Evaluation of Hydrogen Production and Utilization under Various Renewable Energy Curtailment Scenarios

Transcription

1 Evaluation of Hydrogen Production and Utilization under Various Renewable Energy Curtailment Scenarios Prepared for the: State of Hawaii Department of Business, Economic Development and Tourism Under Supplemental Contract No. 1 for Contract #60514 for the NETL Grant Project Subtask 1.5 Deliverable Electric Vehicle Charging / Hydrogen Fuel Cell Vehicles Prepared by Select Engineering Services (SES) Submitted by: Hawai i Natural Energy Institute School of Ocean and Earth Science and Technology University of Hawai i July 2013

2 Evaluation of Hydrogen Production and Utilization under Various Renewable Energy Curtailment Scenarios Purchase Order Number: #Z , Task 1 24 June 2013 Prepared By: Select Engineering Services (SES) 531 Cooke Street Honolulu, Hawaii, For: Hawaii Natural Energy Institute (HNEI) 2800 Woodlawn Drive, Suite 200, Honolulu, Hawaii, P age

3 Acronyms CO2 DOE EV EPA FCV GE NREL OEM KOH PPA SMR SUV Carbon Dioxide Department of Energy Electric Vehicles Environmental Protection Agency Fuel Cell Vehicle General Electric National Renewable Energy Laboratory Original Equipment Manufacturer Potassium Hydroxide Power Purchase Agreement Steam Methane Reformation Sports Utility Vehicle 2 P age

4 Contents 1.0 Executive Summary Introduction/Background Relevant Hydrogen Production Technologies Electrolysis Parameters Hydrogen Compression, Storage and Dispensing Technology Hydrogen Production System Energy Requirements Fuel Cell Vehicle Technologies Analysis Curtailed Energy Captured and Hydrogen Produced Fuel Cell Vehicles Supported Emissions from Hydrogen Production and Use in Fuel Cell Vehicles System Costs Storage Capital Costs Electrolyzer Capital Costs Total Storage and Electrolyzer Capital Costs Electricity Costs Operation & Maintenance Costs Total Yearly Costs Costs per Kilogram Hydrogen and Vehicle Miles Conclusion Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F 35 3 Page

5 Figures Figure 1. Hourly Hydrogen Storage Levels for Base 1 with 100 MW Electrolyzer and 200,000 Kg Storage Capacity Assuming Evenly Distributed Hourly Consumption Tables Table 1. Performance Characteristics for Available FCVs... 9 Table 2. Curtailed Energy Cases Table 3. Maximum Curtailed Energy Captured as a Percentage of total Possible for Each Base Case with Various Electrolyzer Capacities Table 4. Efficiency Ratio (Percent energy captured / electrolyzer capacity) Table 5. Maximum Amount of Hydrogen Production for Each Base Case Using Various Electrolyzer Capacities Table 6. Total Number of Light Duty FCV's Supported with Infinite Storage Table 7. Total Number of Heavy Duty FCV's Supported with Infinite Storage Table 8. Hydrogen Storage Level Requirements for Maximum Number of Vehicles Supported Table 9. Performance Metric Summary for the Various Base Cases and Electrolyzer Strategies Table 10. Additional Thermal Energy Required for a 25% and 50% Reduction in Storage Capacity Table 11. CO 2 Emissions Reduction from Hydrogen Production from Curtailed Renewables in the Above Scenarios Table 12. Displaced Petroleum by Light Duty FCVs Using Hydrogen Produced from Curtailed Renewables in the Above Scenarios Table 13. Total Storage Capital Costs to Support Maximum Number of Vehicles for Each Base Case Table 14. Total Electrolyzer Capital Costs Table 15. Total Capital Costs (Storage + Electrolyzer only) Table 16. Storage and Electrolyzer Capital Costs per Year (Amortized Over 40 Years) Table 17. Electricity Costs Using Maui PPA for Table 18. Estimated O & M Costs per Year Table 19. Total Costs per Year Including Contribution of Electrolyzer and Storage Capital (Amortized over 40 Years), Electricity and Estimated O & M Table 20. Costs per Kilogram of Hydrogen Table 21. Cost ($/Mile) Light Duty Vehicles Table 22. Cost ($/Mile) Heavy Duty Vehicles P age

6 1.0 Executive Summary A study was performed by GE that explores four theoretical generation cases (600, 800, 1000, 1000 MW) of variable renewable energy to be theoretically added to the Oahu electricity grid. These large capacities of wind turbines and photovoltaic panels can create copious amounts of excess power that may go unused. A hydrogen production and storage facility can be effectively used to reduce the amount of curtailed renewable energy from the four theoretical generation cases. Additionally hydrogen can be used as a fuel in fuel cell vehicles (FCVs) which can serve to replace conventionally fueled vehicles on the road. The ability of multiple hydrogen infrastructure configurations to reduce the amount of curtailed renewable energy was studied considering hydrogen producing electrolyzer capacities between MW storage vessel capacities between 65, ,000 kg and the number of FCVs that could be supported Electrolyzer capacity ranges were determined with the goals of minimizing the amount of curtailed renewable energy and maximizing the number of FCVs supported given the magnitude and availability of curtailed renewable energy from the GE report. Both lightduty and heavy duty FCVs were considered in this analysis. The amount of reduced petroleum and emissions that result from replacing conventionally fueled vehicles with FCV and hydrogen produced from renewable energy was derived. System costs were evaluated and presented as $/kg and $/mile driven for the various elements of a hydrogen production/storage facility and expanded over an expected plant lifetime of 40 years. Installed electrolyzer capacity is the key driver to the amount of hydrogen that could be produced from the amount of energy available and thus the amount of FCVs. It was found that a 1 MW electrolyzer will capture less than 1% of all curtailed energy for all base cases. Increasing electrolyzer capacities resulted in approximately linear hydrogen production up to 50MW where at least 25% of available curtailed energy is captured for all Base Cases. An electrolyzer capacity of 200MW would capture up to 90% of available curtailed energy, thus capacities of 50, 100, 150 and 200 MW were considered in the study. The most efficient electrolyzer capacity per percentage curtailed energy used was Base 4, 50MW. This configuration only supported a minimal amount (~4700) of vehicles and used 38% of the available curtailed energy for that base case. The storage requirements ranged from 65,080 kg (Base 2, 50MW) to 559, 280 kg (Base 4, 200MW). The least amount of storage required to support the maximum number of vehicles for a given electrolyzer size is consistently Base 2, with an average of ~9 kg/vehicle. The next lowest storage requirement per vehicle (Base 3) is more than two times this average at ~21 kg/vehicle. The highest case is Base 4 averaging ~33 kg/vehicle. 5 P age

7 Decreasing storage requirements lowers the capital cost. It was also found that refueling times have less than 3% influence on required storage levels for the given situation and that the reduction of available curtailed energy in the September to November time frame is the main driver to determine required amount of storage to support a given number of vehicles. By reducing hydrogen storage amounts up to 50%, up to 18% of the energy required hydrogen production energy would need to be derived from thermal generation but if the thermal generation was installed for this purpose only, it would sit idle for 2/3 to 3/4 of the year If only curtailed renewable energy is used, these various electrolysis capacities supported vehicle numbers ranging from 4,025 (Base 1, 50MW) to 23,620 (Base 2, 200MW) for lightduty FCVs and 196 to 1,152 for heavy duty FCVs. Calculated via simply replacing fossil fuel powered vehicles on the road, for the analyzed cases, between approximately 22.5 million and 131 million tons of CO 2 could be reduced per year by the use of FCVs. Gasoline reduction for the analyzed cases ranges from approximately 2.5 million gallons to approximately 12.1 million gallons. Electrolysis systems are modular for the sizes evaluated in this report and thus large differences in cost are not expected between centralized and distributed (forecourt) systems. Although large up front capital costs 1 will be incurred to purchase electrolyzers and storage capacity, the price per mile over a 40 year plant life (including electricity and estimated O & M costs 2 ) are most favorable in Base Case 2. Values range from $0.15 (200 MW) to $0.23 (50 MW) per mile, which is in the range of an average conventional fossil fuel powered vehicle ($0.19/mile). 2.0 Introduction/Background Installing up to 1 GW of renewable energy capacity to supply Oahu with electricity raises many issues, including how grid operators manage the variable nature of renewables. This problem is compounded if any large electric load such as an electric vehicle (EV) fleet is also added to the Oahu grid and not controlled. Hydrogen can be used as an effective energy storage medium that can help mitigate problems with supply variability, increase the amount of energy captured from variable renewable energy sources, and be used to power FCV's with similar or better performance characteristics to EVs, although FCV's are not yet readily available. 1 Between lowest $80.5 Million (Base 2, 50MW) and highest $461.7 Million (Base 4, 200MW) 2 Anticipated costs not included are deprecation, equipment replacement, utilities other than electricity, engineering design, equity financing, taxes, land, permitting and fees, decommissioning and profit. 6 P age

8 A specific analysis was performed on the viability and associated costs of a hydrogen infrastructure (i.e. electrolyzers and storage vessels) on Oahu that consumes would be curtailed renewable energy to generate fuel for FCV fleets. Also, values were derived for the amount of petroleum and CO 2 emissions that would be displaced by implementing FCV s. 2.1 Relevant Hydrogen Production Technologies Hydrogen gas can be produced on demand in various ways such as using electrical energy and a feedstock such as water, natural gas or potassium hydroxide. This report focuses on the production of hydrogen using electrolysis. A more energy efficient method of hydrogen production using Steam Methane Reformation (SMR) is not considered in this analysis as natural gas is not yet available on Oahu. Also, the process is not considered suitable to the given application which requires a fast ramp rate of the system due to the intermittent and quickly variable nature of renewable energy. Additionally, SMR emits the pollutant CO 2, unlike electrolysis. Further details of the respective electrolysis methods, and a comparison, are provided in Appendix A Electrolysis Parameters Energy consumption and cost were of primary interest in the evaluation of electrolyzers for this study. The largest PEM electrolyzer readily available produces 65 kg/day (~0.174 MW) of hydrogen but this system is to be scaled up to 736 kg/day (~ 2 MW) within two years with a target cost of approximately $1 Million/MW. The largest alkaline electrolyzer available produces 1623 kg/day of hydrogen and is priced at about $1.3 Million/MW. These values are consistent with values presented in case studies 3 completed by the DoE. Original Equipment Manufacturer (OEM) published 4,5, 6,7 energy consumption figures (kwh/kg) for both of the evaluated electrolyzer technologies are within 20% (above) of DoE's published 8 current energy consumption rate of 50 kwh/kg. This consumption rate includes the additional system equipment, or "balance of plant" 9, required for the electrolyzers to operate as designed hydrogen.com/home/?pid= pressure electrolysers.html 7 solutions/industrial hydrogen generators byelectrolysis/outdoor installation/hystat trade Includes pumps, fans, computer control and safety system monitoring electronics etc. 7 P age

