Supply-side options to reduce land requirements of fully renewable electricity in Europe

Tim Tröndle

doi:10.1371/journal.pone.0236958

. 2020 Aug 6;15(8):e0236958. doi: 10.1371/journal.pone.0236958

Supply-side options to reduce land requirements of fully renewable electricity in Europe

Tim Tröndle ^1,^2,^*

Editor: Baogui Xin³

PMCID: PMC7410258 PMID: 32760117

Abstract

Renewable electricity can fully decarbonise the European electricity supply, but large land requirements may cause land-use conflicts. Using a dynamic model that captures renewable fluctuations, I explore the relationship between land requirements and total system cost of different supply-side options in the future. Cost-minimal fully renewable electricity requires some 97,000 km² (2% of total) land for solar and wind power installations, roughly the size of Portugal, and includes large shares of onshore wind. Replacing onshore wind with offshore wind, utility-scale PV, or rooftop PV reduces land requirements drastically with only small cost penalties. Moving wind power offshore is most cost-effective and reduces land requirements by 50% for a cost penalty of only 5%. Wind power can alternatively be replaced by photovoltaics, leading to a cost penalty of 10% for the same effect. My research shows that fully renewable electricity supply can be designed with very different physical appearances and impacts on landscapes and the population, but at similar cost.

1. Introduction

Europe has the potential to generate all its electricity from renewable sources [1–3]. The potential provides a possibility to decarbonise the European electricity system, which is a necessary step to reach the European Commission’s goal of becoming a climate-neutral economy by 2050 [4]. Compared to the predominant forms of electricity supply based on fossil and nuclear fuels, land requirements of renewable electricity are high [5–8], however. A transition to renewable electricity will therefore increase the total land requirements of electricity supply and it may even do so by orders of magnitude.

While renewable electricity is an indispensable option to mitigate global climate change, its high land requirements have the potential to cause conflicts locally where it is built. This is for three reasons. First, it may compete with other uses of land. Of the main two current technologies of renewable electricity, photovoltaics and wind turbines, only the latter allows for limited dual use of land: for technical reasons, spacing between turbines is large and, as a result, that land can be used for agriculture [7]. Second, renewable generation infrastructure has the potential to economically devalue land on which it is built, and also neighbouring land. There is conflicting evidence whether wind power in sight of properties impacts property values, and while the majority of studies do not find statistically significant impacts, some others find losses in property values of up to 15% [9]. Third, wind [10] and solar power [11] are sometimes perceived as negatively impacting the landscape, depending on place attachment and the aesthetics of the previously undisturbed landscape [10,12,13], and location and density of structures [10].

While the acceptance of the energy transition is generally high and the majority of the population does not feel disturbed by wind and solar installations [14,15], local opposition has hindered and delayed local renewable electricity projects in the past [16,17]. Opposition may continue in the future, considering the large expansion of impacted land area moving towards fully decarbonised electricity supply [15]. This led some authors to the conclusion that fully renewable electricity–while being theoretically possible–will not be feasible in Europe [5,7].

Strategies to mitigate negative impacts associated with the land requirements of renewable electricity include location and placement of generation infrastructure [10] and technology choice to reduce land requirements [18]. If proven effective, these strategies can not only reduce negative side-effects, but also increase the feasibility of electricity systems with large shares of renewable electricity by reducing opposition.

In this article, I explore the relationship between land requirements and total system cost of fully renewable electricity systems in Europe with different supply sides. Renewable supply technologies have vastly different land requirements, with two orders of magnitude between the land requirements of bioenergy with dedicated farming for crops, the technology with the largest land requirements, and solar electricity, which has the lowest [7,8]. Previous research shows that fully renewable electricity supply in Europe is possible in many different ways, with very different shares of supply technologies, and that cost differences between supply options can be low if designed right [19–21]. However, while several studies have assessed cost [1,22–26] and land requirements [1,3,27,28] of supply technologies and entire electricity systems, only one study has assessed the relationship between the two on the system level [18]. The system perspective is central to renewable electricity systems as it takes into account not only the supply side but also technologies to handle fluctuations of the supply side. In their case study of Alberta, Canada, the authors find higher total system cost for lower land requirements even though they allow for large amounts of electricity from non-renewable sources. No study has assessed the relationship between land requirements and cost on the system level using only renewable resources in Europe. I address the relationship in this study by determining cost-effective ways to reduce land requirements of fully renewable, future electricity systems in Europe through supply technology choice.

To find cost-effective ways to reduce land requirements, I do the following: I use a nationally resolved, dynamic model of the European electricity system to determine total system cost and land requirements of fully renewable supply. I find that, while there is a trade-off between cost and land requirements, systems with vastly different requirements for land can be built with small cost penalties.

2. Data and methods

To identify cost-effective ways to reduce land requirements by supply technology choice, I generate total system cost and total land requirement data for fully renewable electricity supply in Europe with different shares of supply technologies using a model-based approach. I generate the data in two steps. First, I generate cost-minimised system designs for 286 different shares of supply technologies. Second, I determine total system cost and land requirements for each system design. Using the Monte Carlo method respecting uncertainty of technology parameters, I create 100,000 samples for each system design. In total, I end up with ~29 million observations of pairs of cost and land requirements. All data procedures and all analysis steps are publicly available [29] as a Snakemake workflow [30].

The following subsections explain all data generation and analysis steps in more detail.

2.1 Capacity shares of supply technologies

To understand how supply technology choice can mitigate land requirements of renewable electricity in Europe, I enforce different capacity shares of technologies in the system designs that I am analysing. The geographic scope of this study includes most countries with member organisations in the entso-e: EU-27 without Cyprus, the United Kingdom, Switzerland, Norway, and Western Balkan countries. I focus on four dominant wind and solar supply technologies: on- and offshore wind, and ground- and roof-mounted photovoltaics (PV). I analyse all 286 possible combinations based on ten different shares per technology, from 0% through 10% and 20% up to 100% (Fig 1). The shares are applied to the European level, but also to the national level, meaning that each country in Europe has to meet shares individually. Furthermore, I assume each country to be net self-sufficient, generating enough electricity annually to fulfil its domestic electricity demand but able to trade with other countries to balance renewable fluctuations.

Fig 1 — Pixels in each panel represent one system design. System designs are in no particular order. Shares of the same pixel in all panels always add to 100%.

Other than these four supply technologies, the system designs furthermore contain hydroelectricity with and without reservoirs and bioenergy, all of which can generate renewable electricity to a limited extent as well. While I do not restrict their capacity shares, they are both restricted by their generation potential: hydroelectricity is limited to the amount that can be generated using current capacities, and bioenergy is limited to the amount of bioenergy that can be generated from residuals (see System design model). Both contribute to electricity supply, but only to minor extents. Hydroelectricity with reservoirs and bioenergy are used to balance renewable fluctuations as well.

Each country in Europe can potentially generate enough electricity from rooftop PV, utility-scale PV, and onshore wind to cover national demand together with limited generation from hydroelectricity and bioenergy (see “System design model” and “Discussion”). Thus, system designs with 100% rooftop PV, 100% utility-scale PV, or 100% onshore wind are possible. The situation is different for offshore wind: while the generation potential for all Europe is large enough to cover European demand, only countries with shores can build offshore wind farms. In countries without shore I replace offshore wind with onshore wind; i.e. when all countries with shores have to enforce a 40% capacity share of offshore wind, and a 20% capacity share of onshore wind, all countries without shores have to have a 60% capacity share of onshore wind. Countries without shores, or with insufficient offshore potentials are: Austria, Bosnia and Herzegovina, Switzerland, Czech Republic, Hungary, Luxembourg, North Macedonia, Serbia, Slovakia, and Slovenia.

Based on the enforced supply capacity shares, the system design model determines absolute installed capacities in each country for all supply technologies, but also for all storage and cross-border transmission capacities.

2.2 System design model

The system design model determines cost-minimal system designs for Europe. The model is a network flow model with the electricity transmission grid at its core [31]. Each country in Europe is modelled as a node on the network and all nodes are connected through the transmission grid. The model has a 3h resolution and simulates one full weather year to cover renewable fluctuations. It is a linear optimisation model that optimises system design and operation simultaneously. The model is implemented using the Calliope model framework [32] and has been used in a former publication. It is described in full detail in ref. [21].

On the supply side, the model contains four main renewable technologies to generate electricity: on- and offshore wind, and utility-scale and rooftop PV. In addition, hydroelectricity with and without reservoirs, and bioenergy can generate electricity. Capacities are limited by their technical potentials, which I derive from ref. [3] for wind and solar power, and from ref. [21] for hydroelectricity. For hydroelectricity, I assume no further expansion from today is possible, and thus its technical potential equates to today’s capacity (see Section 1 in S1 File for a discussion of the impact of this decision). For wind and solar power, I allow any capacity up to their technical potential. Generation profiles are based on weather data [33–35] and taken from the same sources as the potentials. Based on the enforced supply capacity shares, the system design model determines cost-minimal capacities of supply technologies.

To balance fluctuating hydro, solar, and wind generation, system designs contain battery storage, hydrogen storage, pumped storage hydroelectricity, and bioenergy. All storage technologies are modelled as single storage tanks with efficiencies, i.e. there is no flow of energy in any other form than electricity. Battery storage can discharge for a maximum of four hours at full power, while hydrogen storage can discharge for at least four hours. Bioenergy is limited by the amount of fuels that can be produced from residuals in each country per year [36], i.e. I do not assume dedicated farming for energy crops. The limited fuel supply and high fuel cost make bioenergy a technology mainly used for balancing, rather than for supplying electricity. Bioenergy and storage capacities other than pumped hydro are not restricted in any way. I assume pumped storage hydroelectricity can not be expanded significantly and it is thus limited to today’s capacities [21]. The system design model determines cost-minimal capacities of all balancing technologies.

All supply, balancing, and cross-border transmission capacities have costs: overnight installation costs, annual maintenance costs, and variable costs (Table 1). The technology costs are important determinants of the magnitude of capacities chosen by the system design model. Their values are future projections and assume all technologies are deployed at large scale. In particular, this means that I ignore any forms of transitional effects stemming from technological or financial learning. Together with the technology lifetime, I determine annuities from these cost components. I uniformly assume cost of capital to be 7.3%, which has been the historic average [37].

Table 1. Assumptions on installation and maintenance cost of electricity infrastructure.

Technology	Overnight cost (€/kW)	Overnight cost (€/kWh)	Annual cost (€/kW/yr)	Variable cost (€ct/kWh)	Lifetime (yr)	Source
Utility-scale PV	520	0	8	0	25	Ref. [38] Table 7
Rooftop PV	880	0	16	0	25	Ref. [38] Table 9
Onshore wind	1100	0	16	0	25	Ref. [38] Table 4
Offshore wind	2280	0	49	0	30	Ref. [38] Table 5
Biofuel	2300	0	94	6	20	Ref. [38] Table 48, ref. [38]
Hydropower run of river	0	0	169	0	60	Ref. [38] Table 14
Hydropower with reservoir	0	0	101	0	60	Ref. [38] Table 12
Pumped hydro storage	0	0	7	0	55	Ref. [39]
Short term storage	86	101	1	0	10	Ref. [39]
Long term storage	1612	9	14	0	15	Ref. [39]
AC transmission^{^}	900	0	0	0	60	Ref. [38] Table 39

Open in a new tab

^AC transmission overnight cost is given in €/kW/1000km.

While future cost is uncertain, I am using expected values in the deterministic system design model. To cover the aspect that future cost is not known exactly, I handle cost uncertainties in the generation steps that follow the system design phase.

2.3 Cost uncertainty

Cost of almost all components of future renewable electricity systems can be expected to fall compared to today. Cost falls with deployment as production processes get improved, product understanding increases with the use, and financing can be provided with lower overheads. Exactly how much cost will fall with deployment is not known, however. To cover this uncertainty, I am using minimum and maximum estimates of cost [38] of the four supply technologies analysed in this study. Because I do not have any other information about how likely any cost developments are, I am following the principle of maximum entropy and I am assuming a uniform distribution between minimum and maximum estimates, see Table 2.

Table 2. Uncertain input parameters.

Parameters with uniform distribution are represented by their min and max values. Parameters with normal distribution are represented by their mean and standard deviation.

Name	Description	Distribution	Min/Mean	Max/Std	Source
Cost onshore wind	Cost scaling factor of onshore wind.	uniform	0.727	1.545	Ref. [38] (Table 4)
Cost offshore wind	Cost scaling factor of offshore wind.	uniform	0.785	1.434	Ref. [38] (Table 5)
Cost rooftop PV	Cost scaling factor for rooftop PV.	uniform	0.864	1.136	Ref. [38] (Table 9)
Cost utility-scale PV	Cost scaling factor for utility-scale PV.	uniform	0.538	1.115	Ref. [38] (Table 7)
Land requirements wind	Onshore wind capacity density [W/m²].	normal	8.820	1.980	Ref. [8]
Efficiency utility-scale PV	Module efficiency of utility-scale PV.	uniform	0.175	0.220	Ref. [40]
Land requirements utility-scale PV	Share of land that is covered by PV modules.	uniform	0.400	0.500	Refs. [7, 41, 42]

Open in a new tab

I only consider uncertainty in cost of on- and off-shore wind, and utility-scale and rooftop PV, first because these are the technologies whose cost-effectiveness I am assessing in this study. Second, because cost has little impact on the system design. Because I enforce supply shares in the system design model, the extent to which supply technologies are deployed is determined by the enforced shares, not by cost. Changes in cost of other technologies could lead to different designs, for example if hydrogen storage cost is much higher than the expected value, hydrogen may be replaced with bioenergy. For this reason, I use expected values only (see Table 1) for all other components other than the four wind and solar supply technologies.

2.4 Land requirements

To determine land requirements of supply technologies, I assume capacities of technologies always to require the same amount of land and therefore apply a proportional constant to installed capacities: the inverse of capacity density, given in square meters per Watt. As the range of capacity density values given in the literature is large for onshore wind and utility-scale PV, I am using a stochastic approach here as well.

