Jesse and Alex, this is brilliant, and it’s interesting your caution that VRE maximum production at any instant

can’t exceed about 55-60 percent of total demand without risking system stability

is interpreted by the American Wind Energy Association as a positive:

Western Grid Can Withstand Disturbances Under High Renewables Penetration

Is 55% “high penetration”? IPCC doesn’t think so, and their prescription for eliminating non-essential fossil fuel combustion by 2100 makes the question of how we might best replace remaining fossil generation an urgent one.

If it’s with advanced nuclear there’s no reason, except in poor or politically unstable areas of the world, to continue to devote $billions in public resources to wind and solar – nuclear is as clean, it’s more stable, it occupies a fraction of the geographic footprint. It’s not as nimble as CCGT, but we’ve only begun to scratch the surface of applications which might divert overgeneration to liquid fuel synthesis or desalination. Unlike CCS, which also “eats its own lunch” as fossil use declines, nuclear is proven, and transmission uses existing infrastructure.

As Fukushima fades in memory, I suspect the tide will shift dramatically in the next five years away from renewables to nuclear. The main challenges are ones of public perception and expense, neither of which would be obstacles in the face of determined leadership.

Yes, if one decides to set prices by establishing an artificial market with a marginal-price pricing system, that is a decision that discriminates in favor of fueled power sources and discriminates against equipment-intensive harvest of freely available renewable energy.

But the decision how to set prices to be paid for production of energy is not an empirical law of nature, it is a policy decision.

If the average prices in marginal-price markets drops when harvest of freely available renewable energy is paid for on a fixed tariff that is adequate to cover the opportunity cost of providing the equipment, which is often the case for high quality windpower resources and sometimes the case for utility grade solar, then the decision to require the harvested renewable energy to be sold in the marginal-pricing market is a decision to increase the overall cost of electricity to the final consumer.

So what it is that this piece is describing is how a particular policy choice for price regulation places a lower limit on penetration of renewable energy than some alternative policy choices for price regulation.

“The capacity factor threshold implies that wind may eventually be able to provide on the order of 25-35 percent of a power systems’ electricity, while solar may top out at 10-20 percent in most regions.”

The capacity factor threshold implies no such thing for wind. What the capacity factor threshold implies is that, depending upon how we regulate the payment for the power produced by wind, there may be a substantial premium available for increasing the capacity factor of wind. And, of course, the capacity factor is in part a design choice … a taller tower and a larger blade diameter, geared lower for the same nameplate windturbine capacity yields a higher capacity factor, location in an area with less volatile wind yields a higher capacity factor, and long distance transmission with any variable energy resource with harvest that is not highly correlated yields a higher capacity factor.

If we insist on regional autarky, then the threshold in places like the Pacific Northwest (due to volatility) and the Southeastern US (due to wind speeds) may well be in the 25-35 percent range … but then the direct response to that is to not insist on regional autarky, and to allow them to import from wind resources where the capacity factor in the region can be designed to be in the 40%-50% range on a regional basis and in the 45%-55% range when transmission is provided to allow regional cross-haul.

“This is precisely why we both laud the growth of wind and solar, but are very concerned when conversations about decarbonizing the power system become overly focused on a “renewables-only” path forward. Wind and solar will be important contributors to a high-energy, low-carbon planet. But they can’t do the job alone — far from it. Other nonvariable renewable power technologies like biomass and geothermal face different, but potentially even greater obstacles in the form of resource availability in the case of geothermal and land footprint in the case and bioenergy.”

It is, however, important to avoid the all-or-nothingitis that seems to plague reaction to serious study of “renewables only” paths. An objection along the lines of “we cannot do that” may take an energy source off the table. An objection along the lines of “there are limits to how much of that we can do” means that energy source provides a slice of the required energy, and we may need to continue looking for additional slices.

Biofuels in particular is plagued by efforts to provide a “easy transition” plug and play liquid fuel replacement for gasoline and diesel in powering our massively inefficient transport system. And inefficiencies multiply together to amplify the land footprint required to provide those “plug and play” gasoline or diesel replacement biofuels … the energy inefficiency of our gasoline powered passenger and diesel powered freight transport, times the energy inefficiency of converting biomass into suitable gasoline and diesel substitutes, times the energy inefficiency of the biomass production of corn kernels, the most commonly used liquid biofuel feedstock.

But 100% capacity, 20%-30% supply dispatchables including biomass, the dispatchable component will include an existing ~5% for hydro, and an additional 2%-5% from new hydro power, a range of smaller niche dispatchables such as biogas for another 1%-5% … and the land footprint of biomass production for biocoal for on the order of 5%-10% of electricity supply is far more reasonable than the land footprint of corn ethanol production for any substantial share of our current transport energy supply.

Alex and Jesse, thanks for an important discussion on renewable penetration limits.

A related topic I believe would be worthwhile to discuss is optimal ownership and market models for sustainable energy.  It seems to me that the “deregulation” and “privatization of plant ownership” that much of the electricity industry has undergone is really only suitable when most generation is gas-fired.  

For a capital intensive power plant that has low operating cost and long life (e.g. hydro, nuclear, solar, and to a lesser extent wind), once the construction loans are paid off, the plants become cash cows for their owners, if the power is sold on the merchant market (for long-term PPAs or sub-hourly markets).  This is because the market electricity price must on average be high enough to support the needed growth and plant replacements, but the same price is paid to new and old plant owners.

If instead the sustainable power plants are owned by regulated public utilities, the low cost energy from older plants is averaged in with the more expensive power from the newer plants.  In the long term, this public ownership could result in wholesale electricity prices of only one half of that produced by a deregulated market (e.g. assuming half the electricity is free).  Public ownership is still compatible with project development by private companies, as contracts could be written such that ownership is transferred to the utility several years after construction (thus the private company would own the project during the financially risky development phase, and the public utility would own it during the cash cow phase).  As Joris mentioned recently, private companies won’t be happy to take ownership of risks that are caused by the public and their government (as is often the case with nuclear project delays), so the private developer likely won’t be advantageous in this case.

