Content uploaded by Mike B Roberts
Author content
All content in this area was uploaded by Mike B Roberts on Apr 09, 2018
Content may be subject to copyright.
Spatial Analysis of Solar Potential in Adelaide
Prepared for the Australian PV Institute
by Jessie Copper, Mike Roberts and Anna Bruce, UNSW Sydney – April 2018
Key Findings
Our analysis shows that Adelaide CBD could generate more than 25% of its electricity
needs from its own rooftops, with the installation of 129 MW of solar on CBD rooftops.
Using the average results from our 4 methods:
• There is potential to install 129MW of solar photovoltaics on CBD rooftops
• There is potential for 32 times the existing PV deployment
• 43% of the total roof area could accommodate 516,000 solar panels
• this could generate 174GWh annually
o meeting 26% of the CBD energy demand
o supplying the equivalent of 34,000 SA households
o avoiding 77,000 tonnes of CO2 emissions
• CBD electricity customers could save up to an estimated $54 million per year
Analysis of 3 case study buildings in Adelaide’s CBD suggests potential solar PV capacities of:
350kW on the Centrepoint Building,
840 kW on the Central Market
and 1300 kW on the central building of Adelaide Convention Centre.
Page | i
Executive Summary
Introduction
There is significant potential for rooftop solar PV in Australia. Rooftop solar PV is a key energy
technology because it is leading the transition to consumer uptake of low-carbon demand-side
energy technologies, which are providing new opportunities for consumer engagement and new
clean energy business models to emerge. However, there is a lack of good information in the public
domain about the potential for rooftop solar to contribute to low-carbon electricity generation in
Australia’s cities. This type of information is important for policymakers and planners, and to
encourage public support for rooftop solar.
This research uses the data and methodologies behind the APVI Solar Potential Tool (SunSPoT),
developed by researchers at UNSW, to estimate the Solar Potential in the Adelaide CBD. The report
includes:
1. An assessment of PV Potential in Adelaide CBD
2. An estimate of the potential impact of rooftop PV on local electricity consumption and emissions
3. Identification of rooftops with the largest PV potential (area available) in the CBD
4. Three case studies of PV Potential on landmark buildings in Adelaide
Summary Results: Adelaide CBD
The useable area suitable for PV deployment across Adelaide’s CBD was calculated using two
different methods and two different datasets. The calculation takes account of the orientation and
slope of the rooftop, as well as the average insolation and the degree of shading.
Conservative and average results are presented in the body of the report. The average of 2
methodologies applied to 2 different datasets suggests that 43% of the total roof area in the CBD is
suitable for PV deployment. This area could accommodate over 516,000 solar PV panels, with a
generating capacity of 129 MW.
There is an estimated 4.0MW of PV capacity currently installed on Adelaide CBD rooftops, which
represents only 3% of the estimated potential capacity.
Annually, this could supply 174 GWh of electricity, approximately 26% of the total electricity demand
of the CBD (higher than most other Australian capitals), or the annual electricity demand of 33820
average South Australian households.
The equivalent CO2 emission savings are 77 kilotonnes per year.
The financial benefits of solar PV are highly specific to characteristics of the building and of the
electricity demand being met, as well as to contemporary electricity retail market conditions.
However, based on typical small business tariffs, we estimate the potential savings on electricity bills
to be in the region of $54million per year.
The rooftops with the largest PV potential in Adelaide have been mapped (Figure 1 below, with
more detailed images in Appendix B – Detailed Maps of Rooftops with Large Solar Potential
Page | ii
large - med - small
Figure 1: Rooftops with Largest PV Potential in Adelaide CBD
Summary Results: Cities Compared
Table 1 shows a summary of our results from other Australian capital cities. Although differences in
the data available in different jurisdictions mean that direct comparisons should be used cautiously,
the results suggest that Adelaide has the potential to supply the greatest proportion of its CBD load
from rooftop solar.
Table 1: State Capitals Compared
Usable
rooftop area
Potential Installed
Capacity (MW)
Potential Annual
Generation (GWh)
Estimated %
of Load
Brisbane
45%
188
241
11%
Melbourne
38%
461
548
11%
Sydney
40%
619
777
22%
Canberra
50%
68
98
17%
Adelaide
43%
129
174
26%
Page | iii
Summary Results: Case Studies
Case studies of specific landmark buildings - the Centrepoint Building, Central Market and the central
building of Adelaide Convention Centre - were carried out.
