Are all grids are created equal? (Or are some grids are more equal than others?)
Even when we talk about “the grid” in the singular, most power systems are really a constellation of islands and semi‑islands learning how to run on renewables and storage.
Earlier this week, I outlined the case from Australia.
But, what about other islanded grids? Some are obvious – small island states in the Pacific or Caribbean. Others are carved out of huge continental networks, but behave like islands in practice because physics, politics or market rules limit how much help they can get from their neighbours.
That is where some of the most useful transition lessons are emerging.
At one end of the spectrum are true geographic islands. Hawaii is now the canonical example. The last coal plant on Oahu was closed in 2022 and replaced much of its capacity with a 185 MW battery that supports a rapidly growing fleet of solar projects.
Coal is gone, oil is shrinking (down to about two-thirds of its electricity in early 2026), and storage has become a core provider of grid services rather than a bolt‑on. In a place where every litre of fuel arrives by ship, the economics of that shift are as important as the emissions.
Puerto Rico is trying, but with resilience first, decarbonisation close behind. After Hurricane Maria exposed the fragility of its centralised grid, the rebuild has leaned heavily on solar‑plus‑storage microgrids.
FEMA is funding a nearly $100 million microgrid for Vieques and Culebra that brings 15 MW of solar and 11.6 MWh of storage to existing diesel plant, specifically to keep power on when the main system fails.
Non‑profit and community projects are backing at least ten more microgrids around critical services such as clinics and supermarkets. Here the story is less about hitting a national renewable percentage and more about making sure the lights stay on through the next storm, with less diesel burned each year.
Across the Pacific islands – from Fiji and Samoa to Tonga, Vanuatu and Palau – the pattern is similar. These systems still rely heavily on diesel, and Australian proponents are boosting to import Australian rubbish for waste-to-energy – but the economics of solar and batteries are better.
Every litre of fuel saved is foreign exchange retained and risk reduced. The technologies are familiar – PV, batteries, flexible engines, smarter controls – but they are being assembled in ways that fit small, often fragile grids. In that context, “energy transition” is as much about sovereignty and balance of payments as tonnes of CO₂.
Europe is usually described in terms of large synchronous areas and cross‑border interconnectors, yet it too contains grids that operate more like islands. Ireland and Northern Ireland share a system linked to Great Britain only by HVDC cables; from a stability and planning standpoint it behaves like a distinct islanded grid.
Over the last decade that system has shifted decisively towards wind. Wind already supplies roughly a third of its electricity, and coal at Moneypoint closed midway through last year. Batteries have entered the market to provide fast system services.
Ireland’s experience is a live demonstration that a small, windy grid can run reliably with very high variable renewable shares supported by a modest gas fleet and growing storage.
The Baltic states show a more geopolitical form of islanding. Estonia, Latvia and Lithuania historically operated as part of the Russian‑controlled BRELL ring. Their decision to desynchronise and connect to the Continental Europe synchronous area required them to prove they could keep their systems stable without Russian support.
At the same time, they have rapidly deployed utility-scale BESS as a critical component of synchronizing their power grids with continental Europe in February 2025, and are growing renewables and planned large offshore wind farms in the Baltic Sea.
In this case, islanding is both physical and political: shifting away from dependence on Russian fossil fuel and grid services toward a more autonomous renewable system anchored in the EU.
The same “hidden islands” appear inside the big North American grid. On a map of high‑voltage lines, the United States looks fully meshed; in reality, it is split into three main synchronous areas: the Eastern Interconnection, the Western Interconnection and ERCOT in Texas.
ERCOT behaves like a de facto islanded grid, with only limited DC links to neighbours and its own market rules. In the last few years Texas has seen one of the fastest battery build‑outs anywhere.
By early 2026 ERCOT had close to 14 GW of battery capacity online, almost double the level a year earlier, alongside tens of gigawatts of wind and solar.
That build‑out is rapidly changing the role of gas peakers and reshaping price patterns, showing how a large, quasi‑island grid can lean on storage and solar to manage variability and extreme events.
Within North America there are also literal islands and remote regional systems: Alaska’s patchwork of microgrids, remote Canadian communities, and thousands of microgrids on military bases, campuses and industrial sites designed to “island” during disturbances.
These are not just resilience experiments. They are test beds for technology combinations – high‑penetration renewables, long‑duration storage, advanced controls – that can then scale into the main interconnections.
China looks monolithic at first glance: one national giant, knitted together by ultra‑high‑voltage AC and DC lines. Internally, though, grid companies still recognise distinct regional systems.
Remote regions such as Xinjiang, Tibet and Hainan have historically operated with weaker ties to the coastal load centres and, in some cases, their own balancing and dispatch logic. But large-scale transmission lines for renewable energy are being built to feed eastern demand hubs.
Xinjiang in particular has surged ahead in wind and solar, at times generating more renewable energy than local demand and relying on long‑distance DC lines to export the surplus (and ecological benefits, too).
In that sense it behaves like a resource‑rich island at the edge of a continent, grappling with congestion, curtailment and the challenge of turning raw renewable potential into reliable, monetised capacity.
There are also compact systems that act like conceptual islands.
Singapore has limited domestic renewables and no room to waste. So it is investing in efficiency, flexible gas, regional imports and early‑stage hydrogen projects, while rapidly transforming its grid to incorporate renewables, exploring floating solar and storage, and aiming for 6 GW of low-carbon electricity imports by 2035 and solar deployment of at least 2 GW by 2030.
It shows what a high‑income, resource‑constrained system can do when it accepts that it will always be an energy island in resource terms, even if it is physically connected to neighbours.
Across all these cases, the common thread is not geography but autonomy. Islanded and semi‑islanded systems cannot pretend the old fossil fleet will always come to the rescue.
They must plan for independence: high shares of renewables by necessity; storage and flexible demand as core tools; grid codes and markets tuned to fast response; and a clear understanding of what happens when tie‑lines fail.
For larger interconnected systems, their experience is a preview. Even on a continental grid, there will be moments – cyber-attacks, extreme weather, geopolitical shocks – when parts of the system must stand on their own.
Jurisdictions that have already learned to operate as islands, whether literal or metaphorical, will not be starting from scratch when the rest of the world is suddenly forced into island mode.
If you would like to join more than 29,000 others and get the latest clean energy news delivered straight to your inbox, for free, please click here to subscribe to our free daily newsletter.
