Content uploaded by Dirk Oudes
Author content
All content in this area was uploaded by Dirk Oudes on Apr 29, 2021
Content may be subject to copyright.
Renewable and Sustainable Energy Reviews 145 (2021) 111101
1364-0321/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Next generation solar power plants? A comparative analysis of frontrunner
solar landscapes in Europe
D. Oudes
a
,
b
,
*
, S. Stremke
a
,
b
a
Wageningen University & Research, Environmental Sciences Group, Landscape Architecture Chair Group, PO Box 47, 6700 AA, Wageningen, the Netherlands
b
Academy of Architecture, Amsterdam University of the Arts, 1011 PG, Amsterdam, the Netherlands
ARTICLE INFO
Keywords:
Utility scale solar energy (USSE)
Ground mounted solar system
Photovoltaic landscape
Solar park
Energy landscape
Energy transition
Ecosystem services
Case study
ABSTRACT
Solar power plants transform the existing landscape. This landscape change raises concerns about visual impact,
land use competition and the end-of-life stage of solar power plants. Existing research stresses the need to address
these concerns, arguing for a combined spatial arrangement of solar power plant and landscape: solar landscape.
Solar landscapes share the aim to achieve other benets (e.g. reducing visibility, habitat creation) in addition to
electricity generation, yet empirical evidence on solar landscapes is scarce. This comparative analysis of 11
frontrunner cases aims to contribute to the understanding of solar landscapes, by studying the spatial properties
visibility, multifunctionality and temporality. Visibility is reduced in all cases. In ve cases, however, visibility is
partly enhanced in combination with recreational amenities. Between 6 and 14 provisioning, regulating and
cultural functions were found in the cases. Functions were located beneath arrays, between arrays and adjacent
to photovoltaic patches. Temporal considerations were identied in most cases, yet only two cases introduced
new landscape features to enhance future use of the sites after decomissioning. Across cases, this case study
shows how contemporary concerns about solar power plants, such as visual impact, land use competition and the
end-of-life stage are addressed. Although the cases altogether present a portfolio of measures responding to
societal concerns, the full potential of the three key properties is yet to be explored. Furthermore, this
comparative analysis highlights the need to address emerging trade-offs between spatial properties and to discern
between different types of solar landscapes. The used analytical framework may supplement the assessment of
solar power plants to examine not only negative, but also positive impacts.
1. Introduction
Solar power plants (SPP) have been constructed at an increasing rate
over the past decades [1]. These power plants, consisting of
ground-mounted photovoltaic (PV) arrays and electrical infrastructure,
transform the landscape [2–7]. Landscape is here dened as ‘an area, as
perceived by people, whose character is the result of the action and
interaction of natural and/or human factors’ [8]. SPP not only transform
existing landscape patterns, that is the size, shape, arrangement and
distribution of individual landscape elements [9], but also how the
landscape is perceived by inhabitants and other landscape users [10,11].
These landscape transformations raise societal concerns about visual
impact, land use competition and the end-of-life stage. Visual impact is a
key concern with respect to SPPs [2,12,13]. SPPs can have visual impact
due to their scale, color, pattern and articiality [2,14–16] and, as a
consequence, inuence perception adversely [11]. Furthermore, SPPs
require land previously occupied by other uses and therefore increase
land use pressure. SPPs can, for example, result in the loss of agricultural
land [17,18] and also affect habitats, as vegetation is degraded or
removed [17,19,20] and soil is moved or covered [18]. These land use
changes can be substantial in a short period of time [21] and require
recovery time for vegetation and soil [19]. The common life-span of SPP
is 20–30 years, due to the life expectancy of the modules [22]. Concerns
about the end-of-life stage of SPPs are whether decommissioning will
take place [23] and if so, what the state of the resulting landscape will be
[24]. All these three groups of concerns have a clear spatial dimension
Abbreviations: SPP, solar power plant; PV, photovoltaic; LAOR, land area occupation ratio; GCR, ground coverage ratio; CICES, Common International Classi-
cation of Ecosystem Services.
* Corresponding author. Wageningen University & Research, Environmental Sciences Group, Landscape Architecture Chair Group, PO Box 47, 6700 AA, Wage-
ningen, the Netherlands.
E-mail address: dirk.oudes@wur.nl (D. Oudes).
Contents lists available at ScienceDirect
Renewable and Sustainable Energy Reviews
journal homepage: http://www.elsevier.com/locate/rser
https://doi.org/10.1016/j.rser.2021.111101
Received 15 July 2020; Received in revised form 24 February 2021; Accepted 11 April 2021
Renewable and Sustainable Energy Reviews 145 (2021) 111101
2
and can result in negative responses of local inhabitants and other
landscape users towards SPP [2,11,12,25]. Consequently, these re-
sponses may threaten the progress of the energy transition [26,27].
Existing research points to the need of SPP to address societal con-
cerns, by attending to three key properties: visibility, multifunctionality
and temporality. Visibility refers to whether an SPP is observable by
landscape users from a certain location [13,28]. Visibility can be
changed, for example, by using vegetation for screening or adjusting the
size of the SPP to the characteristics of the host landscape [2,11,29].
Multifunctionality refers to the capacity of a certain area of land to serve
multiple purposes and fulll several needs at the same time [30–32].
Electricity generation in SPP can be combined, for example, with
ecological restoration [24,33–35] and outdoor education [2,24]. Tem-
porality is a relatively new, emerging topic in energy landscape research
and refers to the dynamic character of SPP [4,23]. Elements introduced
during the SPP construction have the potential to enhance the future
landscape or inhibit certain developments after decommissioning of the
solar infrastructure [4,24]. Temporality is also relevant in the context of
recycling energy landscapes: renewable energy technologies are intro-
duced at sites formerly used for conventional energies. In Nijmegen in
the Netherlands, for example, an SPP is built on a site previously
occupied by a coal-red power plant.
Others have recently introduced the concept of ‘photovoltaic land-
scape’ or ‘solar landscape’ that encompasses a joint approach between
SPP and landscape [2,36]. This approach involves a combined spatial
arrangement of SPP and landscape where solar infrastructure is adapted
(e.g. height of arrays, distance between arrays) and ‘landscape features’
are included (e.g. hedgerows, wildower meadows). While contempo-
rary spatial arrangements of SPPs are optimized for energy and/or
economic benets, spatial arrangements of solar landscapes aim to
achieve other benets (e.g. reducing visibility, habitat creation) in
addition to electricity generation [13,16,24,29]. For this paper, we
make use of and build upon the novel concept of solar landscapes to
examine SPPs that pay attention to visibility, multifunctionality and
temporality.
However, few studies have investigated the visual, functional and
temporal properties of constructed cases of solar landscapes. Lobaccaro
et al. [36] is the only study that examines spatial properties of built solar
landscapes. They partly address visibility and multifunctionality and do
not discuss temporality. Anyhow, most studies overlook the spatial
arrangement of SPP and landscape [14,37,38], focus on a single prop-
erty [16,35,39], or present theoretical discussions on what solar land-
scapes can be or should be [2,29,33], but not on what solar landscapes
are. The following research question is central to this paper: what are the
visual, functional and temporal properties of frontrunner solar landscapes in
Europe?
This research aims to contribute to the growing body of knowledge
on solar landscapes by analyzing and comparing the spatial properties of
constructed solar landscapes in Europe. This study used expert consul-
tation and desk-study to identify so-called ‘frontrunner’ SPPs. Insights in
the innovative properties of these frontrunner cases constitute a vital
contribution to the debate on how societal concerns about SPP can be
resolved. Due to the novelty of the topic, an analytical framework was
developed for the case study, based on a literature review. This frame-
work focusing on visibility, multifunctionality and temporality of SPP
may also enrich environmental impact assessments and multi-criteria
decision analyses of SPP in response to prominent societal concerns.
Furthermore, a better understanding of frontrunner cases, in combina-
tion with the cultivation of solar landscape vocabulary, is believed to
support policy and decision makers, SPP developers, designers and other
stakeholders to conceive solar landscapes supported by landscape users
[2,11,40].
The second section of this paper presents the methods and materials.
The framework for the case analysis is presented in section three. The
results and discussion section rst presents the solar infrastructure and
landscape feature properties, followed by visibility, multifunctionality
and temporality. The paper is concluded in section ve.
2. Methods and materials
2.1. Case-study approach
This study examines the spatial properties of built solar landscapes.
We adopted a case-study approach in our research, as this allows for the
description of a contemporary phenomenon in its spatial context [41].
We used a multiple embedded case design to document and compare a
high variety of spatial properties across all cases [41].
2.2. Case selection
Our research focuses on the Netherlands, the United Kingdom, Ger-
many and Italy, as these countries have shown increasing attention for
solar landscapes. In addition, the travel distance and language of these
countries allowed us to study the cases within the time and resource
provided. We aimed to study cases of SPP that were recognized for being
at the front of addressing societal concerns and providing functions
additional to electricity generation. We identied so-called ‘frontrunner
cases’ through recognition in the form of awards granted, for example by
solar industry, and expert judgement. We reached out to photovoltaic
and environmental design experts using personal contacts and
approaching photovoltaic developer and environmental design associ-
ations.