Table 2 presents the potential array capacity, expected annual energy production and estimated
carbon offsets for each system. The potential 2.4MW of PV generating capacity that could be
accommodated on these three buildings could save an estimated 1.5 kilotonnes of carbon emissions
each year and could supply the equivalent of 637 households.
Table 2: Carbon offset and household energy equivalents
Site
PV
Capacity
Expected Annual
Generation Emissions Offset Average ACT
household equivalent
(kWpeak) (MWh)
(Tonnes CO
2
-e /
year)
Centrepoint
Building 348 452 201 88
Central Market
837
1120
499
218
Convention
Centre - Central
Building
1303 1705 759 331
Totals
2487
3277
1458
637
The potential PV arrays for each building are shown in Figure 2, and in Figures 20 - 22 of the main
report.
(i) Centrepoint
Building
(ii) Central Market
(iii) Convention Centre
Figure 2 : Potential PV arrays on Case Study buildings
Page | iv
Spatial Analysis of Solar Potential in Adelaide
Contents
Key Findings ............................................................................................................................................. i
Executive Summary ................................................................................................................................. ii
Introduction ........................................................................................................................................ ii
Summary Results: Adelaide CBD ........................................................................................................ ii
Summary Results: Cities Compared .................................................................................................. iii
Summary Results: Case Studies ......................................................................................................... iv
Introduction ............................................................................................................................................ 1
Introduction to the Solar Potential Tool ................................................................................................. 1
Assessment of the PV Potential in Adelaide CBD ................................................................................... 3
Method 1: Insolation Limit .................................................................................................................. 3
Method 2: NREL’s Hillshade and Orientation ..................................................................................... 4
Input Data Source: AAM 3D Building Model vs. LiDAR data ............................................................... 6
Calculation of PV Capacity and Annual Yield ...................................................................................... 7
Calculation of Contribution to Total Load ........................................................................................... 8
Existing PV Capacity ............................................................................................................................ 9
Calculation of CO2-e Emission Reductions ........................................................................................ 10
Estimation of Financial Savings ......................................................................................................... 10
Results – PV Potential in Adelaide CBD................................................................................................. 11
Case Studies of Landmark Buildings ..................................................................................................... 13
Assessment of Roof Area .................................................................................................................. 13
Calculation of PV Capacity and Annual Yield .................................................................................... 15
Calculation of Emissions Offset ......................................................................................................... 15
Case Study Results ............................................................................................................................ 16
Case Study Illustrations ..................................................................................................................... 17
References ............................................................................................................................................ 20
Appendix A – Comparison between APVI SPT Simple PV Performance Method vs. Detail Hourly
Simulation of PV Performance in NREL’s System Advisor Model ......................................................... 21
Appendix B – Detailed Maps of Rooftops with Large Solar Potential ................................................... 22
Page | v
Introduction
There is significant potential for rooftop solar PV in Australia. Rooftop solar PV is a key energy
technology because it is leading the transition to consumer uptake of low-carbon demand-side
energy technologies, which are providing new opportunities for consumer engagement and new
clean energy business models to emerge. However, there is a lack of good information in the public
domain about the potential for rooftop solar to contribute to low-carbon electricity generation in
Australia’s cities. This type of information is important for policymakers and planners, and to
encourage public support for rooftop solar.
This research uses the data and methodologies behind the APVI Solar Potential Tool http://pv-
map.apvi.org.au/potential, developed by researchers at UNSW, to estimate the Solar Potential in the
Adelaide CBD. The report includes:
1. An assessment of PV Potential in Adelaide CBD
2. An estimate of the potential impact of rooftop PV on local electricity consumption and emissions
3. Identification of rooftops with the largest PV potential (area available) in the CBD
4. Three case studies of PV Potential on landmark buildings in Adelaide
Introduction to the Solar Potential Tool
The APVI Solar Potential Tool (SunSPoT) is an online tool to allow electricity consumers, solar
businesses, planners and policymakers to estimate the potential for electricity generation from PV
on building roofs. The tool accounts for solar radiation and weather at the site; PV system area, tilt,
orientation; and shading from nearby buildings and vegetation.
Figure 3 APVI Solar potential Map (SunSPoT)
Page | 1
At a city level, an insolation heatmap layer (Figure 4b) allows identification of the best roofs, while
the shadow layer (Figure 4c) allows the user to locate an unshaded area on a rooftop. The tool
allows users to select any building within the mapped area, outline a specific roof area and
automatically generate an estimate of potential annual electricity generation, financial savings and
emissions offset from installing solar PV.
