California could hit the solar wall

[ According to a Stanford University article below this introduction (followed by excerpts from two California Energy Commission reports), if California uses mainly solar power to meet a 50% Renewable Portfolio Standard (RPS), on sunny days, for most of the year, more power would be generated mid-day than needed 23% of the time, and over-generation by solar PV from 42 to 65%. On average, nearly 9% of solar or other electricity generation would have to be be shut down. In other words, California could hit the solar wall.

Why would California mainly use solar rather than wind power as well? “Unlike the Midwest, California has a modest technical potential for wind and many of the best sites are already developed [my comment: any sites not developed yet are too expensive or far from the grid]. California does have a large offshore-wind resource but high costs and technological challenges remain. Importing electricity generated by onshore wind from neighboring states is promising, but some imports will require new high-voltage transmission lines that may take a decade to plan, site, permit, finance and build.”

Overall, California has already developed MOST of the best sites (NREL 2013):

“Prime renewable resources include wind (40% capacity factor or better), solar (7.5 DNI or better), and discovered geothermal potential. All other renewable resources are non-prime. California’s remaining options for easily developable in-state utility-scale renewables could be limited by 2025. Wind, geothermal, biomass, and small hydro projects under contract (either existing or under construction) are about equal to the total developable potential estimated for each of these technologies in California’s renewable energy zones. 

Solar projects to date, however, exceed the amount of developable prime and borderline prime resources estimated to exist within California’s zones.  This suggests that California’s remaining solar resource areas tend have less solar exposure than what has already been developed and might be less productive.”

Less productive = more expensive. This report also points out that California will have 44.3 million people in 2025, requiring renewable generation to grow even further. ]

The consequences of too much solar power include:

  • Since solar power provides energy when least needed (mid-day rather than peak morning and late afternoon hours), massive amounts of power from national gas power plants (tens of gigawatt hours of energy) would need to QUICKLY ramp up as the sun’s power rapidly fades in the afternoon, requiring large expensive natural gas back-up plants. But natural gas is finite, so that’s a temporary solution. The only other commercial source of dispatchable energy is (pumped) hydropower, but there are few spots to put new dams, and existing dams are limited much of the year due to drought, fisheries, agriculture, and drinking water. Geothermal, nuclear, coal are not dispatchable — they are considered baseload power running 24 x 7. They can’t ramp up and down in less than 6 to 8 hours because that damages their equipment.  When built they expected to generate a certain percent of power per day to pay back their cost, so when they have to shut down quickly because solar and wind power have first rights to provide power.  This can drive electricity prices negative, but since ramping down can damage their equipment, coal, natural gas, and nuclear plants lose money to continue generating power. This is why many nuclear and coal power plants are shutting down — they are losing money.  But since solar and wind are unreliable, intermittent, and unpredictable, the electric grid needs them to be ready to fill in when the sun goes down and the wind dies.
  • Utility scale energy battery storage is dispatchable, but far from commercial.
  • Large-scale curtailment of solar PV during times of over-generation reduce the value of solar capacity additions to investors.
  • Real-time pricing during times of over-generation could limit or eliminate the net-metering advantage of PV on residential and commercial-scale installations.

California peak and off peak demand in California. Solar produces power when it’s least needed: from 7 am to 4 pm during Off Peak and Super Off Peak time frames.  Adding more solar power makes the problem worse, requiring even more solar PV power and other plants to shut down more often.

[ California solar generation has reached the point where it’s producing so much power at the time of day when it’s least needed that it has to be shut down during the sunniest time of the year.  This is because year round, solar generates power when there is the least demand, and the least power when demand is highest.

Notice in the figure above that peak demand occurs after 4 p.m., which according to the California ISO, is “when the sun is setting and solar output is declining. During July and August supplies are even more limited during peak hours”. Except for July and August on the weekends, supply surplus occurs during “super off-peak” hours from 10 to 4 PM – which is when solar generation is at its highest. In addition, surplus conditions occur this same time period in March and April weekdays while weather is still mild and there’s no need for air-conditioning.

