NextFin News - China’s power planners are confronting a forecasting problem that goes beyond a simple demand upgrade: the economy is becoming more electric, more digital and less predictable by the rules that once made power planning straightforward. Ren Yuzhi, director-general of the planning department at the National Energy Administration, said China now faces greater uncertainty in predicting energy demand because structural changes and the rapid expansion of new industries are reshaping consumption patterns. He also said demand over the past five years exceeded the government’s expectations.
That matters because the country is entering the 2026-2030 planning period with electricity demand still rising fast enough to complicate every part of the grid equation. The National Energy Administration says China is expected to add around 600 billion kilowatt-hours of electricity consumption each year during that period. The growth will be driven by computing power, electric-vehicle charging demand and steady gains in power use by other industries and households.
The broader backdrop shows why the warning is so important. The International Energy Agency says China will remain the single largest contributor to global electricity demand growth through 2030, accounting for close to 50% of the increase worldwide. It also says China alone is expected to add demand equivalent to the total electricity consumption of the European Union today, with average annual growth of 4.9% through 2030.
In other words, China is not struggling to find demand. It is struggling to map it. A slower percentage growth rate than the last decade still translates into a very large absolute increase because the base is already huge. Once computing power, EV charging and other new loads start growing together, the old industrial model of power forecasting becomes too blunt to catch where the next bottlenecks will emerge.
The result is a more complicated planning job for the government and a more demanding infrastructure challenge for the market. Power planners now have to think not only about how much electricity China will use, but where that electricity will be consumed, at what times of day, and how much transmission, storage and generation flexibility will be needed to support it.
That shift is already visible in the language of officials. The National Energy Administration has framed computing power and EV charging as major contributors to the next phase of electricity growth, which means the demand story is increasingly being written by technology adoption rather than only by factories and households. For grid operators, that changes the type of forecasting error that matters most: not just the annual total, but the timing, geography and intensity of new load.
The challenge is structural, not temporary. AI workloads can expand in bursts, data centers can cluster in specific provinces and EV charging patterns can change quickly as fleets and incentives evolve. Those characteristics make demand more difficult to smooth into a single national trend. They also make grid investment more sensitive to assumptions that may have been adequate in the past but are now losing precision.
China’s planners are therefore dealing with a system in which the headline direction is obvious — demand keeps rising — while the path is less legible than before. That is a meaningful change for a country that has long relied on central forecasting to guide generation buildout, transmission investment and industrial policy. If those forecasts are noisier, every downstream decision becomes a little harder to calibrate.
The Forecasting Problem Is Structural, Not Temporary
The immediate issue is that China’s energy system is being pulled by new sectors faster than the planning process can fully normalize them. Computing power is the clearest example. Unlike steel mills or chemical plants, digital infrastructure can scale in discrete jumps, and those jumps are often tied to chip supply, policy support and corporate spending plans rather than to broad industrial output alone.
Electric vehicles add another layer of uncertainty. Charging demand depends on fleet growth, charging behavior, local incentives and grid access, which makes it harder to project with the same tools used for more traditional industrial loads. The National Energy Administration’s own framing suggests these are no longer peripheral contributors; they are central drivers of the next phase of electricity growth.
That is why the forecast itself is becoming a policy variable. If planners underestimate the speed or geography of demand growth, transmission and distribution bottlenecks can appear before new capacity is in place. If they overestimate it, capital can be misallocated into projects that will not be fully used. Either error is costly, which is why the planning department’s caution is significant even though the overall demand trend remains upward.
One way to read the warning is that China’s energy system is moving from a predictable industrial load profile to an infrastructure profile shaped by digital services and electrified transport. That transition matters because the old model was built around relatively steady demand blocks; the new one is more uneven, more regional and more dependent on technology adoption cycles.
“China is expected to see an annual increase of around 600 billion kilowatt-hours (kWh) in electricity consumption during the 15th Five-Year Plan period (2026-2030).”
That line from the National Energy Administration is important not because it is startling, but because it is so large and so specific. A 600 billion-kWh annual increase is the kind of number that turns forecasting into a grid-design problem. It implies that marginal errors in planning could have oversized consequences for transmission, reserve margins and the pace of clean-energy integration.
Why the Old Model Is Losing Power
China’s old electricity planning model was designed around a familiar mix of heavy industry, household consumption and slower-moving economic cycles. The new mix is more dynamic. AI workloads can add concentrated demand in particular regions. Data-center operators can expand quickly, but power infrastructure cannot. EV charging can scale with policy support, but its load shape is highly sensitive to when and where people plug in.
That mismatch is why Ren Yuzhi’s warning is best understood as a planning problem rather than a one-off forecast revision. Officials are not simply estimating megawatt-hours; they are estimating the geography and timing of a rapidly changing economy. The faster the technology stack expands, the more the electricity system has to act like anticipatory infrastructure instead of a reactive utility grid.
The International Energy Agency’s outlook reinforces the scale of the challenge. If China is responsible for close to half of global electricity demand growth through 2030, then any error in Chinese demand modeling ripples far beyond the country’s borders. It can affect fuel markets, equipment orders, grid spending and the global trajectory of power-sector emissions. The world’s biggest incremental demand center is also becoming less easy to read.
That does not imply an imminent shortage or a demand shock. It does imply that the policy response is likely to center on grid flexibility, transmission buildout, renewable integration and closer coordination between computing infrastructure and power supply networks. The forecasting gap itself is becoming a policy driver, because the system has to be built with more room for error than before.
The implication for investors is not a trade signal but a structural one: Chinese power demand is still rising, yet the mix of demand is shifting quickly enough to make official projections less precise. That increases the importance of equipment suppliers, grid developers, storage providers and operators that can adapt to uneven load growth, while raising the bar for any forecast that assumes yesterday’s industrial pattern still describes tomorrow’s demand.
What Investors and Policymakers Should Watch Next
The next catalysts are likely to come from policy guidance on grid coordination, data-center siting, EV charging infrastructure and any revisions to power-demand assumptions in the 15th Five-Year Plan period. Those decisions will show whether planners view the forecasting gap as a technical nuisance or as a structural feature of a more digital economy.
For the broader market, the lesson is that China’s power demand story is not becoming less important; it is becoming harder to model. That makes official forecasts less about precision and more about range management. If the rise of AI and EVs keeps widening the gap between historical patterns and current consumption, then energy planning in China will increasingly be about managing uncertainty rather than eliminating it.
Explore more exclusive insights at nextfin.ai.
