled by a former NFL linebacker). The powerhouse went offline after a September 2023 fire. In July 2025 FERC ordered Hom Paper to restore the powerhouse or surrender its exemption; the owner’s lawyer wrote that operating it was “no longer economically feasible,” then sought repeated extensions while trying to offload the plant to another small entity (HydroMine Cheboygan LLC, a Wyoming firm). During the limbo, record spring flooding in April 2026 brought the high-hazard dam within ~6 inches of overtopping; an emergency multi-agency scramble (Consumers Energy, Army Corps, DNR, HydroMine) restarted the idle turbine just in time. FERC then invoked Section 31 of the Federal Power Act (civil-penalty authority). A state dam-safety watchdog called it “the same pattern” as Boyce and Au Train. Note the structural identity to Bar Mills: a split public/private ownership, an inoperable powerhouse held by a thinly-capitalized LLC, and an owner seeking to exit an asset it deems uneconomic. ▸ Boyce Hydro / Edenville Dam, Michigan (the catastrophe). The canonical case of an under-capitalized owner. Boyce Hydro (principal Lee Mueller) acquired the dam in 2004 with problems disclosed; FERC ordered spillway upgrades and, when they weren’t completed, revoked Edenville’s hydropower license in September 2018 after more than a decade of violations — transferring oversight to the state. On May 19, 2020, after heavy rain, the Edenville Dam failed, cascading to fail the Sanford Dam downstream; ~10,000 residents were evacuated and damage exceeded $200 million. Boyce filed Chapter 11 within weeks. A federal judge found Mueller personally liable for ~$119M in environmental damage (Nov 2023); his later personal bankruptcy was denied as bad-faith. The lesson for Bar Mills: once a license is gone and the owner is judgment-proof, liability collapses onto the public — a Michigan Court of Claims ruled the state itself not liable, leaving thousands of flood victims without a solvent defendant. License revocation/surrender is exactly the moment that exposure crystallizes. ▸ Morrow Dam, Michigan (single-purpose LLC + contaminated neighbor). Morrow Dam on the Kalamazoo River is owned by STS Hydropower LLC, a subsidiary of Eagle Creek Renewable Energy (ultimately Ontario Power Generation). FERC flagged the spillway gates for repair in 2017; the company delayed, then in October 2019 declared an “emergency” drawdown that it ran for over a year, releasing ~370,000 cubic yards of sediment down ~30 miles of river, burying habitat up to 12 feet deep and trapping people and wildlife. Michigan’s AG sued in 2022 (pleading Eagle Creek as the “alter ego” of STS — a direct attempt to pierce the single-purpose-LLC veil), and the parties settled in 2026 for $25 million — but the settlement required no river cleanup, and the company had already abandoned remediation after addressing under 1% of the sediment. Relevance to Bar Mills: a thin single-purpose LLC operating an aging structure, an owner that walked away from cleanup, and a state forced to litigate the corporate veil to reach a solvent parent — with the river left largely un-restored. (Morrow also sits near a PCB Superfund stretch, echoing Bar Mills’ adjacency to the Rogers Fibre Mill Superfund site.) Industry backdrop: a 2020 survey found ~30% of hydro dam owners are considering decommissioning rather than relicensing, with economics the primary driver. Many FERC-regulated dams are old, uneconomical and have become liabilities — so the volume of aging structures changing hands is rising. Does Brookfield itself show this pattern? No completed case of Brookfield abandoning a surrendered structure to a municipality was found. But three Brookfield-specific patterns make the Hollis concern credible rather than far-fetched: 1. It uses the exact spin-out structure now. In the 2025 sale of four lower Kennebec dams to The Nature Conservancy ($168M), Brookfield asked FERC to transfer each dam’s license to newly created Brookfield-owned holding companies as a step to facilitate the sale. Spinning dams slated for disposal into single-purpose holding entities is precisely the structure that worries Hollis. The Kennebec buyer (a well-funded conservancy forming the Kennebec River Restoration Trust) is strong — but the same structure with a weaker buyer is how a Cheboygan happens. 2. It sheds these assets explicitly to escape cost. Brookfield told Maine regulators the Kennebec sale was motivated by an “increasingly complex, expensive and uncertain regulatory landscape,” calling required fish passage “prohibitively expensive” and “arguably unattainable.” Cost-avoidance is the stated disposal motive — consistent with holding a non-productive powerhouse only until offloading is convenient. 