Because solar PV is so seasonal, it provides from 2% in winter, to 10% in summer of California’s daily needs — but not when most needed, and at times, far more than what is needed, so solar PV and/or other power generation has to be shut down.  Additional solar PV power only makes the problem worse.  Solar thermal with energy storage would help, but it’s mostly “smoke and mirrors”, less than a quarter have storage, and most of the time produce less than half a percent of daily power needs for California.

Related articles

2016-4-7 Texas and California have too much renewable energy. The rapid growth of wind and solar power in the states is wreaking havoc with energy prices. MIT Technology Review.

Alice Friedemann  author of “When Trucks Stop Running: Energy and the Future of Transportation”, 2015, Springer and “Crunch! Whole Grain Artisan Chips and Crackers”. Podcasts:  KunstlerCast 253, KunstlerCast278, Peak Prosperity]

Benson, S., and Majumdar, A. July 12, 2016. On the path to deep decarbonization: Avoiding the solar wall. Stanford University.

Photo: Close-up of solar modules. Credit: NREL

Photo: NREL

While you might not have been paying attention, California’s electric grid has undergone a radical transformation. Today, more than 25 percent of the electricity used in California comes from renewable energy resources, not including the additional 10-15 percent from large-scale hydropower or all the electricity generated from roof-top solar photovoltaic (PV) panels. Compared to only five years ago, solar energy has grown 20-fold (Figure 1). In the middle of a typical springtime day, solar energy from utility-scale power plants provides an impressive 7 gigawatts (GW) of grid-connected power, accounting for about 20 percent of the electricity used across the state (Figure 2).

Last year, the total electricity from both in-state and out-of-state resources was 296,843  gigawatt-hours (GWh), including out-of-state generation from unspecified sources. An estimated 8 percent of California’s electricity consumption was generated from wind farms and 6 percent from solar power plants connected directly to the California Independent System Operator (CAISO) grid. This rapid growth is great news for the nascent renewable energy industry and can serve as proof-point for the scale-up of renewable energy.  In addition to these utility-scale renewable energy power plants, California has an additional 3.5 GW of solar “self-generation” on the customer side of the meter that offsets demand for electricity from the grid when the sun is shining. And, a growing third source is “community solar,” where residents and businesses invest in small, local solar plants.

Figure 1. Total electricity generation in California from solar and wind energy directly connected to the CASIO grid (California Energy Almanac).

By law, in 2020 California plans to have 33% of its electricity sourced from renewable sources and 50% by 2030 under its renewable portfolio standard (RPS) requirements. Many in-state renewable energy projects are in the pipeline, including nearly 9 GW of solar PV and 1.8 GW of new wind projects that have received environmental permits.  Power purchase agreements have already been signed for at least 1 GW of new solar projects. If all of these permitted projects were developed, California would have about 16 GW of solar-generating capacity by 2020.

While wind has provided a significant portion of California’s renewables to date, the majority of new additions for meeting the 2020 33% RPS requirement is forecast to come from direct grid-connected solar PV. California has an enormous and high-quality solar resource, with an estimated technical potential of more than 4,000 GW for utility-scale solar and 76 GW for rooftop solar. Unlike the Midwest, California has a modest technical potential for wind and many of the best sites are already developed. California does have a large offshore-wind resource, some of which is now in permitting, but high costs and technological challenges remain. Importing electricity generated by onshore wind from neighboring states is promising, but some imports will require new high-voltage transmission lines that may take a decade to plan, site, permit, finance and build.

Figure 2. California energy mix on May 29, 2016. Note that renewables provide more than 40% of the power during the middle of the day. Of this, more than 30% is from solar power (CAISO Daily Renewables Watch).

California has a major effort – the Renewable Energy Transmission Initiative (RETI) – that has successfully identified and built lines to meet its RPS requirements. This year, California launched a new phase of RETI to develop the additional transmission, both in-state and out-of-state, needed for the 50-percent RPS. A recent study supporting California Senate Bill 350 implementation (which includes the 50-percent RPS) showed from $1 billion to  $1.5 billion annual savings by adding major  transmission lines that would bring more out-of-state wind energy into California.

