The Cost of Filling Up the Tank with Electricity
Why Californians are getting gouged at the plug.
The summer driving season is underway. We’re hearing a lot about the high cost of gasoline, nationally and in California. In relative terms, electricity to charge electric vehicles can look like a bargain. But if you dig into the numbers, utilities’ rates may be much higher than they need to be. Getting the rates right needs to be an important part of the state’s plan if consumers are to meet the California’s zero-emission vehicle target by 2030.
To start with let’s compare electricity and gasoline prices today using a recent briefing note from the International Council on Clean Transportation (ICCT). An electric vehicle owner in southern California who charges at home and has signed up for a time-of-use rate plan pays 12 cents per kilowatt-hour at night. The ICCT argues that the cost to fuel an electric vehicle should be compared to the sort of highly efficient gasoline vehicle that the driver most likely would have purchased. Following their guidance, the southern California electric vehicle owner is paying the equivalent of just $1.85 per gallon. Things are looking pretty sunny for this driver.
While it may look like California electric vehicle drivers are getting a great deal, they aren’t necessarily. The prices paid by California’s electric vehicle drivers are higher than they should be. This is likely slowing electric vehicle adoption.
The California Public Utilities Commission Tackles Rates
The state’s utilities regulator, the California Public Utilities Commission (CPUC), is not traditionally an agency focused on transportation policy. However, the agency is taking a turn behind the steering wheel to help the state accelerate towards an electrical transportation system. The CPUC is starting by focusing on charging infrastructure. Last month the CPUC approved utilities’ requests to invest over $700 million in electric vehicle charging infrastructure.
The CPUC is now taking a look at the prices that consumers pay to charge up their vehicles in their homes and at commercial facilities. This review kicked off at a forum earlier this month, starting a process that will be very important to watch, for consumers and the power system more broadly.
Forum presenters described how the prices that households and businesses pay for electricity are set through decades-old, complex regulatory processes overseen by state utility regulators. The recipe involves financial analyses, equity considerations, policy priorities and politics, which are all stirred and mixed until, voila, out comes a set of retail rates. Economic principles are sometimes sprinkled in during the process, but are hard to discern in the final concoction. Most troubling, the process results in a wide gap between the prices that households pay for electricity and the cost to produce the electricity.
The blue, yellow and green lines show the averages of the hourly prices across every day in 2017. Prices tend to follow shape of daily demand, net of renewable energy production, shown in red. SOURCE: CAISO, 2017 Annual Report on Market Issues & Performance
According to data reported in June by the California Independent System Operator (CAISO), wholesale electricity prices averaged a just $43 per megawatt-hour ($0.66 per gallon equivalent) in 2017. Nighttime prices were even lower, as shown below in a chart from the CAISO. That means southern California electric vehicle drivers were paying at least triple the cost of producing electricity.
Illuminating new research by Severin Borenstein and James Bushnell, presented at the Energy Institute’s 2018 POWER Conference analyzes the gap in detail across the entire US. They point out that a full comparison between wholesale and retail prices needs to also consider the environmental damage caused by power generation. California’s power plants are not very polluting, so this doesn’t add much to the wholesale costs. They emphasize that the analysis also needs to account for electricity that is lost during the electricity delivery process. These costs are important, but aren’t big enough to justify the existing large gap.
The gap is actually much larger than triple for many consumers. At the forum, we heard that an estimated three-quarters of electric vehicle owners haven’t even enrolled in a time-of-use rate that lets them pay lower prices at night. Instead they are on the standard residential pricing plan. For customers of the two largest utilities these standard plans charge between 25 and 28 cents per kilowatt-hour, that’s equivalent to $3.85 to $4.35 per gallon. It’s even more extreme in San Diego where households that use more than 130% of a baseline amount could be paying as much as 48 cents per kilowatt-hour or the equivalent of $7.42 per gallon. The low adoption of the less expensive time-of-use rate shows the downside of depending on optional rates.
How California Sets Rates
Why are retail prices so much higher than wholesale electricity prices? At least two reasons. First, the utilities collect a lot of revenue to pay for all of their costs. These include the costs to build, operate and maintain its infrastrastructure, interest to lenders and a rate of return to shareholders. The costs of state-mandated programs — such as energy efficiency, subsidies for low-income households, and some early investments that were made in new renewables technologies — are also significant in California.