3. It has minimized obligations on an uneconomic asset before. In New York, Brookfield suspended payments to the Hudson River–Black River Regulating District for the head the Conklingville Dam provides to its E.J. West plant — cutting that revenue ~97% — triggering demand letters and a FERC dispute. A live example of Brookfield minimizing financial obligations when a hydro asset’s economics turn. Counterweight (in fairness): in the Kennebec deal Brookfield agreed to keep maintaining all four facilities under long-term operating agreements through the multi-year decommissioning — i.e. retaining responsibility through transition rather than dumping it. And unlike a Boyce Hydro, Brookfield has a real balance sheet and significant reputational exposure in Maine (it generates ~87–90% of the state’s hydropower). Its hundreds-of-entities structure cuts both ways: it could isolate liability in a shell, but Brookfield also has assets and standing that a failed micro-operator never did. Bottom line: the feared outcome is a documented industry failure mode (Cheboygan, Boyce/Edenville, Morrow), it is enabled by the single-purpose-LLC structure Brookfield is actively using (Kennebec), and Brookfield has shown willingness to shed uneconomic hydro obligations. What is absent is a completed Brookfield abandonment-to-a-town. That argues for a preventive posture: secure transfer-surviving financial assurance in the Bar Mills surrender order now, while FERC still has jurisdiction — not after it ends. Novel stakeholder approaches in FERC surrenders (precedents to draw on) Stakeholders have developed a real toolkit for protecting public interests in surrender/decommissioning dockets. The mechanisms below are drawn from actual FERC orders and litigation, several in Maine and New England, and map directly onto the Bar Mills concern. 1. Co-licensee retention / a creditworthy substitute (the shell-company antidote). In the Klamath decommissioning, PacifiCorp sought to transfer four dams to a newly-formed non-profit (KRRC) purely to decommission them. In July 2020 FERC flagged the exact Hollis concern — that it would not be in the public interest for an under-capitalized transferee to bear all liability, since costs could fall on state or federal authorities once the former owner is off the hook — and conditioned the transfer on PacifiCorp remaining a co-licensee. PacifiCorp refused to stay on; the impasse was resolved by a November 2020 Memorandum of Agreement in which the States of California and Oregon joined KRRC as co-licensees instead, plus a $45M addition to a $50M cost-overrun contingency fund. So the final structure substituted creditworthy public co-licensees (two states) for the original owner. Application to Bar Mills: insist that any transfer of the retained powerhouse/parcel keep a financially-responsible party on the hook — ideally Brookfield or a creditworthy parent as co-obligor, or failing that a funded substitute — so liability cannot evaporate into a shell. 2. Bonds, letters of credit, escrow as financial-assurance conditions. FERC’s formal financial-assurance options are bonds; an individual trust/escrow/remediation fund; or dam-failure insurance, and a stakeholder can ask that surrender be conditioned on one of these, sized to the cost of maintaining or removing the retained structures. A useful illustration of the instruments is the Marseilles project — though note it is an under-construction 10-MW licensing case in LaSalle County, Illinois, not a surrender. FERC imposed a Financial Assurance Plan article there specifically because the project’s primary representative was also the representative of the Sanford project in Michigan (one of the two dams that failed near Midland in 2020), citing that operator’s history of insufficient funding for ordered dam-safety repairs. The article required a registered engineer’s itemized cost estimate to remove or secure facilities for public safety if the licensee could not complete or operate the project, and the licensee satisfied it with a construction-completion bond plus a letter of credit from project financiers. Two transferable lessons for Bar Mills: FERC will demand assurance when an operator’s track record warrants it, and an engineer-itemized make-safe/removal cost backed by a bond-plus-LOC is a concrete, fundable instrument. Caveat: this practice is rare and has typically applied to new construction licenses rather than surrenders, so Marseilles shows the tools exist — it is not a guarantee FERC imposes them on a surrendering licensee. The stronger surrender-specific lever for the shell-company problem is the Klamath co-licensee condition above. 