📺 Decarbonize Deep Dive 012 | ~22 min

DEEP DIVE: Big Tech Builds Its Own Power Company — The Co-Location Playbook

On June 5, 2026, Google and Intersect Power broke ground on the Meitner Energy Center in the Texas Panhandle: a data center co-located with more than 1 GW of new wind, solar, and battery storage, engineered so the servers and the generation switch on together — without leaning on the grid for the baseline load. Three months earlier, Alphabet had closed a $4.75 billion acquisition. Not of more solar panels — of Intersect Power itself, the developer that builds the clean energy. A technology company bought a power company.

This is the corporate-structure sequel to DD008's interconnection-queue crisis. When the grid can't connect you for five-to-seven years and your GPUs can't wait eighteen months, the rational move is to build your own power plant next to your own data center — and buy the company that builds it. The largest electricity consumers in America are becoming some of its largest generators and developers. After thirty years of unbundling, the AI build-out is re-verticalizing the power industry.

But the model leaves a real fight unresolved. These campuses are still grid-connected, and variable renewables plus storage are not the same as firm 24/7 power — so the grid (or on-site gas) still backstops them. That has opened a regulatory battle, templated by the 2025 FERC fight over a co-located Amazon–Talen nuclear deal in PJM, over whether co-located mega-loads are free-riding on infrastructure everyone else pays for.

Key topics:

• The Meitner Energy Center: more than 1 GW of co-located wind, solar, and storage next to a Google data center, energized together

• Why Alphabet bought Intersect Power for $4.75B — the difference between buying power and buying a power company

• The interconnection bottleneck: ~2.6 TW backlogged, 5–7 year waits, and why AI's 18-month clock can't tolerate it

• Co-location vs. the PPA: how putting electrons and servers behind one point of interconnection changes everything

• The economics: time-value of power, internalized developer margin, near-zero marginal cost, 24-7 carbon-free attributes

• The firmness gap: why these campuses still need grid or gas backup — and what that means for "clean"

• The regulatory fight: cost-shifting, the FERC/PJM co-location rulings, and who pays for the backup grid

• The pattern across Microsoft, Amazon, Meta, and Stargate — and why Google's is the cleanest example

• What it means for utilities: adapt as backup/transmission partners, or cede the fastest-growing load on the system

The question for the rest of this decade isn't whether AI gets powered. It's who ends up owning the power — and the answer, increasingly, is the technology companies themselves.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#aidatacenters #cleanenergy #google #intersectpower #datacenterpower #energytransition #behindthemeter #colocation #gridinterconnection #renewableenergy #batterystorage #hyperscaler #powergrid #ercot #decarbonization #energypolicy

📺 Decarbonize Deep Dive 011 | ~22 min

DEEP DIVE: The One Big Beautiful Bill's Accidental Clean Energy Boom — and the Cliff Below It

The One Big Beautiful Bill Act was designed to end the IRA era. Instead, it created the largest clean energy construction sprint in American history. Between 216 and 240 GWdc of solar capacity is being legally locked into IRA tax credit eligibility before the July 4, 2026 deadline — a volume equal to projected US solar installations through the end of the decade, compressed into a single 18-month race.

Here is the paradox: an anti-renewable bill set a construction deadline, and that deadline created a mandatory forcing function that no market signal could replicate. Developers are pouring concrete at unprecedented speed. A record 86 GW of new US generating capacity is expected in 2026. The Georgetown Environmental Law Review called it potentially "the most significant short-term expansion of clean energy infrastructure in US history." The bill that was supposed to kill renewables accidentally supercharged them.

But the cliff is as real as the boom. The 10-year US solar output projection under the OBBBA is 17% below the pre-OBBBA baseline. Battery storage retains ITC eligibility through 2033 — giving storage a 7-year advantage that solar and wind do not have. The IRS eliminated the 5% safe harbor rule in January 2026, requiring actual physical construction — not procurement. Transformer lead times of 2–4 years are locking out late entrants. And $222 billion in outstanding announced clean energy investment is racing against a calendar it was never designed for.

