Ei Benchmarkingmodell RP5 (2026)

Web news article by Energimarknadsinspektionen (Ei), published 2026-05-29. Announces the specific DEA-based benchmarking model that will be used for the TOTEX cost-efficiency incentive in RP5 (2028–2031). A follow-up to the May 2025 inriktning (Source - Ei Inriktning intäktsramar 2028-2031 (2025)) and the May 2026 efficiency incentive announcement (Source - Ei Effektiviseringsincitament Webb (2026-05-12)). Directly resolves the previously open question of how Ei will adjust the benchmark for structurally heterogeneous DSO conditions.

Summary

Ei will use DEA (Data Envelopment Analysis) as the benchmarking framework — the same method used in the current regulation — but with four structural changes for the TOTEX era:

1. All costs included (except myndighetsavgifter)

All cost items are included in a single benchmarking cost variable: capital costs, controllable operating costs, subscription costs to overlying and adjacent networks, and network loss costs. Only regulatory fees (myndighetsavgifter) are excluded — they are neither controllable nor substitutable.

Rationale: a company with a higher-voltage connection to the overlying grid may have higher capital costs but lower subscription costs; combining both enables fair cross-company comparison. Similarly, a company investing in a lower-loss transformer reduces long-run loss costs but increases capital costs — both dimensions must be visible in the benchmark.

2. Elområde correction — price normalization in benchmarking

The raw source leaves this as an open choice among three options, not a settled “both”: Ei considers the most appropriate method to be normalizing, in the benchmarking, either network loss costs, subscription costs, or both to a common electricity price. Electricity price differences between areas (elområden) are outside DSO control and would otherwise create systematic cost differences unrelated to efficiency.

Key methodology: the normalization applies only in the benchmarking comparison. The efficiency incentive (in SEK) is then applied to the DSO’s actual, unreduced costs — not the normalized benchmark costs. This preserves a real financial incentive to manage actual costs while enabling fair peer comparison.

3. Förläggningsmiljö correction — installation environment adjustment

Different types of soil, terrain, and urban density drive significantly different acquisition costs for underground cables and substations. A DSO in an urban or difficult terrain environment would appear inefficient in benchmarking compared to a DSO in rural flat terrain, even if both are equally well-managed.

Ei will apply schablonized percentage deductions per installation environment type (förläggningsmiljö) for underground cable acquisition values — reducing urban/difficult values toward a “landsbygd normal” baseline in the DEA model. A similar adjustment is planned for nätstation types. As with the elområde correction: deduction in benchmarking, incentive on actual unreduced acquisition values.

4. Ledningslängd added as structural factor

Both number of substations (nätstationer, already in current model) and line length (ledningslängd) are used as structural factors for customer density. For sparse rural networks where line length is a more meaningful predictor of cost structure than substation count, this improves benchmark fairness. Grid companies themselves confirmed this in consultation responses.

Production variable supplement

Delivered energy at gränspunkter (grid connection points to overlying network) is added to the existing production output variable (delivered energy to end customers at high voltage). This was identified as a gap in the current model.

Key methodological principle

All corrections follow the same two-step logic:

• In benchmarking: costs or structural factors are normalized to enable fair cross-company comparison
• Efficiency incentive applied: to actual, unreduced costs — preserving the genuine incentive to manage real cost levels

This avoids the critique that corrections “flatten” efficiency differences: the comparison is fair, but the financial reward or penalty is based on actual performance.

What remains open

An earlier version of this page listed three items here (revenue-frame impact cap, kalkylränta beta/peer-selection criteria, enstegsmetoden as the TOTEX method) — none appear anywhere in this raw article. They were apparently carried over from a different, related Ei source without attribution and have been removed pending a correct citation; this article itself states no open questions beyond the model’s own iterative development described above.

Relevance

• Ei: resolves the data gap on adjustment methodology for heterogeneous DSO conditions. Updates the RP5 TOTEX benchmarking model specifics.
• Distribution System Operator: affects how Swedish DSOs’ costs will be compared and incentivized from 2028 (this article gives no DSO count — the “~170” figure in an earlier version was unsourced and has been removed)
• Flexibility Market: the corrections ensure the lösningsneutralitet principle applies even for DSOs with structurally different cost bases — all will face the same incentive to choose the cheapest solution (grid or flexibility)