Working Paper No. 013 · Domain D-1 · D-3 · D-4
Scenario A: nothing changes. Scenario B: three corrections initiated. The central finding is not technical — it is temporal.
ACI's diagnostic work identifies a weakness common to most warning analyses: they describe the problem but leave the decision-maker without a reference point for comparison. A warning without an alternative scenario is difficult to act on — it can be acknowledged, filed, and deferred.
This paper addresses that gap directly. Two scenarios are constructed for the same point in time — 2030 — under two different institutional conditions. The comparison makes the cost of delay visible in concrete terms rather than abstract warnings.
Two conditions govern the construction of both scenarios.
First: institutional delay is treated as a structural constant, not a variable. Finland's historical energy policy decision horizon — from identified need to implemented change — is 3–8 years. OL3 (Olkiluoto Unit 3) required two decades from decision to operation. The Loviisa fuel supplier change took years despite being a familiar process. Electricity market law changes typically require 3–5 years from proposal to entry into force. This is not a criticism; it is a structural feature of a governance system designed for consensus, legal certainty, and democratic participation. The scenarios take it as given.
Second: both scenarios use the same stress test. SP-001's calibrated P99 event: 72 hours of low-wind conditions, outdoor temperature −15°C, data centre load at full capacity. This is a rare but not exceptional winter configuration. It occurs with low frequency but non-negligible probability in any given year.
Both scenarios depart from the same observed conditions.
Electricity consumption has grown from 83 TWh toward a projected 100–110 TWh as data centres and electrification of heating expand demand. Wind power now supplies over 50% of domestic production in annual terms but is fully weather-dependent. Hourly price volatility is among the highest in the EU. Dispatchable reserve has not grown proportionally to demand. Battery storage covers approximately 9.5 minutes of Black Period energy requirements.
District heating prices have risen 32% since 2020. Electricity transmission fees are rising as infrastructure investment programmes are passed through to consumers. Energy poverty affects 7–15% of households by multi-indicator measures. For the lowest income quintile, necessity costs — housing, energy, food — already consume approximately 57% of net income.
Sovereign debt interest payments are 3.5 billion euros and rising. Defence expenditure is 6.7 billion euros and committed to grow substantially. The budget deficit is 12 billion euros. Fiscal space is historically narrow.
Kuopio Energy's SMR project is in pre-feasibility phase. No investment decision has been made, no plant supplier selected, no contingency plan published. Haapaniemi 2 approaches operational end-of-life in 2035. Fingrid reports adequacy in megawatts. Not in megawatt-hours. No institution owns the complete picture.
Scenario A describes the trajectory if institutional action continues at historical tempo. This is not a scenario of failure or negligence — it is the expected outcome of rational sectoral governance in the absence of coordinating intervention.
Wind capacity has grown as planned. Annual production balance approaches self-sufficiency when measured at year-level aggregation — this is the figure most commonly reported. At hourly resolution, the University of Oulu (AIMS Energy 2025) finds Finland electricity-negative 61% of annual hours.
Dispatchable reserve has not grown proportionally. Global supplier backlogs have extended delivery timelines; equipment ordered in 2026 is not fully operational by 2030. Kuopio Energy's SMR is, in the best case, in licensing review awaiting a parliamentary in-principle decision. Haapaniemi 2 approaches end-of-life. Electric boilers are the likely interim measure — adding thermal load to the electricity system precisely at winter peak.
System endurance under SP-001's stress test: approximately 38 hours at current reserve extension rate. The 72-hour threshold has not been reached.
An anticyclone settles over Scandinavia. Wind output falls below 10% of installed capacity. Outdoor temperature reaches −17°C for three consecutive days. Data centres draw at full capacity under contractual priority arrangements.
Spot prices rise. In the first hours, 50–100 €/MWh. As reserves approach limits and import capacity saturates — the same anticyclone affects neighbouring systems — prices reach 200–500 €/MWh for a period of hours. Households on spot-indexed contracts face heating bills that are multiples of normal. Households on district heating face no immediate bill shock — their annual statement arrives months later.
For the lowest income quintile, necessity costs consume 65–70% of net income. Fixed-price electricity contracts carry 15–18 cent/kWh risk premiums. District heating bills are 30–50% higher than 2025. Real wages are flat or marginally negative in real terms. No effective automatic protection mechanism exists for district heating captive customers.
DA-001 trajectory: Concern moving toward Danger. Signals S1–S5 all active. The intervention window has narrowed substantially but has not closed. Political pressure is growing but institutional response follows the crisis rather than anticipating it.
Scenario B describes the trajectory if three identified corrections are initiated in 2026–2027. They are not complete by 2030 — institutional delay is a constant, not a variable. But they have changed the system's trajectory and diagnostic state.
Same anticyclone. Same temperature. Same data centre load. But the system's starting position is different.
Endurance under the stress test: approximately 52 hours. Industrial demand flexibility is active for 36 hours. The critical window — when the system is most exposed — is compressed from 72 hours to approximately 20 hours at the tail.
The price spike still occurs. The 72-hour event still exceeds the system's reserve duration. But the severity and duration of the most acute phase is substantially reduced. Social and economic costs are lower — not zero.
