Marginal power

[eledille]

In any case, as an economist I always evaluate activity at the margin, not on average. If I work another hour, will I earn enough extra money to offset the lost enjoyment of the hour elsewhere? If I eat another doughnut, will that make me happier (or just fatter)? If I charge my EV at midnight tonight, how much more carbon will be emitted? This marginal thinking leads to right action in nearly all circumstances.

I completely agree. However, the same goes for every load, because any load has the power to reduce CO2 emissions at the marginal rate by disconnecting. That's why the blame must be shared by all who continue to use electricity, and not just the latest added load or class of load.

I'm not arguing against the concept of marginality, I'm arguing that you can't use it to assign the blame, i.e. the highest CO2 emissions rate, to the newest type of electric device. All electricity consumers must continually evaluate whether their consumption is acceptable, because reducing it will reduce CO2 emissions at the marginal rate. But if you want to calculate the CO2 emissions of one type of electricity consumer, then you have to use the grid average emissions rate, because all other consumers are just as guilty of not removing their load. Said another way, dividing the consumers into high or low emissions groups based on when they appeared is meaningless as only the next change, positive or negative, will make a difference.

About carbon tax: Is the following correct? A carbon tax would make low-carbon sources preferable to the utility, and to the extent that carbon must be used it will increase the price of electricity. That would force some of the least useful load out. This fits in nicely with my idea that all consumers must share the blame (and consequently the grid average CO2 rate is correct for calculation of EV emissions): All consumers end up paying the carbon tax equally according to their usage, in the form of increased electricity price. If the return on the expense is too low, they will cut their consumption. The price is set on the margin, but affects all consumption equally (disregarding the lag due to long term contracts, but new long term contracts will be based on the new spot price).

 

[JRP3]

@Norbert: I agree, and there's an elegant, simple solution: a carbon tax. Unless one makes the choice affect the bottom line, one can't expect the utility to make any other choice. In fact, in most parts of the U.S., it would be a tariff violation for the system dispatcher to take any action other than to dispatch the least-cost available resource.

Because of low NG prices and the desire to reduce coal use we'll probably see more and more NG use. However, recent studies suggest NG, especially from fracking operations, may be as CO2 heavy as coal. Nuclear is probably running full out in most locations, and if you use reserved hydro at night for additional load it's not available during the day so NG will be used. So what other choices are there? I guess additional solar can be used during the day to offset potential NG use but you still need enough alternative idle backup to cover it's loss during heavy cloud coverage or rain and snow. Similar problem with wind, you need the same backup capacity. More nukes as well if they can indeed throttle. Also solar concentrators with molten salt storage where applicable, if it's cost effective.

 

[Robert.Boston]

About carbon tax: Is the following correct? A carbon tax would make low-carbon sources preferable to the utility, and to the extent that carbon must be used it will increase the price of electricity. That would force some of the least useful load out. This fits in nicely with my idea that all consumers must share the blame (and consequently the grid average CO2 rate is correct for calculation of EV emissions): All consumers end up paying the carbon tax equally according to their usage, in the form of increased electricity price. If the return on the expense is too low, they will cut their consumption. The price is set on the margin, but affects all consumption equally (disregarding the lag due to long term contracts, but new long term contracts will be based on the new spot price).

Yes, you describe the impact well, though I'll make two emendations:

1. the (wholesale) price of electricity will reflect the carbon intensity of the marginal unit, not the average fleet. This follows naturally from the market approach to pricing electricity, in which the (wholesale spot) price is set by the offer of the highest-cost resource providing power. (There are some footnotes that belong here, but we'll skip them.)
2. The revenue collected from the carbon tax goes somewhere. If it were handed back to consumers on a head-count basis, for example, it would exactly offset the average consumer's carbon tax payments, but consumers who incurred lower carbon tax costs will get a bonus, while those with higher carbon tax costs will face a penalty. Such a system would reinforce the virtuous behavior you describe, namely the incentive to reduce low-value use of high-carbon energy.

 

[Robert.Boston]

Because of low NG prices and the desire to reduce coal use we'll probably see more and more NG use.

That's already happening at a staggeringly fast pace. There are many gas-fired units operating full throttle this spring that expected to be on reserve shut-down until summer.

However, recent studies suggest NG, especially from fracking operations, may be as CO2 heavy as coal. Nuclear is probably running full out in most locations, and if you use reserved hydro at night for additional load it's not available during the day so NG will be used. So what other choices are there? I guess additional solar can be used during the day to offset potential NG use but you still need enough alternative idle backup to cover it's loss during heavy cloud coverage or rain and snow. Similar problem with wind, you need the same backup capacity. More nukes as well if they can indeed throttle. Also solar concentrators with molten salt storage where applicable, if it's cost effective.