9 2.2 Hydrogen Compression, Storage and Dispensing Technology Hydrogen fuel stored onboard a FCV is compressed to provide acceptable driving ranges. The onboard storage pressures are typically either 350 or 700 bar (the FCV industry moving towards 700 bar). Fueling station storage pressures are typically either 450 or 850 bar in order to facilitate cascade filling, a procedure that involves using a pressure differential to "fast fill" the vehicle (as oppose to compressing from a lower storage pressure when the vehicle is connected). The estimated total number and frequency of fills are major contributing factors to the inherent design of a single fueling station. This study evaluates the broad hydrogen infrastructure on Oahu and thus assumes a constant refueling frequency, where hourly aggregate fuel consumption is distributed evenly throughout the year, and 850 bar storage. 2.3 Hydrogen Production System Energy Requirements Throughout this report, the worst case all in energy consumption (80 kwh/kg) to produce and deliver hydrogen is used. This is made up of ~60 kwh/kg for electrolysis (and associated electrolysis balance of plant equipment) and ~20 kwh/kg for compression, storage, dispensing of the hydrogen (includes energy conversion inefficiencies and computer control systems, instrument air compressors, lighting etc. energy usage). 2.4 Fuel Cell Vehicle Technologies Light duty FCVs are new to Oahu and are not yet commercially available. Current availability in the rest of the US is limited to lease commercial fleets. Large scale adoption is hindered by the small number and distribution of fueling stations (10 public stations in the US 10 as of June 2013) and the costs of the vehicles 11. Heavy duty FCVs are currently in use in the form of municipal transit buses in places such as British Columbia and California. Published data for available FCVs is given in Table 1. Throughout this report, a fuel economy of 55 mi/kg is used for light duty vehicles (as observed in 2012 fuel cell vehicle performance testing on Oahu) and 8 mi/kg is used for buses (heavy duty). In an attempt to compare FCVs with EVs, an annual range of 11,000 miles was assumed for light duty vehicles 12 and 32,800 miles for heavy duty vehicles 13. Refueling stations in Hawaii are currently limited to US military bases and are not publically accessible FCV vehicle price to consumers cannot accurately be predicted at this time. Toyota has repeatedly publically estimated that its small SUV will be priced "somewhere between $50K and $100K" HNEI Oahu Electric Vehicle Survey Technical and Descriptive Report, Market Trends Pacific; January Typical Honolulu Public Transit Bus travel 8 P age

10 Table 1. Performance Characteristics for Available FCVs Make Model Year Fuel Fuel Tank Economy Range Capacity Pressure mi/kg mi kg Psi Vehicle Class Type of Fuel Cell Honda 14 FCX Clarity Midsized Car PEM Mercedes Midsized Benz 15 F Cell Car PEM General Motors Fuel Cell EV SUV PEM New Fuel Cell Heavy Flyer Bus FC3 Duty PEM New Flyer 7 Fuel Cell Heavy bus FC2 Duty PEM 3.0 Analysis The feasibility of utilizing curtailed renewable energy to generate hydrogen for FCV use was assessed according to the amount and availability of curtailed renewable energy. Various configurations of renewables, electrolyzer and storage capacities were evaluated to determine the amount of FCVs that could be supported by the generated hydrogen, and the resultant costs per kilogram of hydrogen. Fueling times and rates for FCVs were varied in the analysis, but little effect (less than 3%) on system requirements was observed. Thus, for simplicity, a constant hourly refueling rate was calculated by evenly dividing the aggregate fuel requirement for all FCVs for a year. It was initially assumed that only as available curtailed renewable energy would be used to produce hydrogen. The effects of varying levels of storage were then evaluated. The analysis resulted in: i. the capacity of hydrogen production equipment necessary to utilize portions of the curtailed energy for four different curtailed energy cases, ii. the amount of FCVs that such systems could support, iii. the amount of displaced petroleum, iv. the amount of emissions from hydrogen production and from use in FCVs, and v. the costs of the hydrogen production, storage and fueling systems clarity/specifications.aspx fuels fleet 9 P age

11 Four curtailed energy cases, provided from a previous GE study on the integration of renewables on Oahu, were evaluated for hydrogen production. These cases, shown in Table 2 as Base 1 through Base 4, represent the addition of various combinations of wind and solar renewable energy capacities onto the Oahu electricity grid. Table 2. Curtailed Energy Cases System Capacity (MW) Total Curtailed Energy Solar (GWh) Cases Total Wind PV Base Base Base Base The GE Study provided hourly average curtailed power values for each month of the year for each Base Case. Due to the resolution of the data, the power was assumed to be constant for each hour for the purpose of energy calculations (i.e. each day has 24 different power values and curtailed power within each hour is assumed to be constant for each hour's duration). The amount of hydrogen produced from the curtailed renewable energy and subsequently, the number of vehicles that could be supported, was determined based on the following three inputs, each with underlying assumptions: 1. The capacity of electrolyzers operating on the Oahu grid, assuming a. electrolyzers, which typically can ramp to full capacity in a matter of seconds, exactly follow the hourly curtailed power profiles, b. a control network exists that matches distributed or centralized electrolyzer load to hourly curtailed power, c. an electrical network exists that is able to deliver curtailed renewable energy to the hydrogen infrastructure, and d. the amount of electric energy required to produce dispensed hydrogen (by weight) is 80 kwh/kg The capacity of hydrogen storage available, assuming a. hydrogen is compressed and stored as it is produced and b. if the storage is full, no additional hydrogen is produced. 3. Vehicle performance limitations, assuming a. fuel cell vehicle efficiency is 55 mi/kg for light duty and 8 mi/kg for buses 18, and 17 See Section 2 10 P age

12 b. the aggregate FCV fuel requirement is distributed evenly throughout the year such that hourly refueling rates are constant (due to the previously mentioned negligible effects that varying fueling frequency has on hydrogen storage requirements). 3.1 Curtailed Energy Captured and Hydrogen Produced The amount of curtailed energy that could be captured by a hydrogen production facility and the subsequent amount of hydrogen that could be produced from this energy was calculated. The size of aggregate electrolyzer and storage capacity were two factors that limited the amount of curtailed energy that could be captured. The hourly curtailed power values from the GE study were evaluated to determine the upper and lower limits for electrolyzer capacities to be analyzed. It was found that power magnitudes less than 1 MW occurred 50 75% of the time in the four Base Cases (Appendix B), although this range of power resulted in less than one percent of total available energy over a whole year for all Base Cases. A 1 MW electrolyzer, running at full capacity when the curtailed energy is larger than 1 MW, captures less than 1% of the available curtailed energy in all Base Cases. A lower limit for yearly captured curtailed energy of 25% was chosen as below this quantity wasn't satisfactory for this analysis 19. Capturing this amount of energy necessitates an aggregate electrolyzer capacity of at least 50 MW. The upper limit for aggregate electrolyzer capacity was chosen to be 200 MW. This electrolyzer capacity was found to capture upwards of 75% of curtailed energy in all Base Cases and greater than 90% in Base 4. Beyond this size, increased electrolyzer capacity results in diminishing returns of captured curtailed energy (except in Base 2) due to a majority of the hourly power values being less than 200 MW. The effects of varying electrolyzer capacities were determined by evaluating capacities in steps of 50 MW between 50 and 200 MW. Smaller electrolyzer capacities resulted in less energy captured, while larger capacities resulted in reduced electrolyzer usage 20. The maximum amount of curtailed energy that could be captured by these various electrolyzer strategies, given in Table 3, was calculated under the assumption that storage capacities were not a limiting factor and system efficiency was 80 kwh/kgh See Table 1 19 Although it was found that hydrogen output for electrolyzer capacity below 50 MW tended to scale linearly with captured curtailed renewable energy. 20 See Section P age

13 Table 3. Maximum Curtailed Energy Captured as a Percentage of total Possible for Each Base Case with Various Electrolyzer Capacities Maximum Curtailed Energy Captured as % of Available by Electrolyzer Capacity Total Available Energy Case 50MW 100MW 150MW 200MW (MWh) Base % 62.0% 78.6% 89.2% 209, Base % 47.0% 63.6% 76.3% 735, Base % 59.1% 76.5% 87.9% 400, Base % 64.0% 80.5% 90.7% 261, Table 4. Efficiency Ratio (Percent energy captured / electrolyzer capacity) Efficiency Ratio Case 50MW 100MW 150MW 200MW Total Available Energy (MWh) Base , Base , Base , Base , While the Base MW strategy captured the most available energy and Base MW strategy captured the highest percentage of energy, they are not the most efficient outcomes. These results indicate that the amount of curtailed energy captured does not scale linearly with increasing electrolyzer capacities at this magnitude. When comparing the different Base Cases, the highest ratio of percentage of energy captured to electrolyzer capacity (Efficiency Ratio) is consistently Base 4. This fact is due to Base 4 power magnitudes being consistenly higher than other Base Cases. When comparing the electrolyzer capacities, a 50 MW electrolyzer system yields the highest Efficiency Ratio resulting in the Base 4 yielding the highest Efficiency Ratio (0.76%/MW energy captured to electrolyzer capacity), see Table 4. This configuration, while allowing the electrolyzer system to be used to its greatest potential, may not yield the amount of fuel necessary for a wide adoption of FCVs because it will result in only 4700 light duty, or 229 heavy duty vehicles being supported, and 62% of curtailed energy not being captured. The maximum amount of hydrogen that could be generated from the curtailed energy captured by the various electrolyzer capacities (listed in Table 3 Graphed in Appendix C) is shown in Table 5. Again, storage capacities were assumed to be infinite and system efficiency was assumed to be 80 kwh/kgh P age

