SM-006 — The 2028 Convergence Window:
ENTSO-E Empirical Validation of Structural Stress

Purpose and Methodological Note

The ACI synthesis series (SM-001 through SM-005) established Finland's compound energy risk through structural analysis: declining dispatchable capacity (CHP phase-out, DT-002), growing rigid inelastic load (data centres, DT-004), structural import dependence on SE1, and the 2027–2032 intervention window identified in DA-001. These findings were derived from institutional data, market structure analysis, and physical system modelling.

This memo adds a distinct evidential layer: empirical validation from live operational data. The ENTSO-E Transparency Platform provides hourly cross-border physical flow records for BZN|FI for the full year 2026, together with generation adequacy margins and water reservoir levels. The central finding of SM-006 is methodological as much as analytical: the structural thesis was formed before the data was examined, and the data confirms it.

The Four Empirical Findings

The ENTSO-E dataset (87,600 hourly flow records, ten bilateral corridors, full year 2026) yields four findings that directly map to the SM series structural claims.

Finding 1 — Capacity margin: the system is already at the boundary

The ENTSO-E Generation Adequacy forecast for BZN|FI 2026 records a capacity margin of −3,300 MW. This is not a projected future shortfall — it is the current operational state of the Finnish power system: installed and available domestic generation is insufficient to cover forecast peak demand without import.

The same dataset records the single highest SE1+SE3→FI import hour in 2026 at 3,298 MW (31 January, 11:00). The two figures — the adequacy gap and the peak import flow — are numerically identical. The system is running at its physical import limit during peak stress. There is no remaining margin.

Finding 2 — Hydro storage: the primary buffer is depleted seasonally

Water reservoir storage (BZN|FI) fell from approximately 3,600,000 MWh at the start of 2026 to a seasonal minimum of approximately 2,100,000 MWh around weeks 14–16 (April). The drawdown of ~1,500,000 MWh over the winter period represents aggressive utilisation of Finland's primary dispatchable storage resource.

This is the physical expression of the adequacy gap: in the absence of sufficient import capacity to fully cover the −3,300 MW deficit on every hour, the system uses its hydro reservoir as a buffer. By April, that buffer is substantially depleted. Recovery requires spring snowmelt and rainfall — a weather-dependent process outside system control.

Finding 3 — Import dependency: structural and directional

Three findings are embedded in this table. First, Finland is a structural importer from Sweden: SE1 and SE3 together provide +4,771 GWh net over the year, with no month of reverse flow during winter. Second, Finland simultaneously exports to Estonia: −1,699 GWh net, meaning Finland serves as a transit node — receiving power from Sweden and forwarding it south. Third, and most structurally significant: Russia is zero. Every gigawatt-hour that Russian interconnection previously provided — approximately 10% of Finnish consumption at peak — has been permanently removed from the supply balance without equivalent domestic replacement.

Finding 4 — Estonian oil shale: a conditional and seasonal resource

The EE→FI flow is not uniformly directional. While Finland exports to Estonia on a net annual basis (−1,699 GWh), Estonia's Eesti Energia oil shale (põlevkivi) capacity produces reverse flows during specific market conditions: the peak recorded EE→FI hours in 2026 reached 998 MW in March and April. These flows do not occur during the deep winter peak (January–February mean: EE→FI max 217–356 MW); they occur in the shoulder season when Estonian thermal capacity is still warm and Finnish import demand has partially softened.

This confirms that oil shale does not serve as a reliable winter peak resource for Finland. It functions as a marginal hour arbitrage asset: when Estonian thermal capacity is running and Finnish prices are briefly elevated, the corridor activates. The winter peak — when Finland most needs external supply — is served by SE1 hydro and wind, not by Baltic thermal generation.

The Convergence Structure: 2027–2030

Each of the four empirical findings describes the 2026 state. The convergence thesis concerns what happens next. Four simultaneous structural changes — two on the Finnish demand side, two on the SE1 supply side — are scheduled to materialise within the same 24-month window.

The convergence is not a sequence of independent shocks. It is a structural transformation in which the two sides of the Finnish adequacy equation move simultaneously in the wrong direction: domestic supply falls, and the external buffer that currently compensates also falls. The −3,300 MW margin that already leaves no room becomes a deeper deficit at exactly the moment the compensating import capacity is absorbed by SE1 industrialisation.

Why 2028 Is the Convergence Point

The designation "2028 Convergence Window" is not a precise forecast — it is a structural inference from the alignment of four independent timelines. The reasoning is as follows.

By 2028, Stegra will be ramping toward full capacity, HYBRIT commercial volumes will be growing, the largest Finnish data centre installations from the current pipeline will be drawing from the grid, and the CHP units approaching end-of-life will have entered their final operational years without committed replacement. The SE1 surplus that currently provides the 3,298 MW peak import capacity will be competing with 3–4 GW of new industrial demand in the same bidding area.