If, instead, California continues to rely mostly on solar resource for meeting the 2030 50-percent RPS, the total statewide solar-generating capacity would reach 30 to 40 GW under peak production, according to a report by Energy and Environmental Economics Inc. (E3). Under these conditions, on a sunny day, for most of the year, California would be generating more electric power than it needs during the middle of the day from solar energy alone. E3 calculates that this large amount of overgeneration could be a problem 23 percent of the time, resulting in curtailment of 8.9 percent of available renewable energy, with marginal overgeneration by solar PV of 42-65 percent. In other words, California could hit the solar wall. And this does not even consider that midday demand is likely to decrease due to the installation of additional residential and commercial solar PV systems “behind the electricity meter.”

Consequences of hitting the solar wall

Just a decade ago it would have been nearly unthinkable that during the middle of the day solar energy could provide more electricity than an economy as large as California’s needs. But supportive policies, rapid scale-up and decreasing costs make this possibility a reality today. While from some perspectives this is very encouraging, in reality, there are consequences for hitting the solar wall. For example:

  • Reliance on so much solar energy would require rapid ramping capacity for more than 10s of GW of natural gas power plants from 4:00-6:00 p.m., when the sun is going down and electricity demand goes up as people return home.
  • Large back-up capacity from natural gas plants or access to other sources of dispatchable electricity would be required for days when the sun isn’t shining.
  • Zero marginal-cost solar generation could squeeze out other valuable low-carbon electricity sources that can provide baseload power. For example, natural gas combined cycle plants, geothermal energy and nuclear power that cannot operate during these times at zero marginal cost.
  • Large-scale curtailment of solar PV during times of over-generation, which will reduce the value of solar capacity additions to investors.
  • Real-time pricing during times of over-generation could limit or eliminate the net-metering advantage of PV on residential and commercial-scale installations.

There is no doubt that California’s solar energy potential is invaluable, but we must take steps to avoid the solar wall.  Fortunately, these issues are being recognized and addressed at many levels in California.

Avoiding the solar wall

Numerous approaches to avoiding the solar wall are available today, and in the future more options will exist as we develop new technologies, policies and markets to take advantage of large solar-energy resources that exist around the world. In the short term, key actions include:

  • Develop a renewable energy-generation mix that is well-balanced among solar, wind and other forms of renewable generation. The right generation mix will be region specific, but for California should include increasing wind generation to provide nighttime power. [my comment: what other renewable generation?  To reach an 80 to 100% renewable grid, most of the power has to come from solar and wind with a little help from geothermal and hydropower]
  • Support regional generation markets across wide geographic areas to balance the variability of renewable generation. California has created an energy imbalance market with participants in Nevada, Wyoming and Oregon. Expansion of regional markets is being studied as part of the implementation of Senate Bill 350, California’s 50-percent RPS law.
  • Ensure adequate capacity of rapid ramping natural gas plants  to provide reliable supply during the morning and evening hours as the sun rises and sets. [my comment: natural gas is finite! Conventional natural gas peaked in 2001 in the U.S., shale gas is peaking both economically now and geologically by 2020, and we have only 4 Liquefied Natural Gas (LNG) import terminals].
  • Expand use of load shifting through real-time pricing to incentivize using power during daytime hours when large amounts of solar power are available.
  • Encourage daytime smart charging of electric vehicles to take advantage of abundant and zero marginal-cost solar generation. Achieving this will require workplace charging stations and new business models. With transportation at about 40% of the state’s energy use, electrification of the transport sector could have the dual benefits of eliminating tailpipe emissions and providing demand for abundant and low-cost solar energy. [My comment: the math and computer algorithms to have a smart grid are far from existence, batteries aren’t much better than they were 210 years ago, and trucks can’t run on batteries. ]
  • Increase energy storage to avoid curtailment of solar over-generation during peak production periods. For now, few financial incentives exist for large-scale pumped-hydropower or compressed air storage projects [my comment: that’s not the problem!  There are very few places left to put pumped-hydro and no spots at all to put compressed air facilities, unless they’re above ground, which is crazy expensive]. Levelized costs of small-scale storage in batteries range from about $300 to more than $1,000/megawatt-hour (MWh) depending on the use-case and the technology. These are expensive compared to pumped-hydro storage at $190 to $270/MWh. For comparison, gas peaker plants have a levelized cost of $165 to $218/MWh. The business case for battery storage will be limited until prices come down significantly. Both R&D and scale-up will be needed to reduce costs. [my comment: utility scale battery storage is FAR from commercial, and only sodium sulfur (NaS) batteries have enough material on earth to store half a day of world electricity (see Barnhart, Charles J. and Benson, Sally M. January 30, 2013. On the importance of reducing the energetic and material demands of electrical energy storage. Energy Environ. Sci., 2013, 6, 1083-1092)]
  • [ My comment: Furthermore, utility scale battery storage is far from being commercial. Using data from the Department of Energy (DOE/EPRI 2013) energy storage handbook “Electricity storage handbook in collaboration with NRECA”, I calculated that the cost of NaS batteries capable of storing 24 hours of electricity generation in the United States came to $40.77 trillion dollars, covered 923 square miles, and weighed in at a husky 450 million tons.
    Sodium Sulfur (NaS) Battery Cost Calculation:
    NaS Battery 100 MW. Total Plant Cost (TPC) $316,796,550. Energy
    Capacity @ rated depth-of-discharge 86.4 MWh. Size: 200,000 square feet.
    Weight: 7000,000 lbs, Battery replacement 15 years (DOE/EPRI p. 245).
    128,700 NaS batteries needed for 1 day of storage = 11.12 TWh/0.0000864 TWh.
    $40.77 trillion dollars to replace the battery every 15 years = 128,700 NaS * $316,796,550 TPC.
    923 square miles = 200,000 square feet * 128,700 NaS batteries.
    450 million short tons = 7,000,000 lbs * 128,700 batteries/2000 lbs.
    Using similar logic and data from DOE/EPRI, Li-ion batteries would cost
    $11.9 trillion dollars, take up 345 square miles, and weigh 74 million tons. Lead–acid (advanced) would cost $8.3 trillion dollars, take up 217.5 square miles, and weigh 15.8 million tons.
  • Use electrolysis to produce hydrogen fuel to augment the natural gas grid, generate heat and power with fuel cells, or power hydrogen vehicles. [my comment: hydrogen is the least likely energy solution, even more unlikely than fusion]. Also, compared to storing electricity in batteries, hydrogen-based storage systems that combine electrolysis and  fuel cells are about three times less efficient.  In addition, today, these technologies are expensive, and significant cost reductions will be required to make them competitive alternatives.
  • For the longer term, scientists are developing new methods to produce fuels from renewable energy. The SUNCAT Center and  the Joint Center for Artificial Photosynthesis are developing new materials to produce “zero net carbon fuels” from carbon dioxide, water and renewable energy that can be used for transportation or backing up the electric grid. While we don’t know if and when the needed breakthroughs will occur, the game-changing potential of net zero carbon fuels would unlock the full potential of solar energy and break through the solar wall.  [My comment: before reading further, if the fuel isn’t DIESEL to keep trucks running, then what’s the point? And given that we’re at peak oil, peak coal, and peak natural gas, we don’t have the time for breakthroughs to occur. You’d want to prepare at least 20 years ahead of time].

Taking full advantage of the power from the sun

The global potential of solar energy is enormous and surely it can play a major role in a deeply decarbonized future energy system. In Thomas Edison’s words, “I’d put my money on the sun and solar energy. What a source of power! I hope we don’t have to wait until oil and coal run out before we tackle that.”*

We have work to do, but we are well on the way.  Who would have imagined just five years ago that solar energy would provide 6% of California’s electricity and is on track to double, triple or go beyond? But we need to be smart – avoiding running into the solar wall by balancing the generation mix, expanding regional markets, creating real-time markets to increase demand during solar peak-generating periods and creating new electricity demand, such as day-time charging for electric vehicles. In the longer term, electricity storage, hydrogen generation and zero net carbon fuels will further unlock the potential of solar energy.

*As quoted in Uncommon Friends : Life with Thomas Edison, Henry Ford, Harvey Firestone, Alexis Carrel & Charles Lindbergh(1987), by James Newton, p. 31

Meier, Alexandra von. (California Institute for Energy and Environment). 2010. Challenges to the Integration of Renewable Resources at High System Penetration. California Energy Commission. Publication number: CEC-500-2014-042.

