Power rationing can interrupt polysilicon and aluminum production because both supply chains depend on large, uninterrupted electricity loads, turning grid restrictions into physical output constraints.
The impact differs by material: polysilicon shortages raise solar-input costs, while aluminum curtailments tighten metal availability and expose downstream buyers to premiums, delayed deliveries, and wider substitution pressure.
Why Electricity Curtailment Becomes a Supply Shock
Power rationing is not merely a temporary increase in electricity prices. It removes usable output from the production schedule, first reducing operating rates and, if restrictions persist, forcing shutdowns or equipment idling. Reuters reported in 2021 that power rationing would constrain industrial activity until demand weakened enough to restore equilibrium in the domestic electricity market [S3].
Polysilicon production is exposed because the process requires sustained high-temperature, high-energy equipment operation, while aluminum smelting depends on uninterrupted electrolysis to sustain the molten aluminum bath. A short interruption can therefore have consequences beyond the hours of curtailed power, particularly where restart, furnace temperature, and process-control requirements are significant [S3][S4].
The solar-industry supply chain is broader than the finished module. It includes silicon and aluminum raw materials, specialized wire saws and furnaces, warehouse labor, truck drivers, construction crews, and photovoltaic installers [S4]. A grid restriction at the smelter or polysilicon plant reaches the module market through every downstream stage that still requires a continuous flow of material.
Polysilicon Price Transmission to Solar Buyers
At the end of September 2021, polysilicon was approaching $30 per kilogram, compared with an all-time low of $6.80 per kilogram in June 2020, an increase of more than 338% in 16 months [S1]. That historical episode demonstrates how electricity constraints can combine with constrained capacity and speculative purchasing to produce a much larger market response than the lost kilowatt-hours alone would suggest.
The price transmission was not complete. A commonly used M6 mono PERC module increased from $0.21 per watt in January 2021 to $0.24 per watt in September 2021, a 14% increase, as wafer, cell, and module manufacturers absorbed part of the higher polysilicon cost [S1]. The remainder of the increase was therefore passed through the manufacturing chain, while project developers faced the combined effect of higher module prices and schedule pressure.
Power shortages in China limited the production of metals, including silicon and materials used in solar panels [S4]. Polysilicon and aluminum are thus connected, rather than isolated, supply risks: a curtailment can reduce silicon feedstock, while aluminum restrictions affect frame, casing, and mounting-system availability. Procurement teams assessing either material should treat electrical access, not just feedstock tonnage, as part of available capacity.
Aluminum Availability and Downstream Cost Exposure
Aluminum supply was already highly price-sensitive before the 2021 rationing episode: its price rose 80% from May 2020 to August 2021, reaching a 10-year high, while logistics constraints left more material in Asia and tightened supply for US buyers [S4]. Rationing adds a physical-production constraint to that landed-cost problem by reducing the metal available for allocation.
Aluminum is used in the external casing of solar panels, while steel is a mainstay of photovoltaic racking [S4]. Curtailed aluminum output can therefore affect more than one component in a solar project. A buyer may face a higher metal index price, regional scarcity, longer lead times, and processing premiums at the same time, each of which changes the delivered cost of the component rather than merely raising the exchange-traded metal price.
The practical distinction is between a temporary allocation shock and a production loss. If curtailed smelters retain their molten process state and restart after power returns, the market effect may remain largely financial and logistical. Extended outages interrupt the process, raise restart requirements, and remove output permanently until the smelter returns to its normal production rate [S3]. The longer the rationing lasts, the more likely the constraint becomes a physical depletion of available aluminum rather than a temporary scheduling inconvenience.
Demand Weakness Can Mask or Amplify Power Risk
The market did not face the same supply-demand balance in 2026 as it did during the 2021 polysilicon shortage. Solar capacity additions were forecast to fall to 649 GW in 2026 from 655 GW in 2025, while Chinese polysilicon producers were described as facing a “perfect storm” of oversupply and weakening global solar demand [S2]. That combination changes how buyers should interpret a new curtailment event.