The first severe winter — climatically defined as an extended cold and low-wind period lasting 10+ days — that occurs after 2027 will find the Nordic system in a structurally different position from the winters documented in this analysis. The buffer that currently just barely covers the −3,300 MW Finnish gap will no longer be available at the required scale.

Finland's Production Profile as SGFA/MESA Foundation

A secondary analytical finding from examining the ENTSO-E installed capacity data (BZN|FI 2026) is that Finland's production profile is structurally well-suited to the SGFA/MESA architecture described in SM-003 and SP-002 — not as coincidence, but as a structural consequence of the same features that create the risk.

Wind onshore at approximately 9,000 MW installed capacity is the largest single resource by nameplate — but it is weather-dependent and creates the variability that requires balancing. Nuclear at ~4,200 MW provides the stable baseload foundation. Hydro run-of-river and reservoir at approximately ~3,400 MW combined provides the primary dispatchable flexibility layer, now confirmed to be drawing down its storage reservoir over winter. Biomass at approximately ~2,000 MW provides the most reliable, weather-independent, continuously dispatchable generation in the portfolio.

The SGFA architecture targets precisely the biomass-CHP layer: converting existing municipal nodes from single-output heat-and-power facilities to multi-layer resilience assets with chemical storage, PtX capability, and VPP integration. The adequacy data confirms why this layer is critical: it is the only dispatchable resource that is neither weather-dependent (wind), import-dependent (SE1), nor subject to planned outage cycles (nuclear). In a system running at the edge of its import capacity, the biomass-CHP layer is the only domestic resource with genuine flexibility under peak stress conditions.

Revised Assessment: SM-001 Thesis Confirmed

SM-001 (Finland 2030 — A Compound Risk Diagnosis) identified five simultaneous structural pressures converging toward a compound risk configuration. The energy system pressure — the first and most time-sensitive — was characterised as follows: endurance capacity in structural decline, rigid inelastic load growing, import dependency on a single external buffer (SE1), and the buffer itself under growing internal SE1 competitive pressure.

The ENTSO-E 2026 operational data confirms each element of this characterisation with live numerical evidence:

"A thesis formed before examining the data, subsequently confirmed by independent operational evidence, occupies a structurally stronger epistemic position than post-hoc rationalisation of observed data."

This distinction matters for ACI's institutional position. The institute's methodology — diagnostic-first, structural reasoning before data retrieval — has produced a confirmed prediction. The ENTSO-E evidence does not modify the thesis. It validates it.

What SM-006 Does Not Claim

Three limitations are explicit. First, this analysis does not establish that a supply crisis will occur in 2028. It establishes that the structural conditions for such a crisis are forming, that the 2026 operational state is already at the boundary, and that four concurrent changes scheduled for 2027–2030 make the boundary conditions more adverse. Whether a crisis materialises depends on weather — specifically, on whether an extended cold and low-wind event coincides with the convergence window.

Second, the Stegra and HYBRIT electricity consumption figures used here are planning-phase estimates. Both projects have experienced schedule delays (Stegra's original late-2025 production deadline passed without production start). Delay is more likely than acceleration for first-of-kind industrial projects at this scale. Delay is partially mitigating — each year of delay extends the window before SE1 surplus compression occurs. However, delay does not remove the structural convergence; it defers it.

Third, this memo does not account for potential SE1 generation capacity additions in the same period: new Swedish nuclear (SMR target 2035+), additional wind, or demand response from Stegra and HYBRIT themselves (both have incentive to reduce consumption during high-price hours). These factors are real and partially offsetting. The SE1 system is not passive. The claim here is about the structural direction and the 2027–2032 window, not about a deterministic outcome.

Decision Track Implications

The empirical findings of SM-006 bear directly on four active ACI Decision Tracks.

DT-001 (Capacity Mechanism — Finland) gains urgency. The −3,300 MW adequacy gap is not a modelled projection — it is the documented 2026 operational state. A Finnish capacity mechanism designed to close this gap must be sized for a deficit that is already present, not one that may emerge. Delays in mechanism design extend the period during which the system operates without structural margin.

DT-002 (CHP Phase-out — Finland) gains specificity. The biomass-CHP layer is confirmed as the only fully dispatchable, weather-independent domestic resource in Finland's current generation portfolio. Its phase-out removes the system's primary flexibility reserve at exactly the moment that SE1 import capacity is contracting due to industrial demand growth. The replacement investment timeline (SGFA conversion: 2–4 years per node) must begin before 2027 to maintain coverage through 2030.

DT-003 (Hydrological Monitoring) gains empirical grounding. The 1,500,000 MWh winter drawdown confirms that hydro reservoir state is a material operational variable for system adequacy, not merely a hydrological indicator. Real-time reservoir monitoring integrated into WEM §12 provides actionable advance warning for the adequacy gap before it becomes acute.

DT-004 (Data Centre Grid Connection Terms — Finland) gains structural context. The −3,300 MW gap exists before the data centre pipeline adds its 2–3 GW of rigid inelastic load. Connecting data centres to a system that is already at its adequacy boundary without demand-response obligations compounds a deficit that is already fully loaded.