BNEF's renewables costs

The chart below shows the average global costs over time for different renewable generation technologies.   Offshore wind is more expensive than onshore, for obvious reasons, but has the advantage that winds are more reliable on water than they are on land.  Interestingly, tracking solar, where the solar panel rotates during the course of the day to face the direction of the sun, which you'd expect to be more expensive than fixed solar, is not.  The extra yield from a "squarer" insolation profile more than compensates for the extra expense of motors to rotate the panels.  Tracking solar is also better than fixed-tilt solar because output jumps to its maximum just after sunrise and lasts until just before sunset, which means output is better attuned to the daily demand profile, especially the morning peak.

So the cheapest global electricity comes from single-axis tracking solar, then onshore wind, then fixed solar, then offshore wind.  The green line shows a simple average of all four types, and you can see how it has fallen steadily over the last 12 years, falling from $256/MWh in H2 2009 to $52/MWh in H1 2021, or by 80%.  The recent uptick in LCOEs is driven by a three-fold jump in polysilicate prices (for solar) and a doubling of steel prices (for wind).   These are both cyclical, and will partly unwind as economic growth slows after the post-pandemic rebound.  The jump in polysilicate prices is particularly interesting, hinting at a massive build out of solar in China and globally.

Just like Lazard (whose data I have been using for a few years now) and IRENA, BNEF shows the same strong downward trends in the cost of renewables.   Lazard estimate the average cost of new coal at $112/MWh in the US, and the marginal/operating cost of coal at $41/MWh, though that will have risen this year because of the jump in the coal price.  This compares with the average for  onshore wind, and tracking solar of $40/MWh.   Lazard's calculation for the marginal cost of gas in the US is $28/MWh, but gas in the US is less than  half the price of gas outside the US.  For example, gas in Europe  has reached US$12.51/MBtu compared with $4.40 in the US.  And the US natural gas price is up 70% over the last year.   

The moral of this tale is obvious, but I'll tell you anyway:  coal is finished.  Because output from gas power stations can be ramped up more rapidly than from coal, to match supply shortfalls from renewables, gas is still "safe" for now.  Until battery prices halve again.

Offshore wind could power the world

A sailing boat passes the Kentish Flats offshore windfarm. Photograph: Gareth Fuller/PA
Source: The Guardian

There are several carbon-free ways to generate electricity: hydro, onshore wind, offshore wind, solar PV (with either fixed or variable tilt), concentrated solar power (CSP), green methane and hydrogen (i.e., methane and hydrogen manufactured using wind and solar), biomass, nuclear fission (too expensive, too polluting) and nuclear fusion (not yet functional, except possibly for this).    The grid of the future will likely use all these methods, except for nuclear fission, because they complement each other.  The wind blows when the sub doesn't shine, offshore wind is less variable and stronger than onshore, CSP can deliver power 24/7, seasonal storage using green methane/hydrogen will cover week-long periods when wind and solar are low. 

From The Guardian:

Erecting wind turbines on the world’s best offshore sites could provide more than enough clean energy to meet global electricity demand, according to a report.

A detailed study of the world’s coastlines has found that offshore windfarms alone could provide more electricity than the world needs – even if they are only built in windy regions in shallow waters near the shore.

Analysis by the International Energy Agency (IEA) revealed that if windfarms were built across all useable sites which are no further than 60km (37 miles) off the coast, and where coastal waters are no deeper than 60 metres, they could generate 36,000 terawatt hours of renewable electricity a year. This would easily meeting the current global demand for electricity of 23,000 terawatt hours.

“Offshore wind currently provides just 0.3% of global power generation, but its potential is vast,” the IEA’s executive director, Fatih Birol, said.

The study predicts offshore wind generation will grow 15-fold to emerge as a $1tn (£780bn) industry in the next 20 years and will prove to be the next great energy revolution.

The IEA said earlier this week that global supplies of renewable electricity were growing faster than expected and could expand by 50% in the next five years, driven by a resurgence in solar energy. Offshore wind power would drive the world’s growth in clean power due to plummeting costs and new technological breakthroughs, including turbines close to the height of the Eiffel Tower and floating installations that can harness wind speeds further from the coast.

