DENVER AMENDMENT PROPOSAL 2015 INTERNATIONAL CODES

Transcription

1 DENVER AMENDMENT PROPOSAL 2015 INTERNATIONAL CODES UPDATED DEADLINE: JANUARY 9, 2015 NOTE: Each proposed Denver amendment to the International Codes must be justified. In order to be considered, the amendment proposal must address climate, clarity and/or cost. 1) Submitter Information Name: Mike Salisbury Date: 1/8/15 Organization (if Southwest Energy Efficiency Project applicable): Phone: E mail address: msalisbury@swenergy.org 2) *Signature: Mike Salisbury * I hereby grant and assign to City and County of Denver all rights in copyright I may have in any authorship contributions I make to City and County of Denver in connection with this proposal. I understand that I will have no rights in any City and County of Denver publications that use such contributions in the form submitted by me or another similar form and certify that such contributions are not protected by the copyright of any other person or entity. Note: electronic signatures are acceptable. 3) Indicate which International Code (I Code) you propose to amend (please use acronym): IRC If your code proposal requires an amendment to another I Code, please indicate which other I Code and section(s) is impacted, and provide the necessary language within the proposal form. See section below for list of names and acronyms for the International Codes. 4) Be sure to format your proposal and include all information as indicated on pages 2 3 of this form. 5) Send your proposal to Community Planning and Development (CPD), attention: Jill Jennings Golich, via at jill.jenningsgolich@denvergov.org. An e mail submittal should include an electronic version saved with a file name as follows: 2015_public_amendment_proposal_Codname_SectionReference, where Codename_SectionReference is replaced using the code and section being proposed (for instance 2015_public_amendmnet_proposal_IBC_Section312 ). The only formatting necessary is BOLDING, STRIKEOUT AND UNDERLINING. Please do not provide additional formatting such as this will be done by CPD. Please use a separate form for each proposal submitted unless as allowed in 3) above. Note: All amendment proposals received will receive an acknowledgment, and will be posted on the Building Code website to allow the public review. Please check here if a separate graphic file is provided. Graphic materials (graphs, maps, drawings, charts, photographs, etc.) must be submitted as separate electronic files in either PDF, JPEG or TIF format (300 DPI minimum resolution; 600 DPI or more preferred) even though they may also be embedded in your submittal. Acronym DBC AP IBC IECC IEBC IFC IFGC IMC Code Name Denver Building Code Administrative Provisions International Building Code International Energy Conservation Code International Existing Building Code International Fire Code International Fuel Gas Code International Mechanical Code Acronym IPC IRC NEC AMENDMENT PROPOSAL Code Name International Plumbing Code International Residential Code National Electrical Code

2 Please provide all of the following items in your amendment proposal. Your proposal may be entered on the following form, or you may attach a separate file. However, please read the instructions for each part of the amendment proposal. Code Sections/Tables/Figures Proposed for Revision: Note: If the proposal is for a new section, indicate (new). (NEW) IRC Section R324 Electric Vehicle Charging Options Proposal: Show the proposal using strikeout or underline format. At the beginning of each section, include one of the following instruction lines: Add new text as follows R324.1 Electric Vehicle Charging Options. For one or two family dwellings and townhouses, provide a minimum of: a. One 208/240 V 40 amp, grounded AC outlet, for each dwelling unit; or b. Panel capacity and conduit for the future installation of a 208/240 V 40 amp, grounded AC outlet, for each dwelling unit. The electrical outlet or conduit termination shall be located adjacent to the parking area. R324.2 For residential occupancies where there is a common parking area, provide one of the following: a. A minimum number of 208/240 V 40 amp, grounded AC outlets equal to 5 percent of the total number of parking spaces. The outlets shall be located within the parking area; or b. Panel capacity and conduit for future installation of electrical outlets. The panel capacity and conduit size shall be designed to accommodate the future installation, and allow the simultaneous charging, of a minimum number of 208/240 V 40 amp, grounded AC outlets, that is equal to 5 percent of the total number of parking spaces. The conduit shall terminate within the parking area. When the application of the 5 percent results in a fractional space, round up to the next whole number. Exception: If the electric panel is located in the parking area conduit does not need to be installed.

