Until recently, it was uncertain how much solar power was contributing to electricity demand, because the ‘behind the meter’ nature of net-metered solar power systems means their production simply appears as ‘negative load’. Now by combining some publicly available resources, we can clearly see that PV is having a non-trivial midday impact, and in some circumstances representing more than one quarter of power consumption.
Through work contributing to the APVA’s Development of an Australian Solar Map”, over a six month period SunWiz recorded an aggregated feed from thousands of real-time monitored PV systems (via PVoutput.org), and multiplied it by the CER-reported volume of solar capacity at the time, and then divided it by the AEMO-published electricity consumption volumes in the National Electricity Market.
SunWiz assisted the APVA with the ARENA-Funded Project “Development of an Australian Solar Map”, which presents a live view of similar information – this ARENA-funded website produced by the APVA provides live information on Australia’s solar contribution, alongside noteworthy historical events.
This graph shows the proportion of NEM consumption in each state that was met by solar in each hour of the day across the period February-September 2013. The blue line shows the maximum contribution and the orange line shows the minimum contribution. You can see that contribution of solar power in SA reaches over 28% at some points in the sample period, and there is a healthy average of 11% midday contribution. Queensland’s solar power systems are contributing an average of 7% of NEM demand, an reaching up to 12% on some days. Contribution is lower in Victoria and New South Wales, and lower still in South Australia.
It is worth noting that there are inaccuracies that derive from the methodology: state-wide live performance data from PVoutput.org was captured and then multiplied by the amount of solar capacity installed at the time, then divided by the sum of NEM consumption and solar output to reach the proportion of the NEM produced by solar. As a result there are some obvious data anomalies due to errant system recordings after dark in PVoutput.org. Its worth noting that PVoutput.org users produce higher than average performance, and that solar production includes all systems, whether NEM connected or not. Balancing these factors is the fact that capacity data lags installed capacity.
Another graph, shows all data points for each state over the sample period (to see the graph of monthly solar energy contribution for each state from February – September 2014, subscribe to SunWiz’s special mailout here). In this graph, SA and Queensland are shown to have the largest solar contributions. Interestingly, Queensland has a period of reliably-high solar contribution over the months of August and September. After dipping through the winter months, it is clear that contribution is increasing in most states; a function of the weather and of continued installations adding further power generation capacity.
In terms of energy contribution (as opposed to midday power output), solar power contributed 1-2% of NEM energy requirements each month, but this varies considerably by state. Another graph (again, viewable only by subscribing SunWiz’s special mailout here), shows the monthly contribution of solar power to NEM energy consumption in each state, with almost 5% of South Australia’s NEM energy consumption being provided for by solar power.
Some interesting characteristics of distributed energy are also revealed by the data. Because not every system faces the same direction, and because shade can block some parts of a region while leaving others sunny, the solar power output across a state is rarely perfectly aligned.
A further reduction from the nameplate power results from the quantum physics of solar panels, which operate less efficiently at hotter temperatures, and thus invariably produce less peak “in-field” power than their rated “laboratory” capacity.
The graph below shows the characteristics of combined solar power output for the monitored systems in each state, expressed in terms of the combined output divided by the combined nameplate capacity – this is considered the solar production as a percentage of the rated output. The upper line is the maximum solar output across the sample period; the junctions between each shaded brown areas represent the 75th-percentile, median, and 25th percentile of combined solar output.
From this we can see that Tasmania has the greatest coincidence of solar power operation, achieving over 80% of nameplate capacity at one point, most likely influenced by its small geographic size. Queensland and NSW, though less likely to produce highly-coincident solar output from their distributed systems, are more likely than other states to produce in excess of 50% of rated capacity at midday. This is likely due to the geographic distribution of systems – it is unlikely to be sunny across the whole large state at once, but also unlikely to be cloudy across the entire state. In Queensland over the sample period, there was a 75% probability that midday solar power output exceeded 40% of the nameplate capacity.
Applying this information to September’s Solar Sunday, we can see that Queensland Victoria and South Australia had coincident days of maximum solar contributions to NEM consumption (28%, 7% and 12% respectively), and NSW was close to its best. At least part of the reason for the extraordinary solar contribution to NEM consumption was that consumption is low on Sundays. It certainly helped that each state other than Tasmania had coincident upper-quartile solar production – in other words it was sunny across the whole NEM at once.
This SunWiz Insights data is republished here with permission, to celebrate Australia’s 3GW milestone. To obtain SunWiz Insights on a monthly basis, subscribe here.