The AWEA’s quarterly market reports include a list of new wind projects showing the name of the owner and the power purchaser.  Most of the time this seems to be different companies (presumably 20 year power purchase agreements).  Relative to the alternative, this is a strategy that will impose a high cost on rate-payers, 20 years down the road, and it will become more important as sustainable electricity penetration increases.

Pushing down the cost of sustainable electricity is a prerequisite for power-to-fuel technology, and power-to-fuel is likely to be an important part of decarbonizing transportation, decarbonizing rural home heating, as well as assisting with integration of variable renewables in a low fossil fuel grid.

—

To Bruce’s point, note that the public ownership model does not solve the problem of the “capacity factor limit”.  This is because even though with this model wholesale prices are set by the average production cost, each new power plant, wind farm, or transmission line must be cost-justified on an incremental basis, so that its value can be compared to incremental storage, demand response, or other types of generation.  When calculating its value, the incremental system-wide curtailment should be completely assigned to the newest plant. As variable renewables approach their “capacity factor limit”, this incremental value will drop rapidly, just as it does in the market case.

“As long as people really want wind & solar, subsidies will make it happen.”

It’s remarkable to me that the subsidies have been tolerated in Germany as long as they have; it is my understanding that the subsidy cost is plainly stated on home electric bills.  In the US, public support for residential solar seems to be based on the illusion of saving money and a shamelessly over-stated “value of solar” sales pitch.

“The point about declining market value is in my opinion easily the most important one.”

That is the market value when the market is constructed in a particular way.

After all, if wind is given a feed-in tariff and is required to be purchased first before purchasing the fueled energy source in a marginal price market, and the average cost of power that results is lower than if a smaller amount of wind that is commercially viable selling power on the basis of marginal pricing is provided, then the feed-in tariff at that rate is clearly not a subsidy.

Rather, if that feed-in tariff arrangement is the least cost arrangement for energy customers, it would be the insistence that windpower always sell into the marginal pricing market at the current spot market price that is an implicit subsidy to the fueled powerplant sellers.

Great article, which – for what it’s worth – I will be linking to often because it is so informative and crucial.

One quibble though. The following sentence:

“First, as a growing body of scholarship concludes, the marginal value of variable renewable energy to the grid declines as the penetration rises.”

This seems to imply that the decline of the marginal value of intermittent energy sources with rising penetration is a previously unknown phenomenon only recently discovered by scholars. This is not the case, in my experience. At any rate, this issue (the economics of intermittent sources of power) was in my academic curriculum when I studied energy engineering back in the ’90’s, and it was not presented as ‘new’ even then.

As such, the sentence does something which I tend to cringe at: it appears to imply that scholars are simpleton dupes who are chronically behind the curve and only able to inform society about crucial energy issues long after they manifest themselves.

This is not the case. If academic knowledge of power generation economics appears to lacking in the public discourse on energy matters, it is because this knowledge has been deliberately suppressed, miscommunicated and eviscerated by parties having an interest in misleading the public for private gain.

Great article, Alex and Jesse. 

One question came to my mind, prompted by the following line from your post:

“It is thus cheaper for consumers to ramp the baseload power plants down to their technical minimum output, but then curtail any wind or solar beyond that point. And if those baseload plants are emissions-free nuclear stations, this strategy is both less costly and just as good for the climate.“

Wouldn’t this decreased utilization of baseload plants (dubbed “profile costs” by Hirth) have some real economic and technical consequences for plant operators? Or are you suggesting that baseload plants would be run continuously at their technical minimum output level in a >50% VRE system?

Bruce, what form of renewable energy is “freely available”? 3MW GE wind turbines cost $2 million installed, and high-voltage DC transmission is ~$1 million/mile. It’s not a policy decision which turns that into a marginal cost for consumers, but an empirical law of human nature. Money costs money – i.e., investment nearly always requires financing, and an associated periodic repayment schedule.

We could repay those loans by charging customers a flat fee, or absorb the liability into property taxes, but that certainly isn’t fair to people who use very little electricity at all. We could go door-to-door and ask for donations, but experience with financing my kids’ sports teams has taught me that’s a difficult way to raise money (every drive ended well shy of $2 million). Loans serve to spread out the financial burden and serve as a buffer against risk, so marginal pricing is here to stay.

With renewables-only scenarios, it’s more a case of nothing-or-nothingitis. The problem is this: for every watt of VRE, we will always need a watt of dispatchable energy to back it up (grid-wide storage and schemes which rely on parlaying geographically-distributed variable resources into a dependable whole are hopelessly optimistic). When a single method of clean, dispatchable generation is available to us, why are we bothering with renewables at all?

The point being made here is that while high renewable penetration is technically feasible, it is also uneconomic. Nothing we have seen in Germany so far suggests that is not the case. Indeed the increasing disparity between decling wholesale electricity prices, and increasing retail electricity prices in Germany, suggests that Jesse’s analysis here is correct.

Jesse says, “watch out, because if we install a lot of renewables it’s going to be very expensive!”

And Bruce responds, “That’s not true, because we could structure the market differently, in which case there are other ways that we could install a lot of renewables that would be very expensive too!”

Am I missing something here? Did long-distance transmission suddently become free? Is storage free? Does “regulating the payment” to wind producers not cost us anything? Have the laws of physics been repealed? Because both they and the NREL tell us that the energy available in the wind scales with the square of rotor radius, while the cost of the turbune scales with its mass, which scales with the cube of rotor radius — which means that building larger turbines costs more per Watt-hour delivered.

…{a number of important points before the “–“}…

“To Bruce’s point, note that the public ownership model does not solve the problem of the “capacity factor limit”.  This is because even though with this model wholesale prices are set by the average production cost, each new power plant, wind farm, or transmission line must be cost-justified on an incremental basis, so that its value can be compared to incremental storage, demand response, or other types of generation.”

Neither the public ownership model, or private average cost rate-basing pricing, “solve the problem of the capacity factor limit”, when stated as a dichotomy, but they each have a different limit.