3 Supporting Information: Purpose: Add new requirements to the code that will help the City and County meet three of its 2020 Sustainability Goals for the Community. Reasons: Electric vehicles are a growing component of the light duty transportation sector both nationwide and in Denver. The promotion of electric vehicles helps the city meet three of its 2020 Sustainability Goals for the Community. Compared to a gasoline vehicle, electric vehicles reduce criteria pollutants (Air Quality Goal), greenhouse gases (Climate Goal) and energy use (Energy Goal). Please see the attached report from the Colorado Electric Vehicle and Infrastructure Readiness Plan (CEVIRP) which details the environmental and energy use benefits of electric vehicles in Colorado. This addition to the code will remove several barriers to electric vehicle acceptance in the City and County of Denver. One barrier is the higher incremental cost of electric vehicles which can be exacerbated by the need to install a new electric panel and wiring for a home charging station. Retrofitting a home to install a charging station can cost significantly more than preparing the home for electric vehicle charging during new construction. Owners of multi family housing, which makes up 45% of Denver s residential building stock, may be unable or unwilling to install a station at the property. Alameda Station, a new apartment complex with 338 units in Denver s Baker neighborhood, provides electric vehicle charging stations for residents. Providing residents of multi family housing an option to charge vehicles overnight at their residence will play an important part in opening up larger populations to the potential of owning an electric vehicle. The addition of these requirements for new construction, or major renovation projects, to the adoption of the 2015 I codes is an optimal time. The building code is an excellent area to address these barriers because making parking areas electric vehicle ready during initial construction is much less expensive than retrofitting the property at a later date. Retrofitting may require upgrades to the electrical panel and trenching and cutting to run additional wiring to the garage or parking area. Substantiation: An Electric Vehicle Readiness Study prepared by ConSol for the California Department of Housing and Community Development provides a number of estimates on the cost of preparing buildings for charging stations during construction versus having to retrofit at a later date. It estimates that the average cost of retrofitting a single or multi family home for Level 2 charging would be $3,500. The option of installing conduit to facilitate a future charging station at the time of construction is estimated to cost $50 for a single family home and $110 (per installation) for a multi family dwelling. The installation at the time of construction of a 40 amp, 240 volt dedicated branch circuit with a receptacle for a future charging station along with labor to install is estimated to cost $300 for single and multi family dwellings. KB Home, a major home builder in California and Colorado, has estimated that the cost of pre wiring a new home for Level 2 charging would cost approximately $250. The estimates for the cost of electric vehicle charging readiness correspond with those given by Ron Flax, the Boulder County official overseeing the implementation of the County s building code requirements for new single family homes to either have conduit, pre wiring or an actual charging station. A number of other jurisdictions have adopted codes that address electric vehicle charging readiness in the residential code as detailed in the paper An Electric Vehicle Charging White Paper from Fairfax County, Virginia. Examples of jurisdictions that have adopted new code language in the residential sector include: Los Angeles, CA, Boulder County, Lancaster, CA, Mountlake Terrace, WA, Palo Alto, CA, Rolling Hills Estates, CA, Santa Clara County, CA, Sunnyvale, CA, Salt Lake City, UT, Montgomery County, MD, Vancouver, BC and the State of Hawaii. Several municipalities including: Sunnyvale, Lancaster, Palo Alto and Vancouver require greater than 10% of new multi family units be EV ready (Palo Alto requires that 100% of new residential units be EV ready). Jurisdiction Single Family Multi Family Boulder County, CO X Los Angeles, CA X X City of Lancaster, CA X X Mountlake Terrace, WA X X Palo Alto, CA X X

4 Rolling Hills Estates, CA X Santa Clara, CA X X Sunnyvale, CA X X Vancouver, British Columbia X X Salt Lake City, UT X State of Hawaii X Montgomery County, MD X Bibliography: Autoblog California homebuilder offers to pre wire homes for electric vehicle charging. homebuilder offers to pre wire homes for electric veh/ ConSol Electric Vehicle Readiness Study. 2F%2Fwww.hcd.ca.gov%2Fcodes%2Fcalgreen%2Fev_readiness_report_complete.pdf&ei=LtGuVLKUK8jwoASGjYCoAg&usg= AFQjCNE23rGLJN_iuXiUIdt7CiIgd2N6Aw&sig2=kYwPlQ5mPCIr3sSb2aeBxw&bvm=bv ,d.cGU Fairfax County Electric Vehicle Charging White Paper. 2F%2Fwww.fairfaxcounty.gov%2Fplanning%2Fev_policy_white_paper_4_10_2014.pdf&ei=is6uVNSpFsiKoQTD9IHwCQ&usg =AFQjCNESCPcUcJ4gs4jYjnf2_LwpTzHqeg&sig2=_0BPjonD8rUkaIyf C4ZFg&bvm=bv ,d.cGU IMT Residential IMT Alameda Station. Jorgenson et al Emissions Changes from Electric Vehicle Use in Colorado. Appendix to the Colorado Electric Vehicle and Infrastructure Readiness Plan, Denver Metro Clean Cities, 2012 (Attached)

5 Referenced Standards: List any new referenced standards that are proposed to be referenced in the code and provide a minimum of one electronic copy. Should the amendment proposal be recommended for inclusion in the amendment package, you must provide two hard copies.

6 None Impact: Discuss the impact of the proposed amendment and answer the three questions below on the impact of the amendment proposal. The proposed amendment will ensure that new parking areas are prepared for electric vehicle charging. Effect of the proposed amendment on the cost of design: Increase Reduce No Effect Effect of the proposed amendment on the cost of construction: Increase Reduce No Effect Is the amendment proposal more or less restrictive than the I Codes? More Less Same Departmental Impact: To be filled out by CPD/DFD Note: The department shall indicate one of the following regarding the impact of the amendment proposal: Effect of the proposed amendment on the time to review: Increase Reduce No Effect Effect of the proposed amendment on the cost of enforcement/inspection: Increase Reduce No Effect

7 Emissions Changes from Electric Vehicle Use in Colorado Jennie Jorgenson 1,2, Jennifer Detlor 3, Gregory Brinkman 2, and Jana Milford 1 1 Department of Mechanical Engineering, University of Colorado at Boulder 2 National Renewable Energy Laboratory 3 University of Wisconsin at Madison November 7, 2012 Executive Summary Compared to vehicles operated on gasoline, plug-in hybrid electric vehicles and all-electric vehicles (EVs) have the potential to reduce petroleum use and to reduce emissions of several air pollutants. In part, electric vehicles shift emissions from the vehicle tailpipe to the power plants that produce the electricity needed to charge their batteries. Consequently, the net emissions impact of replacing gasoline-fueled vehicles with EVs depends on the type of power plants used to produce that electricity. This study uses well-to-wheels life cycle assessment to compare the energy use and emissions of light duty gasoline vehicles with those of EVs that could be in use in Colorado in the year The well-to-wheels analysis considers energy use and emissions from the stages of production or extraction of the feedstock for vehicle fuel, fuel processing, fuel transport and distribution, and vehicle operation. For gasoline vehicles, this means accounting for impacts of crude oil extraction and refining, delivery of gasoline to the gas station, vehicle refueling, and gasoline consumption in the vehicle. For EVs, the well-to-wheels assessment accounts for extraction and transport of natural gas, coal or other fuels used to generate electricity, electricity production, and ES-1