1
We asked the experts to provide us with the names of SPPs that
provided benets besides to electricity generation, such as ecological
restoration, recreation or aesthetics.
The expert contact and the desk study on SPP awards resulted in a
longlist of over 30 cases. A quick-scan was used to identify their main
spatial properties. Based on the quick-scan, we selected cases that
complied to two criteria that are key to solar landscapes. First, the case
needed to demonstrate a combined spatial arrangement of SPP and
landscape. Second, the case needed to include new landscape features in
addition to solar infrastructure, for example water retention areas, op-
portunities for recreation or habitat patches. For each case, these criteria
were evaluated using design maps or project documentation and
conrmed by satellite imagery or eld visits. We diversied according to
spatial properties, as well as landscape type and project scale; variety in
the latter two are expected to increase the variety of spatial properties
[13,29]. Ultimately, 11 cases were selected (Fig. 1 and Table 1).
2.3. Research process
For each case, we performed a spatial analysis [42,43] and studied
accompanying project documentation. The spatial and document anal-
ysis was subsequently veried by eld observations. To start, the spatial
analysis was conducted using a case-study protocol, to strengthen con-
sistency of the analysis by the multiple researchers involved [41]. This
protocol was tested and further rened by analyzing two contrasting
cases. The properties that were used to guide the spatial analysis are
presented in section 3. Results of the analysis were presented in maps,
text and tables. Data used for the mapping were design maps as well as
recent and historical satellite imagery.
Next, project documentation was used to conrm and specify the
spatial arrangement of SPP and landscape. The document analysis was
mainly based on project reports and websites that were collected until
June 2020. This data was occasionally complemented by insights from
1
Associations in Germany: German Solar Association (BSW) and German
Association of Landscape Architects (BDLA). The Netherlands: Holland Solar,
Netherlands association for garden- and landscape architecture (NVTL) and
Dutch association of urban designers and planners (BNSP). Italy: Italian Asso-
ciation of Landscape Architecture (AIAPP). United Kingdom: Solar Trade As-
sociation and Landscape Institute.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
3
case informants. Finally, intermediate results of spatial and document
analysis were enhanced and veried by eld observations that took
place from May until October 2019.
The results of the individual cases were synthesized to identify
similarities and differences across cases [41]. Maps, textual descriptions
and numerical date were aggregated using tables and examined along
the categories of the framework for case analysis (section 3). Aggre-
gating the data of all cases helped to prole the individual cases, specify
the framework for case analysis and subsequently enhance the cross-case
synthesis in an iterative manner.
3. Framework for case analysis
The framework for case analysis was developed deductively (draw-
ing from literature) and inductively (drawing from the cases) through
multiple iterations of application and reection. The framework was
used to analyze the spatial properties of the embedded cases. The larger
host landscape was analyzed as well, as this forms the backdrop for the
spatial properties. Solar infrastructure and landscape features refer to
physical changes in the landscape that can, to some extent, be examined
independently [e.g. 38,44]. Contrastingly, visibility, multifunctionality
and temporality are emergent properties: properties of the whole revealed
by interactions between individual characteristics [45,46]. These
properties of solar landscapes were analyzed by jointly examining solar
infrastructure and landscape features [2,36] (Fig. 2). This section rst
introduces the solar infrastructure and landscape feature properties
(3.1), followed by the procedure for the study of emergent properties
visibility, multifunctionality and temporality (3.2).
3.1. Solar infrastructure and landscape features
The spatial analysis started by identifying landscape type and pre-
vious land use function. These properties of the host landscape informed
the subsequent analysis of solar infrastructure and landscape features.
Solar infrastructure of SPP is discussed extensively in the literature
[e.g. 2,16,29]. We created an overview of properties found in literature
and specied these with the ndings of the case analysis. For solar
infrastructure, the spatial properties are grouped in three nested levels:
the system as a whole, the patch as distinct group of arrays, and the array
as specic object (Table 2).
Literature reports on both potential and realized landscape features
of SPP [2,20,34,36]. We used the main categories identied in the
literature to group the individual features found in the cases (Table 2),
namely ecological, recreational and educational, agricultural and water
retention features.
3.2. Emergent properties of solar landscape
3.2.1. Visibility
The combined spatial arrangement of SPP and landscape affects the
visibility of the solar infrastructure [16,29,39,50]. To investigate this
relationship, rst the existing and new landscape features at the edge of
the solar landscapes were analyzed. The edge is dened as the space
between solar infrastructure and the project boundary (Fig. 3a). Second,
the part of the solar infrastructure visible to on-road observers was
analyzed [28] and subsequently expressed in the degree of visibility. The
degree of visibility is the part of the outer edge of the solar infrastructure
visible to observers, as seen from the rst line of observation (Fig. 3a).
The rst line of observation is the set of roads or paths closest to the edge
Fig. 1. The 11 selected cases. Scale of the images varies, see Table 1 for actual size of the cases (source satellite imagery: Google Earth and Kadaster).
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
4
of the case. We distinguish between visible, partly visible and invisible,
based on visibility levels as presented in Ref. [13]. Visibility from a
larger distance and for on-site observers [28,29] was examined during
the eld observations but not included here to allow a comprehensive
comparison of the 11 embedded cases.
3.2.2. Multifunctionality
Solar landscapes provide multiple services and functions [2,3,51].
We use the term function as it indicates a capacity to deliver a certain
service. In this research, we aimed to identify deliberately added func-
tions with a certain expected service. The quantication and assessment
of services is beyond the scope of this comparative analysis of 11
frontrunner cases. The Common International Classication of
Ecosystem Services (CICES) was used to systematically identify and
describe functions [52]. For each case, a list of deliberately added
functions was identied in project documentation and subsequently
veried during eld observations. These lists were discussed and
adjusted during multiple workshops among involved researchers to
ensure cross-case consistency. Using CICES, we analyzed the presence
and number of functions identied in the cases. Three types of multi-
functionality were identied: array multifunctionality (beneath arrays),
patch multifunctionality (on patch area and not underneath arrays) and
adjacent multifunctionality (next to patches) (Fig. 3b).
3.2.3. Temporality
Landscapes change through time, largely driven by societal demands
and expressing changing societal values [53]. The demand for renew-
able energy results in the introduction of energy technologies that
transform landscapes within a relatively short period of time; this is why
the development of SPP is considered dynamic. The life-span of SPPs is
relatively short (20–30 years) compared to other, more permanent en-
ergy technologies, such as nuclear power plants [4]. Others have studied
the construction and operation/maintenance stages of SPP [54,55]. This
study adopted a wider temporal perspective and focused on the former
state of the host landscape (i.e. before construction), the case during
operation and maintenance stage and the decommissioning stage
(Fig. 4). Project documentation was used to identify if and how tem-
porality was considered in these three stages: (1) inclusion of existing
features of the host landscape in the case, (2) active management of
landscape features during operation and maintenance stage and (3)
plans for the decommissioning stage.
4. Results & discussion
The results are presented and discussed in three parts. The rst two
parts present the solar infrastructure (4.1) and landscape feature (4.2)
properties. The third part (4.3) takes the perspective of the solar land-
scape as a whole and discusses the visibility, multifunctionality and
temporality of the examined cases.
4.1. Solar infrastructure
4.1.1. System layout and host landscape pattern
We found that the way the system layout responded to the host
landscape differed between cases with a former agricultural use and
those with a browneld use. In the nine cases with a former agricultural
use, the system size and plot size were key factors in the way the system
layout responded to the pattern of the host landscape. In only one case
the system was entirely located within a single plot (Monreale). In the
other eight cases, solar infrastructure was distributed over multiple
plots, whether the cases were small (e.g. Sinnegreide, 12 ha) or large (e.
g. G¨
ansdorf, 181 ha). In these multiple-plot cases, the plots of the host
landscape either remained (almost) completely intact (ve cases) or
were aggregated into a single larger plot (three cases). For some cases,
although parcellation remained intact, the individual plots are poten-
tially not always recognized as such by observers. Recognition of
Table 1
General information on the 11 cases.