Nukes run flat out when they're available, so no extra energy there. The gas/coal CO2 balance is an important fact that needs to be sussed out, but as I recall a high fraction of the gas losses were in the distribution network; most gas-fired power plants (and certainly nearly all of the larger ones) take gas directly off interstate pipelines.

 

[JRP3]

I thought much of the CO2 from NG was associated with releases from drilling/fracking operations. Actually I guess it's methane release, which is worse than CO2, so focusing only on CO2 may be misleading. However, this piece suggests some possible flaws in the initial NG analysis that claimed NG as worse than coal: Is Natural Gas Really Worse Than Coal? A Case of Activist Science Versus Real Science? - Hit Run : Reason Magazine

 

[JRP3]

Basically as soon as I establish that I plug in my EV every night it becomes 'base load' and ceases to be 'marginal load'. And therefore gets put in with everyone else. When everyone starts running their AC in May/June at night that isn't 'marginal load' it is base load. The base load just happens to be different than in January, but the power company can project it.

I think there is some confusion as to what is base load and what amount of base load is powered by what fuel. Nuclear is always base load power, some of base load power is provided by hydro, coal, and NG. So when you plug in your EV the additional demand is not going to be met by nuclear power even if it somehow becomes part of base load. During the day hydro is probably running full out with some in reserve for peaking loads, and maybe FR. So in large hydro areas some EV power may come from hydro, but I'm looking at a nationwide average, not special conditions. In most areas your daytime EV will be charged with either NG or coal, no matter how long you leave it plugged in or keep plugging it in, until they build more nuke or hydro plants, neither very likely for a while. At night hydro power is also curtailed even further so they can back some up for daytime use. NG and coal are also curtailed, NG the most. Plugging in at night will most likely mean more NG or coal is turned on, or left running if you consider your EV as part of base load. The bottom line is even if you charge your EV every night, if you decide not to charge it one night there are probably going to be fewer emissions from NG or coal that evening. That emission differential is the real cost of charging your EV.

 

[Norbert]

More like to justify building more nuclear since wind cannot practically replace a coal plant.

My understanding is that if there is NG (for daytime demand), that can allow wind power to be used in the night, with that NG as a back-up.

There are several states which do well without much coal, and they don't all replace it with nuclear.

EDIT: And, that seems to be what is currently happening in the US, not just my idea.

 

[JRP3]

You can't look at a state grid as an island, you have to see what they may import as well from their entire grid.

 

[Norbert]

You can't look at a state grid as an island, you have to see what they may import as well from their entire grid.

On the entire grid, coal was down to 38% as of Jan 2012, and it will go down further. Coal is leaving the building.

 

[JRP3]

Let's hope. Coal went down because NG production was so high and NG was cheap. Because of depressed NG prices production plans have been curtailed, which could drive NG prices back up. Most of the coal plants are still sitting there, able to be restarted....

 

[Robert.Boston]

Let's hope. Coal went down because NG production was so high and NG was cheap. Because of depressed NG prices production plans have been curtailed, which could drive NG prices back up. Most of the coal plants are still sitting there, able to be restarted....

...until 2015. The new EPA regulations (MATS) are going to require best-available control technologies on all coal-fired generation. In the mid-Atlantic region alone, there's about 15,000 MW of coal units being retired, in part because of the lousy economics against natural gas, but in large part because those lousy margins don't justify spending big $$ on pollution control equipment. So, even if natural gas prices go up a good bit (against all expectations), the coal plants simply won't be around.

 

[stopcrazypp]

I think there is some confusion as to what is base load and what amount of base load is powered by what fuel. Nuclear is always base load power, some of base load power is provided by hydro, coal, and NG. So when you plug in your EV the additional demand is not going to be met by nuclear power even if it somehow becomes part of base load.

So what you are talking about is new installed capacity (not really much to do with base load and marginal load, since you aren't counting the EV as baseload even if it gets "established" and actually raises the baseload for the utility). If that is the case, then coal shouldn't even be in the picture, because as far as I know, almost all new plants being installed in the US are NG now because the fuel is cheap and relatively clean. And most peaking plants are NG and not coal (NG turbines start quickly).

I think it's far simpler to just assign the average mix to ALL your usage and ignore base vs marginal. It's not exactly clear what applicance counts as 100% base vs what counts as 100% marginal. After all, when you install a new appliance, no one cares about base vs marginal (but for EVs suddenly it's considered 100% marginal).

I'm not sure what the original argument was about, but if you are using the baseload vs marginal load accounting method, on average, EV demand (actually any new demand) will likely be met by NG. You can opt for the green energy programs so your EV usage gets counted for newly installed renewable energy capacity. Or you can install your own solar panels.

