7 ms·

A solar power researcher discusses the “duck curve” he helped discover

62 points by curtis 9y ago

ZeroGravitas 9y ago
The headline and intro have much more drama than the interview with the expert.
chis 9y ago
David Roberts When you curtail the coal plant, though, you still have the fuel. You’re not wasting it. Paul Denholm But what fraction of the LCOE [levelized cost of energy] of a mine-mouth coal plant is the fuel? It’s a small fraction. The cost of the plant is the major component of the LCOE, so when I curtail generation from that coal plant, I’m losing the opportunity to recover the cost of investing in that really expensive asset. I never considered that running a coal plant below capacity is the same as dumping solar panel energy, economically speaking.
- Retric 9y ago
  Yea, this is vastly better than most discussions of the topic. Talking about seasonal storage is kind of silly costs are low enough that you can build say 20% more solar and wind than you need on average and they are still cheaper than gas/nuclear/coal. On top of that hydropower is very dispatchable which acts a lot like seasonal energy storage because you are storing months worth of river flow.
  - jaggederest 9y ago
    I'm just thinking of fun ways to build businesses around dispatchable, quick-ramping energy sinks that could serve in lieu of curtailment. Obviously energy intensive businesses like aluminum smelting, chemical production, etc, but also things like refrigeration units running slightly more midday to produce chilled water and then idling until ~10pm, battery backups at datacenters being half way drained and then refilled off-peak, etc.
    - nate_meurer 9y ago
      This is a significant element of DR (demand response). Domestic waters are another attractive target, and there some pilot programs around the U.S. fleshing out the details.
      - nate_meurer 9y ago
        Sorry, that should read "domestic water heaters"
  - Obi_Juan_Kenobi 9y ago
    Hydropower isn't easy, either. The stored water is often used for agriculture, and you have to manage flow rates to keep levies and river systems healthy. My intuition is that the power and agricultural water demand roughly match up, both high in the summer. For levies and fish, you can't just dump or cut off water quickly. Ramping flow up creates turbidity that can harm fish. Cutting it off leaves saturated levies without support and leads to sloughing and failure. I don't know whether these limits would actually matter for an energy management scenario, just pointing out that there are certainly some limits.
    - Retric 9y ago
      The specifics very much depend on the specific area. The northwest US for example has an abundance of water. The southwest on the other hand has to manage water of multiple years and has far less slack in the system. As to flow rates, that's a issue for sudden changes in output over minutes not seasonal shifts in demand. An average flow rate of +/- 10% for a month is a non issue for river ecosystems as long as it's within normal variation, but can represent a lot of stored energy over that month.
keebEz 9y ago
For new technologies, adoption curves can vary wildly, and for different reasons. When it comes to clean tech, PV and wind specifically, the adoption is dominated by sunsetting investments over a long horizon. Other types of tech have much faster adoption cycles due to the lower cost of substitution. For each peak-hour kWh from PV comes at an increased cost of non-peak-hour kWh from non-PV (whether storage or generation).
neverdark 9y ago
For anyone interested in this topic, I’d highly recommend checking out some of the papers by Denholm. His papers are usually quite approachable, but very thorough and well written.
nrook 9y ago
I dunno, this so-called expert sounds like a quack.
eksu 9y ago
They should do dynamic pricing throughout the day and then there would be a market for users to use smart devices to schedule load, when to charge batteries vs when to discharge batteries, etc.
- tantalor 9y ago
  Pump all the water up hill.
  - HankB99 9y ago
    I wonder if that was a tongue in cheek comment or if you knew this is already done. https://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power_Plant https://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power...
    - maxander 9y ago
      My goodness- built in the sixties, stores several GWh, and has 70% efficiency. Enough of those would solve the curtailment problem altogether. Not sure how a technology that relies on the availability of titanic quantities of fresh water is going to play out in California, though. :(
      - jaggederest 9y ago
        It doesn't need to be fresh, you could dam a certain bay and put generator/pumps in the wall. Pump a cubic mile of water back and forth as needed, and you get tidal generation for free, too.
        Zeta_Function 9y ago
        You can do the same thing with compressed gas as well, or really anything. Gas and fluid do make the most sense in terms of directly driving generator turbines though. The only issue with gas is thst you have to deal with thermal effects from expansion and compression.
nthompson 9y ago
Isn't this problem caused by fix axis solar? If all solar panels were on trackers, wouldn't the curve change form and partially mitigate the problem?
- maxander 9y ago
  Look at the x-axis of that chart- the peak nighttime load is happening at something like 8:30pm. Even at times of the year the sun is still up by then, it's not going to get you much no matter how you swivel.
- wtallis 9y ago
  Better tracking of the sun won't significantly increase production at noon, but could help with mid-morning and mid-afternoon production. But that only makes the duck curve steeper, and the existing infrastructure needs a smoother curve.
  - nthompson 9y ago
