SM-002 (DT-Series System Map, v1.4) identifies Freely Allocatable Capacity (FAC) as the correct risk metric for Finland's electricity system under pre-allocation conditions. The central finding is structural: as PPA penetration grows, spot market prices discover scarcity in the residual, not in the system. Crises do not appear as production shortfalls — they appear as access asymmetry. Consumers with contracted supply continue to receive power; spot buyers face prices that reflect structural stress.
SM-002 also confirms two component findings: Es (domestic buffer capacity) is in structural decline — CHP decommissioning is empirically supported (r=−0.89, p=0.001). Ex (SE1 import buffer) is not structurally declining but is conditionally vulnerable to synchronous collapse during systemic external shocks.
This memo asks: what architecture responds to these findings within a realistic intervention window? The answer is not a new technology — it is a conversion of what already exists.
SM-002 defines FAC as total dispatchable capacity minus PPA-committed capacity, minus must-run baseload, minus scheduled outages. FAC is the share of production that remains freely allocatable at the moment of stress — and it is not currently measured by any public instrument.
A critical correction from external review applies here: municipal ownership does not automatically create FAC. A municipal energy company can sign long-term PPAs, optimise for cash flow, or lock capacity for decades — exactly as a commercial operator can. The distinction is not ownership type. The distinction is contract architecture and operational mandate.
FAC is an institutionally enforced property, not an emergent consequence of public ownership. It requires three explicit conditions to hold:
| Condition | Requirement | Current Finnish status |
|---|---|---|
| (A) Optimisation function change | Node mandate shifts from maximum revenue to: stress availability + reserve eligibility + heat/grid dual role. This changes the objective function, not just the ownership structure. | Implicit in municipal culture but not formalised. Kuopio Energia and peers operate commercially within regulated structures — no explicit availability mandate exists. |
| (B) Regulatory recognition of security of supply | EU state aid rules require a "justified mechanism" framing, not a market exemption. CRMA logic, RED III flexibility provisions, and capacity mechanism normalisation (Germany, France, Italy) create a viable regulatory pathway — but it must be explicitly invoked, not assumed. | EU direction is supportive (energy security > pure market efficiency post-2022) but no Finnish instrument currently formalises this for CHP conversion. |
| (C) PPA pre-emption structurally blocked | If the converted node signs long-term PPAs with industrial buyers, FAC = zero regardless of ownership. The contract constraint — a PPA cap or prohibition on stress-capacity pool participation — must be binding, not advisory. | No such constraint currently exists in Finnish energy law or Fingrid reserve market rules. This is the most critical gap. |
For FAC to be institutionally enforced rather than assumed, the SGFA consortium structure requires four binding elements beyond the SP-002 financial model:
The Finnish regulatory context makes this achievable without new legislation at EU level. The Huoltovarmuuskeskus (HVK) mandate already covers energy security obligations for critical infrastructure. A municipal CHP node designated as a strategic energy asset under the Security of Supply Act (1390/1992) can operate under an availability obligation that de facto caps PPA exposure. This is the existing instrument — SGFA conversion would formalise it for the converted node class.
The EU framing is "justified mechanism," not "market exemption." Capacity mechanisms are now normalised across major EU markets. An SGFA node designated as a resilience capacity instrument — not a market participant seeking exemption — fits within the established state aid framework for security of supply investments. The distinction matters for regulatory durability: a mechanism is robust; an exemption is temporary.
The SGFA conversion argument rests on a concrete physical asset: approximately 15–20 existing municipal biomass-CHP plants in Finnish regional centres. These are not hypothetical investments — they are operating facilities with established grid connections, trained personnel, district heating networks, and regulatory relationships. Haapaniemi 2 in Kuopio is the archetypal case documented in WP-013 and SM-001, but it is one instance of a broader pattern.