The capacity factor limit in the above piece is composed of two different elements: the changes in the value of the contribution of renewable power to the portfolio of energy producers, and the changes in payment awarded relative to the value of the contribution made. At high penetration of very low marginal cost producers that have (internal and external) average total cost of power that is competitive with higher marginal cost producers, the result is a lower average cost of power to the final consumer, and that result is translated into a lower economic value for all energy in the system, low and high marginal cost producers alike.

But since this occurs when the average total cost is competitive, because of the way that dispatch bidding markets set such very low prices when only very low marginal cost producers on the merit-order curve are chosen, when a sustainable average cost pricing system would not charge different prices to customer, it implies that the difference in price is not a difference in long term economic value, but rather reduction in the payment awarded relative to long term economic value.

The sustainable long term cost as a component in a system does not change with difference pricing systems, and in all cost optimizing studies of prospective all-renewable or high-renewable energy portfolios, the harvest of renewables will not be 100%, as over-supply and curtailment as a source of capacity has a variety of trade-offs with dispatchable renewable sources and with storage. So any pricing system based on average system costs will (and should) have a limit on how much harvest of variable renewables it should pay for. But legacy Capacity Factor offers only the crudest of first order approximations on what share it ought to be rewarding, since it ignores substitution and complementarity.

And for all technologies where capacity factor is in part determined by design, the optimal low penetration design is likely to optimize for levelized cost of energy supply while the optimal high penetration design might trade that off for higher capacity factor, so the available Capacity Factor for newly built harvest of variable renewables will typically be higher than the average installed Capacity Factor. 

“At any rate, this issue (the economics of intermittent sources of power) was in my academic curriculum when I studied energy engineering back in the ’90’s, and it was not presented as ‘new’ even then.”

The critique that the mis-named “deregulated electricity markets” underpays very low marginal cost producers with fixed costs as a very large share of total costs goes back into economic debates on utility pricing was a standing point in the economic literature on regulated utilities in Industrial Organization when I was in grad school in the early 90’s, so it would be in the literature by the 70s at the lastest, and I’d expect that it would more likely date back to the 50’s.

When I was first introduced to the topic in the 90’s, reservoir hydro was the high fixed cost, low variable cost energy source example … past a certain total capacity, marginal pricing alone offers less reward to reservoir hydro than average cost pricing does. For low average variable cost dispatchables, the under-payment by marginal pricing can be partly compensated for by establishing a capacity market, and if all very low marginal cost power is dispatchable, and if we ignore the massive external costs of fossil-fueled power, the end result is close enough to a cost-optimized system for people to argue that the efficiency benefits marginal pricing markets outweighs any inefficiency in the final composition of energy sources.

No, the point being made here is that whether or not high renewable penetration is economic, when electricity is prices by spot markets for the dispatch of electricity, it becomes non-financial.

After all, the baseline is the price of power under marginal pricing, not the cost of power in various cost-optimized systems with different characteristics, so it cannot determine whether or not high renewable penetration is economic, since at the level of long term economic cost and benefit, the pricing system is a matter of economic policy choice.

To get the German result, it would also be necessary to put rules in place that allow industrial users to avoid paying for some of the fixed costs of the power that they are using, so that there is a cross-subsidy from the retail electricity rates to industrial electricity rates.

“If you have a competitor that produces at lower marginal cost, you’ll need to ramp down when you’re power is not needed. That’s fine in principle. It may mean some baseload plants can no longer recover their costs and will have to exit the market until equilibrium returns.“

This is much less of an issue in an abstract economic model in which it is assumed that resources can flow freely between different sectors … but the nature of “baseload” fueled power plants is that they have higher fixed costs relative to their total costs, and after they have been constructed, much of that fixed cost is a sunk cost in real terms, and a debt servicing cost in commercial terms.

Provided price paid for power is above average variable cost, they would not exit the market, but would continue to operate at a loss. And the range between replacement break-even (including a normal profit) prices and average variable cost of operation prices is quite wide for high fixed costs power plants.

If rate-based, they are an “oops” that keeps rates higher than they would otherwise be. If privately owned, rather than being pushed out of operation, their owners may instead be pushed into bankruptcy, and when the plant is liquidated, it is bought at current residual value for the fixed cost remaining after the operating costs are deducted from available revenues, and so for the new owners will again be at break-even or better.

“Jesse says, “watch out, because if we install a lot of renewables it’s going to be very expensive!”

And Bruce responds, “That’s not true, because we could structure the market differently, in which case there are other ways that we could install a lot of renewables that would be very expensive too!”

Am I missing something here?”

Evidently … if we do anything it is going to be expensive, including if the choice is to do nothing. Indeed, doing nothing different at all will have the highest cost of them all.

But you have mis-characterized Jesse’s article. It says that if we install a low of renewables, including renewables that are presently substantially cheaper than available fossil fuel alternatives when counting full economic costs, then under marginal pricing for electricity. the marginal value of the renewables will decline as more is installed, which creates a limit on how much renewables can be installed on a commercial basis.

The reason that the marginal value will decline is for two fundamental reasons: (1) no matter what variable renewables are added to the portfolio, they have a point of diminishing returns as pure energy supply, and past that point, the average total cost (per kWh) rises and (2) under marginal pricing, the more low marginal cost energy sources you have on the grid, the smaller the price charged relative to the average total cost (per kWh).

(1) would be there on any average total cost pricing system as well, but (2) would go away with average total cost pricing. And so with average total cost pricing the limit would be higher than with marginal pricing.

As far as the “other expensive things”, if paying for some particular ability to do things reduces the total system cost of variable renewables at a given level of penetration, the fact that the current pricing system is not set-up to pay for that ability tells us more about the limits of the current pricing system than the value of those abilities. An appropriate approach to valuing those abilities is to build a cost-optimized system that includes them, and a cost-optimized system that omits them, and consider the change in total system cost.

“Bruce, what form of renewable energy is “freely available”? 3MW GE wind turbines cost $2 million installed, and high-voltage DC transmission is ~$1 million/mile.”

Yes, and if the energy to make that $2m wind turbine spin was not freely available, not only would the total cost of harvesting that energy would be substantially higher, but the marginal cost of the harvest would be substantially higher, and variable costs would be a substantially higher share of total costs, and the “operate at a loss” price range would be substantially smaller.