    I imagine both steeper and flatter. The flatter curve seems to be easier to handle within my naive mental model of how utilities work.
kumarvvr 9y ago
I hope we can find an efficient way to store solar energy. That will solve many problems with solar power generation. Lithium ion batteries can only do so much. On a global scale, they are too small to have a meaningful impact.
- ghouse 9y ago
  They already have a meaningful impact today in the eastern US. In the PJM RTO, batteries (and fly wheels) perform frequency regulation (discussed elsewhere in these comments).
  - dragontamer 9y ago
    Frequency regulation is measured in minutes of storage though, not hours. LiIon / Flywheels are great to store minutes of energy cheaply. In effect, LiIon / Flywheels are hundreds-of-MW-hrs size at best. But unfortunately, the sun disappears for hours. A bigger solution is needed. Pumped hydro works for sure, but even at GW-hr scale (and yes, there are pumped-hydro plants at GW-hr size), its not going to be enough.
Jedd 9y ago
I wonder why solar thermal isn't considered as a partial solution -- in terms of providing power well after sunset, and thereby flattening out that evening peak.
- ghouse 9y ago
  It is considered a partial solution. There are large solar thermal power plants that can store heat in the form of molten salt, and then later using that heat to flash water to steam to spin a turbine. However, this is, in most scenarios today, more expensive than simply spinning up a combustion turbine to provide power well after sunset.
  - stergios 9y ago
    Areva Solar sells thermal power plants where the working fluids are oil and water. A challenge with solar thermal is the lower working temperature and pressure, and the inability to drive traditional size power turbines efficiently.
hyperpallium 9y ago
funfact: the sun is just as bright at dawn and dusk as midday (though, lateral atmosphere blocks more UV-B). The dip in the duck's back is because the angle of incidence changes. Facing solar panels towards the sun throughout the day, to maintain the angle of incidence at the same as midday (i.e. 0°) flattens and lengthens the demand curve. This sun-tracking is important, not to maximize power output, but for this effect on the demand curve, reducing storage and grid need.
brandmeyer 9y ago
> We’re just going to need the right intelligence in the system, sending the right economic signals [i.e., charging varying rates throughout the day, more when power is expensive, less when it is cheaper]. Hell. Yes. Most of the deregulated markets already have a futures market price scheduled in 15 minute intervals that clears 24 hours in advance. Expose end customers to that day-ahead market and we can get all kinds of flexibility.
- sounds 9y ago
  I am cautiously optimistic that varying rate schedules will help. It is unfortunately true that most residential users will not pay enough attention to either curtailing their own use or -- and I think this is the big one -- shenanigans by the electric utility. If the result is an effective rate hike, residential users will oppose it. It's not even that hard to get residential users to love it, but it seems no one is interested in solving that problem.
csense 9y ago
I'm disappointed he didn't say anything about the frequency stabilization issue. A system based mostly around turbines has the inertia of thousands, maybe millions of tons of spinning metal across the entire grid, which buffers enough energy for control systems (for example controlling how much fuel to inject) to respond to sudden demand changes while maintaining stability. If everything's solid-state solar, following demand is harder.
- ghouse 9y ago
  The term of art is "frequency response" for "stabilization." This is the system's ability to respond to shocks. Though, the need for frequency response is driven primarily by very large "traditional" power plants. For example, in WECC, the single largest contingency is the loss of a 4GW large nuclear complex. So, should the cost of maintaining this frequency response be borne by the system, or by that specific contingency? Smaller power plants, and smaller transmission lines have smaller contingencies. Following demand is a different function called frequency regulation. Regulating frequency down (curtailing output) is easy for solid-state inverters (including solar), though regulating up (increasing output) is, as you note, more difficult for generation (but very easy for load).
legohead 9y ago
Do most solar installations not include batteries?
- ghouse 9y ago
  No, most do not. This is true for both utility projects and retail (behind the meter) projects. However, with the falling cost of batteries, and shifting price signals in the market, this is beginning to change -- in Germany, the US, and Australia primarily. Utilities in the US have been changing their rates (tariffs) to decease the cost of energy (kWh), but increase the cost of demand (kW) as a way to (dissuade solar adoption / align cost of service -- depending on your perspective). But within increased demand charges, this permits batteries to offer more value than shifting the time of use of energy.
  - francisofascii 9y ago
    Utilities are also increasing customer changes while keeping kWh charges low.
ghouse 9y ago
For those interested in seeing the California "duck curve" in real time, the CAISO publishes it here: http://www.caiso.com/TodaysOutlook/Pages/default.aspx http://www.caiso.com/TodaysOutlook/Pages/default.aspx
dchichkov 9y ago
Strange. I'm looking at major California provider plans - https://www.pge.com/en_US/residential/rate-plans/how-rates-work/find-my-best-rate-plan.page https://www.pge.com/en_US/residential/rate-plans/how-rates-w... And the peak time (and pricing) is 3 p.m. to 8 p.m. PG&E recommends to "shift usage to the least expensive hours: before 3 p.m. and after 8 p.m. every day".

