Series: Energy, Materials & Industrial Systems Domain: Electricity Systems Program: Technology, Production & Society Code: MT-EMIS-2026-09-16-electricity-systems Edition: September 16, 2026 Information cutoff: 2026-09-16, 12:30 BRT Overall confidence: high on current capacity, queue and investment conditions; medium on the pace at which regulatory and grid-enhancing measures can release capacity.
Central assessment. The central constraint in many electricity systems is shifting from the ability to add generation to the ability to connect, transmit and balance it. Record renewable additions, rapidly growing storage, data-centre loads and electrification are increasing the value of grid capacity, flexibility and electrical equipment. The economic question is no longer only how cheaply a megawatt can be generated; it is whether the system can connect that megawatt, move it to load, balance it in time and procure the equipment required to do so.
1. Executive assessment
The global electricity system entered 2026 with a widening asymmetry.
On the supply side, renewable capacity is expanding at record scale. IRENA reports that 692 GW of renewable power capacity were added in 2025, taking global renewable capacity to 5,149 GW. Renewables represented 85.6% of total net power-capacity additions during the year. These figures measure installed capacity, not actual electricity generation, but they establish the speed at which the physical generation fleet is changing. IRENA, Renewable Capacity Statistics 2026
On the network side, the build-out is slower. The IEA estimates that more than 2,500 GW of renewable generation, storage and large-load projects are stalled in grid connection queues worldwide. Annual grid investment is around USD 400 billion and would need to rise by roughly 50% by 2030 to meet expected electricity demand. IEA, Electricity 2026 — Executive Summary
This does not imply that all queued projects are economically viable or will be built. Connection queues contain speculative, overlapping and immature proposals as well as advanced projects. The relevant signal is the scale of the mismatch between the speed of project development and the speed at which networks can absorb new generation and load.
The consequence is a change in where scarcity sits. In systems with abundant generation proposals, a connection right, transformer slot, substation upgrade, transmission corridor, dispatchable resource or flexible demand contract can become more economically important than another generation project with no path to the grid.
2. The system has to be analysed as a system
Electricity capacity is not equivalent to electricity availability.
A useful analytical chain is:
Each stage can become the binding constraint.
A solar plant can be inexpensive and fully financed but still have low system value if its connection is delayed for years. A data centre can have capital and land but remain constrained by interconnection. A country can have strong renewable resources but face curtailment if transmission expansion and flexibility lag.
This is why the series treats an electricity matrix as more than percentages of hydro, solar, wind, nuclear, gas or coal. The relevant system includes:
- installed capacity and actual generation;
- peak demand and load shape;
- transmission and distribution capacity;
- cross-border interconnections;
- storage and other flexibility resources;
- dispatchability and reserve margins;
- connection queues and curtailment;
- investment, financing and regulated returns;
- manufacturing capacity for transformers, cables and switchgear;
- permitting, land, standards and workforce.
The same nominal generation mix can therefore produce very different reliability, cost and investment outcomes across countries.
3. Grid queues are now a capital-allocation problem
The IEA's 2026 assessment places more than 2,500 GW of generation, storage and large loads in grid queues globally. A large queue is not automatically evidence of underinvestment in wires alone. Queue design can itself create congestion when developers have weak incentives to withdraw low-probability projects or when studies are processed sequentially.
The more important finding is that a meaningful share of advanced projects may be connectable without waiting for full conventional reinforcement.
The IEA estimates that 1,200–1,600 GW of advanced-stage projects could potentially be enabled through a combination of more flexible connection arrangements and grid-enhancing technologies. Roughly 750–900 GW could be connected through non-firm agreements that allow output or consumption to be curtailed under specified conditions. The remainder could come from measures such as dynamic line rating, advanced power-flow control, reconductoring and voltage uprating. IEA, Electricity 2026 — Executive Summary
This matters economically because it separates two problems:
- physical scarcity, where new infrastructure is genuinely required; and
- utilisation scarcity, where existing infrastructure is not being used as efficiently as it could be.
The capital response should differ. Building a new transmission corridor can take many years and billions of dollars. Software, sensors, reconductoring, topology optimisation or non-firm connection rules can sometimes release capacity faster, although they cannot substitute for structural expansion where the network is genuinely undersized.
