Editorial review 2026-09-12 · Jurisdiction: Sweden — flood-risk ordinance and Boverket’s PBL flood-tillsyn commentary
A flood-portal colour is not a hall certificate — cycle-3 maps are planning support
Is a Swedish hall already safe or already blocked because a flood-portal tile is blue, or because a dated plan paper has actually tested the plot against a named scenario? Förordningen (2009:956) om översvämningsrisker is Sweden’s Floods Directive paper. MSB’s September 2024 implementation report, used here, records the three-step cycle: identify areas with significant flood risk, publish threat and risk maps, then write risk-management plans. That report says MSB in 2024 identified 26 areas with significant flood risk for cycle 3 (2022–2027), that threat maps are produced by MSB and risk maps by the county administrative boards, and that cycle-3 threat and risk maps sit in the Översvämningsportal from December 2025. The same report uses Falun as a cycle-2 map example. That is a map-example town. It is not a hall flood certificate for EcoDataCenter’s Falun campus page. Myndigheten för civilt försvar’s January 2026 guidance, publication MCF0026, tells the 17 county boards that hold those 26 areas to prepare or revise risk-management plans during 2026–2027. Ordinary consultation lasts at least two months and is preferably placed between November 2026 and April 2027. The plans shall be finished and decided by 22 December 2027 and then apply for 2028–2033. The authority became Myndigheten för civilt försvar on 1 January 2026; the portal page, last reviewed 26 May 2026, still explains that some digital services keep the MSB letterhead while the information stays current. Boverket’s assignment page, retrieved 12 September 2026, says the 2017–2018 tillsynsvägledning is under review through July 2027, with a remiss for new general advice planned for autumn 2027. Boverket’s starting-points table, still published as tillsyn support, puts new cohesive development and socially important activity against a calculated highest flow or level for lakes, watercourses and the sea, and against a 1-in-100 annual probability for cloudburst. Those are supervisory starting points under revision. They are not a campus flood certificate and not a reserved megawatt. Microsoft’s 16 November 2021 note names Gävle, Sandviken and Staffanstorp. Google names Horndal. atNorth’s 23 February 2026 SWE04 note names Sollefteå. evroc’s 3 February 2025 note names Arlandastad. None of those pages is a threat map, a risk-management plan or a detaljplan flood test. The detaljplan hub and the municipal-planning page already own neighbouring papers. This page only asks which published flood object the hall file is sitting on. This is orientation, not counsel.

General orientation. Assess the specific project separately. Applies to: Readers placing a flood-safe, climate-proof or portal-colour sentence on a Swedish hall siting file.
Reading order
Seat the ordinance cycle before any coloured tile. Identification, threat map, risk map and risk-management plan are four different objects. Write the cycle, the map class and the date on one line before you say the hall is safe or blocked.
Then keep the 26-area list on its own row. MSB’s September 2024 paper names a national count for cycle 3. It does not, in the pages used here, publish a hall-by-hall overlay. A town that is not on that list is not automatically dry. A town that is on that list is not automatically a refused campus.
Then keep Boverket’s tillsyn table on its own row. The 1-in-10-000 highest-flow starting point and the 1-in-100 cloudburst starting point are published for länsstyrelsen review of detaljplaner. Boverket says a site-specific assessment is still required. The same authority is rewriting those starting points through July 2027. Do not paste the table onto a groundbreaking slide as a certificate.
Then keep named project rows as geography. Falun in the 2024 paper is a cycle-2 map example. EcoDataCenter’s Falun page is an operator paper. Microsoft, Google Horndal, SWE04 Sollefteå and evroc Arlandastad stay operator papers. This register does not publish a threat-map cell for those cadastral units.
- Identification, maps and the 2028–2033 plan are different objects.
- Twenty-six cycle-3 areas are a national count, not a hall overlay.
- Boverket’s table is tillsyn support under revision.
- A portal colour is planning support, not a certificate.
What a flood map and a tillsyn table actually measure
A threat map measures a modelled inundation under a named scenario. The portal page warns that layers and climate adjustments differ between mappings and that the report for each mapping holds the scenario. A blue tile without that report is an incomplete quote.
A risk map measures consequences inside an identified area. County boards produce those maps. They are not a bygglov and not a reserved feeder.
A risk-management plan measures objectives and measures for 2028–2033 once the county board has decided it. Until 22 December 2027 those cycle-3 plans are still being written. A 2026 groundbreaking slogan cannot sit on a 2028 plan that does not yet exist.
Boverket’s table measures a supervisory starting point for whether a detaljplan may be inappropriate with regard to flood risk. The county board still makes a site-specific assessment. The assignment page records that the starting points are being rewritten because uncertainty grows when horizons pass 2100.
What named primary sources show
Ordinance 2009:956 remains the statutory flood-risk paper. MSB’s September 2024 report remains the dated cycle-3 count, the map split and the Falun cycle-2 example. MCF0026 from January 2026 remains the dated 17-board, two-month-consultation and 22 December 2027 clock. The portal page, last reviewed 26 May 2026, remains the dated map library.