More work is required to move from the status quo to a system with 33 percent of intermittent renewables. Work must proceed simultaneously on multiple fronts.

The complex nature of the grid and the refining temporal and spatial coordination represented a profound departure from the capabilities of the legacy or baseload system. Any “smart grid” development will require time for learning, especially by drawing on empirical performance data as they become available. Researchers concluded that time was of the essence in answering the many foundational questions about how to design and evaluate new system capabilities, how to re-write standards and procedures accordingly, how to create incentives to elicit the most constructive behavior from market participants and how to support operators in their efforts to keep the grid working reliably during these transitions.

Research needs for temporal coordination relate to: resource intermittence, forecasting and modeling on finer time scales; electric storage and implementation on different time scales; demand response and its implementation as a firm resource; and dynamic behavior of the alternating current grid, including stability and low-frequency oscillations, and the related behavior of switch-controlled generation. Different technologies, management strategies and incentive mechanisms are necessary to address coordination on different time scales.

A challenge to “smart grid” coordination is managing unprecedented amounts of data associated with an unprecedented number of decisions and control actions at various levels throughout the grid.

Renewable and distributed resources introduce space or location (spatial) and time (temporal) constraints on resource availability. It is not always possible to have the resources available where and when they are required

New efforts will be required to coordinate these resources in space and time within the electric grid.

Time lag between solar generation peak and late afternoon demand peak.  The availability of solar power generally has an excellent coincidence with summer-peaking demand. However, while the highest load days are reliably sunny, the peak air-conditioning loads occur later in the afternoon due to the thermal inertia of buildings, typically lagging peak insolation by several hours.

Limited forecasting abilities. Rapid change of power output is especially problematic when it comes without warning.

On the technical side:  

  • Long-distance a.c. power transfers are constrained by stability limits (phase angle separation) regardless of thermal transmission capacity
  • Increased long-distance a.c. power transfers may exacerbate low-frequency oscillations (phase angle and voltage), potentially compromising system stability and security
  • Further expansion of long-distance power transfers, whether from renewable or other sources, will very likely require the increased use of newer technologies in transmission systems to overcome the dynamic constraints.

CEC. December 2016. Tracking resource flexibility. California Energy Commission.

The rapid growth in renewable resources in California has also brought new challenges for grid operators. As discussed in the Renewables Tracking Progress page, wind and solar resources have grown tremendously over the last decade. Solar in particular increased from a little more than 400 megawatts (MW) in 2001 to more than 7,000 MW in 2015. Rooftop solar photovoltaic (PV) has also seen dramatic growth with 4,400 MW installed statewide, nearly 2,000 MW of which was installed in 2014 and 2015. Maintaining the reliability of the electricity system while integrating larger amounts of variable wind and solar generation requires more flexible resources to balance supply and demand.

The continued projected growth of intermittent renewable generation to meet California’s 33 percent Renewables Portfolio Standard (RPS) by 2020 spurred several studies to determine the extent to which the system operator needs additional flexible capabilities to accommodate late afternoon upward ramps in energy demand.1 These studies and current system operating data also highlight the extent to which overgeneration has become a concern.2

Any challenges in addressing intermittent generation at 33 percent are increased when planning to achieve 50 percent. Furthermore, because of expected changes in the natural gas-fired dispatchable fleet, the California Independent System Operator (California ISO) is concerned that it needs greater operational control over flexible capacity than is available through California Public Utilities Commission (CPUC) rules or existing California ISO tariffs.

A standard one-hour time resolution was sufficient to match large amounts of renewable resources with firming resources that can compensate for the intermittency of renewables. However, operational concerns in the California electrical system are increasingly focused on much shorter time scales. For example, there may be plenty of reserve generation capacity but a lack of fast-responding resources that can follow a rapid change in generation and load.

Analyses to date suggest that flexible capacity has to address variability in load and power production in three time scales: (1) seconds-to-minutes, (2) 5-10 minutes, and (3) multi-hour.

Variations in the seconds-to-minutes time scale can be addressed by expanding the existing regulation service, such as using automatic generation control on existing generators. Storage is increasingly seen as a possible solution to these regulation concerns.