3. The standard financial-assurance reopener article (a contested stopgap). After the 2020 Edenville/Sanford failures, FERC opened a rulemaking (Docket No. RM21-9, Jan 2021) on whether to require financial assurance from hydro licensees — driven by the gap that most existing licenses predate any financing requirement, and that even newer financing articles cover normal O&amp;M, not major safety or decommissioning costs. The Edenville repair estimate alone exceeded $300 million. The NOI floated three instruments: (a) bonds for safety/O&amp;M; (b) an industry-wide or individual trust / escrow / remediation fund; or (c) dam-failure insurance — sized to risk factors like dam age, condition, hazard class, reservoir size, floodplain development and climate stressors (every one of which points the wrong way for an old, non-operational, high-hazard structure beside a Superfund site). Because that rulemaking stalled (industry urged limiting any rule to small “mom-and-pop” projects), FERC began inserting a placeholder into transfer orders to preserve its options — the “Reservation of Authority to Require Financial Assurance Measures” article: “the Commission reserves the right to require future measures to ensure that the licensee maintains sufficient financial reserves to carry out the terms of the license and Commission orders pertaining thereto.” It has appeared in transfer orders including the Vermont Tissue Mill Dam (to North Bennington Hydroelectric), Hoosick Falls, Racine, Cornell, Blackstone and Aquenergy. Crucially, the article does not itself impose a bond or trust — it only preserves FERC’s ability to require one later, so approving the transfer doesn’t foreclose action. It is genuinely contested. Commissioner Danly dissented repeatedly, arguing the reopener violates the Federal Power Act by amending the license to reserve authority in a transfer proceeding; even other commissioners doubted the reopener alone is sufficient, and there is an underlying dispute about FERC’s authority to impose new requirements on existing licensees at all. No final RM21-9 rule has issued, so the article remains a working stopgap, not settled law. Takeaway for Bar Mills. The reopener is useful but weak on its own — it merely keeps FERC’s option open, a future commission could decline to use it, and the surrender timing problem still bites (jurisdiction ends at surrender). So a stakeholder should not rely on the reopener as the protection; the stronger move is to seek an actual instrument from the NOI menu — a bond, a funded trust/escrow, or insurance sized to the project’s risk — written into the surrender/decommissioning order itself, with the reopener as a backstop rather than the front line. 4. Purpose-built, funded trusts — with the obligation transferred to the funded party (the gold standard). On the Penobscot, a nonprofit (the Penobscot River Restoration Trust — the Penobscot Nation plus conservation groups) paid the owner, PPL Corporation, ~$24M to buy the Veazie, Great Works and Howland dams (money flowed to the owner, not from it), and the Trust separately raised ~$60M total (about half federal, via NOAA / the 2009 Recovery Act) to fund purchase and removal. Crucially, the FERC license and the surrender/removal obligation transferred to the funded Trust — the seller was not released into a vacuum; a capable successor was bound first. At Veazie the Trust later removed the powerhouse, built a riverside park, and offered the remediated, unencumbered parcel to the town. FERC itself drew the liability lesson: in a later surrender it cited “liability concerns that arose in Penobscot” and an inexperienced transferee as reasons to require a co-licensee structure. Lesson for Hollis: the obligation should move only to a party that already holds the funds to perform it. 5. Municipal-reliance arguments to shape the plan (double-edged). Aclara / Somersworth (Salmon Falls River, ME–NH; FERC-approved 2023, upheld by the D.C. Circuit Jan 2025) is nearly identical to Bar Mills — a non-operational plant (penstock failed 2011) whose owner surrendered rather than pay for fish passage and proposed to leave the dams in place. FERC allowed surrender without removal partly because two municipalities relied on the impoundment for water supply and firefighting. Lesson for Hollis (note the Canal Road dry hydrant): a town can invoke reliance interests to shape the outcome — but “leave in place” is exactly what creates the orphaned-structure risk, so any leave-in-place result should be coupled with a funded maintenance obligation. 