Lind et al TSO-DSO Coordination Meshed-to-Meshed (2023)

Lind, L., Cossent, R., Frías, P. — “TSO-DSO Coordination for the Procurement of Balancing and Congestion Management Services: Assessment of a meshed-to-meshed topology.” Peer-reviewed academic paper (Universidad Pontificia Comillas, IIT), published in Sustainable Energy, Grids and Networks (ScienceDirect, 2023); open-access preprint on arXiv (2209.02360). Funded by the EU Horizon 2020 CoordiNet project (grant 824414), with input from Vattenfall Eldistribution.

Journal: Sustainable Energy, Grids and Networks (ScienceDirect, paywalled)

Preprint: arXiv:2209.02360 (open access)

Case study: Swedish power system, 1 year, Nordic 32 transmission model + 70 kV subtransmission grid of Uppsala

What the paper does

Proposes a techno-economic dispatch model (day-ahead → congestion management → balancing, as sequential Mixed Integer Programming problems in GAMS) to compare four Coordination Schemes (CSs) for how a TSO and a subtransmission-operating DSO jointly procure balancing and congestion-management flexibility from DERs, explicitly modelling a meshed-to-meshed topology — i.e. both the transmission grid and the DSO’s subtransmission grid are meshed (not the textbook radial-distribution assumption), with multiple TSO-DSO interface substations between them. This matters because, per the paper’s own literature review (citing Eurelectric, 2013), 19 EU member states have DSOs operating at 110 kV or higher, versus only 6 where the DSO is confined to LV/MV assets — the meshed-to-meshed case is the European norm, not an edge case.

The four CSs tested:

• Common (joint or separate balancing/CM) — a single market clears both TSO and DSO needs together
• Multi-level (OPF-based or PTDF-based LFM) — the DSO runs its own local congestion market first; unused bids pass upward to the TSO market

The PTDF-based Multi-level LFM is explicitly built to replicate a real Swedish implementation — the paper cites Ruwaida et al. (2022), “TSO-DSO-Customer coordination for purchasing flexibility system services: Challenges and lessons learned from a demonstration in Sweden” (IEEE Trans. Power Systems), as the empirical basis.

The case study — and its direct link to this wiki

The subtransmission network modelled is the 70 kV Uppsala grid, one of the demonstration sites of the Swedish CoordiNet project — the same demonstration already documented in CoordiNet and TSO-DSO Coordination — The Central Design Problem. The paper explicitly frames the underlying real-world mechanic: “In Sweden, the regional DSO operates the subtransmission network and is already faced with the need for procuring distributed flexibility, as they are subject to subscription limitations at their interfaces with the TSO.” This is the same subscription mechanism (abonnemang mot överliggande nät) the wiki already documents — the paper calls it a “virtual congestion” (a regulatory power-flow limit at the substation, distinct from a thermal/physical constraint) and shows it functions as a cost-shifting mechanism from the TSO to the DSO: in the “no-flexibility” scenario, the DSO’s entire modelled cost (1,674 k€/year) is subscription-penalty exposure, not physical grid reinforcement.

TSO-DSO interface detail (Fig. 2 of the paper; the transformer ratings and subscription levels are in the figure image, not in the arXiv text, and rest on the Ruwaida et al. source below): the Uppsala 70 kV subtransmission grid connects to the transmission grid via two interfaces — 2×315 MVA transformers (205 MW subscription level) and a 250 MVA transformer (87 MW subscription level) — illustrating concretely what a “meshed-to-meshed, multiple-interface” TSO-DSO boundary looks like in practice, as opposed to the single-substation boundary the wiki’s Electric Grid Structure diagram simplifies to.

FSPs modelled: 8 flexibility service providers connected to the Uppsala grid, drawn directly from CoordiNet’s actual Swedish demonstration roster — a 20 MWh/5 MW battery, district heating, office/commercial/multi-family buildings, and two industrial loads, bidding 8–20 €/MWh.

Key quantified findings

• The Common CS (single joint market) produces the lowest overall cost of flexibility procurement across TSO+DSO — confirming prior literature; Multi-level (sequential DSO-then-TSO) markets cost more in aggregate because the DSO’s LFM activation isn’t always optimal from the TSO’s perspective
• The DSO’s own cost drops sharply when local flexibility is available: from 1,674 k€/year (no-flexibility, all subscription penalties) to 802 k€/year (PTDF-based LFM base case) — a 52% reduction
• Scalability: scaling the Uppsala FSPs’ available capacity to just 1.6× the base case eliminates subscription penalties entirely for the DSO
• Demand growth deferral: a 10 GWh/year increase in activated local flexibility allows the grid to absorb 145 GWh/year of additional demand without triggering non-served-flexibility events — i.e. local flexibility can defer subtransmission reinforcement in the face of load growth
• Distributed generation effect: adding two wind farms to the DSO’s grid (replication scenario) cut the DSO’s local-market cost by up to 98% — DG at the subtransmission level offsets import needs through the TSO-DSO interface, sharply reducing subscription-penalty exposure

Relevance to existing wiki topics

• Electric Grid Structure — grounds the wiki’s abstract “TSO/DSO boundary is a substation, not a fixed voltage” claim in a concrete, quantified Swedish case with multiple substations at one interface, directly relevant to the ongoing Dir. 2026:83 two-vs-three-level and Svk-takeover question
• CoordiNet — the Uppsala demonstration this paper models is the same one already documented there; this source adds the academic/quantified layer on top of CoordiNet’s own final report
• TSO-DSO Coordination — The Central Design Problem — direct evidence for the “multi-level DSO complication” and “chicken race” sections: this paper quantifies exactly what the subscription-penalty mechanism costs a subtransmission DSO, and how much local flexibility offsets it
• Svenska kraftnät — the paper’s finding that virtual congestion functions as a TSO→DSO cost-shifting mechanism bears directly on any future debate about Svk absorbing regionnät assets
• Source - Ruwaida et al TSO-DSO-Customer Coordination Swedish Demonstration (2022) — the real-world Swedish LFM implementation this paper’s PTDF-based model replicates, now ingested; confirms this paper’s Uppsala substation figures (205/87 MW) from the primary source