Deep decarbonisation of buildings energy services through demand and supply transformations in a 1.5°C scenario

When analysing the emissions from the buildings sector, it is important to consider not only direct emissions from on-site combustion of fossil fuels, but also the impact that buildings energy demand exerts on energy supply emissions. Emissions from energy use in buildings can be decomposed straightforwardly between the influence of useful energy $\left( {{\text{UE}}} \right)$, the conversion intensity $\left( {\frac{{{\text{FE}}}}{{{\text{UE}}}}} \right)$ and the emission intensity $\left( {\frac{{{\text{C}}{{\text{O}}_2}}}{{{\text{FE}}}}} \right)$ (equation (1)). Useful energy corresponds to the amount of energy that comes out of a conversion appliance and is available to provide an energy service. In the case of space heating, useful energy is the energy delivered by the boiler to the room to be heated. Final energy is the energy bought on markets or collected by consumers (e.g. electricity or traditional biomass). The ratio between final and useful energy gives the conversion intensity. To compute the resulting emissions, one needs to take the emission intensity into account (equation (2)). The latter can be disaggregated into the shares of each energy carrier (${\text{s}}{{\text{h}}_{{\text{ec}}}}$), and the individual emission intensity of energy carriers ${\left( {\frac{{{\text{C}}{{\text{O}}_2}}}{{{\text{FE}}}}} \right)_{{\text{ec}}}}$:

$$\begin{equation}{\text{C}}{{\text{O}}_2} = {\text{UE}} \times \frac{{{\text{FE}}}}{{{\text{UE}}}} \times \frac{{{\text{C}}{{\text{O}}_2}}}{{{\text{FE}}}}\end{equation} \tag{ 1 }$$

$$\begin{equation}\frac{{{\textrm{C}}{{\textrm{O}}_2}}}{{{\textrm{FE}}}} = \sum\limits_{{\textrm{ec}}} {{\textrm{s}}{{\textrm{h}}_{{\textrm{ec}}}}} {\left( {\frac{{{\textrm{C}}{{\textrm{O}}_2}}}{{{\textrm{FE}}}}} \right)_{{\textrm{ec}}}}.\end{equation} \tag{ 2 }$$

From this decomposition, four strategies emerge to decrease buildings emissions (table 1), which we can regroup into two broad categories: reductions in energy demand, and reductions in the carbon content of energy.

In this study, we are interested in estimating the individual contributions of each of these four detailed strategies on the evolution of buildings emissions, especially in the context of mitigation scenarios. To that end, we decompose the change in emissions following equation (3) (details in supplementary note 6 (available online at stacks.iop.org/ERL/16/054071/mmedia))

$$\begin{equation}{{\Delta}}{{\text{CO}}_2} = {\text{u}}{{\text{e}}_{{\text{eff}}}} + {\text{con}}{{\text{v}}_{{\text{eff}}}} + {\text{f}}{{\text{s}}_{{\text{eff}}}} + {\text{s}}{{\text{d}}_{{\text{eff}}}}.\end{equation} \tag{ 3 }$$

With ${\text{u}}{{\text{e}}_{{\text{eff}}}}$ the effect of useful energy, ${\text{con}}{{\text{v}}_{{\text{eff}}}}$ the effect of the conversion efficiency, ${\text{f}}{{\text{s}}_{{\text{eff}}}}$, the effect of electrification and fuel switching and ${\text{s}}{{\text{d}}_{{\text{eff}}}}$ the effect of supply decarbonisation. The effect of energy demand reduction is summarised by combining the effects of useful energy and of the conversion efficiency.

We use the IAM REMIND (Luderer et al 2013, ADVANCE 2016) to analyse the role of buildings energy demand in the context of the whole energy system. REMIND is a general equilibrium model which includes representations of the economic, energy and climate systems. It is used to conceive of possible pathways to curb climate change, but also to assess the social and environmental implications of these pathways. REMIND represents 12 regions ³ covering the global energy demand and GHG emissions. In REMIND the macroeconomic output is a function of the inputs labour, capital and aggregated energy services. The aggregated energy services derive from the energy consumption in three sectors: buildings, industry and transport. Each sector requires final energy carriers to provide the sector-specific energy services. The economic output is used for consumption, trade, investments into the macroeconomic capital stock, and energy system expenditures. The energy supply system provides the energetic inputs required by the economic system. The energy supply explicitly represents vintage capital stocks for more than 50 conventional and low-carbon energy conversion technologies and tracks energy flows from primary through secondary to final energy. The macroeconomic and the energy system modules are hard-linked via final energy demand and costs incurred by the energy system. Energy production and final energy demand are determined by market equilibrium (supplementary note 1).

