Breaking News! California Electricity Prices are High

In case you missed it, a recent investigative piece in the LA Times unearthed the shocking fact that California retail electricity prices are high,  about 50% higher than the national average. The article’s main focus is on the fact that California has a lot more installed nameplate generation capacity then has historically been the norm. There are several causes identified in the piece.  Deregulation of the market in the late 1990’s is pointed to as a culprit. Somewhat inconsistently, the construction of regulated, rate-based plants also takes much of the blame. One factor that was barely mentioned, however,  was California’s renewable electricity policy.

The story of how California’s electric system got to its current state is indeed a long and gory one going back at least to the 1980’s. The system still suffers from some of the after effects of the 2000 era crisis.  The Long Term Procurement Process (LTPP) put in place in the wake of the crisis, and overseen by the CPUC, has been criticized from many sides.

However, since the power crisis of the early 2000’s settled down, the dominant policy driver in the electricity sector has unquestionably been a focus on developing renewable sources of electricity generation. As is well known (outside of the LA Times apparently), California has one of the country’s most aggressive renewable portfolio standards (RPS).  The RPS requires each firm that sells electricity to end-users to procure an increasing fraction (33% by 2020, 50% by 2030) of the energy they sell from renewable sources.

Desert Solar Array
The Times article’s focus on generation capacity does (a bit unwittingly) provide a nice starting point for a discussion about the cost and implications of this renewable energy policy. The policy, while undoubtedly effective at reducing the carbon intensity of the power sector, has also been quite disruptive to the economics of the sector.  It is forcing a rapid (and early) replacement of conventional sources with renewable, but variable, generation sources such as solar and wind. Since 2010, about 80% of new capacity has come from renewable sources and it’s likely that much of that capacity would not have been built if not for the RPS.  (Much of the remaining 20% has been coming online to replace the retired SONGS nuclear plant or capacity slated for retirement due to environmental issues with their water cooling  processes.)

newcap2010

New Capacity in California ISO by fuel type

Proponents of strong renewable standards have pointed to the fact that new contracts for renewable energy carry price tags that are (at worst) only modestly above those for a new conventional natural gas power plant. However comparing the cost of a brand-new solar plant to that of a brand-new gas plant overlooks two important facts.  First, renewable, variable output sources offer very different operational capabilities than conventional sources.  Second, right now we don’t really need new capacity of any kind, and are in fact struggling to find ways to compensate the generators that are already here.

The renewable portfolio standard provides an interesting contrast to the federal mileage standards on vehicles. Both require the replacement of older legacy, high-carbon sources with newer,  lower-carbon ones. However automobile standards work by requiring people to buy more fuel-efficient cars when they decide to buy a new car. Renewable portfolio standards require utilities to buy low-carbon energy by a certain deadline rather than when they are deciding to “trade-in” their old power plants. In California at least, the result has been a much more rapid turnover of legacy sources to the newer, cleaner ones. Another implication, however is the fact that the system now has a large amount of what can appear to be excess capacity. This is because renewable policies are rapidly forcing new “green” capacity into a market that was more or less fully resourced before the mandates really started taking effect.

I don’t mean to imply that the “replace it now” approach is definitively worse.  Research has shown that standards applied only to new purchases can inefficiently extend the lifetimes of older technology, from cars to power plants.  This can significantly dilute the environmental benefits of a technology mandate.  In contrast, instead of extending the lifetimes of old plants, the RPS is in effect forcing the early mothballing of legacy capacity. This improves the environmental impact, but also increases costs, sometimes in subtle ways. The effect grows larger with stricter mandates. At higher percentages, the RPS starts to displace increasingly newer (and cleaner) sources of generation.  The economic effects can be mitigated by allowing for renewable energy generated elsewhere in the country to count toward RPS compliance, but California has largely rejected such policies.

Largely due to the RPS, we have a surge of new, low marginal cost energy, flooding into a wholesale market that already had enough generic energy, thereby driving down wholesale prices. Since wholesale prices cannot support the cost of this much generation (new and old), increasingly the gap must be  made up through rising margins between wholesale and retail prices.  Utilities and other retailers have to pay  high market prices for new renewables instead of being able to “buy low” on the wholesale market.  Because all retailers face the same regulation, they pass these costs on to end users. And this doesn’t even consider the costs of new transmission, most of which is being added to boost the power system’s ability to access and absorb large amounts of renewable energy. Transmission costs, which are also charged through to electricity end users as part of the retail prices cited in the Times article, will continue to grow in coming years. The Tehachapi transmission project alone is projected to cost over $2 Billion.