8 transmission and distribution of electricity to the vehicle charging station. For plug-in hybrid electric vehicles, both gasoline and electricity pathways are considered. The well-to-wheels life cycle assessment was conducted using Argonne National Laboratory s Greenhouse Gases, Regulated Emissions and Energy Use in Transportation (GREET) model, with key inputs tailored for Colorado in the year The primary scenarios examined assumed aggressive penetration of EVs into the vehicle fleet, so that 7.5% of vehicle miles traveled statewide in 2020 would be powered by electricity. The scenario assumed about 56% of EV mileage would be driven by plugin hybrids, split evenly between hybrids with nominal 10 and 40-mile all-electric ranges, and the remainder by all-electric vehicles. EVs were assumed to use 0.35 kwh of electricity per mile driven in electric mode. As noted above, previous studies have found that the mix of generating plants used to provide electricity for EV charging is a critical factor in determining how net EV emissions compare to those from gasoline vehicles. To address that factor, detailed unit commitment and dispatch modeling was performed for the projected electric power sector in Colorado in Dispatch modeling uses a least-cost approach to determine which generating units will be used to meet electricity demand on an hour-by-hour basis. We used the PLEXOS dispatch model with projections from the Western Electricity Coordinating Council (WECC) for the composition and characteristics of the generating fleet available to serve Colorado customers in 2020 and their forecast hourly demand for electricity. The WECC projection of the power plant fleet was altered to reflect current plans for power plant fuel switching and shutdowns and Colorado s renewable portfolio standards. ES-2

9 PLEXOS was run to model hourly dispatch for the year 2020 for a base case with negligible EV penetration, and then rerun for cases with added electricity demand for EV charging. The difference between these cases was examined as the incremental impact of EVs. Additional year-long model runs were conducted to examine sensitivity to coal and natural gas prices, the projected level of base electricity demand (before EV charging demand is added), and the profile of EV charging demand throughout the day. In addition to determining the effects on the power system, we also used the dispatch modeling results to estimate the increase in power plant emissions of sulfur dioxide, nitrogen oxides, and carbon dioxide that would result from new demand for EV charging. In the aggressive EV penetration scenario, with 7.5% of light duty vehicle miles traveled in Colorado in 2020 powered by electricity, the load on the electric generating system would increase by about 2% compared to the demand in WECC s base forecast. Our dispatch model results suggest that natural gas generation would produce about 44% of the extra electricity needed to supply the EVs, wind about 32% and coal about 24%. In a case with somewhat less aggressive EV penetration, such that about 5% of light duty vehicle travel is electric-powered, the load on the system would increase by about 1.2%. In that case our results suggest the extra demand would be met with a roughly even split of generation from natural gas, coal, and wind. In all of the cases we considered, the projected capacity of the power system in 2020 was adequate to accommodate increased demand for EV charging, with no additional unplanned infrastructure required. ES-3

10 The generating mix used to supply incremental electricity demand for EV charging shows little sensitivity to the pattern of charging over the day, but is somewhat sensitive to natural gas and coal price assumptions. Of the factors we considered, the incremental generating mix was most sensitive to the base-level demand forecast. If base level demand is reduced by 10%, more spare coal capacity exists in the system, and the model dispatches this resource to meet new demand for EV charging. With a 10% increase in base level demand, little spare coal capacity is available and the extra demand for charging is met mainly with natural gas. This pronounced shift is a consequence of examining the incremental impact of EVs as a new source of demand for electricity. In contrast, the total use of coal or natural gas is not that different across cases. Well-to-wheels energy use and emissions impacts of EVs in Colorado in 2020 were estimated using dispatch modeling results for the aggressive EV penetration scenario, with 44% of electricity for EV charging coming from natural gas plants, 32% from wind, and 24% from coal. Life cycle impacts were determined for a composite EV by weighting results for plug-in hybrid and all-electric vehicles based on their assumed share of travel. The EV impacts were then compared to those for average light-duty gasoline vehicles expected to be in use in Colorado in The PHEVs and gasoline vehicles are assumed to use gasoline blended with 10% ethanol by volume. Figure ES-1 shows well-to-wheels energy consumption for the composite EV and average gasoline vehicles, broken down by fuel coal, natural gas, petroleum or other. The other category includes nuclear power and renewables used to generate electricity, and biomass used to produce ethanol. In this case, EVs are found to be ES-4