The representation of buildings energy demand has been considerably strengthened in order to reproduce the energy efficiency dynamics in this sector. In REMIND, energy efficiency results from endogenous decisions to invest into energy end-use capital—e.g. efficient air conditioners or LEDs fixtures—, and thereby to reduce energy consumption and expenditures. The optimal ratio between end-use capital and energy consumption depends on the price of capital, the price of energy as well as the elasticity of substitution between both factors—a parameter characterising the ease of substituting capital intensive technologies for energy intensive technologies. Formally, the trade-off is represented through a constant elasticity of substitution function. The elasticity of substitution plays a central role in determining the response of efficiency improvements to changes in relative prices. We calibrated the elasticities of substitution based on technological data (supplementary note 3). These investment dynamics concern three energy service categories—appliances and lighting, insulation, and space cooling. In addition, the choice of conversion technologies like electric resistances or heat pumps for space heating as well as for water heating and cooking are determined by a multinomial logit on the basis of the capital and operating costs of each technology (supplementary note 5). A relatively cheap net present value will translate into a high market share for a conversion technology.

The model also represents barriers to the economically efficient implementation of energy efficiency measures. Representing these barriers is essential to account for real-world market failures, as well as to enable a meaningful discussion of currently implemented and conceivable future policies to address these barriers. We follow the approach of representing under-investment into efficiency measures in terms of an implicit discount rate, as already established in the scientific literature (Koomey et al 2001, Wilkerson et al 2013, Capros et al 2016). Implicit discount rates are the discount rates that make observed purchasing decisions coherent with decisions taken according to the net present value of alternatives (Hausman 1979, Train 1985, Schleich et al 2016). It is therefore a convenient way to integrate behaviours that do not seem economically rational into models assuming rational agents. As energy efficient technologies have lower operating costs but higher initial capital costs, high discount rates give inefficient technologies a competitive edge over efficient ones. Following this approach, policies alleviating barriers are mimicked in REMIND via their impact on the implicit discount rate.

In REMIND, the macroeconomic discount rate is computed endogenously. To model the implicit discount rate in buildings, we impose a tax on the end-use capital and recycle the tax revenues in a lump-sum fashion. This pro rata tax increases the macroeconomic discount rate additively and results in an end-use specific implicit discount rate. For instance, if the macroeconomic discount rate is 7% and the tax on end-use capital is 10%, the full discount rate on end-use capital will be 17%. Early studies showed large ranges of estimates for implicit discount rates (Hausman 1979, Train 1985, Sanstad et al 1995). More recent studies suggested lower estimates (Cohen et al 2015), which can be explained by improvements in estimation techniques (Houde 2014, Cohen et al 2015), or by the effect of existing information policies (Min et al 2014). We have therefore opted for target implicit discount rates (table 2) that are consistent with the lower bound of early estimates. More information on the buildings module can be found in the supplementary information (notes 2–5).

Using an IAM allows investigating buildings energy demand in the global context of the energy system and in a framework consistent with the 1.5°C climate target. But these benefits come at the cost of some limitations regarding the buildings representation in REMIND. First, the module represents buildings energy demand as an aggregate for the whole sector and does not distinguish for instance between residential and commercial buildings, or between urban and rural demand. This aggregation presents the important advantage of reducing the computational requirements of the model, but sub-sectoral dynamics, policies and results cannot be displayed as a consequence. Second, and related to this, the choice of energy carriers is, as in other IAMs (e.g. van Ruijven et al 2010, Eom et al 2012) represented through a multinomial logit function. However, multinomial logit functions are only an imperfect representation for the drivers behind the heterogeneity of the market. As a result the energy carrier choice as a response to changes in prices might not fully represent the technically optimal solution. Nevertheless, these limitations do not affect the main results from this study, which remain general in scope.

