The function of energy systems is to provide energy services to end users. The main energy uses are for heat and electricity in industry and buildings and for transport. The industrial, residential and transport sectors together accounted for 70 % of the total global electricity consumption in 2019, and these sectors are also responsible for approximately 60 % of the worldwide carbon dioxide (CO2) emissions . The decarbonisation of the energy system is thus a key driver of overall decarbonisation efforts.

In end-use sectors, the tipping dynamics in wind and solar may initiate further decarbonisation of the energy system through electrification of energy demand. These developments are typically supported by government subsidies and regulation of the energy performance of buildings, and households that electrify their demand may do so for both environmental and financial reasons. Moreover, the fast cost reductions as observed in wind and solar are more likely to occur in smaller and modular technologies . In the consumer end-use sector, there are several other small and modular technologies that may reach cost-parity in the short term, like electric vehicles, household batteries and heat pumps .

The transportation sector is a relevant example of these advancements. The increasing prevalence of electric cars, along with other electricity-powered alternatives such as e-bikes, e-scooters and other mopeds, indicates the key role of batteries in novel modular demand and the significant contribution to sector-wide decarbonisation. In addition to facilitating emission-free mobility, the batteries in electric vehicles can also support the grid infrastructure during periods of ample electricity generation from renewable sources by functioning as modular storage systems.

Stationary household batteries are specifically attractive in places where feed-in tariffs for solar energy into the grid are much lower than the retail price for energy from the grid. The large-scale adoption of household batteries may influence the decarbonisation of the energy system in two ways. First, it reduces curtailment of household solar PV generation, better matching renewable energy supply with demand. Second, it reduces grid congestion during peaks in solar generation (reinforcing feedback on the left in Fig. ). Increasingly, grid congestion is a barrier to further grid integration of renewables. However, few countries currently have strong incentives in place to stimulate demand to synchronise with the availability of renewable energy supply.

Another area of electrification of the residential sector is heating and cooling systems (Fig. ). Heat demand is often met using natural gas boilers. Based on analysis, natural gas accounts for 42 % of global heating energy demand, with a 40 % share of the heating mix in the European Union and over 60 % in the United States. When low-carbon heat sources like waste heat are available, this can be a preferred option. When this is not the case, electrification of heating demand through heat pumps can lead to a large reduction in energy demand. This shift to low-carbon heat sources requires additional changes in technologies and infrastructure in houses and neighbourhoods. The availability of low-cost and abundant renewable energy supply is a key enabler here, and early adopters are often households that generate excess electricity with their own rooftop PV systems. Another important enabler is increased insulation (also to reduce overall heat demand). For instance, the adoption of solar PV coupled with efficiency and sufficiency measures may both yield increased adoption of heat pumps, as shown at the top of Fig. . Barriers to the electrification of heating are the lack of technologies for heat storage and the cumbersome installation process. A more radical and politically challenging behavioural change would be to provide incentives to live in smaller homes or to have higher occupancy per dwelling, for example in planning decisions.

While for many regions in the world renewable energy potential exceeds demand, a fast energy transition faces constraints regarding the availability and sustainable sourcing of materials and personnel . Most scenarios, therefore, envision a reduction in demand where consumption of energy should be brought in line with what can be sustainably produced in the short term. Indeed, reducing energy demand is key in 1.5 °C pathways , especially in wealthy countries. At the same time, energy access and service provision will need to grow for many less-developed countries and for poor people everywhere to ensure decent living standards and wellbeing . Although we observe a decoupling of energy demand and income in some places, in general, household energy demand grows with income. Pro-environmental attitudes and behaviour have also been correlated with income, further complicating the challenge of how to reduce income inequality and material and energy consumption to sustainable sufficiency levels . Moreover, individuals with high socio-economic status are responsible for a large share of emissions . These individuals could have a large positive impact if they reduce greenhouse gas emissions, becoming role models of low-carbon lifestyles, investing in low-carbon businesses and advocating for stringent climate policies . Such approaches are also discussed in the context of energy justice and equitable energy demand reduction .

