JSDEWES: Material efficiency and global pathways towards 100% renewable energy systems – system dynamics findings on potentials and constraints

For a projection of demand, the World Energy Outlook (WEO) 2015 by the International Energy Agency (IEA) [13] provided narratives and data. From the WEO, we chose the New Policies Scenario with ambitious assumptions regarding energy efficiency performance of the sectors transportation, industry, and housing for our model. In this scenario, the WEO simply considers sufficiency to mean less demand for energy from every sector, i.e., less transportation, less production, less energy for housing, and less need for electricity.

While the WEO only projects to the year 2040, the energy demand value for 2040 was used to simulate out to the year 2050. After 2050, the energy demand was kept constant until 2100 to look ceteris paribus at the continued demand for resources from repowering renewables every 20 years (for wind energy) or every 25 years (for photovoltaics).

The model simulates monthly time steps, capturing variation between winter and summer for the different regions. However, it uses a lookup function to consider the daily variations of sunshine and wind. This enables simulating increasing demand for energy storage and expansion of the electricity grid to allow electricity transport from sunnier and windier parts to other regions. Since with the rising proportion of renewable energy, the peak production exceeds the local demand and energy would otherwise remain unused.

The model includes hydropower and biomass according to the WEO, plus offshore and onshore wind and photovoltaics (PV) as renewable energy. The model considers a world market for power-to-liquid/gas (P2L/G) for the storage and re-electrification of surpluses of renewable energy and the remaining demand for synthetic fuels (also as a feedstock for the chemical industry),

The model differentiates global regions according to the WEO, except that Europe is separated into Germany and the rest of Europe so that the model can be validated with other studies on Germany's path towards renewable energy, in particular the later mentioned RESCUE study [14]–[16].

The raw materials needed for renewable energy generation are differentiated at the aggregated level into precious metals, semi-precious metals, rare earth metals, industrial minerals, and non-ferrous non-precious metals. Specific materials like copper, silver, neodymium, iron, and aluminium are explicitly modelled. Other sectors' demand for those materials was added to the model based on a review of relevant literature (e.g., [17], [18]).

The potential substitution of greenhouse-gas-intensive steel, concrete and gypsum through biotic materials can be based on various assumptions. On a highly aggregated level, simple substitution of the same mass (i.e., million tons by million tons) was used to derive the proportion of energy need relative to the total need for the industry sector that most heavily relies on these materials. The assumption is based on the argument from the steel and concrete industry [19] that they are using lightweight construction materials. Although their lifespan for a single application might be longer, one could use wood in a cascade, i.e., sequentially (re)used for different purposes, from massive wood to different varieties of chipboard and cellulose.

The capacities to implement renewable energy (production of renewable energy technology components; transportation; installation of renewable energy technologies; land area for installation) are set to very high levels so that they do not act as constraints. That allows for a focus on the potential needs for raw materials. However, one can use the model with more precise estimates of actual capacities for different applications.

As outputs, the model assesses the need for resources, the depletion of known reserves, the associated CO₂ emissions, and the economic effects on the different regions derived from energy production prices and regional value creation.

The ICARE Energy Model is quite comprehensive. It is highly aggregated to look at magnitudes and trends of developments but with robust quality. A greater level of detail would require more and significantly more complex assumptions, e.g. if the wood substitutes steel from windmills or aluminium from PV constructions.

The ICARE LULUCF Model assesses world regions according to the WEO using data from the Food and Agricultural Organization's (FAO) online database (http://www.fao.org/faostat/en/#data/QC). It assumes steady values of current land use for impermeable areas (i.e., urban), forest areas, pastureland, arable land, and other areas and divides agricultural land into conventional and organic agriculture. On a highly aggregated level, the potential need for food for current and future generations is set to 500 kg per year (according to FAO) [20], regardless of whether it is from animal products or plant-based. Each region's proportion of animal food then defines the area needed for fodder production. The yield for each region is the average taken from FAO data distinguishing between organic and conventional farming practices. As outputs, the model delivers the associated GHG emissions and the theoretical number of people that can be fed.

The model offers parameters to define future scenarios of afforestation or deforestation as well as increasing or decreasing shares of:

• organic farming,

• consumption of animal products,

• food waste,

• use of wood,

• use of yields for non-food purposes,

• sealing of land areas for settlements and infrastructure, and

• cultivation of yet unused areas.