The scenarios run in this study (table 3) evaluate the impact of two types of policies. First, a standard carbon pricing policy is implemented in order to rein in global warming below 1.5°C ('1.5°C' scenario). The carbon price is adjusted so that the global emissions remain within the limits of a 1.5°C carbon budget as presented in the IPCC Special Report on the Global Warming of 1.5°C (Rogelj et al 2018). Carbon pricing is the central tool for achieving climate targets as it would allow—barring other market failures—to identify the most economical solutions. Carbon pricing encourages higher energy efficiency by raising the cost of energy. But it does not address market failures pertaining to energy efficiency such as the split incentives between landlords and tenants or the lack of information on the energy consumption of appliances. We therefore design a second scenario ('EG') which represents the impact of lifting efficiency barriers on energy efficiency investments. While we do not model explicit efficiency policies (e.g. information policies), we model the impact such policies could have in terms of reductions in the implicit discount rate of various end-uses and technologies. In real-world conditions, these policies are extremely varied as they address very diverse barriers: policies targeting market failures (e.g. rental contracts allowing landlords to raise rents following an efficiency investment, or labelling programs), policies based on building standards which are a popular mean of raising energy efficiency but do not necessarily follow economic optimality principles, feedback campaigns (Allcott and Rogers 2014, Asensio and Delmas 2015), etc. In addition to these two first scenarios, we run a third scenario ('1.5°C-EG') which combines the assumptions from the '1.5°C' and 'EG' scenarios. Finally, the 'Baseline' scenario serves as a counter-factual to assess the impact of the various policies. The 'Baseline' scenario, on which all others are based, follows the SSP2—Middle of the Road—economic and demographic projections (Dellink et al 2017, KC and Lutz 2017, O'Neill et al 2017), which project a global population of more than 9 billion people in 2050 and a growth in global income per capita from US$2005 11 500 in 2015 to US$2005 25 000 in 2050.

Both the climate policy and the reduction of efficiency market failures are extremely ambitious in the model. For instance, the global carbon price starts at 100$/tCO₂eq in 2025 and the 'EG' scenario assumes that all barriers will be lifted by 2025. The goal of these scenarios, and the stylized policies in place, is to depict a world that would be on track with the 1.5°C target and undertake all necessary policies in order to achieve this goal at the lowest possible cost, including through the removal of efficiency barriers.

The scenarios clearly show (figure 1(a)) that lifting efficiency barriers has an impact on final energy demand that is of similar scale to carbon pricing policies. The combined effect of carbon pricing and reduced efficiency barriers ('1.5°C-EG') reaches 31% of demand reduction by 2050 in comparison to the 'Baseline'. While scenarios with no or single-focus policies show an increase in energy demand until mid-century, the '1.5°C-EG' scenario is the only one to curb energy demand to its 2015 level by 2050 (−3%). By contrast, the demand increases by 41% until 2050 in the 'Baseline' scenario and by 16%–21% in the single-focus policy scenarios. In the short term, the impact of carbon pricing on energy demand is particularly important as the energy supply has had no time to complete its decarbonisation. As a response to the introduction of carbon prices, final energy prices increase by 26%–58% in 2025 compared to the baseline. In the 'EG' scenario on the other hand, prices drop by 27% (supplementary note 10).

Zoom In Zoom Out Reset image size

The reaction to policies varies strongly across end-uses. Space heating and space cooling are the most sensitive end-uses. They are affected by both improvements in the building shell and by efficiency gains in technologies converting final into useful energy (air conditioners, boilers, etc). The compound effect of buildings envelope upgrades and conversion efficiency improvements yields stark decreases in the demand for these end-uses (−29% for space cooling and −40% for space heating by 2050 compared to the baseline). The other end-use to strongly react to policies is appliances and lighting (−37% by 2050). However, unlike space heating and cooling, this drop is predominantly led by efficiency policies. The 'EG' scenario translates into a decrease of 33% against 6% if only carbon pricing is implemented. The reasons behind this are, one the one hand, the assumed stronger market inefficiencies in the demand for appliances and lighting technologies, and on the other hand, the lower effect of climate policies on the long-term price of electricity, which is the main energy carrier used for appliances. Cooking and water heating have the least possibilities to reduce energy demand in the model and therefore show the least reductions (−18% in '1.5°C-EG'). In particular, the energy efficiency policies do not address the large-scale use of traditional biomass, which is energetically inefficient. The recourse to inefficient traditional biomass in REMIND is driven more by income than by energy prices and discount rates. Reducing the use of inefficient traditional biomass for more efficient modern fuels would therefore necessitate policies affecting income rather than policies addressing market failures in the efficiency markets as in the 'EG' scenario.