The 5-10 minute flexibility requirements address discrepancies between the 5-minute real-time market schedules and actual loads or generation encountered during these intervals.

Multi-hour ramps up and down have been a feature of California’s electrical system for decades, but the introduction of large amounts of renewable capacity with strong diurnal cycles exacerbates these traditional patterns, especially in winter and spring months, and is the focus of flexible capacity efforts.

Scheduling renewables in smaller time intervals, such as the real-time market, can reduce the amount of reserves required since the opportunity for differences between forecast and actual generation is reduced from an hour to a shorter time interval. Also, expanding the geographic footprint of the market can help in two ways. First, greater diversity of renewable resources can reduce the coincidence of production patterns. Second, loads in larger regions can help absorb excess production and generating resources may be able to assist with upward ramping requirements.

The California ISO popularized a graphical depiction of the “net load curve” 6 (the “duck chart”) that dispatchable generating resources must satisfy each hour.

Figure 1 illustrates the extent to which resources must be available to ramp up or down to satisfy this need. A net load curve shares many features with a total load curve but superimposes the hour-by-hour variability of wind and solar generation. The ramps up and down in the net load curve have become sharper and more exaggerated faster than anticipated given the rapid increase in behind-the-meter solar PV and progress toward the 2020 RPS goal.

By definition, a net load curve is total load less the production of wind and solar generating facilities. It can be computed with data of any time increment, most commonly hourly or for 1-minute increments.

Figure 1: The Duck Has Landed  Source: Fowlie, Meredith, The Duck has Landed, Energy at Haas, U.C. Berkeley, May 2, 2016. California ISO Hourly Data, March 28-April 3, Years 2013-2016

In 2013, the California ISO projected that net energy demand after subtracting behind-the-meter generation (net load) could be as low as 12,000 MW by 2020 and that meeting peak demand may require ramping up 13,000 MW in three hours. Two days in 2016 illustrate that the grid is already experiencing unprecedented operational fluctuations that grid operators were bracing for in 2020. On May 15, 2016, the net load reached a minimum of 11,663 MW, and on February 1, 2016, the three hour ramp was 10,892 MW, with the peak shifting to later hours in the day.

Overgeneration is the condition represented by the “belly” of the duck curve. Overgeneration exists when net load falls below the minimum generation level of other resources that must be on-line. The analyses submitted in the CPUC’s 2014 Long-Term Procurement Plan (LTPP) rulemaking identify spring months with high wind and solar production coupled with low loads as the prime time for overgeneration conditions to be encountered.7

See California ISO report for a summary of overgeneration issues and its study results.

Some options to solve overgeneration suggest a need for more flexible generating facilities from either a physical or contractual perspective. Overgeneration also can be solved by curtailing renewable generation, retrofitting existing natural gas plants to reduce minimum generation levels, building load through demand response programs when overgeneration conditions are expected, shifting load using system condition-dependent TOU rates, or by exporting power outside the California ISO balancing authority area, and so forth.8 The development of a regional grid is another important tool to help integrate renewable resources beyond what can be achieved with the Energy Imbalance Market (EIM). The EIM started in 2014 with PacifiCorp but continues to expand with NV Energy joining in 2015, and Arizona Public Service and Puget Sound Energy joining on October 1, 2016. The EIM is a mechanism to balance deviations in supply and demand and dispatch least-cost resources every five minutes. With the EIM, excess energy in the California ISO balancing area can be transferred to other areas in real time. If not for energy transfers facilitated by the EIM, the California ISO would have curtailed 272,000 MWh of renewable energy in the first two quarters of 2016, equivalent to 116,000 metric tons of carbon emissions. PacifiCorp has


California ISO analyses completed in April 2016 show that the problem of rapidly increasing net load ramps is most severe in the winter months of November through March.10 Figure 2 provides an estimate of the maximum ramp over 180 minutes by month for three historical years and 2017 based on renewable projects now in the pipeline.11 Figure 2 shows that maximum monthly 180-minute ramps were relatively uniform throughout the year historically but become much larger into the future for the eight non-summer months. The implication is the need for flexible resources to satisfy this increasing ramp for these non-summer months, the opposite of the traditional capacity planning focus on summer peak months of July to September.

