Source - Energiforsk 2024-1043 DNDP Analys och Flexibilitet (2024)

Full citation: Hartvigsson, E. & Steen, D. (Endre Technologies AB). Analys av nätutvecklingsplaner och flexibilitetsmöjligheter för digitalisering. Energiforsk Report 2024:1043. Stockholm: Energiforsk AB, April 2024. Program: Elnätens hållbara teknikutveckling och digitalisering.

Open access: Published by Energiforsk. Endre Technologies AB is a Swedish DSO software and analytics firm.

Summary

This report examines how DSOs currently produce network development plans (DNDPs / nätutvecklingsplaner) and identifies where digitalization and automation could improve the process. The study combines an international benchmarking component (interviews with DSOs in the UK and Portugal) with a methodological contribution: an automation index framework for DNDP sub-tasks, and a “what-if” flexibility forecasting method based on EV charging behavior scenarios.

The report is part of Energiforsk’s program Elnätens hållbara teknikutveckling och digitalisering and was produced by Endre Technologies AB (Hartvigsson and Steen). It is directly referenced by three subsequent reports in the same program (2026-1151, 2026:1157, 2026:1168), confirming its foundational status in the cluster.

Key claims

International benchmarking

• UK and Portugal DSO interviews (7 people): the two countries chosen because they have among Europe’s most mature DNDP practices, including established flexibility need forecasting routines.
• Excel dominates DNDP work across both countries, creating problems with version control, documentation, reproducibility, and knowledge transfer. The absence of purpose-built tooling is a structural gap.
• FTE burden: approximately 1 FTE per 40,000–80,000 grid connections is required for DNDP production and maintenance. For context, a medium-sized Swedish DSO with 50,000 connections could expect ~1 FTE dedicated to this process.

Automation index

The report proposes an automation index to prioritize which DNDP sub-tasks should be targeted for digitalization. Each sub-task is scored on two dimensions:

• Priority / resource intensity — how much time and effort the sub-task currently consumes
• Automatability — how feasible it is to automate with available data and methods

The combined score produces an automation index (0–9 scale in the report). Results:

The flexibility need forecast ranks highest because: (a) it is time-consuming in its current state (manual load flow calculations per scenario), (b) the inputs (load data, EV scenarios, network topology) are machine-readable, and (c) the output (MW × hours by location) is well-defined.

What-if flexibility forecasting method

Rather than producing a single point forecast of flexibility need, the report proposes a scenario-based “what-if” approach that makes the behavioral assumptions explicit and allows stakeholders to assess likelihood themselves:

The method defines EV charging behavior scenarios:

1. Direct charging — all EVs charge immediately on arrival home (uncontrolled; worst case for evening peak)
2. Price-optimized charging — EVs charge in off-peak hours responding to hourly electricity price signals (moderate peak reduction)
3. Grid-friendly charging — EVs charge in actively managed slots optimized for grid capacity (best case; requires DSO-side or aggregator-side coordination)

For each scenario, a load profile is generated at the relevant substation or grid segment. The resulting flexibility need is expressed as:

• MW of peak reduction required
• Hours/year the need is active

Stockholm area case study

The method is applied to a case study in the Stockholm area, demonstrating the range:

This enormous range (22× in peak kW, 71× in hours) illustrates that the behavioral assumption is the dominant uncertainty in flexibility need forecasting — not the technical model. A DSO that presents only one scenario either over- or underestimates its need by a factor of 10× or more.

Flexibility value context: The report references a comparable capacity upgrade cost of approximately ~4 SEK/kW (transformer upgrade, ~50-year lifetime), establishing a benchmark against which flexibility service value can be assessed. This figure is independently confirmed in Energiforsk 2026-1151.

Transparency argument

The report argues that publishing scenario assumptions alongside forecasts — rather than publishing a single number — is both technically more honest and practically more useful: it allows municipalities, customers, and aggregators to form their own views on which scenario is most plausible, and to make connections (“we’re planning to deploy X smart charging stations — that would push toward the grid-friendly scenario”). The transparency argument supports the broader standardization case made by Energiforsk 2026:1157.

Relevance to the wiki

This source directly informs or strengthens:

• Distribution Network Development Plan — methodology for the flexibility need forecasting step; automation index as a framework for understanding where DNDP digitalization effort is best placed; Stockholm case study data
• Flexibility Need Assessment — the what-if scenario approach is a practical alternative to deterministic load-vs-limit forecasting; complements the Endre probabilistic method and RISE AMI tool already documented
• Distribution System Operator — FTE burden benchmarks; Excel-dominated tooling gap; automation as DSO capacity-building need
• Flexibility Market — the scenario range (20–450 kW, 6–427 hours) from a single location shows how sensitively flex market revenue depends on customer behavior assumptions — a key uncertainty for any business case
• Demand Response — EV charging behavior is the primary uncertainty in near-term flexibility need; the three scenario types (direct / price-optimized / grid-friendly) correspond to implicit, partially-implicit, and explicit DR implementations

Internal links to other Energiforsk 2024–2026 program sources:

• Source - Energiforsk 2026-1151 Effektauktioner med Värmepumpar (2026) — cites this report; uses same ~4 SEK/kW benchmark
• Source - Energiforsk 2026-1157 Nationell Metod Effekt och Kapacitetsprognoser (2026) — references this report as related work; addresses the same DNDP comparability and standardization problem at national scale
• Source - Energiforsk 2026-1168 AI-modeller Prognostisering Efterfrågan El (2026) — PREDATOR project extends the “analyze current DER inventory” step this report identifies as needing automation

Data gaps

• The case study is limited to Stockholm; generalizability to rural networks with different EV penetration curves has not been tested
• The ~4 SEK/kW benchmark is stated but not derived in detail; the source (transformer upgrade amortization) is not fully documented
• No survey of Swedish DSOs specifically — the benchmarks come from UK and Portugal; Swedish DNDP practice may differ