Measured against modelled yield, 220 rooftop systems, one year

Renewable energy research poster example

A renewable energy research poster earns its place when it tests a number people rely on. This Australian A0 landscape board checks the design model behind rooftop solar quotes against a year of meter data from 220 systems, and finds the model overstates yield by 3.1% overall and by 7% on south facing roofs.

Create a poster with OneCraftA0 Landscape, printed at 1189 by 841 millimetres

The whole board

The poster at full size, exactly as it prints. Every number, citation and caption on it was written for this example, so the layout is being judged on real content.

SU
Measured against modelled yield on 220 Australian rooftop solar systems: where the design model is too hopeful
C. Doyle ¹, R. Thakur ¹, E. Lindgren ²
1 Solar Systems Laboratory, Saltbush University · 2 School of Electrical Engineering, Saltbush University
Abstract
Installers quote a yearly yield from a design model, and owners judge the system against that number. We compared a full year of 15 minute meter data from 220 rooftop systems with the yield the same model predicted for each one. Across the fleet the model overstated yield by 3.1%. The gap grows as panels turn away from north: 2.0% for north facing arrays, 2.3% west, 4.8% east and 6.9% south, where measured yield was 1,020 kWh per kW against 1,090 modelled.
N
E
W
S
0%2%4%6%8%
Figure 1. How far the model overstated yearly yield, per cent of measured kWh per kW, for N, E, W and S facing arrays, 2025.
Source: network meter data, 220 systems, 2025.
Results
Measured yield beat the model on only 9 of 220 systems. The gap by orientation: north 1,480 against 1,510 kWh per kW, east 1,250 against 1,310, west 1,290 against 1,320 and south 1,020 against 1,090. Weighted by system count, the fleet produced 1,335 kWh per kW against 1,376 modelled, a 3.1% shortfall. East facing arrays lost more than west because morning cloud was more common in this year.
METHODS
Systems. 220 residential systems of 5 to 10 kW in one Australian metro area: 96 north facing, 48 east, 52 west and 24 south.
Period. Calendar year 2025, with every system in service for the whole year and no inverter swaps.
Meter data. Export and generation at 15 minute intervals, 35,040 a system, from the network meter.
Model. One commercial design model run with each site address, tilt, azimuth and panel data, plus 2025 weather.
Comparison. Yearly kWh per kW installed, measured against modelled, by orientation and by shading class.
COMPLETION BY ARM
North facing, 96 systems: intervals logged
99.1%
East facing, 48 systems: intervals logged
98.7%
West facing, 52 systems: intervals logged
98.9%
South facing, 24 systems: intervals logged
98.2%
Shading losses
The model gap by shading class: unshaded, 118 systems, 1.6%; light, 58 systems, 3.2%; moderate, 30 systems, 5.6%; heavy, 14 systems, 9.3%. Shade explains part of the south gap but not all of it: unshaded south arrays still fell 5.1% short. Orientation itself needs a correction in the quote, not only shade.
Why check the model
Most Australian rooftops cannot face due north, so more east, west and south facing arrays are being sold every year. The design model behind a quote was tuned mostly on north facing systems. If it is less accurate on other roofs, the owners of those roofs are being promised more than they will get, and the error lands on the households with the least favourable roofs. This board tests that.
7%
model overstates south facing yield
1,090 kWh per kW modelled against 1,020 measured, 24 systems, 2025
Key finding
A quote for a south facing roof should come with a lower number than the model gives.
Fleet yield 3.1% below the model across 220 systems
South facing arrays 6.9% below the model, north 2.0%
Heavy shade adds the largest gap, 9.3% on 14 systems
Meter records over 98% complete for every orientation
REFERENCES
1.Harwood, J. and Pillai, S. (2024) Orientation bias in residential yield models. Solar Energy Practice, 22(3), 201 to 219.
2.Nkemelu, A. (2025) Reading 15 minute network meter data for rooftop solar. Energy Data Notes, 7(1), 12 to 30.
3.Doyle, C. (2025) Shade classes for rooftop audits. Saltbush Solar Working Paper 3, 2025.

Block by block

What each block on the board is for, in the order a reader walks it.

