NewEnergyNews: TODAY’S STUDY: The Risk Of Natural Gas Vs. The Risk Of Wind


Gleanings from the web and the world, condensed for convenience, illustrated for enlightenment, arranged for impact...

The challenge now: To make every day Earth Day.


  • SoCalEdison’s Newest Plan To Mitigate Wildfires

  • Weekend Video: New Energy Means New Jobs
  • Weekend Video: Better Communication About The Climate Crisis
  • Weekend Video: VW Affirms Driving Is Ready To Go Electric

  • FRIDAY WORLD HEADLINE-The Climate Crisis Is The World’s Biggest Worry – Survey
  • FRIDAY WORLD HEADLINE-Record New Energy Global Growth In 2020


  • TTTA Wednesday-ORIGINAL REPORTING: The Search For A Successor Solar Policy
  • TTTA Wednesday-Local Governments Still Driving New Energy

  • Monday Study: PG&E’s Plans To Mitigate Wildfires
  • --------------------------


    Founding Editor Herman K. Trabish



    Some details about NewEnergyNews and the man behind the curtain: Herman K. Trabish, Agua Dulce, CA., Doctor with my hands, Writer with my head, Student of New Energy and Human Experience with my heart




      A tip of the NewEnergyNews cap to Phillip Garcia for crucial assistance in the design implementation of this site. Thanks, Phillip.


    Pay a visit to the HARRY BOYKOFF page at Basketball Reference, sponsored by NewEnergyNews and Oil In Their Blood.

  • ---------------
  • ORIGINAL REPORTING: The Differences Between Energy Markets
  • Biden Admin To Ensure Jobs Plan Protects Equity – DOE Head

    Monday, April 24, 2017

    TODAY’S STUDY: The Risk Of Natural Gas Vs. The Risk Of Wind

    Using Probability of Exceedance to Compare the Resource Risk of Renewable and Gas-Fired Generation

    Mark Bolinger, March 2017 (Lawrence Berkeley National Laboratory)

    Executive Summary

    Of the myriad risks surrounding long-term investments in power plants, resource risk is one of the most difficult to mitigate, and is also perhaps the risk that most-clearly distinguishes renewable generation from natural gas-fired generation. For renewable generators like wind and solar projects, resource risk manifests as a quantity risk—i.e., the risk that the quantity of wind and insolation will be less than expected.

    For gas-fired generators (i.e., a combined-cycle gas turbine or “CCGT”), resource risk manifests primarily as a price risk—i.e., the risk that natural gas will cost more than expected. Most often, resource risk—and natural gas price risk in particular—falls disproportionately on utility ratepayers, who are typically not well-equipped to manage this risk. As such, it is incumbent upon utilities, regulators, and policymakers to ensure that resource risk is taken into consideration when making or approving resource decisions, or enacting policies that influence the development of the electricity sector more broadly.

    This paper presents a new framework, grounded in statistical concepts related to probability of exceedance (and confidence intervals more broadly), to incorporate resource risk into decision-making processes. This framework recognizes that the same probability of exceedance concepts that are regularly used to characterize the uncertainty around annual energy production for wind and solar projects can also be applied to natural gas price projections, allowing one to develop a probabilistic range of projections for not only wind and solar capacity factors, but also natural gas prices.

    Importantly, these probability distributions have markedly divergent characteristics. Renewable resource risk is symmetrical about the mean or “P50” projection and declines when considered over longer time horizons (due to mean reversion in the inter-annual variability of the resource). In contrast, natural gas price risk is asymmetrical (skewed towards higher prices) and increases when considered over longer time horizons (reflecting the fact that it is easier to project where natural gas prices will be three months from now than three years from now). Converting these distinctly different probability distributions into directly comparable levelized cost of energy (“LCOE”) terms reveals that even when gas-fired generation is competitive with, or cheaper than, wind and solar power on an expected or P50 basis—the basis on which these resources are most often compared—comparisons that are instead based on worse-than-expected outcomes (e.g., P25 or P1) often reach the opposite conclusion: that wind and solar are cheaper than gas-fired generation.

    Figure ES-1 illustrates this concept by comparing the 25-year LCOE of a new wind project in the United States (without the benefit of the production tax credit or “PTC”) to that of a new CCGT across P-levels ranging from P50-P1 and over time horizons ranging from one to 25 years. The range of time horizons along the x-axis warrants additional explanation to avoid confusion. Every data point shown on Figure ES-1—regardless of where it falls along the x-axis—represents an LCOE that is calculated over a 25-year period (in nominal dollars). These 25-year LCOEs are based on modeling inputs that are held constant in all cases, with two exceptions—the wind project’s capacity factor and the CCGT’s levelized fuel costs vary by P-level and by time horizon. The x-axis simply represents the time horizon (in number of years) over which these two important, but uncertain, inputs into the 25-year LCOE calculation are considered.