62 points by curtis 9y ago

The headline and intro have much more drama than the interview with the expert.

chis 9y ago

David Roberts When you curtail the coal plant, though, you still have the fuel. You’re not wasting it. Paul Denholm But what fraction of the LCOE [levelized cost of energy] of a mine-mouth coal plant is the fuel? It’s a small fraction. The cost of the plant is the major component of the LCOE, so when I curtail generation from that coal plant, I’m losing the opportunity to recover the cost of investing in that really expensive asset. I never considered that running a coal plant below capacity is the same as dumping solar panel energy, economically speaking.

Retric 9y ago
Yea, this is vastly better than most discussions of the topic. Talking about seasonal storage is kind of silly costs are low enough that you can build say 20% more solar and wind than you need on average and they are still cheaper than gas/nuclear/coal. On top of that hydropower is very dispatchable which acts a lot like seasonal energy storage because you are storing months worth of river flow.
- jaggederest 9y ago
  I'm just thinking of fun ways to build businesses around dispatchable, quick-ramping energy sinks that could serve in lieu of curtailment. Obviously energy intensive businesses like aluminum smelting, chemical production, etc, but also things like refrigeration units running slightly more midday to produce chilled water and then idling until ~10pm, battery backups at datacenters being half way drained and then refilled off-peak, etc.
  - nate_meurer 9y ago
    This is a significant element of DR (demand response). Domestic waters are another attractive target, and there some pilot programs around the U.S. fleshing out the details.
    - nate_meurer 9y ago
      Sorry, that should read "domestic water heaters"
- Obi_Juan_Kenobi 9y ago
  Hydropower isn't easy, either. The stored water is often used for agriculture, and you have to manage flow rates to keep levies and river systems healthy. My intuition is that the power and agricultural water demand roughly match up, both high in the summer. For levies and fish, you can't just dump or cut off water quickly. Ramping flow up creates turbidity that can harm fish. Cutting it off leaves saturated levies without support and leads to sloughing and failure. I don't know whether these limits would actually matter for an energy management scenario, just pointing out that there are certainly some limits.
  - Retric 9y ago
    The specifics very much depend on the specific area. The northwest US for example has an abundance of water. The southwest on the other hand has to manage water of multiple years and has far less slack in the system. As to flow rates, that's a issue for sudden changes in output over minutes not seasonal shifts in demand. An average flow rate of +/- 10% for a month is a non issue for river ecosystems as long as it's within normal variation, but can represent a lot of stored energy over that month.