14 Table 5. Maximum Amount of Hydrogen Production for Each Base Case Using Various Electrolyzer Capacities Maximum H2 Produced (kg) by Electrolyzer Capacity Case 50MW 100MW 150MW 200MW Infinite Base 1 (210 GWh) 957,048 1,625,607 2,060,133 2,338,269 2,625,000 Base 2 (736 GWh) 2,370,643 4,319,686 5,844,894 7,013,680 9,200,000 Base 3 (401 GWh) 1,699,282 2,958,541 3,832,798 4,404,116 5,012,500 Base 4 (262 GWh) 1,242,309 2,091,670 2,630,087 2,963,435 3,275,000 When storage components (cylinders, compressors, etc.) are incorporated with the electrolysis system, a hydrogen system acts as an energy storage medium capable of capturing a large portion of the curtailed renewable energy for each Base Case that may have otherwise been wasted. Once stored, this hydrogen can then be used for on demand vehicle refueling. Limiting storage capacities were found to have significant effects on FCVs supported and excess thermal energy required, as discussed in the next section. 3.2 Fuel Cell Vehicles Supported The amount of both light duty and heavy duty (bus) FCVs that could be supported using hydrogen produced from curtailed renewable energy was calculated. Assumptions and boundary conditions were made to narrow the scope of the analysis. It was assumed that light duty vehicles have a fuel economy of 55 miles/kgh 2 and travel 11,000 miles per year. Fuel cell buses were assumed to have a fuel economy of 8 miles/kgh 2 and travel 32,800 miles per year. At first, hydrogen production was limited to only curtailed energy from renewables. Hydrogen storage levels were added as a constraint. If the storage was full and more curtailed renewable energy was available, no hydrogen would be produced. An hourly analysis was then performed to determine the effects of various vehicle refueling schedules. It was found that, due to the availability of storage, no significant change in vehicles supported (maximum <1%) or storage requirements (maximum 3%) occurred if the vehicles fueled evenly throughout the day or within a four hour window between 5:00 pm and 9:00 pm. Thus, it was assumed that hydrogen consumption was distributed evenly throughout the year. An example of hourly storage levels with a generation configuration of Base MW electrolyzer capacity and 200,000 kg of storage (equating to ~6500 light duty vehicles 13 P age

15 supported) fueling evenly throughout the year is provided in Figure 1. This example shows the large reduction in stored hydrogen over the September to November time inherent to each Base Case. The long duration of reduced availability of the curtailed energy, combined with another "low point" in the mid February timeframe, were the determining factors for the quantity of hydrogen storage required to support a given number of vehicles using only curtailed renewable energy. Evaluating only a single year, results in full storage tanks at the end of the year. A multi year analysis could help determine if more vehicles could be incrementally supported over time with the given infrastructure, but if the curtailed energy quantities were identical, no more vehicles could be supported without increasing storage capacity. Figure 1. Hourly Hydrogen Storage Levels for Base 1 with 100 MW Electrolyzer and 200,000 Kg Storage Capacity Assuming Evenly Distributed Hourly Consumption The maximum number of FCVs that could be supported using the hydrogen produced by a given electrolyzer capacity, using only curtailed renewable energy, were calculated by assuming that storage levels were infinite. The results are directly related to the amount of curtailed renewable energy available in each Base Case and the aggregate electrolyzer capacity. It was found that vehicle numbers ranged from 4,025 to 23,620 for light duty FCVs and 196 to 1,152 for heavy duty FCVs, as shown in Table 7. Please see Appendix D for graphs of total number of light duty and heavy duty miles and storage capacities for each Base Case and electrolyzer capacity. 14 P age

16 Table 6. Total Number of Light Duty FCV's Supported with Infinite Storage Light Duty Vehicles Supported Case 50MW 100MW 150MW 200MW Base Base Base Base Table 7. Total Number of Heavy Duty FCV's Supported with Infinite Storage Heavy Duty Vehicles Supported Case 50MW 100MW 150MW 200MW Base Base Base Base The amount of vehicles possible for heavy duty FCVs is significantly less than what is possible for light duty vehicles because heavy duty vehicles have nearly 1/7 th of the fuel economy and travel 3 times as far in a year. The largest amount of vehicles supported occurred in Base 2 with a 200 MW electrolyzer capacity, while the smallest amount of vehicles occurred in Base 1 with a 50 MW electrolyzer capacity, as is shown in Table 7. This result is due to the largest amount of energy available in Base 2 and the largest amount of that energy captured by a 200 MW electrolyzer system. Given the amount of FCVs, the minimum levels of storage capacity were calculated for the above number of vehicles such that only curtailed renewable energy was used. Storage levels were similar for both light duty and heavy duty FCVs and ranged from 65,000 to 559,000 kg as is shown below in Table 8. Table 8. Hydrogen Storage Level Requirements for Maximum Number of Vehicles Supported Light Duty Vehicle H 2 Storage Requirements (Kg) Case 50MW 100MW 150MW 200MW Base 1 100, , , ,710 Base 2 65, , , ,670 Base 3 83, , , ,700 Base 4 106, , , , P age

17 As previously mentioned, the large capacity of storage required is primarily driven by a large lull in curtailed renewable energy in the GE model in the September to November time frame and the assumption that vehicle fueling and hydrogen production begin simultaneously at the beginning of the year. Base 1 required the largest storage to vehicle ratio (an average of 34 kg/vehicle) of all the Base Cases (Appendix E). Within each Base Case, larger storage levels were required for larger electrolyzer capacities because larger electrolyzer capacities can support more vehicles and thus storage levels are more susceptible to changes in available curtailed renewable energy. An absolute maximum number of FCVs that can be supported was calculated by assuming a Perfect Tracking scenario for Base 2 (largest amount of available curtailed energy) in which all energy available is converted to hydrogen and consumed in vehicles. The maximum number of light duty FCVs possible is 45,975 vehicles and the maximum number of heavy duty FCVs is 2,240 vehicles. These vehicle fleets require a 553 MW electrolyzer capacity and storage capacity of 766,560 kg. The Perfect Tracking scenario results in more vehicles than the electrolyzer limited strategy described above, because all available curtailed renewable energy is converted to hydrogen and consumed in FCVs. It is not limited by electrolyzer or storage capacity. Also, all hydrogen produced is consumed by the end of the year. It is noted that the evaluation techniques used in this section of the report could be applied to derive storage requirements for other mixes of fleet vehicles, such as those used by the U.S. Department of Defense. In an effort to determine which Base Case and electrolyzer capacity yields the most efficient configuration, a comparison was made between the maximum vehicles possible, the total curtailed energy of each Base case, and the various electrolyzer capacities. The metric was calculated as 1000 A summary of the values of these metrics is provided in Table 9. Table 9. Performance Metric Summary for the Various Base Cases and Electrolyzer Strategies Minimum Metric Scenarios Maximum Metric Scenarios Average FCV Type Metric Value Base Case Electrolyzer Capacity Metric Value Base Case Electrolyzer Capacity Metric Value Buses MW MW Light Duty Vehicles MW MW P age

18 For both light duty and heavy duty FCVs, the metric decreased as electrolyzer size increased. This means that if the electrolyzer capacity is doubled, the vehicles supported may only be increased by a factor of 1.7. This is due to reduced electrolyzer usage with larger electrolyzer capacities, as discussed in Section 3.4. Base 1 (500MW of wind and 100MW of solar) consistently yields the highest metric of all the cases except for a 200 MW capacity electrolyzer, in which Base 4 is higher. An electrolyzer capacity of 50 MW in Base 1 yields the highest metric of all strategies. Though this strategy requires a larger amount of storage relative to other Base Cases (13% above average), it provides the most vehicle miles per electrolyzer and renewable energy capacity. The previous analysis has considered storage levels large enough to prevent the need for any additional electricity beyond curtailed renewables. A study was performed on lightduty vehicle strategies to determine the effects of reducing storage levels and allowing for electricity to be sourced from places other than curtailed renewables (it should be noted that energy purchased from the grid may not experience the same reduced costs that renewables do). Storage levels were reduced by 25% and 50% from the minimum levels required to produce hydrogen from only curtailed renewables, while the number of vehicles expected to be supported was kept constant. The results, shown in 17 P age

19 Table 10, give the amount of energy required from additional sources (typically thermal generation) to satisfy the same number of supported vehicles. 18 P age

20 Table 10 (Graphed in Appendix F) also shows the percentage of total energy to produce hydrogen that comes from non renewable sources for each strategy. 19 P age

21 Table 10. Additional Thermal Energy Required for a 25% and 50% Reduction in Storage Capacity Light Duty Vehicle 25% Storage Reduction >> GWh uncovered 50MW 100MW 150MW 200MW Base Base Base Base Percent from non renewable sources Base 1 5.4% 6.2% 6.7% 7.0% Base 2 3.1% 3.3% 3.8% 3.4% Base 3 2.5% 3.2% 4.3% 5.1% Base 4 4.9% 5.9% 7.0% 7.9% Light Duty Vehicle 50% Storage Reduction >> GWh uncovered 50MW 100MW 150MW 200MW Base Base Base Base Percent from non renewable sources Base % 15.8% 16.6% 17.1% Base % 13.8% 17.3% 16.9% Base 3 9.0% 7.7% 9.7% 11.3% Base % 12.8% 16.1% 18.4% The above analysis shows that the amount of non renewable energy (as a percentage of total energy used for hydrogen production) required for a 25% and 50% reduction in storage increases with increasing electrolyzer capacities. The large amount of vehicles expected to be supported by larger electrolyzer capacities resulted in the storage levels being more sensitive to changes in curtailed renewable energy and thus can lead to larger requirement for additional energy. The highest additional energy requirement per Base Case for a 25% reduction in storage capacity occurred for Base 4 at 6.4% on average and 7.9% maximum, with a 200 MW electrolyzer capacity. Base 1 had the highest additional energy requirement, on average, for the 50% storage reduction scenario, while Base 4 had the highest single value with a 200 MW electrolyzer capacity at 18.4% of total energy used to produce hydrogen. Bases 1 and 4 each have much less installed renewable power capacity and thus more hydrogen must be produced from thermal generation sources than in Bases 2 and 3. The costs associated with installing additional thermal generation capacity to cover a reduction in hydrogen storage was not analyzed as part of this study. It is noted that it would be 20 P age