Figure 4: (a) Aerial photograph (b) Insolation heat map, (c) Winter shadow layer
The data behind the APVI SPT were generated as follows:
1. Three types of digital surfaces models (DSMs)1 (3D building models, XYZ vegetation points
and 1m ESRI Grids), supplied by geospatial company AAM, were used to model the buildings
and vegetation in the areas covered by the map.
2. These DSMs were used as input to ESRI’s ArcGIS tool to evaluate surface tilt, orientation and
the annual and monthly levels of solar insolation falling on each 1m2 unit of surface.
3. Insolation values output by the ArcGIS model were calibrated2 to Typical Meteorological
Year (TMY) weather files for each of the capital cities and against estimates of insolation at
every 1 degree tilt and orientation from NREL’s System Advisor Model (SAM).
This project expanded the data and methodologies behind the Solar Potential in order to estimate
the Solar Potential in the Adelaide CBD region.
1 Digital surface models provide information about the earth’s surface and the height of objects. 3D building
models and vegetation surface models have been used in this work. The ESRI Grid is a GIS raster file format
developed by ESRI, used to define geographic grid space.
2 Calibration was required in order to obtain good agreement NREL’s well-tested SAM model and measured PV
data.
Page | 2
Assessment of the PV Potential in Adelaide CBD
This section of the report details the methodology and the results of the geospatial analysis of PV
potential across Adelaide CBD.
The assessment of the PV potential in Adelaide’s CBD, expanded on the initial work undertaken for
the Adelaide region of APVI’s SPT. The analysis made use of the following data sources:
1. The three sources of input DSMs data from AAM; and
2. City of Adelaide LiDAR data – 2010 dataset sourced from the City of Adelaide
The general steps in the methodology are illustrated in Figure 5. To test the sensitivity of the
estimated PV potential two input data sources and two rooftop suitability methods were assessed.
The two input data sources used to calculate the tilt, aspect, solar insolation and determine suitable
roof planes were 1) the DSM and 3D building models from AAM and 2) the 2010 City of Adelaide
LiDAR data covering Adelaide CBD. The two methods utilised to determine suitable rooftops were 1)
based on a minimal level of surface insolation and 2) NREL’s PV rooftop suitability method based on
hillshade and surface orientation. Both methods also required a minimum contiguous surface area of
10m2 for a roof plane to be determined suitable. This limit was defined to ensure a minimum 1.5kW
PV system for any plane defined as suitable.
Figure 5: Major process steps for the calculation of rooftop PV potential
Method 1: Insolation Limit
The first method utilised to determine suitable roof planes was based on a minimum level of
insolation. The minimum value was set at an annual average insolation of 3.99 kWh/m2/day. This
limit was calculated as 80% of the expected level of annual insolation for a horizontal surface in
Input Data Source:
AAM or LiDAR
Calculation of roof surface
Tilt and Aspect Calculation of Hillshades
Calculation of surface
Insolation
Identification of Unique
roof surfaces
Assessment of rooftop
suitability:
a) Insolation
b) NREL Hillshade & aspect
Minimum criteria of 10m2
of contigous area
Calculation of PV Capacity
and Yield per suitable roof
plane
Region aggregation to
Sydney City Suburbs
Page | 3
Adelaide, calculated as 4.98 kWh/m2/day, using the default TMY weather file for Adelaide contained
within the National Renewable Energy Laboratories (NREL) System Advisor Model (SAM). This limit
was applied to the Solar Insolation Heat Map which was developed and calibrated as part of the
APVI SPT methodology [1, 2].
Figure 6 presents an example application of the insolation limit in practice, displaying an aerial image
(left), the insolation heat map (centre) and the classified insolation layer (right); classified as either
above (white) or below (black) the insolation limit. As for each method in this report, a 10m2
contiguous area was required for a roof plane to be determined suitable. Figure 7 presents the roof
planes that were identified to meet both the insolation and 10m2 contiguous area criteria for the
example presented in Figure 6.
Figure 6: Example application of the Insolation limit. Areal image (left); Insolation heat map
(centre); and classified Insolation layer (right)
Figure 7: Example application of suitable planes (hatched areas) by the Insolation limit method.
