3 ms·

I used to be a reactor operator and never saw any evidence they were hard to throttle. That's why they are used for warship propulsion: loads in battle are not

I used to be a reactor operator and never saw any evidence they were hard to throttle. That's why they are used for warship propulsion: loads in battle are not predictable by their nature, what with changes in course, catapult launches, etc.

hwillis 8y ago
You're both right- technically, nuclear reactors are on of the fastest-ramping sources available. A reactor in SCRAM can drop down to <10% of full output in minutes, and many reactors are capable of running at continuously at reduced power. However, zero reactors are capable of doing so economically. The cost of a nuclear plant is in its construction, upkeep, and staff. Those costs are the same at 110% and 30% output, but revenue isnt. In order to turn a profit a nuclear plant needs to be running at 80%+, 24/7. Ultimately that means nuclear needs batteries almost as much as solar and wind, since only ~30% of grid energy is constant 24/7. This should be no surprise- all three technologies have nearly no reliance on fuel, and are funded pretty much upfront. If you have no marginal costs then your only rational option is 100% output.
- mdavis6890 8y ago
  There is a flaw in your reasoning: By definition if a plant needs to be throttled then it's surplus power generation has negative (or zero) utility/value and therefore does not contribute to it's profitability whether it's running or not. Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it.
  - hwillis 8y ago
    I'm not making an argument from semantics- whether or not you approve of the above explanations, they are the reason nuclear power is commonly said to be incapable of load-following. If you want a detailed and reputable investigation, the OECD has one[1]. The inability to load-follow is not a design flaw; any reactor built in the last ~40 years can ramp an order of magnitude faster than coal or gas-steam, and will run very happily at anywhere from 50-100% power, ramping up and down multiple times per day. Many can ramp at hundreds of megawatts per minutes. In fact all of them can ramp from 100%-10% output in about a minute, since that's legally required for reactor SCRAM. That's as fast as an average gas turbine[2] and at the high end it's faster than the fastest gas turbine. Nuclear plants are the most agile nonrenewable plants on the grid but the fact is that a nuclear plant operating <80% output does not generate a return. > Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it. I get what you're saying, but it's wrong. The grid is not a free market and excess generation is not legal. Electricity markets are based around contracts and producing too much is just as heavily penalized as producing too little. If you generate so much excess that you bring grid frequency or voltage out of spec, people start going to jail. To rephrase: If a nuclear plant throttles, its costs are constant and its revenue is lower. If it doesn't throttle, its costs are higher and its revenue is lower. Nuclear sees rising revenue/kWh by throttling, but not enough. Because of all that nuclear is not profitable in short term contracts. Generally they aren't used for ancillary services either, but that's more due to conventions- for instance they are used for frequency control in France. [1]: https://www.oecd-nea.org/ndd/reports/2011/load-following-npp.pdf https://www.oecd-nea.org/ndd/reports/2011/load-following-npp... [2]: https://www.wartsila.com/energy/learning-center/technical-comparisons/combustion-engine-vs-gas-turbine-ramp-rate https://www.wartsila.com/energy/learning-center/technical-co...
- wbl 8y ago
  Nuclear plants have high capital costs and low marginal costs for each bit of energy they make. That doesn't mean they can't produce all the power: France is at 80% from nuclear.
  - hwillis 8y ago
    France reduces their penalty from load following by being the largest power exporter in Europe. Most grids drop down to ~30% of peak demand at night, while France only drops to around 75%[1]- almost exactly how much nuclear generation it has, because those nuclear plants are running at 100% as much as possible and exporting power. In 2015 nuclear made up 72% of their generation but only 40% of consumption- 40% of their consumption results in such an oversupply of nuclear that they have to export that much. If anyone else in Europe tried this, or if the US tried it, it would be prohibitively expensive. Even in France, it's still very expensive- the EDF is 36 billion euro in debt[2]. French nuclear gets 9 cents per kWh in operating funds, slightly lower than the US- but the US is MUCH larger, and distribution costs are included in that. And again, the EDF is heavily in debt. Nuclear doesn't readily beat natural gas on price, and the areas that it's better than renewables are rapidly shrinking. End of the day we'll need ~40% nuclear in the US, but that's not much growth compared to how much renewables need to grow. [1]: http://clients.rte-france.com/lang/an/visiteurs/vie/courbes.jsp http://clients.rte-france.com/lang/an/visiteurs/vie/courbes.... [2]: https://www.edf.fr/en/the-edf-group/dedicated-sections/finance/investors-analysts/credits/net-financial-debt-and-cash-flow https://www.edf.fr/en/the-edf-group/dedicated-sections/finan...
- FlyMoreRockets 8y ago
  > The cost of a nuclear plant is in its construction, upkeep, and staff. Those costs are the same at 110% and 30% output, but revenue isnt. In order to turn a profit a nuclear plant needs to be running at 80%+, 24/7. The same can be said about wind and solar power.
  - hwillis 8y ago
    ...and I literally did, in the next paragraph.
wbl 8y ago
I know nothing about the subject, but my understanding was that iodine pits could be a problem for thermal spectrum reactors if they had to suddenly raise power output after dropping it.
Retric 8y ago
Let’s be more specific. Compare to hydro which can often do 0-100% or 100%-0 in minutes and has little opportunity cost for doing so. It’s relatively cheap to add max capacity even if you can’t increase average capacity much etc. On the other hand increasing nuclear max capacity is as expensive as increasing maximum sustainable capacity. Nuclear also has issues going from sub 5% and ramping to 100% with little notice. Thermal stress is an issue ramping up and down frequently over the life of the system. 2x cycles per day is possible but would adds noticeable lifetime costs, 30x per day would cause huge issues. Which means you need to look at ramping as an expense each time you do so which causes incentives to avoid doing so. PS: Civilian and military reactors are also significantly different designs with different goals. Military reactors are more expensive per GW of capacity, making them a poor though relevant comparison.