Second, California’s policymakers have decided that for residential customers, virtually all of the revenue will be collected from each customer in proportion to their electricity use. The alternative would be to collect some of the revenue through a fixed monthly fee.
Many other services that households depend on such as garbage and recycling collection, home internet and streaming music take the opposite approach. Customers pay a fixed monthly fee for a certain size garbage bin or level of service, but aren’t charged per pound of garbage, per Gigabyte downloaded or per song listened to. Curiously, policymakers have taken the opposite approach for electricity.
Electric vehicles highlight the downside of the way California sets residential electricity rates. Households are paying far more to charge their vehicles at home than it costs to generate and deliver the incremental electricity, even when the additional pollution is priced in. This disconnect has the potential to slow electric vehicle adoption or, at the very least, raise the cost of meeting the state’s electric vehicle goals.
Closing the Gap
Broadly speaking there are two ways to close the gap between retail and wholesale prices. One is to lower the utilities’ revenue needs. The other is to restructure rates by increasing fixed monthly charges and lowering the amount collected for each kilowatt-hour consumed. Both would be difficult.
Any move to restructure rates will need to pay attention to distributional impacts and fairness. As prior Energy Institute analysis has shown, subsidies for electric vehicles are heavily skewed toward the wealthy. Regulators should keep these considerations in mind.
While challenging, rate reform could generate big returns for the state’s environment and economy. It’s encouraging that the CPUC is looking at this issue.
Categories
Andrew G Campbell View All
Andrew Campbell is the Executive Director of the Energy Institute at Haas at the University of California, Berkeley. At the Energy Institute, Campbell serves as a bridge between the research community, and business and policy leaders on energy economics and policy.
Energy Institute Blog 
Your analysis fails to include the (1) EV state and federal tax credits, (2) any of the utility and state subsidies for utility, commercial and home energy charging stations, (3) EV avoidance of gas taxes for building and maintaining roads, (4) utility system upgrades to the local distribution and other control systems necessary to accomodate EV demand requirements, and (5) costs to address public safety (fire and ambulance) equipment and procedure upgrades . All of these five categories contribute to the EV “cost per mile” evaluation. Focusing only on gasoline prices and utility rates substantially understates and misstates EV costs.
The cost of filling up the tank with electricity from the grid should include a Power Quality Rating (PQR) that considers harmonics, power factor, balance across the phases, etc at the time of fill- up. “Reactive power is considered to be an extremely negative phenomenon. Huge resources are being wasted to withstand the reactive power in power lines today.” PQR is the best way for the State to lower emissions based on a PQR pricing plan. The closer the PQR is to 100% the more “wattful” the power, the less heat in joules, and the better the rate design should be in for truly efficient fill-ups. PQR is not just for EV’s lowering emissions, PQR is for all connected to a AC grid to lower emissions.
Tucker,
How would we price power quality?
Reactive power is value is highly localized, which means that prices will vary by the customer’s location on the distribution grid.
A study done by Tabors Carimanis for the NY REV Proceeding included a model that calculated locational, time-dependent retail electric energy prices which included the effects of reactive power. This is the one of the most innovative pieces of work that I have seen in years – yet it has been largely ignored.
Google:
White Paper on Developing Competitive Electricity Markets and Pricing Structures, Prepared for: New York State Energy Research and Development Authority (NYSERDA) and New York State Department of Public Service, Prepared by:
Tabors Caramanis Rudkevich (TCR), April 2016 NYSERDA Contract: 64271
mcubed,
“First, if short-run energy prices represented the long-run market values, then the California utilities would have never built any of the fossil plants that were authorized between 2002 and 2010.”
In ISOs with capacity markets (e.g, PJM) generators do not have to rely on energy revenues to recover their investments; they expect to make up the shortfall through the future capacity payments they will receive. Although the CAISO has no capacity market, all of the generators are either rate-based (thus guaranteed to recover their costs if they were prudently incurred) or are under long-term contracts with the utilities.
In the US only ERCOT requires generators to fully recover their costs through energy (and ancillary services) revenues. But ERCOT has a scarcity pricing mechanism in place that allows energy prices to go as high as $9,000/kWh).
So you first statement is clearly wrong and the rest of that paragraph is irrelevant.
“I know exactly why Diablo, the QFs and the other resources you list continue to run despite NOT being economic run within the CAISO merit order stack.”