Land requirements of onshore wind in this study are those of the wind turbines together with the technically necessary spacing between turbines. While the spacing can be used for agriculture, it excludes other land uses and the spacing also does not reduce visual impacts. Thus, I include spacing in the land requirements of wind turbines in this study (see Section 3 in S1 File for an analysis excluding spacing). The land requirements estimates of wind and solar power furthermore include all additional infrastructure necessary: substations, access roads, and service buildings.

Theoretically, the capacity density of onshore wind can be high: based on technical specifications, it is up to 19 W/m² for the best turbines and ~10 W/m² on average [43]. However, in deployed wind farms, the capacity density is lower, with values between 2–10 W/m² [7,8,44,45]. This can have different reasons: the capacity density depends on the placement of turbines in the wind park and it depends on the technical specifications of the wind turbine. More importantly, because land is not the most important cost component of wind farms, farms are not necessarily build in a way to maximise capacity density. In this study, I am using a capacity density estimate taken from ref. [8] based on measurements in the US. Here, the authors found 8.8 W/m² on average with a standard deviation of ~2 W/m², see Table 2.

Uncertainty about land requirements of utility-scale PV is similarly high. Around 40%–50% of the area of solar farms is covered by modules [7,41,42], while the rest of the land is used for inverters, power lines, spacing, and roads. In addition, the technology used, orientation, and efficiency of the PV modules have great impact on land requirements as well as they determine the capacity installable on the area covered by modules. Theoretically derived capacity densities using today’s module efficiencies are in the range of 80–100 W/m² [42], but measurements from the US show much lower estimates of 20–30 W/m² only [7,8,44]. A recent study from Germany shows, however, that the capacity density of German utility-scale PV has increased drastically over time: in less than 20 years, it increased by factor 3 to ~70 W/m² in 2018 [46]. The authors explain the trend not only by increasing module efficiencies, but also by more economic use of land. These findings show that theoretically derived capacity densities may actually be more accurate for future projections than historic measurements, and I am therefore using capacity densities derived from theory. I am assuming that land is covered to 40–50% by modules [7,41,42], and that module efficiency is between 17.5–22% [40], see Table 2. Assuming a uniform distribution for both, this leads to an expected capacity density value of utility-scale PV of ~88 W/m².

The remaining two supply technologies, rooftop PV and offshore wind, have no land requirements whatsoever. Rooftop PV is built on existing structures, and offshore wind is not built on land. This makes them promising options for reducing total land requirements of renewable electricity systems.

Hydroelectricity, bioenergy, storage, and transmission will all require land which I am not analysing in detail in this study. The reasons why I ignore land requirements of each of the components are the following: First, I do not analyse land requirements of hydroelectricity in detail, because system designs in this study all contain the exact amount that is installed today. Thus, hydroelectricity’s land requirements in all cases are the same, and equal to today’s. Today’s requirements are not small, however. They are dominated by the extent of water reservoirs [7], which span roughly 50,000 km² (1% of total land) in Europe [47]. Not all of the reservoirs in Europe are used for electricity generation, and even less are used exclusively for it [47], so this total number can be seen as an upper bound of the land requirements of hydroelectricity, even after dam, power house, and access roads are added.

Second, land requirements of bioenergy are very small as long as only residuals are used for fuel production. When dedicated energy crops are farmed, bioenergy has the lowest capacity density of all renewable technologies [8]. The by far largest contribution to its land requirements stems from fields for crop farming, however. Because I allow only residuals to be used for electricity generation, land requirements include the power plants only, which leads to a capacity density in the order of 10⁴ W/m² [7] and thus 2–3 orders of magnitude larger than solar and wind power. This makes bioenergy’s contribution to total system land requirements insignificant and I am therefore ignoring it in this study.

Third, land requirements of electricity storage depend on the amount of electricity that must be stored. Commercial suppliers offer 1 MW / 1 MWh battery storage systems in standardised container enclosings today [48], leading to a capacity density in the order of 10⁵ W/m² and 10⁵ Wh/m². Power-wise, a capacity density of this magnitude makes the land requirements of battery storage insignificant compared to the one of solar and wind power, even if spacing, roads and further infrastructure must be added. Energy-wise, capacity density cannot be compared to solar and wind power, and the total land requirements depend on the amount of electricity that must be stored in batteries. This is equally true for hydrogen storage. Here, the energy-wise capacity density depends on how hydrogen is stored. Hydrogen has a low energy density of 3 kWh/m³ if stored uncompressed at normal conditions. It can be stored underground in salt caverns, or overground in steel tanks. Capacity density is lowest if hydrogen is stored in such overground tanks in uncompressed form. Together with a conservative height of the tanks of 2 meters, this equals 6,000 Wh/m². This conservative estimation is worse than the one for battery storage. Because much more electricity is anticipated to be stored as hydrogen rather than in batteries, total land requirements of hydrogen storage may be high, if it is not stored in compressed forms, in tanks taller than 2 meters, or underground, and if large amounts must be stored. The latter is not the case for the system designs I am considering in this study (see “Storage and flexible generation require small amounts of land” in results section), and thus I am ignoring land requirements of battery storage and hydrogen storage.

Lastly, the transmission grid already has significant land requirements, which will likely increase in fully renewable systems. Currently, there are 220,000 km of AC transmission lines in the study area (based on a model of the transmission grid originally created in ref. [49] and extended in ref. [50]). With an estimated 13.5 m buffer zone on each side [51], this leads to 6,000 km² (0.12%) of land required. I am not assessing the land requirements of transmission in the fully renewable scenarios of this study because its spatial resolution, the national level, is too low to determine necessary land for transmission. Two former studies with subnational resolution indicate, however, that fully renewable electricity requires only 20% larger transmission capacity when countries are net self-sufficient [21], as they are in this study, and that a transmission system twice as large as today’s or larger is likely not beneficial, even if countries are not self-sufficient [52]. The land requirements of the transmission grid in a fully renewable electricity future is therefore likely well below 0.3% of total land.

2.5 Stochastic model

I use technology cost and technology land requirement parameters to derive total system cost and total land requirements of solar and wind power in all 286 system designs stemming from the system design phase. I do this in two steps. First, I sample 100,000 times from the input uncertainties using Saltelli’s extension of the Sobol sequence [53] to derive a sufficiently large sample set of the seven-dimensional input space. Second, for each sample and each system design I derive total land requirements of solar and wind power by multiplying their installed capacities with the inverse of their capacity density. Similarly, I derive total system cost by scaling technology cost of solar and wind from the system design with the factors given from the input sample. This leads to 100,000 output observations for each system design and therefore to ~29 million observations of pairs of cost and land requirements, which I am using to analyse cost effectiveness of different supply technologies.

3. Results

3.1 Renewable electricity supply with vastly different land requirements

Among almost all of the ~29 million observations, cost of electricity in all Europe is between 0.06 and 0.10 EUR per kWh consumed while land requirements of solar and wind power are between 0% and 3% of total European land (Fig 2). The observations show that land requirements of European electricity systems can vary by more than an order of magnitude while their cost does not exceed twice the lowest cost.

Fig 2 — The observations include all possible supply share combinations, including systems supplied, apart from hydroelectricity and bioenergy, exclusively from onshore wind, utility-scale PV, and rooftop PV, or with high shares of offshore wind. The ranges furthermore contain technology cost and technology land requirement parameters from the full range of their uncertainty. a, System cost relative to electricity demand. b, Land requirements of solar and wind power, relative to total land in Europe.

When I reduce uncertainty distributions to their expected values (their means), I find that there is a trade-off between expected land requirements of renewable electricity and its expected cost. Among all 286 system designs with different supply shares, a system with only onshore wind and utility-scale PV has the lowest expected cost of around 0.07 EUR per kWh consumed and requires ~2% of Europe’s total land (~97,000 km²)–an area roughly the size of Portugal (see S1 Fig in S1 File for data on the national level). Cost is minimal when both technologies contribute 50% to the total capacity of wind and solar technologies. While higher shares of utility-scale PV decrease land requirements, they also increase cost (right flanks in Fig 3A and 3B). Higher shares of onshore wind increase both cost and land requirements. A system design with only onshore wind has the highest expected land requirements (top corner of Fig 3B). Rooftop PV has the largest potential to decrease land requirements, as it requires no additional land, but it also increases cost the most (left corners of Fig 3A and 3B).

Fig 3 — Cost and land requirements are relative to the case with minimal cost of ~0.07 EUR per kWh consumed which requires ~2% of Europe’s total land. All cases include hydroelectricity of today’s capacity and bioenergy from residuals next to three solar and wind technologies. Expected values are the means of uncertainty distributions. **a,b,** Total system cost (a) and land requirements (b) of cases with utility-scale PV, onshore wind and rooftop PV as supply side options. **c,d,** Total system cost (c) and land requirements (d) of cases with utility-scale PV, onshore wind, and offshore wind as supply side options.

Electricity system designs with offshore wind in addition to onshore wind and utility-scale PV have lower cost when they do not include rooftop PV (Fig 3C). The potential of offshore wind to decrease land requirements is smaller than the one of rooftop PV, but only slightly (Fig 3D). While offshore wind requires no additional land–similar to rooftop PV–it is not available in every country in Europe and is replaced by onshore wind in these places. Compared to onshore wind, the land requirements of all other three supply technologies–offshore wind, utility-scale PV, and rooftop PV–are lower and thus, system designs with large shares of any of these alternatives have smaller total land requirements, albeit at higher cost.

3.2 Offshore wind reduces land requirements most cost-effectively

The rather large expected land requirements of the cost-minimal case can be reduced most cost-effectively by replacing onshore wind with offshore wind. In this way, total land requirements of renewable electricity can be reduced by 50% (to 1% of total land) for a cost penalty of 5% (Fig 4A). This cost penalty corresponds to 0.22 EUR per m² and year and comes at a share of offshore wind of ~25%. Land requirements can be decreased further, in total by 85%, with higher shares of offshore wind. However, cost increase sharply for the last minor reduction in land, for which utility-scale PV must be phased out (left-most point in Fig 4A).

Reducing land requirements with utility-scale (Fig 4B) and rooftop (Fig 4C) PV has higher cost. To reach the same reduced land requirements of 50% below the cost-minimal case higher shares of utility-scale PV lead to a cost penalty of ~9%, corresponding to 0.41 EUR per m² and year. Cost rises progressively however, and early decreases of land requirements have very low cost. In total, 80% of the cost-minimal land requirements can be removed with a cost penalty of 23% using utility-scale PV only.

The highest expected cost comes with phasing-in rooftop PV. Here, reducing land requirements by one square meter requires 0.75 EUR per year when reducing cost-minimal requirements by 50% (to 1% of total land, see Fig 4C)–this is a cost penalty of 17%. Similar to utility-scale PV, the cost increases with higher rooftop shares and the largest increase can be explained by the technology that is phased-out: the first half of rooftop PV replaces onshore wind, while the second half replaces utility-scale PV at a much higher cost. The increase of offshore wind, utility-scale PV, and rooftop PV shares always reduces expected land requirements of fully renewable electricity systems, albeit at different cost.

3.3 Cost penalties of 20% or less are most likely even for low land requirements

Uncertainty in technology cost and land requirements leads to high uncertainty in the cost penalties for renewable electricity with lower land requirements. To ensure land requirements are below 1% of total European land, cost penalties can be as large as 40%, but are most likely below 20% for all supply technology options (Fig 5). For offshore wind and utility-scale PV a cost penalty below 20% can be expected in 75% of the cases. In a quarter of all cases, there is no cost penalty necessary at all, because the cost-optimal case has land requirements of 1% or lower. With lower thresholds, cost penalties become more likely and also larger. For a threshold of 0.5% of European land, a cost penalty of 20% is still more likely for the more cost-effective technologies offshore wind and utility-scale PV.

Fig 5 — Cost penalties arise from higher than cost-minimal shares of one of three supply technologies: offshore wind, utility-scale PV, or rooftop PV. The uncertainty distribution of cost penalties is displayed using letter-value plots. Letter-value plots are an extension to boxplots for large data. Dark grey lines indicate the median value of the cost penalties, and the widest boxes above and below the median visualise the 25–75% quantiles. Each following box contains half as many observations as the box closer to the median. The extreme 1% of the observations are considered outliers and marked with rhombs.

Uncertainty does not alter the order of cost-effectiveness of the three supply technologies with low land requirements: rooftop PV is always the least cost-effective technology. Offshore wind is most cost-effective, but only when large amounts of onshore wind are to be replaced (land area thresholds of 1% or smaller in Fig 5). In these cases, offshore wind is more cost-effective than utility-scale PV. For medium land thresholds (1.5%), expected value of cost and its distribution are nearly the same for both technologies. Above that, as long as land area is to be reduced only little, utility-scale PV is the most cost-effective option.

3.4 Low land requirements require low shares of onshore wind

While land needs of supply technologies are uncertain, onshore wind is in any case the technology with the highest requirements for land if spacing is included (see Section 3 in S1 File for an analysis excluding spacing). Offshore wind, utility-scale PV, and rooftop PV are therefore no-regret options to reduce the spatial extent of renewable electricity generation on land. In 50% of the cases, a land threshold of 1% of total European land can only be reached if the capacity share of onshore wind is 40% or lower (Fig 6), and if there are no additional land requirements from utility-scale PV. When utility-scale PV exists as well, onshore wind capacity must be even lower, and if utility-scale PV is the only alternative, onshore wind capacity must be as low as 10%. Renewable electricity with low requirements for land can only be reached by low shares of onshore wind.

Fig 6 — Visualised shares are the maximum shares among system designs with only utility-scale PV or only rooftop PV and offshore wind in addition to onshore wind and given uncertainty. Uncertainty stems from the uncertainty of how much land onshore wind and utility-scale PV require. The uncertainty distribution of capacity shares is displayed using letter-value plots. Letter-value plots are an extension to boxplots for large data. Dark grey lines indicate the median value of the cost penalties, and the widest boxes above and below the median visualise the 25–75% quantiles. Each following box contains half as many observations as the box closer to the median. The extreme 1% of the observations are considered outliers and marked with rhombs.