    For example, at year 12 on the x-axis, wind’s 25-year LCOE range reflects 12-year P50 and 12-year P1 capacity factors used as inputs to the 25-year LCOE calculation; similarly, the range of gasfired LCOE reflects 12-year P50 and 12-year P1 gas price projections that are levelized over 12 years and then used as the fuel price inputs in the 25-year LCOE calculation.

    In Figure ES-1, wind (without the PTC) is more expensive than gas-fired generation on a P50 basis over all time horizons of less than 24 years (the two P50 curves converge at 24 years). But on a P25 basis, the cost of wind falls below the cost of gas-fired generation for all time horizons longer than 15 years.

    This “break-even” point—where the wind and gas-fired LCOE curves for each P-level cross—drops to 10, 8, and 2 years for P10, P5, and P1 levels, respectively.

    In other words, Figure ES-1 presents an illustrative example where wind, without the PTC, is not ostcompetitive with new gas-fired generation (except over a 24-year or longer time horizon) when evaluated on a P50 basis as is typically done. But when considering the possibility of worse-than-P50 outcomes (i.e., higher than-expected natural gas prices and/or a lower-than-expected wind resource), wind looks more competitive—particularly the lower the P-level and the longer the time horizon—and in many cases is cheaper than gas-fired generation.

    The “wedges” that begin where the respective wind and gas-fired LCOE curves at each P-level cross and then widen over longer time horizons illustrate wind’s “hedge value,” which increases with both the level of risk aversion (assumed to be negatively correlated with the P-level—i.e., a lower P-level suggests greater risk aversion) and the time horizon.

    Figure ES-2 shows much the same story for a utility-scale solar photovoltaic project. In this example, solar (with the 30% investment tax credit or “ITC”) is always more expensive than gas-fired generation on a P50 basis, regardless of time horizon shown.

    But, as with wind, worse-than-P50 comparisons reveal solar to be more competitive: the solar and gas-fired P25 LCOE curves converge at a 25-year time horizon, while the P10, P5, and P1 curves show solar’s hedge value starting to accrue at progressively shorter time horizons.

    Another related way to interpret Figures ES-1 and ES-2 is that higher-than-expected gas prices are riskier than lower-than-expected wind or solar output. This suggests that from a ratepayer perspective, we should perhaps be more concerned about gas price risk than about wind or solar resource risk. In other words, in a case where two scenarios—one focusing on higher-than-expected gas prices and another focusing on lower-than-expected wind or solar resources—may be considered to have the same probability (i.e., the same P-level), the resulting impact of the high gas price scenario may be more harmful to ratepayers than the impact of the low wind/solar resource scenario.

    Although the discussion surrounding Figures ES-1 and ES-2 has so far focused on LCOE comparisons at distinct P-levels, by definition, each P-value has an associated probability, thereby enabling a more formal probabilistic assessment. For example, although probability of exceedance does not necessarily imply probability of occurrence, the P50 outcome can nevertheless be thought of as carrying a 50% weight, while the P1 outcome can be given a 1% weight, with all other P-values that fall in between these two extremes (e.g., P49, P48, P47…P4, P3, P2) weighted accordingly (i.e., 49%, 48%, 47%...4%, 3%, 2%). Hence, within this framework, one can easily “probability-weight” the full range of outcomes across the full P50-P1 spectrum, or even some subset thereof—e.g., perhaps just the P50-P25 range for those who are less risk averse.

    The probabilistic nature of this new framework is one of its key advantages over previously proposed approaches to account for the price stability benefit of wind and solar power. Other advantages include its fairness (recognizing that wind and solar also face resource risk), familiarity (probability of exceedance is already widely used within the energy industry), simplicity (just a few key inputs are needed to set up these comparisons), and flexibility (this framework caters to any level of risk aversion over any time horizon).

    Of course, cost is only one side of the equation (value being the other), and few if any resource decisions within the electricity sector are made based on LCOE alone. Instead, the cost of competing resources must be considered along with the value that each provides, which is most often determined by sophisticated models that endogenously assess energy and capacity value as well as integration and transmission costs—all in addition to the LCOE of the generator itself. In this sense, it should be recognized from the start that this report is focused on just one side of a two-sided coin.

    IFTTT Recipe: Share new blog posts to Facebook connects blogger to facebook


    Post a Comment

    << Home