Source - FlexAbility Delrapport 4 (2025)

Full title: Den sociala potentialen för flexibilitet (The social potential for flexibility)

Part of: FlexAbility — sub-report 4 of 5

Authors: USER research group, Uppsala University

Funded by: Energimyndigheten (Swedish Energy Agency), project FlexAbility

Date: October 2025

Length: 28 pages

Related reports: Source - FlexAbility Delrapport 1 (2025), Source - FlexAbility Delrapport 2 (2025), Source - FlexAbility Delrapport 3 (2025)

Summary

The fourth FlexAbility report addresses the social dimension of flexibility — what Swedish households actually think, feel, and prefer regarding flexible electricity use and delegating control to an aggregator. Unlike the first three reports (which covered resource potentials, market development, and DSO economics), this one is sociological and behavioral research conducted by Uppsala University’s USER group.

The study has two components: a large survey of 49,813 electricity consumers, and Discrete Choice Experiments (DCE) with active aggregator customers.

Study design

Survey

Potentiella flexkunder = DSO customers (Jämtkraft, Skövde Energi, Kraftringen) — not yet active in automated flex Aktiva flexkunder = CheckWatt customers — already participating in automated flex (most have solar, EVs, home batteries)

The 14.4% vs 3.3–3.9% response rate gap is attributed to far higher interest and engagement among the active flex group.

Sample representativeness — important caveat

The sample is not representative of the Swedish population. Overrepresented groups:

• Men (95% of CheckWatt respondents, 73% of DSO respondents)
• Middle-aged and older people
• Highly educated (high vs low education)
• Pensioners and high-income earners
• People of Swedish national origin
• Small households
• Villa owners (single-family homes)

This means results likely overestimate the population-level social potential, especially the potentiella flexkunder findings, since even in that group the respondents are likely more engaged with energy topics than the average customer.

Discrete Choice Experiments (DCE)

Conducted only with CheckWatt customers (active flex users) — the rationale being they have higher “cognitive capacity” to evaluate business model attributes from actual experience. Three DCEs: heating, EV charging, home batteries. Each had five attributes at two levels. Method: conditional logit model to estimate odds ratios (OR) for each attribute.

Current control of flexible assets (Tabell 4)

Among all respondents with each type of asset:

Home batteries are the outlier: 65% of home battery owners use an aggregator — by far the most aggregated asset class. The likely explanation: people acquire home batteries primarily in order to engage with an aggregator’s business model (e.g., FCR market revenue via CheckWatt). This suggests the business model is already well-established for batteries, but has not yet propagated to heating, EV charging, or solar.

Belief in flexibility’s sustainability value (Tabell 5)

On a 1 (not at all) to 7 (very much) scale:

Active users are significantly more convinced on both dimensions. Causality is unknown: do they believe because they have participated, or did they participate because they already believed? The report does not resolve this.

Propensity for flexible electricity use (Tabell 6)

Potentials are cautious about delegation (2.96/7) — barely above neutral. Actives are enthusiastic (4.93/7). The gap is large and has implications for market development: unlocking potentials requires a genuine behavior change, not just product availability.

What affects households’ ability to be flexible (Tabell 7)

Factors affecting household capacity for flexible electricity use (1–7 scale):

The top barriers are asset access and interest — not knowledge or scheduling. Other household members are the least constraining factor.

Drivers and barriers for delegating to aggregator (Tabell 9)

Linear regression analysis of what affects propensity to delegate control to an aggregator:

Key counterintuitive finding: lack of knowledge about aggregator options is not statistically significant in the regression. Even people with limited awareness of how aggregation works have an intuitive sense of whether they want it. This directly challenges the common policy assumption that awareness campaigns are the primary lever for unlocking the market.

High-income households are significantly more likely to delegate — attributed to more flexible assets, greater risk appetite, and higher tolerance for loss of control relative to economic gain.