3.5 Storage and flexible generation require small amounts of land

Systems with different shares of solar and wind capacity require different balancing capacity in terms of electricity storage, bioenergy, and transmission. Balancing needs are moderate for cases with balanced mixes of supply technologies (Fig 7). When supply is strongly biased towards one technology, flexibility needs rise, and in some cases, they rise sharply. Exclusively- or almost exclusively-solar cases require high amounts of short-term (battery) electricity storage. In extreme cases, storage capacities alone are able to fulfil the largest part of Europe’s peak demand. Short-term storage capacities in these cases are combined with very high magnitudes of bioenergy capacity of up to 50% of peak demand to balance solar’s seasonal fluctuations. Cases with mainly wind require much less bioenergy capacity and short-term storage capacity, but more long-term storage capacities to balance wind fluctuations between days and weeks. In addition, they require around 2.5 times larger cross-border transmission capacity than more balanced systems. While some of these numbers are very high, especially for cases with single supply technologies, there is no reason to believe these balancing capacities could not be built.

Fig 7 — All designs are exclusively supplied from hydroelectricity of today’s capacity and different shares of three additional technologies each: onshore wind and utility-scale PV in all cases, combined with either rooftop PV (top row) or offshore wind (bottom row). Each technology is varied from 0–100% of the total capacity of the three technologies. **a–d,** Storage power capacity (a), storage energy capacity (b), cross-border transmission capacity (c), and bioenergy capacity (d). Not shown are hydroelectricity capacities which are kept constant in all cases (36 GW run of river, 103 GW / 97 TWh reservoirs, 54 GW / 1.3 TWh pumped hydro storage).

Balancing capacities require land as well and thus add to the land requirements of the entire electricity system. In its current state, the transmission grid uses less than 0.2% of total land (see Methods). For the system designs in this study I can not determine land requirements of the transmission grid, as the spatial resolution is too low to generate estimations.

The land requirements of all other balancing technologies are very small, however. When considering 10⁵ Wh/m² for battery storage capacity, a conservative 6,000 Wh/m² for hydrogen storage capacity, and 10⁴ W/m² for bioenergy capacity (see Methods), total land requirements of all three flexibility technologies are always below 1,800 km² (0.04% of total European land). Within this estimate, the by far largest contribution comes from hydrogen, for which I use an upper-bound estimation (stored uncompressed in overground tanks). If hydrogen is stored land efficiently in underground caverns, flexibility needs of all three technologies can rise by orders of magnitude without making a significant contribution to total land requirements of fully renewable electricity systems.

4. Discussion

I show that there is a trade-off between land requirements and total system cost of fully renewable electricity in the future in Europe, but that reducing land requirements by changing supply-side technologies does not necessarily lead to substantial cost penalties. The expected land requirements of a system design with minimal expected cost is 97,000 km² (2% total European land). Such a low-cost system is supplied, apart from hydroelectricity and bioenergy from residuals, only from onshore wind farms and utility-scale photovoltaics (PV). Its expected land requirements can be reduced by replacing onshore wind with offshore wind, utility-scale PV, or rooftop PV. Offshore wind is the most cost-effective option of these three possibilities. It decreases cost-minimal land requirements by 50% for an expected cost penalty of only 5%. Utility-scale and rooftop PV lead to the same effect for cost penalties of 9% and 17%. All three technologies can reduce land requirements more than 50% for higher cost penalties by replacing larger amounts of onshore wind capacity.

Because future cost and land requirements of wind and solar power are not known with certainty, total system cost and total land requirements of renewable electricity supply in Europe is uncertain as well. Ensuring land requirements lower than 1% of total European land (50% of the cost-minimal case) can thus require cost penalties as large as 40%. Despite these uncertainties, three main findings are robust: First, onshore wind always requires the most amount of land and thus a switch to any other technology to reduce land is a no-regret option. Second, offshore wind is always the most cost-effective option, followed by utility-scale PV, and rooftop PV. Third, reducing land requirements of fully renewable electricity in Europe does likely not come with high cost: cost penalties of 20% or less are most likely for a system with low land requirements of 1% of European land through sufficient shares of offshore wind, utility-scale PV, or rooftop PV.

Considering all uncertainty and all possible system designs, land requirements of solar and wind power are in the range of 0–3% of total European land. Significant contributions to the land needs of electricity supply can be expected from the transmission grid and hydroelectricity (<0.3% and <1%, see Methods). The total land requirements of fully renewable electricity are thus likely within the range of 1.3–4.3%.

4.1 Comparison to previous studies

Comparing my results to findings from previous studies shows that there is some uncertainty about the potential of rooftop PV to reduce land requirements, as well as some uncertainty about the land requirements and therefore cost-effectiveness of utility-scale PV. There are no findings, however, that question the potential or the cost-effectiveness of offshore wind.

First, rooftop PV generates up to 1,800 TWh/yr in this study. Other estimations for the potential of photovoltaics on roofs and facades are lower: 680 TWh/yr for only rooftop PV [54] and 1200–2100 TWh for rooftop PV and PV on façades [2]. Should the lower estimations be correct, very high shares of rooftop PV as considered in this study may not be possible. In that case, rooftop PV could reduce smaller amounts of land requirements only. High shares of rooftop PV are in any case less attractive due to high cost and high balancing requirements, as I show in my results.

Second, up to 1,800 TWh/yr are generated by utility-scale PV when its capacity share is the highest. While this does not exceed potential estimations in the literature [2,28], there are conflicting estimations about how much land utility-scale requires (see also Methods). While this study uses measurements from ref. [46], ref. [55] states that areas larger than the fenced areas of PV farms must be considered, leading to capacity shares two times smaller than in this study. Whether such areas should be included is questioned [7]. Using their capacity shares would reduce the cost-effectiveness of utility-scale PV.

Third, with potential estimations as large as 40,000 TWh/yr (~10 times current electricity demand) [56] or even 50,000 TWh/yr [57], the potential of offshore wind to reduce land requirements is not questioned by previous findings in the literature.

Lastly, one study finds total land requirements of a system based on PV which are more than ten times larger than in this study: 8% of total European land [28]. The large deviation can be explained mainly by two differences: First, by the above-mentioned lower capacity densities given in ref. [55]. Second, by their finding that large overcapacities are necessary to handle renewable fluctuations: in the most extreme case of Finland, this leads to 7 times the required capacity. In my study, fluctuations are handled by continental balancing through the transmission grid and by flexible generation from bioenergy. As a result, I find only three countries require overcapacities, and overcapacities never exceed 1.15 times the required capacity. The handling of renewable fluctuations explains the largest part of the different findings of the two studies. This shows the importance of an analysis on the system level, including not only supply but also balancing.

4.2 Limitations and outlook

The high-level perspective on land requirements in this study allows to understand the full spatial extent of electricity supply infrastructure on European land and its trade-off with system cost. Land requirements of the different technologies, however, are not always directly comparable. For example, while solar photovoltaics does not allow for any other land use–at least not as long as agrophotovoltaics is unavailable at large scale [58]–the vast spacing between wind turbines does allow for agriculture. Thus, the two technologies compete differently with other uses of land. In addition, building onshore wind farms on land that is already developed decreases the impact of the newly constructed infrastructure [59]. Offshore wind of course requires no land but competes with other uses of offshore areas and can have visual impacts if the turbines are close to the shore. Because I analyse total land requirements in this study, I cannot account for these qualitative differences. However, I mitigate this limitation by making options to reduce land requirements technology-specific.

Further, not only total land requirements are important, but also the exact location and technical parameters: wind turbines impact landscapes stronger when they are larger and when they are placed on exposed locations like hilltops. Analysing the visual impact of renewable electricity on landscapes has not been done on this level of detail so far. In this study, I use land requirements of solar and wind installation as a proxy for visual impacts.

By assessing the trade-off between total system cost and land requirements from a continental perspective and for fully renewable electricity supply, I omit any form of transitional or distributional aspects in this study. While this allows me to show options Europe has for its future supply, further research has to be done on the transition towards these options and on the question of how cost and land requirements are geographically and societally distributed. In particular, this includes the distributional imbalance that reducing land requirements has local impacts on land, but continental impacts on system cost. Current electricity markets cannot resolve this imbalance.

4.3 Conclusion

My findings show that supply technology choice is an effective way to reduce land requirements of fully renewable electricity systems in Europe. Systems with vastly different land requirements can be designed, and their cost must not vary much as long as land requirements are reduced cost-effectively. Instead of relying strongly on onshore wind, which is likely the cost-minimal solution, electricity can be generated offshore at large scale and be transported to demand centres using a sufficient transmission grid. The expansion of both, onshore wind farms and transmission grid can be limited by alternatively generating solar electricity locally. Such a solar-centred electricity supply is enabled by flexible generation from bioenergy to cope with seasonal fluctuations. These findings increase the solution space for a European energy transition and allow to integrate more diverse stakeholder positions than is possible with cost-minimised electricity system designs.

Supporting information

S1 File

(PDF)

Click here for additional data file.^{(1.1MB, pdf)}

S1 Data

(NC)

Click here for additional data file.^{(7.6MB, nc)}

S1 Code

(ZIP)

Click here for additional data file.^{(274.9KB, zip)}

Acknowledgments

I would like to thank Johan Lilliestam and Stefan Pfenninger for valuable comments on earlier drafts of this article.

Data Availability

The applied model of the electricity system and all result data are available from Zenodo doi:10.5281/zenodo.3707812.

