Source - Energiforsk 2026-1151 Effektauktioner med Värmepumpar (2026)

Full citation: Månborg, S. et al. (Profu AB + Bengt Dahlgren AB). Effektauktioner med värmepumpar. Energiforsk Report 2026-1151. Stockholm: Energiforsk AB, January 2026. Program: Elnätens hållbara teknikutveckling och digitalisering.

Open access: Published by Energiforsk. Lead author: Samuel Månborg (Profu); heat pump modelling by Bengt Dahlgren AB. Cites Source - Palm et al LFM Drivers and Barriers (2023) and Source - Energiforsk 2024-1043 DNDP Analys och Flexibilitet (2024) directly.

Summary

This report tests a VCG (Vickrey-Clarke-Groves) auction algorithm as a procurement mechanism for heat pump demand response in the context of DSO local flexibility markets. Using a simulation of 137 heat pumps in Kristianstad (calibrated from UK real-world monitoring data and normalized to Swedish conditions), the report calculates VCG auction prices for a 50 kW flexibility need and assesses the mechanism’s advantages and drawbacks compared to standard marginal pricing.

The report is a methodological contribution to flexibility market design — specifically: how should DSOs price and procure flexibility when markets are thin (few participants), bids are infrequent, and the standard “truthful bidding” assumptions of marginal pricing may not hold?

Key claims

The VCG mechanism

VCG (Vickrey-Clarke-Groves) is a class of incentive-compatible auction mechanisms in which:

• Each bidder is paid based on the social value they contribute, not their own bid
• The payment rule incentivizes truthful bidding — it is never in a bidder’s interest to misrepresent their true cost
• The mechanism maximizes total welfare (sum of surpluses) at equilibrium
• It avoids the Winner’s Curse (the tendency for winners to overpay under marginal pricing in thin markets)

The VCG mechanism is related to the Vickrey sealed-bid second-price auction (Clarke and Groves extended it to multi-item settings). It is considered theoretically optimal for social welfare but is complex to implement; William Vickrey received the Nobel Prize in Economics in 1996 for related work (the prize was awarded in 2020 posthumously in the context of auction theory).

How VCG payment works in practice: In a marginal pricing (pay-as-cleared) auction, all winners receive the same clearing price (the marginal bid’s price). In VCG, each winner i is paid the externality they impose on others — the reduction in welfare that the other winners would experience if winner i were excluded. In competitive markets (many bidders), VCG payment converges to the marginal price. In thin markets (few bidders), VCG payment can be substantially lower — the winner is paid their contribution to welfare, not the market-clearing price.

Simulation design

• 137 heat pumps in Kristianstad municipality (population ~35,000; southern Sweden)
• Data source: UK HeatpumpMonitor dataset (real-world monitoring of heat pump operation)
• Normalization: UK data adjusted to Swedish conditions (climate, building envelope, usage patterns) using SVM (Support Vector Machine) regression — a machine learning approach that maps the UK operating profiles to Swedish equivalents without requiring a full building physics simulation
• Flexibility need simulated: 50 kW (representing a typical distribution substation congestion need in a medium-density area)
• Bid curves: Continuous (not discrete block bids); each heat pump submits a bid based on its predicted power consumption in the next hour; bids are scaled by predicted hourly power and updated infrequently (not per-hour, reducing administrative burden)

VCG price results

For a 50 kW need:

Comparison: Grid investment alternative ≈ ~4 SEK/kW (transformer upgrade, consistent with Source - Energiforsk 2024-1043 DNDP Analys och Flexibilitet (2024)).

Total annual payout for full Kristianstad DSO flexibility need (2–10 MW scale):

• 2 MW need → ~1,600 SEK/year
• 10 MW need → ~16,000 SEK/year

These are very low absolute numbers — reflecting both the low VCG price and the limited hours of activation. The report acknowledges that this creates a structural problem: at these payment levels, there is no revenue incentive for an aggregator to build and maintain the bidding infrastructure on behalf of heat pump owners.

VCG vs marginal pricing: when does it matter?

The report’s argument for VCG in the heat pump context:

1. Heat pump flexibility markets are likely to remain thin for the foreseeable future — few aggregators, infrequent auctions
2. Standard marginal pricing rewards “Winner’s Curse” behavior — bidders who anticipate winning shade up their bids; strategic behavior raises the clearing price
3. VCG gives truthful bidding as a dominant strategy — reduces strategic behavior regardless of other participants’ strategies
4. The social welfare maximization property means VCG selects the least-cost resource regardless of gaming attempts

Challenges and limitations

The report explicitly acknowledges four significant challenges:

1. Cartel manipulation risk: If a small group of heat pump aggregators coordinates, they can artificially inflate VCG payments by misrepresenting their cost structures. In thin markets this is a realistic concern. The report suggests monitoring and market design constraints as mitigations but does not resolve the problem.