Method 2: NREL’s Hillshade and Orientation
The second method utilised to determine suitable roof planes was the method developed by NREL to
assess the technical potential for rooftop PV in the United States [3]. NREL’s method makes use of
ArcGIS’s hillshade function to determine the number of hours of sunlight received on each 1m2 of
Page | 4
roof surface, across 4 representative days within a year i.e. the winter and summer solstices and the
two equinoxes; similar to the shadow layers of APVI’s SPT as illustrated in Figure 4.
To determine which areas met the shading criteria, NREL’s method defines that roof surfaces must
meet a minimum number of hours of sunlight. The limit for any location can be determined by
calculating the number of hours a rooftop would need to be in sunlight to produce 80% of the
energy produced by an unshaded system of the same orientation [3]. For the location of Adelaide
the value was determined to be 13.69 hours across the 4 representative days.
In addition to the hillshade limit, NREL’s method also excludes roof planes based on orientation. In
NREL’s method all roof planes facing northwest through northeast (i.e. 292.5 - 67.5 degrees for
northern hemisphere locations) were considered unsuitable for PV. For southern hemisphere
locations the equivalent exclusion would be orientations southeast through southwest (i.e. 112.5 –
247.5 degrees) as per Figure 8. Again, as for each method in this report, a 10m2 contiguous area is
also required by NREL’s methodology.
Figure 8: Rooftop azimuths included in final suitable planes for the Southern Hemisphere
Figure 9 presents an example application of NREL’s hillshade and orientation limit in practice. For
this particular example there is reasonable agreement between the surfaces determined as suitable
for PV deployment from the two methods i.e. Figure 7 vs Figure 9. This is not always the case as
evident in the example presented in Figure 10, which illustrates how the insolation limit method can
define roof planes orientated southeast through southwest as suitable planes if the annual
insolation meets the limit requirement.
Page | 5
Figure 9: Example application of the hillshade limit (left) with the suitable planes overlayed (right)
Figure 10: Comparison between roof planes defined as suitable by the insolation method (both -
yellow) and NREL’s hillshade and orientation method (Left – orange)
Input Data Source: AAM 3D Building Model vs. LiDAR data
The other variable that affected the sensitivity of the estimated PV potential was the input data
source. Two input data sources were available for use in this analysis:
1. The DSMs and 3D building models from AAM, which were utilised to generate the APVI SPT,
2. The 2010 City of Adelaide LiDAR data dataset.
The application of the PV potential analysis was applied identically to both input data sources.
Generally, Figure 11 demonstrates that there is general agreement between the roof planes
identified as suitable via the two input data sources. However, the figure also illustrates how the
analyses undertaken with the LiDAR data set excludes a greater proportion of roof surfaces.
Page | 6
Figure 11: Example of good agreement between the two input data source for large buildings.
Aerial image (Left), AAM 3D buildings with Insolation limit method (centre); Adelaide LiDAR with
Insolation limit method (Right)
Calculation of PV Capacity and Annual Yield
After suitable roof planes have been identified, the PV capacity and annual yield for each roof
surface can be calculated. The DC PV capacity (otherwise known as system size) was calculated as
per APVI’s SPT methodology [1] using the DC size factor and array spacing methodologies [4]. The
relevant equations for this method can be found here.
Generally, the method assumes a fixed DC size factor of 156.25 W/m2 (i.e. a 250W module with
dimensions of 1m x 1.6m) for flush mounted arrays, and a variable DC size factor for rack mounted
PV arrays. For rack mounted arrays, the DC size factor is a function of the PV array tilt and
orientation and the tilt and orientation of the underlying roof surface. Figure 12 presents the
equivalent useable roof area, which is analogous to the DC size factor, for a 15 degree tilted north
facing PV array in Adelaide, as a function of the tilt and orientation of the underlying roof surface.
For an absolutely flat roof, Figure 12 indicates a useable area of 69%, analogous to a DC size factor of
108 W/m2. In comparison, NREL’s method assumes a fixed ratio of module to roof area of 70% for
flat roof surfaces.
As per NREL’s method to calculate the PV potential in the United States [3], this analysis has
assumed that rack mounted arrays will be installed on flat and relatively flat roof surfaces. For
consistency with NREL’s method, flat roofs have been defined as roof surfaces with a tilt <= 9.5
degrees and the tilt angle of the rack mounted arrays were defined as 15 degrees.
Similarly, for tilted roof surfaces > 9.5 degrees, an additional module to roof area ratio of 0.98 was
assumed in the NREL method to reflect 1.27cm of spacing between each module for racking clamps.
This assumption was also applied in this study.