“It’s not a policy decision which turns that into a marginal cost for consumers, but an empirical law of human nature. Money costs money – i.e., investment nearly always requires financing, and an associated periodic repayment schedule.”

Its a policy decision (1) whether to require financing (we do much of our highway investment in the US without charging any financial costs) (2) whether to require financing at market rates for public or private debt (we do most financed road construction at the former, while we do financed railroad construction at the latter), (3) whether to tax the interest income or not (we tend not to tax interest payments for debt to finance road construction but do tax interest payments for debt to finance railroad construction), (4) and what pricing system to apply to the payment for the power that will refund the capital costs, whatever they are based on what policy decisions are made for (1-3).

The argument from some kind of “empirical law of human nature” into a single result when translating capital costs into marginal costs to the final customer can only be based by ignoring a number of policy choices. Typically this is done by assuming that the status quo policy choice that has been made previously is some kind of “natural” state of affairs.

And this kind of assumption is indeed rooted in an empirical law of human nature: it is a common human tendency to come to treat the habitual rules of behavior in the social settings we experience as if they are “natural”, even if they are entirely arbitrary. For Americans, writing their given name first, family name last, and having a middle name that they can in most cases optionally write down or not seems “natural”, while for Chinese, it is “natural” to write the family name first, and having some part of your given name that you can optionally omit is quite “unnatural”.

But the cross-subsidy to industry means that while the retail price is higher than it would be in the US with a similar pattern of deliberately high feed-in tarrifs as an infant industry policy, there are also fewer industry groups with a substantial interest in subsidizing the process of whipping up opposition to the policy.

And on the retail side, once you’ve made the investments in energy efficiency that make sense at the higher retail rates, the savings available for any given reduction in the retail rate has been effectively reduced. So the same increase in retail rates that would cause a substantial disruption for a substantial proportion of the population in the US does not translate into the same potential windfall for a similar proportion of the population in Germany if the system were to be removed.

Bruce, you use highway costs as an example, yet users of highways do in fact pay a marginal cost for use of highways through fuel taxes.

An empirical law of human nature that holds true in the U.S., China, or anywhere else: if you use a larger share of something that incurs a cost, you will be obligated to pay more for your share. That’s marginal pricing, and that’s here to stay.

Jessie and Alex,

Great piece but I disagree with this important comment from your post:

“Can’t energy storage help avoid curtailment and keep VRE shares growing? Yes, but only somewhat.

Storage isn’t free after all, and storage owners will make their money on the spread between the price they buy power at and the price they sell at later in the day.“

Storage from EVs is free in an environment with real time pricing of electricity.  EVs can run algorythms where they only partially charge for the next day leaving the remaining charge to be done at times when the grid most needs a sink for overproduction.  In the case of wind, wind power is night peaking so this fits very well with the emerging usage patterns.  Further, appliances such as dishwashers could be set up to take pricing signals from utilities to run at optimal times at night.  In both cases the users incur no charge for their actions and the incentive lies in reduced cost for the commodity.  It is a win win since it firms up the floor on pricing of the VRE. 

Over the summer I hope to have time to write up why it makes sense that EVs will have over capacity in their batteries.  However the punch line is that, as long as interest rates are low, if someone plans for a 200,000 mile lifetime of their car, they can either buy two battery packs over the lifetime of the car or get it all up front, providing the added value of longer range and the ability to benefit more from load shift arbitrage to reduce their cost of energy. 

The extra weight of an oversized battery does not cost much in energy due to regenerative braking.  Anyone who has a Prius knows that it gets comparable and sometimes better milage in town than on the highway.  This is, in part because you get back your kinetic energy from regenerative braking. 

I base 200,000 miles on anecdotal evidence that it seems to be becoming the norm.  My Prius was retired at 230k miles only because it was totaled in an accident.  In talking to others with Prius’s I am hearing comparable experience.  And mechanics pretty routinely tell me that they mostly see Prius’s for oil changes, alignment, new tires and not much else.  Since, by most accounts EVs are expected to be more reliable than ICE vehicles, I am going with 200k miles as reasonable expectation.

Before the naysayers go off on that, I know, Tesla has had some difficulty with their motor bearings.  The Model-S is the first volume production vehicle for a company that is new to the automotive space.  It is a high performance vehicle with extreme torque on those bearings.  This is very solvable and likely a non-issue in non-performance vehicles such as Leaf, Bolt and Tesla Model-3

Clayton,

“Storage from EVs is free in an environment with real time pricing of electricity.”

The TESLA 10 kWh unit:

– Is designed to provide power up to 2 kW during a power outage.

– Can be cycled up to 50 times per YEAR for buying energy during low cost periods, storing it, and using the energy during high cost periods*.

– Has an installed cost, with other equipment, of about $5,000 under lease, or $7,140, if bought. 

http://www.bloomberg.com/news/articles/2015-05-01/solarcity-taking-orders-for-tesla-batteries-starting-at-5-000

Assuming a 50% charge/discharge, and a 90% AC to DC inverter efficiency, and allocating half of the 8% DC-to-DC loss to the charging side (the TESLA unit has a round-trip DC-to-DC efficiency of 92%, per spec sheet), it would take 0.5 x 10/(0.9 x 0.96) = 5.787 AC kWh of off-peak grid energy to charge 5 DC kWh into the unit.

During on-peak hours, one would get back 0.5 x 10 x 0.96 x 0.90 = 4.32 AC kWh to use in the house, for a (1 – 4.32/5.787) x 100% = 25.3% loss of energy!!

In Southern California, base rates are $0.11, off-peak, and $0.46, on-peak. Such rates likely are THE best-case scenario in the US. But this rate ratio is only for 6 months.

The off-peak cost would be 5.787 x 0.11. The on-peak avoided cost would be 4.32 x 0.46, for a profit of $1.35/day for 50 days a year!!! The monetary gain of this arbitrage is miniscule.

The cost of financing, PLUS any costs for O&M, PLUS any capacity degradation due to cycling are ignored.