Funding Statement

The work of T.T. was supported by a European Research Council grant (TRIPOD, grant agreement number 715132). http://erc.europa.eu The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Jacobson MZ, Delucchi MA, Bauer ZAF, Goodman SC, Chapman WE, Cameron MA, et al. 100% Clean and Renewable Wind, Water, and Sunlight All-Sector Energy Roadmaps for 139 Countries of the World. Joule. 2017;1: 108–121. 10.1016/j.joule.2017.07.005 [DOI] [Google Scholar]
2.Ruiz P, Nijs W, Tarvydas D, Sgobbi A, Zucker A, Pilli R, et al. ENSPRESO—an open, EU-28 wide, transparent and coherent database of wind, solar and biomass energy potentials. Energy Strategy Reviews. 2019;26: 100379 10.1016/j.esr.2019.100379 [DOI] [Google Scholar]
3.Tröndle T, Pfenninger S, Lilliestam J. Home-made or imported: On the possibility for renewable electricity autarky on all scales in Europe. Energy Strategy Reviews. 2019;26 10.1016/j.esr.2019.100388 [DOI] [Google Scholar]
4.European Commission. COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE EUROPEAN COUNCIL, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE, THE COMMITTEE OF THE REGIONS AND THE EUROPEAN INVESTMENT BANK A Clean Planet for all A European strategic long-term vision for a prosperous, modern, competitive and climate neutral economy. Brussels: European Commission; 2018. Nov. Available: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52018DC0773 [Google Scholar]
5.MacKay DJC. Sustainable Energy without the hot air. UIT; 2008. [Google Scholar]
6.McDonald RI, Fargione J, Kiesecker J, Miller WM, Powell J. Energy Sprawl or Energy Efficiency: Climate Policy Impacts on Natural Habitat for the United States of America. Añel JA, editor. PLoS ONE. 2009;4: e6802 10.1371/journal.pone.0006802 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Smil V. Power Density: A Key to Understanding Energy Sources and Uses. Cambridge, Massachusetts: The MIT Press; 2015. [Google Scholar]
8.van Zalk J, Behrens P. The spatial extent of renewable and non-renewable power generation: A review and meta-analysis of power densities and their application in the U.S. Energy Policy. 2018;123: 83–91. 10.1016/j.enpol.2018.08.023 [DOI] [Google Scholar]
9.Brinkley C, Leach A. Energy next door: A meta-analysis of energy infrastructure impact on housing value. Energy Research & Social Science. 2019;50: 51–65. 10.1016/j.erss.2018.11.014 [DOI] [Google Scholar]
10.Molnarova K, Sklenicka P, Stiborek J, Svobodova K, Salek M, Brabec E. Visual preferences for wind turbines: Location, numbers and respondent characteristics. Applied Energy. 2012;92: 269–278. 10.1016/j.apenergy.2011.11.001 [DOI] [Google Scholar]
11.Sánchez-Pantoja N, Vidal R, Pastor MC. Aesthetic impact of solar energy systems. Renewable and Sustainable Energy Reviews. 2018;98: 227–238. 10.1016/j.rser.2018.09.021 [DOI] [Google Scholar]
12.Devine-Wright P, Howes Y. Disruption to place attachment and the protection of restorative environments: A wind energy case study. Journal of Environmental Psychology. 2010;30: 271–280. 10.1016/j.jenvp.2010.01.008 [DOI] [Google Scholar]
13.de Vries S, de Groot M, Boers J. Eyesores in sight: Quantifying the impact of man-made elements on the scenic beauty of Dutch landscapes. Landscape and Urban Planning. 2012;105: 118–127. 10.1016/j.landurbplan.2011.12.005 [DOI] [Google Scholar]
14.Knebel A, Hartman J, Kajimura R. Opinions on renewables—a look at polls in industrialised countries. Berlin: Renewable Energies Agency; 2016. Mar. Available: https://www.unendlich-viel-energie.de/media/file/427.AEE_RK29_%20Internationale_Akzeptanzumfragen_EN.pdf [Google Scholar]
15.Setton D. Social Sustainability Barometer for the German Energiewende: 2018 Edition Core statements and summary of the key findings. Potsdam: Institute for Advanced Sustainability Studies; 2019. Jul. 10.2312/iass.2019.014 [DOI] [Google Scholar]
16.Stokes LC. Electoral Backlash against Climate Policy: A Natural Experiment on Retrospective Voting and Local Resistance to Public Policy. American Journal of Political Science. 2016;60: 958–974. 10.1111/ajps.12220 [DOI] [Google Scholar]
17.Späth L. Large-scale photovoltaics? Yes please, but not like this! Insights on different perspectives underlying the trade-off between land use and renewable electricity development. Energy Policy. 2018;122: 429–437. 10.1016/j.enpol.2018.07.029 [DOI] [Google Scholar]
18.Palmer-Wilson K, Donald J, Robertson B, Lyseng B, Keller V, Fowler M, et al. Impact of land requirements on electricity system decarbonisation pathways. Energy Policy. 2019;129: 193–205. 10.1016/j.enpol.2019.01.071 [DOI] [Google Scholar]
19.Schlachtberger DP, Brown T, Schramm S, Greiner M. The benefits of cooperation in a highly renewable European electricity network. Energy. 2017;134: 469–481. 10.1016/j.energy.2017.06.004 [DOI] [Google Scholar]
20.Neumann F, Brown T. The Near-Optimal Feasible Space of a Renewable Power System Model. 2019. [cited 28 Oct 2019]. Available: http://arxiv.org/abs/1910.01891 [Google Scholar]
21.Tröndle T, Lilliestam J, Marelli S, Pfenninger S. Trade-offs between geographic scale, cost, and infrastructure requirements for fully renewable electricity in Europe. Joule. 2020. doi: Forthcoming [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Bussar C, Moos M, Alvarez R, Wolf P, Thien T, Chen H, et al. Optimal Allocation and Capacity of Energy Storage Systems in a Future European Power System with 100% Renewable Energy Generation. Energy Procedia. 2014;46: 40–47. 10.1016/j.egypro.2014.01.156 [DOI] [Google Scholar]
23.Connolly D, Lund H, Mathiesen BV. Smart Energy Europe: The technical and economic impact of one potential 100% renewable energy scenario for the European Union. Renewable and Sustainable Energy Reviews. 2016;60: 1634–1653. 10.1016/j.rser.2016.02.025 [DOI] [Google Scholar]
24.Pleßmann G, Blechinger P. How to meet EU GHG emission reduction targets? A model based decarbonization pathway for Europe’s electricity supply system until 2050. Energy Strategy Reviews. 2017;15: 19–32. 10.1016/j.esr.2016.11.003 [DOI] [Google Scholar]
25.Brown TW, Bischof-Niemz T, Blok K, Breyer C, Lund H, Mathiesen BV. Response to ‘Burden of proof: A comprehensive review of the feasibility of 100% renewable-electricity systems’. Renewable and Sustainable Energy Reviews. 2018;92: 834–847. 10.1016/j.rser.2018.04.113 [DOI] [Google Scholar]
26.Child M, Kemfert C, Bogdanov D, Breyer C. Flexible electricity generation, grid exchange and storage for the transition to a 100% renewable energy system in Europe. Renewable Energy. 2019;139: 80–101. 10.1016/j.renene.2019.02.077 [DOI] [Google Scholar]
27.Fthenakis V, Kim HC. Land use and electricity generation: A life-cycle analysis. Renewable and Sustainable Energy Reviews. 2009;13: 1465–1474. 10.1016/j.rser.2008.09.017 [DOI] [Google Scholar]
28.Capellán-Pérez I, de Castro C, Arto I. Assessing vulnerabilities and limits in the transition to renewable energies: Land requirements under 100% solar energy scenarios. Renewable and Sustainable Energy Reviews. 2017;77: 760–782. 10.1016/j.rser.2017.03.137 [DOI] [Google Scholar]
29.Tröndle T. Code related to "Tröndle (2020)–Supply-side options to reduce land requirements of fully renewable electricity in Europe". Potsdam: Institute for Advanced Sustainability Studies; 2020. 10.5281/zenodo.3956531 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Koster J, Rahmann S. Snakemake–a scalable bioinformatics workflow engine. Bioinformatics. 2012;28: 2520–2522. 10.1093/bioinformatics/bts480 [DOI] [PubMed] [Google Scholar]
31.Tröndle T. euro-calliope. Potsdam: Institute for Advanced Sustainability Studies; 2020. 10.5281/zenodo.3949794 [DOI] [Google Scholar]
32.Pfenninger S, Pickering B. Calliope: A multi-scale energy systems modelling framework. The Journal of Open Source Software. 2018. [cited 3 Jun 2019]. 10.21105/joss.00825 [DOI] [Google Scholar]
33.Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P, Kobayashi S, et al. The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society. 2011;137: 553–597. 10.1002/qj.828 [DOI] [Google Scholar]
34.Pfenninger S, Staffell I. Long-term patterns of European PV output using 30 years of validated hourly reanalysis and satellite data. Energy. 2016;114: 1251–1265. 10.1016/j.energy.2016.08.060 [DOI] [Google Scholar]
35.Staffell I, Pfenninger S. Using bias-corrected reanalysis to simulate current and future wind power output. Energy. 2016;114: 1224–1239. 10.1016/j.energy.2016.08.068 [DOI] [Google Scholar]
36.Ruiz Castello P, Sgobbi A, Nijs W, Thiel C, Dalla Longa F, Kober T, et al. The JRC-EU-TIMES model. Bioenergy potentials for EU and neighbouring countries. 2015. Dec. 10.2790/39014 [DOI] [Google Scholar]
37.Steffen B. Estimating the Cost of Capital for Renewable Energy Projects. Rochester, NY: Social Science Research Network; 2019. Apr. Report No.: ID 3373905. Available: https://papers.ssrn.com/abstract=3373905 [Google Scholar]
38.JRC. ETRI 2014 –Energy Technology Reference Indicator projections for 2010–2050. JRC; 2014. Available: https://setis.ec.europa.eu/related-jrc-activities/jrc-setis-reports/etri-2014 [Google Scholar]
39.Schmidt O, Melchior S, Hawkes A, Staffell I. Projecting the Future Levelized Cost of Electricity Storage Technologies. Joule. 2019;0. 10.1016/j.joule.2018.12.008 [DOI] [Google Scholar]
40.Wirth H. Recent Facts about Photovoltaics in Germany. Freiburg im Breisgau: Fraunhofer ISE; 2020. January p. 98 Available: https://www.ise.fraunhofer.de/en/publications/studies/recent-facts-about-pv-in-germany.html [Google Scholar]
41.Turney D, Fthenakis V. Environmental impacts from the installation and operation of large-scale solar power plants. Renewable and Sustainable Energy Reviews. 2011;15: 3261–3270. 10.1016/j.rser.2011.04.023 [DOI] [Google Scholar]
42.Wirth H. Aktuelle Fakten zur Photovoltaik in Deutschland. Fraunhofer: ISE; 2017. Aug. Available: https://www.ise.fraunhofer.de/de/daten-zu-erneuerbaren-energien.html [Google Scholar]
43.McKenna R, Hollnaicher S, Ostman v. d. Leye P, Fichtner W. Cost-potentials for large onshore wind turbines in Europe. Energy. 2015;83: 217–229. 10.1016/j.energy.2015.02.016 [DOI] [Google Scholar]
44.Miller LM, Keith DW. Corrigendum: Observation-based solar and wind power capacity factors and power densities (2018 Environ. Res. Lett. 13 104008). Environ Res Lett. 2019. 10.1088/1748-9326/aaf9cf [DOI] [Google Scholar]
45.Nitsch F, Turkovska O, Schmidt J. Observation-based estimates of land availability for wind power: A case study for Czechia. Energ Sustain Soc. 2019;9: 45 10.1186/s13705-019-0234-z [DOI] [PMC free article] [PubMed] [Google Scholar]
46.ZSW, Bosch & Partner. Vorbereitung und Begleitung bei der Erstellung eines Erfahrungsberichts gemäß § 97 Erneuerbare-Energien-Gesetz–Teilvorhaben II c: Solare Strahlungsenergie. German Federal Ministry for Economic Affairs and Energy; 2019. March p. 169 Available: https://www.erneuerbare-energien.de/EE/Redaktion/DE/Downloads/bmwi_de/zsv-boschundpartner-vorbereitung-begleitung-eeg.pdf?__blob=publicationFile&v=7 [Google Scholar]
47.Halleraker J, van de Bund W, Bussettini M, Gosling R, Döbbelt-Grüne S, Hensman J, et al. Working Group ECOSTAT report on Common understanding of using mitigation measures for reaching Good Ecological Potential for heavily modified water bodies. Ispra: JRC; 2016. p. 104 10.2760/649695 [DOI] [Google Scholar]
48.Aggreko Y. Cube—Energy storage–Product Guide. 2020. [cited 29 Feb 2020]. Available: https://www.aggreko.com/-/media/Aggreko/Files/PDF/Energy-Storage/YCubeProductGuideFinal.pdf [Google Scholar]
49.Wiegmans B. Gridkit: Gridkit 1.0 ‘For Scientists’. Zenodo; 2016. 10.5281/ZENODO.47263 [DOI] [Google Scholar]
50.Hörsch J, Hofmann F, Schlachtberger D, Brown T. PyPSA-Eur: An open optimisation model of the European transmission system. Energy Strategy Reviews. 2018;22: 207–215. 10.1016/j.esr.2018.08.012 [DOI] [Google Scholar]
51.Stevens L, Anderson B, Cowan C, Colton K, Johnson D. The footprint of energy: Land use of U.S. Electricity production. Strata; 2017. Jun. [Google Scholar]
52.Hörsch J, Brown T. The role of spatial scale in joint optimisations of generation and transmission for European highly renewable scenarios. 2017 14th International Conference on the European Energy Market (EEM). Dresden, Germany: IEEE; 2017. pp. 1–7. 10.1109/EEM.2017.7982024 [DOI]
53.Herman J, Usher W. SALib: An open-source Python library for Sensitivity Analysis. JOSS. 2017;2: 97 10.21105/joss.00097 [DOI] [Google Scholar]
54.Bódis K, Kougias I, Jäger-Waldau A, Taylor N, Szabó S. A high-resolution geospatial assessment of the rooftop solar photovoltaic potential in the European Union. Renewable and Sustainable Energy Reviews. 2019;114: 109309 10.1016/j.rser.2019.109309 [DOI] [Google Scholar]
55.de Castro C, Mediavilla M, Miguel LJ, Frechoso F. Global solar electric potential: A review of their technical and sustainable limits. Renewable and Sustainable Energy Reviews. 2013;28: 824–835. 10.1016/j.rser.2013.08.040 [DOI] [Google Scholar]
56.Caglayan DG, Ryberg DS, Heinrichs H, Linßen J, Stolten D, Robinius M. The techno-economic potential of offshore wind energy with optimized future turbine designs in Europe. Applied Energy. 2019;255: 113794 10.1016/j.apenergy.2019.113794 [DOI] [Google Scholar]
57.IEA. Offshore Wind Outlook 2019: World Energy Outlook Special Report. Paris; 2019 Nov p. 98. Available: https://www.iea.org/reports/offshore-wind-outlook-2019/geospatial-analysis
58.Weselek A, Ehmann A, Zikeli S, Lewandowski I, Schindele S, Högy P. Agrophotovoltaic systems: Applications, challenges, and opportunities. A review. Agron Sustain Dev. 2019;39: 35 10.1007/s13593-019-0581-3 [DOI] [Google Scholar]
59.Diffendorfer JE, Dorning MA, Keen JR, Kramer LA, Taylor RV. Geographic context affects the landscape change and fragmentation caused by wind energy facilities. PeerJ. 2019;7: e7129 10.7717/peerj.7129 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0236958.r001

Decision Letter 0

Baogui Xin

29 Apr 2020

PONE-D-20-07241

Supply-side options to reduce land requirements of fully renewable electricity in Europe

PLOS ONE

Dear Mr. Tröndle,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We recommend that it should be revised taking into account the changes requested by the reviewers. Since the requested changes includes Major Revision, the revised manuscript will undergo the next round of review by the same reviewers.

We would appreciate receiving your revised manuscript by Jun 13 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Baogui Xin, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This is an interesting paper covering and important topic on how multicountry/regional energy strategies impact land cover change and costs of electricity. I think the overall methodology is sound, even if done at fairly low spatial resolution. There is a time and place for these sorts of broad, generalized models, and I think the intersection of land cover change and energy production is one of them. We need this work done in the US! Thus, I’m supportive of the work being published.

My main concern with the existing manuscript boils down to parameter inputs for the model. Regarding costs, I’m not entirely sure what these costs represent and how they relate/translate to the price of electricity for consumers, which seems to be what really matters. This needs to be described in more detail instead of just citing the sources of the data. Also, the discussion section needs to tie the models results back to the real world better. I’ve seen numerous reports on the large price tag in Germany for their renewables (Energiewende). It would be nice if you could bring the model results back to what is happening in Europe/Germany with prices. You concluded prices won’t go up much, yet Germany seems to suggest your results are wrong. This needs apparent contradiction needs to be added to the discussion to make your simulations more relevant.

Regarding energy capacity. I really think you should run your simulations with and without the spacing between turbines. Energy capacity of wind is quite high when you remove spacing between turbines and simply use estimates of the actual land disturbed by the facility. As I mention in my detailed comments below, the land between turbines is used and the reality is that the existing landscape drives turbine placement, more than turbines drive land use. The point is that turbines are often placed on the landscape in areas where the existing land use can continue..so it is not wasted space. I think your results will be fundamentally different if you do this. Such a large (expected) difference in outcomes means your model is likely VERY sensitive to changes in energy capacity. The values I pulled from the literature for wind energy without the spacing, are quite high…much higher then solar.

Detailed comments.

Line 17. “that must not be large”. Replace with “with cost penalities between 5 and 10% depending on the scenario.” or just “small cost penalties”.

Line 36. Devaluate to devalue.

38. does not to do not.

61. replace must not with do not.

79. Delete ‘To be able “. Start with To identify…

168. Rooftop PV. Is this residential? If so, how is the cost calculated? I don’t know how it works in Europe, but in the US, individual households finance their rooftop systems. This can be done in many ways…from leasing the panels, the buying the fully, to taking out a loan to cover their costs. What does the 880 represent in this case? For some households, the PV reduces electricity costs and, after the loan is paid off, represents a form of tax-free income. Does the 880 represent costs to the government if they were going to provide funding for residential PV? I have further comments below (rel. to line 359( about costs. You need to add some detail here about what exactly these costs represent and why they a reasonable input for your model. For example, it seems like cost to consumers is the most ideal cost to include in your model, but perhaps this is impossible to get.

182. delete ‘in the following’

183. I think it’s good to only consider uncertainty on wind and solar while holding the other technologies constant. However, it needs to be explained better. I think you structure the model so that cost does not affect supply shares, which is fine given the goal of understanding how different energy strategies change both land cover and costs. It seems to me the main reason for doing this is to deal with the uncertainty you have in future costs of the technology. The uniform distribution is good.