4. Flexibility is becoming part of the generation economics
As variable solar and wind grow, the value of a unit of generation increasingly depends on when it produces and what the system can do with that output.
The IEA expects the share of global electricity generation from solar PV and wind to rise from about 17% today to 27% by 2030. At the same time, large concentrated loads such as data centres, electric vehicles and heat pumps are changing demand patterns. IEA, Electricity 2026 — Executive Summary
Battery storage is one response. Utility-scale batteries are expanding rapidly and falling costs are improving their economics, but batteries do not remove the need for networks or all forms of dispatchability. Their value depends on sensitivity to long-term interest rates, cycling, location, market design and the shape of the system's imbalance.
Short-sensitivity to long-term interest rates storage can shift solar output across hours; it is not automatically a substitute for multi-day, seasonal or transmission constraints. The correct comparison is therefore not "renewables versus storage" or "generation versus grids." It is the marginal value of each additional investment inside a constrained system.
5. Capital is abundant in energy, but not evenly matched to system needs
The IEA expects total global energy investment to reach about USD 3.4 trillion in 2026, with roughly USD 2.2 trillion directed to renewables, nuclear, grids, storage, low-emissions fuels, efficiency and electrification. IEA, World Energy Investment 2026
The relevant issue is composition.
Generation assets often have clear unit economics, standardised technologies and shorter development cycles. Transmission and distribution projects face regulated returns, public planning, permitting, land acquisition, long equipment lead times and political allocation of costs. That creates a structural tendency for generation proposals to arrive faster than network capacity.
The timing mismatch is an industrial and institutional constraint, not only an engineering constraint.
6. Distribution of capacity and power
The Marginal Thinking question is: who gains capacity when grid scarcity rises?
Potentially scarce positions are not limited to electricity generators. System value can shift toward actors controlling or supplying:
- transmission corridors and interconnection rights;
- substations and connection capacity;
- power transformers and high-voltage cables;
- grid-enhancing hardware and software;
- battery storage at constrained nodes;
- dispatchable generation and demand response;
- engineering, procurement and construction capacity;
- permitting and regulatory processes that shorten time to connection.
At the country level, the ability to turn cheap generation into reliable delivered electricity can affect industrial competitiveness. A jurisdiction with low-cost renewable resources but slow grid expansion may fail to translate resource abundance into manufacturing advantage. A jurisdiction with a more expensive generation mix but strong networks, equipment supply and predictable connection rules can sometimes offer a more valuable power environment to industry.
7. What would change the assessment
The thesis that grids and flexibility are becoming the binding constraint would weaken if several indicators moved together:
- global connection queues declined materially after adjusting for project quality;
- grid investment accelerated faster than generation investment for several years;
- procurement times and prices for transformers and cables normalised;
- curtailment and congestion costs declined despite continued renewable and load growth;
- interconnection times for large new loads shortened materially;
- large additions of flexible demand, storage and grid-enhancing technologies absorbed new capacity without equivalent network expansion.
One year of improvement would not be sufficient. The underlying mismatch is the accumulated result of multi-year differences in development cycles.
8. Implications for the series
Electricity research in Energy, Materials & Industrial Systems will therefore track five linked layers:
| Layer | What matters |
|---|---|
| Generation | capacity, actual generation, cost, dispatchability |
| Network | connection, transmission, distribution, interconnection |
| Flexibility | storage, demand response, dispatchable supply, market design |
| Industrial base | transformers, cables, switchgear, conductors, engineering capacity |
| Capital and institutions | investment, regulated returns, permitting, planning, financing |
The purpose is not to identify a preferred generation technology. It is to identify the physical and institutional constraint that determines the next unit of reliable power.
Sources
- International Energy Agency — Electricity 2026
- International Energy Agency — Electricity 2026: Executive Summary
- International Energy Agency — World Energy Investment 2026
- IRENA — Renewable Capacity Statistics 2026
Method note. Installed capacity, actual generation, connection-queue capacity and investment are different measures and are not added together. Queue data are treated as an indicator of system pressure rather than as a forecast that every queued project will be built.