Key topics:

• What the OBBBA actually did to clean energy tax credits — the timeline, the technology asymmetries, and the FEOC restrictions

• The July 4, 2026 construction deadline: what "begin construction" now legally means after the IRS eliminated the 5% safe harbor

• Why 216–240 GWdc of solar is being safe-harbored — and how that compares to a full decade of projected US solar installations

• Battery storage's separate runway: why storage gets ITC through 2033 while solar and wind face the cliff

• The four bottlenecks killing late projects: transformers, interconnection queue, FEOC compliance, and labor competition

• The project finance anatomy: how the ITC works as a tax equity financing mechanism, not a subsidy

• The 17% long-term solar gap: what the 10-year US solar forecast actually shows under OBBBA

• State-level responses: which state renewable portfolio standards are positioned to absorb what federal credits cannot sustain

• Where battery storage goes from here: record 24 GW of US installations in 2026, AI data center demand, and the LFP dominance story

• Three things to watch by end of 2026: actual construction starts vs. announced intent, IRS enforcement posture, and state policy responses

This is a portrait of an industry in a timed sprint — aware of the cliff ahead, racing to get as much in the ground as possible before the deadline designed to stop it.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#CleanEnergy #IRA #OBBBA #SolarEnergy #WindEnergy #BatteryStorage #EnergyTransition #ClimatePolicy #InvestmentTaxCredit #InflationReductionAct #OneBigBeautifulBill #USEnergy #FEOC #TaxEquity #RenewableEnergy #Decarbonization #SolarFarm #GridStorage #EnergyPolicy

📺 Decarbonize Deep Dive 010 | ~22 min

DEEP DIVE: The 2026 Green Steel Inflection — CBAM, Boden, and the Economics That Finally Work

On January 1, 2026, the EU's Carbon Border Adjustment Mechanism entered full enforcement. On the same continent, Stegra's Boden plant in northern Sweden is entering commissioning — the world's first full-commercial-scale DRI facility running on green hydrogen, backed by €3.5 billion in capital and offtake contracts from Volkswagen, Mercedes, and BMW.

These two events — CBAM and Boden — are the first time that green steel has had both a proven technology at commercial scale and a market mechanism that changes the competitive economics. For twenty years, the decarbonization pathway for steel was clear in theory and unbuilt in practice. 2026 is when that changes.

Steel is responsible for 7–9% of global CO₂ — roughly 2.3 billion tonnes per year. The blast furnace has been the dominant technology for 150 years. The chemistry requires a reductant, and that reductant has always been coal. Replacing it with green hydrogen — and the liquid steel with a near-zero-carbon product — is what Boden is now proving at commercial scale.

Key topics:

• Why steelmaking is structurally hard to decarbonize — the chemistry problem, not the combustion problem

• How DRI-EAF (Direct Reduced Iron — Electric Arc Furnace) replaces the blast furnace, step by step

• What Stegra Boden actually is — scale, H₂ volumes, renewable power supply, steel grades, and what it answers

• CBAM full enforcement: what €70–140/tonne on imported high-carbon steel does to the competitive math

• The green hydrogen cost equation — 55 kg H₂/tonne DRI, and why €2–3/kg is the break-even threshold

• Why automaker offtake contracts at €100–300/t premium are what makes Stegra financeable — buyer pull, not spot competition

• The blast furnace majority — why 70% of global steel is still BF-BOF and what it takes to transition China's fleet

• Boston Metal's MOE as the alternative pathway — direct iron oxide electrolysis, no H₂ supply chain required

• The scale reality: optimistic 2030 scenario is 50–80 Mt of green steel — still less than 4% of global production

• Three things to watch through end of 2026: Boden ramp-up yields, CBAM first full-year certificate submissions, EU ETS Investment Booster allocations

The Boden commissioning is the event. The CBAM enforcement is the market mechanism. 2026 is when the answer to "can green steel actually work?" stops being theoretical.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#GreenSteel #CBAM #DRI #EAF #Stegra #GreenHydrogen #SteelDecarbonization #IndustrialDecarbonization #EUClimate #CarbonBorderAdjustment #Decarbonization #ClimatePolicy #HydrogenEconomy #BotonMetal #MOE #BlastFurnace

📺 Decarbonize Deep Dive 009 | ~22 min

DEEP DIVE: The Largest Oil Shock in History — and the First With a Real Off-Ramp

The Strait of Hormuz has been closed since 28 February 2026. Eleven weeks in, the IEA calls it the largest oil supply disruption in the history of the global oil market — more than 10 million barrels a day removed, roughly twice the size of the 1973 and 1979 shocks. Brent peaked near $126 in April and sits around $100 in mid-May, up about 58% year-to-date.

