Solar and wind grid system value in the United States: The effect of transmission congestion, generation profiles, and curtailment

•The causes of wind and solar value decline vary by type region•Both transmission congestion generation timing cause decline•Solar is primarily due to timing•Wind cost declines have outpaced over the past decade Substantial adoption energy essential meet decarbonization goals. As supply these resources increases, electricity during sunny or windy hours in relation average electricity. Left unchecked, this might put practical limits on expansion threaten We evaluate decline, calculated from capacity potential revenues at plants across US. show that dominant (output profile, congestion, curtailment) varies between solar, region. Policy makers, regulators, researchers can use results inform decisions about which mitigation strategies pursue, choosing technological strategies, such as storage altered plant design, infrastructure policy regulatory strategies. The generated sources increases. This has varied time regions, indicating mitigate will need be carefully targeted. To help guide development we empirically determine ?2,100 within United States wholesale markets using local prices plant-specific profiles. how each loses (or gains) because its output curtailment. In regions where account for roughly 20% generation, 30% 40% below regional a flat profile all plants. Solar reductions are most sensitive both region dependent. Curtailment was not major source reduction. Variable renewable (VRE), referring power study, cost-effectively decarbonizing grid. VRE provides greater portion electricity, low marginal VRE. larger than future learning.1Green R. Léautier T.-O. Do costs fall faster revenues? Dynamics renewables Entry into Electricity Markets. TSE Working Paper, 15–591.http://tse-fr.eu/pub/29543Date: 2015Google Scholar,2Das S. Hittinger E. Williams Learning enough: diminishing increasing abatement solar.Renew. Energy. 2020; 156: 634-644Crossref Scopus (12) Google Scholar Other studies expect substantial declines3Creutzig F. Agoston P. Goldschmidt J.C. Luderer G. Nemet Pietzcker R.C. underestimated climate change.Nat. 2017; 2: 17140Crossref (346) Scholar, 4Victoria M. Haegel N. Peters I.M. Sinton Jäger-Waldau A. del Cañizo C. Breyer Stocks Blakers Kaizuka I. et al.Solar photovoltaics ready sustainable future.Joule. 2021; 5: 1041-1056https://doi.org/10.1016/j.joule.2021.03.005Abstract Full Text PDF (63) 5Beiter Cooperman Lantz Stehly T. Shields Wiser Telsnig Kitzing L. Berkhout V. Kikuchi Y. Wind driven innovation experience with further horizon.WIRE’s Energy Environ. https://doi.org/10.1002/wene.398Crossref (10) 6Wiser Rand J. Seel Beiter Baker Gilman Expert elicitation survey predicts 37% 49% 2050.Nat. 1–11Google but highlight key challenge more generally, system integration questions7Mills Changes Economic Value Generation High Penetration Levels: Pilot Case Study California. Lawrence Berkeley National Laboratory, 2012https://eta-publications.lbl.gov/sites/default/files/lbnl-5445e.pdfCrossref 8Mills A.D. R.H. Strategies economic high penetration California.Appl. 2015; 147: 269-278https://doi.org/10.1016/j.apenergy.2015.03.014Crossref (53) 9Hirth optimal share variable renewables: variability affects their welfare-optimal deployment.Energy 36https://doi.org/10.5547/01956574.36.1.6Crossref (87) 10Veers Dykes K. Barth Bottasso C.L. Carlson O. Clifton Green Holttinen H. al.Grand challenges science energy.Science. 2019; 366https://doi.org/10.1126/science.aau2027Crossref PubMed (212) Scholar). possibility losing race declining puts goals risk led calls ambitious research efforts cut technology11Sivaram Kann needs target.Nat. 2016; 1: 1-3Crossref direct support.12Davidson D.J. Exnovating transition.Nat. 4: 254-256Crossref (70) Although there no consensus exact provide, it clear relatively means penetrations feasible scope heavily influence VRE’s deployment potential. mechanics complex, producing varying periods. evidenced variety found empirical Europe,13Cludius Hermann Matthes F.C. Graichen merit order effect photovoltaic Germany 2008–2016: estimation distributional implications.Energy Econ. 2014; 44: 302-313Crossref (219) 14Sapio effects space time: A regime switching model Italian price.Energy Pol. 85: 487-499Crossref (20) 15Benhmad Percebois Photovoltaic feed-in impact prices: case Germany.Energy 2018; 119: 317-326Crossref (32) 16Gürtler Paulsen forecasts day-ahead intraday 75: 150-162Crossref (35) 17Di Cosmo Malaguzzi Valeri L.M. Wind, storage, interconnection generation.Energy 69: 1-18Crossref (14) 18Janda Slovak market price photovoltaics.Energy 122: 551-562Crossref (13) 19Frade Vieira-Costa J.V. Osório G.J. Santana J.J. Catalão J.P. Influence spot market: comparative study based real data.Energies. 