Education strongly predicts propensity to sign aggregator contracts — associated with greater understanding of and trust in new technical solutions.

Villa owners are more likely than apartment dwellers — because they have more controllable assets (heat pumps, EV chargers, batteries). Apartment dwellers have typically little individual control over heating and must rely on collective solutions.

Probability of signing an aggregator contract (Tabell 10)

Discrete Choice Experiments: contract attribute preferences

Heating (Tabell 11 — active flex users only)

Odds ratios (OR) from conditional logit model — higher = more attractive:

Autonomy is the #1 factor for heating: regaining control (OR 3.15) outweighs compensation (OR 2.67). Heating is comfort-critical, making loss of control more threatening than for other assets.

Surprise: unlimited activations are preferred over “once per day” — households don’t object to more frequent control interventions as long as override rights are preserved.

EV charging (Tabell 12 — active flex users only)

Compensation is the #1 factor for EV charging (OR 7.16 — by far the strongest attribute across all three DCEs). EV charging is treated as a transactional service: discomfort is acceptable at sufficient compensation. Autonomy remains important (OR 5.37) but is subordinate to money.

Surprise: more frequent charging pauses (every session) are preferred over less frequent (every 5th) — consistent with the heating finding. Households trust the system more if they can intervene frequently, not less.

Home batteries (Tabell 13 — active flex users only)

Compensation-first (OR 5.17), autonomy second (OR 2.52). Shorter notice periods preferred (12-month lock-in is unattractive). 24/7 control availability preferred over nights-only.

Trust in aggregator types

Independent aggregators receive slightly higher trust than retailers (elhandlare) and car manufacturers (biltillverkare) across all three DCEs. However, the trust differential is small (OR ~1.06–1.10) and other factors dominate. Trust functions as a necessary but not sufficient condition: households need baseline trust to consider an offer, but compensation and autonomy determine whether they accept it.

Conclusions

1. Economic incentives are the primary driver

The strongest motivation is earning money (5.58/7 overall), followed by cutting electricity costs (5.22/7). Sustainability motivations are present but weaker — and are a secondary factor once economic incentives are established.

2. Autonomy is non-negotiable

The ability to temporarily regain control of the controlled asset is critical across all three applications. For heating it is the single most important factor. Without an override right, no economic compensation reliably makes the offer attractive.

3. Knowledge deficit is not the main barrier

Lack of knowledge about aggregation options has no statistically significant effect on propensity to delegate. People have a sufficient intuitive understanding to form a preference. Information campaigns alone will not unlock participation.

4. Product design must be differentiated by asset

• Heating: autonomy-first; even unlimited activations are acceptable if override rights exist; comfort/indoor temperature effect is a major deterrent
• EV charging: compensation-first at very high OR (7.16); more transactional; shorter charging delays preferred
• Home batteries: compensation-first (OR 5.17); flexible contract terms (short notice period) matter; anytime activation preferred over nights-only

5. The social potential is real but unevenly distributed

The market is currently concentrated in villa owners (with heat pumps, EVs, batteries), highly educated, and predominantly male households. Apartment dwellers are largely excluded because they lack individual control over heating and shared infrastructure. Collective solutions (via housing associations / bostadsrättsföreningar) are needed to reach this group.

6. Policy implications

• Transparent and fair compensation rules are needed for household trust
• Flexibility contracts should guarantee override rights — regulatory framework should support this
• Information campaigns should target lower-education groups specifically, but with realistic expectations of impact
• Subsidies or incentives for low-income households may be needed to prevent the social potential from reinforcing existing inequalities
• Collective flex solutions for multi-family housing require specific regulatory support

Study limitations

• Sample is not representative (biased toward men, highly educated, villa owners, high income)
• DCE conducted only with active flex customers (early adopters) — not generalizable to the full population
• Results should be treated as indicative of early-adopter preferences, not population averages
• The 14.4% vs 3.3–3.9% response rate differential may have distorted the active/potential comparison

Relevance to wiki