Time for variable pricing for feeding into the grid

February 12th, 2018, Published in Articles: Energize

For some time there have been proposals to charge customers variable prices based on when energy is consumed. Such schemes, which come in many forms and flavours including time-of-use pricing (TOU), real-time pricing (RTP), dynamic pricing and critical peak pricing (CPP) were originally intended to manage peak demand by encouraging consumers to shift some of their usage away from peak-demand hours.

While that is still a good reason to introduce them, even more pressing reasons have emerged in recent years notably the rapid rise of variable renewable generation which literally overwhelms many a network when wind and solar generation peaks and there is little demand resulting in negative prices.

Solar generation, now among the cheapest forms of generation nearly everywhere, has become particularly troublesome in places where it is a significant part of the generation mix, certainly on sunny parts of the day. Although large-scale generators face variable prices on the wholesale market, small-scale distributed solar is typically fed into the low-voltage local distribution network and receives a fixed credit regardless of the time it is generated.

This may not sound like a big deal, but it is in places like Hawaii, California or Arizona where distributed solar PVs may be heavily concentrated in certain parts of the network – say certain suburbs where nearly all customers live in detached homes with big roofs. Just as a high concentration of electric vehicles (EVs) all charging their batteries at the same time exerts extra demand on the distribution network, the same happens when many distributed solar generators feed into the network.

Fig. 1: Distribution of negative prices March, April and May 2012 to 2017 (CAISO).

With so many consumers becoming prosumers – which refers to those who produce and feed net kWh into the network at certain times, say sunny hours of the day, why not apply the same to electrons feeding the network just as they would to those withdrawn from it?

The same logic applies, even more so to kWh injected since there are increasing numbers of prosumers causing havoc to the distribution network with so much fed into the network, a problem exacerbated by the continuous drop in the cost of rooftop solar PVs, on the one hand, and rising retail tariffs, on the other.

The problem has already become acute in certain parts of the network in places like Hawaii where prosumers tend to be heavily concentrated in some neighbourhoods. In these areas, the distribution network can literally get flooded with solar-generated power during sunny hours of the day, frequently more than the capacity of the network to handle it.

Moreover, there is a strong correlation between the excessive mid-day distributed and utility-scale solar generation since both peak during the sunniest hours of the day. This increasingly depresses mid-day prices in wholesale electricity markets, occasionally sending prices to negative territory.

Fig. 2: The duck curve now appears on networks everywhere (CAISO).

Both California and Texas are experiencing increased frequency of negative prices which can be directly attributed to the rise of solar and wind, respectively.

In this context, a recent proposal by the Victorian energy and economic regulator, the Essential Services Commission (ESC) to introduce a peak period tariff of
29(US)c/kWh for rooftop solar exports back into the grid, is timely, innovative and ground-breaking. If enacted, it could help change the way prosumers think about investing in distributed solar generation while encouraging battery storage.

In mid-December 2017, the ESC proposed a scheme that recognizes the value of distributed generation fed into the grid at times of peak demand, while discounting it at times when such generation is not particularly valuable or needed – including times when the network is already saturated with solar or wind generation. The ESC’s proposal also includes a value of 2,5c/kWh for social cost of carbon (SCC) following an ordinance issued by the Victorian government, effectively putting a $25/MWh price on carbon, which was included in the tariff from last year, in another Australian first.

The social cost of carbon, which used to be pegged at $49/ton under the Obama Administration has magically been cut to around $1 to $6/t under the Trump administration, whose views on climate change and carbon emissions widely diverge from those of his predecessor.

Under ESC’s proposal, distributed solar fed into the network between 15h00 and 21h00 during weekdays can fetch 29c/kWh while its value would fall to 10,3c/kWh from 07h00 and 15h00, and 7,2c/kWh from 20h00 and 07h00. The ESC envisions the “time-of-use” solar tariff to become compulsory following a transition period during which retailers can offer a flat FiT (feed-in-tariff) of 9,9c/kWh, below the current 11,3c/kWh tariff.

Fig. 3a: Variable generation solar power (CAISO).

The proposal is in response to the continued adoption of rooftop solar PVs at record rates while Victoria, Queensland and South Australia plan to add significant amounts of new utility-scale solar over the next few years. Australia is unique among many countries in the fact that thus far its solar installations have been mostly if not entirely of the distributed variety – but this is changing as more utility-scale solar is under construction and/or planned.

ESC’s proposal draws on by modelling of wholesale prices by ACIL Allen, a consulting firm, which showed dramatic drop in prices during the mid-day hours due to the additional solar capacity while resulting in significantly higher prices in the evening hours after the sun sets. ESC believes that the proposed variable FiTs will encourage consumers to shift demand, while various commentators have noted its potential to encourage investment in storage.

Commenting on the proposed variable FiT, James Clinch, manager of regulatory reform at the ESC, said a “critical peak” price for solar exports was also considered, but subsequently dropped due to uncertainty over the impact of the changing wholesale price profile impact on retailers’ contracting practices. Depending on their design, critical peak tariffs could provide rewards above
29c/kWh in periods when wholesale prices soar, such as times when they hit $14 000/MWh – allowed under Australia’s energy-only National Electricity Market (NEM).

Clinch noted that the projected growth of distributed and large-scale solar will – for the first time – result in the weighted average wholesale price when solar generation is at its peak to fall below average wholesale price for 2018-19 period. The former is expected to be 6,8c/kWh; the latter 9,1c/kWh.

According to Clinch, analysis conducted through recent inquiries shows that one outcome is that customers will have an incentive to export more at peak periods and that has overall efficiency benefits “Essentially, they are making more efficient use of their solar assets and supplying energy at time of high demand”, he says.

Fig. 3b: Variable generation wind power (CAISO).

Prosumers with solar panels can shift load, orient their panels to the west – rather than the north to produce more later in the day when it is more valuable – or install battery storage to shift more of the generation into the high premium hours. The variable FiT scheme would certainly boost the value proposition for investing in storage.

Commenting on the ESC’s proposal, Bruce Mountain, the director of carbon and energy markets (CME), points out that the Australian solar uptake – unlike that of the US – is not associated with income. In fact, the data shows that Australians mostly go solar because it is so much cheaper than grid supply.

Mountain is also skeptical of arguments that solar will drive higher network costs, noting that distribution feeders are invariably capable of handling much higher distributed rooftop solar power injections than they now face and that localised voltage issues attributable to distributed solar can easily be dealt-with. Moreover, Mountain maintains that distributed solar is only likely to have a meaningful impact on wholesale prices in Victoria many years in the future.

While Mountain recognises the attractiveness to economists and regulators of consumption and production prices that have greater temporal differentiation he urges that careful thought be given to the complexity that arises from temporally differentiated production and consumption prices. Retail markets, like other markets, work well when customers are able to easily identify the offers that best suit them. Complexity has a cost and retail pricing policies must account for this in considering arrangements that are in customers’ interests.

Complexity certainly matters, and this is a factor the Commission is considering as part of the transition to time-varying feed-in tariffs, according to Clinch.

The thrust of the Victorian proposal makes perfect sense and its timing is spot on. With the rapid growth of mid-day solar generation from both distributed and utility-scale installations depressing mid-day wholesale market prices it is becoming clear that solar generation is not particularly needed or valuable during such episodes.

Regulators around the world should adopt similar regulations – not just for electrons consumed but increasingly for those injected into the network. The time of flat rates in either direction is over.


This article was first published in the February 2018 edition of EEnergy Informer, and is republished here with permission.

Send your comments to energize@ee.co.za