11 Energy Consumption (Btu/Mile) more efficient overall (well-to-wheels), requiring about 60% of the energy per mile that is needed by gasoline vehicles. As expected, EVs are found to offer significant reductions in petroleum use but to require more coal and natural gas consumption than gasoline vehicles. 7,000 Well-to-Wheels Energy Consumption 6,000 5,000 4,000 3,000 2,000 Other Natural Gas Coal Petroleum 1,000 - Average Gasoline Composite EV Figure ES-1. Well-to-wheels energy consumption for electric vehicles in the aggressive EV penetration case, compared to that for average light duty gasoline vehicles. Figure ES-2 shows well-to-wheels emissions rates for the composite EV and average light duty gasoline vehicles. The estimated greenhouse gas emissions per mile for EVs are about 55% of those for gasoline vehicles. Emissions of volatile organic compounds, carbon monoxide, and nitrogen oxides are also estimated to be lower with EVs than with gasoline vehicles. On the other hand, well-to-wheels emissions of sulfur ES-5

12 Emissions (grams/mile) dioxide are about 10% higher for the composite EVs compared to gasoline vehicles, due mainly to the use of coal to generate a portion of the electricity for EV charging. Well-to-Wheels Emissions Average Gasoline Composite EV Figure ES-2. Well-to-wheels emissions for electric vehicles in the aggressive EV penetration case, compared to those for average light duty gasoline vehicles. For display purposes emissions rates for GHG are divided by 1000; those for CO are divided by 10. In addition to the main cases shown in the figures, GREET was also used to estimate well-to-wheels energy use and emissions for EVs with the electricity generating mix for EV charging determined from the dispatch modeling sensitivity cases with reduced or increased reference demand; for summer conditions (as opposed to year-round); and for a case with reduced PHEV fuel economy during operation on the gasoline engine. Although the magnitude of the estimated energy and emissions impacts varies across these cases, well-to-wheels emissions of greenhouse gases, carbon monoxide, volatile organic compounds and nitrogen oxides are consistently lower with EVs than with the comparison gasoline vehicles. ES-6

13 Introduction Electric vehicles (EVs) have the potential to reduce transportation-related emissions of several significant air pollutants, including the greenhouse gas carbon dioxide 1 as well as volatile organic compounds (VOCs) and nitrogen oxides (NOx) 2, which are precursors of ground-level ozone. Although EVs have no tailpipe emissions while operating on the electric motor, air pollution emissions are associated with generation of electricity needed to charge EV batteries. The net emissions impact of displacing gasoline-fueled vehicles with EVs thus depends on the mix of electric power plants used to charge the vehicles, among other factors. 3 This study estimates net emissions changes from displacing light duty gasoline vehicles with electric vehicles (including plug-in hybrids as well as all-electric vehicles) in Colorado in the year 2020, using projections of future development of the state s light duty vehicle and electricity generation fleets. A unit commitment and dispatch model was applied to estimate how electricity generation and corresponding power plant 1 Stephan, C. H. and J. Sullivan. Environmental and Energy Implications of Plug-in Hybrid- Electric Vehicles. Environmental Science and Technology 42 (4) (2008): EPRI, Environmental Assessment of Plug-In Hybrid Electric Vehicles, Vol. 2: United States Air Quality Analysis Based on AEO-2006 Assumptions for Electric Power Research Institute, Palo Alto, CA, 2007; Parks, K., P. Denholm, and T. Markel. Costs and Emissions Associated with Plug-In Hybrid Electric Vehicle Charging in the Xcel Energy Colorado Service Territory. NREL/TP , National Renewable Energy Laboratory, Golden, CO, 2007; Brinkman, G.L., P. Denholm, M.P. Hannigan, and J.B. Milford. Effects of Plug-in Hybrid Electric Vehicles on Ozone Concentrations in Colorado, Environmental Science and Technology, 44 (2010): Elgowainy, A., J. Han, L. Poch, M. Wang, A. Vyas, M. Mahalik, and A. Rosseau, Well-to- Wheels Analysis of Energy Use and Greenhouse Gas Emissions of Plug-In Hybrid Electric Vehicles. ANL/ESD/10-1, Argonne National Laboratory, June 2010; Anair, D. and A. Mahmassani. State of Charge: Electric Vehicles Global Warming Emissions and Fuel-Cost Savings Across the United States, Union of Concerned Scientists, Cambridge, MA, June Note that EVs could also be powered from distributed solar generators (e.g., residential roof top collectors) but for the purpose of this study all EVs are assumed to be powered from the grid. 1

14 emissions might change in response to introduction of EVs into Colorado s light duty transportation sector. The resulting emissions increases were compared to emissions from the light duty gasoline vehicles that EVs would displace, in order to estimate net emissions impacts. In addition to considering direct emissions from vehicle or power plant fuel combustion, a well-to-wheels analysis was conducted with Argonne National Laboratory s Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model 4 to estimate the overall life cycle impacts of EV use versus conventional gasoline-powered vehicles. Methods Unit Commitment and Dispatch Modeling Theoretically, power utilities strive to serve electricity demand in the least costly manner by adjusting the commitment of generating units to the load. Considering this, optimization tools like unit commitment and dispatch models can determine the least expensive way to deliver power. Such models must consider the physical constraints of the power transmission system as well as the limitations of each generating unit. For instance, each generating unit has a minimum and maximum generating capacity, startup and shut-down rate, and minimum up- and down-time. Each unit also operates under a specified efficiency with a given maintenance schedule. Unit commitment and dispatch models consider these constraints and solve a mixed-integer program to minimize the cost of electricity by considering the costs of fuel, operation and maintenance, and start-up. 4 Wang, M., Y. Wu and A. Elgowainy, Operating Manual for GREET: Version 1.7. ANL/ESD/05-3, February