Who said it is not economic to run them? They’re infra marginal units so it is economic to run them. The rest of this paragraph is so convoluted I don’t want to even try to unravel it. But the last two sentences (quoted below) are interesting enough to be worthy of a reply.
“Again, you haven’t answered why we don’t do the same in the housing market by using the nightly rental rate on AirBNB for a spare bedroom to value a house? The answer is obvious for the same reason it’s obvious that the CAISO imbalance market fails to provide the full value of the power asset.”
A spare bedroom rented out as an AirBNB is a byproduct. The rental revenues simply make a contribution to the owner’s cost of home ownership. But suppose the owner considered buying a second house with the intention of dedicating it to AirBNB rentals. In that case the expected future rental income would, in fact, determine the value of the house, at least as a rental property.
Your analogy is a good one. The CAISO “imbalance market” (a poor label since the CAISO dispatches all generators and runs a settlements system that includes all of those generators, even the ones with long-term contracts for their full output. But, as I have already said, the generators do not depend on their energy revenues alone to determine their market values because of the makeup payments they receive from their contracts or from being rate based. That doesn’t mean that the CAISO real-time LMPs do not closely represent the marginal cost of energy.
“I’m not aware of any other states getting ahead of California on residential rate design….”
Look up the tariffs of Commonwealth Edison and Oklahoma Gas & Electric. Both have offered true dynamic rates for years. Commonwealth has a voluntary hourly energy rate that is indexed to the hourly day-ahead LMPs. Not quite real-time rates but pretty close. And the customers who have selected this tariff love it.
As for residential TOU rates, many utilities outside of California have offered them for years, e.g., Arizona Public Service, Connecticut P&L, Florida P&L, Georgia Power, Central Maine Power, Detroit Edison, Northern State Power, Nevada Power, and Duke Energy Carolinas. This is only a partial list drawn from a study I did last year.
You lament, just if the utilities could have just passed through all of their exorbitant costs while doing nothing to mitigate those costs, it all would have been better (or at least the utility shareholders would have avoided bankruptcy), even if it would have ruined the state economy.”
Pray tell, what could the utility have done to mitigate their costs of having to buy virtually all of their energy from the wholesale spot market? Nothing! However, if the CPUC has allowed them to pass through those costs their retail customers would have cut back on their usage, just as they did when the CPUC finally did (belatedly) raised the retail rates. The reduced usage would have caused wholesale market prices to fall and would have imposed a discipline on the rapacious generators (most of whom were Texas-based).
And it seems to me that the $7 billion hit California took came damn close to ruining the state economy so the CPUC wasn’t much help was it.
“No, the real cause was the failure of the CPUC (and more importantly, the State Legislature in June 2000) to allow the utilities to sign long term bilateral power contracts….”
Now here we do agree. Forcing the utilities to buy from the spot market (presumably in order to ensure market liquidity) was insane. I suspect that Jeff Skilling had a hand in that decision, just as he did in convincing California to adopt an inefficient market design that prohibited the CAISO from dispatching all but a few “load-balancing” generators in economic order. But once that decision was made the CPUC could still have mitigated the damage by raising the price caps on retail rates.
So on that positive note, let’s call it a day and stop sniping at each other.
Robert, you wrote “Although the CAISO has no capacity market, all of the generators are either rate-based (thus guaranteed to recover their costs if they were prudently incurred) or are under long-term contracts with the utilities.” So you acknowledged I am correct–the CAISO market does NOT reflect the full market value of generation. California needed to use separate transactions to bring new resources on line. That value is reflected in a separate set of markets, not the CAISO market. So obviously the rest of my discussion about other hidden transactions is on point. (Given that there have been several utility bankruptcies in Texas, I doubt that California wants to follow that model. And that would require a fundamental change in California law in repealing AB 57.)
QF short run avoided cost (SRAC) payments are often well above the CAISO price, yet they still are must run. They are not “infra marginal” by your definition of costs being below the MCP. They are simply self scheduled at zero because the market transaction that set their price occurred in a different marketplace. Again, the CAISO imbalance market price is irrelevant to valuing that resource.
On AirBnB, few would value an entire house based on the AirBnB rental rate for a spare bedroom (and it is only the spare bedroom, not the entire house, that is being bid on the CAISO imbalance market.) A spare bedroom has very different characteristics of the rest of the house and excludes the kitchen, bathrooms and garage.