keebEz 9y ago

For new technologies, adoption curves can vary wildly, and for different reasons. When it comes to clean tech, PV and wind specifically, the adoption is dominated by sunsetting investments over a long horizon. Other types of tech have much faster adoption cycles due to the lower cost of substitution. For each peak-hour kWh from PV comes at an increased cost of non-peak-hour kWh from non-PV (whether storage or generation).

neverdark 9y ago

For anyone interested in this topic, I’d highly recommend checking out some of the papers by Denholm. His papers are usually quite approachable, but very thorough and well written.

nrook 9y ago

I dunno, this so-called expert sounds like a quack.

eksu 9y ago

They should do dynamic pricing throughout the day and then there would be a market for users to use smart devices to schedule load, when to charge batteries vs when to discharge batteries, etc.

tantalor 9y ago
Pump all the water up hill.
- HankB99 9y ago
  I wonder if that was a tongue in cheek comment or if you knew this is already done. https://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power_Plant https://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power...
  - maxander 9y ago
    My goodness- built in the sixties, stores several GWh, and has 70% efficiency. Enough of those would solve the curtailment problem altogether. Not sure how a technology that relies on the availability of titanic quantities of fresh water is going to play out in California, though. :(
    - jaggederest 9y ago
      It doesn't need to be fresh, you could dam a certain bay and put generator/pumps in the wall. Pump a cubic mile of water back and forth as needed, and you get tidal generation for free, too.
      - Zeta_Function 9y ago
        You can do the same thing with compressed gas as well, or really anything. Gas and fluid do make the most sense in terms of directly driving generator turbines though. The only issue with gas is thst you have to deal with thermal effects from expansion and compression.

nthompson 9y ago

Isn't this problem caused by fix axis solar? If all solar panels were on trackers, wouldn't the curve change form and partially mitigate the problem?

maxander 9y ago
Look at the x-axis of that chart- the peak nighttime load is happening at something like 8:30pm. Even at times of the year the sun is still up by then, it's not going to get you much no matter how you swivel.
wtallis 9y ago
Better tracking of the sun won't significantly increase production at noon, but could help with mid-morning and mid-afternoon production. But that only makes the duck curve steeper, and the existing infrastructure needs a smoother curve.
- nthompson 9y ago
  I imagine both steeper and flatter. The flatter curve seems to be easier to handle within my naive mental model of how utilities work.

kumarvvr 9y ago

I hope we can find an efficient way to store solar energy. That will solve many problems with solar power generation. Lithium ion batteries can only do so much. On a global scale, they are too small to have a meaningful impact.

ghouse 9y ago
They already have a meaningful impact today in the eastern US. In the PJM RTO, batteries (and fly wheels) perform frequency regulation (discussed elsewhere in these comments).
- dragontamer 9y ago
  Frequency regulation is measured in minutes of storage though, not hours. LiIon / Flywheels are great to store minutes of energy cheaply. In effect, LiIon / Flywheels are hundreds-of-MW-hrs size at best. But unfortunately, the sun disappears for hours. A bigger solution is needed. Pumped hydro works for sure, but even at GW-hr scale (and yes, there are pumped-hydro plants at GW-hr size), its not going to be enough.