| Characteristic | Status | SGFA relevance |
|---|---|---|
| District heating network | Existing — 100,000+ residents per major node | Heat anchor (TN-001 Layer 1). Eliminates greenfield risk for thermal integration. |
| Grid connection | Existing — transmission-connected | VPP participation (FCR/aFRR/mFRR) requires no new connection. Reserve market access from day one. |
| Biomass fuel supply chain | Existing — forest industry residues, municipal waste streams | Chemical storage layer: biogas from organic waste streams enables 72h+ endurance independent of weather and import. |
| Biogenic CO₂ stream | Existing — flue gas from biomass combustion is near-pure CO₂ | PtX integration: CO₂ + electrolytic hydrogen → synthetic methane or methanol. OGAS2 simulator (Kuopio node): CAPEX 120–180 M€, IRR 14–20%. |
| Operating licence and regulatory relationships | Existing | Conversion permit process substantially shorter than greenfield. 2–4 year build timeline achievable. |
| End-of-life status | Haapaniemi 2 type: approaching end-of-life 2030–2035 | Conversion window is open and narrowing. Investment decision before 2028 required for 2032 operation. |
The aggregate effect is significant: 1–3 GW of reserve-eligible balancing capacity (SP-002 §4), available within the 2027–2032 window, built on existing infrastructure, under municipal ownership that preserves FAC.
Sweden is the natural reference case — similar climate, similar market structure, similar nuclear heritage, and Finland's primary SE1 import counterpart. The comparison reveals a structural asymmetry that SM-002's SE1 regime switch framework makes legible.
| Dimension | Sweden | Finland |
|---|---|---|
| Nuclear trajectory | Ringhals 1+2 closed 2019–2020. Six reactors remain, all 1980–1985 vintage, approaching 60-year limit 2040–2045. SMR target 2035 — capacity gap 2035–2045. | OL3 operational 2023. OL1/OL2 extended to 2038. Single-unit concentration risk (1600 MW per outage). No SMR timeline confirmed. |
| Existing biomass-CHP infrastructure | Significant but geographically concentrated in southern Sweden. District heating penetration high but CHP share declining under EU carbon pressure. | 15–20 operating municipal CHP nodes in regional centres (Kuopio, Joensuu, Jyväskylä, Seinäjoki, Oulu and others). Highly distributed. Finnish forest industry residue supply chain mature. |
| SE1 industrialisation pressure | Hybrit (Luleå), Stegra (formerly H2 Green Steel), Nordic data centre pipeline — all drawing from SE1 before Finland. Vattenfall PPA book growing rapidly. SE1 FAC declining. | Receives SE1 exports when available; faces SE1 Competitor regime when SE1 industrial draw exceeds hydro surplus (A.3 state function, SM-002). |
| SGFA deployment potential | Architecture is applicable but greenfield investment required in most locations. Longer permitting timelines. PPA market more mature — commercial pressure to pre-commit capacity. | Conversion of existing assets preferred. Municipal ownership culture stronger. Forest industry CO₂ stream available. 2027–2032 window achievable. |
| Intervention window | New nuclear 2035+. Wind capacity growing but baseload gap widening 2025–2035. | SGFA conversion 2027–2032 — matches the critical endurance gap identified in DA-001 and WP-013. |
WP-002 (Distributed Resilience Design) establishes a three-level resilience framework. SGFA nodes map directly onto this structure in a way that neither large nuclear nor offshore wind can:
| DRD level | Condition | SGFA node behaviour | Nuclear / wind behaviour |
|---|---|---|---|
| L1 — Normal market operation | Spot market allocates normally. Price signal functional. | Node participates in FCR/aFRR/mFRR reserve markets. CHP supplies heat and electricity. PtX absorbs surplus wind. | Normal generation at market price. |
| L2 — Market stress | Price signal distorted. PPA layer pre-empts spot allocation. SM-002 Shift III conditions. | Node continues operating on biogas reserve. Kunnallinen omistus means no PPA pre-commitment — FAC remains available. Municipal customers supplied at avoided-cost price regardless of spot. | Nuclear: output unchanged but revenue accrues to contracted counterparty. Wind: intermittent, cannot guarantee availability. |
| L3 — Grid stress / partial islanding | Transmission disruption. Import unavailable. WP-002 islanding scenario. | Node operates in islanded mode on biogas reserve + local heat supply. 72h+ endurance from chemical storage (WP-001 Black Period threshold). District heating network provides thermal continuity. | Nuclear: requires grid connection to operate. Wind: grid-dependent. Neither provides islanding capability without additional investment. |
The islanding capability is the critical L3 differentiator. A 150–250 MW SGFA node serving a Finnish regional centre can maintain essential services — hospital heat, emergency shelter, water treatment — for 72+ hours without grid connection, using chemical energy stored in the biogas reservoir. This is the DRD property WP-002 identifies as the most critical for national resilience under adversarial or compound stress scenarios, and the one that no market-optimised energy investment is currently providing.