GENERAL SOLAR INFRASTRUCTURE HOST LANDSCAPE
Cases Latitude Year of
construction
Country Power
(MWp)
Size
(ha)
Energy density
(MWp/ha)
Land Area Occupation
Ratio (LAOR)
Technology Landscape type Previous land use
1. G¨
ansdorf 48′48′12 2009 Germany 54,0 180,9 0,30 22% Fixed tilt Open agricultural Agriculture: highly productive arable
land
2. Kwekerij 52′03′24 2016 Netherlands 2,0 7,1 0,28 16% Fixed tilt Semi-open bocage landscape Agriculture: low grade, tree nursery
3. Valentano 42′35′19 2011 Italy 6,0 17,6 0,34 23% Fixed tilt Open agricultural Agriculture: highly productive arable
land
4. Southill 51′51′31 2016 United
Kingdom
4,5 18,1 0,25 16% Fixed tilt Semi-enclosed valley side farmland Agriculture: extensive, low grade
5. Hemau 49′02′10 2002 Germany 4,0 18,0 0,22 20% Fixed tilt Enclosed, agricultural landscape
with large evergreen forests
Browneld: military ammunition
depot within production forest
6. Laarberg 52′06′43 2018 Netherlands 2,2 6,4 0,35 21% Fixed tilt Semi-open bocage landscape Agriculture: intensive grassland and
corn production
7. Sinnegreide 53′26′04 2018 Netherlands 11,8 12,0 0,98 53% Fixed tilt Open agricultural Agriculture: grassland
8. Mühlenfeld 51′27′51 2013 Germany 3,5 24,4 0,14 10% Fixed tilt Semi-open bocage landscape Browneld: gravel mining and
nature development
9. Midden-
Groningen
53′10′48 2019 Netherlands 103,0 121,2 0,85 61% Fixed tilt Open peat landscape Agriculture: arable and grassland
10. Monreale 37′52′07 2010 Italy 5,0 28,0 0,18 13% Single-axis
tracker
Undulated open agricultural
landscape
Agriculture: extensive, wheat and
olive groves
11. Southwick 50′52′50 2015 United
Kingdom
48,0 83,4 0,58 35% Fixed tilt Enclosed, mixed farmland/
woodland
Agriculture: arable and grassland
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
5
individual plots can occur if there is high vegetation along the plot
border and/or a eld margin; a zone between plot border and PV patch.
In the two cases with a former browneld use, the system layout was
adjusted to site specic elements of the previous land use function, for
example a gravel mining pit in Mühlenfeld.
4.1.2. Patch conguration and density
The system layout consists of multiple PV patches that are each
congured within a specic plot. Five different types of patch congu-
rations were found (Fig. 5). Most cases consisted of a single congura-
tion. In the responsive conguration, the size of the PV patch
Fig. 2. Solar infrastructure and landscape feature properties refer to physical changes that can be examined independently. Visibility, multifunctionality and
temporality are emergent properties of the solar landscape as a whole.
Table 2
Framework for the analysis of the host landscape, solar infrastructure and landscape features.
Category Sub-category Property Description Literature
Host landscape Landscape type Open/enclosed, parcellation/plot sizes, existing landscape infrastructure/features,
urban settlements.
[2,36]
Previous land use Previous land use(s) at the site [19,47]
Solar infrastructure System Layout The number, size and position of the patches as part of the solar system. [11,29]
Response to
parcellation
The response of the system layout to the original parcellation. [29]
Patch Conguration Size, position and alignment of the of patch within parcellation. [2,11,16,29,36]
Density Density of the array within a patch. Indicator is the ground-coverage-ratio (GCR),
which is the array length (L) divided by the row-to-row pitch (R)
[2,11,48]
Array Orientation Orientation or azimuth of the arrays. Traditional orientation (east-west or north-south)
results in a stripes pattern, but other types of patterns are possible if the azimuth is
varied.
[2,29]
Dimensions Dimension of array, determined by: tilt of modules, total height of the array from the
ground; length (l) of array; width of array; layout of array (orientation of modules and
number of rows);
[36]
Concurrence Presence of multiple PV technologies or types of modules in a single case [14]
Materials Color of modules, materials used in supporting structure. [16,29,49]
Landscape features Ecological Feature Features that support ecological functions, for example patches of wildowers or
hedgerows.
[24,30,33,34]
Recreational and
educational
Feature Features that support recreational and education functions, such as community
gathering spaces and outdoor classrooms.
[2,24,36]
Agricultural Feature Features that support agricultural functions, such as grazing or orchards. [2,24,33]
Water management Feature Features that support hydrological functions, such as water retention areas. [30,33,34]
Fig. 3. a) Visibility of the solar infrastructure is expressed by the ratio of the outer edge of the solar infrastructure that is visible, partly visible or invisible (based on
visibility levels as presented in Ref. [13]). b) Multifunctionality beneath the arrays (array multifunctionality), on the patch area (patch multifunctionality) and next to
patches (adjacent multifunctionality).
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
6
predominantly matched the plot size. For example, in Sinnegreide, PV
arrays with various widths were used to cover the entire plot; the
original parcellation remained intact. Contrastingly, in the irresponsive
conguration, the patch shape was mainly self-referential, which,
dependent on the plot shape, can result in left-over spaces. In G¨
ansdorf,
constructed in 2009, the limited exibility in array width of that time
can have contributed to this conguration.
In the split conguration, the patch responded to the shape of the plot,
yet only partially covered the plot area (25–50%). This partial coverage
resulted in a perceived split of the original plot. In Southill for example,
only the south-western part of the plot was used for arrays, the
remainder of the plot consisted of landscape features. The xed size of
the single-axis tracker system employed in Monreale seems to have
resulted in roughly equal patch sizes.
In the islands conguration, a single patch was divided into sub-
patches and the patch shape was almost entirely self-referential. The
plot was only for a small part (30–45%) covered by arrays. In the
Kwekerij and Laarberg for example, this conguration resulted in mul-
tiple small PV patches dispersed across the plot. These congurations
corroborate the proposals for patch variations by Scognamiglio [2]. The
irresponsive, split and island conguration increase the spatial hetero-
geneity of the landscape, dependent on the previous land-use. In a host
landscape with monofunctional agricultural plots, these congurations
increase the variety of functions within a single plot, countering agri-
cultural upscaling often seen in the countryside [56]. However, some of
these congurations are less aligned with landscape parcellation and
recent research has shown this can negatively inuence perception [11].
For browneld cases, the patch conguration coincided with the
strategy for the system layout: site specic elements determined the
conguration. In the case of Hemau for example, the patch was shaped
around the existing (elevated) bunkers and identied hotspots for
biodiversity were also excluded for electricity generation. This fth,
Fig. 4. Temporal properties: former state of the host landscape, case during operation/maintenance stage and decommissioning stage.
Fig. 5. Five types of patch congurations. Main determinants for the patch congurations are alignment to plot and coverage of the plot by the PV patch. Case names
between brackets indicate a certain conguration was identied, but it was secondary to another, primary conguration.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
7
incidental conguration, was also found in other cases in addition to
another, primary conguration. In these cases, elements were high-
lighted that otherwise remained invisible, for example underground
infrastructure became visible as a blank space between arrays.
The patch density is determined by the width of the array and the row-
to-row-pitch, expressed by the ground-coverage-ratio [48] and ranged
for the cases between 0,35 and 0,84 (Table 3). Existing research points to
two consequences of patch density: visual impact and impact on land use
[2,11].
Potential visibility affected patch density in three cases: the Kwe-
kerij, Sinnegreide and Southwick. In these cases a lower, secondary
ground-coverage-ratio (GCR) was found where a high visibility of the
solar system was expected. For Southwick though, we could not conrm
a causal relationship. Although Scognamiglio [2] stresses that a low
patch density is pivotal to increase multifunctionality, no relationship
between multifunctionality and patch density was found (see also 4.3.2).
Cases with multifunctionality beneath arrays or on the patch area had a
GCR ranging from 0,35 to 0,73, covering almost the entire spectrum of
GCR found in the cases.
4.1.3. Array orientation, dimensions and materials
On the level of arrays, we found that in all cases the orientation of the
PV arrays was optimized for maximum solar energy generation. This
optimization was for 10 cases east-west oriented arrays facing south, and
for one case with a single-axis tracker north-south oriented arrays
(Table 4, see also Table 1). In other words, the type of pattern was the
same for all the cases: parallel stripes [2]. To relinquish energy opti-
mization and vary the azimuth is considered a key feature of solar
landscapes. A variable azimuth can improve ecological performance or
allow the solar infrastructure to align with the landscape pattern [2,11,
16] that, in turn, can result in new patch congurations. In addition,
non-optimal azimuth angles reduce peak loads on the electricity grid
and allow for a more exible integration into the landscape [1,57]. If
business models can incorporate these benets, non-optimal azimuth
angles can also result in an improved alignment of array and landscape
pattern.
2
The dimension of the arrays is specic to each case, although the
height was variable in three cases (Table 4). In two cases (the Kwekerij
and Sinnegreide), arrays were found with two different heights. Arrays
with a lower height were closest to where most observers were expected
(see also table GCR). In Monreale the difference in height of the arrays
was caused by partial ground levelling.
The color of the arrays in the cases was the blue commonly seen in
SPP. However, the rapid development of colored modules in the built
environment may also permeate to solar landscapes [58]. Only in
Hemau, modules were three different shades of blue, as at the time of
construction (2002) suppliers were not able to deliver the requested
amount of modules from a single type of module. Consequently, Hemau
is also the only case where concurrence was identied. The same applies
for the type of supporting structure used: all cases except Hemau used
metal structures, while Hemau used a wooden structure.