Most coverage asks one question: how high do prices go? Decarbonize Weekly asks the one that matters more — does an oil shock this size accelerate the energy transition, or stall it?

History says oil shocks accelerate structural change. What makes 2026 different is that, for the first time, the alternatives — EVs, heat pumps, grid storage, electrified freight, sustainable fuels — are mature enough to absorb the demand a shock displaces. But maturity is not destiny: the same shock that pulls demand toward electricity also raises the cost of the capital needed to build the electric supply. Which force wins is a policy choice, not a market outcome.

Key topics:

• The 1970s template — how the 1973 and 1979 shocks reshaped energy policy, and why the structural responses (efficiency standards, building codes, the strategic petroleum reserve) outlasted the price spike by decades

• Demand destruction vs. demand substitution — why 2026 is the first oil shock where switching fuels, not just using less, is a large and available channel

• The accelerant in the data — European BEV sales +29.4% in Q1 2026, EVs roughly 1 in 4 new cars sold globally, solar leading global energy demand growth for the first time

• The other half of the story — why the same shock starves the transition of capital: higher-for-longer interest rates, fossil supply as the highest-return trade, and the affordability ceiling on transition politics

• The tie-breaker — why locking in demand-side structural change while the price signal is doing the persuading is what made 1973 matter for decades

• The aviation exception — SAF at 0.8% of jet fuel, no European e-SAF project at final investment decision, and why the shock makes the gap more visible without making it smaller

• The transatlantic split — a tightening EU with the machinery to convert a price signal into policy vs. a deregulating US relying on the price signal alone

• The realistic 2026 outlook — base, upside and downside cases, and the four signals that will tell you which one is unfolding

The technology question is settled. Whether 2026 becomes the transition's inflection point or a missed decade turns entirely on capital allocation — and that is a choice being made right now.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#OilShock #EnergyTransition #StraitOfHormuz #OilCrisis #BrentCrude #Electrification #EVs #SAF #EnergySecurity #1973OilCrisis #RenewableEnergy #CleanEnergy #Decarbonization #EnergyPolicy #ClimateTech #DecarbonizeWeekly

📺 Decarbonize Deep Dive 008 | ~22 min

DEEP DIVE: The Gas Verdict — PJM Just Decided How AI Will Be Powered

On April 29, 2026, PJM Interconnection — the largest grid operator in North America — announced the results of the first cycle of its reformed interconnection queue. After four years of effective closure, 811 generation projects representing 220 gigawatts entered the queue. The composition is the story:

• 106 GW natural gas (48%)

• 67 GW battery storage (30%)

• 18 GW nuclear (8%)

• 15 GW stand-alone solar (7%)

• 9 GW solar-storage hybrid (4%)

• 5 GW wind (2%)

Combined with the parallel Texas behind-the-meter buildout — 58 GW of gas in planning, half of it dedicated to data centers on private grids that bypass the public grid entirely — this is the verdict on how American AI infrastructure will be powered through 2032. Not with geothermal. Not with small modular reactors. Not with renewables-plus-storage. With natural gas.

Key topics:

• The PJM Cycle 1 numbers and what they mean — first reformed queue with first-ready, first-served filtering

• Why gas wins on the five variables: speed of build, capacity factor, capital intensity, site control, federal regulatory environment

• Texas behind-the-meter — Pacifico's 7.7 GW GW Ranch, the Lancium / Crusoe / Stargate clusters, and the private-grid pattern

• Why the 67 GW of storage in the queue is mostly a complement to gas, not a replacement

• Why the 18 GW of nuclear is smaller than it sounds — uprates, restarts, aspirational SMR pipeline

• The state climate plan implications for Maryland, Virginia, New Jersey, Illinois — RPS targets calibrated against assumptions Cycle 1 has just broken

• The federal regulatory dismantling that locked the gas verdict — 2025 budget reconciliation, EPA 111(b) rollback, 45V cancellation, PTC/ITC acceleration removed

