Slide 1Slide 2Slide 3Slide 4Wind Resources in New YorkSlide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 13Our approachMarginal emissions displacement profile: New York electricity market (Z = non-hydro generation)Putting early results in contextSlide 17Slide 18Greenhouse Gas Emissions Reductions from Wind Energy : Location, Location, Location?Duncan CallawaySNRE & Department of Mechanical [email protected] FowlieFord School of Public Policy &Department of [email protected] support from the University of California Energy Institute and the University of Michigan Energy Science, Technology, and Policy Award Program“The issue of how much generation of emissions is displaced by production of electricity generation through wind energy is complex, but it needs to be understood properly.... Emissions displacement analysis is essential for policy and regulatory decisions.”National Academy of Sciences (2007)How are these emissions displacement estimates used in the policy process?• Policy implementation EX: Used to distribute incentives to operators based on avoided emissions (e.g. the NOx Budget Program)• Policy design/ evaluation EX: Estimating (ex ante and ex post) emissions reductions achieved through policy interventions that promote renewables development.Estimating emissions displacementA seemingly simple calculation is complicated by three stylized facts: 1.Resource performance profiles vary across sites.2.Emissions displacement profiles vary across regions.3.Correlation between resource availability and emissions displacement rates.Annual average wind speed at 50 meters- Red = 11 m/s- Orange = 6.5 m/s - Blue = 3 m/sWind Resources in New YorkData from AWS Truewind5Stylized example : Wind resource profilesAverage production (% of peak capacity)*Profile 1: average capacity factor : 30%Profile 2 : average capacity factor : 30%Stylized example : Estimating emissions displacement•Next- estimate how system emissions would respond to the addition of a new wind resource.•Anticipate how the electricity market will respond to an increase in intermittent, low marginal cost power production at a given site.•This requires some understanding of how electricity markets work….Stylized example : Emissions displacement profilesRegion ANatural gas always on marginMarginal emissions rate (lbs CO2/MWh) •US average emissions rate of 2249 lbs CO2/MWh for coal generation.•US average emissions rate of 1135 lbs CO2/MWh for natural gas generation.113511352249 off-peak peak off-peak Region BCoal off peakNatural gas on peak off-peak peak off-peak Marginal emissions rate (lbs CO2/MWh) 8Stylized example : Estimated emissions displacementProfile 1 Profile 2Region A204 tons/day*(1135 lbs/MWh)203 tons/day*(1135 lbs/MWh)Region B301 tons/day*(1674 lbs/MWh)346 tons/day*(1939 lbs/MWh)* These estimates assume a 50 MW development lbs CO2/MWh) off-peak peak off-peak lbs CO2/MWh)off-peak peak off-peak 9Emissions reductions from wind energy: Location, location, location? •Do emissions displacement benefits vary significantly across wind sites in the U.S.?•What explains variation across locations a. variation in wind profiles?b. variation in emissions displacement profiles?c. correlation between (a) and (b) ?•What are possible policy implications of the variation we document?Research objectives•Possibly improve upon methods currently used to estimate marginal emissions displacement rates.•Estimate, with unprecedented precision, the marginal emissions displacement potential at wind sites across the Eastern U.S.•Investigate the relative importance of inter-temporal correlations in resource availability and marginal emissions displacement profiles.Past and present approaches to estimating emissions displacement•Regional system average emissions rates often used to estimate marginal operating emissions displacement rates (US DOE, AWEA).•Late 1990s/early 2000s, most serious studies used grid dispatch simulation models (ISO NE, US EPA, OTC, NESCAUM).•Significant efforts underway to develop /improve methods to estimating emissions displacement benefits. (CEC, 2004; ISONE, 2006; Gil and Joos, 2007; Keith et al., 2004; NAS, 2007; NREL, 2007; Schiller et al., 2007; US DOE, 2006; US EPA, 2004; WRI, 2007; etc.).•We are taking a regression-based approachm indexes market region, h indexes hour of day, t indexes time,Z is a proxy for changes in wind, X includes variables that affect dispatch /system operations.•Data:–Hourly, unit level emissions and generation from units in three regional markets (NY, NE, PJM).–Forecast load and RT load from ISOs and FERC 714–Imports / exports from ISOs–Other covariates (temp, fuel prices, etc)14Our approachmhtmhtmhtmhmhmhtZXe )'(15Marginal emissions displacement profile: New York electricity market (Z = non-hydro generation) NYISO system average OTC estimateLocationAverage CO2 emissions rate (lbs/MWh)EU ETSCO2 value1 ($/MWh)RGGICO2 value2 ($/MWh)Current policy incentive3 ($/MWh)Z=combustion generationProfile a 1162 $11.94 $1.78$35Profile b 1178 $12.11 $1.8116Putting early results in context1 The EU ETS price: approx. $22.61 / metric ton CO2 in October 20082 The RGGI price: $3.07 / ton CO2 in September 20083 All new wind eligible for Federal Production Tax Credit ($20/MWh). The weighted average price that NYSERDA paid for RPS Attributes for the most recent solicitation is $15 per Megawatt hour. 4 Source: FERC Oversight Electric Power Market SummariesEstimated marginal emissions displacement profiles by regionNYMISO EASTPJMNEY- axes measure lbs CO2/MWhNext steps…..•Continue to refine our model using data from New York and other Eastern states.•Take the model to the data- estimate MOERs using 3TIER data.•Generate site-specific estimates of emissions displaced by new wind development.•Address policy questions surrounding efficacy of production-based