4.1.4. Reections on solar infrastructure across frontrunner SPPs
Southwick illustrates that combining solar infrastructure with land-
scape occurs at multiple scales: on the system level, the size of the
existing plots determined the system layout; on the patch level, indi-
vidual patches matched the shape of the plots. Even more, the existing
parcellation remained visually recognizable as existing hedgerows were
combined with a sufcient eld margin around the PV patch. This
spatial arrangement required additional space, resulting in the trade-off
of a decreased maximum amount of arrays (lower LAOR value).
4.2. Landscape features
4.2.1. Ecological features
Several ecological features were found in the cases: patches of dry or
wet vegetation, vegetative buffers, built structures for roosting, nesting
and hibernating, wildlife permeable fencing and some cases incorpo-
rated existing vegetation into the system layout (Table 5).
Vegetative patches were identied in all cases, for example wild-
ower elds or shrubs. Vegetative buffers were found in seven cases,
often combined with screening function at the edge of the case (see also
4.3.1). Buffers were for example hedgerows, tree rows or reed zones. In
one case, Hemau, an existing monoculture forest patch was removed to
avoid shadow on the arrays. The presence of vegetative patches and
buffers in the cases reects the growing evidence that SPP contribute to
local biodiversity of [24,34,35,59]. Several similarities in ecological
features were found, independent of landscape type: hedgerows, or-
chards and ower elds were found in many cases. Landscape features
dependent on landscape type and other contextual characteristics
become especially important when SPP become a more familiar phe-
nomenon in the landscape [60]. In ve cases, built structures for
roosting, nesting and hibernating, such as beehives or insect hotels, were
identied. In nine cases wildlife permeable fencing was realized by
either lifting the fence or by the addition of small mammal gates. These
ndings show that in most cases landscape fragmentation is addressed
[20,36]. In ve cases existing vegetation was retained, such as hedge-
rows or solitary trees, while it is not uncommon that existing vegetation
is removed [17,20]. In retaining vegetation, these cases address the loss
of identity elements, or fragmentation of the countryside [17].
4.2.2. Recreational and educational features
Recreational and/or educational features were identied in 9 of the
11 cases, conrming the potential suggested in earlier research [2,24].
All recreational and educational features were located next to a PV
patch, and not beneath or between the arrays as has been identied in
the Solar Strand, USA [2]. Recreational and educational features were
for example lookouts, benches and information panels. The Kwekerij,
Laarberg and Mühlenfeld seemed to actively enable recreation by add-
ing multiple recreational facilities and connecting the case to a local
recreational network. The other cases seemed to be addressing occa-
sional or accidental on-site observers. Recreational features were absent
in Midden-Groningen, Monreale and Southill.
In the large-scale cases G¨
ansdorf, Midden-Groningen and Southwick,
Table 3
Patch density of the cases expressed by the ground-coverage-ratio (GCR). In
three cases, two different array types were found, resulting in two values for the
GCR. The GCR is calculated by dividing the array length (L) by the row-to-row
pitch (R) [48].
Cases GCR of primary
array type (L/R)
GCR or secondary
array type (L/R)
Location of secondary
array type
1. G¨
ansdorf 0,45 n/a
2. Kwekerij 0,44 0,41 Most visible patches for
nearby inhabitants.
3. Valentano 0,49 n/a
4. Southill 0,63 n/a
5. Hemau 0,35 n/a
6. Laarberg 0,52 n/a
7. Sinnegreide 0,84 0,69 Most visible patch near
road.
8. Mühlenfeld 0,44 n/a
9. Midden-
Groningen
0,73 n/a
10. Monreale 0,40 n/a
11. Southwick 0,63 0,57 Most visible patch in
west compartment.
2
A recent example in the Netherlands is the project ‘Energy garden’ Assen-
Zuid: https://www.nmfdrenthe.nl/wij-werken-aan/energieneutraal-drenth
e/energietuin-assen-zuid/(in Dutch).
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
8
the space between the patches was occasionally publicly accessible
(Table 6). In G¨
ansdorf and Midden-Groningen, this access was the
consequence of practical considerations (land ownership and mainte-
nance respectively), while in Southwick the patch shape was deliber-
ately adjusted to maintain an existing path. Across cases, no roads or
paths formerly accessible were removed or cut-off. Moreover, in the
Kwekerij a path network was created between the patches and access
within the fence is possible on a daily basis. This study shows that solar
landscapes are able to maintain or increase landscape connectivity [61].
4.2.3. Agricultural features
Nine cases included agricultural features, ranging from small fruit
tree orchards to substantial olive groves. This high presence of agricul-
tural features may point to addressing the loss of agricultural land [17].
In Monreale, a large olive grove was located next to the solar system, and
the left-over spaces within the solar system were planted with olive and
almond trees. In G¨
ansdorf, the case comprised a part of the former arable
land. In three cases (Laarberg, Hemau and Midden-Groningen) sheep
were kept inside for grazing. In ve cases (G¨
ansdorf, the Kwekerij,
Southill, Laarberg and Sinnegreide) small-scale agriculture targeting the
local community (fruit orchards, vegetable gardens) was identied.
4.2.4. Water management features
Local water management was found in ve of the cases. This study
identied water retention areas, in addition to techniques of water
recuperation [36]. Water retention areas were part of two cases (Laar-
berg and the Kwekerij). In Laarberg, water run-off from a (future)
business area can be stored beneath PV arrays, and the solar infra-
structure was adjusted to allow for temporary ooding (above-ground
cables). In Monreale, rain water recuperated from the PV patches, was
stored in a basin to be used for the adjacent olive grove. In two other
cases (Sinnegreide and Valentano) waterways were enhanced or
recovered.
4.2.5. Reections on landscape features across frontrunner SPPs
Laarberg includes multiple categories of landscape features:
ecological, recreational, agricultural and water retention features have
been combined with electricity generation on only 6,4 ha. Southill, on
the contrary, displays focus on a single category: ecological restauration
is central and to this end human access is limited.
Furthermore, in Southill and Hemau spaces suitable for electricity
generation have been deliberately kept free to achieve ecological ob-
jectives. In other words, spatial arrangement of solar infrastructure is
adjusted and even sub-optimal to accommodate other objectives. The
Kwekerij and G¨
ansdorf are examples of synergy between functions:
recreational and ecological values are increased by locating strips of
wildowers next to roads and pathways.
4.3. Solar landscape
Three emergent properties that arise from the combined spatial
arrangement of SPP and landscape - in this paper and elsewhere
conceptualized as solar landscape - are presented in this section: visi-
bility, multifunctionality and temporality.
Table 4
Array orientation, height, materials and concurrence.
Array orientation Array height Array materials Concurrence
Cases Adjustment to
plot
Optimum for solar energy
generation
Consistent/
variable
Color modules Supporting
structures
1. G¨
ansdorf x consistent Blue Metal no
2. Kwekerij x variable Blue Metal no
3. Valentano x consistent Blue Metal no
4. Southill x consistent Blue Metal no
5. Hemau x consistent Blue (three
shades)
Wood Yes, three types of modules and
array types
6. Laarberg x consistent Blue Metal no
7. Sinnegreide x variable Blue Metal no
8. Mühlenfeld x consistent Blue Metal no
9. Midden-
Groningen
x consistent Blue Metal no
10. Monreale x variable Blue Metal no
11. Southwick x consistent Blue Metal no
Table 5
Ecological features found in the cases (x =new; [x] =enhanced, not completely new; (−) =removal).
Ecological features
Cases Total Patch of dry
vegetation
Patch of wet
vegetation
Vegetative
buffer
Built structures for roosting, nesting
and hibernating
Wildlife permeable
fencing
Retaining existing
vegetation
1. G¨
ansdorf 3 x x x
2. Kwekerij 5 x x x x x yes
3. Valentano 3 x x x
4. Southill 4/
[1]
x [x] x x yes
5. Hemau 3/(1) x/(−) x x
6. Laarberg 5/
[1]
x x [x] x x yes
7. Sinnegreide 3 x x x
8. Mühlenfeld 3 x x x yes
9. Midden-
Groningen
3 x x x
10. Monreale 3 x x x
11. Southwick 5/
[1]
x x/[x] x x yes
Total 10/(1) 5 7/[3] 5 10 5
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
9
4.3.1. Visibility
This section presents the visibility of the solar infrastructure based on
an analysis of the existing and new landscape features at the edge of the
solar landscape.
4.3.1.1. The edge. The edge of the cases consisted of existing eye-level
vegetation (appendix A) and new edge measures (appendix B), and in
each case the solar infrastructure was completely surrounded by a fence.
Existing eye-level vegetation, such as forest patches or hedgerows, were
found in eight cases, with ve cases consisting of over 60% existing eye-
level vegetation along the edge. New edge measures consisted of land-
scape features, for example hedgerows or a reed zone. Three types of
measures were applied in the cases: removal of existing landscape fea-
tures, enhancing existing landscape features and new landscape features
(appendix B).
