SE2 versus SE3 siting — a heavier Stockholm-area queue cell is not a Sundsvall spare megawatt

Reading order

Seat the bidding area and the paper class before you compare slogans. SE2 Sundsvall and SE3 Stockholm are wholesale-price names. A 2 September 2026 data-centre application cell is a TSO queue paper. A municipal welcome, a detaljplan and a regional-network letter are other papers. Write the issuer, the date and the object on one line before you argue that one mid-Sweden zone is “easier”.

Then keep the two application cells on their own rows. 1 320 MW in SE2 and 5 169 MW in SE3 are printed data-centre consumption cells before reservation. They are not reserved halls, not IT load and not a spare-power table. A blank SE4 cell on the same table is still a blank cell. Do not invent a mid-Sweden leftover by subtracting SE2 from SE3.

Then keep operator geography on its own rows. Microsoft’s 16 November 2021 notice names Gävle and Sandviken, which sit in SE2, and Staffanstorp, which sits in SE4. EcoDataCenter Falun sits in SE3. Google Horndal sits in Avesta and SE3. atNorth SWE04’s 23 February 2026 Sollefteå notice sits in SE2. evroc’s Arlandastad purchase sits in the Stockholm area and SE3 as a land paper. Those sentences do not convert the TSO cells into campus reservations.

Then keep the open area-review clock on its own row. The government’s 29 May 2026 assignment, with a 29 January 2027 report date, is open work. ENTSO-E’s April 2025 bidding-zone study did not propose changing Sweden’s four areas. Neither paper redraws SE2 against SE3 for a hall file.

• Bidding-area name and paper class before a mid-Sweden siting slogan.
• 1 320 MW and 5 169 MW are application cells, not reserved halls.
• Operator geography stays geography, not a queue conversion.
• The 2026–2027 review is open work, not a redrawn pair.

What an SE2 clock and an SE3 clock actually measure

Energimarknadsinspektionen’s four names measure wholesale-price constraints, not plot offtake. SE2 Sundsvall and SE3 Stockholm answer which day-ahead price a location sits under. They do not tell a network owner to reserve megawatts on a cadastral unit in Sollefteå, Gävle, Falun, Avesta or Arlandastad.

Svenska kraftnät’s 2 September 2026 area table measures TSO-queue applications until reservation and förhandsbesked. The data-centre consumption cells — 1 320 MW in SE2 and 5 169 MW in SE3 — stay in that definition. The table’s own note keeps them before reservation. Treating the heavier SE3 cell as proof that Stockholm has no room, or the lighter SE2 cell as proof that Sundsvall has spare, rewrites the letterhead.

Operator notices measure existence and intention, not the TSO cell. SWE04’s 300 MW sentence is an announced mega-site paper in Sollefteå. EcoDataCenter’s dated Falun megawatt sentences stay Falun operator papers. Microsoft’s three municipalities stay geography. Google Horndal’s groundbreaking stays an operator event. evroc’s land and building-rights purchase stays a land paper. None of those rows is a reserved TSO megawatt on the September table.

The government assignment of 29 May 2026 and ENTSO-E’s April 2025 study measure review work. They do not print a new SE2/SE3 boundary for a hall file. Svenska kraftnät’s indicative capacity map remains context. It is not the property-level connection document.

What named primary sources show

Energimarknadsinspektionen remains the dated source for the four area names. Svenska kraftnät’s 2 September 2026 requested-capacity table remains the dated source for the data-centre application cells, including 1 320 MW in SE2 and 5 169 MW in SE3, and for the sentence that those figures run until reservation and förhandsbesked.

The 29 May 2026 government assignment and ENTSO-E’s 28 April 2025 bidding-zone release remain the dated review sources. Operator pages for Microsoft’s municipalities, Meta Luleå, EcoDataCenter Falun, Google Horndal, atNorth SWE04 and evroc Arlandastad remain geography and intention papers.

No cited source here publishes a reserved megawatt that converts the SE2 or SE3 application cell into a named hall offtake.

A matrix instead of a mid-Sweden slogan

Draw one page with object rows and evidence columns. SE2 price, SE3 price, SE2 application cell, SE3 application cell, operator geography and reserved offtake are different columns. Empty cells stay empty. A heavier SE3 cell is a queue footnote, not a Stockholm refusal.

Do not invent a Sundsvall spare or a Stockholm blockade from the two printed cells. Those remain application megawatts from one dated table.

Common misreads

The first misread is to treat the heavier SE3 cell as proof that Stockholm or Falun cannot connect. The cell is applications before reservation. The second is to treat the lighter SE2 cell as Sundsvall surplus. The third is to subtract the two cells and call the difference spare mid-Sweden power.

The fourth is to move Microsoft’s Gävle–Sandviken geography onto the SE2 cell as reserved megawatts. The fifth is to treat SWE04’s announced 300 MW as the SE2 cell. The sixth is to treat Horndal or Arlandastad as a reserved SE3 offtake. The seventh is to collapse this page into the SE1-versus-SE2 page or the SE3-versus-SE4 page.

What to ask next

Ask which object anyone is quoting — SE2 price, SE3 price, the 1 320 MW cell, the 5 169 MW cell, or a named operator municipality — and which dated letterhead you actually hold. Ask whether a reserved connection paper exists, or only an application cell. Ask whether an announced campus megawatt has been pasted onto the TSO table.

Then open the electricity-areas page, the SE1-versus-SE2 page, the SE3-versus-SE4 page and the grid-connection page. If the operator will not name the paper class, record the gap. Do not fill it from a mid-Sweden slide.

1. Is the live paper a wholesale area or a TSO application cell?
2. Has 1 320 MW been treated as Sundsvall surplus?
3. Has 5 169 MW been treated as a Stockholm blockade?
4. Has SWE04’s 300 MW been pasted onto the SE2 cell?
5. Has a reserved offtake been invented from operator geography?

What this does and does not prove

This page proves that a dated Svenska kraftnät area table already prints data-centre application cells of 1 320 MW in SE2 and 5 169 MW in SE3, and that those cells are not reserved plot megawatts, Sundsvall surplus or a finished redraw of the two mid-Sweden areas.

It does not prove that a named SE2 or SE3 hall can connect next year. It does not invent a leftover by subtracting the cells. It does not treat SWE04’s announced megawatt as the SE2 cell.

Empty cells stay empty. Until a later primary source names a reserved megawatt for a named SE2 or SE3 hall against that table, the siting clock stays a method, not a census.