Landsvirkjun has agreed to sell Elkem 127 MW of Icelandic power for ten years from April 2029. The contract has no price. It has a floor, and above that floor the electricity price moves with the market price of ferrosilicon, the alloy the smelter buying the power exists to make. That turns a fixed cost into a variable one, and it is the reason the deal is interesting.
Key takeaways
- Landsvirkjun signed a corporate power purchase agreement with Elkem for 127 MW of renewable supply in Iceland, announced on 22 October 2026, running ten years from April 2029 on a take-or-pay basis.
- From April 2029 the electricity price will include a fixed floor price and a partial linkage to the market price of ferrosilicon, so the tariff is a formula rather than a figure and no rate is disclosed.
- The agreement supersedes the previous Landsvirkjun and Elkem contract, which ran from 1 April 2019 to 31 March 2029, so the replacement was agreed roughly 29 months before the existing deal expires.
- At 127 MW this is the largest stated capacity offtaken among the 29 Nordic hydro power purchase agreements Enerdatics records, and only seven of those 29 state a capacity at all.
- Not one of those 29 Nordic hydro agreements carries a disclosed tariff, and Landsvirkjun is the supplier on four of the five Icelandic power purchase agreements Enerdatics records in total.
What does linking power to ferrosilicon actually do?
It moves the risk of a bad market from the smelter to the power producer. Ferrosilicon is an iron and silicon alloy used mainly in steelmaking, and electricity is one of the largest items in the cost of making it. A smelter on a fixed power price has a fixed cost that does not care what its product sells for, so when the ferrosilicon price falls the margin is crushed from both ends. Link the power price to the product price and the largest input cost falls when revenue falls.
The floor is what stops that becoming a one-way bet. A fixed floor price means Landsvirkjun is guaranteed a minimum regardless of where ferrosilicon trades, so the producer carries the upside and downside of the alloy market only above a level it has already secured. Elkem gets relief in a downturn; Landsvirkjun gets participation in a boom and protection in a slump.
The take-or-pay term completes the split. Take-or-pay means Elkem pays for the contracted 127 MW whether or not it draws the power, so volume risk sits with the buyer even though price risk is shared. A smelter that idles a furnace still pays. Read together, the structure says Landsvirkjun was willing to accept commodity exposure it does not control, in exchange for a floor and a guaranteed volume.
Why renegotiate 29 months before the old contract expires?
Because both sides need the certainty earlier than the expiry date. The existing agreement runs to 31 March 2029 and the new one starts the following day, so supply is continuous and nothing was at risk of lapsing. What changes is planning. A smelter deciding whether to keep furnaces running, upgrade them or close them is making a decision measured in years, and it cannot make it against an unpriced 2029.
The producer has the same problem in reverse. Landsvirkjun's portfolio is roughly 2 GW of hydropower, 155 MW of geothermal and 1.8 MW of wind, and 127 MW committed for a decade is a meaningful block of it. Iceland's heavy industry has long been the anchor load on that system, and locking in a large industrial buyer well ahead of time is how a hydro producer decides what else it can sell and whether it needs to build.
The alternative load is already visible. Enerdatics records Landsvirkjun signing a 12 MW supply agreement with Borealis Data Center in June 2026, and a contract with Reykjavik Data Centre as far back as March 2020. Data centres pay for firm power and are growing; smelters are the incumbent load and are cyclical. Agreeing the industrial renewal early fixes the base before the newer demand competes for it.
What can anyone tell about the price?
Nothing, and the Nordic record says that is normal rather than evasive. Across the 29 Nordic hydro power purchase agreements Enerdatics records, covering Norway, Sweden, Finland and Iceland, not one discloses a tariff. Only seven state a capacity offtaken. Durations range from three to 17 years and are the most consistently disclosed term in the dataset, which tells you what the parties consider commercially harmless to publish.
Elkem itself illustrates the pattern. Enerdatics records it as the offtaker on four of the 29, the other three all Norwegian: a 43.58 MW seven-year agreement with Statkraft for the Bjolvo plant in June 2026, a ten-year contract with Nord Trondelag Electricity in July 2025, and a seven-year agreement with Hafslund in December 2024. Four hydro offtakes in under two years, across two countries, none with a published price.
Enerdatics records only five Icelandic power purchase agreements in total, of which Landsvirkjun supplies four, and none of the five carries a tariff. The only other Icelandic contract at comparable scale is ON Power's 150 MW geothermal supply to the Nordural smelter in April 2025, agreed for five years. Against that, a ten-year 127 MW industrial contract is the longest large commitment on the Icelandic record, and the ferrosilicon linkage is the only pricing information any of the five discloses.
What does the deal signal for Nordic industrial power?
That producers in long, cheap hydro markets are being asked to take commodity risk to keep industrial load. Nordic hydro systems are structurally long on power and the prices show it, which is why aventron sold its entire 143 MW Norwegian hydro fleet in August 2026 and cited persistently low prices expected to stay low. A producer facing that outlook has little leverage to insist on a flat tariff, and a floor plus upside participation is a better trade than a low fixed price for ten years.
The second signal is about what industrial buyers now want from a contract. Matching the power price to the product price is a hedge against the buyer's own market, not against the electricity market, and it only works where the producer knows the buyer's commodity well enough to underwrite it. That is a narrower trade than the financially settled contracts dominating corporate procurement elsewhere, where Cargill contracted 87 MW of South Dakota wind on a virtual basis and received credits rather than electrons. Expect more commodity-linked structures where a single large smelter and a single large producer sit on the same grid and have run out of other options.
Frequently asked questions
What price is Elkem paying for the power?No tariff is disclosed. From April 2029 the price includes a fixed floor and a partial linkage to the market price of ferrosilicon. None of the 29 Nordic hydro power purchase agreements Enerdatics records carries a disclosed tariff.
What does take-or-pay mean here?Elkem pays for the contracted 127 MW whether or not it draws the power, so the volume risk sits with the buyer. Price risk is shared, because the producer is guaranteed a floor but exposed to the ferrosilicon market above it.
When does the new contract start and what does it replace?It runs ten years from April 2029 and supersedes the previous Landsvirkjun and Elkem agreement, which was in effect from 1 April 2019 through 31 March 2029.
Ready to get deal-ready answers in seconds? Try Enerdatics Leap AI and access verified intelligence across M&A, financings, PPAs, projects and energy market developments through natural language.
.png)
.jpeg)