22 required for only 3 4 months of the year, thus if installed for this purpose only, would sit idle 3/4 2/3's of the year. 3.3 Emissions from Hydrogen Production and Use in Fuel Cell Vehicles Initially electrolysis was assumed to use only renewable energy. Since FCVs emit only water, no harmful emissions would be released in the production of hydrogen or driving of FCVs. In fact, harmful emissions would be reduced to the extent that FCV miles displace miles from conventionally fueled vehicles. Calculations were performed using data published by the Environmental Protection Agency (EPA) and DOE to determine the amount of petroleum and CO 2 emissions that would be displaced by various light duty FCV configurations only from vehicle consumption of fuel. Emissions generated and energy used during production of petroleum based fuels was not included. Table 11 and Table 12 show the amount of CO 2 and petroleum reduced in a year, respectively, calculated by replacing light duty gasoline vehicles with FCVs for the amount of vehicles supported by each Base Case and electrolyzer capacity. Table 11. CO 2 Emissions Reduction from Hydrogen Production from Curtailed Renewables in the Above Scenarios Emissions Reduction (kilograms) 21 50MW 100MW 150MW 200MW Base 1 22,484,110 38,879,355 48,822,639 54,604,267 Base 2 43,041,011 78,736, ,342, ,944,019 Base 3 33,516,686 59,352,464 79,322,823 93,818,789 Base 4 26,254,737 45,247,526 59,827,284 70,077,804 Table 12. Displaced Petroleum by Light Duty FCVs Using Hydrogen Produced from Curtailed Renewables in the Above Scenarios Petroleum Reduction (gallons) 22 50MW 100MW 150MW 200MW Base 1 2,068,781 3,577,321 4,492,210 5,024,182 Base 2 3,960,238 7,244,588 9,968,698 12,140,275 Base 3 3,083,897 5,461,068 7,298,557 8,632,342 Base 4 2,415,719 4,163,261 5,504,756 6,447, EPA study on CO 2 emissions 8,887 gco2/gal mpg, as supplied by HNEI 21 P age

23 Vast amounts of petroleum and airborne pollution would be displaced by any of the hydrogen system scenarios. This is largely due to using renewable energy to produce the hydrogen and the nearly zero emissions (except water) of a FCV. As shown in Section 3.2, if storage levels were reduced, then excess energy would be needed to support the same amount of vehicles over the year. The amount of energy needed is in the GWh range and typically occurs in the span of 3 4 months. Therefore, energy would likely need to be sourced from large fossil fuel generators which would result in increased harmful emissions. It is also possible, depending on the number of vehicles required and hydrogen system configuration, that additional generation capacity would need to be installed to support certain FCV infrastructures. 3.4 System Costs A top level evaluation of costs was completed. The evaluation included the three largest cost elements of a hydrogen production and dispensing facility, namely, capital and operation & maintenance (O&M) of the storage and electrolysis systems and the reoccurring electricity costs. Consideration was not given to costs associated with anticipated depreciation, equipment replacement, utilities other than electricity, engineering design, equity financing, taxes, land, permitting and fees, decommissioning and profit 23. Anticipated prices for 2020 were used Storage Capital Costs Storage capital costs were determined based on the Hydrogen Storage Cost Analysis, Preliminary Results report published by Strategic Analysis for DOE 24. A storage system capital cost of $468/kg was taken from the Strategic Analysis Sensitivity Analysis, given an assumption that 500,000 units will be produced in To support the number of vehicles detailed in Tables 6 and 7 (Referencing Table 8 storage requirements for maximum number of vehicles) storage capital costs would range from $30.5M (Base 2, 50 MW) to $261.7 (Base 4, 200MW), as shown below in 23 Further in depth financial analysis would be required before investment P age

24 Table P age

25 Table 13. Total Storage Capital Costs to Support Maximum Number of Vehicles for Each Base Case Total Storage Capital Costs ($ Million) Case 50MW 100MW 150MW 200MW Base Base Base Base Electrolyzer Capital Costs The largest electrolyzer available on today's market is 3.5 MW and although currently priced at $1.3 Million/MW, is expected to fall to ~$1 Million/MW by Therefore a 200 MW capacity will require an approximate $200M capital outlay, as shown below in Table 14. Maximum electrolyzer size is not expected to increase substantially. Due to the relatively large amount of electrolysis capacity required to capture the quantities of curtailed energy analyzed in this report, the systems would need to be modular (in the form of multiple 3.5 MW units). Since centralized and distributed (forecourt) systems will be composed of the different amounts of the same components, they are not expected to be largely different in cost per MW. This was confirmed via comparing DOE models of both centralized and distributed systems that resulted in less than 1 % difference 25 in costs. Table 14. Total Electrolyzer Capital Costs Total Electrolyzer Capital Costs ($ Million) Case 50MW 100MW 150MW 200MW All Total Storage and Electrolyzer Capital Costs Table 15 below shows the total storage and electrolyzer capital costs for the maximum number of vehicles that can be supported. Table 15. Total Capital Costs (Storage + Electrolyzer only) Total Capital Costs ($ Million) Case 50MW 100MW 150MW 200MW Base It is noted that this comparative analysis did not consider location specific factors that may impact the economic attractiveness of each option on Oahu (i.e. costs associated with land purchase, road transport, piping installation, electrical supply system enhancements, etc.), because the required information is not readily available. 24 P age

26 Base Base Base Table 16 below shows the storage and electrolyzer capital costs per year amortized over a 40 year expected plant lifetime. Table 16. Storage and Electrolyzer Capital Costs per Year (Amortized Over 40 Years) Capital Cost/Year ($ Million Over 40 Years) Case 50MW 100MW 150MW 200MW Base Base Base Base Electricity Costs A PPA between Maui Electric Company and a wind farm company 26 was used to determine the cost of electricity. Typical PPAs of this nature define a price for a set amount of energy. If more than this set amount of energy is purchased, a reduced rate is then offered on future energy purchase. For this analysis, the rates for 2020 from the PPA were used. The first 83 GWh of energy delivered will cost $ per MWh and any additional energy will cost $57.17 per MWh. Shown below in Table 17 are the total electricity costs per year, again based on the maximum number of vehicles supported ( Table 6 and Table 7). Table 17. Electricity Costs Using Maui PPA for 2020 Electricity Costs/Year ($ Million ) 50MW 100MW 150MW 200MW Base Base Base Base MECO and Auwahi Power Purchase Agreement # ; June 15, P age

27 3.4.5 Operation & Maintenance Costs A very top level approximation of operating costs is shown below in Table 18. This was calculated simply via 10% of capital (plus electricity) costs and although not a standard method of calculation, is considered a reasonable proportion on experience. Table 18. Estimated O & M Costs per Year Estimated O & M Costs /Year ($ Million ) Case 50MW 100MW 150MW 200MW Base Base Base Base Total Yearly Costs Shown below in Table 19 is the total yearly costs including contribution of electrolyzers and storage (amortized over 40 Years), electricity and estimated O & M. Table 19. Total Costs per Year Including Contribution of Electrolyzer and Storage Capital (Amortized over 40 Years), Electricity and Estimated O & M. Total Costs / Year ($ Million) Case 50MW 100MW 150MW 200MW Base Base Base Base Costs per Kilogram Hydrogen and Vehicle Miles Figure 19 below shows costs per kilogram of hydrogen using total costs above and amount of hydrogen produced in a year as per Table 8. Table 20. Costs per Kilogram of Hydrogen $/Kilogram Case 50MW 100MW 150MW 200MW Base Base Base Base P age

28 Table 21 below shows costs per mile for light duty vehicles, noting that Base 2 (100, 150 and 200 MW) are below average conventional vehicle costs at $0.19/mile. Table 21. Cost ($/Mile) Light Duty Vehicles $/Mile Light Duty Case 50MW 100MW 150MW 200MW Base Base Base Base Table 22 below shows costs per mile for heavy duty vehicles. Table 22. Cost ($/Mile) Heavy Duty Vehicles $/Mile Heavy Duty Case 50MW 100MW 150MW 200MW Base Base Base Base Conclusion This study evaluated the installation of hydrogen production, storage and dispensing infrastructure on Oahu, to reduce the amount of curtailed renewable energy from four theoretical electricity generation cases: Base MW wind, 100 MW Solar PV; Base MW wind, 300 MW Solar PV; Base MW wind, 500 MW Solar PV and Base MW wind, 300 MW Solar PV. It was shown that a hydrogen infrastructure can significantly reduce the amount of curtailed renewable energy. The use of hydrogen produced from curtailed energy serves to reduce air pollution via replacing petroleum based fuel usage on Oahu. When stored as a compressed gas, the hydrogen can be used to fuel FCVs, as well as materials handling vehicles, backup power systems, unmanned aerial vehicles etc. Large capital investments are required to capture a useful amount of the curtailed energy. These capital investments are greater than would be required for hydrogen generation systems that run solely on grid 27 P age

29 power, due to the reduced availability and additional storage requirements. Although the initial investment is greater, when preliminarily 27 evaluated over a 40 year cycle, the costs per mile driven for an FCV are less than an average gasoline vehicle ($0.19/mile) for Base Case 2, 100, 150 and 200 MW installed electrolyzer capacity. The analysis resulted in several key facts for the following four categories: 1) electrolyzer capacities, 2) storage capacities, 3) supported FCVs and 4) system costs. 1) Analysis of various electrolyzer sizes specifically showed that: For an aggregate electrolyzer capacity between 1 and 50 MW, the number of vehicles supported and energy captured was approximately linear. This was not the case above 50 MW installed capacity. Consequently, calculating the best strategies for electrolyzer capacity required the use of stepwise, iterative evaluation techniques. 200 MW of installed electrolyzer capacity will capture approximately 90.7% of the curtailed energy for Base 4 (highest) and 76.3% for Base 2 (lowest). This fact is important because it is the reason that 200 MW capacity was used as an upper limit for the analysis. To capture at least 25% (considered an acceptable amount to justify investment) of the available curtailed energy for each base case (lowest, Base 2), at least 50 MW of electrolyzer capacity is required. The highest amount of curtailed energy that can be captured with a 50 MW electrolyzer capacity is 38% (Base 4). A 1 MW electrolyzer will capture less than 1% of all curtailed energy for all base cases. As mentioned, curtailed energy captured above 50 MW electrolyzer capacity does not scale linearly. The best efficiency is Base 4, 50 MW (at 0.76%/MW) but this will 27 Using only capital costs, PPA electricity prices and estimated O & M costs. This does not include anticipated deprecation, equipment replacement, utilities other than electricity, engineering design, equity financing, taxes, land, permitting and fees, decommissioning and profit etc. 28 P age

