Three years is a long time in hydrogen strategy. The Netherlands’ 2023 National Energy System Plan projected roughly 460 petajoules of gaseous hydrogen demand in 2040. Its 2026 National Energy System Plan update cuts that central estimate to about 200 PJ, a reduction of roughly 57%. This is especially interesting as few European countries committed as broadly to the hydrogen-economy thesis as the Netherlands. Germany went further on scale, but the Dutch built policy around production, imports, pipelines, storage and manufactured demand at the same time. The applications disappearing from the forecast are as revealing as the headline number. Hydrogen is largely gone from dispatchable electricity generation, has been pushed out of low- and medium-temperature industrial heat by direct electrification, barely appears in buildings beyond pilots, and remains only a small direct energy carrier in road transport, aviation and shipping. Much of the synthetic-fuel production previously imagined inside the Netherlands also shifts toward imports of energy-rich intermediates and finished fuels from regions with cheaper renewable electricity. At the same time, electricity moves much more firmly to the centre of the Dutch energy system. The new plan has electricity supplying 52% of final energy by 2040, up from roughly 25% today, alongside extensive electrification of transportation, buildings and industry. That figure understates the transformation because final energy measures what reaches the customer rather than the useful work produced. Electric drivetrains waste far less energy than combustion engines, while heat pumps deliver several units of heat for each unit of purchased electricity. The Dutch plan’s own assumptions have 71% to 88% of road-transport activity electrified depending on vehicle class, about 64% of households using all-electric or hybrid heat pumps, and industrial electricity consumption rising substantially. A system with electricity at 52% of final energy can therefore have a much larger share of its actual mobility, heat and industrial work electrified. The full TFIE Strategy Briefing analysis asks what has to happen for that remaining growth to appear, what happens if it does not, and whether infrastructure planned around a much larger hydrogen future is being tested against the downside strongly enough. The hydrogen correction looks dramatic against that increasingly electric system, but 200 PJ is still not a small Dutch hydrogen market. Current gaseous-hydrogen demand is about 130 PJ, overwhelmingly in refining, ammonia and chemicals. On a lower-heating-value basis, that is roughly 1.1 million tonnes of hydrogen per year. The revised 2040 central case is about 1.7 million tonnes, while the discarded 460 PJ vision implied roughly 3.8 million tonnes. After removing more than half of the previous forecast, the Netherlands still assumes hydrogen demand grows by about 54% from today. The reassessment is not the product of an ideological turn against hydrogen. The economics changed as projects, financing and competing technologies provided better evidence. Earlier Dutch analyses entertained renewable hydrogen at roughly €1.70 to €4.70 per kilogram around 2030. The 2026 plan now cites bids into the first European Hydrogen Bank auction averaging about €9.80 per kilogram and Dutch 2040 estimates around €6.70 to €6.80. Those numbers make it much harder to justify using hydrogen where electricity can perform the same service directly, particularly for industrial heat, buildings, road transport and electricity generation. The plan explicitly concludes that hydrogen is too expensive to serve as either a temporary or structural alternative to electrification for low- and medium-temperature process heat. As a note, those prices are in the range I have been predicting for years, having done technoeconomic workups on projects on multiple continents to figure out if hydrogen could be both green and cheap. That is a somewhat useful example of what technology planning looks like when comparative evidence is allowed to change the forecast. The Netherlands has not simply lowered a hydrogen target. It has removed hydrogen from applications where batteries, heat pumps, electric industrial equipment or imported industrial feedstocks increasingly make more sense. Hydrogen remains important in existing industrial chemistry and in whatever other applications can support its delivered cost, but that is a materially different proposition from an expanding hydrogen energy system spanning transport, heat, power generation and synthetic fuels. The problem is that the Netherlands appears to have stopped the correction too early. Cutting the 2040 forecast from 460 PJ to 200 PJ looks aggressive until the remaining number is tested against the evidence in the plan itself. Current demand is only about 130 PJ, refinery output is expected to fall roughly by half, renewable ammonia increasingly loses to imports, synthetic fuels increasingly move to lower-cost production regions (reality around biofuels is lagging reality around hydrogen still), and hydrogen has already been pushed out of most heating, power generation, buildings and road transport by direct electrification. The plan also documents renewable-hydrogen costs far above those assumed when the larger hydrogen economy was conceived. Yet after incorporating all of those downward pressures, its central case still requires hydrogen demand to grow by more than 50% from today. The NPE acknowledges that demand could fall below 200 PJ if mandates weaken, Dutch production remains expensive, electrification keeps winning or the assumed new industrial activity fails to appear, but then declines to quantify that outcome. Instead it models 200 PJ as the base case and 300 PJ as the upside. On the evidence assembled in the NPE itself, that is too timid a revision: the quantitative work cuts away most of the applications that were supposed to create hydrogen growth, then preserves the growth conclusion anyway. That matters because the missing downside is not an academic scenario. Hydrogen infrastructure is being financed against expectations of future customers and throughput. Delayed demand can sometimes be handled by spreading early network costs over later users; permanently lower demand cannot. Once the plan’s own sector assumptions are allowed to flow through to the aggregate denominator, declining hydrogen demand becomes a scenario that deserves to be modelled explicitly, not left as an acknowledged possibility outside the quantitative analysis. The 2026 update is a major improvement on 2023, but its hydrogen forecast has not yet caught up with the evidence that produced the update in the first place. Read the full TFIE Strategy Briefing analysis for the missing downside case, the interaction with changing EU hydrogen policy, and what lower demand could mean for infrastructure already being developed around it.