The 10 kWh unit providing up to 2 kW of “quiet, clean” energy for a few hours of outage is quite a stretch, based on need, and is making no economic sense, unless grossly subsidized; 2 kW for an upscale mansion is insufficient and multiple wall-hung units would be required.

A PV solar system owner doing some energy “shifting” with a 7 kWh unit is quite a stretch, based on need, and makes no economic sense, unless grossly subsidized.

Utilities could be using multiples of the TESLA 100 kWh units for stabilizing their distribution grids that have large numbers of PV systems. The US DOE, getting on that bandwagon to look progressive, will be subsidizing those applications for “demonstration purposes”.

“Storage from EVs is free in an environment with real time pricing of electricity. …only partially charge for the next day” 

“…as long as interest rates are low…”  “ …the added value of longer range and the ability to benefit more from load shift arbitrage to reduce their cost of energy“

Clayton, as we’ve discussed before, I agree that many EVs (half?) will have enough capacity that their owners will be indifferent to the exact time the vehicle is charged, as long as it is fully charged by the next morning.  This makes EVs supportive of baseload energy sources, with the only cost (i.e. not free) being the use of 6 hour chargers instead of 10 hour chargers.

But the idea that EV owners will accept their EV being only partially charged after being plugged-in all night is quite different.  There will have to be a strong monentary incentive for this to work (to help motivate the needed battery size expansion, and to devote that extra capacity to grid service rather than owner peace of mind).  Also, there are extra hurdles that make this route to grid energy storage more expensive than utility-owned batteries:  

• utilities always get lower interest rates than individuals.
• utilities get economies of scale when buying and maintaining batteries.
• car companies/dealers always get higher profit margins on up-grades than on base models.
• batteries for EV applications have size and weight constraints that likely add cost and shorten service life compared to stationary applications.

So if batteries are not affordable for utility applications, then up-sizing vehicle batteries to support the grid probably won’t work either.

Redundant comment removed.

Willem,

I agree with much of what you said.  I would encourage you to reread my post, I don’t think we are in disagreement.

I was responding to the original post’s contention that wind power will have to be curtailed and pointing out that EV charging can be managed in such a way as to favor periods of wind overproduction.  So there is no cost, i.e. the EV owner is doing what they would have done anyway, just adding some flexibility to timing.  This will reduce the need to curtail wind particularly in areas like Texas where the wind blows more at night and is decorrelated from the load.

Nathan,

If we are discussing next week, next month or next year I agree with much of what you are saying.  In talking about 10 – 20 year time frames I disagree with much of what you are saying. 

“But the idea that EV owners will accept their EV being only partially charged after being plugged-in all night is quite different.  There will have to be a strong monentary incentive for this to work (to help motivate the needed battery size expansion, and to devote that extra capacity to grid service rather than owner peace of mind).“

I disagree.  Lets look at some of your premisses. 

1) People will not accept partial charge.  Most people fill up their gas tanks on an as needed basis.  It is common for people to go down to 1/4 tank before filling.  Why would a commuter, with a 200+ mile battery, who plugs in every night, care if they had a 75% charge rather than full charge?

2) High capacity chargers too expensive.  As EVs become common I think that many will opt for them.  They will certainly come down in cost as any manufactured product does with volume.  I would expect comparable to putting in a plug for an electric stove.  Seems likely that new construction will come EV ready.  Installing the EV connect point will probably be in the “noise” as far as wiring cost for the home.

The rest of your comments are not really applicable to what I have in mind.  Hope to have time to write this up and submit for clearer conveyance of my ideas.

For the purposes of this post and comment thread, EVs will certainly have some degree of ability to load shift allowing for reduced curtailment.  If I am right this will reduce curtailment substantially and allow for a non-trivial increase in wind penetration.  If you are right it will not improve things sufficiently to be of value. 

Bottom line it is free storage and due to decorrelation of EV charging it will reduce wind curtailment.  If the grid gets smarter, it will be more effective, if EV battery size continues to increase, it will be more effective.  It is an interesting approach that has only begun getting discussed.  In the past most have been talking V2G and as Willem says (and I agree) that has some serious downsides and a much longer timeline to be useful if ever. 

‘Bruce, you use highway costs as an example, yet users of highways do in fact pay a marginal cost for use of highways through fuel taxes.”

Marginal cost is the additional cost of producing an additional unit of output. There is no system in place at all to charge highway users the marginal cost of their use of the road.

“A” marginal cost implies that there are multiple marginal costs … and there are, since the marginal cost for a short-run decision where some factors of production are fixed inputs for that particular decision will not necessarily be the same as the marginal cost for a long-run decision where all of the factors of production are subject to change …

… but whatever the decision horizon, a marginal cost is always the cost at the margin, and it is straightforward that the marginal cost to the motorist and heavy truck road user is quite clearly a policy choice, and is not determined by the marginal cost imposed on the road system by their use of the road, given that the real marginal cost to the motorist has been declining steadily over the past several decades, while the real marginal cost to the system has been increasing, given that even while the US has not been maintaining the roads that it has, it continues to build more, and the US is on the increasing-cost portion of the highway construction cost curve.

“The marginal cost to the user” will therefore be determined by the tariff system in place, and there is nothing about the physical characteristics of any given electricity supply that determine which tariff system is in place.

A system where the price paid for the wholesale electricity supply is determined in a marginal-pricing auction market will be biased toward paying a smaller share of the total average break-even costs of a low marginal cost energy producer and a larger share of the total average break-even costs of a high marginal cost energy producer. That comes directly from the logic of break-even and shut-down prices for profit-seeking producers.

Where the physical realities of renewable energy production enter in is that spending money on equipment to harvest energy that is freely available at the turbine blade, or at the panel is an activity that is intrinsically a high fixed cost share, low variable cost share activity, and so the marginal cost of the energy harvested is a small fraction of the average total cost of the energy harvested.