Jedd 9y ago

I wonder why solar thermal isn't considered as a partial solution -- in terms of providing power well after sunset, and thereby flattening out that evening peak.

ghouse 9y ago
It is considered a partial solution. There are large solar thermal power plants that can store heat in the form of molten salt, and then later using that heat to flash water to steam to spin a turbine. However, this is, in most scenarios today, more expensive than simply spinning up a combustion turbine to provide power well after sunset.
- stergios 9y ago
  Areva Solar sells thermal power plants where the working fluids are oil and water. A challenge with solar thermal is the lower working temperature and pressure, and the inability to drive traditional size power turbines efficiently.

hyperpallium 9y ago

funfact: the sun is just as bright at dawn and dusk as midday (though, lateral atmosphere blocks more UV-B). The dip in the duck's back is because the angle of incidence changes. Facing solar panels towards the sun throughout the day, to maintain the angle of incidence at the same as midday (i.e. 0°) flattens and lengthens the demand curve. This sun-tracking is important, not to maximize power output, but for this effect on the demand curve, reducing storage and grid need.

brandmeyer 9y ago

> We’re just going to need the right intelligence in the system, sending the right economic signals [i.e., charging varying rates throughout the day, more when power is expensive, less when it is cheaper]. Hell. Yes. Most of the deregulated markets already have a futures market price scheduled in 15 minute intervals that clears 24 hours in advance. Expose end customers to that day-ahead market and we can get all kinds of flexibility.

sounds 9y ago
I am cautiously optimistic that varying rate schedules will help. It is unfortunately true that most residential users will not pay enough attention to either curtailing their own use or -- and I think this is the big one -- shenanigans by the electric utility. If the result is an effective rate hike, residential users will oppose it. It's not even that hard to get residential users to love it, but it seems no one is interested in solving that problem.

csense 9y ago

I'm disappointed he didn't say anything about the frequency stabilization issue. A system based mostly around turbines has the inertia of thousands, maybe millions of tons of spinning metal across the entire grid, which buffers enough energy for control systems (for example controlling how much fuel to inject) to respond to sudden demand changes while maintaining stability. If everything's solid-state solar, following demand is harder.

ghouse 9y ago
The term of art is "frequency response" for "stabilization." This is the system's ability to respond to shocks. Though, the need for frequency response is driven primarily by very large "traditional" power plants. For example, in WECC, the single largest contingency is the loss of a 4GW large nuclear complex. So, should the cost of maintaining this frequency response be borne by the system, or by that specific contingency? Smaller power plants, and smaller transmission lines have smaller contingencies. Following demand is a different function called frequency regulation. Regulating frequency down (curtailing output) is easy for solid-state inverters (including solar), though regulating up (increasing output) is, as you note, more difficult for generation (but very easy for load).

legohead 9y ago

Do most solar installations not include batteries?

ghouse 9y ago
No, most do not. This is true for both utility projects and retail (behind the meter) projects. However, with the falling cost of batteries, and shifting price signals in the market, this is beginning to change -- in Germany, the US, and Australia primarily. Utilities in the US have been changing their rates (tariffs) to decease the cost of energy (kWh), but increase the cost of demand (kW) as a way to (dissuade solar adoption / align cost of service -- depending on your perspective). But within increased demand charges, this permits batteries to offer more value than shifting the time of use of energy.
- francisofascii 9y ago
  Utilities are also increasing customer changes while keeping kWh charges low.

ghouse 9y ago

For those interested in seeing the California "duck curve" in real time, the CAISO publishes it here: http://www.caiso.com/TodaysOutlook/Pages/default.aspx http://www.caiso.com/TodaysOutlook/Pages/default.aspx

dchichkov 9y ago

Strange. I'm looking at major California provider plans - https://www.pge.com/en_US/residential/rate-plans/how-rates-work/find-my-best-rate-plan.page https://www.pge.com/en_US/residential/rate-plans/how-rates-w... And the peak time (and pricing) is 3 p.m. to 8 p.m. PG&E recommends to "shift usage to the least expensive hours: before 3 p.m. and after 8 p.m. every day".