• The transatlantic split — EU Hydrogen Bank's €0.44/kg auction announced May 7 vs. US 45V cancellation

• Realistic 2026-2032 buildout projection — gas at 80-110 GW, storage 50-70 GW, solar 30-45 GW, then a sharp drop

• What this means for 2030 NDC math — roughly 100 Mt/yr of CO2 added to the power sector from AI load alone

• Three things to watch through end of 2026: PJM Cycle 2 composition, Texas BTM regulation, Maryland PSC response

The clean firm power story (geothermal, SMR, nuclear restarts) is real but second-tier in the 2026-2032 buildout window. AI training capacity needs to be online before any of those scale. PJM Cycle 1 is the math made visible.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#AIDataCenters #PJM #InterconnectionQueue #DataCenterPower #NaturalGas #BehindTheMeter #Decarbonization #CleanEnergy #USPowerGrid #GridReliability #ClimatePolicy #IRA #45V #StateRPS #ClimateTech #DecarbonizeWeekly

📺 Decarbonize Deep Dive 007 | ~22 min

DEEP DIVE: The Quiet Reset That's Defining the Real Hydrogen Economy

Between mid-2024 and April 2026, the green hydrogen industry experienced its first real correction. Companies cancelled close to 60 major projects in 2025 alone — more than 4.9 million tonnes per year of would-have-been capacity. BP exited Australia, Oman, and the UK. Air Products halted three US projects. ArcelorMittal walked away despite €1.3B in committed subsidies. Lhyfe suspended a >100 MW project in April 2026.

But this is not the death of clean hydrogen. It's the death of speculative clean hydrogen. While the export mega-projects collapsed, FIDs in 2025 actually grew ~20% — on a different shape of project.

Key topics covered:

• The 60-project cancellation wave: who died and why

• The 4 economic gates every hydrogen project must clear (most failed at "bankable offtake")

• Why the cost curve disappointed: stack cost-down, balance-of-plant inflation, capacity factor revisions

• The survivor pattern — Stegra (€1.4B closed April 2026), Plug Power's Quebec deal, Topsoe SOEC, Sunfire alkaline

• The EU Hydrogen Mechanism (April 30, 2026): what it is and why it might actually work

• Hy24 quadrupling its Enagás Renovable stake — the PE survivor signal

• Why captive industrial use cases (steel, ammonia, methanol, refining) quietly worked while transport and heat didn't

• 45V regulatory turbulence, additionality, time-matching: the policy story that broke US economics

• Realistic 2030 outlook: 25-40 GW operational vs the 150 GW announced pipeline

• 5 things to watch in 2026-2027

The hydrogen industry that exists in 2030 will be smaller than the 2022 forecasts suggested but materially more bankable. It is anchored to captive industrial demand, modular sizing, sovereign-backed offtake, and customers who absorb the green premium through their own product price. That is a real, durable, financeable industry — and it's the one that will actually decarbonize hard-to-abate sectors in time to matter.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#GreenHydrogen #HydrogenEconomy #ProjectCancellations #BankableOfftake #Stegra #PlugPower #EUHydrogenMechanism #45V #ClimateTech #IndustrialDecarbonization #Decarbonization #DecarbonizeWeekly

📺 Decarbonize Deep Dive 006 | ~22 min

DEEP DIVE: The Unexpected Winner in the 24/7 Clean Power Race

Enhanced geothermal has crossed from science project to commercial reality in 2024–2026 — and the buyers driving the transition are not utilities chasing climate mandates. They're hyperscale AI data centers who need firm, carbon-free, always-on power and cannot wait for new nuclear.