Among fueled power plants, it would seem to be nuclear power that is most heavily discriminated against by marginal pricing of power, since its fuel costs tend to be a lower share of its average total cost of energy than for fossil fuel energy sources. And this systemic discrimination will only become stronger if the full economic costs of CO2 emissions are internalized into the price of the fossil fuel, further increasing the fossil fuel shutdown price as a percentage of breakeven price.

“If renewable energy is ever to become truly subsidy independent and earn its keep in electricity markets, that means there is a natural stopping point at which a marginal increment of wind or solar will become unprofitable. The market revenues earned by these VREs will eventually fall far enough that it’s no longer worth deploying more.

This is also why the idea of reaching “grid parity,” or a levelized cost equal to the prevailing market price, is pretty meaningless. As soon as wind or solar penetration grows, the goal posts move further away due to this merit-order or market price effect. Wind and solar costs will have to keep falling to secure greater penetration levels and remain profitable at the ever lower and lower market prices caused by increasing VRE penetration.”

Note that if we do not choose to assume that an artificially constructed marginal pricing market is somehow more “natural” than any other price regulation system, this translates into a sharp critique of relying on marginal pricing.

These two paragraphs taken together say that even when the renewable source of energy is the lowest cost source of energy, the artificially constructed marginal pricing market will prevent an economically optimal roll-out of variable renewables, because compared to fossil-fueled power, “too much” of that reduction in cost will be handed over to consumers for the fixed costs of the renewable power to be paid for.

And this is not just variable renewables … since nuclear-fueled power tends to have lower variable cost share and higher fixed cost shares than fossil fuels, this discrimination by the system selected to regulate wholesale power prices also discriminates against nuclear-fueled energy. This discrimination is strongest for nuclear fueled plants that are operating as continuous output baseload suppliers (which would be part of the reason why the French have their system of partial load-following operation of their nuclear plants, with freshly refueled plants ramping production up and down on a planned schedule and transitioning to baseload operation as their fuel is consumed).

Bruce, cost of fuel is a liability that’s factored into an energy producer’s bid, just like the principal on the loan used to purchase a producer’s wind turbines, or the interest on that loan, or the maintenance requirements of a nuclear reactor. I don’t know why you’re breaking it down into whether the fuel is free or not, because no one who’s buying or selling energy cares.

Both before and after a wind turbine is paid for, it’s an asset which a producer can use to generate energy. Utilities who are buying that energy pay more for more of it, just like water (or toll road access, or tomatoes). In other words: even if their expenses are not priced marginally, the product they’re selling is.

So what market design do you suggest?

Right.  So with the deregulated margin-priced markets (wherein merchant power producers bid to sell electricity, bids are accepted starting with the lowest, until demand is met, then all accepted bids receive the same unit price based on the highest accepted bid), fossil fuel can not be driven to zero, even with a carbon tax.  The sustainable electricity component should not be so large that prices fall to zero very often.

In a simplified analysis, with one fossil fuel bidder again one/two alternative provider (estimates based on the “capacity factor rule of thumb” and baseload fraction):

• Solar would blend about 20/80 with fossil fuel
• Wind would blend 30/70 with ff
• Wind +solar  would blend 25/12/63 with ff
• nuclear would blend 60/40 with ff
• nuclear+solar would blend 60/5/35 with ff

Adding a carbon tax allows all options to penetrate higher against fossil fuel, since more frequent zero-price events can be tolerated.

As discussed elsewhere, adding night-time charging of EV gives the nuclear options a nice boost, and the wind options a small boost.  Operating the nuclear plants as combined-heat-and-power for district heat reduces fossil fuel use in the non-electric heating sector.  Adding dedicated high temperature nuclear plants for industrial process heat can reduce fossil fuel use in that sector.

The really annoying thing about this market design is that market over-crowding (e.g. too rapid a build-out, or falling demand) causes all non-fossil producers to loose money!  Effectively, the threat of future efficiency improvments becomes yet another mechanism that encourages the market to keep using fossil fuels.

I’m pretty much agnostic between different non-discriminatory pricing systems with a carbon price on a regular upward track toward the full opportunity of GHG emissions.

Discriminatory pricing systems in favor non-fossil-fueled power are not as desirable as non-discriminatory systems with a transition away from allowing costly GHG emissions to be performed free of charge … but substantially better than discriminatory pricing systems that favor fossil fueled power.

Rate basing of power on an average replacement cost basis, setting prices on the basis of long term power purchase agreement markets, modifying marginal-pricing markets so that non-fueled or low-fuel-cost power sources have a baseline feed-in tariff, or modifying marginal pricing markets so that non-fueled or low-fuel-cost power sources have a fixed feed-in tariff are all pricing systems that can eliminate the bias of marginal pricing systems toward fossil-fueled power.

There are advantages and disadvantages to each, and I view the magnitude of the advantages and disadvantages in choosing between them as secondary to the disadvantage of aiding the dependence on a fossil fueled status quo that is a long term policy of committing suicide as advanced industrial economies.

“I agree that many EVs (half?) will have enough capacity that their owners will be indifferent to the exact time the vehicle is charged, as long as it is fully charged by the next morning. This makes EVs supportive of baseload energy sources, with the only cost (i.e. not free) being the use of 6 hour chargers instead of 10 hour chargers. But the idea that EV owners will accept their EV being only partially charged after being plugged-in all night is quite different.”

Note that smaller capacity may well imply greater indifference as to the exact time the vehicle is charged … since the incremental cost of a faster charge for a pluggable hybrid with a smaller battery is likely to be smaller than the incremental cost of a faster charge for a substantially larger battery in a pure EV.

But for those with the larger capacity batteries, if they can set and forget a standard charging regime to charge up to 80% when RTP tariffs are higher and 100% when RTP tariffs are lower, and the over-ride THAT they want to have it fully charged and WHEN they want it charged is easy enough, more capacity makes them more likely to routinely trade-off partial charging for lower electricity rates.

” Also, there are extra hurdles that make this route to grid energy storage more expensive than utility-owned batteries:”

None of those extra hurdles offset the fact that the buyer of the EV (or pluggable hybrid) is buying the battery in any event, and the incremental advantages available only has to be sufficient to get them to change the way that they charge that battery, and does not have to be sufficient to cover the replacement cost of the battery.