192-199. The only problem with ref. 8 for an estimate of energy density is that many, if not all of the studies regarding onshore wind didn’t actual measure the area used. Instead, they estimated it. I only know of a few studies used aerial photography to directly digitize then estimate the area transformed by onshore wind. I don’t think these corrected for capacity factor. They may be worth comparing to ref. 8. Diffendorfer has a table of estimates in (ha/MW) similar to ref. 8. For example, Diffendorfer estimated an average of 0.93 ha/MW, which (if I did my conversions correctly) is 1.075 MW/ha or 107.5 W/m^2…much higher than 8.82 used in this paper. These values represent just the disturbance caused by the wind farm, not the entire project area.

Diffendorfer, J.E., and R.W. Compton. “Land Cover and Topography Affect the Land Transformation Caused by Wind Facilities.” PLoS ONE 9, no. 2 (2014): e88914.

Jones, N.F., and L. Pejchar. “Comparing the Ecological Impacts of Wind and Oil & Gas Development: A Landscape Scale Assessment.” PLoS ONE 8, no. 11 (2013): e81391

Jones and Pejchar estimated 247m2/317 MMBTu.

I’m not sure I would use “the wind turbines together with the technically necessary spacing between turbines”. The area between turbines is not useless. It can be agriculture, pasture, and habitat for many species. . You are correct that some land uses are not compatible with wind energy..homes between turbines would be very dangerous. But, at least in the US, turbines are only placed on areas where land use is compatabile…in farmlands, or pastures, or natural areas. The wind energy is not excluding the use of land…it’s exactly the opposite. Certain uses of the land excludes where wind energy goes. You could limit the potential area for wind energy in the EU by masking out urban areas, areas of building density above X/ha, certain distances to roads, areas near airports or weather radar (turbines affect radar). There are a number of GIS based wind energy potential maps out there.

Furthermore, your point about visual impacts suggests you may want to use an area even larger than the spacing between turbines because the visual impacts of wind facilities go out quite far. This would would suggest even higher land requirements.

204-206. Are these #’s truly theoretical maximums? If so, then the papers I cited above, which directly measured land use from wind facilities, seems to contradict them.

228. Given how you measure capacity density for solar, it seems like estimate for wind that only include the real land disturbance might be more comparable. In most utlility scale PV systems I know of, areas between panels are cleared and not useable..similar to the areas around the base of wind turbines.

Ultimately, you might want to consider using both the ‘footprint’ only calculation for wind as well as a large-scale estimate (as you’ve done). Other’s have done this (Denholm) and it gives a more clear picture of wind energy’s spatial impacts.

272. “Lastly, the transmission grid already has significant land requirements….”

281. Change to low spatial resolution. Higher spatial resolution = smaller units of area per pixel. 1m^2/pixel is higher resolution than entire countries. Your study is very low spatial resolution.

289. “by applying” do you mean you multiplied total energy * 1/capacity density? Please use equations when necessary to better explain your methods. I realize it was an MC approach, but the underlying process that happens at each iteration can be described. Also, how did you deal with the outputs. Did you simply develop summary statistics (means and sd) for each scenario?

299. Figure 2 is a nice summary. What is of most interest to me are those scenarios that generate both a low cost and a low land use. Thus a scatterplot of land use vs cost would be a nice panel to add. You could then describe the %’s of the energy mixes for either extreme point…lowest cost and lowest land use vs highest cost and highest land use. Just seeing the shape of joint distribution would be helpful to me as it might show trade-offs or system bounds…for example there are no cases of low land use and low cost (a pet hypothesis of mine…particularly if you want to place wind in areas with minimized environmental impacts…it will cost more).

299. I’d make the ranges be 0% and ~3% and ~0.06 and ~0.10 as the graphs show higher and lower values.

299. You don’t need the sentence “these ranges include…” nore the next sentence. You state the figure represents all 29 million cases.

305. I don’t disagree with your statement here but think some readers my wonder why 0-3% is a ‘vast’ difference in land use and others will think that a doubling of their electric bill is a HUGE increase! In my world, an energy company has a very large fight if they try to raise electricity rates by just 3%...this really hurts low income households. So, it would be good to try to put these modelled changes in context..both social (for cost) and perhaps environmental? (for land use).

311. Figure 3 would better match the text if the cost plots were in Eur per KWH and the land use plots in % of Europe’s land total. Right now the scales don’t match what’s described in the text. I can see the logic for the current ‘difference from minimum’ scale.Perhaps % change from minimum would be better here since you describe % changes in the text quite often.

Your results will change considerably if you used the estimate of energy capacity for wind energy that considers just the land used, not the space between turbines.

340. You are correct, but somewhere you should acknowledge that offshore wind also has impacts…these are not on land obviously, but there is a growing literature on impacts from offshore wind on marine ecosystems and birds. I’m not sure but offshore sites likely restrict industrial fishing? There is also visual/social and cultural issues if the turbines can be seen from shore.

359. Cost becomes important here. See my comments above about cost of PV. Are your costs estimates similar to the levelized cost of electricity (LCOE) or do they vary by who is paying for the energy? Residential PV is paid for (typically) by the homeowner, while cost of industrial wind energy could be calculated as the cost to consumers, the price it costs the wind energy company to generate X capacity, etc. The broader point being that residential PV may not have the high cost when viewed from the households perspective over the life of the panels.

529. While I’m not entirely sure what you mean by the impacts on landscapes, the following papers might be relevant:

“Geographic Context Affects the Landscape Change and Fragmentation Caused by Wind Energy Facilities [PeerJ].” Accessed April 2, 2020. https://peerj.com/articles/7129/.

Jones, N.F., L. Pejchar, and J.M. Kiesecker. “The Energy Footprint: How Oil, Natural Gas, and Wind Energy Affect Land for Biodiversity and the Flow of Ecosystem Services.” BioScience, 2015. http://bioscience.oxfordjournals.org/content/early/2015/01/22/biosci.biu224.abstract.

“Monitoring Wind Farms Occupying Grasslands Based on Remote-Sensing Data from China’s GF-2 HD Satellite—A Case Study of Jiuquan City, Gansu Province, China - ScienceDirect.” Accessed April 2, 2020. https://www.sciencedirect.com/science/article/abs/pii/S092134491630163X?via%3Dihub.

“Energy Development in Colorado’s Pawnee National Grasslands: Mapping and Measuring the Disturbance Footprint of Renewables and Non-Renewables | SpringerLink.” Accessed April 2, 2020. https://link.springer.com/article/10.1007%2Fs00267-017-0846-z.

532. How much would your results change if you could model transmission lines and how much that contributes to overall land requirements for a given energy scenario? I’m not sure if more offshore would mean more transmission, but it could.

Reviewer #2: This paper addresses a large concern with renewable energy development, land use requirements.

Since I do not have the expertise to address the economic sections of this manuscript, I will focus on the land use requirements. Therefore I will provide more high-level comments than detailed.

1) I would have like to see more about the context of the foot print and cost in European countries (e.g., impact to natural lands)

2) There has been a lot of work done on this matter in USA. Please explain why the "Energy Sprawl" work was not cited in this paper?

3) Would it have been possible to give summary regarding different countries? It seems like Europe was considered one large mass. How do these results like up with European renewable energy policies to meet future energy demands?

4) On the above note, the author assumes no new hydropower dam development. Is this realistic?

5) It seems the author included roads and transmission lines into the land use estimate for solar and wind but not hydroelectic. This does not seem valid.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Aug 6;15(8):e0236958. doi: 10.1371/journal.pone.0236958.r002

Author response to Decision Letter 0

4 Jul 2020

# Author’s response to the review on the manuscript PONE-D-20-07241

First submitted to

*PLOS ONE*

on 12 March 2020

Re-submitted in revised form after major revisions on 04 July 2020.

I would like to thank the editors for the consideration of my manuscript and the reviewers for their review and the valuable comments. I updated the manuscript based on the comments of the reviewers and hope that I addressed all issues to the satisfaction of the reviewers and editors.

In addition to the comments from the reviewers I did the following:

* Applied minor fixes to the model of the underlying electricity system.

* Increased the temporal resolution of the model from 4h to 3h.

Both improvements increase the fidelity of the results but had only minor impacts not the results.

In the following, I address all comments by the reviewers one by one.

---

I would like to thank the reviewers for taking the time to review my manuscript and for providing useful comments that helped to improve it. Below, I copy all comments of the reviewers in bold and add my responses underneath.

## Reviewer #1:

> Main comment 1: This is an interesting paper covering and important topic on how multicountry/regional energy strategies impact land cover change and costs of electricity. I think the overall methodology is sound, even if done at fairly low spatial resolution. There is a time and place for these sorts of broad, generalized models, and I think the intersection of land cover change and energy production is one of them. We need this work done in the US! Thus, I’m supportive of the work being published.

Thank you.

> Main comment 2: My main concern with the existing manuscript boils down to parameter inputs for the model. Regarding costs, I’m not entirely sure what these costs represent and how they relate/translate to the price of electricity for consumers, which seems to be what really matters. This needs to be described in more detail instead of just citing the sources of the data. Also, the discussion section needs to tie the models results back to the real world better. I’ve seen numerous reports on the large price tag in Germany for their renewables (Energiewende). It would be nice if you could bring the model results back to what is happening in Europe/Germany with prices. You concluded prices won’t go up much, yet Germany seems to suggest your results are wrong. This needs apparent contradiction needs to be added to the discussion to make your simulations more relevant.

The aspects that you are mentioning here are important to the energy transition in Europe and elsewhere. I agree that transitional and distributional (be it geographically or between different types of investors or users) effects would be very relevant and should be further researched. However, I decided to not include any form of transitional or distributional effects in this study as I think it widens its scope too much.

The price developments in Germany that you describe contain both transitional and distributional aspects. They contain transitional aspects because Germany has been investing into solar power at a point in time when solar power’s cost was still very high, knowing that cost will decrease with deployment. Prices in Germany also contain at least two forms of distributional aspects. First, had Germany not invested into solar power but rather waited for another country to do so, electricity prices in Germany would be lower today. Parts of the cost of the energy transition would have been redistributed to another country. Second, the renewable electricity levy included in electricity prices in Germany strongly depends on who is paying for them. Households do pay, some businesses do, larger businesses do not. This distributional aspect of electricity prices is driven by political decisions, not by technical or economic aspects of renewable electricity.

In this study, I am not considering any of these transitional or distributional aspects of electricity prices. Doing so would require a significant shift of the focus of this study and a different research design. Instead of prices, I am assessing total system cost of electricity (by that avoiding distributional aspects) of a fully renewable electricity system (by that avoiding transitional aspects and with it, distributional aspects between countries). This lets me understand the trade-off between total system cost of fully renewable electricity and their land requirements on a long term. The study answers the question of cost of _not_ using land for electricity generation in future electricity supply. It does not answer questions of how to reach these kind of systems (transitional aspects) or who would profit (distributional aspects). I am looking forward to reading about these aspects in further studies.

I fully agree with you that the manuscript must be transparent about these choices, and I therefore applied the following changes:

1. In several places in the main text and the abstract I made clearer that the study is about future electricity supply and therefore does not reflect current cost.

2. In several places in the main text and the abstract I qualify the vague term “cost” with “total system (cost)” in order to be more precise.

3. In the “System design model” section I added the following sentence:

“Their values are future projections and assume all technologies are deployed at large scale. **In particular, this means that I ignore any forms of transitional effects stemming from technological or financial learning.**”

4. I made the caption of Table 1 more precise:

“Assumptions on installation and maintenance cost of electricity infrastructure.”

6. I added a paragraph to the “Limitations and outlook” section:

“By assessing the trade-off between total system cost and land requirements from a continental perspective and for fully renewable electricity supply, I omit any form of transitional or distributional aspects in this study. While this allows me to show options Europe has for its future supply, further research has to be done on the transition towards these options and on the question of how cost and land requirements are geographically and societally distributed. In particular, this includes the distributional imbalance that reducing land requirements has local impacts on land, but continental impacts on system cost. Current electricity markets can not resolve this imbalance.”

> Main comment 3: Regarding energy capacity. I really think you should run your simulations with and without the spacing between turbines. Energy capacity of wind is quite high when you remove spacing between turbines and simply use estimates of the actual land disturbed by the facility. As I mention in my detailed comments below, the land between turbines is used and the reality is that the existing landscape drives turbine placement, more than turbines drive land use. The point is that turbines are often placed on the landscape in areas where the existing land use can continue..so it is not wasted space. I think your results will be fundamentally different if you do this. Such a large (expected) difference in outcomes means your model is likely VERY sensitive to changes in energy capacity. The values I pulled from the literature for wind energy without the spacing, are quite high…much higher then solar.

You are right in assuming that the study results are sensitive to the decision whether to include spacing of wind turbines into the analysis. Its biggest impact is on the usefulness of utility-scale PV: without spacing, land requirements of onshore wind are smaller than the ones of utility-scale PV and thus replacing onshore wind capacities with utility-scale PV capacities increases not decreases total land requirements.

Nevertheless, I think that including spacing into the analysis and the conclusions of this analysis is the right choice. While I agree with you that wind development is driven by landscapes and existing land uses and not the other way round, the spacing of wind power does in any case exclude other land uses, has the potential to devalue land, and is visually impacting landscapes. As these are the main impacts that I am considering in this analysis, I decided to focus on land requirements including spacing.

However, I now also show results from an analysis that does _not_ include spacing. I show these results in Section 3 of Supporting Information S1 together with a short discussion of the implications of the choice of including or excluding spacing.

> Detailed comments.

> Line 17: “that must not be large”. Replace with “with cost penalities between 5 and 10% depending on the scenario.” or just “small cost penalties”.

I changed this to “with only small cost penalties”.

> Line 36: Devaluate to devalue.

Corrected accordingly

> Line 38: does not to do not.

Corrected accordingly.

> Line 61: replace must not with do not.

Thank you for this comment. Applying this replacement would lead to a sentence that does not express what I intended to express though. Cost _can_ vary a lot, but it must not vary a lot if the system is designed right.

Therefore, I changed the sentence in the following way: “[…] cost differences between supply options can be low if designed right.”

> Line 79: Delete ‘To be able “. Start with To identify…

Changed accordingly.