30 support only 4700 light duty vehicles (or 229 heavy duty) with 62% of the availably energy not being used 28. Not considering storage and consumption limitations (number of vehicles), 200 MW installed electrolyzer capacity would result in a highest hydrogen production mass of 7,013,680 kg (Base 2) and lowest case 2,338,269 kg (Base 1) 2) Analysis of various storage capacities specifically showed that: It made essentially no difference when the FCVs were refueled with regards to time frame on a daily/weekly basis. Refueling times have minimal effects on required storage levels of hydrogen (maximum 3%) and thus associated costs etc. to accommodate higher refueling frequencies for the same number of vehicles is minimal. The reduction of available curtailed energy for all base cases in the September November time frame (most severe in November) has a significant effect on the size of storage required and consequently the number of vehicles that can be supported in the first year 29. End of year storage levels were not taken into consideration but do not increase the number of vehicles that can be supported without increasing storage if the next year has an identical energy production profile. Storage requirements 30 ranged from 244,670 kg to 100,240 kg for these highest (Base 2, 200MW) and lowest (Base 1, 50MW) supported FCV cases. The least amount of storage required to support the maximum number of vehicles for a given electrolyzer size is consistently Base 2, with an average of ~9 kg/vehicle. This is less than half the requirement of Base 3, the next lowest at average ~21 kg/vehicle. The highest is Base 4 at average ~33 kg/vehicle. This is important when evaluating capital costs. 3) Analysis of the overall results from the various configurations showed that: At 200 MW electrolysis capacity, supported vehicle numbers ranged from 4,025 (Base 1, 50MW to 23,620 (Base 2, 200MW) for light duty FCVs and 196 to 1,152 for heavy duty FCVs. 28 (100% (0.76%/MW x 50MW)). 29 This was not an artificial constraint as the curtailed energy data supplied by GE was calculated from real world, location specific irradiance and wind data, gathered over an extended time period. 30 Storage requirements are a result of the maximum number of vehicles that could be supported using the given electrolyzer capacity and curtailed energy fluctuation so that the station would not run out of hydrogen in the lull periods. 29 P age

31 A "perfect tracking" scenario where all hydrogen and curtailed energy is used yields 45,975 light duty FCVs and 2,240 heavy duty FCVs but requires 553 MW of electrolyzer capacity and 766,560 kg of storage (Base 2), which is not considered a feasible approach as a large amount electrolyzer capacity would sit idle for extended periods and capital costs would be extreme. Comparing electrolyzer strategies (installed capacity) for the Base Cases which proportions maximum number of vehicles supported to the total amount of available curtailed energy and electrolyzer capacity (a metric to allow numerical comparison), reveals that doubling electrolyzer capacity only increases number of vehicles supported by a factor of 1.7. This is due to the reduced electrolyzer usage (e.g. less energy is available from curtailed renewables in the MW range, than the MW range). If storage levels are reduced and vehicle fueling requirements remain the same, the resulting energy gaps that require additional energy from the grid to produce hydrogen do not have a linear relationship between the reductions in storage. A larger installed electrolyzer capacity (and thus larger number of vehicles supported) results in storage levels being more sensitive to changes in curtailed renewable energy and thus can lead to larger requirements for additional energy. For a 25% reduction in storage capacity, the highest amount of additional energy from the grid is 7.9% (Base 1). For a 50% reduction in storage, the highest amount of additional energy from the grid is 18.4% (Base 4). This amount 31 of energy would likely come from thermal energy generators such as petroleum fired power plants. If additional capacity was installed for this purpose only, it would sit idle 3/4 2/3's of the year. Calculated via simply replacing fossil fuel powered vehicles on the road 32, between approximately 22.5 million 33 and million tons of CO 2 could be reduced per year by the use of FCVs. Gasoline reduction for the analyzed cases ranges from approximately 2.5 million gallons (Base 4) 35 to approximately 12.1 million gallons (Base 2) 36. 4) Analysis of the resulting costs of these hydrogen production, storage and dispensing facilities showed that: and GWh respectively mpg, as supplied by HNEI MW electrolyzer capacity and 244,000 kg's of hydrogen storage Base MW installed capacity and approximately 100,000 kg's of storage Base MW installed electrolyzer capacity MW installed electrolyzer capacity 30 P age

32 Due to the relatively large amount of electrolysis capacity required to capture the quantities of curtailed energy analyzed in this report, the systems would need to be modular. Consequently, centralized and distributed systems are not expected to be largely different in cost. This was confirmed via comparing DOE models of each which resulted in less than 1 % difference. It is noted that this does not take in to consideration Oahu specific land, electric supply infrastructure and transportation cost differences, but does consider capital, deprecation, equipment replacement, energy, utilities, engineering design, equity financing, taxes, land, permitting and fees, O & M, decommissioning, profit etc. and also takes into consideration expected planned and unplanned downtime. For the maximum number of vehicles that can be supported for each base case and electrolyzer capacity, storage capital costs range from lowest $30.5 Million (Base 2, 50 MW) to highest $261.7 Million (Base 4, 200 MW). Electrolyzer costs are expected to be $1 Million/MW by Total capital costs for electrolyzer and storage range from lowest $80.5 Million (Base 2, 50 MW) to highest $461.7 Million (Base 4, 200 MW) Electricity Costs per year range from lowest $17.3Million (Base 1, 50MW) to $46M (Base 2, 200MW) Including capital (amortized over 40 years), electricity and O & M, the lowest costs per year is $21.7 Million (Base 1, 50MW) and $59.3 Million (Base 2, 200MW). In reference to the maximum number of light duty vehicles supported for each base case and thus costs per mile driven, Base case 2 is consistently the most favorable, ranging from $0.15 (200MW) to $0.23 (50MW). The highest is cost is $0.41/mile (Base 1, 50 MW). The cost analysis shows that costs per mile driven reduces with increased installed electrolyzer capacity due to a reduction in electricity rates and an increased number of vehicles supported. Overall, a large number of FCV's can be supported with the use of curtailed energy which can in turn displace a large amount of airborne pollutants (up to 131 Million tons of CO 2 ). Although Base 2 has the least usage of total energy available for the scenarios analyzed (50 200MW, 25.8% and 76.3% respectively), it is by far the best configuration to support the greatest amount of vehicles with the least amount of storage, thus the most cost efficient. Although an estimated capital investment of between ~$80.5 and ~$315 Million would be required for this best configuration, the price per mile over a 40 year plant life (including 31 P age

33 electricity and estimated O & M costs) ranges from $0.23 to $0.15, respectively, which is in the range of an average conventional fossil fuel powered vehicle ($0.19/mile). 32 P age

34 5.0 Appendix A Hydrogen production from electrolysis is an electrochemical process in which an electric current is passed through water using electrodes. The electric energy splits the chemical bonds in the water molecule producing pure oxygen at the anode and hydrogen at the cathode. H 2 O ½ O 2 + H 2 The process can be scheduled such that it is synchronized with the excess renewable energy that would have normally been curtailed. Hydrogen is also easily stored and dispensed and thus can be made available for future use. The two commercial methods of electrolysis are Proton Exchange Membrane (PEM) and Alkaline. Regardless of the method, the overall electrochemical reaction is the same. PEM Electrolysis PEM technology is mature in small capacities but several companies are presently scaling up their units in order to produce larger scale output electrolyzer units (up to 2 MW). In a PEM electrolyzer, the electrolyte is a solid ion conducting membrane, as opposed to the aqueous solution in the alkaline electrolyzers. The membrane allows the H+ ion to transfer from the anode side of the membrane to the cathode side, where it forms hydrogen. The membrane also serves to separate the hydrogen and oxygen gasses, as oxygen is produced at the anode on one side of the membrane and hydrogen is produced at the cathode on the opposite side of the membrane. PEM Hydrogen Production at the Cathode 2 H + + 2e H 2 PEM Oxygen Production at the Anode H 2 O ½ O H + + 2e Alkaline Electrolysis Alkaline electrolysis is the most mature electrolysis technology available. Electrolyzers, which can be either unipolar or bipolar, can be obtained in large sizes (up to 3.5 MW). Unipolar alkaline electrolysis involves submerging electrodes connected in parallel in an aqueous electrolytic solution, such as potassium hydroxide (KOH). The electrolyte is used because of its high conductivity which improves hydrogen and oxygen production. A membrane is placed between the cathode and anode electrodes that serves to separate the hydrogen and oxygen molecules during electrolysis, but allows the transfer of ions. The 33 P age

35 bipolar design resembles a filter press where electrode cells are connected in series and separated by a membrane. Hydrogen is produced on one side of the cell, oxygen on the other. The potassium is not used in the reaction. Alkaline Hydrogen Production at the Cathode 4 H 2 O + 4e 2H 2 + 4OH Alkaline Oxygen Production at the Anode 4OH O 2 + 2H 2 O + 4e Electrolysis Method Comparison Alkaline electrolysis is by far the most mature technology having been in commercial production since the 1920's. Although mature, proven and readily available in the marketplace, the technology has some drawbacks. These include requiring corrosive potassium hydroxide to operate, along with operating at a reduced efficiency to PEM electrolyzers (up to 16% 37 ). Large alkaline electrolyzers have not typically been used in an environment where rapid hydrogen production rate changes are required and thus, have previously been designed with a slower ramp rate than PEM's, but research and development in this area is ongoing. The world's leading manufacturer of large electrolyzers, NEL Hydrogen (Norway), now advertises a 3 second ramp rate which, for the purpose of this study, would be sufficient. PEM electrolyzers are a far less mature technology than alkaline (at large scales, i.e. MW sizes) but can be load following to the millisecond and a 50 millisecond ramp rate has been demonstrated at the United States National Renewable Energy Laboratory (NREL) 38. A PEM electrolyzer producing the same amount of hydrogen as an alkaline electrolyzer will typically have a smaller footprint, which can have practical implications when selecting a certain technology. 37 K. Harrison (NREL), Hydrogen Works Conference, San Diego, CA, Feb 17 19, PEM R & D Electrolysis Webinar (NREL), Dr. Katherine Ayers 34 P age