6. State dam-abandonment law as backstop — and its trap. Maine DEP recently denied an owner’s petition to give up three high-hazard dams (Bucksport / AIM) as “deficient in numerous ways.” But officials warn the statute could leave residents or municipalities stuck with the dams and all O&amp;M costs — a forfeiture process never once carried to completion in 20+ years. The Bucksport filing exposes the precise trap: an owner that sold off water-supply rights and other encumbrances, then tried to hand the town a “highly encumbered” dam, which the town argued violates the requirement that a dam be offered for no consideration and in sound condition. Defensive lesson: insist the Bar Mills structure cannot be offloaded encumbered or stripped, and that abandonment-to-town is not a permissible exit without funding. The universal timing rule. A license may be surrendered only by mutual agreement, and FERC may impose conditions the licensee must fulfill before surrender is complete — but on surrender, FERC’s jurisdiction ends and it cannot compel new measures afterward (“a step for any successor agency to take”). Every mechanism above must therefore be secured before the surrender order issues. Once granted, the leverage is gone. Where the owner helped pay: a spectrum of contribution A central question for Hollis: have dam owners actually contributed money (or a bond) to protect the town, trust, or successor — rather than just selling and exiting? Yes, and the precedents span a wide range. Where on this spectrum the Bar Mills outcome lands is a matter for negotiation while the docket is open. ▸ Owner funds most of the removal (Klamath — the strongest example). PacifiCorp (a Berkshire Hathaway Energy utility) contributed $200 million toward the ~$450M four-dam removal — raised via a ratepayer surcharge collected over a decade and held in dedicated trust accounts — with California adding $250M from a water bond. PacifiCorp, California and Oregon also split a $45M cost-overrun contingency fund. Customers were protected by a hard cost cap: ratepayers bore no liability for overruns. The utility paid because removal was cheaper than the fish-passage upgrades relicensing would have required — the same economics Brookfield cites at Bar Mills and the Kennebec. ▸ A downstream beneficiary funds removal via mitigation (Edwards Dam, Kennebec, 1999). After FERC ordered the Edwards Dam removed, much of the removal cost was paid by Bath Iron Works — a downstream shipbuilder expanding into sturgeon habitat — which paid into the dam-removal settlement as environmental mitigation. A model where a third party with its own regulatory needs funds the work, sparing both the town and the dam owner. ▸ Buyer brings a capital reserve for the successor (Kennebec, 2025 — the live Brookfield deal). TNC’s $168M purchase of Brookfield’s four Kennebec dams expressly includes “a capital reserve for the new restoration trust,” with a further ~$140M to be raised to operate and remove them. Here the buyer/successor is capitalized up front — though note Brookfield is the seller receiving value, not a contributor to removal; its commitment is to keep operating the dams during the multi-year transition. ▸ Owner pays nothing; the successor/public funds it (Penobscot). At the other end, PPL was simply paid ~$24M to sell; the Trust (largely federal money) funded removal. The owner contributed assets to the deal and fish-passage costs elsewhere, but no removal funding. This is the floor — and the one closest to what an unprotected Bar Mills transfer could resemble. Takeaway for Hollis. Owner contribution is precedented and negotiable, not unheard-of. The strongest protections combine (1) an owner/parent funding contribution or surety sized to a registered engineer’s make-safe/removal estimate, (2) a hard cost cap or indemnity shielding the town from overruns, and (3) transfer of the obligation only to a pre-funded successor. The leverage to obtain these exists only while Brookfield needs FERC’s surrender approval — i.e., before the surrender order issues. Honesty caveat: this list is built from Brookfield’s project websites, the Federal Register, FERC court-case pages, and Maine news coverage — not from a complete docket-by-docket sweep of FERC eLibrary, which my tools can’t fully query. It captures the prominent, publicly reported Maine cases; there may be smaller surrenders (e.g. minor exemptions or sub-1-MW projects) elsewhere in Brookfield’s ~38-dam Maine fleet or other states that aren’t surfaced here. A definitive list would require a direct FERC eLibrary search on each Brookfield licensee entity (Brookfield White Pine Hydro, Black Bear Hydro Partners, Merimil, Hydro Kennebec, Erie Boulevard Hydropower, and others). Confirmation status Hydro identifications (Smoky Mountain) are confirmed across buyer, offtaker, lender and Brookfield-fund sources. Wind locations for 2015, 2021 and Shepherds Flat are confirmed. “Not disclosed / not identified” entries reflect genuine absence of public detail. The FERC decommissioning cases are confirmed as reported but the list is not guaranteed exhaustive (see caveat above). Sources: Brookfield