Page | 7
Figure 12: Percentage of useable roof area as a function of roof tilt and orientation for a 15 degree
North facing array in Adelaide
The PV yield was calculated using APVI’s SPT methodology as detailed here. This method multiplies
the calculated DC PV capacity by the average annual level of insolation calculated on the roof surface
and by a derating factor of 0.77. The derating factor accounts for all the typical PV losses of
temperature, soiling, wiring, mismatch, manufacturing module tolerance and inverter efficiency. This
simplified method shows good agreement with detailed hourly PV performance simulations
undertaken in NREL’s SAM as illustrated in Appendix A.
Calculation of Contribution to Total Load
The potential contribution of rooftop PV generation to electricity load in the CBD area was estimated
by comparison to the annual energy consumption seen at the zone substations in SA Power
networks’ Adelaide Central Region (ACR) which includes the CBD area for which rooftop PV was
modelled. These substations and loads are listed in Table 3, and mapped in Figure 13. The total
annual demand for these substations is 674GWh, but it should be noted that these substations may
feed some areas outside of the mapped CBD. Due to lack of information about which customers are
connected to different feeders in the distribution network, it is not possible to accurately estimate
the load in the CBD. Nevertheless, this figure can be used to give a sense of the scale of PV
contribution to load in the Adelaide CBD area.
Table 3: Load Data from SA Power Networks’ ACR Substations 2016-17 [5]
Zone Substation
Annual Load (GWh)
Coromandel Place 66/11kV
169
East Terrace 66/11kV
148
East Terrace 66/33kV
39
Hindley Street 66/11kV
160
Hindley Street 66/33kV
29
Whitmore Square 66/11kV
129
TOTAL ACR
674
0
5
10
15
0
30
60
90
120
150
180
210
240
270
300
330
359
Roof Tilt Angle (°)
Roof Orientation from North (°)
0.9-1
0.8-0.9
0.7-0.8
0.6-0.7
0.5-0.6
0.4-0.5
0.3-0.4
0.2-0.3
0.1-0.2
0-0.1
Page | 8
Figure 13: SA Power Networks’ Adelaide Central Region [6]
The annual yield was also compared to the average 2014 electricity demand of a South Australian
household, being 5145 kWh [7].
Existing PV Capacity
In order to assess the potential for additional rooftop PV in the Adelaide CBD, and associated
emissions reductions and electricity savings, existing PV capacity in the area was estimated. The CBD
area covered by this assessment falls within the postcode area 5000 (see Figure 14 ). Using the
Clean Energy Regulator’s database of PV systems registered under the Renewable Energy Target
scheme (accessed via the APVI’s Solar Map[8]), which is a near complete record of PV systems
installed in Australia, the installed PV capacity in this postcode area is given in Table 4. The total
existing PV capacity in the CBD is therefore estimated to be around 4.0 MW.
Table 4: Existing PV Capacity in Adelaide Postcode 5000 [9]
POA
5000
PV less than 10kW (kW)
1397
PV 10kW to 100kW (kW)
2492
PV bigger than 100kW (kW) 151
Total PV Capacity (kW) 4040
Page | 9
Figure 14: Adelaide CBD and Postcode Areas
Calculation of CO2-e Emission Reductions
The annual CO2-equivalent emission reductions are calculated by multiplying the estimated annual
yield by an appropriate emissions factor for South Australia as sourced from the 2017 National
Greenhouse Account Factors[9]. The relevant value for South Australia was 0.49 kg CO2-e/kWh
which is reduced by 0.045 kg CO2-e/kWh to account for the embodied carbon emissions from the
manufacture, installation, operation and decommissioning of the PV systems. (Note that this value
is lower than other jurisdictions due to the relatively high proportion of renewables in the South
Australian energy generation mix.)The value of 45 g CO2-e/kWh of electricity produced was sourced
from the PV LCA Harmonization Project results found in [10], which standardised the results from 13
life cycle assessment studies of PV systems with crystalline PV modules, assuming system lifetimes of
30 years.
Estimation of Financial Savings
As well as depending on the size and orientation of the PV panels and efficiency of the PV system,
the financial benefits of rooftop solar PV are highly specific to the particular energy user and to
market conditions. Bill savings depend on the amount of generated electricity that is self-consumed
(avoiding electricity purchase costs), the amount exported to the grid (in exchange for feed-in tariff)
and on the available electricity retail tariffs for import and export. However, we are able to make
some broad estimates for potential savings, based on typical values for commercial tariffs.