And “imbalance market” is the correct label for the CAISO energy markets. It’s that it highlights my point is what you have a problem with. As I have repeatedly pointed out, only a small portion of the generation portfolio is actually bid into the imbalance market rather than self scheduled by the LSEs in California. As I pointed out, Europe treats these transactions for what they are–transmission scheduling mechanisms, not product valuation.
On residential rates, California also has offered VOLUNTARY residential TOU rates for decades. That’s not different than the other states that you list. However, California is among the first moving to DEFAULT TOU rates.
“Pray tell, what could the utility have done to mitigate their costs of having to buy virtually all of their energy from the wholesale spot market?” You ignore that the IOUs cooked themselves into their own soup when they helped set up restructuring. First, they could agreed to use long-term sales and contracts to set the market value instead of the PX daily price (note, it wasn’t the CAISO price to start). The round of plant auctions in 1997 and 1999 showed that the true values were substantially higher than what the short-term markets were showing. That would have ended the CTC period much sooner and the IOUs could have signed bilateral contracts that greatly reduced their exposure. Second, they could have valued their portfolios to end the CTC earlier and again would have been freed sign bilaterals that avoided the market exposure. PG&E’s hydro system was valued at $2.8 billion in May 2000 and that would have fulfilled PG&E’s remaining CTC balance at that point. That is right before the beginning of the crisis. The incentive to manipulate the NP15 market would have disappeared (and likely the SP15 market would have been favorably protected.) So clearly, the IOUs had these tools, and also had agreed to take on the ENTIRE risk of collecting the CTC at the outset. PG&E went back on that deal in March 2001 and a weak governor let them get away with it.
“And it seems to me that the $7 billion hit California took came damn close to ruining the state economy so the CPUC wasn’t much help was it.” Yes, the CPUC wasn’t much help, as I pointed out that the CPUC allowed the IOUs to pass through their exorbitant costs. So you agree with me that it was poor public policy to allow that cost pass through instead of having shareholders to absorb that cost as they agreed to do in AB1890.
Mcubed, you said:
“So you acknowledged I am correct–the CAISO market does NOT reflect the full market value of generation. California needed to use separate transactions to bring new resources on line. That value is reflected in a separate set of markets, not the CAISO market. So obviously the rest of my discussion about other hidden transactions is on point.”
Our initial difference of opinion was over whether the CAISO LMPs represent the marginal cost of energy. That is totally different issue than determining the market value of a generator. Except in ERCOT, energy revenues (net of variable running costs) are only one factor contributing to a generator’s market value. So you are beating a dead horse.
“QF short run avoided cost (SRAC) payments are often well above the CAISO price, yet they still are must run. They are not “infra marginal” by your definition of costs being below the MCP. They are simply self-scheduled at zero because the market transaction that set their price occurred in a different marketplace. Again, the CAISO imbalance market price is irrelevant to valuing that resource.
It may be true that a QF’s SRAC is above CAISO LMPs. But if the utility has a take-or-pay contract with the QF the marginal cost to the utility IS zero so that is the logical price to offer into the CAISO market. Thus the QF is truly an infra marginal unit. Indeed, many QFs are must-run units because the byproduct they produce with the waste heat cannot be varied (e.g., district heating). But, again, my comments had nothing to do with valuing the QF.
I’ll skip the AirBNB analogy because It sheds little light on our difference of opinion.
“As I have repeatedly pointed out, only a small portion of the generation portfolio is actually bid into the imbalance market rather than self scheduled by the LSEs in California.”
So are you saying that the CAISO has no control over these resources? Then how can it be sure that the total of the LSE schedules don’t violate some transmission system constraint?
What you have just described is the arrangement California originally implemented using Scheduling Coordinators to self-schedule their units into the PX. Back then the CAISO only determined whether the totality of these schedules constituted a feasible solution. If it didn’t the CALISO sent it back to the SCs for modification and this cumbersome, iterative process continued through multiple cycles until a feasible solution (but generally not an economically optimal) solution was discovered. The CAISO was prohibited from modifying the schedules even if it identified an uncommitted lower cost resource that could be dispatched instead of a higher cost one that was scheduled to run. All of that changed when the CAISO market was restructured following the 2000-01 meltdown. Today the CAISO has control over all of the units that are committed to run.