The big unknown factor in the EV battery scenario is how many privately owned individual motor vehicles we have. The more share cars / riderless cabs / etc. we have, the smaller the opportunity, which rests on the massive material inefficiency of our present local passenger transport system organized around the private ownership of “one size fits all” motor vehicles.

“[…] and if all very low marginal cost power is dispatchable, […]”

There’s the rub.

Solar/wind energy is not and never will be dispatcheable, which means their value is equal at best to the value of the fuel costs they displace.

In comparison with coal fired power generation, (unsubsidised) renewables are far more expensive than the coal fuel costs, which means coal power beats renewables on a level playing field.

Making matters worse, solar/wind are not continuous, which means they imply chronic fossil fuel burning for backup. So not only does solar/wind fail to compete with coal, they also fail to eliminate fossil fuel burning even if they would compete. And since other technologies (AKA nuclear) exist which can eliminate fossil fuel burning, the full external cost of residual fossil fuel burning for backup must increasingly be assigned to solar/wind energy producers, making them even more expensive than they are already.

In order for renewables to be economically sustainable, they depend on permanent and increasing financial subsidies. The pro-renewables propaganda community today is however agitating against such subsidies, by claiming that renewables are almost at ‘grid-parity’, which would seem to imply that subsidies are no longer needed. As such, the pro-renewables propaganda community is the worst enemy of the renewables industry. If politicians believe this propaganda, they will shut down the subsidies, which will kill the renewables industry overnight.

“Bruce, cost of fuel is a liability that’s factored into an energy producer’s bid, just like the principal on the loan used to purchase a producer’s wind turbines, or the interest on that loan, or the maintenance requirements of a nuclear reactor. I don’t know why you’re breaking it down into whether the fuel is free or not, because no one who’s buying or selling energy cares.”

Except neither the principle nor the interest on the loan used to purchase a producer’s wind turbine is factored into the minimum market price at which an energy producer will sell on a marginal pricing market … just as the neither the principle nor the interest on the loan used to purchase a coal fired power plant or a natural gas turbine is factored into the minimum market price at which an energy producer will sell on a marginal pricing market.

A rational energy producer would rather earn $0.05 to help with $1 in fixed costs than to earn $0.00, so as long as the price covers the marginal cost, it is rational to produce at a loss on total cost of production.

“Both before and after a wind turbine is paid for, it’s an asset which a producer can use to generate energy. Utilities who are buying that energy pay more for more of it, just like water (or toll road access, or tomatoes). In other words: even if their expenses are not priced marginally, the product they’re selling is.”

The words you say at the beginning of this passage, before “in other words”, do not translate into “the product is priced marginally”, because “paying more for more of it” applies to regulated average cost pricing or to baseline feed-in tariffs or to technology-specific feed-in tariffs or to long term contracted prices just as much as applies to marginal pricing.

So, if I am reading you right, pricing will not close out fossil fuel electricity production. Until the cost of carbon pollution becomes really high. Therefore we need to either price carbon really high and / or legislate against GHG emissions. That allows the market / pricing to provide the incentive for CCS. And if that is not achievable then no FF’s. Time to stop pussy footing around this issue. When meat processing plants used our rivers to dispose of their wate products the result was local, visible and unacceptable. FF electricity generation waste (GHG’S) is global, invisible and unacceptable.

“The sustainable electricity component should not be so large that prices fall to zero very often.”

Precisely.

If the variables are cheaper than the dispatchables, or else the dispatchables are operating under a budget that is substantially less than 50% of total energy supply, or a combination of the two …

… then in a cost-optimized portfolio of sustainable energy, the variable renewables should be large enough that a marginal pricing market would have prices that are often falling to zero.

A retrogressive feed-in tariff for windpower avoids that problem. In a retrogressive system, the feed-in tariff declines with the amount of wind capacity already in place, so it might roughly line up with the real diminishing returns to a variable renewable as its penetration increases … but it in turn may have problems in offering commercial returns to storage or dispatchable loads in line with their benefit to the system.

A baseline feed-in tariff would pay a set feed-in tariff rate to all qualifying producers (harvest of renewables, low/no carbon producers, etc.) up to some threshold level, such as up to their average energy output … and the balance of energy produced would be sold on the marginal pricing market. That would reduce the windfall gain when selling into peak demand periods, but eliminate the “giving their lunch away to the consumers for free” effect. That would also retain a price differential for transmission, storage, dispatchable demand, and for trading off higher equipment costs for higher capacity factors, without seeing total revenues drop to negligible levels.

Bas posted almost the same comment, word-for-word, on Atomic Insights.

“Darius” is the previously-banned Bas troll.  Please fix this oversight.

“Solar/wind energy is not and never will be dispatcheable, which means their value is equal at best to the value of the fuel costs they displace.”

First, note that it is not when they were paid the value of the fuel costs that they displace that they get in trouble in the above analysis, it is when they are paid far below the value of the fuel costs they displace, because they have displaced all of them.

Second, note that the conclusion does not follow.

And third note that the full economic cost of fossil fuels is a price sufficiently high to ensure that a majority of fossil fuel reserves stay in the ground, so even if the cost of the fossil fuel that it displaced were the ceiling on the value of the variable renewables, that is a ceiling that is substantially above the breakeven replacement cost of a complement of windpower and solar PV.

“Making matters worse, solar/wind are not continuous, which means they imply chronic fossil fuel burning for backup.”

It evidently implies no such thing in general. Under the status quo conditions, we will resort to fossil fuel burning for backup to variable renewables for the same reason we would resort to even more fossil fuel burning if we did not have the variable renewables … because we allow the fossil fuel to be burned without paying for the privilege of using the atmosphere as a GHG dumping ground.

But if we assume that we will continue to allow fossil fuel to be burned without paying for the privilege of using the atmosphere as a GHG dumping ground … then we are assuming into the argument at the outset the policy stance that leads to the conclusion.

Which is begging the question … building the conclusion into the assumptions that have been made.

Not “pricing”, but rather “marginal pricing alone”.