4.3.1.2. Reducing visibility of solar infrastructure. In all cases, the visi-
bility was deliberately reduced, either through siting within existing
vegetation or through new edge measures with screening function [16,
39]. The highest ratio of a visible edge in a single case was 30%
(Mühlenfeld). Contrastingly, in three cases clear views on the solar
system were almost absent (Fig. 6).
Southwick, Laarberg, Mühlenfeld, Hemau and Southill combined
low visibility with few new landscape features and many existing
landscape features. Existing vegetation, sometimes enhanced, was used
for screening purposes. This combination supports the notion that
careful site selection is an important aspect to achieve low visibility
without the need for many new edge measures [29,39]. Patch congu-
ration also inuenced visibility. In Southill for example, positioning the
patch into the lower lying part of the plot reduced the visibility from
higher located roads [16,28,39].
Monreale, Midden-Groningen, Sinnegreide, Valentano and G¨
ansdorf
combined a low amount of existing eye-level vegetation along the edge
(<11%) with a high degree of new screening features (85–100%).
Introduction of eye-level vegetation that is not typically found in open
landscapes may have an adverse effect on the landscape character [11,
16,39]. In some cases, screening measures provided other functions as
well. For example, in G¨
ansdorf an orchard was planted to reduce visi-
bility from the road and at the same time produce fruit.
4.3.1.3. Enhancing visibility of the solar infrastructure. The overall
reduction of visibility was contrasted by measures that deliberately
enhanced visibility. In G¨
ansdorf, the Kwekerij, Mühlenfeld, Sinnegreide
and Laarberg, features were added that provided visitors with a clear
view of the solar system (Fig. 7). In the rst three cases, the solar
infrastructure can be seen from a lookout, while the latter two cases
feature an area at the edge of the case that provided amenities for vis-
itors to stay for a short period of time. These ve cases showed a com-
bination of two strategies with respect to visibility: in general, visibility
is reduced, but at a specic point visibility of the solar infrastructure is
enhanced. The latter strategy seems to reect ‘embracing visibility of
energy facilities’, which can be part of a place branding approach [62].
This research shows that the cases addressed visibility [5,11,13,14,17,
18], and at the same time aimed to reframe visibility from a mainly
negative impact into a potential positive impact.
4.3.2. Multifunctionality
Solar landscapes that provide functions additional to electricity
generation can be considered multifunctional. In this section, we further
detail the multifunctionality of the cases by examining the presence and
number of functions, as well as three types of multifunctionality. The
section is concluded with reections on the assessment of
multifunctionality.
4.3.2.1. Presence and number of functions. The studied cases provide a
Table 6
Recreational and educational features and accessibility in the cases.
Recreational and educational features Accessibility
Cases Total Lookout Information
panel
Benches Picnic
tables
Community
gathering site
Playground
features
Car
parking
Bicycle
parking
Charging point
electric
bicycles
Walking
path
Node in local
recreational
network
No
access
Access
between
patches
Access
within
fence
1. G¨
ansdorf 1 x x
2. Kwekerij 9 x x x x x x x x x x
3. Valentano 1 x x
4. Southill 1 x x
5. Hemau 1 x x
6. Laarberg 4 x x x x x
7.
Sinnegreide
2 x x x
8. Mühlenfeld 5 x x x x x x
9. Midden-
Groningen
0 x
10. Monreale 0 x
11.
Southwick
2 x x x
Total 3 5 2 3 1 1 1 2 1 3 4 7 3 1
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
10
multitude of different functions. Of the 65 functions in the CICES model
of ecosystem services, 18 were found in the cases (appendix C). The
function Providing habitats for wild plants and animals (2.2.2.3), was the
only function identied in all cases, besides Solar power (4.3.2.4). Two
other functions were identied in nine out of eleven cases: Pollinating our
fruit trees and other plants (2.2.2.1) and Screening unsightly things
(2.1.2.3). Small-scale agricultural functions, such as grazing sheep
(1.1.1.3), food production (1.1.1.1) were found in nine cases. These
functions conrm that the cases aim to mitigate impacts of SPP identi-
ed in earlier research, such as habitat destruction and fragmentation,
decrease of wildlife and biodiversity [19,20,63] and land use impact or
loss of productive land [18,19,47,64].
From the 18 identied functions, four were provisioning, seven
regulating and seven were cultural functions. All three types of functions
were found in all cases. The total number of functions ranged from 6 to
14 (Fig. 8).
No clear relationship was found between the number of functions
and the land area occupation ratio (LAOR, see Table 1) [2]. Cases with a
high LAOR (highest ratio found was 61%) still supported multiple
functions, although these cases represented the lower end of the range of
functions (Fig. 9).
4.3.2.2. Three types: array, patch and adjacent multifunctionality. Func-
tions were located beneath arrays (array multifunctionality), on the
patch area (patch multifunctionality) and adjacent to patches (adjacent
multifunctionality) (Fig. 10). Array and patch multifunctionality allow
for interactivity between functions (e.g. sheep nding shade under ar-
rays) and were identied in 8 out of 11 cases (Fig. 11). Adjacent mul-
tifunctionality was identied in all cases and was often a form of
multiple land use or co-location with little interaction with the solar
infrastructure [31,32]. These ndings are in line with, and further
specify earlier research on solar landscapes; earlier research identied
multifunctionality applied to solar infrastructure and as multiple land
use within the project boundary [36].
Fig. 6. Visibility of the solar infrastructure as observed from road infrastructure closest to the case, the rst line of observation.
Fig. 7. Measures enhancing visibility: lookout in G¨
ansdorf (a), Mühlenfeld (b, picture by Florian Becker) and the Kwekerij (c), and benches near a clear view to the
solar infrastructure in Laarberg (d, picture by Coos van Ginkel) and Sinnegreide (e). All pictures by authors unless otherwise indicated.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
11
On average, the cases contained 28,9% adjacent multifunctionality,
19,8% patch multifunctionality and 11,6% array multifunctionality,
totaling to 60,4% (Fig. 11). In seven cases, over 70% of the land surface
was allotted to multifunctionality. In three of these cases (Valentano,
Hemau and Midden-Groningen), this high share is for a large part caused
by harvesting the meadow beneath and between the arrays by or for
livestock. Specically for Midden-Groningen, multifunctionality is ar-
ranged as a sharp spatial distinction between the high-density PV
patches and livestock (array and patch multifunctionality), and
ecological features in the edge (adjacent multifunctionality). The large
share of multifunctionality in the other four cases is explained by a
diverse set of features: wildower elds, recreational amenities and
water retention (the Kwekerij), livestock grazing and water retention
(Laarberg), elds of wildowers and ne grasses (Southill) and an olive
grove and wet ecological corridor (Monreale). The cases with high
shares of array and patch multifunctionality indicate the potential to
increase multifunctionality without adversely affecting land used for
electricity generation. High shares of adjacent multifunctionality were
found in Monreale, Southill, the Kwekerij, Valentano, Laarberg and
Hemau. With adjacent multifunctionality, however, land otherwise
available for electricity generation is used for other functions. This latter
type of multifunctionality therefore reduces the overall land use energy
intensity of the solar landscape [2].
4.3.2.3. Assessment of multifunctionality. The number of functions and
the land surface allocated to multifunctionality are useful indicators to
compare SPP on multifunctionality, yet they do not assess functions.
Assessment of ecosystem functions and services needs to provide insight
in their effectiveness, management [35] and comparison to the baseline
situation [47]. Such an assessment requires integrated approaches that
make use of a mix of methods and tools on multiple scales of analysis
[3]. Without advancing such assessment methods for solar landscapes,
cases may emerge that bear the promise of multifunctionality, but only
deliver minor provisioning, regulating or cultural benets. Current as-
sessments of SPP often make use of performance indicators based on
installed capacity or electricity generation [64,65]. Using these in-
dicators, most of the cases in this study will be outperformed by SPP that
are optimized for electricity generation. These assessments and their
associated indicators will need to be supplemented by other indicators
that capture multifunctionality.
4.3.3. Temporality
Temporality in the cases was addressed in 8 out of 11 cases by
attention for landscape elements and patterns present in the host land-
scape, active management during operation and maintenance stage and
landscape plans for the decommissioning stage.
In ve cases, landscape elements and patterns that were part of the
former state of the host landscape were included in operation and main-
tenance stage, with the potential to extend into decommissioning stage.
These efforts can result in ‘remnants of the past’ and carry symbolic and
historical value [53]. Elements were often vegetation, such as hedge-
rows or trees, but also former military bunkers were preserved (Hemau).
Fig. 8. The number of functions in each case, divided over provisioning, regulating and cultural functions.
Fig. 9. The number of functions compared to the land use energy intensity,
expressed by Land Area Occupation Ratio. 1=G¨
ansdorf; 2=Kwekerij;
3=Valentano; 4=Southill; 5=Hemau; 6=Laarberg; 7=Sinnegreide;
8=Mühlenfeld; 9=Midden-Groningen; 10=Monreale; 11=Southwick.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
12
In two cases (G¨
ansdorf and Midden-Groningen) existing parcellation
was explicitly considered to maintain landscape character during the
operation and maintenance stage.