11: 2974Crossref 20Figueiredo N.C. da Silva P.P.d. “Merit-order effect” power: volatility determinants.Renew. Sustain. Rev. 102: 54-62Crossref (19) 21Pereira Horta volatility: Iberian market.Int. 38: 794-813Crossref (9) 22Acar B. Selcuk Dastan S.A. river hydroelectricity Turkish prices.Energy 132: 1298-1319Crossref 23Maciejowska Assessing level – quantile regression approach.Energy 104532Crossref (36) 24de la Nieta A.S. Contreras Quantifying day–ahead Spanish case.Energy 90: 104841Crossref (11) 25Sirin S.M. Yilmaz B.N. technologies analysis merit-order effect.Energy 144: 111660Crossref (16) Australia,26Csereklyei Z. Qu Ancev Australia.Energy 131: 358-369Crossref (45) States.27Woo C.K. Horowitz Moore Pacheco spot-market variance: Texas experience.Energy 2011; 39: 3939-3944Crossref (237) 28Woo Schneiderman Ho Olson Alagappan Chawla Toyama Zarnikau Merit-order divergence California’s real-time markets.Energy 92: 299-312Crossref (109) 29Woo C.-K. Kadish Wang hydro-rich Pacific Northwest.IEEE Trans. Power Syst. 2013; 28: 4245-4253Crossref (49) 30Gil H.A. Lin PJM market.IEEE 3945-3953Crossref (64) 31Barbose Heeter Mai Bird Bolinger Carpenter Heath Keyser D. Macknick al.A retrospective benefits impacts U.S. portfolio standards.Energy 96: 645-660Crossref (84) These mostly focus concept related “merit effect,” quantifies deflation overall output. There only limited number specifically values differ values. Studies Texas32Tsai C.-H. Eryilmaz Effect ERCOT nodal 76: 21-33Crossref California33Bushnell Novan Setting sun: markets. Bureau Research, 2018https://www.nber.org/papers/w24980#:?:text=The%20Impacts%20of%E2%80%A6-,Setting%20with%20the%20Sun%3A%20The%20Impacts%20of,Energy%20on%20Wholesale%20Power%20Markets&text=Focusing%20on%20the%20California%20electricity,in%20utility%2Dscale%20solar%20capacityCrossref Scholar,34Prol J.L. Steininger K.W. Zilberman cannibalization California market.Energy 104552Crossref (38) (versus value), interesting dynamics other types (for example, improved flexible gas combustion turbine combined cycle plants). Brown Sullivan35Brown P.R. O'Sullivan F.M. Spatial temporal variation markets.Renew. 121: 109594Crossref electric nodes. They also evaluated 2010 2017. provide important insights US leave questions unanswered. Specifically, existing peer-reviewed do national assessment assess relative importance different underlying decline. topics first step toward designing effective understand root location technology. address gaps providing an grid-system utility-scale contiguous term “value” reflect revenue environment. analyze differences decade. then assess, plant, what difference this, decompose observed three separate causes: decomposition loosely builds concepts modeling efforts9Hirth Scholar,36Hirth Integration power.USAEE working paper no. 12-150. 2012; https://doi.org/10.2139/ssrn.2187632Crossref complements evaluations costs37Ueckerdt Hirth Edenhofer System LCOE: renewables?.Energy. 63: 61-75Crossref (329) 38Hirth Ueckerdt revisited framework variability.Renew. 74: 925-939Crossref (255) 39Hirth Why coal: economics 37https://doi.org/10.5547/01956574.37.3.lhirCrossref (50) 40Heptonstall P.J. Gross R.J. systematic review integrating grids.Nat. 6: 72-83Crossref (34) identify strategies.8Mills Scholar,41Bistline J.E. technical operations under large-scale 64: 363-372https://doi.org/10.1016/j.eneco.2017.04.012Crossref 42Odeh R.P. Watts Impacts spatial diversification value: Chilean market.Renew. 