Marginal decision making as economically optimal always rests on an assumption that the process is viable, since there is never any guarantee that marginal costs/benefits at the marginal cost = marginal benefit boundary are high enough to cover full average costs, and rests on an assumption that there is some individuals or organization engaged in design of complex systems, since complex systems cannot be designed using marginal information alone.

But assumptions made when building an economic model do not have any real world weight when a marginal pricing system is put into place unless there is some actual real world process to enforce them.

Unfortunately, the fundamental underlying real world process that enforces the termination of unsustainable systems is that “and then the system collapsed, because it was unsustainable”. If we want to avoid the pain of that process running to its conclusion, we need to add constrains on the marginal pricing that ensures that it moves toward the limits of system viability, and if we want complex new systems to be designed to replace the long term unviable systems that have already been developed, we have to make sure that the new systems that do what we want get rewarded for doing so.

Bruce, wouldn’t a carbon tax accomplish the same thing as a baseline feed-in tariff – make low/zero emissions generation competitive – and do it more efficiently? You would  have to establish a rate,  threshhold, and retrogression structure for each technology based on some calculation of “avoided carbon value” and “benefit to the system”, which will be a magnet for litigation.

Seems like addressing unemployment by paying people who find work instead of penalizing companies which move jobs overseas.

Bruce – thanks, that’s a good explanation and I see your point.

http://www.wua-wien.at/images/stories/publikationen/renewable-energy-ver...

The cost of nuclear vs. RE based on the GreenX model.

If the comment is relevant and made in good faith it is not clear to me why he should not repost it.

“Bruce, wouldn’t a carbon tax accomplish the same thing as a baseline feed-in tariff – make low/zero emissions generation competitive – and do it more efficiently?”

It wouldn’t change the way that a marginal pricing system discriminates in favor of high variable cost producers and against low variable cost producers … so a non-discriminatory pricing system would accomplish the same amount of GHG emissions reduction at a lower carbon tax rate.

Thanks.  I don’t see anything on your list I like better than the regulated public utility model for nuclear and long-term PPAs for wind & solar, but thanks.

Clayton,

To set up, with subsidiies, homeowners to charge their EVs with their wall-hung TESLA units with “excess/free” wind energy that might otherwise be curtailed (actually a very small percent of all wind energy IS curtailed) would cost about $7000 per homeowner, at the expense of those who are providing the subsidy.

More such schemes and the US would soon be uncompetitive

SHIFTING AWAY FROM LOW-COST FOSSILS TO EXPENSIVE RE

The more we shift to expensive RE, the more we shift the US wholesale price of the energy mix on the grid from the current 5 c/kWh (kept low because of an abundance of inexpensive, domestic natural gas) to about 10 – 15 c/kWh.

That trend of increasing wholesale prices would be more visible, if many of the RE changeover costs were actually charged to the US energy system. Instead, they are “socialized” by POLITICIANS by means of taxes, fees, surcharges, feed-in tariffs, bond issues, grants, etc., because they do not want to be blamed for raising the cost of electricity and harm their re-election chances.

NOTE: A perfect example of such deceptive follies is the wood chip-fired, Montpelier District Heating Plant in Vermont, a money-losing project based on taking from deluded and plucked Peter, to do favors for well-connected and well-off Paul.

Those various costs, due to increasing RE in the US, will have a MAJOR impact on making much more expensive ALL goods and services, not just energy, as is already happening in Germany, although many of its RE proponents and politicians blame it on other factors; somewhat like Miss Piggy: MOI?

In fact, rich Germany, THE economic engine of the EU, has experienced slowing economic growth, due to the growing expense of its ENERGIEWENDE, during the past five years. The economies of poorer EU countries are significantly affected by the German economic slowdown.

Germany and other EU countries losing part of the very lucrative Russian market and throwing billions each year into a black hole, a.k.a., Ukraine, is an additional headwind.

http://theenergycollective.com/willem-post/338781/high-renewable-energy-costs-damage-germanys-economy

http://theenergycollective.com/willem-post/368081/russian-gas-exports-and-western-encroachments-russia

“To set up, with subsidiies, homeowners to charge their EVs with their wall-hung TESLA units with “excess/free” wind energy that might otherwise be curtailed (actually a very small percent of all wind energy IS curtailed) would cost about $7000 per homeowner, at the expense of those who are providing the subsidy.”

Except there is nothing in what Clayton said that referred to charging EV’s with wall-hung Tesla units.

It’s, rather, about time-shifting the charging of EV’s, given that most EV’s will be plugged in for longer than required to recharge them for the power used since the previous day, and almost all pluggable hybrids will be plugged in for longer than required to fully charge them for the morning.

“The more we shift to expensive RE, the more we shift the US wholesale price of the energy mix on the grid from the current 5 c/kWh (kept low because of an abundance of inexpensive, domestic natural gas) to about 10 – 15 c/kWh.”

On the time scale of the roll-out of renewable to the levels where these factors have an impact, which is over a decade, its not likely that we will still have natural gas at $0.05/kWh. And the full economic cost of exhausting all of the “cheap” natural gas that we have available to us is likely to be catastrophic collapse of the current economic system, which some might feel to be more costly than an extra $0.05/kWh – $0.10/kWh.

As I said, I’m pretty agnostic between that model and the other three … if given a chance to move toward any of the four, I’d likely back it, unless it came with a pretty bad poison pill attached.

Changing the market design (from marginal-price bid-clearing, to one of the other suggestions, variations on regulated) can help prevent price collapse which would occur well before variable renewables reach the capacity factor limit, but they don’t solve the other underlying problem: that the output from new generators would be corrolated with that of the existing generators (adding more energy when there is too much, and adding little when more is needed).

So while the market design argument delays the effects that are cited by analysis by Hirth, the curtailment effects predicted by Kreifels (see this study) would still occur.  Hence, somewhere around the capacity factor limit, even if solar and wind continue to drop in cost, it becomes more attractive to keep burning fossil fuel (e.g. with CC&S) or burn more biomass rather than adding more solar and wind along with the needed storage, transmission, and curtailment.