> Line 168: Rooftop PV. Is this residential? If so, how is the cost calculated? I don’t know how it works in Europe, but in the US, individual households finance their rooftop systems. This can be done in many ways…from leasing the panels, the buying the fully, to taking out a loan to cover their costs. What does the 880 represent in this case? For some households, the PV reduces electricity costs and, after the loan is paid off, represents a form of tax-free income. Does the 880 represent costs to the government if they were going to provide funding for residential PV? I have further comments below (rel. to line 359( about costs. You need to add some detail here about what exactly these costs represent and why they a reasonable input for your model. For example, it seems like cost to consumers is the most ideal cost to include in your model, but perhaps this is impossible to get.

The cost in Table 1 is installation and maintenance cost of electricity infrastructure. In that respect, it does not matter who is paying for the cost. I update the caption of the Table to make this clearer.

In the entire study, I am not considering cost of electricity of consumers, but rather total system cost — the total cost of electricity the European society has to pay for. With this focus, it does not matter whether rooftop PV is residential or not, as long as this does not impact total system cost. In my model, it does not.

You are of course correct in saying that rooftop PV can reduce cost of electricity of residential households, but this is usually due to many non-technical aspects like subsidies, levies, fees, and taxes. Such subsidies, levies, fees, and taxes have no impact on total system cost which I am assessing in this study.

Please see my response to your Main Comment 2 for a longer discussion and for the changes I applied in response to your comments on cost.

> Line 182: delete ‘in the following’

Changed accordingly.

> Line 183: I think it’s good to only consider uncertainty on wind and solar while holding the other technologies constant. However, it needs to be explained better. I think you structure the model so that cost does not affect supply shares, which is fine given the goal of understanding how different energy strategies change both land cover and costs. It seems to me the main reason for doing this is to deal with the uncertainty you have in future costs of the technology. The uniform distribution is good.

I agree with you that this paragraph was not clear enough. I added a longer explanation of why cost is not determining deployed capacity:

“Second, because cost has little impact on the system design. Because I enforce supply shares in the system design model, the extent to which supply technologies are deployed is determined by the enforced shares, not by cost.”

> Lines 192-199: The only problem with ref. 8 for an estimate of energy density is that many, if not all of the studies regarding onshore wind didn’t actual measure the area used. Instead, they estimated it. I only know of a few studies used aerial photography to directly digitize then estimate the area transformed by onshore wind. I don’t think these corrected for capacity factor. They may be worth comparing to ref. 8. Diffendorfer has a table of estimates in (ha/MW) similar to ref. 8. For example, Diffendorfer estimated an average of 0.93 ha/MW, which (if I did my conversions correctly) is 1.075 MW/ha or 107.5 W/m^2…much higher than 8.82 used in this paper. These values represent just the disturbance caused by the wind farm, not the entire project area.

> Diffendorfer, J.E., and R.W. Compton. “Land Cover and Topography Affect the Land Transformation Caused by Wind Facilities.” PLoS ONE 9, no. 2 (2014): e88914.

> Jones, N.F., and L. Pejchar. “Comparing the Ecological Impacts of Wind and Oil & Gas Development: A Landscape Scale Assessment.” PLoS ONE 8, no. 11 (2013): e81391

> Jones and Pejchar estimated 247m2/317 MMBTu.

> I’m not sure I would use “the wind turbines together with the technically necessary spacing between turbines”. The area between turbines is not useless. It can be agriculture, pasture, and habitat for many species. . You are correct that some land uses are not compatible with wind energy..homes between turbines would be very dangerous. But, at least in the US, turbines are only placed on areas where land use is compatabile…in farmlands, or pastures, or natural areas. The wind energy is not excluding the use of land…it’s exactly the opposite. Certain uses of the land excludes where wind energy goes. You could limit the potential area for wind energy in the EU by masking out urban areas, areas of building density above X/ha, certain distances to roads, areas near airports or weather radar (turbines affect radar). There are a number of GIS based wind energy potential maps out there.

> Furthermore, your point about visual impacts suggests you may want to use an area even larger than the spacing between turbines because the visual impacts of wind facilities go out quite far. This would would suggest even higher land requirements.

These are important observations. I understand there are in fact three questions here:

1. Should spacing be included?

2. Are the numbers correct considering spacing?

3. Should the number be even higher because I assume visual impacts?

1. I argue that it is correct to include spacing in this analysis (but I also added results and a discussion of excluding spacing). See my response to your Main Comment 3.

2. The publications you cite find indeed much lower land requirements (higher capacity densities) than the ones I am using. While I am using ~8.8 W/m2 mean, Differndorfer et al. find 108 W/m2. However, their number represents the direct surface impact which excludes spacing. Jones et al find 143 W/m2 habitat loss (when considering a capacity factor of 30%) which I did not consider at all in this study and which, I assume, does not include spacing either.

In fact, some other studies that include spacing find capacity densities lower than the one I am using here (Miller and Keith, 2018) (Nitsch et al., 2019). The reason I chose (van Zalk and Behrens, 2018) as the main reference for this study is that it combines several sources, and that it provides an uncertainty estimate which covers lower or higher values reported in other publications.

3. Yes, to correctly measure visual impact, the areas would need to be higher than the ones I am assuming in this study. The visual impact depends on the exact location of the infrastructure, for example to account for the elevation and thus visibility. Such a detailed assessment of visual impact is unfortunately not possible when considering all of Europe and when applying an energy system model. I therefore decided for using land requirements as defined in the manuscript as a proxy for visual impacts.

To address your comment, I amended the discussion on visual impacts in the second paragraph of the “Limitations and outlook” section. Please also see my response to your Main Comment 3 for further changes.

van Zalk, J., & Behrens, P. (2018). The spatial extent of renewable and non-renewable power generation: A review and meta-analysis of power densities and their application in the U.S. *Energy Policy*, *123* 83–91. https://doi.org/10.1016/j.enpol.2018.08.023

Miller, L. M., & Keith, D. W. (2018). Observation-based solar and wind power capacity factors and power densities. *Environmental Research Letters*, *13*(10), 104008. https://doi.org/10.1088/1748-9326/aae102

Nitsch, F., Turkovska, O., & Schmidt, J. (2019). Observation-based estimates of land availability for wind power: a case study for Czechia. *Energy, Sustainability and Society*, *9*(1), 45. https://doi.org/10.1186/s13705-019-0234-z

> Lines 204-206: Are these numbers truly theoretical maximums? If so, then the papers I cited above, which directly measured land use from wind facilities, seems to contradict them.

Please see my response to your comment on Lines 192–199. In short, I don’t think the papers contradict my numbers because they measure different quantities. However, land requirements of wind power are controversially debated and to accommodate the uncertainty originating in this debate, my study contains a global uncertainty analysis which includes values that are smaller and values that are larger than the “best estimate” I took from (van Zalk and Behrens, 2018).

> Line 228: Given how you measure capacity density for solar, it seems like estimate for wind that only include the real land disturbance might be more comparable. In most utlility scale PV systems I know of, areas between panels are cleared and not useable..similar to the areas around the base of wind turbines. Ultimately, you might want to consider using both the ‘footprint’ only calculation for wind as well as a large-scale estimate (as you’ve done). Other’s have done this (Denholm) and it gives a more clear picture of wind energy’s spatial impacts.

For both solar and wind power I estimate the full extent of the project area that is used for electricity generation and I use this as a proxy for landscape impact.

I agree with you that other measures are possible and that they would lead to different results. See Section 3 in Supporting Information S1 for an analysis that excludes spacing in wind farms.

Please see also my responses to your Main Comment 3 and to your comment on Lines 192—199.

> Line 272: “Lastly, the transmission grid already has significant land requirements….”

Changed accordingly.

> Line 281: Change to low spatial resolution. Higher spatial resolution = smaller units of area per pixel. 1m^2/pixel is higher resolution than entire countries. Your study is very low spatial resolution.

You are absolutely right. Changed accordingly.

> Line 289: “by applying” do you mean you multiplied total energy * 1/capacity density? Please use equations when necessary to better explain your methods. I realize it was an MC approach, but the underlying process that happens at each iteration can be described. Also, how did you deal with the outputs. Did you simply develop summary statistics (means and sd) for each scenario?

Yes, your assumption is correct, as is your remark that this part is not precise enough. The MC approach is not more than what you describe: for each input sample I derive an output sample of land requirements using the equation you are giving. For cost, I do something similar: I scale cost of the system design with the input sample (scaling factor from Table 2).

Regarding your question about outputs: I do not aggregate outputs. Fig 2 shows all ~29 million samples of the outputs.

To address your comment, I made minor changes to this paragraph so that the method is described more precisely.

> Line 299a: Figure 2 is a nice summary. What is of most interest to me are those scenarios that generate both a low cost and a low land use. Thus a scatterplot of land use vs cost would be a nice panel to add. You could then describe the %’s of the energy mixes for either extreme point…lowest cost and lowest land use vs highest cost and highest land use. Just seeing the shape of joint distribution would be helpful to me as it might show trade-offs or system bounds…for example there are no cases of low land use and low cost (a pet hypothesis of mine…particularly if you want to place wind in areas with minimized environmental impacts…it will cost more).

Your questions are indeed valid questions. In my opinion, the insights you are looking for are shown in Fig 3 and 4. Fig 4 shows the scatterplot you are looking for — even if only for expected values. The figure shows that your hypothesis is correct that there are no cases that have low land requirements and low cost at the same time — meaning that there is a trade-off. The same figure also shows low cost and low land requirements (Pareto-optimal) cases in dark blue.

A scatterplot of all cases is not possible due to the amount of points to plot (29 million). In fact, the plotting libraries I use (seaborn and matplotlib) do not even manage to generate a kernel density estimate of a dataset as large. However, I think showing the scatterplot for expected values as I do in Fig 4 is more insightful in any case. The variability in the 29 million cases has two very different sources: uncertainty about model parameters, and supply shares (which are a design choice). By freezing the former in Fig 3 and 4, the analysis can focus on the impact of supply shares and show the trade-offs and bounds you are looking for.

> Line 299b: I’d make the ranges be 0% and ~3% and ~0.06 and ~0.10 as the graphs show higher and lower values.

You are right that this was ambiguous. I changed the beginning of the sentence to “Among almost all of the ~29 million observations, […]” to be more precise.

> Line 299c: You don’t need the sentence “these ranges include…” nore the next sentence. You state the figure represents all 29 million cases.

You are correct that this is redundant information for the attentive reader. I added the information for two reasons. First, readers may choose to skim through the article and this information should help to understand the figure. Second, I wanted to make sure the reader understands that variability in the graph does not stem from uncertainty only, but also from design choices.

To address your comment, I moved the information into the caption of Fig. 1. In this way, it’s still available for the readers that may choose to skim through the article, but it’s less disturbing for the more attentive readers.

> Line 305: I don’t disagree with your statement here but think some readers my wonder why 0-3% is a ‘vast’ difference in land use and others will think that a doubling of their electric bill is a HUGE increase! In my world, an energy company has a very large fight if they try to raise electricity rates by just 3%...this really hurts low income households. So, it would be good to try to put these modelled changes in context..both social (for cost) and perhaps environmental? (for land use).

I fully agree with you that the text should not claim that a doubling of cost is less important to Europe’s population than the ranges of land requirements I found. In fact, I do not know if a doubling in cost is more or less important than a doubling in land requirements. I am not aiming at weighing cost against land requirements in this text. I see this as a political decision, not a conclusion of scientific work.

Instead, what I was trying to say here is that the range is larger. Cost doubles in the extreme cases, but you can easily find cases in which land requirements increase by an order of magnitude (0.3%—3%) or more. To reflect this better and to address your valid remark I changed the sentence to “The observations show that land requirements of European electricity systems can vary by more than an order of magnitude while their cost does not exceed twice the lowest cost.“.

> Line 311: Figure 3 would better match the text if the cost plots were in Eur per KWH and the land use plots in % of Europe’s land total. Right now the scales don’t match what’s described in the text. I can see the logic for the current ‘difference from minimum’ scale.Perhaps % change from minimum would be better here since you describe % changes in the text quite often. Your results will change considerably if you used the estimate of energy capacity for wind energy that considers just the land used, not the space between turbines.

Thank you for spotting this. I think numbers are best given relative here as I’d like the reader to focus on the relative differences between the cases rather than absolute numbers. This is also in line with the main text which mainly focusses on relative differences. But you are right that the main text additionally states absolute numbers which cannot be found in the figure.

To correct this, I added the absolute numbers of the case to which all numbers are given relatively to the figure caption.

> Line 340: You are correct, but somewhere you should acknowledge that offshore wind also has impacts…these are not on land obviously, but there is a growing literature on impacts from offshore wind on marine ecosystems and birds. I’m not sure but offshore sites likely restrict industrial fishing? There is also visual/social and cultural issues if the turbines can be seen from shore.

You are right. The “Limitations and outlook” section now includes the following sentence:

“Offshore wind of course requires no land but competes with other uses of offshore areas and can have visual impacts if the turbines are close to the shore.”

> Line 359: Cost becomes important here. See my comments above about cost of PV. Are your costs estimates similar to the levelized cost of electricity (LCOE) or do they vary by who is paying for the energy? Residential PV is paid for (typically) by the homeowner, while cost of industrial wind energy could be calculated as the cost to consumers, the price it costs the wind energy company to generate X capacity, etc. The broader point being that residential PV may not have the high cost when viewed from the households perspective over the life of the panels.

Please see my response to your Main Comment 2 and Comment on Line 168 for a longer discussion and for the changes I applied in response to your comments on cost.

> Line 529: While I’m not entirely sure what you mean by the impacts on landscapes, the following papers might be relevant:

> “Geographic Context Affects the Landscape Change and Fragmentation Caused by Wind Energy Facilities [PeerJ].” Accessed April 2, 2020. https://peerj.com/articles/7129/.

> Jones, N.F., L. Pejchar, and J.M. Kiesecker. “The Energy Footprint: How Oil, Natural Gas, and Wind Energy Affect Land for Biodiversity and the Flow of Ecosystem Services.” BioScience, 2015. http://bioscience.oxfordjournals.org/content/early/2015/01/22/biosci.biu224.abstract.