 

[JRP3]

...until 2015. The new EPA regulations (MATS) are going to require best-available control technologies on all coal-fired generation. In the mid-Atlantic region alone, there's about 15,000 MW of coal units being retired, in part because of the lousy economics against natural gas, but in large part because those lousy margins don't justify spending big $$ on pollution control equipment. So, even if natural gas prices go up a good bit (against all expectations), the coal plants simply won't be around.

Good to know. Do you have any references I can use? Would be handy info to bring to the table against the coal argument.

 

[JRP3]

So what you are talking about is new installed capacity (not really much to do with base load and marginal load, since you aren't counting the EV as baseload even if it gets "established" and actually raises the baseload for the utility). If that is the case, then coal shouldn't even be in the picture, because as far as I know, almost all new plants being installed in the US are NG now because the fuel is cheap and relatively clean. And most peaking plants are NG and not coal (NG turbines start quickly).

I'm not really talking about any new capacity, just existing capacity. If coal is going to be as quickly regulated out or emission controlled as RB claims then it's really a moot point. However I'm also talking about NG emissions, which are cleaner than coal at the plant but MAY have similar emissions if fracking is used, the jury is still out on that one.

I think it's far simpler to just assign the average mix to ALL your usage and ignore base vs marginal. It's not exactly clear what applicance counts as 100% base vs what counts as 100% marginal. After all, when you install a new appliance, no one cares about base vs marginal (but for EVs suddenly it's considered 100% marginal).

EV's are pretty much the only intermittant load that can quickly hit the grid for hours at those power levels. Nothing else is going to be brought into an established neighborhood in a number of homes in a narrow time frame, and that additional load needs to be met with something dispatchable.

I'm not sure what the original argument was about, but if you are using the baseload vs marginal load accounting method, on average, EV demand (actually any new demand) will likely be met by NG. You can opt for the green energy programs so your EV usage gets counted for in newly installed reneweable energy capacity. Or you can install your own solar panels.

Problem is you can install solar panels without an EV too, so if you install solar panels and drive a Prius instead of an EV the grid doesn't put out any emissions to compensate. I obviously agree that owning an EV gives additional reason to install solar panels so the ultimate effect is a neutralization of some of the additional EV load.

 

[stopcrazypp]

EV's are pretty much the only intermittant load that can quickly hit the grid for hours at those power levels. Nothing else is going to be brought into an established neighborhood in a number of homes in a narrow time frame, and that additional load needs to be met with something dispatchable.

I wouldn't call a couple hour load "intermittent". It seems you are set in calling EVs a marginal load, even though it's entirely possible for EVs to be baseload (which really only means the point of lowest demand during the day; so if you charge during the lowest point in demand, then it's 100% baseload at that point, and a mix of marginal and base during the other points in time). And if utility control of EV charging happens then EVs can become 100% baseload automatically because it is a highly predictable load and utilities can adjust their baseload capacity to account for it (really no different from adjusting for new customers). It seems your core issue is with the intermittent stage between the increase in the base load demand from EVs and the installation of additional "baseload plants" vs just running peaking plants for longer in the meantime to account for the extra demand. As I said before, that additional base load demand will likely be met by natural gas at this point, although wind might also be a good candidate at night.

Problem is you can install solar panels without an EV too, so if you install solar panels and drive a Prius instead of an EV the grid doesn't put out any emissions to compensate. I obviously agree that owning an EV gives additional reason to install solar panels so the ultimate effect is a neutralization of some of the additional EV load.

If you drive a Prius, you already emit at the tailpipe. If you drive an EV with solar panels, it essentially doesn't emit (tailpipe or otherwise). There's a big difference there.

 

[JRP3]

I guess he doesn't comprehend that the US only uses half the peak power from 12 AM to 4 AM and that coal and NG plants slow down electricity production at those times, but do not shut down so we are producing a surplus of energy at that time or hydro and nuclear have to also decrease production (pump the water back up, divert the steam, do something, etc, etc).

That is not exactly correct. For the most part nuclear never ramps down, yes some hydro does get backed up, for daytime use. Coal and NG do ramp down, but they don't produce unused electricity, what they produce is used. If there is more night time demand then they won't ramp down as far. Glenn is essentially correct that marginal load at this point is met with coal or NG.

 

[Norbert]

All true, and all irrelevant to the current discussion of what exists now. I realize it's a narrow focus but the purpose of this discussion, for me, is to get useable data to counter detractors that only focus on today or the very near future. Yes wind capacity is growing but it's not dispatchable and in most places, even CA, there is not so much excess wind at any time that you can claim to be charging an EV from it. In Texas, yes.

If you agreed, then, in consequence, what you call the "current discussion", is the irrelevant part, and it is misleading to give that much way to those pseudo-arguments.