A standard commercial retail tariff in South Australia is Origin’s SA Small Business eSaver General
Supply Tariff which has an energy charge of 44.36 c/kWh with a standard solar feed-in tariff of
11c/kWh paid on exports. (It is important to note that larger businesses will pay significantly less for
their electricity use, with larger charges for other components of their bill, and that FiTs up to
22c/kWh are available for some customers.)
Page | 10
For commercial buildings, self-consumption during the week is likely to be high due to high daytime
loads, but on weekends there is likely to be significant solar export for some types of businesses,
depending on the size of the PV system compared to the load. A 60% self-consumption case is
therefore probably quite conservative for commercial buildings in the CBD.
=
× + ×
Based on these estimates,
×60% +
×40% ×
($) (0.4436 ×60% + 0.11 ×40% ) ×
Results – PV Potential in Adelaide CBD
Table 5 shows the results of the rooftop suitability assessment for the Adelaide CBD for the two data
sources and the two methods outlined above.
Table 5: Detailed results of rooftop suitability calculated using i) AAM DSM and 3D buildings and ii)
Adelaide North 2013 LiDAR dataset from NSW LPI
Data
Source
Method 1 - Insolation Limit (3.99
kWh/m2/day) - 3D Buildings
Method 2: NREL Hillshade
E/NE/N/NW/W (13.69) - 3D Buildings
Total Area
(ha)
Developable
(ha)
% Useable
Capacity
(MW)
Yield (GWh)
Developable
(ha)
%
Useable
Capacity
(MW)
Yield
(GWh)
DSM / 3D
190.34
92.61
48.7%
144.7
196.15
96.1
50.5%
150.20
201.3
LiDAR
68.84
36.2%
107.6
146.76
72.0
37.8%
112.46
151.6
Conservative Results
The most conservative estimate suggests the useable area suitable for rooftop PV deployment
(the ratio between the area of PV panels that could be accommodated and the total roof area) is
36% corresponding to 108 MW of PV potential with an expected annual yield of 147 GWh. These
values were calculated using the LiDAR data as the input data source in conjunction with the
Insolation method.
This corresponds to approximately 22% of the energy used in Adelaide CBD, or the average 2014
annual energy use of 28570 SA households
The equivalent CO2 emission savings are 65 kilotonnes per year with estimated potential financial
savings of $45million, although this is highly dependent on the specific circumstances of the building
occupants.
Page | 11
Average Results
In Table 6, results are presented for the average and standard deviation (Std) of the sensitivity
analysis undertaken by assessing the two input data sources and the two calculation methodologies.
Table 6: Summary of results for Adelaide CBD
Adelaide
Percentage Useable Area
Capacity (MW)
Yield (GWh)
Average
Std
Average
Std
Average
Std
CBD
43.3%
7.3%
128.7
21.8
173.9
28.7
The average of the two methods indicated that an area equal to 43% of the available roof surfaces
could be used to accommodate PV, corresponding to 129 MW of PV potential with an expected
annual yield of 174 GWh.
This corresponds to approximately 26% of the energy used in Adelaide CBD, or the average 2014
annual energy use of 33820 SA households. The equivalent CO2 emission savings are 77 kilotonnes
per year with estimated potential financial savings of $54million, although this is highly dependent
on the specific circumstances of the building occupants. There is an estimated 4.0 MW of existing PV
capacity installed on Adelaide CBD rooftops, approximately 3% of the potential capacity. The
electricity generation and emissions savings calculated would therefore be almost all additional.
The rooftops with the largest PV potential in Adelaide have been mapped (Figure 15 below).
large - med - small
Figure 15: Rooftops with Largest PV Potential in Adelaide CBD
Page | 12
Case Studies of Landmark Buildings
This section of the report details the methodology and the results for a detailed assessment of the
PV potential for 3 landmark Adelaide buildings: the Centrepoint Building, Central Market and the
central building of Adelaide Convention Centre. These buildings were chosen because of their high
public profile and large roof areas. Note that a PV array of approximately 75kW PV was installed on
the Central Market in 2011, but the roof area has much greater potential. Also, the assessment of
the Convention Centre only includes the central of the three buildings, as the more recent
development was not included in the dataset used for this analysis.
The case studies were assessed by combining the same GIS analysis used to assess the PV potential
of Adelaide CBD with a visual assessment of the building roof profiles using aerial imagery. No
structural assessment of the buildings has been carried out.