> “Monitoring Wind Farms Occupying Grasslands Based on Remote-Sensing Data from China’s GF-2 HD Satellite—A Case Study of Jiuquan City, Gansu Province, China - ScienceDirect.” Accessed April 2, 2020. https://www.sciencedirect.com/science/article/abs/pii/S092134491630163X?via%3Dihub.

> “Energy Development in Colorado’s Pawnee National Grasslands: Mapping and Measuring the Disturbance Footprint of Renewables and Non-Renewables | SpringerLink.” Accessed April 2, 2020. https://link.springer.com/article/10.1007%2Fs00267-017-0846-z.

Thank you for pointing me to this literature which I enjoyed reading. In this work I care for the visual impacts on landscapes. In this respect, I found especially (Diffendorfer et al., 2019) relevant, which I am now referencing in this first paragraph of the “Limitations and outlook” section.

Diffendorfer, J. E., Dorning, M. A., Keen, J. R., Kramer, L. A., & Taylor, R. V. (2019). Geographic context affects the landscape change and fragmentation caused by wind energy facilities. *PeerJ*, *7*, e7129. https://doi.org/10.7717/peerj.7129

> Line 532: How much would your results change if you could model transmission lines and how much that contributes to overall land requirements for a given energy scenario? I’m not sure if more offshore would mean more transmission, but it could.

This is an interesting question. An answer backed by model results would require a model that is higher resolved in space, which would increase the computational effort of this study substantially.

To address your valid question, I did two things.

First, I made the estimation of current land requirements of the transmission grid more precise. The previous estimation was an upper bound, because it included a larger area than the one I am assessing and it included circuit length, not line length. As many lines have more than one circuit, this number is too large. I now use a different method and different data, see the last paragraph in the “Land requirements” section. The new estimation of today’s land requirements is 0.12% of total land.

Second, I use two former studies with subnational spatial resolution to give a generous upper bound for the land requirements of the transmission grid in a fully renewable electricity system of 0.3% (same paragraph in the manuscript).

I do not know how land requirements vary with supply shares, but I am confident to say that this variation will likely be in the range between 0.12%—0.3% of total land and thus small compared to the variation stemming from supply technologies.

## Reviewer #2:

> This paper addresses a large concern with renewable energy development, land use requirements.

> Since I do not have the expertise to address the economic sections of this manuscript, I will focus on the land use requirements. Therefore I will provide more high-level comments than detailed.

> Comment 1: I would have like to see more about the context of the foot print and cost in European countries (e.g., impact to natural lands)

In several responses to comments from Reviewer #1 I made changes to the manuscript that made the definition of cost and land requirements clearer. For cost, these changes are mainly in the “System design model” and “Limitations and Outlook” sections. For land requirements, these are mainly in the “Land requirements” section and in the Support Information S1.

> Comment 2: There has been a lot of work done on this matter in USA. Please explain why the “Energy Sprawl” work was not cited in this paper?

Thank you for hinting me to the energy sprawl work in the USA. I am now citing (McDonald et al., 2009) in the introduction, as I found this reference to be most related to this study and the situation in Europe.

McDonald, R. I., Fargione, J., Kiesecker, J., Miller, W. M., & Powell, J. (2009). Energy Sprawl or Energy Efficiency: Climate Policy Impacts on Natural Habitat for the United States of America. *PLoS ONE*, *4*(8), e6802. https://doi.org/10.1371/journal.pone.0006802

> Comment 3: Would it have been possible to give summary regarding different countries? It seems like Europe was considered one large mass. How do these results like up with European renewable energy policies to meet future energy demands?

I agree with you that an analysis on the national level would be very insightful, in particular for decisions to be made on the national level. I abstained from analysing the national level mainly because the study is designed in a way to be insightful for Europe as a whole in the first place. When I analyse the impact of different supply shares, I vary supply shares in all countries in Europe in parallel. One might want to analyse more complex variations in which countries have different supply shares (for example lower solar shares in the Northern countries). However, I did not include such variations and doing so would increase the space of possible solutions a lot.

I further agree with you that a study with strong links to existing national policies would be very insightful, but it would also require scenarios that are informed by these policies and therefore a different research design than mine. In my study, I focus on the solution space of the entire continent.

To address you comment, I created a map visualising the expected land requirements of the case with minimal expected cost, see Figure 1 in S1 Supporting Information. Because all countries are connected, it is difficult to allocate cost to countries and therefore I did not create a similar map for cost.

> Comment 4: On the above note, the author assumes no new hydropower dam development. Is this realistic?

You are correct that this is a conservative assumption. Indeed, a previous study finds that European hydropower generation can be increased by 50% based on technical, economic, and ecological constraints (Gernaat et al., 2017). However, the authors estimate that 85% of this increase would come from run-of-river plants without dams and they ignore social constraints. Further, integrating their results into my model is challenging as the spatial distribution of the generation increase and the associated time series of water in-feed into the new hydro stations are not known to me. Partly because of this challenge, a no-expansion assumption for hydropower is quite common in European models, e.g. (Zappa et al., 2019), (Hörsch et al., 2018), and (Pleßmann & Blechinger, 2017).

I consider the impact of this conservative assumption to be small for my results. While larger capacities of run-of-river hydropower will decrease the amount of solar and wind power necessary, it will do so only in a very limited magnitude (from today ~10% to max 15% of total demand (Publications Office of the European Union, 2019)) and it will have largely the same impact in all 286 cases. Only the small fraction of future hydropower that can be equipped with a dam can impact my results, as more dispatchable hydropower can reduce the need for other forms of flexibility and will thus likely decrease cost especially of solar-centred cases. Because the potential for hydropower with reservoir is small, I consider its impact to be small as well.

Based on your comment, I added a discussion of this aspect to Section 1 of S1 Supporting Information and link to it from the “System design model” section in the main text.

Publications Office of the European Union. (2019). *EU energy in figures: statistical pocketbook 2019*. European Union.

Zappa, W., Junginger, M., & van den Broek, M. (2019). Is a 100% renewable European power system feasible by 2050? *Applied Energy*, *233*–*234*, 1027–1050. https://doi.org/10.1016/j.apenergy.2018.08.109

Hörsch, J., Hofmann, F., Schlachtberger, D., & Brown, T. (2018). PyPSA-Eur: An open optimisation model of the European transmission system. *Energy Strategy Reviews*, *22*, 207–215. https://doi.org/10.1016/j.esr.2018.08.012

Pleßmann, G., & Blechinger, P. (2017). How to meet EU GHG emission reduction targets? A model based decarbonization pathway for Europe’s electricity supply system until 2050. *Energy Strategy Reviews*, *15*, 19–32. https://doi.org/10.1016/j.esr.2016.11.003

Gernaat, D. E. H. J., Bogaart, P. W., Vuuren, D. P. van, Biemans, H., & Niessink, R. (2017). High-resolution assessment of global technical and economic hydropower potential. *Nature Energy*, *2*(10), 821–828. https://doi.org/10.1038/s41560-017-0006-y

> Comment 5: It seems the author included roads and transmission lines into the land use estimate for solar and wind but not hydroelectic. This does not seem valid.

It is correct that the land requirements estimates for solar and wind power include access roads. They do include transmission lines only insofar as they are used on-site to connect to the transmission grid.

The estimations of land requirements of hydro electricity are much more coarse, and this is in particular because hydro electricity is not the focus of this article. Compared to solar and wind capacity, for which large expansions of 500% or more can be expected in Europe, the expansion potential of hydro electricity is small. As I explained in my response to your Comment 4, this small expansion potential stems largely from hydro electricity without reservoir whose land requirements are small (Smil, 2015). A significant land use change due to expansion of hydro electricity is therefore unlikely in Europe.

Because reservoirs are the main driver of hydro’s land requirements, I use their extent as an estimate of the land requirements of Europe. Because by far not all reservoirs are used for electricity generation, and even less exclusively for electricity generation, I consider the estimate based on reservoirs as a conservative upper bound for the land requirements of hydro electricity in Europe.

To address your comment, I made sure these aspects are clearer explained in the article.

1. I amended the description of the land requirements of hydro electricity in section “Land requirements”.

2. I further amended the description of the land requirements of solar and wind power in the second paragraph of the same section.

Smil, V. (2015). *Power Density: A Key to Understanding Energy Sources and Uses*. The MIT Press.

Attachment

Submitted filename: response-to-reviewers.pdf

Click here for additional data file.^{(136.9KB, pdf)}

PLoS One. doi: 10.1371/journal.pone.0236958.r003

Decision Letter 1

Baogui Xin

17 Jul 2020

Supply-side options to reduce land requirements of fully renewable electricity in Europe

PONE-D-20-07241R1

Dear Dr. Tröndle,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Baogui Xin, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

PLoS One. doi: 10.1371/journal.pone.0236958.r004

Acceptance letter

Baogui Xin

24 Jul 2020

PONE-D-20-07241R1

Supply-side options to reduce land requirements of fully renewable electricity in Europe

Dear Dr. Tröndle:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Baogui Xin

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File

(PDF)

Click here for additional data file.^{(1.1MB, pdf)}

S1 Data

(NC)

Click here for additional data file.^{(7.6MB, nc)}

S1 Code

(ZIP)

Click here for additional data file.^{(274.9KB, zip)}

Attachment

Submitted filename: response-to-reviewers.pdf

Click here for additional data file.^{(136.9KB, pdf)}

Data Availability Statement

The applied model of the electricity system and all result data are available from Zenodo doi:10.5281/zenodo.3707812.

[pone.0236958.ref001] 1.Jacobson MZ, Delucchi MA, Bauer ZAF, Goodman SC, Chapman WE, Cameron MA, et al. 100% Clean and Renewable Wind, Water, and Sunlight All-Sector Energy Roadmaps for 139 Countries of the World. Joule. 2017;1: 108–121. 10.1016/j.joule.2017.07.005 [DOI] [Google Scholar]

[pone.0236958.ref002] 2.Ruiz P, Nijs W, Tarvydas D, Sgobbi A, Zucker A, Pilli R, et al. ENSPRESO—an open, EU-28 wide, transparent and coherent database of wind, solar and biomass energy potentials. Energy Strategy Reviews. 2019;26: 100379 10.1016/j.esr.2019.100379 [DOI] [Google Scholar]

[pone.0236958.ref003] 3.Tröndle T, Pfenninger S, Lilliestam J. Home-made or imported: On the possibility for renewable electricity autarky on all scales in Europe. Energy Strategy Reviews. 2019;26 10.1016/j.esr.2019.100388 [DOI] [Google Scholar]

[pone.0236958.ref004] 4.European Commission. COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE EUROPEAN COUNCIL, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE, THE COMMITTEE OF THE REGIONS AND THE EUROPEAN INVESTMENT BANK A Clean Planet for all A European strategic long-term vision for a prosperous, modern, competitive and climate neutral economy. Brussels: European Commission; 2018. Nov. Available: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52018DC0773 [Google Scholar]

[pone.0236958.ref005] 5.MacKay DJC. Sustainable Energy without the hot air. UIT; 2008. [Google Scholar]

[pone.0236958.ref006] 6.McDonald RI, Fargione J, Kiesecker J, Miller WM, Powell J. Energy Sprawl or Energy Efficiency: Climate Policy Impacts on Natural Habitat for the United States of America. Añel JA, editor. PLoS ONE. 2009;4: e6802 10.1371/journal.pone.0006802 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0236958.ref007] 7.Smil V. Power Density: A Key to Understanding Energy Sources and Uses. Cambridge, Massachusetts: The MIT Press; 2015. [Google Scholar]

[pone.0236958.ref008] 8.van Zalk J, Behrens P. The spatial extent of renewable and non-renewable power generation: A review and meta-analysis of power densities and their application in the U.S. Energy Policy. 2018;123: 83–91. 10.1016/j.enpol.2018.08.023 [DOI] [Google Scholar]

[pone.0236958.ref009] 9.Brinkley C, Leach A. Energy next door: A meta-analysis of energy infrastructure impact on housing value. Energy Research & Social Science. 2019;50: 51–65. 10.1016/j.erss.2018.11.014 [DOI] [Google Scholar]

[pone.0236958.ref010] 10.Molnarova K, Sklenicka P, Stiborek J, Svobodova K, Salek M, Brabec E. Visual preferences for wind turbines: Location, numbers and respondent characteristics. Applied Energy. 2012;92: 269–278. 10.1016/j.apenergy.2011.11.001 [DOI] [Google Scholar]

[pone.0236958.ref011] 11.Sánchez-Pantoja N, Vidal R, Pastor MC. Aesthetic impact of solar energy systems. Renewable and Sustainable Energy Reviews. 2018;98: 227–238. 10.1016/j.rser.2018.09.021 [DOI] [Google Scholar]

[pone.0236958.ref012] 12.Devine-Wright P, Howes Y. Disruption to place attachment and the protection of restorative environments: A wind energy case study. Journal of Environmental Psychology. 2010;30: 271–280. 10.1016/j.jenvp.2010.01.008 [DOI] [Google Scholar]

[pone.0236958.ref013] 13.de Vries S, de Groot M, Boers J. Eyesores in sight: Quantifying the impact of man-made elements on the scenic beauty of Dutch landscapes. Landscape and Urban Planning. 2012;105: 118–127. 10.1016/j.landurbplan.2011.12.005 [DOI] [Google Scholar]

[pone.0236958.ref014] 14.Knebel A, Hartman J, Kajimura R. Opinions on renewables—a look at polls in industrialised countries. Berlin: Renewable Energies Agency; 2016. Mar. Available: https://www.unendlich-viel-energie.de/media/file/427.AEE_RK29_%20Internationale_Akzeptanzumfragen_EN.pdf [Google Scholar]

[pone.0236958.ref015] 15.Setton D. Social Sustainability Barometer for the German Energiewende: 2018 Edition Core statements and summary of the key findings. Potsdam: Institute for Advanced Sustainability Studies; 2019. Jul. 10.2312/iass.2019.014 [DOI] [Google Scholar]

[pone.0236958.ref016] 16.Stokes LC. Electoral Backlash against Climate Policy: A Natural Experiment on Retrospective Voting and Local Resistance to Public Policy. American Journal of Political Science. 2016;60: 958–974. 10.1111/ajps.12220 [DOI] [Google Scholar]

