Artelys Svk MARI Impact Assessment (2026)

Title: Impact assessment of Sweden’s accession to the European balancing platform MARI (Swedish file name: Konsekvensanalys av anslutningen till MARI) Author: Artelys (Paul Brière, Jimmy Brossier, Michaël Gabay, Renaud Saltet), on behalf of Svenska kraftnät Language: English. Note: the euro sign is lost in the text extract; the PDF shows € throughout.

A simulation study of what MARI (the European mFRR energy activation platform) would mean for Sweden. Artelys developed both the MARI and the Nordic clearing algorithms, and the study is commissioned by Svk, so it is a TSO-commissioned model result, not an independent evaluation. The figures are simulation outputs under stated assumptions, not observed outcomes.

Method

• Three configurations for Sweden: (1) a stand-alone Swedish mFRR energy activation market, a counterfactual that has never existed; (2) the Nordic mFRR EAM, live since 4 March 2025, which is Sweden’s current situation; (3) MARI, a prospective scenario with all Nordic countries and every country operational in MARI as of December 2025.
• Period and data: 4 March 2025 to 3 March 2026, 15-minute market time units (34,708 simulated, about 300 with missing data), using historical TSO demands and BSP bids. Belgium (joined 21 May 2025) and the Netherlands (2 December 2025) are excluded before their go-lives. Poland, expected to join in December 2026, is modelled only as a provider of transfer capacity, with no bids or demand.
• Cross-border capacity: an indicative 400 MW of HVDC capacity between the Nordic countries and continental Europe open to mFRR activation (a figure Svk gave and is still evaluating), split across cables in proportion to capacity. Baltic Cable is excluded, on Svk’s recommendation. The study says this assumption limits the benefit of coupling the markets. No HVDC ramping or polarity-reversal limits are modelled.
• Algorithm: the MARI clearing function, with tolerance bands disabled in the Nordic runs so all cases are comparable; direct activation, aggregated areas and net technical profiles are not modelled; elastic demand is not modelled either (Svk’s TSOs did not use it before June 2026). Capacity is ATC-based, not flow-based.
• Price conventions: the report states the TSO’s willingness to pay as ±15,000 €/MWh (demand side, inelastic) and the MARI bid price cap as ±10,000 €/MWh in different sections; both appear as printed.

Results

• The 459 M€ is not the value of the 2025 Nordic go-live. Sweden has been in Nordic balancing cooperation since 2002, so a large part of the benefit already existed before March 2025; the stand-alone case is only a reference.
• Distribution of surplus: the TSO’s demand-side surplus rises; BSPs’ offer surplus falls (fewer scarcity events, more competition); internal Swedish congestion rent falls and external congestion rent rises.
• Prices: larger markets bring activation prices closer to zero. Upward price reductions exceed 390 €/MWh in SE3 and SE4 and are below 65 €/MWh in SE1 and SE2; the report says this reflects structural scarcity of upward resources in southern Sweden. About 1.4% of the periods account for 25% of the 505 €/MWh SE4 upward reduction. Downward effects are small and mixed. Going from Nordic to MARI has minimal further effect on average prices; upward price dispersion narrows mostly in the step to the Nordic market, while downward volatility rises slightly in MARI.
• Where balancing comes from: 30% of needs are met by netting between Swedish zones in every configuration. For upward demand in the Nordic and MARI cases, Swedish BSPs cover 22% and imports 45% (44% in MARI), mostly from other Nordic countries. For downward demand, neighbours cover 22% (Nordic) and 29% (MARI). In SE3 and SE4, upward needs are met 25–28% locally and the rest by imports from Denmark, Norway and SE2; downward, SE2 supplies 71% (Nordic) and 61% (MARI), with Poland 9% in MARI.
• Interconnectors to Poland and Lithuania: the balancing capacity open on SE4–PL and SE4–LT is fully used in more than 50% of the period, but they supply only about 2% of Sweden’s upward and 8% of its downward activation (6% Poland, 2% Lithuania), because of the small capacities. SE4–PL supports SE4’s balancing need in about 70% of periods and SE4–LT in about 55%.
• Price spread mitigation (PSM): the Nordic feature that limits price divergence between areas is worth up to 2 M€ in surplus, so it is minor next to the market-size effect. It cuts the number of extreme-price quarter-hours and cuts unforeseeable rejected bids by up to 37 GWh, but joining MARI reduces extreme prices more than PSM does. The report concludes MARI accession would be beneficial even without PSM, and notes PSM may have benefits, such as price interpretability, that were not assessed.

Limits stated by the report

Historical data, the 400 MW assumption, the exclusion of Belgium and the Netherlands early in the period, Poland as a capacity provider only, no tolerance band and no elastic demand. The authors say the simplifications may affect the numbers but not the qualitative conclusions.

Timeline note

The report says accession of the Nordic countries to MARI is planned for Q1 2027. That has since been postponed to an estimated September–October 2027 (Source - Svk Forandrad Tidtabell MARI (2026)); the report predates or does not reflect that notice.

Relevance to wiki topics

• Nordic Balancing Model — quantitative case for the mFRR platform steps
• Balancing Markets — MARI and Nordic mFRR EAM
• Swedish Balancing Market Prices and Volumes — scarcity and price context in the southern zones
• Svenska kraftnät — the commissioning TSO; its 400 MW capacity assumption and the Baltic Cable exclusion