In order to locate the degree of ambition of the '1.5°C-EG' scenario in terms of energy demand reductions, we compare it with scenarios from the 1.5°C scenario database (Huppmann et al 2018). These scenarios underpin the IPCC Special Report on a Global warming of 1.5°C (Rogelj et al 2018) and reflect the latest improvements of IAM scenarios. We here select 1.5°C scenarios which had a reference scenario to be compared with (figure 2). We also added the 'Beyond Two Degrees' scenario from the IEA Energy Technology Perspectives (International Energy Agency 2017). Our '1.5°C-EG' scenario clearly belongs to the most ambitious scenarios available in the database. We can also notice that the current REMIND version with the detailed buildings sector shows more ambition than previous versions (Luderer et al 2018). Supplementary note 8 compares the results with non-IAM scenarios from the literature and confirms the impression that the '1.5°C-EG' is ambitious in terms of energy demand reductions, while not being the most optimistic.

Zoom In Zoom Out Reset image size

Beyond the change in the level of global energy demand, the scenarios project important changes in the regional structure of the demand (figures 3 and 4). While the demand in the Global North is projected to slightly increase or decline, in the Global South, the demand is projected to increase even under the most ambitious policy scenario (+27% compared to 2015). In the Global South, the demand for end-uses like space cooling or appliances and lighting, which run primarily on electricity, will represent a greater share of the demand than in the Global North. Consequently, the share of electricity in the Global South is expected to be larger than in the Global North (65%–82% compared to 42%–61% by 2050, respectively).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Following equation (3), figure 5 shows the decomposition of changes in emissions induced by buildings energy services across the four strategies identified above: reductions in the level of useful energy, improvements in the useful to final energy conversion, switch towards low-carbon energy carriers and decarbonisation of energy carriers. Figure 5(a) clearly shows that the large growth in useful energy of 160% in the 'Baseline' scenario between 2015 and 2050, which reflects the improvement of living standards, exerts an important pressure on emissions (+127%) which is largely compensated by the progress in efficiency (−85%). Overall however, emissions increase by 54%. When only efficiency barriers are lifted (figure 5(a), 'EG'), improved efficiency almost entirely compensates for the impact of useful energy demand growth that would otherwise double emissions. The influence of supply-side factors is much lower than that of demand side factors on the change in emissions in the 'Baseline' and 'EG' scenarios between 2015 and 2050.

Zoom In Zoom Out Reset image size

The opposite is true when considering the change in 2050 emissions between the 'Baseline' and 1.5°C scenarios (figure 5(b)). Demand side factors account at most for a decline in emissions of 18%, while measures decreasing the carbon content of energy contribute as much as a 74% decline in buildings emissions in the '1.5°C-EG' scenario. That is, in this scenario, reducing the carbon content of energy contributes a share of 81% of the total 91% reduction in emissions. According to the decomposition, the 31% reduction of final energy demand in '1.5°C-EG' compared to 'Baseline' translates into a 19% contribution to emission reductions. As the scenario 'EG', which only removes market failures in efficiency markets, does not envisage any policy targeted at CO₂ emissions, fuel switching and supply decarbonisation do not contribute at all to the decline in emissions in this scenario. Accordingly, the achieved emission reductions correspond to the level of energy demand reductions (−16% and −17%, respectively). A strategy that exclusively consists in removing barriers to energy efficiency is thus very unlikely to achieve emission reductions sufficient for the Paris climate targets.

Interestingly, despite the differences in the composition of demand in the Global North and Global South regions, carbon content reductions remain the primary driver of decarbonisation in both cases (figure 5(c)). Carbon content reductions contribute 74% of the decarbonisation in the Global North and 84% in the Global South. As the demand in the Global South is provided by a larger share of electricity, it is clear that the role of supply-side decarbonisation is more important than in the Global North where fossil fuels play a more important role. Conversely, in the Global North, demand reductions lead to a larger proportional emission decrease than in the Global South (23% against 15%). Overall however, despite the fact that emissions in the Global South are almost twice as important as in the Global North in the 'Baseline', in the '1.5°C-EG' scenario, they drop below the level of the Global North.

Looking into the details of the individual factors of equation (1) at the global level (figure 6), we observe that while the reduction in useful energy demand is limited to 10% and the reduction in energy intensity to 23%—leading to an aggregated drop in final energy demand of 31%—, the drop in the emission intensity factor reaches almost 90% in the '1.5°C' and '1.5°C-EG' scenarios. Several energy carriers can almost fully decarbonise (figure 7): electricity and district heating. Residual emissions therefore stem to a large extent from the remaining demand for gases, whose share is still 11% in the '1.5°C-EG' scenario (figure 8). By 2050, the share of electricity is already above 60% in the 'Baseline' and rises to 75% in the most ambitious scenario. Accordingly, 74% of all 2050 buildings emissions stem from electricity in the 'Baseline' (supplementary note 9), making the decarbonisation of electricity a fundamental requirement to reducing buildings sector emissions.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size