Boverket’s assignment page and the 30 November 2022 tillsyn page remain the dated PBL commentary, including the rewrite through July 2027. Named operator rows remain geography.
No cited source here publishes a project-specific threat-map cell, a decided 2028–2033 plan for a named hall or a reserved hall megawatt created by a portal colour.
A matrix instead of one safe-or-unsafe slogan
Draw one page with object rows and evidence columns. Cycle-3 identification, threat map, risk map, 2028–2033 plan, Boverket tillsyn table and a named campus town are different columns. Empty cells stay empty.
Do not invent a hall flood year or a reserved megawatt from the portal. Those remain planning-support facts from dated MSB, MCF and Boverket pages.
| Object | What a source can show | What this page will not invent |
|---|---|---|
| Cycle-3 areas | MSB Sept 2024: 26 areas with significant flood risk | A hall overlay or a dry-town certificate |
| Threat / risk maps | Portal from Dec 2025; layers differ by mapping report | A bygglov or reserved MW |
| Risk-management plan | MCF Jan 2026: decide by 22 Dec 2027; applies 2028–2033 | A 2026 groundbreaking certificate |
| Boverket table | Tillsyn starting points; rewrite through July 2027 | A campus flood certificate |
| Falun map example | MSB 2024 cycle-2 illustration; EcoDataCenter Falun is an operator page | A hall flood class |
Common misreads
The first misread is to treat a portal colour as a hall certificate. The portal is planning support. The mapping report still names the scenario. The second is to treat the 26-area count as a closed list of wet towns. Identification is one cycle step. The third is to paste Boverket’s 1-in-10-000 or 1-in-100 starting point onto a campus slide while the same authority is rewriting the advice.
The fourth is to treat Falun-the-map-example as EcoDataCenter’s flood class. The fifth is to treat a 2026 groundbreaking as proof that a 2028–2033 plan already exists. The sixth is to invent a hall flood year or a reserved megawatt. The seventh is to collapse this page into the detaljplan hub or the contaminated-land page.
What to ask next
Ask which object anyone is quoting — identification, threat map, risk map, risk-management plan or Boverket tillsyn table — and which mapping report names the scenario. Ask whether the plot sits inside one of the 26 cycle-3 areas, and which dated paper shows that.
Then open the detaljplan hub, the municipal-planning page, the contaminated-land page and the översiktsplan-versus-detaljplan page. If the operator will not name the map class and the date, record the gap. Do not fill it from a blue-tile slide. This page is not legal advice.
- Is the quote an identification, a threat map, a risk map or a 2028–2033 plan?
- Which mapping report names the scenario and the climate adjustment?
- Has Boverket’s tillsyn table been pasted on as a hall certificate?
- Has Falun-the-map-example been treated as a campus flood class?
- Has a 2026 groundbreaking been treated as a decided 2028 plan?
What this does and does not prove
This page proves that Sweden already publishes a flood-risk cycle with 26 cycle-3 areas, portal maps from December 2025 and a 22 December 2027 plan clock, and that Boverket’s tillsyn starting points are under rewrite — and that those facts are not a reserved hall megawatt.
It does not prove that a named Swedish hall is safe or blocked. It does not invent a flood year for Horndal or Arlandastad. It does not treat a campus press note as a threat-map cell.
Empty cells stay empty. Until a later primary source names the map class, the scenario and the dated municipal file, the flood clock stays a method, not a census. DataCenterSweden records the method.
Sources
- Ordinance (2009:956) on flood risks — Sveriges riksdag, 2009-10-15. Checked 2026-09-12.
- MSB — The flood-risk ordinance: Sweden’s implementation of the EU Floods Directive (September 2024) — MSB, 2024-09-01. Checked 2026-09-12.
- Myndigheten för civilt försvar — Risk-management plans in Floods Directive cycle 3 (January 2026) — Myndigheten för civilt försvar, 2026-01-01. Checked 2026-09-12.
- Myndigheten för civilt försvar — Flood portal (Översvämningsportalen) — Myndigheten för civilt försvar, 2026-05-26. Checked 2026-09-12.
- Boverket — Flood risk in spatial planning (assignment page) — Boverket, 2026-01-01. Checked 2026-09-12.
- Boverket — Supervisory guidance on flood risk in detailed planning — Boverket, 2022-11-30. Checked 2026-09-12.
- Boverket — Starting points for assessing flood risk — Boverket, 2022-01-01. Checked 2026-09-12.
- Planning and Building Act (2010:900) — Sveriges riksdag, 2010-07-01. Checked 2026-09-12.
- Microsoft opens its sustainable datacenter region in Sweden — Microsoft, 2021-11-16. Checked 2026-09-10.
- EcoDataCenter 1, Falun — EcoDataCenter, 2026-01-01. Checked 2026-09-10.
- Horndal, Sweden – Google Data Center Location — Google Data Centers, 2026-09-01. Checked 2026-09-11.
- New 300MW mega site in Sollefteå, Sweden — atNorth, 2026-02-23. Checked 2026-09-11.
- evroc finalizes land purchase in Arlandastad, Stockholm — evroc, 2025-02-03. Checked 2026-09-10.