Household energy costs are identical to Scenario A. The three corrections address system endurance and institutional transparency — not the distributional asymmetry. DA-007 identifies the instrument gap that Scenario B's corrections do not fill. District heating captive customers are not specifically protected by any of the three corrections.
This is the critical asterisk in Scenario B: system-level resilience has improved. Household-level protection has not changed. These require different instruments.
DA-001 trajectory: Concern — movement toward Danger has slowed. Signals S1 and S2 remain active; S3 has weakened as diagnostic transparency creates political accountability; S4 and S5 unchanged. The intervention window is wider than in Scenario A. The structural problem has not been solved — the conditions for solving it have been created.
| Indicator | Scenario A | Scenario B |
|---|---|---|
| System endurance (hours) | ~38 | ~52 |
| P99 critical hours/year (est.) | 6.5 | ~3–4 |
| Political awareness of endurance gap | Low | Moderate — indexed, public |
| Household protection (district heating) | Unchanged | Unchanged * |
| Contingency plans published | Absent | Mandatory — major operators |
| Intervention window | Narrow | Wider |
| Dispatchable capacity pipeline | Slow | Active — not yet operational |
* The asterisk is the central diagnostic gap: the three corrections improve system-level resilience but do not address household distributional exposure. Household-level protection requires DA-007's instrument conditions — a separate policy action not covered by these three corrections.
Institutional delay is not a governance failure. It is a structural property of systems designed for democratic accountability, legal certainty, and consensus. It becomes a risk when the physical window for low-cost intervention closes faster than the institutional decision cycle allows.
Finland's energy policy history is consistent on this point. The decision to build OL3 was made in 2002; the reactor reached commercial operation in 2023 — twenty-one years. The Loviisa fuel supplier change was a known process that took years. Haapaniemi 2's replacement requirement has been visible for years; no alternative is operationally confirmed.
WP-003's Institutional Termination Time framework identifies the point at which institutional response latency exceeds the remaining system adjustment horizon. When that ratio reaches zero, governance loses causal influence over outcomes regardless of whether resources, authority, or analytical knowledge remain intact.
Three gaps remain unaddressed by this paper.
The first is household-level protection. The comparison table's asterisk — that neither scenario changes household distributional exposure — is the central unresolved problem. DA-007 specifies the structural conditions for an effective instrument. The instrument itself has not been designed or proposed. That requires a political actor with ownership of the problem.
The second is the behavioural model. Households in this paper are passive — they absorb costs but do not adapt. This is analytically tractable but probably wrong. People move, reduce consumption, change heating systems, vote differently. A dynamic model of household response would change both scenarios.
The third is political agency. Who initiates the three corrections? Who calls Fingrid's leadership? Who writes the Energy Authority's guidance letter? Who schedules the cross-ministry coordination? WP-013 identifies what needs to happen. It does not identify who makes it happen. That is DA-008's question.
Note: The Regional State Administrative Agencies (AVI) were dissolved on 1 January 2026. Their relevant functions transferred to the new National Licensing and Supervisory Authority (Luova). This map reflects the post-reform institutional landscape.
●●● owner · ●● supporter/alternative · ● influencer
| Step | Action | Timeline |
|---|---|---|
| 1 | Fingrid board or CEO decides to publish endurance index alongside conventional capacity reports | 3–6 months · K1 initiated |
| 2 | TEM issues guidance letter to Fingrid, Energy Authority, and Luova on endurance reporting, reserve contract duration requirements, and contingency plan obligations | Concurrent · K1 confirmed, K2 and K3 mandated |
| 3 | Energy Authority issues interpretation on district heating operator reporting — contingency plan obligation added to annual reporting requirements, coordinated with Luova | 6–12 months · K3 initiated |
| 4 | Fingrid's next contracting cycle includes duration-tier requirement for demand flexibility — graduated structure (0–6h / 6–24h / 24–48h) with corresponding price differentiation | 12–18 months · K2 initiated |
Four actions. Three institutions. No new legislation required.
Realistic timeline if initiated in 2026: K1 visible 2027 · K3 visible 2027–2028 · K2 visible 2028–2029.
Full effect: 2028–2030.
This paper identifies what needs to happen and names who must act. It does not — cannot — be the actor that makes it happen.
ACI operates as a diagnostic institution: it identifies, names, and publishes. It does not advocate, lobby, or implement. This position has protected the analytical credibility of the work. It is also a structural limit that this paper must name explicitly.
The three corrections described above are low-cost, technically straightforward, and address a documented gap. They have not happened because no institution owns the compound problem. ACI can document the gap. It cannot fill it. Filling the gap requires a political actor willing to take ownership of a problem that currently has no institutional home.
An analysis that identifies the coordination gap but cannot produce the coordinator faces its own version of the ITT condition: the diagnostic is complete, the action window is open, and the causal link between diagnosis and action depends on a variable — political agency — that the diagnostic institution cannot supply.
This annex is included not as an admission of failure but as an honest accounting of what a diagnostic institution can and cannot do. The value of the diagnosis does not depend on ACI being the one to act on it. It depends on someone acting on it.