The next generation of floating turbines capable of operating further from the shore could generate enough energy to meet the world’s total electricity demand 11 times over in 2040, according to IEA estimates.

The report predicts that the EU’s offshore wind capacity will grow from almost 20 gigawatts today to nearly 130 gigawatts by 2040, and could reach 180 gigawatts with stronger climate commitments.

In China, the growth of offshore wind generation is likely to be even more rapid, the IEA said. Its offshore wind capacity is forecast to grow from 4 gigawatts to 110 gigawatts by 2040 or 170 gigawatts if it adopts tougher climate targets.

Birol said offshore wind would not only contribute to generating clean electricity, but could also offer a major opportunity in the production of hydrogen, which can be used instead of fossil fuel gas for heating and in heavy industry.

That this report comes from the IEA is telling.  It has in the past been far too conservative about the cost declines in renewables and about the rise in the penetration of renewables in the grid.  Good to see an analysis which is less favourable to fossil fuels.

There are no technological or financial impediments to de-carbonising our entire electricity grid.  The constraints now are political.

Rooftop PV reaches grid parity in EU

From PV Magazine:

If all the rooftops across the European Union able to host solar arrays did so, 680 TWh could be generated, providing 24.4% of the political bloc’s current electricity consumption.

That is the chief finding of a paper entitled A high-resolution geospatial assessment of the rooftop solar photovoltaic potential in the European Union, published on the ScienceDirect website.

The authors of the study combined geospatial and statistical data to assess the technical potential of rooftops for solar energy deployment on every building in the EU. The model, which also used machine learning, was used to quantify the total available rooftop surface for PV systems.

The methodology helped the researchers identify EU markets where rooftop PV could generate electricity at a very competitive levelized cost of energy.

“Specific countries such as Germany, France, Italy, Spain stand out in the maps as they host the highest economic potential that translates to more options for advantageous investments,” stated the paper, adding, electricity retail prices of €0.30-0.169/kWh meant rooftop solar could offer electricity savings of 49% in Germany, 44% in Spain, 42% in Italy and 23% in France.

Eastern EU member states such as Bulgaria, Hungary, Romania and Estonia, however, were cited as markets with very low retail electricity prices, of €0.095-0.12/kWh.

The rooftop PV analysis identified nine markets where grid-parity is some way off as a result of cheap grid power and all of them are in Eastern Europe: Romania, Poland, Hungary, Czechia, Slovakia, Croatia, Lithuania, Latvia and Estonia.

By contrast, Portugal was highlighted as a market with very favorable conditions, including high solar radiation, good financing availability and high retail electricity prices of around €0.22/kWh.

[Read more here]

Remember that rooftop solar is more expensive than industrial- or utility-scale solar because of economies of scale, and because utility-scale solar can use variable-tilt/tracking solar panels which increase yield by 20% or so, plus produce a "squarer" output profile than fixed-tilt solar because they shift to face the sun during the course of the day.   On the other hand, the price point is also higher, being offset against the retail not the wholesale cost of electricity. 

I have already looked at how onshore wind could provide ten times Europe's electricity needs, and how offshore wind could power all of NE Europe.  Rooftop PV could provide 25%, and large-scale PV even more.  There'll be no shortage of electricity even in a 100% green Europe. 

As can be seen from the map above, it would prolly make sense to put Europe's solar farms in Spain, North Africa, the south of France, Italy, Greece, Turkey and SE Europe, where solar resources are greater.   The offsetting cost would be the construction of HVDC (high-voltage direct current) interconnectors between these regions and northern Europe.  Rooftop solar doesn't have that problem, because it's located right next to the demand for electricity.

Massive solar + storage farm in Australia

Robertstown, South Australia

From PV Magazine:

A massive solar-plus-storage project with a A$1.17 billion price tag (US$822 million) has been waved through by the South Australian government. The facility will feature 500 MW (AC) of solar PV generation capacity collocated with 250 MW/1,000 MWh of battery storage around five kilometers northeast of Robertstown [About 120 kms NE of Adelaide].