Weak demand can initially cushion producers because high inventory and lower orders reduce the urgency of full-rate production. It can also mask deterioration until inventories are consumed. By contrast, a 2021 shortage emerged after major Chinese producers expanded capacity in 2018 and 2019, no new polysilicon capacity came online in 2020, and solar demand recovered faster than expected, triggering a price rally in February 2021 [S1]. The same power restriction can therefore have a much smaller effect when inventories are high and demand is weak than when capacity additions have stalled and downstream buyers are restocking.
For procurement, the key signal is not the headline production loss alone. It is the relationship between curtailed capacity, usable inventory, contracted order backlog, and end-market demand. A 5% cut to a plant with three months of finished stock is operationally different from a 5% cut at an integrated producer with thin inventory and an order book tied to scheduled solar projects.
Criteria Comparison: Polysilicon Versus Aluminum
Polysilicon and aluminum are exposed to the same rationing mechanism, but their direct substitutes, inventory structures, and market signals differ. Comparing them by electrical intensity, inventory buffer, downstream product, and price mechanism gives procurement teams a practical framework for estimating exposure. [S1]
Polysilicon is more directly tied to a narrow solar feedstock chain, while aluminum competes across construction, transport, packaging, and electrical applications. This makes a solar project especially sensitive to polysilicon inventory conditions, but it also makes aluminum procurement vulnerable to broader industrial demand and regional logistics rather than solar demand alone.
Failure Modes That Can Mask a Shortage
Inventory hoarding can create a misleading price signal before physical supply becomes scarce. During the 2021 episode, panic buying and speculative traders hoarding material pushed the April spot price sharply higher, although reports of traders selling inventories appeared by early August [S1]. Inventory accumulated by upstream producers can also depress margins and create a risk of excessive stock rather than sustained scarcity.
A second failure mode is assuming that lower demand automatically repairs supply. Polysilicon oversupply and weakening solar demand were expected to pressure smaller companies in the supply chain in 2026 [S2]. Such an adjustment can remove marginal production capacity, but it does not restore electricity availability. It can also leave buyers dependent on a narrower supplier base if financially weaker producers exit.
Logistics create the third failure mode. More than 80% of global trade was transported by ocean freight in the cited analysis, and a container cost rose from $1,920 in 2020 to more than $14,000 in 2021, an increase above 600% [S4]. Rationing can therefore be followed by shipping disruption: the metal or polysilicon may exist, but it may not arrive within the production window required by wafer, cell, module, or fabrication customers.
Sourcing and Contracting Response
A supply-risk review should separate three stages: nameplate capacity, electrically available capacity, and deliverable material. Nameplate capacity can remain unchanged while a power order lowers the second figure. Deliverable material can fall further when transport, inventory ownership, export controls, or regional premiums intervene [S3][S4].
For polysilicon, buyers should request producer-level operating rates, finished and in-process inventory, and delivery commitments tied to wafer and cell schedules. The historical M6 module price movement from $0.21 per watt to $0.24 per watt illustrates why module delivery dates and input-price escalation clauses need separate review [S1].
For aluminum, contracts should specify the metal index, conversion or fabrication charge, regional premium, and the delivery location. That structure distinguishes a market-price increase from a physical smelter curtailment and prevents the two from being confused. A related analysis of aluminum smelter power contracts and metal availability extends the contracting logic to long-term electricity arrangements and restart risk.
Electricity resilience also belongs in the equipment plan. Plants with backup generation, demand-response arrangements, and protected critical loads can preserve process continuity, but backup systems do not replace a grid-supply plan for high-tonnage production. The decision variable is how many hours of constrained power the plant can tolerate before inventories, order schedules, and restart procedures become binding.
Two trackable signals will indicate whether rationing is moving from a temporary allocation issue into a physical supply shock: producer inventory falling below contracted delivery requirements, and smelter or polysilicon operating rates remaining below normal schedules after grid conditions improve.