Active management of landscape features during operation and
maintenance stage was identied in four cases (G¨
ansdorf, the Kwekerij,
Southill and Hemau). In these cases, monitoring and evaluation was
organized, and consequently enabled decision-making based on chang-
ing monitoring results and contextual circumstances. A distinctive
example is the Kwekerij, where changing demands by local stakeholders
resulted in the addition of a vegetable garden in a later stage. On the
contrary, other cases indicate a lack of active management and appeared
not to be resilient to changing circumstances. In Monreale for example,
the olive grove adjacent to the solar system is currently in a poor state.
This olive grove was supposed to be used for local olive oil production,
but it seems it was not well embedded in the local socio-economic
context. In Southwick, original plans involved wildower elds, graz-
ing sheep and bat boxes. These plans, partially executed, appear to have
been abandoned following a change in the ownership of the SPP.
Plans for the decommissioning stage were mentioned in six cases,
mostly involving reversibility [4]. Three cases (Southill,
Midden-Groningen and Southwick) plan to reverse the site into the
former state of the landscape, although it is not always clear if this
concerns removal of both solar infrastructure and landscape features. In
G¨
ansdorf, rather than decommissioning, the plan is to continue
combining electricity generation with habitat creation and agriculture
by means of agrivoltaïcs. If executed, this plan will result in the recycling
of the existing energy landscape [4]. In the Kwekerij and Monreale,
landscape features in operation and maintenance stage supported the
Fig. 10. a) array multifunctionality in Mühlenfeld (picture by Florian Becker): shade tolerating vegetation and inverters beneath arrays; b) patch multifunctionality
in Laarberg: sheep grazing on the lowered patch area that also functions as water retention area; c) adjacent multifunctionality in G¨
ansdorf: hedgerow and wildower
eld developed next to the PV patch.
Fig. 11. Shares of land surface allotted to array, patch and adjacent multifunctionality.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
13
plans for the decommissioning stage. In the Kwekerij, local inhabitants
benet from the park function in operation and maintenance stage, and
a larger park will be available to them once the solar infrastructure will
be dismantled. In Monreale, cultivation of herbs between the arrays is
supposed to increase soil quality for agricultural use in the decom-
missioning stage. Concluding, while in eight cases the temporal char-
acter of the cases was considered, only two cases used landscape features
to enhance future use of the sites, beyond site restoration [59]. Thus, in
most cases use of landscape features in decommissioning stage is not
explicitly considered, which in turn might adversely affect their
continuation [24]. This unclarity of the decommissioning stage has
already been identied for wind energy and can potentially result in
repowering or abandonment of renewable energy technologies [23].
4.3.4. Reections on emergent properties across frontrunner solar
landscapes
Although most cases pay attention to visibility, multifunctionality
and temporality, the spatial arrangement of each case illustrates varying
degrees of integration between solar infrastructure and landscape fea-
tures. In the Kwekerij, these are entwined to a degree that the case is
neither just a solar power plant nor just a public park: it is a combination
of both. Patches have been congured to allow visitors to walk between
the arrays, height of the arrays has been adjusted to address visibility
concerns of neighboring residents. In G¨
ansdorf however, solar infra-
structure and landscape features are strictly separated: additional
functions are not found within, but next to the PV patches.
Whether landscape features are sustained beyond the decom-
missioning of the solar infrastructure depends on the type of the fea-
tures. Features enhancing landscape character (e.g. Southwick) or
features able to provide a function independent of solar infrastructure in
the future (e.g. the Kwekerij) are likely to be sustained. In Midden-
Groningen, on the contrary, some of the landscape features are unfa-
miliar to the host landscape and their existence will be less certain when
the SPP is decommissioned.
5. Conclusions
This study aimed to contribute to the understanding of solar land-
scapes by examining 11 frontrunner cases across Europe, guided by the
following research question: what are the visual, functional and temporal
properties of frontrunner solar landscapes in Europe?
The examined frontrunner solar landscapes use a combined spatial
arrangement of solar infrastructure and landscape features to address
societal concerns. Solar infrastructure operates on system, patch and
array level and landscape features are categorized as ecological, recre-
ational and educational, agricultural and water management features.
Visibility is reduced in all cases; yet in ve cases visibility is simulta-
neously enhanced in dedicated areas in combination with recreational
amenities. Cases contain between 6 and 14 different functions, although
the share of land allocated to multifunctionality differs greatly between
cases. In addition to electricity generation, habitat creation is identied
in all cases, and in 9 out of 11 cases pollinating, screening and small-
scale agricultural functions are identied. In eight cases the temporal
character is considered in some way, yet only two cases explicitly
introduce landscape features to enhance future use of the sites. Across
the cases, our analysis of spatial properties shows how contemporary
concerns about SPP, such as visual impact, land use competition and the
end-of-life stage are addressed. Next to these empirical ndings, we
draw three main conclusions from this case study.
First, although the cases altogether present a portfolio of measures
responding to societal concerns, the full potential of the three key
properties is yet to be explored. The orientation of PV arrays, for
example, is optimized for maximum electricity generation in all cases.
Alternative array orientation may support maintaining existing land-
scape patterns and, simultaneously, reducing peak load on the electricity
grid. Another example is the presence of similar landscape features
across cases, despite the differences in character of the host landscapes.
Second, despite the additional benets found in the cases, some
(local) trade-offs may still emerge. To illustrate, some of the identied
congurations of PV patches provide space for provisioning, regulating
or cultural functions but can, at the same time, destroy existing land-
scape patterns. Furthermore, a high share of land solely dedicated to
ecosystem functions increases the total land area needed to generate a
set amount of electricity. These examples show the need to assess both
individual properties as well as the SPP as a whole. Where existing
research on additional benets for SPPs is mainly theoretical, the
empirical evidence in this research resulted in properties and initial
indicators to describe, compare and potentially assess additional bene-
ts. Such properties and indicators can become part of environmental
impact assessments, multi-criteria decision analysis and other methods
to asses not only negative, but also positive impacts of SPP. For example,
assessment of enhancing visibility (e.g. through dedicated recreational
areas with clear views on solar infrastructure) may enrich impact as-
sessments that consider visibility as a negative property exclusively. Yet,
other properties related to visual impact, such as frequency of views and
glare still need to be taken into account. Similar, including properties
such as temporality in multi-criteria decision analysis may favor alter-
native proposals of SPP that allow continuation of existing landscape
features.
Third, as individual cases diverge in their attention for certain
properties, further distinctions within the concept ‘solar landscape’ can
be made. To illustrate, some cases focus mainly on visibility and only
marginally on multifunctionality. In addition, some cases focus on
provisioning and regulating functions, while others focus on cultural
functions. A clear distinction between the different types of solar land-
scapes may help to conceive solar power plants appropriate to the site-
specic considerations of local stakeholders and society at large.
Credit author contribution statement
Dirk Oudes - Conceptualization, Methodology, Validation, Investi-
gation, Resources, Data curation, Writing - original draft, Writing - re-
view & editing, Visualization, Project administration. Sven Stremke -
Conceptualization, Methodology, Investigation, Writing - review &
editing, Supervision, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgements
The authors are grateful to the experts that helped to identify the
cases, and to Florian Becker and Coos van Ginkel (both Wageningen
University) and Paolo Picchi (Amsterdam Academy of Architecture) for
their assistance in the data collection and analysis. Furthermore, we
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
14
thank Martine Uyterlinde (Netherlands Environmental Assessment
Agency) and Adri van den Brink (Wageningen University) for reviewing
an earlier version of this paper, and the members of the Dutch National
Consortium ‘Solar energy and Landscape’ and the Solar Research
Program (Wageningen University) for providing valuable feedback
during presentations of intermediate results. Last, we like to thank the
anonymous reviewers for their useful comments and critique.
Appendices.
Appendix A. Share of existing eye-level vegetation (e.g. forest patches or hedgerows) along the edge of the solar landscapes.