Key topics covered:

• How horizontal drilling and completion tech from shale oil and gas unlocked enhanced geothermal (EGS)

• Fervo Energy's Cape Station in Utah — 400 MW, the first multi-hundred-megawatt EGS project ever built

• The Google–Fervo multi-gigawatt framework PPA and what it signals

• Meta's 150 MW pumped geothermal storage deal with Sage Geosystems

• Eavor's closed-loop approach — geothermal without fracturing

• Why 90% capacity factor matters more than cheap solar for AI workloads

• The 70% drilling cost reduction Fervo has already delivered

• Texas, Utah, Nevada — where geothermal actually gets built, and why the oilfield workforce is the secret weapon

• Common misconceptions: seismicity, geography, cost, nuclear comparison

• Realistic outlook through 2030 — projects, competition, and the SMR question

The commercial window for EGS is 2026–2028. The buyers are already signed.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#Geothermal #EGS #AIDataCenters #FirmCleanPower #FervoEnergy #24x7CarbonFree #OilAndGasTransition #Decarbonization #DecarbonizeWeekly

📺 Decarbonize Deep Dive DD005

The IMO Net-Zero Framework Is the First Global Carbon Price on an Entire Industry — and LNG May Fail It

The IMO's Net-Zero Framework is the first globally binding carbon price on an entire industry. Its Well-to-Wake accounting methodology reveals a problem the shipping industry didn't fully price in: LNG — ordered by the thousands as shipping's "bridge fuel" — may be noncompliant by 2030 under real-world methane slip data.

In this deep dive, we break down how the GFI compliance mechanism works, why Well-to-Wake accounting changes everything for LNG, which ships face stranded asset risk, and where green methanol and ammonia now sit in the race.

MEPC 84 (April 27 – May 1, 2026) will be a critical signal for whether the 2028 compliance timeline holds.

📺 Decarbonize Deep Dive 004 | ~18-22 min (evergreen)

DEEP DIVE: AI Is Stress-Testing the Entire Architecture of Power-Sector Decarbonization

For a decade, the corporate clean-power story sounded straightforward — sign PPAs for wind and solar, buy renewable energy certificates, and claim operations on clean electricity. AI data centers are exposing how thin that framework actually was. Hyperscale AI loads are big, demand round-the-clock reliability, and need power on a near-term timeline that the cleanest options (transmission expansion, long-duration storage, SMRs) cannot meet. The result: a structural revaluation of what "clean power" actually means in a market that now rewards firm, local, deliverable electrons over annual accounting.

Key topics covered:

• Why hyperscale AI loads (hundreds of MW per campus, gigawatt clusters) force utilities to rewrite resource planning

• The annual REC vs. 24/7 hourly matching gap — and why corporate climate credibility is shifting from "spreadsheet logic" to "physical electrons"

• The real value and limit of battery storage — improves the problem, but does not eliminate firm-supply needs

• Why natural gas is staging an awkward but rational comeback — dispatchable, lender-familiar, fast to deploy

• Why Microsoft, Google, and Amazon are backing SMR developers — first credible demand anchor for first-of-a-kind nuclear

• The unglamorous bottlenecks that actually decide projects: interconnection queues, transmission expansion, substations, transformers

• Water as a permitting risk — no longer an environmentalist concern, now a capital-market resource risk

• The labor constraint: data center construction, gas plants, transmission, and nuclear all draw from overlapping skilled-worker pools

• How utility integrated resource plans are being forced to absorb large, concentrated, uncertain new demand

• The new ESG bar: hourly matching + local supply + physical reduction — not just annual REC logic

• Near-term winners: gas turbine OEMs, flexible-gas developers, grid equipment suppliers, hyperscaler-anchored SMR developers, premium-reliability storage, water-efficient cooling

• Under pressure: hyperscalers with weak physical-power strategy, utilities trying to socialize grid-upgrade costs, renewable developers stuck in annual-REC framing, regions without transmission/water/permitting capacity

The bottom line: AI is not just increasing electricity demand. It is forcing the power system to reveal what it truly values — and that's a much tougher game than building the cheapest green megawatt-hour.

---

🔗 Website: decarbonizeweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#AIDataCenters #FirmCleanPower #24x7CarbonFree #SMR #SmallModularReactors #NaturalGas #HyperscaleAI #GridInfrastructure #InterconnectionQueue #Decarbonization #EnergyTransition #DecarbonizeWeekly

📺 Decarbonize Weekly Episode 4

In this episode:

1. Europe puts a real price on CBAM carbon costs

2. China launches a large integrated green fuels project

3. Shanghai Electric proves biomethanol bunkering at scale

---

🔗 Website: decarbonizedweekly.com

📧 Contact: [email protected]

🎧 Also on Spotify: search 'Decarbonize Weekly'

#Decarbonization #CleanEnergy #EnergyTransition #NetZero