Title band and three authors
The title names the comparison, the 220 systems and the question in plain words: where the design model is too hopeful.
Abstract and why check the model
The abstract gives the fleet gap and the four orientation gaps. The second section explains why it matters in Australia, where many roofs cannot face north.
Methods
Five entries: the 220 systems by orientation, calendar year 2025, 15 minute meter data at 35,040 intervals a system, the one design model and the comparison in kWh per kW.
Figure 1, model overstatement by orientation
Four bars in per cent, labelled N, E, W and S. The measured and modelled kWh per kW behind each bar are in the abstract and results.
Results and the completeness bars
Results give the four measured and modelled pairs. The bar block, whose heading the design fixes as completion by arm, shows the share of 15 minute intervals logged for each orientation, all above 98%.
Shading losses
The four row shading table from the brief, written as text because this design has no table block: 1.6%, 3.2%, 5.6% and 9.3% from unshaded to heavy shade.
Key number, key finding, points and references
The stat block holds 7% for south facing roofs. The callout says a south facing quote should come in lower than the model, followed by four checked points and three references.

What makes this board work

One gap, plotted directly

Figure 1 plots how far the model overstated yield for each orientation, 2.0%, 4.8%, 2.3% and 6.9%, rather than two near identical bars per roof. The difference is the finding, so the difference is what the chart draws.

The fleet figure reconciles

Weighting the four orientations by 96, 48, 52 and 24 systems gives 1,335 kWh per kW measured against 1,376 modelled, the 3.1% in the abstract. The shading classes, weighted by 118, 58, 30 and 14 systems, land on the same figure.

It separates shade from orientation

The shading section shows shade explains part of the south gap, then notes unshaded south arrays still fell 5.1% short. That stops a reader blaming the trees for a model problem.

Questions people ask

What should a renewable energy poster compare?

A prediction against a measurement, on the same systems and the same period. Here the prediction is the yearly yield a design model gives at quote time and the measurement is a full year of 15 minute meter data. Comparing one site to another tells a reader much less than comparing a promise to what was delivered.

Why does this Australian board treat south facing as the worst case?

Because in the southern hemisphere the sun sits to the north, so a north facing array gets the most direct light over a year and a south facing one the least. The measured yields follow that order: 1,480 kWh per kW north against 1,020 south. The model gap follows it too.

Why plot the gap instead of measured and modelled bars?

Paired bars of 1,480 against 1,510 look almost identical at poster distance, and the eye has to subtract. Plotting the overstatement directly, 2.0% to 6.9%, puts the finding on the axis. The underlying kWh figures are in the abstract and results for anyone who wants them.

What does the completion by arm block mean on a solar poster?

The heading is fixed by this design and cannot be edited from the content. On this board the four bars show how complete the meter record was for each orientation group, between 98.2% and 99.1% of intervals, which tells a reader the yearly totals rest on nearly complete data.

How do I show shading results without a table block?

As a short run of class, count and gap, in reading order. This board writes unshaded 118 systems 1.6%, light 58 systems 3.2%, moderate 30 systems 5.6% and heavy 14 systems 9.3%. Four rows read cleanly in a sentence; eight would not, and would need a table led design.

Will the chart text be readable on a printed A0 board?

At the default sizes, yes. Captions are set at 18 pixels on a 50 DPI working canvas, which prints at roughly 26 point at full A0, and body text at 17 to 22 pixels prints at about 24 to 32 point. Axis labels are smaller, so keep them to short codes like N, E, W and S.

Build your own in about a minute

The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.

Make one like this

Other poster examples

Biomedical engineering research poster example

A biomedical engineering research poster for a device test is an agreement study: how close is the new sensor to the reference, and where does it fail. This A0 landscape board tests a wearable heart rate patch in 40 volunteers, 72,000 paired readings, with a hero line chart of error by activity and the patch within 5 bpm on 96% of exercise readings.

Software engineering research poster example

A software engineering research poster has to make a mined dataset readable from two metres away. This A0 landscape board does it for code review: 41,000 pull requests in 14 open source repositories, with the median wait for a first review falling from 31 hours to 22 over three years, shown as a four point timeline across the top.

Urban planning research poster example

A walkability poster that reports a suburb score is reporting at the level where nothing can be fixed. This board works at the segment instead: 180 block faces across six Australian suburbs, scored on 22 items by two auditors, with 76 segments named as lacking a continuous footpath.

Want the steps in the builder? Read Add charts and diagrams, then choose the template, theme and size. For everything this generator can do, see the poster maker.

Sources

Written and checked by the OneCraft team. Last checked .