hwillis 8y ago

You're both right- technically, nuclear reactors are on of the fastest-ramping sources available. A reactor in SCRAM can drop down to <10% of full output in minutes, and many reactors are capable of running at continuously at reduced power. However, zero reactors are capable of doing so economically. The cost of a nuclear plant is in its construction, upkeep, and staff. Those costs are the same at 110% and 30% output, but revenue isnt. In order to turn a profit a nuclear plant needs to be running at 80%+, 24/7. Ultimately that means nuclear needs batteries almost as much as solar and wind, since only ~30% of grid energy is constant 24/7. This should be no surprise- all three technologies have nearly no reliance on fuel, and are funded pretty much upfront. If you have no marginal costs then your only rational option is 100% output.

mdavis6890 8y ago
There is a flaw in your reasoning: By definition if a plant needs to be throttled then it's surplus power generation has negative (or zero) utility/value and therefore does not contribute to it's profitability whether it's running or not. Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it.
- hwillis 8y ago
  I'm not making an argument from semantics- whether or not you approve of the above explanations, they are the reason nuclear power is commonly said to be incapable of load-following. If you want a detailed and reputable investigation, the OECD has one[1]. The inability to load-follow is not a design flaw; any reactor built in the last ~40 years can ramp an order of magnitude faster than coal or gas-steam, and will run very happily at anywhere from 50-100% power, ramping up and down multiple times per day. Many can ramp at hundreds of megawatts per minutes. In fact all of them can ramp from 100%-10% output in about a minute, since that's legally required for reactor SCRAM. That's as fast as an average gas turbine[2] and at the high end it's faster than the fastest gas turbine. Nuclear plants are the most agile nonrenewable plants on the grid but the fact is that a nuclear plant operating <80% output does not generate a return. > Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it. I get what you're saying, but it's wrong. The grid is not a free market and excess generation is not legal. Electricity markets are based around contracts and producing too much is just as heavily penalized as producing too little. If you generate so much excess that you bring grid frequency or voltage out of spec, people start going to jail. To rephrase: If a nuclear plant throttles, its costs are constant and its revenue is lower. If it doesn't throttle, its costs are higher and its revenue is lower. Nuclear sees rising revenue/kWh by throttling, but not enough. Because of all that nuclear is not profitable in short term contracts. Generally they aren't used for ancillary services either, but that's more due to conventions- for instance they are used for frequency control in France. [1]: https://www.oecd-nea.org/ndd/reports/2011/load-following-npp.pdf https://www.oecd-nea.org/ndd/reports/2011/load-following-npp... [2]: https://www.wartsila.com/energy/learning-center/technical-comparisons/combustion-engine-vs-gas-turbine-ramp-rate https://www.wartsila.com/energy/learning-center/technical-co...
wbl 8y ago
Nuclear plants have high capital costs and low marginal costs for each bit of energy they make. That doesn't mean they can't produce all the power: France is at 80% from nuclear.
- hwillis 8y ago
  France reduces their penalty from load following by being the largest power exporter in Europe. Most grids drop down to ~30% of peak demand at night, while France only drops to around 75%[1]- almost exactly how much nuclear generation it has, because those nuclear plants are running at 100% as much as possible and exporting power. In 2015 nuclear made up 72% of their generation but only 40% of consumption- 40% of their consumption results in such an oversupply of nuclear that they have to export that much. If anyone else in Europe tried this, or if the US tried it, it would be prohibitively expensive. Even in France, it's still very expensive- the EDF is 36 billion euro in debt[2]. French nuclear gets 9 cents per kWh in operating funds, slightly lower than the US- but the US is MUCH larger, and distribution costs are included in that. And again, the EDF is heavily in debt. Nuclear doesn't readily beat natural gas on price, and the areas that it's better than renewables are rapidly shrinking. End of the day we'll need ~40% nuclear in the US, but that's not much growth compared to how much renewables need to grow. [1]: http://clients.rte-france.com/lang/an/visiteurs/vie/courbes.jsp http://clients.rte-france.com/lang/an/visiteurs/vie/courbes.... [2]: https://www.edf.fr/en/the-edf-group/dedicated-sections/finance/investors-analysts/credits/net-financial-debt-and-cash-flow https://www.edf.fr/en/the-edf-group/dedicated-sections/finan...
FlyMoreRockets 8y ago
> The cost of a nuclear plant is in its construction, upkeep, and staff. Those costs are the same at 110% and 30% output, but revenue isnt. In order to turn a profit a nuclear plant needs to be running at 80%+, 24/7. The same can be said about wind and solar power.
- hwillis 8y ago
  ...and I literally did, in the next paragraph.

wbl 8y ago

I know nothing about the subject, but my understanding was that iodine pits could be a problem for thermal spectrum reactors if they had to suddenly raise power output after dropping it.

Retric 8y ago

Let’s be more specific. Compare to hydro which can often do 0-100% or 100%-0 in minutes and has little opportunity cost for doing so. It’s relatively cheap to add max capacity even if you can’t increase average capacity much etc. On the other hand increasing nuclear max capacity is as expensive as increasing maximum sustainable capacity. Nuclear also has issues going from sub 5% and ramping to 100% with little notice. Thermal stress is an issue ramping up and down frequently over the life of the system. 2x cycles per day is possible but would adds noticeable lifetime costs, 30x per day would cause huge issues. Which means you need to look at ramping as an expense each time you do so which causes incentives to avoid doing so. PS: Civilian and military reactors are also significantly different designs with different goals. Military reactors are more expensive per GW of capacity, making them a poor though relevant comparison.