Renewable / BEPC annual reports and Form 6-K interim reports 2015–2025 (SEC EDGAR; bep.brookfield.com); NextEra Energy Partners, Argo Infrastructure Partners, TVA and Brookfield Infrastructure Income Fund disclosures; trade press (Renewables Now, IPE Real Assets, Daily Energy Insider); FERC (federalregister.gov, ferc.gov, MBR/eTariff guidance, court-case pages) and the Bar Mills project site (barmills.brookfieldusprojects.com — FAQ, Draft/Final Study Plans, FERC No. 2194); comparative cases from Bridge Michigan, Detroit News, CBS Detroit, 9&amp;10 News and Interlochen/WCMU Public Radio (Cheboygan); Michigan Advance, ABC12 and Michigan AG (Boyce/Edenville); Bridge Michigan, WWMT, WMUK and Michigan AG (Morrow Dam / Eagle Creek $25M settlement); American Rivers Practitioner’s Guide to Hydropower Dam Removal, Michigan AG filings (Morrow Dam), The Nature Conservancy and Central Maine / Bangor Daily News (Kennebec sale), HRBRRD (Conklingville / E.J. West); stakeholder-mechanism precedents from National Hydropower Association (Klamath co-licensee, Marseilles bond/LOC), FERC RM21-9 financial-assurance docket (NOI, staff presentation, technical conference notice, Federal Register), National Hydropower Association and National Law Review (reopener article and Danly dissent), American Whitewater v. FERC (Aclara/Somersworth, D.C. Cir. 2025), Penobscot River Restoration Trust, Colorado Environmental Law Journal (trust-fund analysis), and Maine DEP / Press Herald (Bucksport dam-abandonment); owner-funding precedents from Oregon CUB, ASCE, KRRC and ORS 757.736 (Klamath ratepayer fund / cost cap), The Revelator and NRCM (Edwards Dam / Bath Iron Works mitigation), and Maine Public / TNC (Kennebec capital reserve); plus Maine municipal and news sources (Town of Hollis, Town of Buxton, Ellsworth American, Mainebiz, Portland Press Herald, NewsCenter Maine). Informational summary, not legal or investment advice.

• Brookfield_Bar_Mills_Discussion_Thread (1).docxOpen original

  Read from the file

  Brookfield Renewable / Bar Mills Research Discussion Thread — Questions &amp; Findings A chronological record of the inquiry, from asset sales through FERC surrender strategy. Compiled June 2026. Companion to the data deliverables (the asset-sales summary and the US-disposals ledger). Informational summary, not legal or investment advice. This record traces how the inquiry developed: it began as a question about Brookfield Renewable’s asset sales and narrowed, step by step, to a specific concern about the Bar Mills dam in Hollis, Maine, and the legal tools available to protect the town. Each section pairs the question asked with what the research found. 1. Brookfield Renewable’s asset sales (2024–25) Q: List and describe the asset sales Brookfield Renewable has done over the past two years. Brookfield runs a capital-recycling (“asset rotation”) strategy: it sells mature, de-risked operating assets to lower-cost-of-capital buyers and redeploys proceeds into development and acquisitions. Reported recycling proceeds were a record $2.8B in 2024 (&gt;$1B net to BEP) and a further record $4.5B in 2025 ($1.3B net). Headline deals: Saeta Yield (Spain/Portugal, ~$1.4B, to Masdar), First Hydro (UK pumped storage, 25%, $350M), Shepherds Flat (Oregon wind), an India wind/solar portfolio (to Gentari); then in 2025 Luminace (North American distributed generation), a “non-core” US hydro portfolio, a 700 MW US DG portfolio, ~$1B of Neoen asset rotation, and a ~2,300 MW US wind &amp; solar portfolio closing into 2026. 2. How Brookfield handles “excess / unwanted” property Q: Is there evidence of how Brookfield handles excess, unwanted properties? The framing didn’t match the evidence. Brookfield’s disposals are deliberate capital recycling — it sells assets when they are in demand and attracting strong valuations, not when they have become problems. Track record since 2020: ~$6B in proceeds at ~22% average IRR and ~2.1x invested capital — returns inconsistent with offloading unwanted assets. No distinct “distressed property” disposal process was found at the renewable entity. The one place language like that appears is at the parent (Brookfield Corporation) regarding troubled office real estate — a different business. 3. End-of-life / written-down / problem assets Q: Dig into sales of end-of-life assets, assets written down in value, and projects with permitting or physical problems. The granular disposal notes surface a few non-core cases (e.g., an 85 MW Brazil biomass portfolio sold at a loss; Brazil hydro/biomass in held-for-sale). But two structural points cut against the broader thesis: most small disposals were gains, not write-downs; and Brookfield reports under IFRS using the revaluation model, so a deteriorating asset is typically marked down through annual fair-value changes in equity/OCI before any sale — meaning distressed assets rarely surface as loss-making disposals. 