However, demand reduction options are often constrained by the existing socio-technical system. It is, for example, difficult for individuals to change their mobility practices if employer preferences regarding workplace presence do not change. The Avoid-Shift-Improve (ASI) framework is often used to identify demand reduction options. Avoid options reduce unnecessary energy consumption, possibly by redesigning service provisioning systems. Shift refers to the switch to already existing competitive, efficient and cleaner technologies and service provisioning systems. Improve refers to efficiency improvements in existing technologies. While improve options are not sufficient to tip the energy system to a decarbonised state, they are an important enabler for options that can. Figure  adds ASI measures as an additional element to the previous figure on feedback loop in electrification of heating and cooling.

The different ASI measures often co-occur. While avoid options have the largest mitigation potential, they often need to be flanked with shift and improve options to be attractive. For example, when people switch from natural gas heating to heat pumps, good insulation (improve) is a pre-condition. Typically avoid and shift options require larger changes in social practices and in the broader socio-technical system compared to improve options. More specifically, options where both behavioural and technological change is required, or that require a substantial change in social and user practices, are typically more difficult to realise and thus difficult as a starting point for tipping dynamics .

Higher prices lead to reduced energy demand, providing evidence for measures like a carbon tax. Natural gas consumption in the EU and in the period August 2022 to January 2023 decreased by 19 % compared to the average gas consumption for the same months in the previous 5 years. However, this also came with increased levels of energy poverty, particularly affecting low-income households in badly insulated homes . Interestingly, the high prices also triggered and opened the opportunity for sufficiency-based energy price interventions. Because of the relationship between income and energy use, a rebound effect may occur (see top right balancing feedback in Fig. ) when technologically or socially induced demand reductions lead to a higher budget and more energy demand .

At the individual and household level, energy behaviour changes, including turning down the thermostat and reducing the demand for hot tap water (e.g. taking shorter showers), are effective strategies . These strategies are most effective when combined with policy support and shift and improve measures. In many countries, regulation regarding the energy performance of buildings and subsidies for energy saving measures are enablers of such behaviour change. More specifically, digital technologies are key to better matching renewable supply with demand to avoid curtailments and grid congestion (load shifting and balancing) but have not yet reached widespread diffusion. Digitalisation in general (at the top-left in Fig. ) can also play a key role in avoiding unnecessary energy demand .

Empirical studies show that informing people about the energy conservation behaviours of their neighbours, combined with the public labelling of energy conservation behaviour as desirable, can lead to significant reductions in energy consuming behaviour . These studies show that a relatively weak form of sanctioning (i.e. showing approval and disapproval of particular behaviour by using thumbs up or down or positive or negative smileys) already has a modest positive effect on energy savings. Peer effects in social network structures can provide inhibiting or supporting conditions for the diffusion of energy conservation practices, depending on the structure of the network and the type of activity .

Changes in the energy behaviour and lifestyle of individuals can make a large contribution to sustainability but are only feasible when supported by changes in the broader socio-technical system . This means that social tipping of energy consumption by individuals, households or organisations is conditioned by a range of factors such as social and cultural norms, ownership and control of resources, technology accessibility, infrastructure design and services availability, social network structures, organisational resources, knowledge and awareness, and sociodemographic characteristics . Social and behavioural change is thus constrained by the existing socio-technical system and people's daily lives and behaviour, or social practices . Social practices approaches shine a light on the culturally embedded routines that reproduce (but also potentially transform) socio-technical energy systems from the bottom up. Crucially, they also point to the differentiation of these practices across social groups (e.g. women versus men, upper class versus working class) .

Avoid options reduce unnecessary energy consumption. However, when avoiding energy use is undesirable from a wellbeing perspective, then shifting the way this activity is done (or finding an alternative means to the same goal) is key. When the demand reductions stem from changes in norms or behaviours with a sustainability motive, the risks of rebound effects are lower. However, different attitudes make some demand-side alternatives difficult to scale up in the population . Often lacking are enabling conditions for just and smooth change. For instance, urban infrastructure or the built environment may prevent people from avoiding using private cars through alternatives like walking, cycling or taking public transport.