15 In this study, the PLEXOS model 5 was used to simulate unit commitment and dispatch to supply projected electricity demand for Colorado in 2020, within the constraints of Colorado s projected electricity generating system for that year. In the next eight years, Colorado s electricity generating fleet is expected to increase wind, solar, and natural gas capacity while decreasing coal capacity. The PLEXOS model was run for the Public Service Company (PSC) and Western Area Colorado and Missouri (WACM) service territories in Colorado, with a sliding 24-hour unit commitment window to model the hourly operation of the electric power system. Projected Demand Profiles for 2020 The Western Electricity Coordinating Council (WECC) is interested in planning for future transmission and generation needs. The Transmission Expansion Planning Policy Committee (TEPPC) provides WECC with forecasts for guiding expansion planning in the western US. TEPPC publishes predicted hourly loads and annual reports for the WECC 10-Year Regional Transmission Plan. 6 This analysis uses information from the TEPPC 2020 Study Report Scenario PC1. Scenario PC1 represents a reference case in which the WECC balancing authorities submitted load forecasts for 2020 and then incorporated energy efficiency programs and related policies. Forecasting the magnitude of electricity demand is a source of uncertainty. Thus, the TEPPC demand level will be referred to as the base case demand level, that is, assumed demand without the incorporation of electric vehicle adoption. We also 5 Hobbs, Benjamin F., Glenn Drayton, Emily Bartholomew Fisher, and Lise Wietze. Improved Transmission Representations in Oligopolistic Market Models: Quadratic Losses, Phase Shifters, and DC Lines. IEEE Transactions on Power Systems, 23.3 (2008): WECC. 10-Year Regional Transmission Plan 2020 Study Report. Western Electricity Coordinating Council, September < 3

16 include a scenario in which additional energy efficiency policies are adopted and the reference demand level decreases by 10%, referred to as the reduced demand scenario. 7 Conversely, we also analyze a scenario in which reference demand increases by 10% (the increased demand scenario ), which represents a future with growth beyond what TEPPC expects. Increased penetration of EVs in 2020 may drive an increase in forecasts beyond these three reference case predictions. SWEEP provided the possible penetration levels of EVs by the year Table 1 details each penetration scenario. Charging demand was estimated assuming 56.6% of EV will be plug-in hybrids (PHEV) and 43.3% will be all-electric battery electric vehicles (BEV). The average efficiency of an EV operating on electricity is assumed to be 2.85 mi/kwh. This efficiency figure is assumed to include transmission, distribution and charging losses. On average, PHEVs are assumed to be driven 45% of their mileage on electricity with the balance on conventional gasoline. Here, we analyze mainly the highest penetration scenario to observe the most dramatic effect on grid operations, and conduct several sensitivity analyses for comparison with this scenario. We also include an analysis of the medium penetration scenario. 7 A recent study by SWEEP estimates that energy efficiency measures could reduce electricity demand in Colorado by 20% by 2020 (H. Geller, personal communication with Jana Milford, July 13, 2012). The lower reduction figure of 10% was used here because the TEPPC PC1 case already incorporates some energy efficiency measures. 8 Michael Salisbury, FEVER Assumptions and Methodologies Underlying Analysis, personal communication to Jana Milford and Jennie Jorgenson, March 13,

17 Table 1: Characteristics of each penetration scenario for Colorado in the year Scenario % of Light Duty Vehicle Miles Traveled Annual EV Charging Demand (GWh) % of projected annual load Baseline 1.9% % Medium 4.7% % Aggressive 7.5% 1, % Both the level of penetration as well as the charging schedule must be considered to fully characterize effects on electricity generation. SWEEP and icast provided the hourly EV demand profiles, 9 which were broken down by service territory, with 59% of total demand for EV charging in the PSCo service territory and 41% in WACM. Figure 1 shows the pattern of vehicle charging by hour for both weekday and weekend as a percent of total daily charging. To address uncertainty in this assumption, we also included a scenario in which the charging profile is flat, meaning that charging is evenly distributed throughout the day (Appendix Scenario D). 9 Michael Salisbury, Load Profile Methodology, personal communication to Jennie Jorgenson and Jana Milford, May 7, The load profiles are based on data from current PHEV and BEV-drivers that has been collected by the ECOtality project ( The ECOtality dataset includes areas with Time-of-Use electricity rates to incentivize night-time charging. 5

18 Figure 1: The expected breakdown of electric vehicle charging for weekends and weekdays. Electricity Generating Units for 2020 The TEPPC report also predicts the expected composition of the generator fleet for WECC. We altered this base fleet by reflecting current plans for power plant fuel switching and shutdowns and Colorado s renewable portfolio standards. The regional model for 2020 reflects generating unit shutdowns, fuel switching, and addition of new generating units that are planned in part to achieve compliance with the Colorado Clean Air-Clean Jobs Act. Table 2 shows the expected generation changes in Colorado. 10 WECC,