4. US focus — orphaned / non-productive assets Q: Focus on the sale of orphaned or non-productive assets, especially in the US (gains or losses). The only US assets Brookfield itself labels “non-core” are hydro plants. Standalone US disposals included a 30 MW hydro plant (2024, sold above book at a $29M gain), a 60 MW battery (2024), and a ~400 MW “non-core” US hydro portfolio (2025). From 2023 onward several US sales routed a portion to Brookfield-affiliated (BAM-managed) funds — the mechanism by which assets leave BEP’s books while staying in the Brookfield group. No US disposal was attributed to end-of-life, permitting, or physical failure; none was booked at a loss. 5. A full US disposal ledger, 2015–2025 Q: Build a US-only ledger across every annual report since 2015, read thoroughly. An eleven-year, US-only ledger was compiled from each year’s disposal and held-for-sale notes (delivered as a separate landscape document). Wind disposals cluster on the West Coast / mountain-West (California, Oregon) with a New Hampshire outlier; 2016–19 and 2022 had no standalone US disposal. The hydro rows were highlighted at your request. A recurring finding: Brookfield discloses location sparingly — most precise locations came from buyers or trade press, not Brookfield’s filings. 6. Identifying the sold hydro assets Q: How can we determine the locations of the sold hydro assets — and can you identify them? By triangulating the buyer, offtaker, lenders, and a sister Brookfield fund’s holdings table, both large hydro portfolios resolved to the same asset: Smoky Mountain (formerly Tapoco) — four dams (Cheoah, Santeetlah, Calderwood, Chilhowee) on the Little Tennessee and Cheoah Rivers across east Tennessee and western North Carolina, FERC Project No. 2169. Brookfield bought it from Alcoa for $600M in 2012, sold 50% to Argo Infrastructure Partners in 2023, and sold the remaining 50% across 2025–26. “Non-core” meant portfolio fit, not poor quality — Smoky has a 10-year TVA offtake and a $435M financing. 7. FERC license non-renewals &amp; decommissioning Q: Find all licenses not renewed or plants decommissioned by a Brookfield entity at FERC — including non-hydro surrenders. These cluster in Maine and are relicensing conflicts. Clearest cases: Bar Mills (FERC No. 2194, Saco River, Hollis — surrender filed 2020; units inoperable for years); the Ellsworth Project (Black Bear Hydro; license expired 2017, running on annual temporary licenses, water-quality certification denied); and the lower Kennebec dams (Shawmut, Lockwood, Hydro-Kennebec, Weston — contested relicensing/removal). On the non-hydro question: FERC license surrender applies only to hydropower, while wind/solar/storage entities hold cancellable market-based-rate and QF authorizations — a section later removed from the memo at your request. 8. The Bar Mills retained-powerhouse risk Q: As part of the Bar Mills surrender, BWPH plans to retain a defunct powerhouse. The risk is they later abandon or sell that surplus property to a buyer that cannot maintain or remove it, and it falls to the small town of Hollis, which cannot absorb the expense. The record supports the concern. BWPH will retain the powerhouse (1956 structure) and leave part of the spillway and the eastern dam in place. The decisive mechanism, in BWPH’s own words: “once the surrender application is approved by FERC, there is no longer a FERC project boundary.” On surrender, FERC’s safety and financial jurisdiction ends entirely — no federal bond, no maintenance trust, no obligation running with the land to a future owner. BWPH’s maintenance duty binds BWPH under the approved plan, not clearly a perpetual covenant against a downstream buyer. The retained structures sit next to the Rogers Fibre Mill Superfund site. Exposure is split: the powerhouse is on the Hollis side; the eastern dam and Superfund parcel are owned by the Town of Buxton. 9. Is the Hollis fear precedented? Q: Are there examples of this feared outcome — by Brookfield or other large operators? (with detail on Cheboygan, Boyce/Edenville, Morrow) Yes — a documented failure mode. Cheboygan, MI is the closest analog: a split complex (state owns the 1922 dam, a thin LLC — Hom Paper XI — owns the powerhouse), offline since a 2023 fire, whose owner called operating it “no longer economically feasible,” won repeated FERC extensions while trying to offload it to a small Wyoming firm (HydroMine), and came within ~6 inches of overtopping in April 2026 — saved only by an emergency multi-agency turbine restart. Boyce / Edenville is the catastrophe: FERC revoked the license in 2018 after a decade of violations; the dam