“You ignore that the IOUs cooked themselves into their own soup when they helped set up restructuring. First, they could agreed to use long-term sales and contracts to set the market value instead of the PX daily price (note, it wasn’t the CAISO price to start).”
I agree with you on this point, The IOUs should never have accepted the obligation to buy most of their requirements from the day-ahead or real time spot markets. But that’s hindsight. Nobody anticipated the perfect storm that occurred in 2000, which was brought about by the confluence of a number of facts and then exploited by the unscrupulous, anti-competitive practices of ENRON, El Paso Gas, and a number of generators. The carrot offered to the IOUs was that their retail rates would be fixed at above-market levels to allow them to recover their stranded costs, many of which were created by the CPUC (e.g. QF procurements at above-market avoided costs administratively set in Standard Offer No. 4). Once those stranded costs were fully recovered the cap on retail rates would disappear and rates would again be cost-based.
“Second, they could have valued their portfolios to end the CTC earlier and again would have been freed sign bilaterals that avoided the market exposure. PG&E’s hydro system was valued at $2.8 billion in May 2000 and that would have fulfilled PG&E’s remaining CTC balance at that point.”
I question whether ending the CTC would have allowed the IOUs to sign bilaterals. My understanding (which may be wrong) is that the prohibition against contracting long was to ensure sufficient liquidity in the short-term market. If so, that was a structural requirement that would not go away with the CTC.
“So clearly, the IOUs had these tools, and also had agreed to take on the ENTIRE risk of collecting the CTC at the outset. PG&E went back on that deal in March 2001 and a weak governor let them get away with it.”
Again, I doubt that the IOUs had the tools you described but they did agree to take on the entire risk of collection. As a result, PG&E was technically bankrupt in 2001 and SCE was not far behind. That’s not my idea of someone going back on a deal; you can’t continue giving something that you no longer have.
Gray Davis wasn’t a weak governor who let PG&E get away with anything. He was a stupid governor who publicly acknowledged that he could end the crisis by raising retail tariffs but refused to do so until there was no other choice.
“So you agree with me that it was poor public policy to allow that cost pass through instead of having shareholders to absorb that cost as they agreed to do in AB1890.”
Hell no, I don’t agree with you. The public was already suffering dearly from rolling blackouts caused by PG&E’s inability to purchase energy. While higher electricity prices suck, what sucks even more is not having electrictricity. The shareholders gave just about all they could.
The California meltdown could have been greatly mitigated with demand response driven by exposing retail customers to the wholesale market prices. Which brings us full circle to the retail tariff design I described way back in our exchange of views.
Robert, you wrote “Our initial difference of opinion was over whether the CAISO LMPs represent the marginal cost of energy. That is totally different issue than determining the market value of a generator.” In California in several CPUC proceedings, and the recent series of blogs by different individuals on this website, the two have been equated. It’s as the core of the utilities’ proposal in the PCIA OIR. And by saying that rooftop solar is not competitive with the LMPs, they are saying that the VALUE of rooftop solar is set by the LMP. This whole debate is about the market value of solar rooftop, not where should it be dispatched in the CAISO DA/RT stack. So we are back to the issue that I raised.
“Nobody anticipated the perfect storm that occurred in 2000, which was brought about by the confluence of a number of facts and then exploited by the unscrupulous, anti-competitive practices of ENRON, El Paso Gas, and a number of generators.” Catherine Wolfram did in her 1997 paper that provided the recipe for withholding that the merchant generators used to manipulate the market.
The CAISO can control resources without them being bid into the market. Submitting a schedule that can be changed (often without regard to economics–see renewables curtailment) and a bid with prices are two very different actions economically. The “contracts for differences” in these PPA scheduling arrangements essentially bifurcates the markets.
You’re arguing both sides on the QF. Either 1) if its operating cost is greater than the MCP then it shouldn’t be run, or 2) as a must-take its market value is set in a different market place.
“I question whether ending the CTC would have allowed the IOUs to sign bilaterals. My understanding (which may be wrong) is that the prohibition against contracting long was to ensure sufficient liquidity in the short-term market. If so, that was a structural requirement that would not go away with the CTC.” If you were familiar with the law, you would know that they were free to sign bilaterals as soon as their CTC collection ended. That’s why SDG&E was in much less trouble. The prohibition only applied to those utilties still collecting the CTC. (It was in a PUC section added as part of the 2000-01 budget trailer.)