Appendix B. In addition to fencing, three types of measures were applied along the edge of the solar landscapes: removal of existing landscape features, enhancing
existing landscape features and new landscape features.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
15
Appendix C. Presence of functions in the cases following the Common International Classication of Ecosystem Services (CICES) [52]
Section Division Code Simple descriptor Absolute
presence
Relative
presence
Provisioning (Biotic) Biomass 1.1.1.1 Any crops and fruits grown by humans for food;
food crops
5 45%
Biomass 1.1.1.2 Material from plants, fungi, algae or bacterial
that we can use
2 18%
Biomass 1.1.3.1 Livestock raised in housing and/or grazed
outdoors
4 36%
Non-aqueous natural abiotic ecosystem outputs 4.3.2.4 Solar power 11 100%
Regulation &
Maintenance
(Abiotic)
Transformation of biochemical or physical inputs to
ecosystems
5.1.2.1 Natural protection 3 27%
Regulation &
Maintenance
(Biotic)
Transformation of biochemical or physical inputs to
ecosystems
2.1.2.3 Screening unsightly things 9 82%
Regulation of physical, chemical, biological conditions 2.2.1.3 Regulating the ows of water in our
environment
6 55%
Regulation of physical, chemical, biological conditions 2.2.2.1 Pollinating our fruit trees and other plants 9 82%
Regulation of physical, chemical, biological conditions 2.2.2.2 Spreading the seeds of wild plants 1 9%
Regulation of physical, chemical, biological conditions 2.2.2.3 Providing habitats for wild plants and animals
that can be useful to us
11 100%
Regulation of physical, chemical, biological conditions 2.2.4.2 Ensuring the organic matter in our soils is
maintained
1 9%
Cultural (Biotic) Direct, in-situ and outdoor interactions with living systems
that depend on presence in the environmental setting
3.1.1.1 Using the environment for sport and
recreation; using nature to help stay t
4 36%
Direct, in-situ and outdoor interactions with living systems
that depend on presence in the environmental setting
3.1.1.2 Watching plants and animals where they live;
using nature to destress
5 45%
Direct, in-situ and outdoor interactions with living systems
that depend on presence in the environmental setting
3.1.2.1 Researching nature 3 27%
Direct, in-situ and outdoor interactions with living systems
that depend on presence in the environmental setting
3.1.2.2 Studying nature 5 45%
Direct, in-situ and outdoor interactions with living systems
that depend on presence in the environmental setting
3.1.2.4 The beauty of nature 6 55%
Indirect, remote, often indoor interactions with living
systems that do not require presence in the environmental
setting
3.2.2.1 The things in nature that we think should be
conserved
7 64%
Cultural (Abiotic) Indirect, remote, often indoor interactions with physical
systems that do not require presence in the environmental
setting
6.2.2.1 Things in the physical environment that we
think are important to others and future
generations
4 36%
References
[1] Comello S, Reichelstein S, Sahoo A. The road ahead for solar PV power. Renew
Sustain Energy Rev 2018;92:744–56. https://doi.org/10.1016/j.rser.2018.04.098.
[2] Scognamiglio A. “Photovoltaic landscapes”: design and assessment. A critical
review for a new transdisciplinary design vision. Renew Sustain Energy Rev 2016;
55:629–61. https://doi.org/10.1016/j.rser.2015.10.072.
[3] Picchi P, van Lierop M, Geneletti D, Stremke S. Advancing the relationship between
renewable energy and ecosystem services for landscape planning and design: a
literature review. Ecosyst Serv 2019;35:241–59. https://doi.org/10.1016/j.
ecoser.2018.12.010.
[4] Pasqualetti M, Stremke S. Energy landscapes in a crowded world: a rst typology of
origins and expressions. Energy Res Soc Sci 2018;36:94–105. https://doi.org/
10.1016/j.erss.2017.09.030.
[5] Carullo L, Russo P, Riguccio L, Tomaselli G. Evaluating the landscape capacity of
protected rural areas to host photovoltaic parks in sicily. Nat Resour 2013:460–72.
https://doi.org/10.4236/nr.2013.47057. 04.
[6] Selman P. Learning to love the landscapes of carbon-neutrality. Landsc Res 2010;
35:157–71. https://doi.org/10.1080/01426390903560414.
[7] Ioannidis R, Koutsoyiannis D. A review of land use, visibility and public perception
of renewable energy in the context of landscape impact. Appl Energy 2020;276:
115367. https://doi.org/10.1016/j.apenergy.2020.115367.
[8] Council of Europe. European landscape convention. Eur Treaty Ser – No 2000;176:
7.
[9] Farina A. Principles and methods in Landscape Ecology. 2006. https://doi.org/
10.1111/j.1442-9993.2007.01854.x.
[10] Delicado A, Figueiredo E, Silva L. Community perceptions of renewable energies in
Portugal: impacts on environment, landscape and local development. Energy Res
Soc Sci 2016;13:84–93. https://doi.org/10.1016/j.erss.2015.12.007.
[11] Bevk T, Golobiˇ
c M. Contentious eye-catchers: perceptions of landscapes changed
by solar power plants in Slovenia. Renew Energy 2020;152:999–1010. https://doi.
org/10.1016/j.renene.2020.01.108.
[12] Wolsink M. Co-production in distributed generation: renewable energy and
creating space for tting infrastructure within landscapes. Landsc Res 2017;6397:
1–20. https://doi.org/10.1080/01426397.2017.1358360.
[13] Apostol D, Palmer J, Pasqualetti M, Smardon R, Sullivan R. The renewable energy
landscape: preserving scenic values in our sustainable future. Abingdon, Oxon:
Routledge; 2017.
[14] Torres-Sibille A del C, Cloquell-Ballester VA, Cloquell-Ballester VA, Artacho
Ramírez M´
A. Aesthetic impact assessment of solar power plants: an objective and a
subjective approach. Renew Sustain Energy Rev 2009;13:986–99. https://doi.org/
10.1016/j.rser.2008.03.012.
[15] S´
anchez-Pantoja N, Vidal R, Pastor MC. Aesthetic impact of solar energy systems.
Renew Sustain Energy Rev 2018;98:227–38. https://doi.org/10.1016/j.
rser.2018.09.021.
[16] Merida-Rodriguez M, Lobon-Martin R, Perles-Rosello M-J. The production of solar
photovoltaic power and its landscape dimension. In: Frolova M, Prados M-J,
Nadaï A, editors. Renew. Energies eur. Landscapes lessons from south. Eur. Cases.
Springer; 2015. p. 255–77. https://doi.org/10.1007/978-94-017-9843-3.
[17] Chiabrando R, Fabrizio E, Garnero G. The territorial and landscape impacts of
photovoltaic systems: denition of impacts and assessment of the glare risk. Renew
Sustain Energy Rev 2009;13:2441–51. https://doi.org/10.1016/j.
rser.2009.06.008.
[18] Tsoutsos T, Frantzeskaki N, Gekas V. Environmental impacts from the solar energy
technologies. Energy Pol 2005;33:289–96. https://doi.org/10.1016/S0301-4215
(03)00241-6.
[19] Turney D, Fthenakis V. Environmental impacts from the installation and operation
of large-scale solar power plants. Renew Sustain Energy Rev 2011;15:3261–70.
https://doi.org/10.1016/j.rser.2011.04.023.
[20] Hernandez RR, Easter SB, Murphy-Mariscal ML, Maestre FT, Tavassoli M, Allen EB,
et al. Environmental impacts of utility-scale solar energy. Renew Sustain Energy
Rev 2014;29:766–79. https://doi.org/10.1016/j.rser.2013.08.041.
[21] Poggi F, Firmino A, Amado M. Planning renewable energy in rural areas: impacts
on occupation and land use. Energy 2018;155:630–40. https://doi.org/10.1016/j.
energy.2018.05.009.
[22] Fthenakis V, Kim HC. Land use and electricity generation: a life-cycle analysis.
Renew Sustain Energy Rev 2009;13:1465–74. https://doi.org/10.1016/j.
rser.2008.09.017.
[23] Windemer R. Considering time in land use planning: an assessment of end-of-life
decision making for commercially managed onshore wind schemes. Land Use Pol
2019;87:104024. https://doi.org/10.1016/j.landusepol.2019.104024.
D. Oudes and S. Stremke
Renewable and Sustainable Energy Reviews 145 (2021) 111101
16
[24] Semeraro T, Pomes A, Del Giudice C, Negro D, Aretano R. Planning ground based
utility scale solar energy as green infrastructure to enhance ecosystem services.
Energy Pol 2018;117:218–27. https://doi.org/10.1016/j.enpol.2018.01.050.
[25] Roddis P, Roelich K, Tran K, Carver S, Dallimer M, Ziv G. What shapes community
acceptance of large-scale solar farms? A case study of the UK’s rst ‘nationally
signicant’ solar farm. Sol Energy 2020;209:235–44. https://doi.org/10.1016/j.
solener.2020.08.065.
[26] Wüstenhagen R, Wolsink M, Bürer MJ. Social acceptance of renewable energy
innovation: an introduction to the concept. Energy Pol 2007;35:2683–91. https://
doi.org/10.1016/j.enpol.2006.12.001.
[27] Batel S, Devine-Wright P, Tangeland T. Social acceptance of low carbon energy and
associated infrastructures: a critical discussion. Energy Pol 2013;58:1–5. https://
doi.org/10.1016/j.enpol.2013.03.018.
[28] Fernandez-Jimenez LA, Mendoza-Villena M, Zorzano-Santamaria P, Garcia-
Garrido E, Lara-Santillan P, Zorzano-Alba E, et al. Site selection for new PV power
plants based on their observability. Renew Energy 2015;78:7–15. https://doi.org/
10.1016/j.renene.2014.12.063.
[29] Stremke S, Sch¨
obel S. Research through design for energy transition: two case
studies in Germany and The Netherlands. Smart Sustain Built Environ 2019;8:
16–33. https://doi.org/10.1108/SASBE-02-2018-0010.