2. Algorithm complexity: VCG requires solving a social welfare maximization problem at each auction clearing — computationally tractable for small markets but potentially complex as the number of bidders and constraint dimensions grows. DSOs would need software support to run VCG clearing.

3. No aggregator incentive: At 0.8–1.6 SEK/kW and 1,600–16,000 SEK/year total payout, there is no commercial case for an independent aggregator to build the bidding infrastructure for heat pump owners. The mechanism only works if participation is bundled with other services (e.g., smart tariff optimization, comfort guarantee services, energy service contracts) that provide additional revenue to the aggregator.

4. Design mismatch with existing markets: Swedish Flexibility Market products (LFM-h/p/e) and platforms (SWITCH, NODES) are designed around pay-as-bid or pay-as-cleared marginal pricing. VCG would require a different settlement calculation that existing platforms do not support.

National heat pump flexibility potential

The report cites Power Circle’s FlexAbility DR1 data for national-level context:

• 5.75 GW of heat pump flexibility potential at hourly timescale by 2030 (nationally, Sweden)
• Literature range: 1–6 GW (various studies; wide range reflects different assumptions about demand response participation rates and simultaneous availability)

For comparison, Sweden’s total DNDP-reported DSO flexibility need at 6–10 year horizon is 1,387–2,523 MW (Source - Ei PM2025-03 DNDP Sammanställning (2025)). The national heat pump potential is thus 2–4× larger than the expected DSO need — suggesting that if participation barriers can be overcome, heat pump flexibility is more than sufficient to meet near-to-medium term DSO needs.

~300,000 connected heat pumps is cited (from Source - Energimyndigheten ER 2025-35 Förbättra Flexibiliteten (2025)) as the approximate threshold at which simultaneous activation could cause grid frequency instability — a cybersecurity and coordination concern independent of the VCG mechanism design.

Relationship to Palm et al. (2023)

The report explicitly cites Source - Palm et al LFM Drivers and Barriers (2023) on the market complexity barrier: the VCG mechanism addresses one part of this (pricing difficulty for FSPs — “I have no idea what price to set on flexibility”) by relieving individual bidders of the need to strategize. Under VCG, a heat pump owner can simply bid their true cost and trust that the mechanism will pay them appropriately, without needing to understand market dynamics. This is a direct design response to the information barrier identified by Palm et al.

Relevance to the wiki

This source directly informs or strengthens:

• Flexibility Market — introduces VCG as an alternative auction mechanism to pay-as-bid/pay-as-cleared marginal pricing; provides design argument for thin/infrequent markets; updates heat pump potential to 5.75 GW
• Demand Response — 5.75 GW national heat pump flexibility potential; bid curve concept for continuous DR; SVM normalization approach for DER data
• Aggregation — no aggregator incentive at VCG prices is a design problem; aggregators need to bundle flex revenue with other services to build a viable business case
• Distribution Network Development Plan — ~4 SEK/kW grid upgrade reference confirms the 2024:1043 figure; VCG as alternative to grid investment for recurring substation-level congestion
• Baseline Methods — VCG replaces the baseline problem for resource-level procurement: because bids are based on predicted power (not historical baselines), there is no counterfactual estimation problem; the mechanism itself verifies truthful bidding rather than relying on baseline measurement

Internal links to other Energiforsk 2024–2026 program sources:

• Source - Energiforsk 2024-1043 DNDP Analys och Flexibilitet (2024) — cited; same ~4 SEK/kW benchmark; both analyze the DSO-side case for flexibility over grid investment
• Source - Energiforsk 2026-1157 Nationell Metod Effekt och Kapacitetsprognoser (2026) — complementary: 2026:1157 addresses how DSOs forecast what they need; 2026-1151 addresses how they procure it once the need is defined
• Source - Palm et al LFM Drivers and Barriers (2023) — cited; VCG design addresses the pricing difficulty barrier identified by Palm et al.

Data gaps

• The simulation uses UK HeatpumpMonitor data normalized to Sweden via SVM — the normalization quality has not been independently validated; Swedish heat pump operational profiles may differ materially
• Kristianstad is a medium-density urban area; results may not generalize to rural networks with lower heat pump density
• The aggregator incentive problem (no revenue at VCG prices) is identified but not resolved — a path to commercially viable aggregation at these prices is absent from the report
• VCG requires platform support that no current Swedish LFM platform (SWITCH, NODES) provides; implementation path is not addressed