The result is the seemingly perverse situation where customer rates are rising while (conventional) generation sources are simultaneously struggling for revenue and threatening to retire. Such conditions are a recurring theme on this blog and are often drivers of significant change. Unfortunately, despite the glut of electrical energy, we will likely still need the conventional capacity to handle the ramping and back-up needs created by the increased reliance on variable sources (wind and solar).

One of the debates lurking in the background is who should be responsible for the cost of these disruptions. Richard Schmalensee has observed that deregulation may make it easier for State policy makers and regulators to ignore wholesale market effects. This is because the assets being stranded today are largely owned by non-utility generation companies in contrast to the late 1990’s when the stranded assets were a joint problem of regulated utilities and their rate payers.

California led the way with developing renewable energy in the 1980’s,  with the deregulation of the power sector in the 1990’s and 2000’s, and now with high-volume renewable mandates since 2010. We are learning a lot about how to physically manage and finance a cleaner energy system. We also need be realistic about the costs of such policies.  When you combine the cost of policies of the past with the aggressive goals for the future, you get retail electricity prices that, yes, continue to be pretty darn high.

 

This entry was posted in Uncategorized. Bookmark the permalink.

31 Responses to Breaking News! California Electricity Prices are High

  1. I wonder what the electrical cost per month is for the average Californian, not merely the rates per kWh, and where those stand. Massachusetts has high electric rates, too, but the average cost per month to a resident is no higher than it is in Wyoming, where the electric rates are much lower. In Massachusetts, my suspicion is that residents are much more efficient in their use of electricity, and that a creaking old grid and trio of utilities have fewer kWh to burden with their fixed costs. Accordingly, price per kWh is higher.

    I wonder if the same is true in California.

    It is clear that as much as utilities and regulators might wish it to be otherwise, renewables cannot be managed in the same manner conventional generation is, and reinventing the grid — meaning replacing it in large measure — is an essential part of their proper operation. Maybe that’s what utilities are doing?

    • Kevin Swartz says:

      People near the cost line hard use any electricity. However, as you go inland (warmer and warmer) the yearly usage goes way up. We also see a higher penetration of roof top solar, as we go inland. I believe on average that Californians are pretty efficient, but don’t have anything to back it up.

    • Our organization puts together the California Green Innovation Index, and you can find the latest version here: http://next10.org/2016-gii. It compares a host of environmental factors, including electricity costs, among the states.

  2. Kevin Swartz says:

    Given the recent news from the CAISO, the importance of having ‘rapid response’ capacity available in the early evening hours is clear. We had a 15GW ramp earlier this month, in the early evening hours. However, if this older gas fired capacity is not dispatched enough, or doesn’t earn enough revenue to remain in business, then we will have to support them artificially somehow.

  3. mcubedecon says:

    If only we could leave those conventional generation assets as stranded investment. Unfortunately, most have long-term PPAs with the IOUs, and the IOUs pass those through to ratepayers, standing on a state law that allows whatever costs are incurred so long as they are deemed “prudent” by the CPUC.

    I think I have a different twist on the underlying thread of this story. I see the real culprit as overprocurement of overly long PPAs with no real risk mitigation strategies. The IOUs have complained that the CPUC has forced them into this box, but the IOUs haven’t demonstrated an ability to accommodate load growth uncertainty, either in generation or distribution planning. A lot of this is due to the lack of incentives for IOUs which leads to a severe principal-agent problem.

  4. rogerl47 says:

    The justification for California’s accelerated push for renewables has always been based on environmental benefits. Given the reported higher California energy costs, how do the environmental benefits compare with the business and personal cost impacts?

    • Robert Borlick says:

      “Given the reported higher California energy costs, how do the environmental benefits compare with the business and personal cost impacts?”

      Great question!

      Of course it depends on how you value the environmental benefits. My guess is that the costs exceed the benefits in California. Why? Because the subsidies flowing to residential solar are far in excess of what it would cost by achieving the same environmental benefits through investing in large-scale solar instead.

      California politics are really screwed up.

  5. Thanks for a much more thorough discussion of the situation in California regarding current capacity & RPS mandates. The LA Times piece was troublesome & extremely simplistic when it came out a couple weeks ago, claiming largely installed capacity alone was the problem but little understanding of how the state got there. Even Moss Landing was a footnote in that piece, while ignoring the cooling water issue you highlighted, or what my fact checking revealed in the irony that a Bay Area NG plant in a NG pipeline/supply constrained state is selling/shipping the electricity to LA.