19 Table 2: Expected changes to fossil fuel-fired generating units in Colorado by Generator Arapahoe 3 Arapahoe 4 Cherokee 1 Cherokee 2 Cherokee 3 Cherokee 4 Cherokee CC Pueblo Airport CC 1 Pueblo Airport CC 2 Valmont 5 WN Clark 1 WN Clark 2 Zuni 1 Zuni 2 Action Shutdown Fuel Switch (Coal to Natural Gas) Shutdown Shutdown Shutdown Fuel Switch (Coal to Natural Gas) New Natural Gas Combined-Cycle Plant New Natural Gas Combined-Cycle Plant New Natural Gas Combined-Cycle Plant Shutdown Shutdown Shutdown Shutdown Shutdown Colorado Renewable Portfolio Standard (RPS) legislation calls for 30% of the state s utility electricity sales to come from renewable sources by the year The RPS also requires 10% renewable energy sales from cooperatives and municipal utilities. However, since the majority of electricity sales will come from utilities, we applied the 30% figure to all electricity sales in the region. The TEPPC 2020 report predicts that all renewable energy capacity growth in Colorado will come from wind and solar power, with the majority being from wind. The model incorporates future wind and solar plants using the methodology developed in the Western Wind and Solar Integration Study (WWSIS) prepared for the National Renewable Energy Laboratory (NREL). 11 Future wind resource profiles were derived from hourly historical data in the Numerical Weather Prediction Model, as described in the WWSIS report. The solar 11 GE Energy. Western Wind and Solar Integration Study. GE Energy, Schenectady, NY, May < 7

20 resource dataset was developed using a cloud cover model. 12 Sufficient wind and solar generation was added to the model to ensure that the Colorado RPS was met. Table 3 shows the projected generating capacity, by fuel type, of PSC and WACM in 2020, with and without aggressive EV penetration. Note that for modeling purposes, the wind generating capacity is increased slightly in the case with added electric vehicles since the generation used to meet the additional demand from EVs must also fulfill the RPS. Table 3: Generating capacity for the PSC/WACM region in Capacity (MW) No EVs With EVs Coal Natural Gas Oil Conventional Hydro Pumped Hydro Wind Solar Thermal Plant Characteristic Data Along with forecasted demand profiles, the TEPPC study includes predictions for fuel prices and thermal plant operation characteristics. In general, thermal characteristics such as variable operation and maintenance costs, maintenance hours, minimum down time, startup energy requirement, and ramp rates largely depend on 12 Perez, Richard et al. Improving the Performance of Satellite-to-Irradiance Models Using the Satellite s Infrared Sensors. Proceedings of the ASES Annual Conference, May After this study was completed, an error was discovered in the TEPPC fleet projections, which misidentified about 70 MW of existing natural gas generating capacity as coal-fired. This corresponds to about a 1% shift in the capacity for each fuel, and is negligible in comparison to other uncertainties associated with the modeling framework and with projections out to

21 technology type. We modeled the power system using thermal characteristics developed for the TEPPC 2020 study. 14 Our model also incorporates load-dependent heat rate information. EPA requires Continuous Emission Monitoring System (CEMS) reporting from generating units, which includes heat rate information for individual units. For most units, heat rate curves were derived from data reported to EPA s Air Markets Program for Average heat rate curves for the same generating technology were used for smaller units that did not have to report to EPA s Clean Air Markets program, and for new or repowered facilities. Fuel Costs The TEPPC study breaks down forecasted gas and coal prices by region. The annual average gas prices are from Henry Hub, but are adjusted monthly in the model as shown in the TEPPC Assumptions Matrix. 16 Coal prices for 2020 remain constant through the year but depend on region. Forecasted oil prices are constant in time and region. Table 4 summarizes the fuel price assumptions. 14 WECC, Assumptions Matrix for the 2020 TEPPC Dataset, 2011, available at < %20TEPPC%20Dataset.pdf> The dataset includes both fuel and non-fuel variable operating and maintenance (O&M) costs, including those associated with operating existing pollution control equipment. Potential increases in O&M costs to meet future control requirements (e.g., for mercury emissions from coal-fired power plants) are not included. Based on EPA s cost estimates for typical coal-fired power plants, the O&M costs for mercury emissions control are generally expected to be small compared to total dispatch costs assumed in this analysis. See US EPA (2010) Documentation for the IPB EPA Base Case v. 4.10, Emissions Control Technology. 15 U.S. Environmental Protection Agency, Air Markets Program Data, < 16 WECC, Assumptions Matrix,

22 Table 4: Fuel price assumptions for PSC and WACM service regions in Fuel Average Price ($/MMBtu) Natural Gas PSC $4.11 Natural Gas WACM $4.09 Coal PSC $1.42 Coal WACM $1.42 Oil $19.99 Due to the current low price of natural gas and uncertainty in the prediction for the price in 2020, we also examined the sensitivity of the results to the price of both natural gas and coal, as discussed more fully in Scenarios A and B of Appendix A. Emissions In addition to determining the effects on the power system, we also analyzed the effect of additional generation on the overall emissions from power plants. To quantify the release of carbon dioxide from fossil fuel combustion, we used regional average emission factors, calculated to be lbs/mmbtu fuel for coal and lbs/mmbtu fuel for natural gas, based on data reported to the EPA s Clean Air Markets Database. 17 The overall sulfur dioxide (SO 2 ) emissions were determined for each generating unit and scenario as the product of an emissions factor determined for each unit and the annual heat input for the unit. Emissions factors (lbs SO 2 /MMBtu heat input) were determined from historical data reported to CAMD for 2010, with adjustments to reflect new emission limitations for certain plants as detailed in Colorado s Regional Haze 17 U.S. Environmental Protection Agency, Air Markets Program Data for 2010, < 10