[30] Lovell ST, Johnston DM. Designing landscapes for performance based on emerging
principles in landscape ecology. Ecol Soc 2009;14. https://doi.org/10.5751/ES-
02912-140144.
[31] Brandt J, Vejre H. Multifunctional landscapes - motives, concepts and perceptions.
In: Brandt J, Vejre H, editors. Multifunct. Landscapes vol. 1 theory, values hist.
Southampton: WIT Press.; 2004. p. 3–32.
[32] Selman P. Planning for landscape multifunctionality. Sustain Sci Pract Pol 2009;5:
45–52. https://doi.org/10.1080/15487733.2009.11908035.
[33] Hernandez RR, Armstrong A, Burney J, Ryan G, Moore-O’Leary K, Di´
edhiou I, et al.
Techno–ecological synergies of solar energy for global sustainability. Nat Sustain
2019;2:560–8. https://doi.org/10.1038/s41893-019-0309-z.
[34] Moore-O’Leary KA, Hernandez RR, Johnston DS, Abella SR, Tanner KE,
Swanson AC, et al. Sustainability of utility-scale solar energy - critical ecological
concepts. Front Ecol Environ 2017;15:385–94. https://doi.org/10.1002/fee.1517.
[35] Randle-Boggis RJ, White PCL, Cruz J, Parker G, Montag H, Scurlock JMO, et al.
Realising co-benets for natural capital and ecosystem services from solar parks: a
co-developed, evidence-based approach. Renew Sustain Energy Rev 2020;125:
109775. https://doi.org/10.1016/j.rser.2020.109775.
[36] Lobaccaro G, Croce S, Lindkvist C, Munari Probst MC, Scognamiglio A, Dahlberg J,
et al. A cross-country perspective on solar energy in urban planning: lessons
learned from international case studies. Renew Sustain Energy Rev 2019;108:
209–37. https://doi.org/10.1016/j.rser.2019.03.041.
[37] Chiabrando R, Fabrizio E, Garnero G. On the applicability of the visual impact
assessment OAISPP tool to photovoltaic plants. Renew Sustain Energy Rev 2011;
15:845–50. https://doi.org/10.1016/j.rser.2010.09.030.
[38] Armstrong A, Ostle NJ, Whitaker J. Solar park microclimate and vegetation
management effects on grassland carbon cycling. Environ Res Lett 2016;11.
https://doi.org/10.1088/1748-9326/11/7/074016.
[39] Apostol D, McCarty J, Sullivan R. Improving the visual t of renewable energy
projects. In: Apostol D, Palmer J, Pasqualetti M, Smardon R, Sullivan R, editors.
Renew. Energy landsc. Preserv. Scen. Values our sustain. Futur. Abingdon, Oxon:
Routledge; 2017. p. 167–97. https://doi.org/10.1111/j.1477-8947.1989.tb00353.
x.
[40] De Marco A, Petrosillo I, Semeraro T, Pasimeni MR, Aretano R, Zurlini G. The
contribution of Utility-Scale Solar Energy to the global climate regulation and its
effects on local ecosystem services. Glob Ecol Conserv 2014;2:324–37. https://doi.
org/10.1016/j.gecco.2014.10.010.
[41] Yin RK. Case study research : design and methods. fourth ed., vol. 5. Thousand
Oaks CA: Sage; 2009. https://doi.org/10.1097/FCH.0b013e31822dda9e.
[42] Steenbergen CM. Composing landscapes : analysis, typology and experiments for
design. Basel: Birkh¨
auser Verlag; 2008.
[43] Frankl P. Principles of architectural history: the four phases of architectural style,
1420-1900. MIT Press; 1968.
[44] Massi Pavan A, Mellit A, De Pieri D. The effect of soiling on energy production for
large-scale photovoltaic plants. Sol Energy 2011;85:1128–36. https://doi.org/
10.1016/j.solener.2011.03.006.
[45] Barrett TL, Farina A, Barrett GW. Aesthetic landscapes: an emergent component in
sustaining societies. Landsc Ecol 2009;24:1029–35. https://doi.org/10.1007/
s10980-009-9354-8.
[46] Klijn JA. Hierarchical concepts in landscape ecology: pitfalls and promises.
Wageningen: DLO Winand Staring Centre; 1995. https://doi.org/10.1111/j.1460-
9568.2009.06623.x.
[47] Hastik R, Basso S, Geitner C, Haida C, Poljanec A, Portaccio A, et al. Renewable
energies and ecosystem service impacts. Renew Sustain Energy Rev 2015;48:
608–23. https://doi.org/10.1016/j.rser.2015.04.004.
[48] Doubleday K, Choi B, Maksimovic D, Deline C, Olalla C. Recovery of inter-row
shading losses using differential power-processing submodule DC-DC converters.
Sol Energy 2016;135:512–7. https://doi.org/10.1016/j.solener.2016.06.013.
[49] Haurant P, Oberti P, Muselli M. Multicriteria selection aiding related to
photovoltaic plants on farming elds on Corsica island: a real case study using the
ELECTRE outranking framework. Energy Pol 2011;39:676–88. https://doi.org/
10.1016/j.enpol.2010.10.040.
[50] Kapetanakis IA, Kolokotsa D, Maria EA. Parametric analysis and assessment of the
photovoltaics’ landscape integration: technical and legal aspects. Renew Energy
2014;67:207–14. https://doi.org/10.1016/j.renene.2013.11.043.
[51] Haines-Young R, Potschin M. The links between biodiversity, ecosystem services
and human well-being. In: Raffaelli D, Frid C, editors. Ecosyst. Ecol.; 2012.
p. 110–39. https://doi.org/10.1017/cbo9780511750458.007. Cambridge.
[52] Haines-Young R, Potschin M. Common international Classication of Ecosystem
Services (CICES) V5.1 and Guidance on the Application of the Revised Structure.
2018.
[53] Antrop M. Why landscapes of the past are important for the future. Landsc Urban
Plann 2005;70:21–34. https://doi.org/10.1016/J.LANDURBPLAN.2003.10.002.
[54] Guerin TF. Evaluating expected and comparing with observed risks on a large-scale
solar photovoltaic construction project: a case for reducing the regulatory burden.
Renew Sustain Energy Rev 2017;74:333–48. https://doi.org/10.1016/j.
rser.2017.02.040.
[55] Guerin T. A case study identifying and mitigating the environmental and
community impacts from construction of a utility-scale solar photovoltaic power
plant in eastern Australia. Sol Energy 2017;146:94–104. https://doi.org/10.1016/
j.solener.2017.02.020.
[56] Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I, Thies C. Landscape
perspectives on agricultural intensication and biodiversity - ecosystem service
management. Ecol Lett 2005;8:857–74. https://doi.org/10.1111/j.1461-
0248.2005.00782.x.
[57] Freitas S, Brito MC. Non-cumulative only solar photovoltaics for electricity load-
matching. Renew Sustain Energy Rev 2019;109:271–83. https://doi.org/10.1016/
j.rser.2019.04.038.
[58] Tsai CY, Tsai CY. See-through, light-through, and color modules for large-area
tandem amorphous/microcrystalline silicon thin-lm solar modules: technology
development and practical considerations for building-integrated photovoltaic
applications. Renew Energy 2020;145:2637–46. https://doi.org/10.1016/j.
renene.2019.08.029.
[59] Sinha P, Hoffman B, Sakers J, Althouse L. Best practices in responsible land use for
improving biodiversity at a utility-scale solar facility. Case Stud Environ 2018;
1–12. https://doi.org/10.1525/cse.2018.001123.
[60] Oudes D, Stremke S. Climate adaptation, urban regeneration and browneld
reclamation: a literature review on landscape quality in large-scale transformation
projects. Landsc Res 2020;45:905–19. https://doi.org/10.1080/
01426397.2020.1736995.
[61] Antrop M. Changing patterns in the urbanized countryside of Western Europe.
Landsc Ecol 2000;15:257–70. https://doi.org/10.1023/A:1008151109252.
[62] Frant´
al B, Van der Horst D, Martin´
at S, Schmitz S, Teschner N, Silva L, et al. Spatial
targeting, synergies and scale: exploring the criteria of smart practices for siting
renewable energy projects. Energy Pol 2018;120:85–93. https://doi.org/10.1016/
j.enpol.2018.05.031.
[63] Lovich JE, Ennen JR. Wildlife conservation and solar energy development in the
desert southwest, United States. Bioscience 2011;61:982–92. https://doi.org/
10.1525/bio.2011.61.12.8.
[64] Horner RM, Clark CE. Characterizing variability and reducing uncertainty in
estimates of solar land use energy intensity. Renew Sustain Energy Rev 2013;23:
129–37. https://doi.org/10.1016/j.rser.2013.01.014.
[65] Martín-Chivelet N. Photovoltaic potential and land-use estimation methodology.
Energy 2016;94:233–42. https://doi.org/10.1016/j.energy.2015.10.108.
D. Oudes and S. Stremke