“As a result, PG&E was technically bankrupt in 2001 and SCE was not far behind. That’s not my idea of someone going back on a deal; you can’t continue giving something that you no longer have. Gray Davis wasn’t a weak governor who let PG&E get away with anything. He was a stupid governor who publicly acknowledged that he could end the crisis by raising retail tariffs but refused to do so until there was no other choice.” Asking for a rate increase to cover the $7 billion instead of having shareholders absorb it is going back on the deal. PG&E could have gone to zero value. There’s a lot of history that I won’t go into here in which its clear that the IOUs took actions that avoided accepting the risk. Bankruptcy means that shareholders lose all value–that didn’t happen at all. Davis had a series of choices to be made before the retail rate increase in April 2001 (many of which I participated in). And perhaps the best would have been for the state to take over PG&E (for a nominal payment) and restructure the utility. (And yes, that option was available at the time.)
“The shareholders gave just about all they could.” No, the shareholders got back all of their power purchase costs through the rates. In true bankruptcy, shareholders would have eaten that cost (which happened to be almost exactly equal to the amount of transition costs that were to be collected through the CTC).
“The California meltdown could have been greatly mitigated with demand response driven by exposing retail customers to the wholesale market prices. Which brings us full circle to the retail tariff design I described way back in our exchange of views.” That required a wholesale rate design that had never been undertaken up to that point, and was not technologically feasible at a reasonable cost in 1997 for residential customers. So your solution wasn’t even feasible at the time.
Mcubed, My final responses:
“Robert, you wrote “Our initial difference of opinion was over whether the CAISO LMPs represent the marginal cost of energy. That is totally different issue than determining the market value of a generator.” In California in several CPUC proceedings, and the recent series of blogs by different individuals on this website, the two have been equated. It’s as the core of the utilities’ proposal in the PCIA OIR. And by saying that rooftop solar is not competitive with the LMPs, they are saying that the VALUE of rooftop solar is set by the LMP. This whole debate is about the market value of solar rooftop, not where should it be dispatched in the CAISO DA/RT stack. So we are back to the issue that I raised.”
The market value of rooftop solar IS almost exclusively determined by the marginal cost of energy at the solar customer’s meter in the hours of solar production and the CAISO LMPs are the starting point for calculating those marginal costs. The LMPs must be grossed up for avoided distribution system losses, which is a small adjustment. Also, the market value of the avoided generation capacity costs must be imputed to the solar facility. This is also small given that solar output is intermittent and residential loads peak in the mid- to late- afternoon when solar production is already falling off.
Lastly, as I suspect you know, rooftop solar is not dispatched by either the CAISO or the utility so they appear only as a reduction in net load. That explains why they do not appear in the DA/RT stack. More fundamentally, my earlier comments regarding the CAISO dispatching all wholesale generation was meant to contradict your assertion that only a few generators participated in “imbalance” market. That’s appears to be how our discussion got off track.
“ Nobody anticipated the perfect storm that occurred in 2000, which was brought about by the confluence of a number of facts and then exploited by the unscrupulous, anti-competitive practices of ENRON, El Paso Gas, and a number of generators. Catherine Wolfram did in her 1997 paper that provided the recipe for withholding that the merchant generators used to manipulate the market.”
Good for Catherine. And for the other people that voiced concern over the original market design, including Bill Hogan and Larry Ruff, both of whom were colleagues of mine. Obviously, the California legislators didn’t listen. But, I admit, my statement was too absolute.
Even so, despite the flawed market design, the 2000-01 meltdown would not have occurred if California didn’t have several back-to-back dry hydro years combined with higher than anticipated load growth and other factors that created the perfect storm. The combined effect was to greatly increase the mark power of the generators, even if they have not colluded and even if El Paso hadn’t removed about one-third of the pipeline capacity delivering natural gas to California.
“The CAISO can control resources without them being bid into the market. Submitting a schedule that can be changed (often without regard to economics–see renewables curtailment) and a bid with prices are two very different actions economically. The “contracts for differences” in these PPA scheduling arrangements essentially bifurcates the markets.”
I think our disagreement here is over semantics. Offering a must-run schedule to the CAISO is effectively the same as offering a zero price. Obviously the CAISO has the authority to change any schedule if a system constraint is about to be violated.
I don’t see where CFD’s have any effect on the CAISO, which (most likely) doesn’t even know the details of these contracts – nor does it care.
