100% renewable electricity at “low cost”: the more costs are not accounted for, the “cheaper” it will get

It is not unusual in alternative energy communication to ignore or minimize its negative sides. This is no different in the Conversation article on solar PV and wind being on track of replacing fossil-fuels within two decades as discussed in the last two posts. Halfway the article there is the only admission that there might be a negative side to solar PV and wind energy:

A renewable grid
PV and wind are often described as “intermittent” energy sources.

When I read this the first time, I had high hopes that real issues would be tackled. That hope was in vein, it was followed by this sentence in full cheering mode:

But stabilising the grid is relatively straightforward, with the help of storage and high-voltage interconnectors to smooth out local weather effects.

Relatively straightforward?!?!

Okay, I can understand that if a country has water power (or can get it from a neighboring country), then yes, I can understand that it is relatively straighfoward to stabilize the grid this way (for example Denmark that can rely on the water power of its big neighbors Norway and Sweden). But the reluctance of for example guide country Germany to build such storage, shows me that there are some issues involved that the authors are not telling us about.

The link goes to a scientific paper by both authors, titled “100% renewable electricity in Australia”. However, the first sentence of the article promises that it is as (un)balanced as the Conversation article itself:


  • PV and wind allow Australia to reach 100% renewable electricity rapidly at low cost

Rapidly at low cost?!?!

That is very optimistic. There are several ways to stabilize a grid and they all have considerable technical and financial challenges. The strategy used by the authors seems to be to overproduce (and curtail renewable power when production is high and consumption is low), combined with building more interconnectors (to bring electricity from where it is abundant to where there is a shortage). Sure, but that would mean building many times the needed capacity and would require strengthening the grid (if there is a shortage in one part of the grid and a surplus of wind in another part, it is necessary to handle the peak production there). That will not be cheap by any means. So how could transition be “rapidly at low cost”?

Something that caught my eye was the assumption of the needed annual consumption. They claim the consumption of electricity is 205 TWh. They also claim that consumption is stable since 2008 and therefor assuming that consumption will be constant in the future too.

205 TWh per year seems rather low. If I look at the BP numbers, electricity consumption in 2016 was 256.9 TWh. In their list of assumptions they stated that they were modeling the Australian National Electricity Market (NEM), which “exclude the much smaller systems that exist in Western Australia, the Northern Territory and remote regions in other states”. That is difference of almost 52 TWh or a tad more than 20% of total electricity consumption of Australia. They are not studying 100% electricity of Australia, but 100% electricity of 80% of Australia. My guess is that they assume that the excluded areas show the same pattern as the other 80%.

Next question: is that consumption really “constant”? If I look at the BP electricity data of the last 10 years for Australia, then I get this:

It indeed looks like there is a leveling off at the end. If I look at the last ten years, then it shows a pause since about 2009:

Is a bit ambiguous though. With some fantasy, yes, the trend is leveling off. But that would depend on the chosen starting point. If we take the last 7 or maybe even 8 years, then consumption could well have been stabilized. If we take the last 3 to maybe even 5 years, then it is increasing. So, it will depend on why electricity consumption dropped suddenly in 2011, why it again increased in 2014 and how stable this increase will be. If the NEM data follows the same pattern as the Australia data, then this leveling off is not exactly clear-cut.

But even if the generation of wind and solar was constant since 2008-2009, then that doesn’t necessarily mean that it will also be the case for the next two decades. Just remember the claim by the same authors in the Conversation article that transition requires electrification of the whole energy sector of the economy, meaning electricity generation has to triple over the next two decades. In that case, electricity consumption would be 615 TWh from which 410 TWh will have to be provided by something else than solar and wind… If their own assumption published in the Conversation article is correct, then 100% renewable electricity will not be reached within two decades.

Not even close.

When I got to point 2.4. (Local generation and demand management), I got a glimpse of how the transition could be done at “low cost”. It said that part of the capacity are small scale systems, usually on urban rooftops. The total capacity of those roof mounted systems is assumed to be 17.3 GW and in the simulation of the authors this would yield 23 TWh of annual generation in two decades. Which is about 11% of their assumed annual electricity consumption of 205 TWh.

Nothing could prepare me for this (my emphasis):

The cost of these systems is absorbed by the building owners, and does not directly affect calculated electricity costs under this model.

Hey, that is a neat trick! 11% of the cost of the annual production that just went poof.

Even assuming that the investments made by those private households should not be included when calculating the cost of the transition, there are other costs involved. Australia had (maybe even still has) feed-in tariffs for residential solar panels. That is a real cost and someone has paid for it (or is still paying for it). Yet it is not accounted for in the paper.

What about the cost of the curtailment? By the way, whose production will be curtailed? Those investors will get less return, so their investment will take longer to pay back. Making investors more reluctant to make such investments in the future. Unless a compensation is paid, but then again, it should be in the calculation. It then is a direct cost that originates from the way the transition is achieved.

In the next paragraphs, another neat trick is employed. The NEM has a reliability standard of no more than 0.002% of unmet load (4 GWh per year) without demand management. Most of the scenarios of the paper meet that standard, but in some scenarios these standards are, ahem, “relaxed” during critical periods (for example cold wet windless weeks in winter that occur once every few years). During these periods the pumped storage reservoirs run down to zero over a few days because there is insufficient wind and PV generation to recharge them, leading to a shortfall in supply. The authors then state that the amount of PV, wind and pumped hydro storage could be increased to cover this shortfall, but this would give rise to a substantial extra investment only for a few days every few years (my emphasis):

In some scenarios, demand management during critical periods is modelled by relaxing the NEM reliability standard. For example, the allowable unmet load might be increased to 336 GWh per 5 year period through contractually agreed load shedding arrangements. In most years demand would be fully met, but every few years this additional shortfall allowance would be utilised. Modern techniques allow cloudy and windless periods to be forecast (thus providing ample warning), and the PHES storages represent a substantial buffer. A portion of the savings in investment in PV, wind and PHES would be available to compensate certain consumers for partial loss of supply for a few days every few years. For example, reducing the overall cost of electricity supply by $2/MWh by allowing an unmet load of 336 GWh per 5 years would save $2 billion per 5 years, which is equivalent to $6000 per unmet MWh.

Well, of course it is! They discovered the whole issue with intermittent energy sources. If they really have to replace fossil-fuels for 100%, then not only an overcapacity has to be build, but a over-the-top overcapacity when one want to account for all possible, even rare, moments when there isn’t enough sun, wind and/or storage. Okay, not every scenario has the same relaxed standards, but is that nuance made clear in the communication of their results?

This were only the few weird assumptions that stood out at first read. Who knows how many other assumptions hiding real costs went into this paper? If not all the costs are included into their calculations, then that would undoubtedly result in a cheaper transition than it is in reality.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s