23 State Implementation Plan. 18 Emissions factors for new natural gas units were determined based on average emissions for similar technology. Nitrogen oxides (NOx) emissions from fossil fuel power plants were determined by multiplying the hourly electricity generation from each unit during each hour of the year by output-based emissions factors determined primarily from 2008 CAMD data. Because power plant NOx emissions can vary significantly depending on combustion conditions, they were calculated using emissions factors that varied from hour-to-hour as a function of the load factor. Distinct emissions rates were determined based on the percent of maximum capacity generation used each hour, with one factor used for operation at levels below 50% of capacity and separate factors for each increment of 10% above 50% load. The quantities of NOx produced each hour were then summed over all hours of the year to determine the total NOx produced for each unit. The emission rates for Comanche 3, which began operating after 2008, were calculated by fraction of total load from the hourly data reported in CAMD for Wygen3, Rawhide1, Rocky Mountain3, and Pueblo Airport emissions were determined by averaging the NOx emission rates of similar units at the same facilities. For all other facilities with missing NOx emission data, combined cycle (CC) and combustion turbine (CT) average emissions rates were used from facilities of about the same maximum capacity generation. In addition, as detailed in the Colorado and Wyoming 19 Regional Haze State Implementation Plans, more stringent NOx limits will be applied between 18 Colorado Department of Public Health and Environment. Regional Haze State Implementation Plan January < 19 Wyoming Department of Environmental Quality. Wyoming State Implementation Plan, Regional Haze: Addressing Regional Haze Requirements for Wyoming Mandatory Federal Class I Areas Under 40 CFR (g). January < 11

24 now and 2020 for nine coal-fired units and two repowered natural gas units in Colorado and two coal-fired units in Wyoming. The historical NOx emission factors for these facilities were multiplied by a unit-specific fraction to reflect the expected reductions in emissions rates. Life Cycle Analysis with GREET To compare well-to-wheels life cycle impacts of gasoline vehicles and EVs, we applied Argonne National Laboratory s Greenhouse Gases, Regulated Emissions and Energy Use in Transportation (GREET) model, version 1_ GREET analyzes the well-to-wheels (WTW) energy use and emissions of alternative fuel and advanced vehicle systems. WTW analysis tracks energy use and emissions from the primary energy source through the vehicle s operation. This analysis is divided into two stages. The well-to-pump (WTP) stage analyzes the production or extraction of the fuel feedstock, fuel processing, and fuel transport and delivery to the refueling or charging station. The pump-to-wheels (PTW) stage analyzes the impacts of vehicle operation. For PHEVs, the WTW analysis is split into three distinct parts: electricity use in charge depleting (CD) operation, gasoline use in CD operation, and gasoline use in charge sustaining (CS) operation. 21 CD mode is vehicle operation that primarily depends on the battery pack, with the potential for some gasoline use for ancillary purposes. For PHEVs, CD mode occurs at start-up until the battery is at its minimum state of charge threshold. The vehicle will then switch to CS mode where the vehicle is operating on gasoline. 20 Wang et al., 2007, with updates as available from through March Elgowainy et al.,

25 The GREET model has been widely used for life-cycle analysis of fuel and vehicle systems, including PHEVs and EVs. 22 The model provides a default description of energy use and supply systems in the US transportation sector, including relevant attributes of electric vehicles (and many other alternative fuel and vehicle technologies), with projections out to GREET allows the user to modify most inputs, including vehicle characteristics that influence pump-to-wheels energy use and emissions, and characteristics of the upstream fuel and electricity production and delivery system that influence the well-to-pump stage of the analysis. For this study, a number of changes were made to the default settings to tailor the analysis to Colorado. The most significant changes made to GREET default values were determined by the results of the dispatch modeling and emissions analysis, or represented shared assumptions within the FEVER project. They included: 1. Electricity Generation: marginal generation mixes for Colorado EV charging scenarios were input based on the dispatch modeling results; 2. Efficiencies and emissions rates for electrical generating units providing electricity for EV charging were input based on the dispatch modeling results; 3. Electric vehicle efficiencies (0.35 kwh/mi) were input based on the shared assumption for the FEVER project; Liquid fuel used in passenger cars was assumed to be conventional gasoline blended with 10% ethanol. 22 Elgowainy et al., 2010; Anair, D. and A. Mahmassani. State of Charge: Electric Vehicles Global Warming Emissions and Fuel-Cost Savings Across the United States, Union of Concerned Scientists, Cambridge, MA, June Michael Salisbury, FEVER Assumptions,

26 5. Emissions rates and fuel consumption for light duty gasoline vehicles in Colorado in 2020 were input based on information provided by the Colorado Department of Transportation and the Denver Regional Council of Governments. 24 The light duty vehicle fleet was assumed to be comprised of 57% passenger cars and 43% light duty trucks, 25 with an overall annual average fuel economy of 22.7 mpg and a summertime average fuel economy of 22.8 mpg. The corresponding emissions rates are shown in Table 5. Table 5: Light duty gasoline vehicle emissions (g/mile). Annual Summer CO CO NOx VOC exh VOC evap Additional changes included changing source locations and transport distances and modes for coal, oil, natural gas, and ethanol 26. For 2020, the shale gas share of natural gas production was assumed to be 40% and the share of ethanol produced from 24 Michael Salisbury, personal communication with Jana Milford, August 13, 2012; Sabrina Williams, Colorado Department of Transportation, personal communication with Jana Milford, August Sabrina Williams, Colorado Department of Transportation, personal communication with Jana Milford, August 24, For petroleum: Energy Information Administration. Petroleum Supply Monthly. July < for natural gas and coal: correspondence from Kathryn Valdez, Xcel Energy, personal communication with Jennie Jorgenson, April 2012; for ethanol: EAI Inc. Denver/North Front Range Fuel Supply Costs and Impacts. Report for the Regional Air Quality Council, March