[pone.0236958.ref017] 17.Späth L. Large-scale photovoltaics? Yes please, but not like this! Insights on different perspectives underlying the trade-off between land use and renewable electricity development. Energy Policy. 2018;122: 429–437. 10.1016/j.enpol.2018.07.029 [DOI] [Google Scholar]

[pone.0236958.ref018] 18.Palmer-Wilson K, Donald J, Robertson B, Lyseng B, Keller V, Fowler M, et al. Impact of land requirements on electricity system decarbonisation pathways. Energy Policy. 2019;129: 193–205. 10.1016/j.enpol.2019.01.071 [DOI] [Google Scholar]

[pone.0236958.ref019] 19.Schlachtberger DP, Brown T, Schramm S, Greiner M. The benefits of cooperation in a highly renewable European electricity network. Energy. 2017;134: 469–481. 10.1016/j.energy.2017.06.004 [DOI] [Google Scholar]

[pone.0236958.ref020] 20.Neumann F, Brown T. The Near-Optimal Feasible Space of a Renewable Power System Model. 2019. [cited 28 Oct 2019]. Available: http://arxiv.org/abs/1910.01891 [Google Scholar]

[pone.0236958.ref021] 21.Tröndle T, Lilliestam J, Marelli S, Pfenninger S. Trade-offs between geographic scale, cost, and infrastructure requirements for fully renewable electricity in Europe. Joule. 2020. doi: Forthcoming [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0236958.ref022] 22.Bussar C, Moos M, Alvarez R, Wolf P, Thien T, Chen H, et al. Optimal Allocation and Capacity of Energy Storage Systems in a Future European Power System with 100% Renewable Energy Generation. Energy Procedia. 2014;46: 40–47. 10.1016/j.egypro.2014.01.156 [DOI] [Google Scholar]

[pone.0236958.ref023] 23.Connolly D, Lund H, Mathiesen BV. Smart Energy Europe: The technical and economic impact of one potential 100% renewable energy scenario for the European Union. Renewable and Sustainable Energy Reviews. 2016;60: 1634–1653. 10.1016/j.rser.2016.02.025 [DOI] [Google Scholar]

[pone.0236958.ref024] 24.Pleßmann G, Blechinger P. How to meet EU GHG emission reduction targets? A model based decarbonization pathway for Europe’s electricity supply system until 2050. Energy Strategy Reviews. 2017;15: 19–32. 10.1016/j.esr.2016.11.003 [DOI] [Google Scholar]

[pone.0236958.ref025] 25.Brown TW, Bischof-Niemz T, Blok K, Breyer C, Lund H, Mathiesen BV. Response to ‘Burden of proof: A comprehensive review of the feasibility of 100% renewable-electricity systems’. Renewable and Sustainable Energy Reviews. 2018;92: 834–847. 10.1016/j.rser.2018.04.113 [DOI] [Google Scholar]

[pone.0236958.ref026] 26.Child M, Kemfert C, Bogdanov D, Breyer C. Flexible electricity generation, grid exchange and storage for the transition to a 100% renewable energy system in Europe. Renewable Energy. 2019;139: 80–101. 10.1016/j.renene.2019.02.077 [DOI] [Google Scholar]

[pone.0236958.ref027] 27.Fthenakis V, Kim HC. Land use and electricity generation: A life-cycle analysis. Renewable and Sustainable Energy Reviews. 2009;13: 1465–1474. 10.1016/j.rser.2008.09.017 [DOI] [Google Scholar]

[pone.0236958.ref028] 28.Capellán-Pérez I, de Castro C, Arto I. Assessing vulnerabilities and limits in the transition to renewable energies: Land requirements under 100% solar energy scenarios. Renewable and Sustainable Energy Reviews. 2017;77: 760–782. 10.1016/j.rser.2017.03.137 [DOI] [Google Scholar]

[pone.0236958.ref029] 29.Tröndle T. Code related to "Tröndle (2020)–Supply-side options to reduce land requirements of fully renewable electricity in Europe". Potsdam: Institute for Advanced Sustainability Studies; 2020. 10.5281/zenodo.3956531 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0236958.ref030] 30.Koster J, Rahmann S. Snakemake–a scalable bioinformatics workflow engine. Bioinformatics. 2012;28: 2520–2522. 10.1093/bioinformatics/bts480 [DOI] [PubMed] [Google Scholar]

[pone.0236958.ref031] 31.Tröndle T. euro-calliope. Potsdam: Institute for Advanced Sustainability Studies; 2020. 10.5281/zenodo.3949794 [DOI] [Google Scholar]

[pone.0236958.ref032] 32.Pfenninger S, Pickering B. Calliope: A multi-scale energy systems modelling framework. The Journal of Open Source Software. 2018. [cited 3 Jun 2019]. 10.21105/joss.00825 [DOI] [Google Scholar]

[pone.0236958.ref033] 33.Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P, Kobayashi S, et al. The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society. 2011;137: 553–597. 10.1002/qj.828 [DOI] [Google Scholar]

[pone.0236958.ref034] 34.Pfenninger S, Staffell I. Long-term patterns of European PV output using 30 years of validated hourly reanalysis and satellite data. Energy. 2016;114: 1251–1265. 10.1016/j.energy.2016.08.060 [DOI] [Google Scholar]

[pone.0236958.ref035] 35.Staffell I, Pfenninger S. Using bias-corrected reanalysis to simulate current and future wind power output. Energy. 2016;114: 1224–1239. 10.1016/j.energy.2016.08.068 [DOI] [Google Scholar]

[pone.0236958.ref036] 36.Ruiz Castello P, Sgobbi A, Nijs W, Thiel C, Dalla Longa F, Kober T, et al. The JRC-EU-TIMES model. Bioenergy potentials for EU and neighbouring countries. 2015. Dec. 10.2790/39014 [DOI] [Google Scholar]

[pone.0236958.ref037] 37.Steffen B. Estimating the Cost of Capital for Renewable Energy Projects. Rochester, NY: Social Science Research Network; 2019. Apr. Report No.: ID 3373905. Available: https://papers.ssrn.com/abstract=3373905 [Google Scholar]

[pone.0236958.ref038] 38.JRC. ETRI 2014 –Energy Technology Reference Indicator projections for 2010–2050. JRC; 2014. Available: https://setis.ec.europa.eu/related-jrc-activities/jrc-setis-reports/etri-2014 [Google Scholar]

[pone.0236958.ref039] 39.Schmidt O, Melchior S, Hawkes A, Staffell I. Projecting the Future Levelized Cost of Electricity Storage Technologies. Joule. 2019;0. 10.1016/j.joule.2018.12.008 [DOI] [Google Scholar]

[pone.0236958.ref040] 40.Wirth H. Recent Facts about Photovoltaics in Germany. Freiburg im Breisgau: Fraunhofer ISE; 2020. January p. 98 Available: https://www.ise.fraunhofer.de/en/publications/studies/recent-facts-about-pv-in-germany.html [Google Scholar]

[pone.0236958.ref041] 41.Turney D, Fthenakis V. Environmental impacts from the installation and operation of large-scale solar power plants. Renewable and Sustainable Energy Reviews. 2011;15: 3261–3270. 10.1016/j.rser.2011.04.023 [DOI] [Google Scholar]

[pone.0236958.ref042] 42.Wirth H. Aktuelle Fakten zur Photovoltaik in Deutschland. Fraunhofer: ISE; 2017. Aug. Available: https://www.ise.fraunhofer.de/de/daten-zu-erneuerbaren-energien.html [Google Scholar]

[pone.0236958.ref043] 43.McKenna R, Hollnaicher S, Ostman v. d. Leye P, Fichtner W. Cost-potentials for large onshore wind turbines in Europe. Energy. 2015;83: 217–229. 10.1016/j.energy.2015.02.016 [DOI] [Google Scholar]

[pone.0236958.ref044] 44.Miller LM, Keith DW. Corrigendum: Observation-based solar and wind power capacity factors and power densities (2018 Environ. Res. Lett. 13 104008). Environ Res Lett. 2019. 10.1088/1748-9326/aaf9cf [DOI] [Google Scholar]

[pone.0236958.ref045] 45.Nitsch F, Turkovska O, Schmidt J. Observation-based estimates of land availability for wind power: A case study for Czechia. Energ Sustain Soc. 2019;9: 45 10.1186/s13705-019-0234-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0236958.ref046] 46.ZSW, Bosch & Partner. Vorbereitung und Begleitung bei der Erstellung eines Erfahrungsberichts gemäß § 97 Erneuerbare-Energien-Gesetz–Teilvorhaben II c: Solare Strahlungsenergie. German Federal Ministry for Economic Affairs and Energy; 2019. March p. 169 Available: https://www.erneuerbare-energien.de/EE/Redaktion/DE/Downloads/bmwi_de/zsv-boschundpartner-vorbereitung-begleitung-eeg.pdf?__blob=publicationFile&v=7 [Google Scholar]

[pone.0236958.ref047] 47.Halleraker J, van de Bund W, Bussettini M, Gosling R, Döbbelt-Grüne S, Hensman J, et al. Working Group ECOSTAT report on Common understanding of using mitigation measures for reaching Good Ecological Potential for heavily modified water bodies. Ispra: JRC; 2016. p. 104 10.2760/649695 [DOI] [Google Scholar]

[pone.0236958.ref048] 48.Aggreko Y. Cube—Energy storage–Product Guide. 2020. [cited 29 Feb 2020]. Available: https://www.aggreko.com/-/media/Aggreko/Files/PDF/Energy-Storage/YCubeProductGuideFinal.pdf [Google Scholar]

[pone.0236958.ref049] 49.Wiegmans B. Gridkit: Gridkit 1.0 ‘For Scientists’. Zenodo; 2016. 10.5281/ZENODO.47263 [DOI] [Google Scholar]

[pone.0236958.ref050] 50.Hörsch J, Hofmann F, Schlachtberger D, Brown T. PyPSA-Eur: An open optimisation model of the European transmission system. Energy Strategy Reviews. 2018;22: 207–215. 10.1016/j.esr.2018.08.012 [DOI] [Google Scholar]

[pone.0236958.ref051] 51.Stevens L, Anderson B, Cowan C, Colton K, Johnson D. The footprint of energy: Land use of U.S. Electricity production. Strata; 2017. Jun. [Google Scholar]

[pone.0236958.ref052] 52.Hörsch J, Brown T. The role of spatial scale in joint optimisations of generation and transmission for European highly renewable scenarios. 2017 14th International Conference on the European Energy Market (EEM). Dresden, Germany: IEEE; 2017. pp. 1–7. 10.1109/EEM.2017.7982024 [DOI]

[pone.0236958.ref053] 53.Herman J, Usher W. SALib: An open-source Python library for Sensitivity Analysis. JOSS. 2017;2: 97 10.21105/joss.00097 [DOI] [Google Scholar]

[pone.0236958.ref054] 54.Bódis K, Kougias I, Jäger-Waldau A, Taylor N, Szabó S. A high-resolution geospatial assessment of the rooftop solar photovoltaic potential in the European Union. Renewable and Sustainable Energy Reviews. 2019;114: 109309 10.1016/j.rser.2019.109309 [DOI] [Google Scholar]

[pone.0236958.ref055] 55.de Castro C, Mediavilla M, Miguel LJ, Frechoso F. Global solar electric potential: A review of their technical and sustainable limits. Renewable and Sustainable Energy Reviews. 2013;28: 824–835. 10.1016/j.rser.2013.08.040 [DOI] [Google Scholar]

[pone.0236958.ref056] 56.Caglayan DG, Ryberg DS, Heinrichs H, Linßen J, Stolten D, Robinius M. The techno-economic potential of offshore wind energy with optimized future turbine designs in Europe. Applied Energy. 2019;255: 113794 10.1016/j.apenergy.2019.113794 [DOI] [Google Scholar]

[pone.0236958.ref057] 57.IEA. Offshore Wind Outlook 2019: World Energy Outlook Special Report. Paris; 2019 Nov p. 98. Available: https://www.iea.org/reports/offshore-wind-outlook-2019/geospatial-analysis

[pone.0236958.ref058] 58.Weselek A, Ehmann A, Zikeli S, Lewandowski I, Schindele S, Högy P. Agrophotovoltaic systems: Applications, challenges, and opportunities. A review. Agron Sustain Dev. 2019;39: 35 10.1007/s13593-019-0581-3 [DOI] [Google Scholar]

[pone.0236958.ref059] 59.Diffendorfer JE, Dorning MA, Keen JR, Kramer LA, Taylor RV. Geographic context affects the landscape change and fragmentation caused by wind energy facilities. PeerJ. 2019;7: e7129 10.7717/peerj.7129 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Supply-side options to reduce land requirements of fully renewable electricity in Europe

Tim Tröndle

Roles

Abstract

1. Introduction

2. Data and methods

2.1 Capacity shares of supply technologies

Fig 1. Capacity shares of all 286 system designs for four supply technologies.

2.2 System design model

Table 1. Assumptions on installation and maintenance cost of electricity infrastructure.

2.3 Cost uncertainty

Table 2. Uncertain input parameters.

2.4 Land requirements

2.5 Stochastic model

3. Results

3.1 Renewable electricity supply with vastly different land requirements

Fig 2. All ~29 million observations of cost and land requirements of fully renewable electricity systems.

Fig 3. Expected cost and land requirements of fully renewable electricity systems with all possible shares of three different supply technologies.

3.2 Offshore wind reduces land requirements most cost-effectively

Fig 4. Cost-effective ways to reduce expected land requirements using supply technologies individually.

3.3 Cost penalties of 20% or less are most likely even for low land requirements

Fig 5. Resulting cost penalties to ensure land thresholds.

3.4 Low land requirements require low shares of onshore wind

Fig 6. Maximal capacity shares of onshore wind to ensure land thresholds.

3.5 Storage and flexible generation require small amounts of land

Fig 7. Flexibility needs of fully renewable electricity systems with all possible shares of three different supply technologies.

4. Discussion

4.1 Comparison to previous studies

4.2 Limitations and outlook

4.3 Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Baogui Xin

Roles

Author response to Decision Letter 0

Decision Letter 1

Baogui Xin

Roles

Acceptance letter

Baogui Xin

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases