12 ms·

Lazard’s Levelized Cost of Energy Analysis: Version 16.0 [pdf]

snake_doc 4y ago
Great to see updated figures. Unfortunate to see average LCOE go up for existing renewables due to the double whammy of price level rise (inflation) and cost of capital rise (interest rate). These 2 factors also make it more difficult to commercialize frontier renewables like green hydrogen.
- tuatoru 4y ago
  General phenomena like inflation (construction cost index rising) and interest rate (cost of capital) rises should affect all technologies proportionally. The notable thing was that CCGT costs aren't rising to the same extent, despite broadly similar capital cost to utility scale PV (page 11). And longer construction time (pages 37 PV and 39 CCGT), which affects the risk premium. I recall reading that there was a polysilicon shortage because of changes in China's energy policy. Not sure what's going on with wind. Nor why CCGT costs are continuing low.
  - snake_doc 4y ago
    Yes, though those effects are much less on an incremental basis for any technology already at scale. The mechanisms still exist, but delay is longer.
  - 7952 4y ago
    Cost of capital might vary between types of project. It is becoming common in battery storage to have sites funded by private equity. And non-recourse debt is common in offshore wind. Different sources of capital have different capabilities to assess risk. A battery storage site is easy to understand so can be funded my smaller groups. Offshore wind requires more modelling, due diligence, and technical understanding so needs the resources a bank has. And a nuclear site is too difficult for anyone but government. Also, inflation can vary across industries. For example wind turbines could be vulnerable to high steel costs. A fast project can better predict staff costs than a slow one.
probablypower 4y ago
Always useful data. Important to note that electrical grids and markets are more complicated than "average levelized cost". The uptake of intermittent resources will increase the need for ancillary services, and skew the economics towards generators that are flexible. For example, you can end up in a situation where half the time the price is zero (even negative) and the other half the price is set by your peaking plants. If you own a wind farm, you're likely to be produce most of your energy in low value periods. This effect worsens as you get more renewables. Hydropower owners dominate in this environment, as they can save their water when the prices are low, and sell at peak capacity when it is high. Norway is making a killing on this, for example. There are also limits to the proportion of renewables you can have generating at any moment in a synchronous area. At present it is likely around 50-60% for most grids (see DS3 project at Eirgrid, where they're closer to 70%). This is largely due to grid inertia and ancillary services. The remaining 30% is interesting. At present the best option is natural gas, to act as peaking, with some old conventional plants hanging in just to eek out value before decommissioning. In the future this may be hydrogen plants and nuclear. In the far future it may be fusion.
- snake_doc 4y ago
  You’re describing energy market price (LBMP) dynamics. LCOE isn’t a price, it’s an cost estimate for the lifetime of an “power plant”. LCOE is an input to deciding whether to build a new power plant in a market. The market’s LBMP is another input. Super simplified, if the market’s average LBMP is lower than LCOE for the plant, the plant shouldn’t be built…
- ZeroGravitas 4y ago
  Grid inertia and ancilliary services are moving to inverter based platforms like renewables/batteries as detailed on pages 20 and 21 of this PDF.
- algo_trader 4y ago
  You have discovered "value adjusted LCoE" and "System LCoE" Also South Australia grid (~1GW) has been running at 90% renewable for months, and recently 99% with some new hardware. It is true the last 10% will be very hard, but luckily we have engineers.
roenxi 4y ago
This is one of those examples of modern madness at a grand scale. We are in a general energy crisis and we can see in the graph that the regulators are still committed to making nuclear unaffordable. I'm just going to preempt the argument because it is very normal. Nuclear isn't getting more expensive [0] because we've discovered the first technology that gets harder to work with as time goes one. That is all regulation. The LCOE is working to double in a decade and that is an external influence on the nuclear industry. [0] See page 9 - Levelized Cost of Energy Comparison—Historical Utility-Scale Generation Comparison
- ZeroGravitas 4y ago
  Worth noting that LCOE analysis generally helps nuclear, since the assumption is that once built it will be able to sell all its power, running at near full capacity for its entire 40 year lifespan. Which is a dubious assumption in a world where the main sources of new power (solar and wind) and storage (batteries) are still dropping in price dramatically and are already at the level of the running costs of existing nuclear plants.
  - acidburnNSA 4y ago
    I disagree. LCOE is a favorite way to make nuclear look awful because it massively weights upfront capital cost (high for nuclear) and completely devalues the fact that nuclear plants can run for 60+ years. Take any LCOE spreadsheet and check the sensitivity to revenue 30 years out and you'll see that it basically doesn't matter.
    - Gwypaas 4y ago
      You should learn about the time value of money. Money today is more valuable than in 50 years so you can't simply ignore the high upfront cost because it makes nuclear impossible on any open market. https://www.investopedia.com/articles/03/082703.asp https://www.investopedia.com/articles/03/082703.asp
      - acidburnNSA 4y ago
steeve 4y ago
For the record, the (widely shared) Lazard LCOE graph has been _heavily_ criticised in energy circles. LCOE not being a reliable or useful information in itself. Indeed, it compares apples to oranges, that is intermittent vs non intermittent sources.
- snake_doc 4y ago
  For the record, it’s usually criticized dependent on the individual’s views on intermittent vs. non-intermittent… Just as green hydrogen is heavily criticized by skeptics vs advocates… It’s not criticized broadly by the entire energy industry…
- paganel 4y ago
  Having Roland Berger as co-authors of this study is not very confidence-inducing, either.
- bjourne 4y ago
  Can you please link to some of that criticism? Lazard's numbers have been criticized because they make nuclear power look noncompetitive, but that doesn't mean that the numbers are wrong.
  - Schroedingersat 4y ago
    You see that's the thing though. From the nuclear stan's point of view nuclear is cheapest and best, and anything that contradicts that is wrong by definition.
    - steeve 4y ago
      Thinking tap water and rain water don't have the same value makes one a nuclear stan?
      - Schroedingersat 4y ago
        Telling the same lies over and over again does.

Great to see updated figures. Unfortunate to see average LCOE go up for existing renewables due to the double whammy of price level rise (inflation) and cost of capital rise (interest rate). These 2 factors also make it more difficult to commercialize frontier renewables like green hydrogen.

tuatoru 4y ago
General phenomena like inflation (construction cost index rising) and interest rate (cost of capital) rises should affect all technologies proportionally. The notable thing was that CCGT costs aren't rising to the same extent, despite broadly similar capital cost to utility scale PV (page 11). And longer construction time (pages 37 PV and 39 CCGT), which affects the risk premium. I recall reading that there was a polysilicon shortage because of changes in China's energy policy. Not sure what's going on with wind. Nor why CCGT costs are continuing low.
- snake_doc 4y ago
  Yes, though those effects are much less on an incremental basis for any technology already at scale. The mechanisms still exist, but delay is longer.
- 7952 4y ago
  Cost of capital might vary between types of project. It is becoming common in battery storage to have sites funded by private equity. And non-recourse debt is common in offshore wind. Different sources of capital have different capabilities to assess risk. A battery storage site is easy to understand so can be funded my smaller groups. Offshore wind requires more modelling, due diligence, and technical understanding so needs the resources a bank has. And a nuclear site is too difficult for anyone but government. Also, inflation can vary across industries. For example wind turbines could be vulnerable to high steel costs. A fast project can better predict staff costs than a slow one.