failed in May 2020, cascaded into Sanford Dam, evacuated ~10,000 people, caused $200M+ damage; the owner went bankrupt and was found personally liable for ~$119M but is judgment-proof — and a court held the state not liable, leaving victims with no solvent defendant. Morrow Dam shows the single-purpose-LLC-plus-Superfund profile: STS Hydropower (a subsidiary of Eagle Creek / Ontario Power Generation) ran a year-long “emergency” drawdown that buried ~30 miles of the Kalamazoo River in ~370,000 cubic yards of sediment; the state had to plead the parent as STS’s “alter ego” to pierce the LLC veil, settling in 2026 for $25M — with no required cleanup. No completed Brookfield abandonment-to-a-town was found. But Brookfield uses the same single-purpose-holding-company structure now (Kennebec), states cost-avoidance as its disposal motive, and has minimized obligations on an uneconomic asset before (suspending ~97% of E.J. West / Conklingville payments). These are what-the-structure-permits precedents — not Brookfield conduct. 10. Novel stakeholder approaches in FERC surrenders Q: What novel approaches have stakeholders used to protect their interests in FERC surrenders? (verified against primary sources) A real toolkit. (1) Co-licensee retention or a creditworthy substitute — at Klamath, FERC required the owner to stay on; when PacifiCorp refused, the States of California and Oregon joined as co-licensees instead. (2) Bonds / letters of credit / escrow sized to a registered engineer’s make-safe estimate — illustrated by the Marseilles construction license (an Illinois under-construction case, not a surrender; FERC imposed it because the representative was tied to the failed Sanford project). (3) The financial-assurance reopener article — see below. (4) Purpose-built funded trusts (Penobscot). (5) Municipal-reliance arguments to shape the plan (Aclara/Somersworth, upheld by the D.C. Circuit in 2025). (6) State dam-abandonment law as backstop and trap (Bucksport, where an owner tried to hand a town a “highly encumbered” dam). On the reopener (mechanism 3), in detail. After the 2020 failures, FERC opened a rulemaking (RM21-9) on requiring financial assurance, floating three instruments — bonds, a trust/escrow/remediation fund, or dam-failure insurance — sized to risk factors (dam age, hazard class, floodplain, climate). The rulemaking stalled, so FERC began inserting a placeholder into transfer orders: the “Reservation of Authority to Require Financial Assurance Measures” article, reserving its right to require such measures later. But the article imposes nothing on its own, Commissioner Danly calls it an FPA violation, no final rule has issued, and surrender ends FERC’s jurisdiction. So the reopener is a weak backstop — the stronger move is to write an actual instrument (bond, funded trust/escrow, or insurance) into the surrender/decommissioning order itself. The universal rule: secure all of it before the surrender order issues, because FERC’s jurisdiction then ends and it cannot compel new measures afterward. 11. PPL, the Trust, and the flow of money Q: Who is PPL — is it related to Brookfield? Who was the Trust? Was the surrendering corporation relieved of obligation? Did PPL give money to the Trust? PPL Corporation is a Pennsylvania-based investor-owned utility, unaffiliated with Brookfield. The Penobscot River Restoration Trust is a nonprofit (Penobscot Nation + conservation groups) formed to buy and remove the dams. The surrendering owner was relieved of obligation only because the license and removal duty transferred to the funded Trust — not abandoned. On the money: it flowed to PPL, not from it — the Trust paid PPL ~$24M to buy the dams and raised ~$60M total (about half federal) for purchase and removal. (Note: PPL also sold its other Maine hydro to Black Bear Hydro / ArcLight, the portfolio Brookfield later came to own — the only thread connecting these names to Brookfield.) 12. Did the owner ever help fund removal or post a bond? Q: Are there examples where the corporate owner helped fund the removal, or bonded, to protect the town / trust / other party? Yes — a spectrum. Owner funds most of removal: at Klamath, PacifiCorp contributed $200M (via a ratepayer surcharge in dedicated trust accounts) toward a ~$450M removal, with a hard cost cap protecting customers from overruns. A downstream beneficiary funds it: at Edwards Dam (1999), Bath Iron Works paid much of the cost as environmental mitigation. Buyer brings a capital reserve: the 2025 Kennebec deal includes a capital reserve for the new trust (though Brookfield is the seller receiving value, committing to operate during transition). Owner pays nothing: Penobscot — the floor. Takeaway: owner contribution is precedented and negotiable. Strongest protections