  6. Robert Borlick says:

    “However comparing the cost of a brand-new solar plant to that of a brand-new gas plant overlooks two important facts.”

    Elaborating on your first “important fact” is that adding more solar to the power system imposes the cost of having to deal with the Duck Curve. Thus, the cost of procuring the fast-ramping resources needed, e.g., combustion turbines or storage, should be allocated to the cost of solar. In that light solar is not so cheap.

    The levelized cost of an intermittent resource, such as wind or solar, cannot logically be compared to the levelized cost of resource that can produce its nameplate output on a 24/7 basis. Yet these simplistic comparisons are routinely done. The appropriate comparison is to include the cost a resource that can “firm up” the output of the intermittent resource (e.g., storage) so that apples are being compared to apples.

    Why is this so hard for people to understand? Or is it that they do understand this but have self-serving reasons for not acknowledging it?

    • Jardinero1 says:

      Hear, Hear!

      • Existence of a “duck curve” is, for me, a consequence of poor grid planning. Consider ERCOT in contrast.

        • mcubedecon says:

          It’s not clear to me that the “duck curve” has a real consequence in a grid in which California is embedded as a minority portion. The duck curve is misleading in representing an isolated portion of the total WECC grid. Much of the rest of the WECC is available to accommodate these swings, of which much (most?) is coal fired generation. The EIM should be able to address much of this swing.

          • Robert Borlick says:

            There is no doubt that the EIM will alleviate the Duck Curve problem – assuming that existing transmission constraints don’t limit the amount of energy that California can export during the daylight hours. Historically California has been transmission-constrained in terms of importing energy from WECC. But alleviating the Duck Curve requires flows in the opposite direction to the historical flows so this may not be a problem. Only a detailed load flow analysis can answer that question. Loop flows can produce nonintuitive effects.

    • Jim Roumasset says:

      Right; maybe the warnings of Joskow about the perils of levelized cost have been falling on deaf ears. Clearly the marginal cost of intermittency increases with the level of the RPS, but how to calculate it w/o running a full stochastic optimization model? Heal concludes from his two 2016 NBER papers that 33% of power demand can be satisfied by solar with modest intermittency costs but without the need for storage and that the same amount of wind power would need one day’s worth of storage plus increased wind capacity, after allowing for spatial diversification, .

    • mcubedecon says:

      I agree that the LCOE is an inadequate measure. However the problem is not one-sided bias against conventional resources. For example the fuel price uncertainty of fossil generation is ignored, and the fluctuation of hydropower availability affects system costs in a non linear fashion. The capacity factors of peak and intermediate load gas generation also is notoriously overestimated.

      • Robert Borlick says:

        Your comment on fuel price uncertainty is very insightful. Ideally the LCOE calculations for all resources should use the certainty-equivalent prices for each fuel. Actually, this is possible to do for about the first 10 years for natural gas because NYMEX forward contracts go out that far.

        However, this produces a conundrum because the certainty-equivalent prices for future gas purchases are likely (but not necessarily) to be lower than the expected values of the spot prices. I say this is likely to be true because natural gas producers, most of whom are heavily debt-financed, would rather have certainty in their future revenues than risk not being able to service their debt if the market price crashes. However, it depends on the risk aversion of the gas purchasers since a futures contract shifts the price risk to them.

        This effect lowers the LCOE of a gas-fired generator relative to a renewable resource, which (by definition) has no fuel cost. Does this make sense? Maybe it does. As a consumer I would rather incur an uncertain cost in the future that is likely to vary directly with the health of the economy than be saddled with a fixed cost because my ability to pay will be higher in the healthy economy.

        I discovered this in a different context years ago but am still not comfortable with it.

        • mcubedecon says:

          Do to the multitude of constraints on gas delivery and consistency of use in electricity, buyers are not contracting for long-term futures on the NYMEX, and the price would almost certainly go up substantially if this was a truly liquid market. We’ve looked at the error around gas price forecasts as a measure forecast uncertainty, and the range is from $2 to $10/MCF, which implies the actual risk premium is quite large. This adds to the true LCOE for gas plants.

          As for price certainty to consumers, apparently quite a few are willing to pay the higher upfront cost for solar rooftops to gain certainty in later years.