111: 442-461Crossref (25) 43Eising Hobbie Möst Future Germany—evidence dependencies?.Energy 86: 104638Crossref (27) 44Bistline Turn down what? operational flexibility markets.IEEE 34: 527-534https://doi.org/10.1109/TPWRS.2018.2856887Crossref 45Hirth flexibility: hydropower.Appl. 181: 210-223Crossref (121) compare recent trends forward-looking studies, context evolve future. take care develop realistic profiles generation—each modeled individually, resulting bias corrected combination records various sources, including hourly available. estimate plant-level curtailment basis characteristics, reports, pricing patterns. following sections, present value, followed detailed broken above-mentioned components. brief comparison trends. conclude discussion relevant planners makers. next explore differs (wind value). investigate drivers curtailment). examine plant’s impacted faces. aggregate information find differences. note our split parts, part focused value. comprises bulk section. Trends treated briefly, solutions Our energy-value motivated fact heterogeneous regions. Finally, briefly series insight likely near-term implications changes (value decline). quantify two “energy value” (the concurrent prices) “capacity additional peak demand hours, thereby offsetting capacity). That said, mentioned above, focuses creates reducing emissions criteria air pollutants greenhouse gases. However, environmental public health large46Fell Johnson J.X. Regional disparities reduction net trade renewables.Nat. 358-365https://doi.org/10.1038/s41893-020-00652-9Crossref (8) Scholar,47Millstein Barbose air-quality States.Nat. 17134Crossref (99) clearly motivating factor much support Currently, though some trading programs exist, external excluded (see Millstein al.47Millstein emission markets). exclude streams ancillary services. Ancillary services total amount needed small. Thus, typically make up less 1% power.48Denholm P.L. Sun T.T. An Introduction Grid Services: Concepts, Technical Requirements, Provision Wind. Renewable 2019https://www.energy.gov/sites/default/files/2019/06/f64/An%20Introduction%20to%20Grid%20Services%20Concepts%2C%20Technical%20Requirements%2C%20and%20Provision%20from%20Wind_72578.pdfGoogle Additionally, participate services, finally, Real-time “physical” prices, sense they depend actual dispatch realized conditions thus representation grid needs. excluding associated forecast uncertainty, “balancing” prior forecasts. comfortable exclusion balancing generally small compared costs.38Hirth Scholar,40Heptonstall summarize, captures internal (i.e., benefits). interested why penetration, capture factors contrast change estimated simulation model. simulations fidelity constraints systems, models alternative scenarios scenarios. Important terms: “value difference” single (as described below) particular year “Value reduction” premium” indicate negative positive difference. decline” describes trend penetration. sum calculate product specific price. Capacity credit rules independent operator (ISO), along profile. plants, generation-weighted taken Similarly, (non-VRE VRE) region, reported totals. case, assumption call “flat block” power, often refer block value.” comparator consistent factor” literature.34Prol Scholar,35Brown Scholar,49Hirth Market cost-competitive?.IET Renew. Gener. 9: 37-45Crossref (67) Additional included experimental procedures. topic fundamentally similar question “integration” costs.37Ueckerdt interpreted rough approximation increased costs, formulation does allow equivalence. Note, al.37Ueckerdt details equivalence Research findings. Throughout sections reference literature rely conceptual similarity discussion. seven markets, (CAISO), Electric Reliability Council (ERCOT), Southwest Pool (SPP), Midcontinent Independent Operator (MISO), Interconnection (PJM), New York (NYISO), England (ISO-NE). For convenience, operators ISOs Operators), although deemed (RTOs); ISO RTO unimportant context. organized geographically locational (LMP) 2019, were 50,000 LMP nodes covering linked neighboring At nodes, fine scale (often 5 min) security constrained unit commitment dispatch. computational tractability speeds available resolution. wind, significantly (Figure 1). CAISO ISO-NE, valuable power. correlated ISO. Figure 2 shows annual ratio value) developed time, 2012 2019. (represented 1.0) higher, SPP wind. Solar’s 1.4 when averaged years 3% (though wide range around average), dropped half, 0.7 higher almost CAISO). Wind’s 0.8 0.6 near 20%. 25% factor. solar’s exceeded demonstrated right panel 2), declined rapidly broader literature.8Mills Scholar,50Dobson Dwivedi Krishnaswamy Garcia Kiran Analysis Geysers geothermal Field, California.Geothermics. 87: 101871Crossref (6) 51Hirth renewables.Energy 218-236Crossref (474) 52Wiser Darghouth Hoen Mills Porter Widiss al.2016 Technolo