Assessment of Roof Area
Firstly, Method 1 above was used to identify developable roof planes: continuous areas greater than
10m2 receiving 80% of the annual insolation for an unshaded horizontal surface (3.99 kWh/m2/day).
Figure 16: Developable Planes with > 3.99 kWh/m2/day
The roof surfaces were then assessed visually, using imagery from multiple sources: aerial plan view
images from Nearmap and Google Earth, multiple viewpoint aerial imagery from Nearmap, and
photographs sourced from the internet. Unsuitable surfaces, including staircases, temporary
structures, and public spaces (roof terraces, platforms, etc.), were identified and excluded from the
usable roof area.
Figure 17: Examples of unsuitable surfaces (a) rooftop terrace, (b) temporary structure, (c)
staircase
Page | 13
Small rooftop obstructions and perimeter walls below the resolution of the GIS data were also
identified and their height was estimated using multiple viewpoint aerial imagery. (see Figure 18)
Figure 18: Estimation of rooftop obstructions
The shading on a PV module at a range of distances from obstructions of different heights was
modelled using the 3D shading calculator in NREL’s System Advisor Model (SAM) and the impact on
annual output for a horizontal PV panel in Adelaide (using the Adelaide RMY weather file from
Energy Plus[11]) was calculated. Figure 19 shows the results for a small range of distances and wall
heights. Using this data, additional roof area proximate to rooftop obstructions was excluded if
estimated annual output was less than 80% of an unshaded horizontal panel.
Figure 19: Nearest distance to obstruction to give 80% annual output
Nearmap’s Solar Tool was then used to arrange 1.6m x 1.0m PV panels on the usable roofspace, with
the roof slope determined from the GIS building slope layer. For sloping roofs, the panels were
positioned flush with the roof in order to avoid self-shading and maximise generation. For flat roofs,
panels were orientated towards North (i.e. between 045°and 315°) at a tilt angle of 5°.
Page | 14
As the assessment was carried out remotely, there may be additional physical constraints on the
available roof area as well as structural restrictions on the potential array size that have not been
considered here.
Calculation of PV Capacity and Annual Yield
The power capacity of the array was calculated using a nominal output of 250W per module
(equivalent to a DC size factor of 156.25 W/m2), and an initial value for the predicted annual energy
output (without accounting for shading losses) was calculated for each orientation and tilt using
SAM’s PVWatts model and a derate factor of 0.77.
To account for shading losses, the average yield (in kWh/kW/day) was calculated using the APVI SPT
method, averaged across all developable roof planes within the building footprint. This yield was
then applied to the calculated array size to give a predicted annual generation accounting for
shading losses. As it is outside the area of the APVI solar potential map, shading losses for Suncorp
stadium were modelled using SAM’s 3D Shading Model.
The annual yield was compared to the average annual electricity demand of a South Australian
households, being 5145 kWh [7].
Calculation of Emissions Offset
The potential CO2-e emissions reductions from the modelled PV systems on the 3 landmark buildings
were calculated by multiplying the indirect (Scope 2) emissions factor for consumption of electricity
purchased from the grid in SA (0.49 kg CO2-e/kWh[9]) by the expected annual energy generation
from the system, and subtracting the estimated embodied carbon emissions from the manufacture,
installation, operation and decommissioning of the PV system (0.045kg CO2-e /kW[10])
Page | 15
Case Study Results
Table 7 shows the potential roof area available for PV installation on each building
Table 7: Available roof areas
Site Total Roof Area
(m2)
Developable
Planes (m2)
Array Area
(m2)
Array Area /
Roof Area
Centrepoint Building
3,936
3,442
2,224
57%
Central Market
7,566
7,315
5,354
71%
Convention Centre - Central
Building
15,722 13,177 8,338 53%
Table 8 shows the projected array capacity and expected annual energy production. The proposed
PV arrays are illustrated in Figure 20 -Figure 22 below.
Table 8: Expected Annual Energy Production
Site PV Capacity
Annual Energy
Production
(w/o shading)
Average Yield
per kW PV
installed
Expected Annual
Energy Production
(adjusted)
(kWpeak) (MWh/year) (kWh/kW/day) (MWh/year)
Centrepoint Building
348
489
3.86
452
Central Market
837
1225
4.01
1120
Convention Centre -
Central Building
1303 1821 3.83 1705
Table 9 presents the estimated carbon offsets for each system and shows that these three buildings
could save an estimated 1.5 kilotonnes of carbon emissions each year and could supply the
equivalent of 637 households, based on the average 2014 electricity demand of an SA household
being 5145 kWh [7].