36 6.0 Appendix B Portion of time that certain magnitudes of hourly curtailed power occur Amount of energy available at various power levels for curtailed renewables 35 P age

37 7.0 Appendix C Maximum amount of curtailed energy captured as a percentage of total available (lines) and subsequent amount of hydrogen produced for each Base case and various electrolyzer capacities (bars) 8.0 Appendix D The Amount of light duty vehicle miles (lines) and corresponding minimum amount of storage required (bars) assuming consistent and evenly distributed hourly consumption of hydrogen for each Base Case with varying electrolyzer capacities 36 P age

38 The Amount of heavy duty vehicle miles (lines) and corresponding minimum amount of storage required (bars) assuming consistent and evenly distributed hourly consumption of hydrogen for each Base Case with varying electrolyzer capacities 9.0 Appendix E Case Storage kg/vehicle 50MW 100MW 150MW 200MW Average Base Base Base Base P age

39 10.0 Appendix F Storage reduction of 25% analysis that gives the amount of excess energy from thermal generation required (bars) to maintain the same amount of vehicle miles as the strategy that uses only curtailed renewables. Lines show the percentage of energy used to produce hydrogen that is from additional sources. Storage reduction of 50% analysis that gives the amount of excess energy required (bars) to maintain the same amount of vehicle miles as the strategy that uses only curtailed renewables. Lines show the percentage of energy used to produce hydrogen that is from additional sources. 38 P age

40 39 P age

41 ANALYSIS OF SMART GRID TECHNOLOGIES Prepared for the: State of Hawai i Department of Business, Economic Development and Tourism Under Supplemental Contract No. 1 for Contract #60514 for the NETL Grant Project Task 1, Subtask 5, Smart Grid Technologies Element Prepared by: Black & Veatch Submitted by the: Hawai i Natural Energy Institute July 2013

42 ANALYSIS OF SMART GRID TECHNOLOGIES TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII B&V PROJECT NO Black & Veatch Holding Company All rights reserved. PREPARED FOR Hawaii Natural Energy Institute 25 JUNE 2013

43 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Table of Contents Legal Notice Executive Summary Scope of Work Report Organization Introduction and Purpose Grid Flexibility Energy Efficiency Demand Response The Smart Grid Enables Technologies That Provide Flexibility Barriers to Increased Penetration of Renewable Generation in Hawaii Solutions to Barriers to Increased Penetration of Renewables in Hawaii Integration of Enabling Technologies Proposed Implementation Roadmap Summary of Overall Recommendations or Proposed Next Steps Conclusion Introduction and Purpose Background Purpose Existing Conditions Grid Flexibility Traditional Electric System Flexibility/Reliability Functions Traditional Sources of Flexibility Flexible Conventional Generation Demand Response Curtailment Sub hourly Scheduling Consolidation of Balancing Areas Enabling Transmission Energy Storage New EMERGING FLEXIBLE RESOURCES A Proactive Approach in Hawaii Section Summary Energy Efficiency (EE) The Value and advantages of energy Efficiency existing Energy Efficiency programs in Hawaii Transmission And Distribution efficiency, voltage conservation and voltage/var optimization BLACK & VEATCH Table of Contents i

44 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 4.4 Smart Metering increases customer driven efficiency Green Button Initiatives Present New OPportunities to Increase Customer Driven efficiency Energy Efficiency creates the need for additional system flexibility Opportunities to Enhance Hawaii s Energy Efficiency Programs, as Identified to DaTE In State Planning Documents Opportunities to Enhance Hawaii s Energy Efficiency Programs, as Identified by third parties Appliance efficiency standards State government initiatives The Interaction between Energy Efficiency & Demand Response Section Summary Demand Response What is Demand Response, What are the types of Demand response Direct Load Control Third Party Aggregation Load Shifting and Scheduled Load Management Rate Design Programs Load Shaping Demand Response For Flexibility Hawaii s Existing Demand Response Programs Opportunities to Enhance Hawaii s Demand Response Programs Measurement and Verification of Demand Response Evaluating the cost effectiveness of DR Programs Customer interactions, communication techniques and advancements in load control Section Summary The Smart Grid Monitoring, Communications and Control Technologies how does the Smart Grid enable Hawaii s Existing Smart Grid, Communications and COntrol Programs Opportunities to Enhance Hawaii s Smart Grid Programs Smart Metering as an enabler Control systems as an enabler Communication systems as an enabler Enabling DR Enabling Energy Storage Enabling Generation and Storage and Reserves Management Enabling Forecasting and Curtailment How microgrids help Section summary Barriers to Adoption of Smart Grid Technologies in Hawaii BLACK & VEATCH Table of Contents ii

45 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 7.1 Barriers Lack of technical and commercial readiness Lack of Utility Preparedness Cost Cost Recovery Lack of customer awareness Customer s lack of access to their own energy information Policy, Regulatory, Statutory barriers Permitting requirements/procedures Section Summary Solutions to Barriers to Adoption of Smart Grid Technologies Increased Penetration of Renewables in Hawaii Incentives, tax credits and rebates that change perceived cost Modification of cost recovery mechanisms Mechanisms that improve technical and commercial readiness Education to improve customer awareness Rule Changes Section Summary Integration of Enabling Technologies How many flexible Resources do we need? How do we select them and will they be ready in time? A straw man order of GRID flexibility Options Proposed Implementation Roadmap What are the steps to make these flexibilities available Solution Costs Section summary Summary of Overall Recommendations or Proposed Next Steps Conclusion Appendix A. Glossary of Abbreviations and Terms... A 1 Appendix B. Power Generation Resource Mixes... B 1 LIST OF TABLES Table 2 1 Current Levels of Renewable Generation on Each Island Table 3 1 Flexibility/Reliability Functions Table Hawaiian Energy Efficiency Program Results Table 4 2 Strategy Recommendations to Meet Energy Efficiency Goals Table 4 3 ACEEE 2012 Energy Efficiency Scorecard for the State of Hawaii Table 10 1 Smart Grid Solution Reference Cost Estimates BLACK & VEATCH Table of Contents iii

46 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII LIST OF FIGURES Figure 1 1 High Level View of the Smart Grid Figure 2 1 Relative Proportion of Conventional and Renewable Generation on Hawaiian Islands Figure 4 1 MDMS Implementation Architecture Figure 4 2 Customer Energy Use Data Presentment Figure 6 1 Estimated Advanced Metering Penetration Nationwide Figure 6 2 DMS Example Integration Scenario Figure 6 3 Potential Applications of Energy Storage Figure 10 1 Strategic Plan Path Forward Figure 10 2 Recommended Smart Grid Solution and Enabling Mechanisms Sequencing BLACK & VEATCH Table of Contents iv

47 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Legal Notice Acknowledgement This material is based upon work supported by the Department of Energy under Award Number DE FG26 08NT This report was prepared for and compiled by the University of Hawaii, Hawaii Natural Energy Institute on behalf of the State of Hawaii Department of Business Economic Development and Tourism (DBEDT). Disclaimer: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. BLACK & VEATCH Legal Notice LN 1

48 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 1.0 Executive Summary Utilities are undergoing significant change to adapt to a landscape that is very different from the one that existed at the dawn of the existing electric power industry. Utilities are striving to accommodate new customer expectations, adapt to changing regulatory requirements, respond to weather and nature events, and invest in increasingly aged infrastructure. Against this background, utilities are also experiencing profound disruption in the generation mix. Throughout North America, natural gas has become a fuel of choice and worldwide, utilities are striving to integrate higher penetrations of renewable energy resources onto the grid. In Hawaii, a goal is to accommodate more renewable energy through various enabling technologies. All utilities are attempting to make this transition in part by implementing new solutions based on advances in technological innovation and the ability to integrate these new solutions together into a smarter grid. Utilities have embraced the ability of the Smart Grid to enable a fundamental migration from a predominantly analogue based industry based on time proven but limited control systems into a two way information rich, digital, highly integrated electric grid. While the term Smart Grid supports many different detailed definitions, each with a particular focus, the term has come to encompass a collection of enabling technologies and advanced applications that provide utilities with new tools and capabilities to operate the evolving electric grid more reliably. A commonly recognized view of the Smart Grid is provided in Figure 1 1. Figure 1 1 High Level View of the Smart Grid 1 1

49 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Smart Grid technologies include solutions that range from advanced control and operating solutions at generation facilities to smart meters and engagement solutions at customers homes. Utilities are implementing a wide range of solutions on the transmission and distribution grid itself, including line sensors and monitors, substation automation, improved distribution management systems, distribution automation and control, and information and data analytic solutions. Each of these solutions has value that can be derived to operate the grid more efficiently, and with more flexibility to provide increased reliability. The importance of the term Smart Grid is the notion of combining these individual enabling technologies or advanced applications together into an integrated set of solutions that in aggregate will transform the grid. Figure 1 1 also illustrates that the Smart Grid, in addition to being a collection of technologies and advanced applications, is also a layered effort to introduce more intelligent control and management of various aspects of the electricity grid. An important concept in understanding how utilities are implementing Smart Grid solutions is that generally, there is a sequenced approach to implementing the various solutions and Figure 1 1 illustrates that sequence. From the perspective of generation resources, the Smart Grid technologies and applications create significant opportunities to enable increased operating flexibility that provides more reliability. Smart Grid allows utility operating practices related to reserves, generation plant management, curtailment, and optimum run decisions to be modified. The communications systems and customer engagement opportunities afforded by smart metering or demand response (DR) provide the opportunity to manage demand in ways that support the variable output of many renewable generation resources. The concept of flexibility and solutions that increase the flexibility of the grid relative to traditional generation capabilities are key components of the value that Smart Grid solutions enable and form the core of the discussion presented in this section of the HREGP report. Flexibility can be described as the capacity of a grid to rapidly modify generation or demand in response to evolving uncertain conditions SCOPE OF WORK This white paper explores the challenges of increasing the amount of renewable generation on the Hawaiian Island s grid to 1 GW and the role that Smart Grid technologies and solutions can play in enabling this fundamental shift in generation resource mix. The report discusses the current barriers to higher penetration of renewable energy and how both traditional and non traditional utility operations and policy are relevant. It further explores a selection of Smart Grid solutions that 1 Beauvais, D.; et. al., "SMART GRID to Balance Renewable Energies Contributing Distributed Energy Resources Varennes Reseach Center, March