“You’re arguing both sides on the QF. Either 1) if its operating cost is greater than the MCP then it shouldn’t be run, or 2) as a must-take its market value is set in a different market place.”
There is no contradiction in those two statements but maybe I was unclear in what I meant to say.
Any resource whose variable running rate is higher than the MCP should be backed off or shutdown and the curtailed energy should be purchased from the spot wholesale market because that will save money. The key issue is who offers the unit into the wholesale market, the unit owner or the utility? If the utility controls the unit and has a take-or-pay contract its costs are higher if it curtails the unit output. Also many (most?) QF’s don’t have the flexibility to change output on command for reasons I cited earlier.
“I question whether ending the CTC would have allowed the IOUs to sign bilaterals. My understanding (which may be wrong) is that the prohibition against contracting long was to ensure sufficient liquidity in the short-term market. If so, that was a structural requirement that would not go away with the CTC.” If you were familiar with the law, you would know that they were free to sign bilaterals as soon as their CTC collection ended. That’s why SDG&E was in much less trouble. The prohibition only applied to those utilties still collecting the CTC. (It was in a PUC section added as part of the 2000-01 budget trailer.)”
As you can see, I conditioned my statement because I was not familiar with all of the details regarding CTC recovery. However, signing bilaterals would have done little to help the utility if I . s retail tariffs remain capped. Furthermore, once the cap was lifted the utility would have been fine even if it continued purchasing from the wholesale spot market because it would have been allowed to directly pass those costs on through higher tariffs.
SDG&E got out of trouble by getting a large rate increase from the CPUC when its CTC was fully amortized – not by signing any long-term power contracts (which were exorbitantly priced by the way).
“Asking for a rate increase to cover the $7 billion instead of having shareholders absorb it is going back on the deal. PG&E could have gone to zero value.”
PG&E was on the way to zero value.
But I don’t think you appreciate how disruptive it is to go through bankruptcy proceedings. Most likely PG&E’s electricity consumers would have suffered from service interruptions. Furthermore, investors would have lost all confidence in the regulatory environment and would have demanded large risk premiums for any new investments in California IOUs. Who do you think would pay those risk premiums?
“There’s a lot of history that I won’t go into here in which its clear that the IOUs took actions that avoided accepting the risk. Bankruptcy means that shareholders lose all value–that didn’t happen at all. Davis had a series of choices to be made before the retail rate increase in April 2001 (many of which I participated in). And perhaps the best would have been for the state to take over PG&E (for a nominal payment) and restructure the utility. (And yes, that option was available at the time.)”
That may be true but we will never know how exercising that option would have turned out. The one good outcome might have been that the CPUC would no longer have jurisdiction over the utility’s rates. That’s one way to limit damage control.
“The shareholders gave just about all they could.” No, the shareholders got back all of their power purchase costs through the rates. In true bankruptcy, shareholders would have eaten that cost (which happened to be almost exactly equal to the amount of transition costs that were to be collected through the CTC).”
When did they fully recover those costs? Retroactively? I am not aware of that happening. If that’s true I stand corrected. Please provide a reference that confirms your statement.
“The California meltdown could have been greatly mitigated with demand response driven by exposing retail customers to the wholesale market prices. Which brings us full circle to the retail tariff design I described way back in our exchange of views.” That required a wholesale rate design that had never been undertaken up to that point, and was not technologically feasible at a reasonable cost in 1997 for residential customers. So your solution wasn’t even feasible at the time”
My comment was referring to the ideal rate design that is feasible TODAY – not back in 1997.
However, California did fund the installation of close to 25,000 smart meters at industrial and commercial customer sites around 1999. Just putting those customers on dynamic rates would have greatly facilitated demand response and reduced wholesale prices. The stupidity of California energy policy is displayed by its willingness to spend taxpayer money on smart meters then not use them to their fullest capacity. Unbelievable!
Lastly, just lifting the cap on residential rates would have induced substantial reductions in demand, which is exactly what happened when the CPUC did finally allow rate increases.
Mcubed, I think we have exhausted his subject. Let’s end this colloquy.
“More fundamentally, my earlier comments regarding the CAISO dispatching all wholesale generation was meant to contradict your assertion that only a few generators participated in “imbalance” market. That’s appears to be how our discussion got off track.” Yes, our difference is over this factual issue. You need to provide empirical evidence that shows that the CAISO strictly follows merit-order dispatch for all resources, and that gas fired generation is used to meet incremental load growth (vs. balancing of intrahour transmission scheduling). I’ll note that PG&E met its reduced bundled load requirement in 2017 almost entirely through reduced CAISO purchases.