10 California ISO,

Figure 2: Comparing Historical and Projected Maximum 3-Hour Ramps by Month

For the first time, the California ISO study for 2017 flexibility requirements included behind-the-meter PV generation. This increases the 3-hour ramps considerably. As noted earlier, the rapid growth in behind-the-meter PV capacity means that the load curve does not remain static, but itself is lower during the middle hours of the day, creating ramping requirements where none would have existed without the behind-the-meter PV.

Flexible Resources.  Since the California ISO assessments assume that the great majority of renewable resources will continue to be “must take,” the California ISO wants to ensure that sufficient flexible capacity will be available to satisfy these growing ramping requirements.

Clearly, the total of nearly 35,000 MW of existing flexible capacity expected in 2017 exceeds the largest California ISO estimate of requirements in 2017. There are three concerns, however, suggesting that the balance between requirements and capabilities is tighter than it might appear in comparing Figure 2 with Table 1.

First, nearly all of the steam turbine capacity is very old, and most of it uses once-through cooling (OTC) technology. Facility owners must satisfy State Water Resources Control Board (SWRCB) OTC policy by retiring or retrofitting the power plants.

Responses to SWRCB information requests reveal that nearly all generator owners plan to comply by retiring, although many would prefer to repower if long-term contracts can be secured from load-serving entities (LSEs). Retiring all of the remaining natural gas steam boiler EFC (8,931 MW) would reduce the remaining EFC of the generating fleet to about 26,000 MW if nothing more was added.

Table 1: Effective Flexible Capacity by Generating Technology and Fuel Type (Megawatts)

Second, much of the fossil-fired generating fleet must shut down for annual maintenance, and the optimal time has typically been in the winter months, when loads have been low. The need for much larger amounts of flexible capacity in winter months means that there are now competing motivations for when to schedule maintenance

Third, even if sufficient physical flexible capacity exists, such resources may not be available to the California ISO when flexibility is needed.

ISO wants to have greater control to ensure that it can dispatch capacity up or down to satisfy net loads. LSE/generator contracts with self-scheduling will still be allowed, but such capacity will not count as flexible. An LSE wishing to continue to self-schedule will be required to satisfy its share of the aggregate, or combined, flexible capacity requirements by nominating18 other capacity that is both physically flexible and can be dispatched up or down by the California ISO.


NREL. August 2013. Beyond Renewable Portfolio Standards: An Assessment of Regional Supply and Demand Conditions Affecting the Future of Renewable Energy in the West. National Renewable Energy Laboratory.

This entry was posted in Photovoltaic Solar, Renewable Integration and tagged , , , . Bookmark the permalink.

4 Responses to California could hit the solar wall

  1. Maura Shannon- Benton says:

    you do not that we are doomed. You do not know, nor can you know about unforeseen events. I agree with much of what you say above. Nobody that’s needs to be convinced is listening to what you are saying.Im all for preaching to the choir, but realistically how many people out of 7 billion are going to read our little discussion

  2. energyskeptic says:

    I know a woman who is bringing solar to Africa. I said that was wonderful, cooking with solar would help to prevent remaining (rain)forests from being cut down to cook with. A few seconds of silence. I’m bring cell phones and solar chargers — people don’t want to cook at noon because they’re at work or busy elsewhere. If even the poorest of the poor can’t be bothered to save the planet by cooking with solar power, what hope is there of the more affluent taking appropriate actions

  3. Joe Clarkson says:

    The cost numbers for batteries have the wrong units. It should be $300 to $1000 per kWh, not MWh.

    It is far less expensive (2 to 3 orders of magnitude) to store energy as heat rather than in batteries. CSP with several days to weeks of storage is perfectly feasible technically and will eventually become financially attractive, especially if gas powered peaking plants are discouraged due to carbon emission concerns.

  4. energyskeptic says:

    Well, tell Sally at Stanford about the Mwh issue.

    Concentrated Solar power/thermal plants cost about $1 billion each, and only a quarter now in operation or being built have energy storage.

    They also use a lot of water without dry cooling, in which case they are less energy efficient (and still use water).

    But if their EROI is positive they’d be better than solar PV as energy storage, but they cost so much I don’t know if many will be built.