          • “As for price certainty to consumers, apparently quite a few are willing to pay the higher upfront cost for solar rooftops to gain certainty in later years.”

            The financial attraction of solar obviously depends upon per kWh cost, costs of competing energy sources, and whether or not consumers have in mind somethingvof the negative externalities of fossil fuels and their Rube Goldberg distribution networks.

            Also, avoided cost of energy depends upon what else consumers want to do. In our case, near Boston, we electrified our heating as well as our cooling with air source heat pumps, and orphaned the oil heat and hot water source with that and an air source heat pump hot water tank. How far solar can get you depends upon what you need and use. We almost never use our big electric oven, using a microwave, a Brevard small oven, and an induction stovetop instead.

          • Robert Borlick says:

            I disagree with your statement that homeowners adopt rooftop solar to gain certainty. If you do a simple Discounted Cash Flow analysis for the investment in rooftop solar in California (at least those areas served by the three IOUs) it is clear that these projects are obscenely profitable because of the high bill savings created by avoiding the two highest pricing tiers in the residential tariffs. I did such a study for the typical solar customer served by Southern California Edison and it produced an after-tax rate of return of about 17 percent! See: http://www.edisonfoundation.net/iei/publications/Documents/IEI_NEM_Subsidy_Issues_FINAL.pdf

  7. Kevin Swartz says:

    I agree. The ‘grid parity’ statement is over simplistic.

  8. JonGone says:

    Always a joy to read Dr. Bushnell’s views. I liked the comparison of the RPS to the fuel economy policy approaches for vehicles.

    I suspect that the high level of renewables deployment, by themselves, have led to a higher level of capacity than has been seen because most renewables have far lower capacity factors than traditional sources (e.g. capacity factors of around 30% for renewables v. well over 50% for a natural gas CC). So you need a lot more capacity to supply the energy system. In a theoretical system that has 100,000 MW of capacity producing with a capacity factor of 60%, the output is 525,600,000. A similar system of 100,000 MW producing at 35% capacity factor produces about 306,600,000 MWh. To have a system that only has a capacity factor of 35% produce as much as the system producing 60% would require having more than 171,000 MW.

    There is good cause for consideration of the impacts of renewables on costs. As noted in the article, a good bit of the cost of the added renewables are being paid for through the RPS, directly or indirectly, and translated to consumer prices/rates. It is worth noting that some of these procurement decisions have long-term implications since many of the contracts are 20 years or longer. Buying today’s technology, paying up-front, and committing the consumer to buy it long into the future. When long-term procurements like this were made for coal or gas contracts and the market prices dropped there was gnashing of teeth. Compounding these direct costs, increased deployment of renewables is pushing traditional capacity out of the market during periods when the renewables are available, because they can bid zero or negative prices, rendering the costs of the traditional sources during remaining hours higher–costs that show up when utilities are required to backfill the dropping output of renewables. These higher costs show up in the form of higher ramp-up, ramp-down costs and through the procurement of higher levels of spinning reserves. However, other costs that cannot plausibly be claimed to be a marginal cost (an approved by CAISO) might not be recovered from these regulated markets and the plant may end up leaving the market only to be replaced by a new plant that can be profitable in this intermittent duty cycle, but may be less cost effective than an alternative that was designed to operate more frequently.

  9. Arthur Winer says:

    This excellent analysis, and the cogent comments in response to it, illustrate once again the sorry state of what passes for investigative journalism at the Los Angeles Times, a once great paper now staffed by “journalists’ who appear to have little or no technical training that would equip them to get to the bottom of complex issues like this one.

  10. Jim Roumasset says:

    Not terribly clear on this, but are you (JB) saying that the utilities are mandated to enter long-term agreements to buy from (high-cost) renewable providers, thus displacing what they could otherwise get from the (marginal cost) wholesale market? What regulatory design would avoid this problem?

  11. Kevin Swartz says:

    I’m not a big believer in 20-20 hindsight. Could it been handled better? Probably. However, shorter term PPA’s mean shorter financing periods, and much higher costs.

    Second issue that should be considered. Between now and 2024, California will be closing some 17.6 GW of coastal fossil units that utilize once through cooling. Here is a link to a descriptive document;

    http://www.energy.ca.gov/renewables/tracking_progress/documents/once_through_cooling.pdf

    Another factor to consider is the WECC assessment of Ca/Mx reserve capacity. Here is a link to that report;

    https://www.wecc.biz/Reliability/2015PSA.pdf

    It shows (on pg 18-19), reserve margins in the 20’s. A 20% reserve margin is not out of line with customary utility practices.