The power station will be built in stages and connected to the Robertstown substation via 275 kV transmission lines. A previous assessment has determined the facility could export energy to the grid without significant restraint but it will potentially incorporate synchronous condensers to support reliability and security of supply.

According to EPS Energy, the Robertstown project is on track to break ground in the middle of next year and generate around 275 jobs during construction and 15 or so full time jobs once operational. When commissioned, the facility will generate enough electricity to power 144,000 homes during its 30-year life.

EPS Energy director Steve McCall said the company hopes to secure finance for the project within months.“We’re working with equity and finance partners right now and that’s looking all very positive,” he said. “We’re also committed to utilizing the regional workforce and local contractors.”

For EPS, the Robertstown project is one of several large scale solar and battery storage schemes in its gigawatt-scale portfolio. The company’s South Australian pipeline includes the Bungama Solar project – a proposed 280 MW generation capacity and battery project near Port Pirie – and the Yoorndoo Ilga Solar project, a 200-400 MW solar capacity and battery facility near Whyalla.

The Robertstown plant is one of two large scale solar and battery plans in the area, along with the Solar River Project which received development approval a year ago. That facility comprises a 200 MW solar generation plant plus 120 MWh of battery storage and is likely to add another 200 MW of solar and a further 150 MWh of battery storage in a second stage if a proposed high-voltage transmission line to Victoria goes ahead.

'250 MW/1,000 MWh' referring to the battery means that the output of the battery bank is a maximum of 250 MW for 4 hours.  Or it could be, half that for 8 hours.  With a capacity factor of 30%, the 500 MW of panels will produce only 150 MW of output, so, if that level of output were maintained into the evening and the night, the battery would supply power for 6 hours and 30 minutes.   This battery is twice the size of the Tesla "big battery" in SA, which when it was installed just two years ago, was the largest battery in the world.

The company says that the solar farm will have tracking solar panels, which "follow the sun":

tracking solar panel systems follow the sun’s movement throughout the day for maximum collection. At the end of the day the panels track back to the east ready for the next operation.

This produces a "squarer" electricity output profile and higher capacity factors compared with a fixed-tilt solar panel (rooftop solar, for example).  In that part of the world (semi-desert inland South Australia), tracking solar will provide pretty much constant from sunup to sunset.  Utilities and the grid operators will love the combination of near-baseload power provide by tracking solar and massive storage.

Australia is galloping towards a 100% green electricity future, despite the denialist right-wing federal coalition government which is firmly wedded to coal, and would quite like some nuclear too, thank you.  Mind you, the state coalition government, after railing against Labor's policy on renewables before it was elected, flipped when it came to power and is now enthusiastic about renewables.

The solar bell curve

With fixed solar panels, for example rooftop panels, the amount of power produced describes a bell, or perhaps better, a sine wave, rising steeply, then more slowly, then achieving a brief peak before falling slowly then faster.  There's a very brief "lip" at dawn and sunset, hence the "bell curve".

Single-axis tracking solar panels follow the sun, so they produce a much "squarer" generation profile.  Compare the two in the chart below (from Australian company, Carnegie Clean Energy):

This has important implications for the usefulness of solar to the grid.  Output rises sharply to its peak and stays there (unless it gets cloudy).  This is very useful for the morning demand ramp.

“The additional generation gained due to the single axis tracking system selected for the project has been evident with the system’s output quickly ramping up in the morning,” Carnegie said in a statement.

It also means that it's easier to project how much storage will be needed to approximate the output from a baseload plant.  Assuming nighttime demand is 2/3rd of daytime demand (as in California, for example), we would need 66% of maximum output for the 12 hours of darkness (on average; it would be different by season).  That means 2/3*12, or 8 hours of storage.  That leaves no safety margin for cloudy days, but if many geographically separated solar farms are attached to the grid, and there are other generation sources in the grid (wind/offshore wind/hydro) then it would be enough.  We will ultimately need more storage as renewables increase their penetration in the grid, but by then, battery costs will also have fallen--they should halve over the next 6 years.  [Update, 3/08/2026:  Battery costs have far more than halved since I wrote this seven years ago]