27 corn to be 95%, based on the most recent national projections from the U.S. Energy Information Administration. 27 These changes were relatively minor in their impact. We examined results for PHEV 10 and PHEV 40 vehicles, with energy use and emissions determined separately for the CS and CD modes. The percentage of miles driven on electricity was assumed to be 30% for PHEV 10 and 60% for PHEV 40. Following Elgowainy et al. (2010) 28, we assumed PHEV 10 and PHEV 40 vehicles would consume 1542 Btu/mile and 187 Btu/mile of gasoline, respectively, while operating in CD mode. Also following Elgowainy et al. (2010), 29 we assumed the fuel economy of PHEV 10 and PHEV 40 vehicles operating in CS mode would be 157% and 122%, respectively, of the average fuel economy for an all-gasoline vehicle. The importance of this assumption was tested through analysis of an alternative case in which PHEV fuel economy in CS mode was set equal to the average for gasoline vehicles. Composite results for the 2020 EV fleet were calculated by weighting results 27 Energy Information Administration. Annual Energy Outlook, June This analysis used GREET s default estimates of emissions from unconventional and conventional natural gas production, processing, and transmission (Burnham et al., 2011). Although Petron et al. (2012) have suggested that emissions rates in northeastern Colorado may have been higher than the GREET default estimates, the GREET default values are used because they are consistent with the central tendency of national values reported in the recent literature, and because this study requires projections for 2020, not estimates of past emissions rates. The upstream natural gas emissions rates may warrant reconsideration as leakage rates become better understood and new practices and regulatory requirements take effect. See for example, Burnham, A., et al. Life-Cycle Greenhouse Gas Emissions of Shale Gas, Natural Gas, Coal, and Petroleum, Environmental Science and Technology, 46(2): , 2011; Petron, G. et al., Hydrocarbon Emissions Characterization in the Colorado Front Range: A Pilot Study, Journal of Geophysical Research, 117, D04304, 2012; Skone, T., Life Cycle Greenhouse Gas Inventory of Natural Gas Extraction, Delivery, and Electricity Production, DOE/NETL-2011/1522, National Energy Technology Laboratory, U.S. Department of Energy, October, Elgowainy et al., Elgowainy et al.,

28 for PHEV 10, PHEV 40 and BEV by 28%, 28% and 44%, respectively, based on the breakdown of vehicle types provided by SWEEP and icast. 30 Results Dispatch Modeling Interpretation of Results The WACM/PSC service region must obey an overall electrical energy balance. Here, consumption refers to demand from power purchasers as well as pump load used for pumped hydro energy storage. This electrical energy balance is shown concisely below: In the following results, the left-hand side of the above equation is referred to as Energy In, while the right-hand side represents Energy Out. Since we could not determine the air quality implications of energy imported from other regions, we ran the base case to determine the overall imports and exports and fixed them at the base case levels for the rest of the scenarios. Dispatch Modeling Aggressive EV Penetration Case Ultimately, we aim to estimate the overall emissions impact of electric vehicle adoption on the state of Colorado. To this end, we must first determine how the electricity sector will meet the incremental demand. As stated above, we first report the effects of the most aggressive penetration of EVs in 2020 (7.5% of light-duty vehicle miles traveled on electricity) to observe the most notable changes to the power 30 Salisbury, FEVER Assumptions,

29 generation sector. Table 6 shows the overall generation differences between a case with no significant EV penetration and a case with aggressive EV adoption. Table 6: Electricity balance in PSC/WACM with and without aggressive EV penetration* Base Case - No EV Penetration Aggressive EV Penetration Generation Type GWh % Generation Type GWh % Coal 47, % Coal 48, % Natural Gas 9, % Natural Gas 10, % Conv Hydro 3, % Conv Hydro 3, % Pumped Storage % Pumped Storage % Wind 16, % Wind 16, % Imports % Imports % Solar 4, % Solar 4, % Total Energy "In" 83,149 Total Energy "In" 84,611 Exports (3,611) 4.3% Exports (3,611) 4.3% Pumped Load (1,124) 1.4% Pumped Load (1,072) 1.3% Load (78,415) 94.3% Load (79,944) 96.1% Total Energy "Out" (83,150) Total Energy "Out" (84,627) *Number of digits displayed for GWh generation was chosen to display differences across cases and does not imply significance for absolute results in individual cases. Table 7 depicts the margin, or the change resulting in the system due to the incremental charging demand from EV adoption. As the table indicates, natural gas generation provides the plurality (44%) of marginal demand. Wind and coal generation also provide a significant portion, 32% and 24% respectively. 17

30 Table 7: Change in total generation by type resulting from aggressive EV adoption. Incremental Difference Generation Type GWh Coal 360 Natural Gas 662 Conv Hydro - Pumped Storage (36) Wind 475 Imports - Solar - Total Energy "In" 1,462 Exports - Pumped Load (52) Load 1,529 Total Energy "Out" 1,477 Figure 2 shows the total demand and generation from the study region. Note that demand and generation are not necessarily equal at any given time due to the imports and exports of the system. Additionally, the dashed lines indicate how demand and generation change with the added electricity demand resulting from EV charging. The time-series graph shows hourly data from four 3-day periods. The graph includes the day with the highest overall demand (July 22), the lowest overall demand (April 19), and two more typical sets of days, including weekdays and weekends. Figures 3 and 4 depict the generation profile for the same days, with and without EV use. 18