50 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII are relevant to meeting the technical, operational, business and policy requirements of transitioning to a more renewable rich resource mix. 1.2 REPORT ORGANIZATION Introduction and Purpose The goal of this report is to provide recommendations for the strategic deployment of the Smart Grid to enable an integrated portfolio of technologies and mechanisms that facilitate higher levels of energy efficiency and renewable energy deployment by providing increased electric system flexibility. While energy efficiency can be largely self standing, flexible resources which provide necessary system flexibility/reliability functions are frequently enabled by the advanced monitoring, controls and information features of the Smart Grid Grid Flexibility Intermittent renewable generation increases the need to supply electric system flexibility/reliability functions with existing traditional and new emerging flexible resources. When flexibility/reliability functions are sold in energy markets they are called ancillary services. Hawaii s Reliability Standards Working Group (RSWG) recently worked with General Electric (GE) to define a set of Hawaii specific ancillary services that could be used to protect reliability, incent renewable generation, and minimize production cost for each Hawaii island utility. 2 Existing traditional and new emerging flexible resources will be enabled and enhanced with the advanced communications and control capability of the Smart Grid. Examples of emerging flexible resources that will be enabled by Smart Grid communications and control capability include fixed and mobile energy storage, new types of demand response, improved forecasting of renewable energy resources, and new methods of reserves management Energy Efficiency Energy efficiency (EE) is a critical component of Hawaii s Clean Energy Initiative, and it is an important part of each of the island utilities Integrated Resource Plans (IRPs). EE is typically considered the most cost effective means to meet future energy demand. EE reduces the need for new generation, thus reducing overall system cost. Many States have recognized this and have provided EE performance standards and other incentives to induce utilities and customers to adopt it on a priority basis. Independent reports by both the Electric Power Research Institute and 2 HNEI, "Report on Business Case in Hawaii for Storage Options," US Department of Energy, Office of Electricity Delivery and Energy Reliability, Honolulu, HI, October

51 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII McKinsey and Company in indicated that energy efficiency presents a vast low cost energy source nationwide for the U.S. By increasing the ratio of renewable supply to grid load, EE reinforces the need for Smart Grid technologies to keep supply and demand in balance. EE also overlaps in some cases (capturing the same opportunities) with DR, and the result is a lean grid requiring increased flexible resources to be reliable. Specific opportunities to strengthen energy efficiency policies and programs in Hawaii include minimizing free ridership, promoting new energy efficiency technologies, enhancing customer experience with smart meters, smart appliances, and green button tools, strengthening appliance efficiency standards, strengthening state government funded R&D initiatives, increasing stringency of building codes, adopting reward structures for utilities exceeding energy efficiency goals, and enhancing cost benefit accounting to include new programs and technologies. Black & Veatch recommends that the annual statewide measurement and verification reports, which currently address a number of energy efficiency programs, should be expanded to include cost benefit accounting for demand response, and for communication, monitoring and controls programs. This would enable capturing of additional savings, and ensure that co benefits of EE measures do not go unrecognized Demand Response Demand Response (DR) solutions and technologies come in many different variants depending on the load or appliances to be controlled, the control employed and the initiating agent or methodology. DR can and should play an important role in the integration of the renewable energy resources. DR is becoming an increasingly valuable tool due to technology innovations and policy directives. It is both technically feasible and economically reasonable for consumers to respond to signals from a utility system operator, load serving entity, Regional Transmission Operator (RTO)/Independent System Operator (ISO), or other DR provider for DR to provide reliability services to the bulk system. The key differences between DR for traditional peak shaving and DR for supporting renewable penetration are the amount of advance notice provided and the required response time. For peak shaving, which is typically based on day ahead or hours ahead forecasts, consumers can plan their response well in advance. Alternatively, the decision to call DR events for renewable generation variability occurs with limited advance notice and therefore requires faster reaction times by consumers in order to maintain grid stability. 3 y_efficiency_in_the_us_economy df 1 4

52 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII DR programs vary in design and focus, some initiatives such as rate programs, which encourage consumers to shift load to off peak periods when wind production is usually at its highest, could be used to avoid curtailment of variable generation or to minimize cycling of large base load power plants. However, DR is not one size fits all and any demand response program should include a portfolio of options designed to deliver necessary results, match customer comfort zones and maximize adoption rates. Utility requirements, business cases, and program options vary greatly. Customer demographics, level of demand and participation of the commercial and industrial sector, and environmental conditions all impact how programs are designed to meet specific market dynamics. The program intent to deliver results for cost, reliability, and renewable generation offset reasons will impact program design, the underlying requirements and technologies used. DR therefore needs to be looked at as a set of solutions that are customized to meet particular utility needs. The availability of complementary or enabling Smart Grid technologies provides the opportunity to integrate the DR into broader holistic and sustainable grid reliability solutions. In addition, regulatory policies must accommodate the variety of viable DR options and support the goals of the DR programs The Smart Grid Enables Technologies That Provide Flexibility The Smart Grid enables flexible technologies necessary for integration of intermittent renewable energy resources. Utilities require both fast acting power system solutions that respond to realtime situations and slower acting centralized analytical solutions that include forecasting and longer term resource management. The level of challenges will depend on the amount, size, location, and voltage levels of the interconnected resources; the capabilities of those resources for dispatch and communication (and similar Smart Grid capabilities); and the ability of the grid to accept generation at multiple distributed sites and to flow power as needed in the opposite direction to that initially designed. In preparing themselves for the increased penetration of renewable Distributed Generation (DG), utilities realize the need to investigate and apply smart grid solutions such as smart metering, smart communications solutions, distributed monitoring and control, and Distribution Management System (DMS) applications. These Smart Grid solutions have far reaching value beyond supporting a higher penetration of DG resources. The various enabling technologies are designed to meet a range of utility operational efficiencies, reliability improvements, and customer service enhancements. Solutions such as DMS, for example, are designed to provide a higher level of awareness of grid status and provide methods to control and respond in more real time fashion. This provides value in identifying and responding to outage conditions, providing more efficient asset optimization, gathering substantially more data for downstream analytic tools, and other new operational processes. Additionally, the value of the disparate smart grid solutions such as Advanced Metering Infrastructure (AMI) or Smart Metering, depending on terminology used, Distribution Management System (DMS), Outage Management system (OMS) and System Control and Data Acquisition (SCADA) are increased through synergies by integrating the solutions together. By proactively embracing these changes, utilities can begin to 1 5

53 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII shape the myriad of planning and operating approaches that maximize the potential long term benefits to the utility and its customers Barriers to Increased Penetration of Renewable Generation in Hawaii A significant barrier to increasingly high penetrations of variable renewables is the lack of information about the relative merits (costs and benefits) of flexible resources. Similarly, the lack of information about the interaction between customer based DR programs and renewable energy is a barrier. The solution to this information gap is the evaluation of actual field demonstrations that prove the solutions, their cost and benefits and the implementation methodologies and customer engagement options. As more demonstrations are conducted that leverage DR technologies and smart communications and control technologies, there will be an opportunity for the industry to develop comprehensive studies that identify cost benefit tradeoffs so utilities can identify and rank alternatives which allow higher penetrations of intermittent renewables Solutions to Barriers to Increased Penetration of Renewables in Hawaii In addition to the broader industry demonstration needs, Smart Grid technology demonstrations on the Oahu grid will be required to validate their effectiveness in reducing renewable curtailment or in reducing peak demand loads. The costs and benefits identified from these technology demonstrations will need to be analyzed and compared. In some cases, new cost recovery mechanisms will need to be approved by regulators or existing mechanisms adjusted to achieve implementation and in some cases incentives will need to be provided to achieve commercialization. Also, process rule changes may be required and it may be necessary to provide education to inform customers, utilities and independent power producers. Key technology demonstrations will be related to smart inverters, distribution system voltage controls, provision of automatic generation control (AGC) to enable regulation, frequency response, reserves management, and overall system integration. Information about the costs of individual technologies and mechanisms are required to develop the priorities and an implementation plan. Once the information is available to establish the plan, the emphasis will switch to understanding and modifying cost recovery mechanisms, establishing appropriate incentives and identifying appropriate rule changes and education mechanisms Integration of Enabling Technologies While the disparate Smart Grid solutions can be evaluated in isolation; a core concept of the Smart Grid is the integration of the various enabling technologies and advanced applications. Each individual solutions delivers defined benefits to utilities in supporting a safer, more reliable, more flexible electric grid with a higher level of awareness and control. The synergies of integrating multiple technologies, especially enabling technologies and advanced applications, provide the 1 6

54 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII ability of the solutions to support the needs and requirements of one another and provide synergies and capabilities that cannot be otherwise derived. This significantly increases the business benefit to utilities, customers and other stakeholders Proposed Implementation Roadmap The proposed implementation roadmap included in this white paper follows a representative sequence of steps to address the barriers to achieving the goals of the HCEI and attain 1 GW of renewable generation in Hawaii: Demonstrate technologies capabilities and effectiveness, Develop inter utility coordination, Evaluate the costs and benefits, Determine cost allocation and recovery mechanism, Obtain regulatory approval, Establish incentives (customer, utility and other stakeholders), Identify and implement rule changes, and Provide necessary education mechanisms. This sequence applies to all technologies and mechanisms used to achieve the electric system flexibility that is discussed in the report Summary of Overall Recommendations or Proposed Next Steps The recommendations provided in this white paper are designed to leverage Smart Grid solutions that have value to the goal of enabling 1 GW of renewable generation onto the Hawaiian Islands grids. This does not represent the entire collection of available Smart Grid solutions. It is possible that advances in certain areas may deliver unanticipated benefits that would allow additional Smart Grid technologies and advanced applications to contribute to the additional electric system flexibility required to support the high quantities of renewable energy. However, this report focuses on those solutions that have demonstrated, or are reasonably anticipated, to contribute the necessary flexibility. The report includes the identification of the Smart Grid technologies and advanced applications as well as a recommendation as to the sequence of how they should be implemented to provide the maximum impact and cost effectiveness. Given that these solutions would both contribute to and require significant changes by HECO utilities, specifically, the time required to implement these changes;the impact on the organization and its employees;and the potential cost. A detailed strategy and business assessment would need to be performed by HECO in order to accurately define these impacts and the actual quantitative and qualitative benefits that would be attained. Additionally, the recommendations do not take into account planned or potential initiatives that 1 7

