Semiconductor and AI data center demand scenarios
The need to expand South Korea's nuclear generating capacity is growing as electricity demand from artificial intelligence (AI) data centers and the semiconductor industry rises rapidly.
In a report dated 26, Kyobo Securities analyst Lee Sang-ho projected that 9.3 GW of nuclear power capacity would be needed by 2040, based on electricity demand from semiconductor clusters and AI data centers.
Including all investment plans for Samsung Electronics and SK hynix semiconductor clusters and SK and GS AI data centers, the related industries' ultimate electricity demand is estimated at 38.4 GW. However, actual demand may vary depending on subsequent capital investment, transmission network construction, customer acquisition, financing and electricity supply conditions.
The report put industrial electricity demand in 2040 at 9.7 GW under a conservative scenario, 22.4 GW under a base scenario and 38.4 GW under an optimistic scenario. The base scenario consists of 15 GW for semiconductors and 7.4 GW for AI data centers. It explained that planned demand should be distinguished from actual electricity consumption because not all announced corporate investment plans will be realized.
In the semiconductor industry, the electricity supply plan for Samsung Electronics' Yongin cluster was identified as a key variable. The analysis indicated that plans to supply a cumulative 6 GW through reinforcement of the existing grid are becoming more concrete. SK hynix's Yongin cluster was assessed as having secured a supply network for 3 GW in its initial phase. The Honam semiconductor cluster was assigned 3 GW, taking into account its initial 2 fabrication plants and first-phase supply plan.
Because semiconductor plants expand production facilities in stages, their entire plans do not translate into electricity demand at once. Actual demand is expected to depend on how quickly investment in subsequent production lines proceeds and whether transmission networks capable of supplying large amounts of electricity are built on time.
AI data centers are another major driver of electricity demand. The report's base scenario included 5 GW of demand from SK-related data centers and 2.4 GW from GS's Donghae data center. However, data center projects can materialize only with support such as customer contracts, investment funding and electricity supply agreements.
Although data centers' computing capacity and electricity consumption are increasing together as AI services expand, it is difficult to assume that all facilities included in business plans will be built on schedule. This is why subsequent investment plans in semiconductors and data centers, along with the securing of electricity infrastructure, are considered key variables in medium- and long-term electricity demand forecasts.
Nuclear capacity estimates and their limitations
The report estimated that 9.3 GW of nuclear power capacity would be needed by 2040 to meet the base scenario's industrial electricity demand of 22.4 GW. Applying a long-term capacity reserve margin of 22% gives required generating capacity of 27.3 GW. It then applied the approximately 34% share assigned to nuclear power among newly required capacity in the 11th Basic Plan on Long-term Electricity Supply and Demand.
Under this estimate, required nuclear capacity is 4.0 GW in the conservative scenario, 9.3 GW in the base scenario and 15.9 GW in the optimistic scenario. The base scenario's 9.3 GW is equivalent to approximately 5 large nuclear reactors and approximately 11 modules of the i-SMR small modular reactor.
However, the 9.3 GW figure is not a direct calculation of the capacity shortfall against the existing Basic Plan on Long-term Electricity Supply and Demand, nor does it represent confirmed nuclear plant orders. It is an estimate of required capacity based on industrial electricity demand forecasts, assuming that supply conditions for each generation source do not change substantially. Actual nuclear construction plans and order volumes may vary depending on government electricity policy, licensing, locations and construction periods.
The report assessed that existing generating facilities alone would struggle to meet all future growth in industrial electricity demand. Its analysis was that increasing existing nuclear plants' capacity factor and pursuing continued operation could supplement supply capability, but would be unlikely to replace entirely the need for new facilities to meet rising demand.
Expanding liquefied natural gas (LNG) generating capacity is another option, but a substantial portion of new capacity could be used to replace aging coal-fired power plants. The report therefore explained that, if industrial electricity demand grows as expected, pressure to secure new generation sources could increase rather than being addressed solely through efficiency improvements to existing sources.
Potential opportunities and investment conditions
If nuclear expansion materializes, medium- and long-term order opportunities could also increase for related industries. Nuclear power involves companies across multiple fields, from design and major equipment manufacturing to construction, operation and maintenance. If new nuclear construction and SMR projects become more concrete, they could affect related companies' orders and earnings.
Doosan Enerbility is identified as a potential beneficiary of expanded supplies of major equipment for large nuclear plants and equipment for SMRs. If nuclear plant construction proceeds, opportunities to supply key equipment could increase, while demand for related equipment could also expand if SMR projects enter commercialization. However, the actual scale of benefits will depend on order schedules, supply contracts and the pace of projects.
For KEPCO Engineering & Construction, a key factor to watch is an increase in design orders for new nuclear plants. As nuclear construction plans become more concrete, opportunities in design and engineering could expand. KEPCO KPS is a company expected to benefit from greater maintenance demand as nuclear plant operations expand and continued operation is pursued. Extending existing plants' operating periods and increasing operating capacity could create scope for more maintenance work.
For Korea Electric Power Corporation, attention should be paid to increased electricity sales associated with rising industrial demand and changes in the generation mix. Higher electricity demand could create opportunities for sales volume growth, but the investment burden of expanding generating capacity and building transmission networks must also be considered. Increased electricity sales do not immediately translate into improved profitability, and electricity purchase costs, tariff policy and the scale of capital investment could affect earnings.
Investment momentum in the nuclear industry is not determined solely by expectations of rising electricity demand. Important factors include whether new nuclear construction is incorporated into the government's electricity supply and demand plans, whether actual orders and licensing proceed, and whether related companies secure supply contracts. For SMRs as well, technology development, demonstration, commercialization and economic viability must come first before substantial earnings contributions can be expected.
Growth in AI and the semiconductor industry is making access to electricity an important condition for corporate investment plans. However, electricity demand forecasts remain subject to uncertainty over whether investment plans will materialize and how quickly infrastructure will be built. When assessing nuclear-related companies, long-term demand expectations should therefore be distinguished from actual orders, contracts and the timing of their contribution to earnings.
Original Korean article: Pinpoint News
English brief by Korea Wind Intelligence, koreawindintel.com