“My comment was referring to the ideal rate design that is feasible TODAY – not back in 1997.” But that portion of our discussion was entirely in the context of what could be done during the 2000-01 crisis, so I’m not sure how that is relevant.
“Just putting those customers on dynamic rates would have greatly facilitated demand response and reduced wholesale prices. The stupidity of California energy policy is displayed by its willingness to spend taxpayer money on smart meters then not use them to their fullest capacity.” As I’ve found out working for large agricultural customers, those meters are much less capable than advertised by the IOUs. Also, the billing system has to also be capable–it wasn’t in 2000.
“Lastly, just lifting the cap on residential rates would have induced substantial reductions in demand, which is exactly what happened when the CPUC did finally allow rate increases.” The cap was in place as a trade off for funding the IOUs stranded costs. Your proposed solution is incomplete as that it doesn’t provide to the retail customers what they would have gotten in return for lifting the cap. Again, you need to show how utility shareholders would bear a much higher risk than what they did, and what they do today. That failure to bear much risk is at the heart of the problem we face today.
“Even so, despite the flawed market design, the 2000-01 meltdown would not have occurred if California didn’t have several back-to-back dry hydro years combined with higher than anticipated load growth and other factors that created the perfect storm.”
I testified on much of this point in the FERC docket on behalf of the California Parties. 2000 was NOT an unusually dry year (it was 90% of average or Below Normal in DWR parlance), 1999 had been an average year, and there was a typical amount of hydropower available on the West Coast. In 2001, AFTER the crisis was resolved by March 2001, there was dry year, yet the market issues disappeared entirely by June 2001. In addition, there was not an unusual amount of load growth. All of that just repeats that fallacious stories presented by the merchant generators in the case. ALL (100%) of the crisis can be attributed to the withholding strategies exercised by the merchants, and was fully demonstrated in the testimony filed by the California Parties. The merchants had already been experimenting in 1999. A study I conducted in April 2000 as part of the hydro divestiture case for the CPUC found that the merchants could exercise significant market power in several thousand hours of the year across a range of hydro conditions. None of what you say here is true related to the cause of the crisis.
Added note: In Water Year 2000, the Columbia River was 101% of average. http://www.nwd-wc.usace.army.mil/PB/PEB_08/docs/WaterMan/WaterMan00/Table%2009.pdf
“I don’t see where CFD’s have any effect on the CAISO, which (most likely) doesn’t even know the details of these contracts – nor does it care. ”
Precisely my point, and the CAISO price is irrelevant to the parties of this transaction for determining market value.
As to whether LMPs or MCPs set the market value, I can’t tell which way you’re spinning on that issue, so I’ll just leave it to my statements about how the CAISO MCP is increasingly irrelevant to determining a market value (there is still a small portion of relevance).
I’m a longtime energy nerd and have a MS in Management Science from Cal and live in PG&E territory. I’ve had PV on my home since 2007 that produced about half of the KWh we consumed before I bought an EV last November. I’m on the E6 time of use rate plan that is favorable to solar households.
Despite having reams of data on my PV system’s average past production — down to the hourly level — it wasn’t possible for me to confidently predict whether switching to the EV-friendly rate plan would save me money or not. I did the calculations based on EXPECTED EV miles driven (about 6000/year), and it wasn’t clear which plan was better for me. After I’ve had the car for a year I’ll run the numbers again and see what I can determine based on ACTUAL miles driven and actual miles/kWh.
I’m also fortunate to live close to a city-owned garage where I can charge for up to 2 hours at a time for $.16/kWh, which is a bit lower than the E-6 Tier 1 off-peak rate. Other EV owners may have access to other bargain rates from employers, public agencies, etc.
Fixed monthly fees are one solution, but more generally you just need to increase inframarginal rates and lower marginal ones. Tiered pricing is already widely practiced for different categories of consumers. You just have to add more categories, especially for residential consumers. Just make sure that the categories are not defined by actual consumption. I’m sure that a profession steeped in instrumental variables can come up with a suitable proxy for (efficient) consumption. I suspect the biggest problem would be legal challenges from those who dispute the category they were placed in.