Wood and crops for material use, and at the end of a cascade, their end-of-life energy use are considered regarding their effect on carbon capturing: first substituting fossil resources and materials (steel and concrete) and later releasing the CO₂ again. The substitution effect and the additional potential for biomass energy use are then transferred into the quantitative ICARE Energy Model.

To support the interpretation of the results from the ICARE Energy Model regarding winners and losers, the much smaller and very abstract ICARE Game-Theoretical Model looks at two fictional regions. One region is developing (called "elsewhere") and provides raw materials, while the other region is rich (called "here") and imports materials that support technology, including renewable ones for energy. Raw materials can deplete and be substituted by another material, possibly through recycling from anthropogenic stocks.

In both regions, there is economic activity through the production and consumption of material products and services, although the potential for consumption growth is greater in the developing region. Both regions also have jobs, purchasing power, and money creation. The latter takes place only in times of growth and is needed in the model because logically, without money creation, the simulation would reach an equilibrium.

The model runs one hundred abstract time steps. The units for the resources, wealth, purchasing power, etc., are also abstract index values used merely to explore the general logic of economic growth and its dependency on resources and value creation. The parameters to define and explore scenarios are the shift of value creation, recycling, material efficiency, productivity, sufficiency, financial investment activity, and material losses for both regions.

Using the qualitative model, we explored the mechanisms and interactions in the climate-resource-nexus based on literature review and expert interviews (see results published in [3]). Figure 1 shows how, starting from that, we developed three different, interlinked quantitative models to explore the global need for energy, the energy transition and its need for resources. The figure also shows how we added data from the ICARE LULUCF model on the potential of biotic resources and how we used these two models to run more than 15 scenarios. As an extra model, it shows the ICARE Game Theoretical Model that investigated economic effects and shifts in value creation via changes in abstract index values. Comprehensive model documentation is available upon request.

The results section first presents the need for resources for the global energy transition, followed by the potential effects of material efficiency and substitution of abiotic for biotic resources. Finally, the results from the game-theoretical modelling of economic growth are shown.

While the resources for a global energy transition would be available, another question is whether the economy could provide for all the materials, production capacities and transportation needed for this transition. The needed increase in installations, as Figure 6 shows, would be massive. Once demand rises, at least for a short period, prices will also rise posing the risk that poorer parts of the world will be unable to afford the transition and continue to use their own fossil energy sources.

Likewise, a constraint could arise due to the difference between 'early-adopter' countries that can use high-grade resources and those that start later and must use more expensive low-grade resources. Later, the access to recycled materials would again be easier for the 'early-adopters' since they repower their installations earlier in the simulation and thus obtain recycled materials from the repowered installations.

Figure 7 compares the four different scenarios for a 100% renewable energy world by considering North America's need to implement onshore wind capacities, which serves as a proxy for all technologies in all regions. According to IEA (2015) [13], the region has capacity targets set so that the mix of onshore/offshore and PV provides 100% renewable energy for the given energy demand. If demand decreases, the capacities for the mix decrease accordingly. Therefore, if a global change in demand through efficiency (scenario S5), sufficiency (scenario S6), or substitution by biotic resources (scenario S8) becomes possible, the difference can be shown with any of the technologies from any region since everything changes proportionally. There is one exception: the Middle East as a region offers a lower potential for biotic materials compared to the other regions.

Scenario S6 implies a further reduction of demand from all energy-related sectors (production, transportation, housing) through a change of lifestyle (sufficiency). Scenario 8 also implies a form of sufficiency since it requires less food waste and a reduction in animal-based food consumption.

Fostering material efficiency (scenario S5) seems to have a small effect on onshore wind capacity installation. It shows the corresponding need for energy rather than materials; however, this need is just a small part of the overall need from all other sectors. The main reason is that the WEO's forecast of energy demand already considered gains in material efficiency, and scenario 5 only adds some rather sophisticated efforts. The graph must be interpreted considering that it does not show ceteris paribus a constant demand with or without material efficiency, instead indicating a massive increase in demand from global economic growth. However, material efficiency remains important to prevent the constraints explained in the previous section. Also, the model considers a reinforcing feedback loop when less need for materials reduces the need for energy, leading to less need for renewables that, again, reduces the need for materials.

The substitution of biotic resources for parts of the global use of steel, concrete, and gypsum (scenario S8 in Figure 7) would lower the demand for renewable energy installations by roughly 10%. One reason is the substitution of GHG-intensive resources; the other is the end-of-life energy use of biotic materials that increases the renewable energy from biomass and thus reduces the demand for wind and solar energy.