The SGFA conversion is not merely compatible with EU climate regulation — it is architected to exploit it. Three regulatory instruments create direct economic value for converted nodes:
Biomass combustion combined with CO₂ capture and storage (BECCS) produces negative emissions under EU accounting. The biogenic CO₂ stream from existing CHP flue gas is near-pure and requires no separation from fossil CO₂ — it is the optimal BECCS starting point. The EU Carbon Removal Certification Framework (CRCF), entering force 2026–2027, creates a revenue stream for verified carbon removals. A converted node capturing 100,000–300,000 tonnes CO₂/year at EU ETS prices of €60–100/tonne generates €6–30M annual revenue from this stream alone.
The Renewable Energy Directive III mandates sectorintegrering — coupling of heat, electricity, and transport fuel sectors. The SGFA architecture is RED III's reference implementation at municipal scale: CHP core (electricity + heat), PtX layer (surplus electricity → hydrogen → synthetic fuel), CO₂ capture (biogenic carbon for fuel synthesis). A converted node simultaneously satisfies the heating sector (district heat), the electricity sector (VPP balancing), and the transport sector (RFNBO synthetic fuel).
Renewable Fuel of Non-Biological Origin (RFNBO) classification requires electrolysis powered by additional renewable electricity (additionality criterion). An SGFA node using surplus wind electricity to power electrolysis meets this criterion — the PtX output qualifies for EU shipping and aviation mandates with premium pricing. SP-002 §3 documents the financial model: RFNBO eligibility converts the PtX layer from a cost to a revenue source, improving node IRR from 8–12% (stress case) to 14–20% (base case).
DA-001 identifies Finland's pre-shortage phase as 2026–2032. WP-013 establishes that the intervention window is open and narrowing — industrial commitments are being made, but infrastructure is not yet locked. SM-002's FAC analysis adds precision: the window is defined by the rate at which PPA pre-commitment absorbs available capacity. Once the PPA layer absorbs 60–70% of dispatchable generation, the spot market can no longer perform its balancing function.
| Technology | Decision-to-operation timeline | 2027 operational? | 2030 operational? | 2032 operational? |
|---|---|---|---|---|
| New nuclear (SMR) | 15–20 years (licence + construction) | No | No | No — earliest 2035 |
| LDR-50 (Steady Energy) | STUK: design not yet construction-ready. Best case 2032. | No | No | Possible — conditional on 2026 investment decision |
| Offshore wind (large-scale) | 8–12 years (permitting + construction) | No | Partial | Partial — does not address endurance gap |
| SGFA conversion (existing CHP) | 2–4 years (conversion of existing asset) | Yes — if investment decision 2025 | Yes — multiple nodes | Yes — full network at scale |
The timeline asymmetry is decisive. SGFA conversion is not merely cheaper or more distributed than new nuclear — it is the only option that fits within the 2027–2032 window that DA-001 identifies as critical. Every year of delay narrows the window further as existing CHP plants approach end-of-life without replacement and PPA penetration grows.
The argument here is not against nuclear energy. Nuclear is well-suited to the 2035–2070 energy system as baseload, low-carbon, weather-independent generation. The argument is about temporal fit: nuclear cannot serve the 2027–2032 endurance gap regardless of political will, regulatory speed, or financing availability. This is a structural constraint, not a political one.