probablypower 4y ago

Always useful data. Important to note that electrical grids and markets are more complicated than "average levelized cost". The uptake of intermittent resources will increase the need for ancillary services, and skew the economics towards generators that are flexible. For example, you can end up in a situation where half the time the price is zero (even negative) and the other half the price is set by your peaking plants. If you own a wind farm, you're likely to be produce most of your energy in low value periods. This effect worsens as you get more renewables. Hydropower owners dominate in this environment, as they can save their water when the prices are low, and sell at peak capacity when it is high. Norway is making a killing on this, for example. There are also limits to the proportion of renewables you can have generating at any moment in a synchronous area. At present it is likely around 50-60% for most grids (see DS3 project at Eirgrid, where they're closer to 70%). This is largely due to grid inertia and ancillary services. The remaining 30% is interesting. At present the best option is natural gas, to act as peaking, with some old conventional plants hanging in just to eek out value before decommissioning. In the future this may be hydrogen plants and nuclear. In the far future it may be fusion.

snake_doc 4y ago
You’re describing energy market price (LBMP) dynamics. LCOE isn’t a price, it’s an cost estimate for the lifetime of an “power plant”. LCOE is an input to deciding whether to build a new power plant in a market. The market’s LBMP is another input. Super simplified, if the market’s average LBMP is lower than LCOE for the plant, the plant shouldn’t be built…
ZeroGravitas 4y ago
Grid inertia and ancilliary services are moving to inverter based platforms like renewables/batteries as detailed on pages 20 and 21 of this PDF.
algo_trader 4y ago
You have discovered "value adjusted LCoE" and "System LCoE" Also South Australia grid (~1GW) has been running at 90% renewable for months, and recently 99% with some new hardware. It is true the last 10% will be very hard, but luckily we have engineers.

roenxi 4y ago

This is one of those examples of modern madness at a grand scale. We are in a general energy crisis and we can see in the graph that the regulators are still committed to making nuclear unaffordable. I'm just going to preempt the argument because it is very normal. Nuclear isn't getting more expensive [0] because we've discovered the first technology that gets harder to work with as time goes one. That is all regulation. The LCOE is working to double in a decade and that is an external influence on the nuclear industry. [0] See page 9 - Levelized Cost of Energy Comparison—Historical Utility-Scale Generation Comparison

ZeroGravitas 4y ago
Worth noting that LCOE analysis generally helps nuclear, since the assumption is that once built it will be able to sell all its power, running at near full capacity for its entire 40 year lifespan. Which is a dubious assumption in a world where the main sources of new power (solar and wind) and storage (batteries) are still dropping in price dramatically and are already at the level of the running costs of existing nuclear plants.
- acidburnNSA 4y ago
  I disagree. LCOE is a favorite way to make nuclear look awful because it massively weights upfront capital cost (high for nuclear) and completely devalues the fact that nuclear plants can run for 60+ years. Take any LCOE spreadsheet and check the sensitivity to revenue 30 years out and you'll see that it basically doesn't matter.
  - Gwypaas 4y ago
    You should learn about the time value of money. Money today is more valuable than in 50 years so you can't simply ignore the high upfront cost because it makes nuclear impossible on any open market. https://www.investopedia.com/articles/03/082703.asp https://www.investopedia.com/articles/03/082703.asp
    - acidburnNSA 4y ago

steeve 4y ago

For the record, the (widely shared) Lazard LCOE graph has been _heavily_ criticised in energy circles. LCOE not being a reliable or useful information in itself. Indeed, it compares apples to oranges, that is intermittent vs non intermittent sources.

snake_doc 4y ago
For the record, it’s usually criticized dependent on the individual’s views on intermittent vs. non-intermittent… Just as green hydrogen is heavily criticized by skeptics vs advocates… It’s not criticized broadly by the entire energy industry…
paganel 4y ago
Having Roland Berger as co-authors of this study is not very confidence-inducing, either.
bjourne 4y ago
Can you please link to some of that criticism? Lazard's numbers have been criticized because they make nuclear power look noncompetitive, but that doesn't mean that the numbers are wrong.
- Schroedingersat 4y ago
  You see that's the thing though. From the nuclear stan's point of view nuclear is cheapest and best, and anything that contradicts that is wrong by definition.
  - steeve 4y ago
    Thinking tap water and rain water don't have the same value makes one a nuclear stan?
    - Schroedingersat 4y ago
      Telling the same lies over and over again does.