Grid Capacity Becomes the Conversion Bottleneck

2026-07-12 · weekly · Planetary Balance Sheet

The week in one paragraph

The first balance-sheet baseline is not a story of civilization running out of a single input. It is a story of conversion capacity tightening: electricity demand is accelerating, but grids, interconnection queues, transformers, local permitting, water access, refining capacity, and risk finance are slower-moving physical systems. The strongest signal is electricity deliverability. The IEA forecasts global electricity demand growth of 3.6% per year from 2026 to 2030, with electricity consumption growing at least 2.5 times as fast as total energy demand. That turns grid capacity into the near-term balance-sheet line item to watch.

Constraint ledger: tightened, eased, still binding

Tightened — electricity delivery and grid queues. More than 2,500 GW of renewables, storage, and large-load projects remain stalled in grid connection queues worldwide, according to the IEA. The same report estimates that meeting electricity demand through 2030 requires annual grid investment to rise roughly 50% from today’s USD 400 billion level. Generation capacity alone is therefore an incomplete metric. The binding system is transmission, distribution, interconnection, flexibility, regulatory process, and grid-equipment supply chains.

Tightened — AI/data-centre load. Data centres are no longer a rounding error in local infrastructure planning. The IEA’s AI-energy update says global data-centre electricity demand grew 17% in 2025, while AI-focused data-centre consumption grew 50%. Its central projection has data-centre electricity use roughly doubling from 485 TWh in 2025 to 950 TWh in 2030. The physical constraint is not only megawatt-hours; it is suitable sites with power, water, chips, capital, grid connections, and community acceptance.

Eased in appearance, still binding structurally — critical minerals. Some battery-metal balances have eased after supply additions and lower prices, but refining security has moved the other way. The IEA reports that the average market share of the top three refining nations for key energy minerals rose to 86% in 2024 from about 82% in 2020, with recent supply growth dominated by Indonesia for nickel and China for most others. Mining investment growth also slowed sharply, to 5% in 2024 from 14% in 2023. Low prices can ease manufacturers today while weakening the future project pipeline.

Eased at the global aggregate — food supply; still binding regionally. FAO describes the global food commodity outlook as still relatively favourable, with cereal production expected to ease from record highs but remain historically elevated. But that surplus is fragile: El Niño risk, energy and fertilizer spillovers, conflict-linked trade disruption, and weather volatility mean the food balance sheet must separate global calories from regional access, affordability, and resilience.

Surpluses, deficits, and capacity gaps

Surplus/cushion: global food commodity availability is not the tightest aggregate constraint this week. Some critical-mineral price pressure also signals near-term material availability in parts of the battery chain.

Deficit: deliverable grid capacity is the clearest deficit. The issue is not simply the number of solar panels, wind turbines, batteries, or data centres proposed; it is whether the grid can connect and move useful power when and where needed.

Hidden deficit: safely managed water access remains a civilizational infrastructure deficit. The 2026 UN World Water Development Report summary says 2.1 billion people still lack safely managed drinking water, with climate change, scarcity, and disasters intensifying inequalities. This matters for households, agriculture, mining, thermal power, semiconductor supply chains, and urban growth.

Balance-sheet stock depletion: ecological capacity remains undercounted. IPBES reports around 1 million species facing extinction, over 85% of wetlands lost, and average native species abundance in major terrestrial biomes down at least 20%. These are not sentimental indicators; they are degraded stocks of pollination, water regulation, fisheries, soil function, disease buffering, and climate resilience.

Emerging trade-offs and second-order consequences

The clean-power buildout can be delayed by the grid it depends on. Decarbonization increases material and grid intensity before it reduces fossil dependence. The downstream cost of rapid electrification is not only more copper, transformers, and permitting conflict; it is curtailment and stranded projects if the connection system is not upgraded.

AI efficiency can increase total demand. More efficient models may lower energy per task, but they can also unlock far more uses. The IEA warns that video generation, reasoning, and agentic tasks may use hundreds or thousands of times more energy per query than simple text generation. The second-order consequence is rebound: efficiency plus mass adoption can raise total load.

Cheap minerals can create future scarcity. Lower lithium or nickel prices help near-term deployment, but if they suppress exploration and new-project finance, the system may trade today’s affordability for tomorrow’s concentration risk.

Insurance repricing transfers physical risk. Allianz reports insured natural-catastrophe losses rising 5–7% annually in real terms. As affordability deteriorates, more losses shift to households and public balance sheets. Adaptation spending that lowers exposure is balance-sheet repair; adaptation that protects expanding exposure may be liability extension.

Long-consequence infrastructure decisions

The longest-consequence decision is grid architecture: what gets connected, where, under what rules, and with what degree of flexibility. Transmission lines, substations, interconnection rules, grid-enhancing technologies, non-firm connection agreements, and distribution upgrades will shape where industry, data centres, EV charging, heat electrification, and clean generation can physically exist. The IEA estimates grid-enhancing technologies and flexible connection approaches could unlock 1,200–1,600 GW of advanced-stage projects now stuck in queues. This is high-leverage because it uses existing corridors and assets while slower buildouts proceed.

Under-watched region or resource

Resource: water as cross-system infrastructure. Water is often treated as local or humanitarian, but it links food, power, mining, cooling, ecosystems, cities, and health. The water constraint deserves more attention than a simple global scarcity metric can provide. The right unit is basin-level usable water under climate volatility, governance quality, and competing demand.

Region watch: emerging Asia, especially India and Southeast Asia, because electricity demand, cooling load, industrialization, food exposure, water stress, and grid expansion intersect there. IEA expects emerging economies to account for nearly 80% of additional electricity consumption through 2030, with India and Southeast Asia becoming more important contributors.

Medium-term outlook change

The medium-term outlook should become more infrastructure-constrained and less commodity-centric. The binding question is not “do we have enough energy/minerals/food globally?” but “can physical systems convert stocks into useful, delivered, affordable services at the needed location and time?” Grid queues, refining concentration, water governance, and insurance affordability now look like first-order determinants of growth and resilience.

The outlook improves where existing systems can be unlocked: grid-enhancing technologies, dynamic line ratings, non-firm connections, recycling, demand flexibility, water governance, and adaptation that lowers actual exposure. It worsens where new demand assumes infinite grid capacity, cheap insurance, frictionless mineral diversification, stable weather, and ecological buffers that can absorb further depletion.

Evidence notes and uncertainties

• Data-centre electricity projections remain uncertain because efficiency, model capability, utilization, disclosure, and capital-market conditions are all moving.
• Critical-mineral balances are not uniform: lithium, nickel, copper, graphite, cobalt, and rare earths have different bottlenecks. Refining concentration is not the same as geological scarcity.
• Food supply data are not interchangeable with hunger, nutrition, affordability, or political stability metrics.
• Climate indicators are physically comparable over time, but the economic consequences depend on exposure, adaptation quality, insurance coverage, and public finance.
• Biodiversity metrics are slower-moving and less frequently updated than energy-market data, but they represent depleted stocks with long recovery times.

Watch list

• Grid queues: do countries unlock existing hosting capacity or only announce new generation?
• Data centres: do power, water, and local infrastructure constraints slow projects before capital does?
• Critical minerals: does low-price relief reduce future investment and increase concentration?
• Food: does El Niño or trade disruption convert comfortable global balances into regional stress?
• Water: which basins show mining, power, agriculture, and urban demand colliding?
• Insurance: where does repricing become withdrawal, and where does public finance absorb the gap?