Table 9: Carbon offset and household energy equivalents
Site
Expected Annual
Energy Production
Emissions Offset
Average ACT
household equivalent
(MWh/year)
(Tonnes CO
2
-e /
year)
Centrepoint Building
452
201 88
Central Market
1120
499 218
Convention Centre - Central
Building
1705 759 331
Totals 3277 1458 637
Page | 16
Figure 21: Central Market, now (inset) and with potential 837kW PV Array
Page | 18
Figure 22: Adelaide Convention Centre central building, now (inset) and with potential 1.3MW PV
array
Page | 19
References
1. Copper, J.K. and A.G. Bruce. APVI Solar Potential Tool - Data and Calculations. 2014
Accessed; Available from:
http://d284f79vx7w9nf.cloudfront.net/assets/solar_potential_tool_data_and_calcs-
2dc0ced2b70de268a29d5e90a63432d7.pdf.
2. Copper, J.K. and A.G. Bruce. Validation of Methods Used in the APVI Solar Potential Tool.
2014 Accessed; Available from: http://apvi.org.au/solar-research-conference/wp-
content/uploads/2015/04/1-Copper_APVI_PVSystems_PeerReviewed.pdf.
3. Gagnon, P., et al. Rootop Solar Photovoltaic Technical Potential in the United States: A
Detailed Assessment. NREL/TP-6A20-65298 2016 Accessed; Available from:
http://www.nrel.gov/docs/fy16osti/65298.pdf.
4. Copper, J.K., A.B. Sproul, and A.G. Bruce, A method to calculate array spacing and potential
system size of photovoltaic arrays in the urban environment using vector analysis. Applied
Energy, 2016. 161: p. 11-23.
5. SA Power Networks. Zone Substation Data. 2016-2017;
https://www.sapowernetworks.com.au/custom/files/zonesubreport/2016-
2017/Zone%20Sub%20Report%20-%20ALL%20REGIONS%20-%2020171023.zip.
6. SA Power Networks. Distribution Annual Planning Report 2016/17 to 2020/21. 2016;
Available from: https://www.sapowernetworks.com.au/public/download.jsp?id=28109.
7. Acil Allen Consulting. Electricity Benchmarks final report v2 - Revised March 2015. 2015.
8. Australian Photovoltaic Institute. Australian PV Institute (APVI) Solar Map, funded by the
Australian Renewable Energy Agency. 2017 Accessed: 8/12/2017; Available from: http://pv-
map.apvi.org.au/postcode.
9. Department of Environment and Energy. National Greenhouse Accounts Factors July 2017.
2017.
10. Hsu, D.D., et al., Life Cycle Greenhouse Gas Emissions of Crystalline Silicon Photovoltaic
Electricity Generation. Journal of Industrial Ecology, 2012. 16: p. S122-S135.
11. Energy Plus Weather Data - Southwest Pacific Region. Accessed; Available from:
https://energyplus.net/weather-region/southwest_pacific_wmo_region_5/AUS%20%20.
Page | 20
Appendix A – Comparison between APVI SPT Simple PV Performance Method
vs. Detail Hourly Simulation of PV Performance in NREL’s System Advisor
Model
Figure 23 presents a comparison between the calculated annual yields using APVI SPT simplified
method versus detailed hourly simulations of PV performance using NREL’s SAM PVWatts module
with default settings. The results highlight the similarity in the calculated values, and demonstrate
how the annual yield can be calculated using a simplified methodology, which requires as input only
the annual or monthly averages of surface insolation in kWh/m2/day. The simplified APVI SPT
methodology enables geospatial calculation of yield for each identified roof surface.
Figure 23: Correlation between APVI SPT simplified method to calculate annual yield from annual
average insolation vs. detailed hourly simulations of PV performance from NREL’s SAM. Results
presented for each 1 degree combination of tilt (0-90°) and orientation (0-360°).
y = 0.9965x + 11.166
R² = 0.9997
0
200
400
600
800
1000
1200
1400
1600
1800
0500 1000 1500 2000
APVI SPT Annual Yeild (kWh/kWp)
SAM Detailed Hourly Annual Yeild (kWh/kWp)
Page | 21
Appendix B – Detailed Maps of Rooftops with Large Solar Potential
high - med - low
Page | 22
high - med - low
Page | 23
high - med - low
Page | 24
high - med - low
Page | 25