While some demand-side behaviour changes are quite swift, a key policy challenge is how to make the new and desired behaviour “stick”. An example is the substantial energy demand reduction in Europe during the winter of 2022/2023, resulting from concerns about high energy prices and the war in Ukraine . Similarly, but at a global system level, in 2020, the world witnessed a reduction in global fossil fuel emissions as a result of COVID-19 lockdowns across the globe. However, emissions rebounded in 2021, reaching levels comparable to those observed in 2019 . These observations reinforce the idea that social tipping is better understood as tipping dynamics rather than tipping points . This is not just because they take some time to evolve but also because different reinforcing processes are needed to provide momentum and to ensure that the change sticks or becomes difficult to reverse on the relevant timescales.

Interestingly, pro-environmental behaviours may induce other pro-environmental behaviours, so changes in behaviour in mobility or food may spill over to energy behaviours . The adoption of household PV for environmental reasons may thus induce other pro-environmental behaviours. When the new sustainable lifestyle and behaviour becomes common and the norm starts to shift, this also increases the political feasibility of strict regulation. There is, for example, public support for measures like incentives towards renewable technology and a ban on the least energy-efficient household appliances (see Fig. ).

The positive feedback loop mechanism of opinion exchange can increase awareness and promote more sustainable lifestyles. However, it can also have a negative effect when contrarians get the majority in a given social group, leading to the amplification and reinforcement of anti-environmental beliefs. For this reason, avoiding opinion polarisation is crucial in climate-related issues to foster cohesion for effective government action . Citizens' environmental consciousness and the formation of their opinions directly affects actions that impact the local and global environment .

The presence of a group with strong anti-environmental beliefs can discourage pro-environmental engagement and support for climate change initiatives. Opinion polarisation makes it challenging to reach consensus and decreases public support for environmental initiatives, posing a challenge for policymakers . To mitigate negative feedbacks and harness the positive cascade effect of opinion dynamics, some governments have implemented policies to incentivise pro-environmental behaviours, while awareness campaigns and education aim to correct misinformation and provide accurate information . When opinions drive clique formation, they can lead to concrete pro-environment actions, such as social movements and public support for climate change initiatives .

Social acceptance and changes in norms and behaviours may have large influence on both direct consumer demand and policy support . Civil society engages with energy transitions in many ways: from adopting energy efficient technology to joining energy cooperatives and from environmental activism to resistance against wind parks . These interactions are driven by (changes in) perceptions, attitudes, motivations, emotions, beliefs, values and norms , sometimes triggered by external events like the oil crisis or nuclear accidents. Some of these factors also may influence the willingness to adopt a certain technology (as in ).

There is extensive literature on the social acceptance of renewable energy infrastructure . One of the most prominent conceptualisations of social acceptance is the social acceptance triangle of , comprising community, market and socio-political acceptance. This draws attention to the fact that community acceptance or local opposition to projects can influence general public or political acceptance and societal demand for renewable energy. From this perspective, demand is not simply the economic behaviour of individuals or households but is a product of societal relations. One potential balancing feedback for renewables deployment is project delays caused by local opposition, which leads to pressure to streamline planning and reduce participation options, which in turn creates more opposition. This dynamic is seen in many EU countries today.

Finally, policy feedbacks are well recognised in political science literature, whereby policy not only stimulates deployment but also creates legitimacy and new interests, leading to increased lobbying and support for policy to support the new industries and further deployment . For example, identifies “green spirals” which resemble tipping dynamics for the reduced use of CFCs for ozone protection. Policies engendered new industrial interests who in turn support new policies. Kelsey also identified that these spirals can transcend domestic politics and scale up to the international level. This is similar to the notion of tipping cascades.