You acknowledged that the "coal plants are getting closed" info is relevant. That's because that is on the relevant level, in the relevant context: looking at the whole system at the grid level.

If you assign 100% coal to EVs, that's like the following situation: Let's say in public transportation, there is more people wanting to take a bus than there is space in the bus, and that each day. Now the solution is to buy an additional bus. But it would be wrong to make the last person, who decided to use that bus and caused it to become full, pay for the whole new bus. You can't make that one person responsible. You have to look at the whole system, and solve it at that level. And at that level, the few MWh added by EVs, at their current market share, are irrelevant to the whole CO2 problem. And even more so, any tiny difference between Prius and EV. *That* is irrelevant.

What matters, is that currently lots of wind power (100 times as much in CA) is being added, and that we start the transition to electric cars today, so that we can take advantage of clean power by the time numbers become significant on both sides. The EV you buy today will become cleaner almost each day, and contribute to a development that will actually be part of a complete solution to the CO2 and oil problems.

If you put all that aside, in my mind you already lost to JP.

 

[Dan5]

That is not exactly correct. For the most part nuclear never ramps down, yes some hydro does get backed up, for daytime use. Coal and NG do ramp down, but they don't produce unused electricity, what they produce is used. If there is more night time demand then they won't ramp down as far. Glenn is essentially correct that marginal load at this point is met with coal or NG.

My point was supposed to be that if in a "perfect" world electricity would be used as needed, not stored. "Surplus" was the wrong term to use. More like you have to ramp down the hydro, NG, and coal in order to meet the minimal needs at that time.
Why store the hydroelectric at night and use coal/NG instead? I realize the hydro is backed up, but why back it up to begin with? Wouldn't it be better to have the NG/coal operating at extremely low loads than back up the hydro (pump the water back up/bypass/etc, which is lost energy)? it's kind of exploiting the numbers game.

The ideal situation would be to run nuclear, wind, and hydro at capacity and supplement it with solar/NG/coal when needed
At least from the graphs for the UK, it looks like nuclear, wind and hydro would be enough to support night-time delivery if operating at capacity. To me it seems like it's more an engineering limitation of the NG/coal plants than anything else and the hydro is trying to make up for the limitations. Can't turn coal/NG completely off every day to meet demand

 

[Dave EV]

To me it seems like it's more an engineering limitation of the NG/coal plants than anything else and the hydro is trying to make up for the limitations. Can't turn coal/NG completely off every day to meet demand

All traditional thermal power plants have minimum/maximum ramp rates and operating efficiencies which are used to determine what plants to ramp up/down and possibly shut off completely. Turning off plants are generally avoided since turning them on typically takes a long time and burns a lot of extra fuel.

In general:

Coal is very slow - it can take a day to ramp up a coal plant from off to on.
Old gas thermal plants are a bit faster - perhaps half a day from off to on.
Gas turbine plants are faster than that - can go from off to on in hours faster.
Modern CCGT gas plants are the fastest - these can often ramp from off to on as fast as a turbine, but have the benefit of being twice as efficiency once running.

Every time one of these plants ramps up/down, efficiency goes down. Typically efficiency stays withing 70-80% of maximum down to 30-40% peak load - below that efficiency tends to drop off significantly, and maximum efficiency typically occurs near peak load.

Cold starts for a power plant are the worst - while ramping up efficiency will be half what what peak efficiency is.

Nuclear plants also suffer from lowered efficiency at less than full output and typically have limited ramp rates as well - not fast to ramp up or down.

 

[rolosrevenge]

My point was supposed to be that if in a "perfect" world electricity would be used as needed, not stored. "Surplus" was the wrong term to use. More like you have to ramp down the hydro, NG, and coal in order to meet the minimal needs at that time.
Why store the hydroelectric at night and use coal/NG instead? I realize the hydro is backed up, but why back it up to begin with? Wouldn't it be better to have the NG/coal operating at extremely low loads than back up the hydro (pump the water back up/bypass/etc, which is lost energy)? it's kind of exploiting the numbers game.

The ideal situation would be to run nuclear, wind, and hydro at capacity and supplement it with solar/NG/coal when needed
At least from the graphs for the UK, it looks like nuclear, wind and hydro would be enough to support night-time delivery if operating at capacity. To me it seems like it's more an engineering limitation of the NG/coal plants than anything else and the hydro is trying to make up for the limitations. Can't turn coal/NG completely off every day to meet demand

The reason hydro is backed down at night is because it's the easiest to back down, and, there is only so much water that will flow down the river in a given year. Once you've reached your water limit, that's all you can do, you can't get more coal or natural gas, or uranium deliveries like with other plants. So you store the water and use it when it is most valuable.