    • mcubedecon says:

      In my case, it’s not hindsight. I was advocating for shorter and fewer contracts in 2006, and even in 2001 when DWR went overboard. We saw lots of new gas plants with shorter contracts and even merchant development from 1997-2003, and the amount of renewable capacity in the queue around 2010 was multiples of the RPS requirement. Clearly investors have been willing to take the risks in building these plants. The CPUC and IOUs should have taken advantage of this, rather than the other way around.

      The CAISO reserve margin is in excess of 30% in 2021 according to the LARS report. Peak load is forecasted to fall between 2017 and 2021.

      • Robert Borlick says:

        Is that reserve margin based on maximum capability or is it adjusted for the ELCC for each individual generating resource? The reason I bring this up is that wind and solar resources typically have ELCCs down in the 20 percent to 30 percent range, whereas nuclear, hydro and fossil plants have ELCCs up around 80 percent to 95 percent of maximum capability.

  12. Jim Roumasset says:

    JB,
    Many economists have claimed that one problem w/ renewable subsidies, including FITs, is that they will lower prices, thus incurring excess burden on the demand side as well as on the supply side. You and The Economist (“A World Turned Upside Down,” Feb 25) are among the first to point out that retail prices can go up, even as wholesale prices go down. Where does it end? Here are some possible outcomes:
    1. To avoid the “death spiral” of rising prices and falling demand (as well as consumption), subsidies from tax and ratepayers are ever increasing, exemplifying black-hole economics.
    2. The technology of storage, renewable generation, grid management and demand-side conservation improve enough to break the vicious circle.
    3. Gas is increasingly used as a “bridge fuel,” and RPS mandates are rescheduled.
    Any guess?
    Jim

  13. How does $171 billion in surplus retail cost strike you? That’s what California’s last two decades of policy “innovation” have cost consumers as of 2015, the last time I calculated our higher rates against the U.S. average. And to compare apples to apples, I did the calculation for Texas (another big state with lots of renewables) and got an even larger gap.
    Here’s the bad news: We got far less carbon reduction than if we had simply set a carbon quota instead of RPS and let utility engineers repower to combined-cycle gas with strategic renewable supplementation where appropriate.
    My column on this: http://sdut.us/2mZnZjn

    • mcubedecon says:

      Did you adjust for the higher usage that would have occurred, both from the lower rates and from reduced energy efficiency financed by those higher rates? This is far from a one-sided equation.

      I agree that the state botched its roll out of restructuring. (We should have used Jesse Knight’s proposed buy out instead.) But to a large degree we can blame that debacle on Sen. Steve Peace who wrote into law a provision that largely prohibited the long-term contracting that would have prevented the market manipulation. (FERC market pricing rules and its failure to step in to stop the manipulation in August of 2000 also was a major cause of the crisis.) It appears you calculated the cost of the energy crisis at $110 billion or 64% of those costs. However, you failed to account for the fact that the DWR contracts actually ended up saving the state money later in the decade when natural gas prices spiked.

      I agree that California could have gotten GHG reductions for less cost by relying much more on a carbon tax or cap & trade (see my testimony to the CARB in 2008 on behalf of EDF). But you did not account for the lower costs of renewables now and going forward induced by the accelerated investment earlier.

      As for San Onofre, you’re simply wrong. PG&E has filed testimony estimating the cost of relicensing Diablo Canyon (which does not have the same operational problems that SONGS had) would cost at least $105 per MWH. SONGS would have cost at least this much. That SCE got away largely scot-free is a travesty of regulatory capture, and Mike Pevey has been revealed as corrupt. Compared to renewable contracts going for as little at $40 per MWH, closing SONGS has been a savings, not a cost.

      Upgrading SCE’s distribution system has nothing to do with the closure of SONGS. PG&E is spending similar amounts and only announced the closure of Diablo Canyon in 2024 last June. Utilities across the country are making similar investments regardless of whether they are closing plants or not.

      As for the consequences of declining usage, why haven’t you raised the obvious solution–making utility shareholders bear the risk just as they do in other industries?

      Your article uses a false premise that technology and consumer preferences are static and that the industry was operating just fine. You have taken a snapshot view without looking at the dynamic situation. It’s like comparing the upfront cost of buying a house to renting for a month–of course renting is cheaper when you look at it that way. Instead you should be looking at how these costs evolved over time.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s