Solar's contribution to peak demand

The release of a strategic priorities report by the energy rule maker, the Australian Energy Market Commission, illustrated that there are plenty of people concerned that the rise of solar power will come at their expense. The report states:

    The recent increase in rooftop solar PV was one of the most frequently raised topics in stakeholder workshops and submissions.

And:

    Stakeholders are concerned that network costs of consumers with solar PV are cross-subsidised by other consumers, due to current inefficiencies in network tariffs.

The risk in all of this is that the AEMC or other government bodies are tempted to address these stakeholder concerns by rushing a stop-gap measure just for solar PV owners without properly considering some of the more fundamental complexities at work.

For example, a quick and dirty measure would be to charge owners of solar PV systems a special fixed charge for use of the network, irrespective of their circumstances. This would help ensure that those who can’t afford, aren’t able or can’t be bothered with reducing their energy consumption aren’t lumped with the vast majority of the sunk investments in the network, while those using solar systems escape these costs.  

Such a fixed charge would be very easy and therefore tempting to implement – no need for a widespread rollout of smart meters and no need to make fundamental reforms to the smeared average electricity prices charged to other electricity consumers.

In these charts of the week we can see how such a simplistic approach may miss some important complexity.

The first chart, below, represents the predominant view that solar PV generation makes a big impact on demand over the middle of the day but fails to reduce the amount of network capacity required, which peaks in the evening for residential consumers. The chart represents how the shape of electricity demand on an individual section of the Energex network with very high installations of solar PV has changed over the last few years (using the date of Tuesday, October 2, except for 2013 where it used September 9). As the chart shows, peak demand, which occurs at around 6pm, has remained pretty much the same even though demand has almost disappeared around the middle of the day.

Source: Mike Swanston of Energex – Presentation given to the All Energy Conference, Melbourne, October 2013.

However, one thing you need to keep in mind is that this is a snapshot in time for spring (not the time of year when demand reaches its highest levels) for one single section of one distributor’s network.

Ideally, the public should be able to readily access demand time profiles for every section of every distribution network so we could sort out where the generation profile of solar made a better match with demand. But alas, such data is not made readily available.

Still, some other data is available which suggests in other cases solar generation can coincide with peak demand on some sections of the distribution network.

This chart below, from Victorian distributor SP Ausnet, models how they estimate the demand profile would change on one of their particular feeders without solar PV, in the light blue line, versus 30 per cent penetration of solar PV where the panels faced north (red line), or where they faced west (purple). You can see that solar PV does makes a noticeable contribution to reducing peak demand, particularly if the panels face west. In fact, SP Ausnet found that panels facing west would make a 100 per cent improvement in the level of peak demand reduction versus north-facing panels.

Change in grid demand profile without solar and with 30 per cent solar facing north or west

Source: SP Ausnet and Oakley Greenwood – Improving energy forecasting at SP Ausnet

What this illustrates is that circumstances for how solar PV affects network infrastructure costs can be quite different. 

Because Victoria has daylight saving, and its lower latitude means the sun sets later in the day during summer than in Queensland, solar PV is better able to shave off residential demand peaks during the early evening than what happens in Queensland. This chart also helps to show that if you alter the orientation of solar PV panels you can quite markedly improve its impact on evening residential peaks.

NSW is also quite different to Queensland, it seems, from what little data we have publicly available. The table below is taken from a report by the network distributor Ausgrid. In that report Ausgrid said solar made little difference to network peaks. But when you delve into the data you find that a reasonable proportion of the solar PV capacity installed on the feeders for which they provided data was generating power during the major summer peak periods. For Pennant Hills, as an example, 24.8 to 40.6 per cent of the solar capacity was contributing power. In Avoca when the peak occurred around 6pm, solar’s output was around 10 to 20 per cent of rated capacity. But on the day when demand was highest, which was February 5, the peak occurred at 4pm and over half of the solar-installed capacity was producing power. Blanket assertions that solar can’t make a contribution to reducing a customer’s load on the network appear incredibly simplistic, if not plain wrong.

Proportion of installed solar capacity producing power during peak periods on selected feeders in Ausgrid network.

Source: Ausgrid - Effect of small solar Photovoltaic (PV) systems on network peak demand

Given this complexity, the AEMC and other government authorities should be very careful before imposing blanket charges on households just for owning a solar PV system, while ignoring more fundamental distortions in our electricity market.