Three nuclear-specific risk factors compound the timing problem for Finland and the wider Nordic system:
Unit concentration risk (OL3). Olkiluoto 3 is a 1,600 MW single-unit EPR — Euroopan's largest operating reactor. At Finnish winter demand of 8,000–13,000 MW, a single OL3 maintenance outage removes 12–20% of system capacity for 50 days. The original plan for two smaller units (OL3 + OL4) would have halved this exposure; OL4's cancellation in 2015 — driven by financing failure, not political opposition — made the concentration risk permanent. WEM §07 documents the 2026 maintenance calendar: OL1 offline 19 April–13 June (55 days), OL3 offline 10 September–30 October (50 days). Each outage is a direct Es shock visible in the WEM Firm Share metric.
Fuel supplier transition risk (Loviisa). Loviisa's two VVER-440 reactors — 507 MW each, commissioned 1977 and 1981 — are undergoing fuel supplier transition from Russian TVEL to Westinghouse. The first Westinghouse fuel batch was loaded in 2024; the TVEL contract runs to 2027 (Loviisa 1) and 2030 (Loviisa 2). A VVER-440 fuel supplier change is a multi-year regulatory and technical process: the reactor was designed around Soviet fuel geometry, and mixed-core operation (TVEL + Westinghouse assemblies simultaneously) requires independent neutronics validation for each campaign. Bulgaria's Kozloduy and Ukraine's fleet completed similar transitions — each took the better part of a decade. Loviisa's extended operating licence runs to 2050, but the fuel transition introduces operational uncertainty through 2030 that is not present in standard Western reactor operation.
Nordic nuclear age cliff (Sweden). Sweden's six remaining reactors — Ringhals 3 and 4, Forsmark 1, 2 and 3, Oskarshamn 3 — were all commissioned 1980–1985. They are approaching 60-year operational limits in the 2040–2045 window. Vattenfall's SMR programme targets first unit operation circa 2035, leaving a capacity gap of approximately a decade between retirement and replacement. During this gap, SE1's export capacity to Finland — already under pressure from Hybrit, Stegra, and Nordic data centre demand — faces structural reduction. SM-002's SE1 Competitor regime (A.3 state function) becomes the default rather than the exception.
The SGFA network addresses three supply security dimensions that centralised generation cannot:
Transmission independence. A node serving its local district heating network and emergency load operates independently of long-distance transmission. Finland's 5.2 billion euro transmission investment programme (2025–2035) is driven partly by the need to balance a system whose dispatchable capacity is concentrated in fewer, larger plants. Distributed SGFA nodes reduce this requirement — each converted node is one fewer balancing problem for the transmission system.
Fuel supply diversity. Forest industry residues, municipal organic waste, and agricultural biogas are domestic fuel sources with no geopolitical exposure. The Loviisa nuclear fuel supplier transition — from TVEL to Westinghouse — illustrates the vulnerability of centralised supply chains to geopolitical disruption. A distributed biogas network has no single point of supply failure.
Demand-side anchor. District heating is the largest controllable thermal load in Finnish cities. A SGFA node operating the district heat network can provide demand response at scale — not by curtailing households, but by adjusting the heat-electricity balance of the CHP unit. This is the flexible load that DT-004 identifies as absent in the data centre investment wave: SGFA provides controllable load, data centres provide rigid load. They are structural complements, not competitors.
This memo makes a structural argument, not a project proposal. Three limitations are explicit:
Biomass is not unlimited. The EU RED III sustainability criteria constrain large-scale biomass use. The SGFA model is not a biomass-maximisation programme — it uses existing forest industry residue streams and municipal organic waste that would otherwise be under-utilised. The scale ceiling is approximately 3–4 GW nationally before competing with forest industry raw material demand. The 15–20 node network described here is well within that ceiling.
FAC is not yet measured. SM-002's central finding — that FAC is the correct metric — is confirmed but FAC itself remains unmeasured in public data. SM-003's claim that SGFA conversion grows FAC rests on the ownership and incentive argument (§2), not on observed data. Fingrid's reserve market data provides a partial proxy; direct FAC measurement requires PPA registry data that does not currently exist.
Coordination is the binding constraint. The 15–20 nodes exist as individual municipal assets, not as a coordinated network. SP-002's SGFA Holding Oy consortium structure is the required institutional innovation — not the technology. The reason this capacity has not been converted already is not technical feasibility but coordination deficit: no institution owns the complete picture, as SM-001 documents for the compound stress configuration more broadly.