combine an owner/parent contribution or surety sized to an engineer’s removal estimate, a hard cost cap or indemnity, and transfer only to a pre-funded successor. One caveat: Klamath’s money came from ratepayers via a regulated utility — a mechanism Brookfield (a merchant generator) lacks, so any Brookfield contribution would more likely take the form of a direct corporate/parent contribution or surety bond than a ratepayer fund. 13. The Kennebec cost-avoidance, in detail Q: Provide more information on the Kennebec cost avoidance. What Brookfield faced. The Shawmut Dam (FERC No. 2322) entered relicensing in 2020 when its 40-year license expired (Jan 31, 2021). A new 40-year license required a Maine DEP Section 401 water-quality certification — which became the chokepoint. The direct hardware cost was modest: FERC estimated upgrading Shawmut’s fish passage at ~$6.3M, of which ~$4.8M could come from Bipartisan Infrastructure Law funding. The real exposure was far larger than that number. Why the true cost dwarfed $6.3M. Three things stacked up. (1) Standards, not just hardware: Maine DMR wanted a nature-like fishway Brookfield called cost-prohibitive; Brookfield wanted a fish lift. DEP adopted fish-passage performance standards Brookfield deemed “unattainable” — meaning it could spend the money and still fail certification. (2) Multi-dam, fleet-wide precedent: the proceedings covered all four lower Kennebec dams; Brookfield warned that the same requirements imposed elsewhere could “threaten the continued operation” of its other dams. The math drove the logic — the four dams produce only ~47 MW of Brookfield’s 615 MW in Maine (~0.43% of state electricity), not worth an open-ended fight that could set precedent across a fleet supplying ~90% of Maine’s hydropower. (3) Litigation and delay: Brookfield sued the DMR and DEP, lost at the Somerset County Superior Court (Jan 2024) and the D.C. Circuit (July 2024), and defended an Endangered Species Act suit — years of spend with no certification to show. How the sale avoids all of it. When DEP issued a draft approval still requiring extensive fish-passage improvements (Oct 2025), Brookfield withdrew its water-quality applications for all four dams, calling the conditions too burdensome and citing the “complicated, expensive licensing process” as a major influence on both the withdrawal and the sale. Its logic: because TNC plans to remove the dams within a decade, fish-passage upgrades were “moot.” So instead of paying for fishways it might never get credit for, Brookfield exits — and the far larger removal cost (part of the ~$140M TNC must still raise to operate and remove the dams) passes to TNC and the new trust. Brookfield’s framing vs. the state’s. In its Oct 16, 2025 letter, Brookfield said the standards “conflicted with state law, would be prohibitively expensive… and were ‘arguably unattainable,’” and that a motivating factor for the sale was “the increasingly complex, expensive and uncertain regulatory landscape.” Maine DEP said it was “taken aback” and “strongly disagrees,” noting the standards incorporated federally recommended endangered-species requirements. Three layers of cost avoided — and the Bar Mills through-line. It helps to separate (a) the modest direct fishway cost (~$6.3M at Shawmut), (b) the large avoided removal cost Brookfield never wanted to bear (now TNC’s), and (c) the avoided open-ended regulatory and fleet-precedent risk — likely the real driver, given the 47-of-615 MW math. The same pattern underlies Bar Mills: a low-value asset facing high regulatory/decommissioning cost, exited by Brookfield — the open question being who holds the expensive end. At the Kennebec it is a well-funded conservancy; at Bar Mills the concern is that it could be the town of Hollis. Deliverables produced in this thread • Asset-sales summary (Brookfield Renewable, 2024–25). • US asset-disposals ledger with locations, 2015–2025 (landscape; hydro highlighted; Smoky Mountain identified; includes the FERC decommissioning cases, the Bar Mills risk analysis, comparable-case precedents, the novel-approaches toolkit, and the owner-funding spectrum). • This discussion-thread record. Open thread, if useful: the current docket status of the Bar Mills surrender (FERC P-2194) — whether the Surrender Application has actually been filed yet — determines how much of the comment window remains open for seeking the financial-assurance conditions described above. Kennebec cost-avoidance detail draws on Press Herald, Bangor Daily News, Maine Public, Central Maine, The Maine Monitor, NewsCenter Maine, and Maine DEP records (Shawmut WQC, FERC No. 2322). Sources for all other findings are listed in full in the ledger document. Informational summary, not legal or investment advice.