Key considerations for policymakers hoping to create tipping dynamics in this way is the sequencing of policies . For the energy transition, similar dynamics can potentially be found with the renewables industry. For instance, the German feed-in tariff for renewables is frequently mentioned as an enabling condition for this feedback . Further, strong pro-environment policies may incentivise firms towards more and R&D and innovation, thereby expanding industrial sectors for low-carbon technologies. In this way, public opinion may also increase support and acceptance for new low-carbon technologies, increasing pressure on policymakers to set goals and strategies for a more sustainable society .

The political sphere can not only trigger social tipping but can also tip itself into a new state, generating a tipping cascade . Indeed, the same applies to any group, organisation or institution that is part of the socio-technical system. For example, civil society could also be a key element in energy system tipping dynamics. Increasing attention is being paid to “prosumerism”, which can be understood as a broad movement towards a decentralised democratic energy model . These and other civil society movements interact with the state, which in turn creates opportunities or barriers to different lines of action for citizens or households, engendering balancing or reinforcing policy feedbacks.

While research on policy feedbacks frequently targets its findings towards policymakers, this knowledge can also be used by civil society or interest coalitions to try to initiate such feedback processes. Indeed, some research from social movements theory identifies movement–policy feedbacks or “opportunity/threat” spirals in which “demands lead to concessions that encourage further demands, and so on” (, p. 228; ). discuss the relationship between the Fridays for Future movement and European states in ways that could align with this idea. Focusing specifically on energy, such feedbacks could help to explain the recent boom of the energy cooperative movement in countries like the Netherlands, for example.

In general, the energy transition requires a system-level transformation of the energy system, which depends on both phasing out fossil fuels and accelerating renewable energy provision. In the previous sections, we highlighted the promise of positive social tipping dynamics in renewables development, while recognising that this is inevitably an incomplete picture without fully considering fossil phase-out. Below we address some of the feedbacks (Fig. ) that strengthen the existing fossil-based energy system (see , for a systematic overview).

First, sources of balancing feedback, lock-in and path dependence of fossil fuel-based energy systems are energy infrastructures, technologies and institutions . These can directly hinder the decarbonisation of the energy system through existing standards and resistance from incumbents and vested interests. Further, renewable energy generation sometimes faces curtailment, and the mismatch of renewable supply with energy demand slows down the replacement of fossil fuels. Indirectly, the availability of cheap energy has stimulated demand for energy-intensive goods and services. Similarly, the high return on fossil fuel investments and the assessment of renewables as risky make it difficult to move capital from fossil to renewables . As an example, in the early 2000s, the UK government provided initial capital grants to boost offshore wind demonstration projects. This has, in turn, built confidence among financial investors, easing access to resources for project developers (i.e. lower interest rates) , stimulating the overall offshore sector.

Second, examples of barriers encountered by renewables are challenges related to intermittency and the need for a flexible and well-managed grid infrastructure to ensure a reliable and stable energy supply. The increasing need to electrify various end-user sectors adds further complexity to the grid management challenge. For instance, the electrification of transportation is experiencing rapid growth, boosted by policy initiatives for the adoption of e-mobility. Similarly, there is a strong policy focus on electrifying heating and cooling systems in residential areas and districts. Moreover, the electrification of demand is not always viable, and the energy transition may negatively impact individuals with restricted financial resources . In addition, many processes that reinforce fossil-fuel-based energy systems, ranging from subsidies to vested interests and existing infrastructures, are still in place. Energy infrastructures are typically built for a lifespan of around 40 years, and changing these infrastructures takes place on the timescale of months to years. Once built, they contribute to stabilising the system state and are a source of path dependence and lock-in.

Social dynamics can also create balancing feedbacks when they mobilise opposition and a lack of societal support for larger-scale solar and wind parks . Therefore, cost-competitiveness is not a sufficient indicator to predict support for technologies for which the main public concerns are about spatial impacts, health and safety, and questions of fairness. This shows that economic tipping dynamics alone are not sufficient to realise rapid decarbonisation.