Even with the assumption of 15% less energy demand globally (sufficiency assumption in scenario S2) compared to the projected energy demand from the WEO, most model simulations (except for scenario S6) reflect a narrative of continued economic growth in most world regions. Only the combined sufficiency and material efficiency scenario S6 with its additional 20% sufficiency could be interpreted as a step towards economic contraction. While Figure 5 showed that, at least for copper, material availabilities would not become a constraint before the year 2100, with unabated consumption, they are likely to become a constraint in the 22^nd century. Therefore, fostering a circular economy is crucial to satisfy raw material demands by increasingly recycling the anthropogenic stock.

However, out of the anthropogenic stock of materials, one can only retrieve what becomes available after the average lifecycle of products and structures. So, if the primary material used increases year by year, the world would inevitably face the risk of exhausting virgin raw material reserves while not yet having the quantities of recycled materials needed to satisfy demands. Therefore, material wealth would collapse in the long run even despite increased recycling efforts (Figure 10 using index values and an abstract time scale).

Because all four scenarios achieve a 100% renewable energy world with zero GHG emissions in 2050 or 2060, it is the path towards achieving zero GHG emissions that matters. Similar to other studies (e.g. [1], [2], [4]–[7], [17], [29]), the present findings show that material efficiency and recycling are key contributors to achieving GHG neutrality in two ways:

1. By reducing GHG emissions directly (avoided or reduced emissions from the production of primary raw materials and resource-efficient goods) or indirectly (reducing the need for energy associated with primary raw materials production);

2. By supporting the 100% renewable energy transition through increasing the availability of needed secondary raw materials, thus fostering resilience against increasing costs and scarcities of primary raw materials.

In the former context (1), the finding of around 7 Gt CO₂ emission reduction effects in 2050 of the material efficiency scenario S5 compared to a baseline with climate action (scenario S2, Figure 2) are of an order of magnitude similar to other studies. For instance, Hatfield-Dodds et al. (2017) [5] arrive at around a 3 Gt reduction in global GHG emissions in 2050 for their Efficiency Plus Scenario. This scenario combines resource efficiency improvements (through technology innovation, raw material taxation and demand changes through green public procurement and regulation) with GHG abatement policies and is compared to the effects of just the GHG abatement settings (the Ambitious Climate scenario). Meyer et al. (2018) [29] find around 2 million Gt reductions in global CO₂ emissions in 2050 under a resource-efficient scenario, which assumes 2% annual efficiency improvements in 75 key industry sectors in Germany and – with a time lag – also in other countries.

In the latter context described above (2), potential constraints regarding the availability of materials needed for the renewable energy transition make a case for keeping materials in cycles longer and recovering materials from existing renewable energy installations [17], [26]–[28]. Hence, circular economy approaches are considered a key lever to support the renewable energy transition globally [30] and in different world regions and countries, for instance, Africa and China [31]–[34]. Several raw materials critical for the renewable energy transition are increasingly considered scarce because of depleting high-grade reserves, making them more expensive [26], [28], [35]. Hence, in addition to improving the availability of (secondary) raw materials needed for the energy transition, recycling also brings down costs associated with raw material needs [36]–[38]. Thus, recycling and material substitution could alleviate price shocks and foster capacities to pursue the renewable energy transition. Additionally, recycling and substitution can diversify the material stocks needed for this transition, e.g., from abiotic to biotic, which might also render national economies more resilient and less dependent on imports.

This shows that pathways towards a 100% renewable energy world by around 2050 seem to be most promising when they increase the resilience of regions or countries towards external shocks related to price or import dependency. Not only are circular economy approaches and recycling relevant in this context, but also reductions in overall energy and material needs without compromising global capacities for socio-economic well-being and justice.

Reducing overall energy demand, even if derived from renewable energy, is essential for at least three reasons. First, there are environmental impacts from PV and wind energy generation installations. For instance, PV installations, which are often located in arid or semiarid regions, can negatively affect soil structure, increase soil loss through floods or storms, and impact local biodiversity, e.g., by blocking migration routes [39]. Wind turbines can negatively affect wildlife and humans regarding mechanical and aerodynamic noise and visual impact [39]. However, environmental impacts of renewable energy generation (in particular regarding GHG emissions and land use, but also largely for water consumption) are perceived to be much lower than those of fossil energy generation [39]–[41].

Second, reducing overall energy demand can also mitigate the need for additional renewable energy capacities. These might prove challenging regarding the scale of needed investments in different country contexts (e.g., [42]) and potentially increased raw material prices. Arising from there are potential constraints to the renewable energy transition if the needed increase in capacities becomes technically or economically unrealistic (see, e.g., [43]) or if repowering is hampered by issues of acceptance or changing access to land for renewables (see, e.g. [44]). International policy support should assist countries in transitioning toward renewable energy systems and mitigating raw-material price increases for relevant technologies.

Lastly, and most importantly, only by reducing overall energy demand will it be possible to achieve and maintain atmospheric GHG concentration compatible with a 1.5-degree world. The required concentration is a maximum of 430 ppm, equivalent to a GHG amount of approximately 3,360 Gt by 2100 (with 128 ppm CO₂ equaling approximately 1,000 Gt of CO₂, see, e.g., [45]–[47]). In December 2021, atmospheric concentrations were already around 417 ppm [48] or approximately 3,258 Gt, leaving only little room for adding further to the stock of CO₂ in the atmosphere. None of the four scenarios modelled in this study arrives at a value lower than 3,700 Gt accumulated globally by 2100. Therefore, even greater reductions in energy demand or earlier achievements of 100% renewable energy appear necessary to stay within a 1.5-degree world, in addition to the modelled combined sufficiency and material efficiency scenario (scenario S6 in Figure 2).

An essential pillar of achieving a 1.5-degree world while delivering on global just socio-economic well-being is reducing CO₂ emissions associated with human lifestyles (see, e.g., [24]), particularly in the Global North. The ongoing major environmental issues stem from the Western lifestyles of affluent households and individuals, particularly housing, mobility, and food choices (see, e.g., [49]–[52]). Investigating lifestyles for a 1.5-degree world, Akenji et al. (2021) [24] argue that high-income countries need to reduce their carbon footprints, mostly relating to food, housing, and transport, by between 69–82%, while upper- and lower middle-income countries need reductions of 23–50% by 2030. Changes in lifestyles are essential to achieve climate neutrality alongside a 100% renewable energy world because they are needed to "keep the demand for renewable energy within acceptable boundaries" ([53], p. 418) and to reduce interlinked energy and material needs as well as GHG emissions.

According to Akenji et al. (2021) [24], lifestyle changes that could achieve such reductions in carbon footprints encompass car-free travelling and transport, minimisation of international flights, vegan, vegetarian or low-carbon protein diets, and renewable energy-based heating and/or cooling. Reduced per capita floor space in housing is also discussed (e.g., [7], [54], [55]).

Therefore, changing lifestyles towards an absolute reduction in the use of energy, land, materials and other natural resources offers great potential to help stay within an environmentally safe and socially just planetary space while ensuring a decent standard of living for humanity [24], [47], [52], [56].

Four different scenarios for achieving a 100% renewable energy world by 2050 were modelled. The model findings show that while a 100% renewable energy world with zero GHG emissions seems feasible, it is important to design a transition pathway that increases the availability of secondary materials for renewable energy generation and fosters value creation in developing countries. In this context, improving recycling is an essential milestone to provide secondary materials and thus reduce the risk of economic constraints related to depleting high-grade raw material reserves (associated with increasing costs). Therefore, material efficiency improvements are an important contribution to 100% renewable energy generation and climate mitigation.

Potentials for GHG emission reductions were greater in material substitution scenarios than in material efficiency scenarios – particularly since substituting abiotic for biotic materials increases sink functions. However, changing diets and reducing food waste are essential to avoid competition between biomass for food vs. material or energy use.

Here, lifestyle changes become relevant, which reduce energy and material demand across different sectors. These reduce GHG emissions to allow staying below 1.5-degree warming. Hence, sufficiency-oriented lifestyle changes are key to reducing energy use and GHG emissions associated with human food, housing, mobility, and product choices. Regarding just and decent living standards worldwide, such lifestyle changes should reduce present inequalities and allow for needed absolute reductions in the Global North while enabling a sustainable catching-up of socio-economic development in the Global South. From a global perspective, these efforts require avoiding an over-reliance on market forces only and creating synergies among paths that improve material efficiency, material substitution, and lifestyle changes. Otherwise, unrestricted free-market policies would make industrialised countries short-term winners and humanity a long-term loser. How the feasible pathways may look like and what mechanisms could be used to forge such international policy requires further investigation to support the needed transition. Helping other countries establish and foster a secondary raw materials economy for domestic use and export should form part of international policy. This trend would allow material efficiency and substitution to lead toward more ambitious GHG emission reductions.