55 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII HECO may already have planned that might provide additional synergies or conflict with the recommendations Conclusion The challenges and barriers to achieving a level of penetration of renewable energy resources in Hawaii in excess of 1 GW requires that significant improvements be made to the electric grids in terms of reliability and flexibility. While improvements in generator operation and control can contribute some of this improvement, it is insufficient to meet the complete needs of the future grid. Regulatory policy changes and increased participation of HECO s customers as well as the renewable generation community can also contribute. But the greatest opportunity for the transformation of the electric grid to accept and even embrace the significant amounts of renewable generation desired are Smart Grid solutions that are being implemented or developed today. The maturity of these solutions varies and the impact on grid flexibility and renewable generation support may not yet be completely demonstrated, but they hold the most significant opportunities for the creation of a new, smarter, more enabled electric grid. 1 8

56 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 2.0 Introduction and Purpose The purpose of this report is to provide a strategic roadmap for the deployment of smart grid technologies to enable the achievement of twin goals related to renewable generation on the Hawaiian Islands electric grids: 1. Meet Hawaii s clean energy goals of achieving 70 percent clean energy by 2030, 40 percent to be provided by renewable energy and 30 percent from energy efficiency, relative to 2008 levels. 2. Support the utilization of 1 GW of aggregate renewable energy generation capacity. This report is a work product of Black & Veatch under contract with the Hawaii Natural Energy Institute, as part of a larger effort in support of the Hawaii Clean Energy Initiative. At the time of this writing, the State of Hawaii has achieved higher levels of renewable energy penetration in some parts of its electrical grid than any other state in the U.S. This means that the State of Hawaii is also by necessity leading the nation in developing solutions to address the challenges associated with high levels of renewable energy penetration in the electric power grid. This report provides recommendations for the strategic deployment of Smart Grid solutions to enable an integrated portfolio of technologies and mechanisms that facilitate higher levels of energy efficiency and renewable energy deployment. While energy efficiency can be self standing, additional flexibility in the electric gird which provides necessary system reliability functions can be enabled by smart grid features such as advanced monitoring, controls and information management. A portfolio of technologies is required because each of these technologies has a different set of functionalities, delivered benefits and associated attributes and the various combinations provide synergistic capabilities at varying costs and capabilities. These attributes must work together to provide electric system reliability functions in an efficient, robust, integrated solution, customized to the unique island operating conditions of the Hawaii environment. The three greatest barriers to achieving high levels of renewable energy deployment are as follows: 1) the typically high capital costs of renewable generation equipment and associated technologies necessary to achieve integration, 2) the intermittency of the renewable resources (predominantly sun, wind, and rainwater used for solar, wind and hydro generation respectively), and 3) the grid reliability impacts of this intermittent power generation. The following portfolio of smart grid technologies can address these barriers: Energy efficiency technologies provide low cost energy services. 2 1

57 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Demand response technologies increase the system s ability to utilize renewable power at the time when it is available, or to schedule the delivery of generated electricity for the times when it is most needed. Monitoring capabilities that provide real time status and alarming to provide indication of the grid situation and the availability of distributed generation resources. Control technologies provide means for utilities to react in real time to the changing situation or disturbance Information technologies provide analytic and forward looking tools to enable effective grid management. These technology solutions can enable the electrical grid s flexibility/reliability functions, under the challenging frequency and voltage conditions that are potentially created by intermittent renewable energy generators. Taken as a whole, this portfolio of technologies can dramatically improve the quality of service experienced by the electric power customer, while also achieving Hawaii s clean energy goals in the most cost effective way. 2.1 BACKGROUND The State of Hawaii has adopted aggressive goals to encourage a shift from the existing reliance on petroleum based energy supply toward an increased focus on renewable energy and energy efficiency. The Hawaii Clean Energy Initiative (HCEI) has a target of 70 percent clean energy by percent to be provided by renewable energy electricity and 30 percent from energy efficiency, relative to 2008 levels. This translates to a renewable energy supply goal of approximately 80 percent of actual electric generation. That level of renewable energy penetration far exceeds current levels. In 2011, 12 percent of Hawaiian Electric Companies electricity sales came from renewable sources 4. This is comparatively high relative to many other US mainland electric utilities but low relative to Hawaiian goals. Achieving the state goals will require new technologies, modified utility operational processes and supportive public policy to enable a viable and efficient transition to the targeted clean energy penetration levels. The aggressive clean energy goals coupled with the unique characteristics of an island based grid means that Hawaii will continue to be at the forefront of efforts to identify, analyze and adopt policies to encourage the optimal use of enabling technologies. Hawaii does not have the luxury of participating in an integrated multi utility grid with its inherent ability to provide mutual support and a robust ancillary services market that comes with these larger grids. Enabling smart grid 4 af2b154da9010vgnvcm bacrcrd&vgnextfmt=default 2 2

58 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII technologies will include: energy efficient technologies, demand response programs, monitoring and control technologies, energy storage and information technologies. 2.2 PURPOSE The purpose of this section of the HREGP is to identify opportunities in expanding energy efficiency solutions and address opportunities and issues surrounding smart grid technologies that will support the integration of 1 GW of renewable energy onto the Hawaiian Island electric grids. Smart grid technologies enable acceptance of renewable energy technology, optimize performance of energy efficient technologies and contribute to the reliability and efficiency of the grid. Specifically they enable increased customer engagement and support various forms of demand response. They also enable and optimize grid control and management, support energy storage and facilitate management of reserves. 2.3 EXISTING CONDITIONS Hawaiian Electric Company and its subsidiaries serve 95 percent of Hawaii s 1.2 million residents. Hawaiian Electric Company (HECO) itself serves Oahu; a subsidiary, Hawaii Electric Light Company (HELCO), serves the Island of Hawaii; and another subsidiary Maui Electric Company (MECO), serves Maui, Lanai and Molokai. The island of Kauai is served by the Kauai Island Utility Cooperative (KIUC) and the islands of Niihau and Kahoolawe are not served by utilities. This white paper focuses on examples and opportunities related to the HECO utility companies, since they are so prominent and are essential to meeting the Hawaii goals, but the information and recommendations presented would be substantially applicable to most island electric grids. The table below shows the defining characteristics of each island and current levels of renewable energy capacity installed. 2 3

59 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Table 2 1 Current Levels of Renewable Generation on Each Island HAWAII ISLAND MAUI OAHU KAUAI 5 Electric Utility HELCO MECO HECO KIUC Area (sq mi) 4, Population 185, , ,207 66,921 Population Density (pop./sq mi) , Firm Generating Capacity , Central Solar (Installed MW)_ Distributed Solar (Installed MW) Wind (Installed MW) Geothermal (Installed MW) Biomass (Installed MW) Waste to Energy (Installed MW) Hydropower (Installed MW) Total Existing Renewable Generating Capacity (MW) Total Potential Renewable Generating Capacity 9 (MW) Percentage of Electricity Sales from Renewable Capacity (2012) 10 (%) 1, , Existing Smart Grid Projects Automated Maui Smart Grid HECO East Federal grant generation Project 12 Oahu to install control program DOE Smart PV Switching advanced to respond to Inverter Project 15 metering changes in Project 13 Grid infrastructure frequency in the Grid Modernizatio information Personal communication with HNEI dated 24 June Additional data from PowerPoint presentation titled Hawaiian Electric Companies (HECO, HELCO, MECO) IRP 2013: Existing System and Conditions. Lisa Gang, IRP Advisory Group Meeting # content/uploads/2011/09/hawaii Clean Energy Initiative Scenario Analysis Summary_March2012.pdf 10 the renewable future 2 4

60 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII grid. Smart Meter Education Program 11 Modernization Projects 14 Maui County Load Management Assessment n Projects 16 It can be seen from the table above that the renewable energy and grid stability issue is different for each island. The islands have a wide range of load demand, potential renewable energy and currently installed renewable energy resources. Oahu has a stronger grid and lower renewable energy penetration but the most load and highest population density. Therefore there is significant potential for Oahu to expand renewable penetration while implementing smart grid technology in the process of working toward the state s renewable energy goals. In contrast, the Big Island has a much lower maximum demand and a low population density, and already has approximately 40 percent renewable energy installed, and cannot accept more intermittent renewable energy without significant grid upgrades and changes in operational methodologies and capabilities. Smart Grid deployments will need to be targeted and strategic on the Big Island, focusing on grid reliability. An important conclusion from Table 2 1 is that the conditions and issues are different for each island and thus the smart grid technologies, prioritization and pace of implementation and grid flexibility and reliability challenges differ. For additional detail regarding each island s power generation resource mix, see the table provided in Appendix B. Hawaiian Electric Co. reported customers on Oahu, Maui and the Big Island installed 92.8 megawatts of photovoltaic generating capacity in 2012, which was more capacity installation than in the previous six years combined smart grid project description/ 13 essf.pdf technologies/smart grid/ 11 essf.pdf 14 Ibid. 16 Ibid

61 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 2 6

62 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Figure 2 1 Relative Proportion of Conventional and Renewable Generation on Hawaiian Islands For many years the state of Hawaii has planned and investigated an inter island undersea transmission cable that will interconnect the electric systems on Oahu and neighbor island(s). This white paper assumes an end state that includes this transmission cable, but the content herein does not include discussion of the transmission interconnection. In the context of existing conditions, it can be seen that each island has different current state and generation needs, and each island has different constraints, within which next steps toward the state s renewable energy goals will take place. The grid on each island will require different levels of flexibility, in order to adapt the existing resource mix to meet expected future demand. It is this paper s hypothesis that smart grid technologies can provide the needed flexibility. 2 7

63 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 3.0 Grid Flexibility Wind and solar energy generation are variable and will impact the electric system s ability to maintain voltage and frequency. The higher the percentage of variable renewable energy and efficiency that exist, the more that the electric system requires solutions that provide flexible response capabilities. The Grid Flexibilitysection is divided into three subsections. The first explains how electric utilities traditionally provide flexibility/reliability functions that allow them to respond to mismatches in supply and demand. The second identifies existing technologies and mechanisms (flexible resources) that utilities use to provide electric system flexibility/reliability functions. The third subsection identifies new flexible resources that utilities will use as they are technically and commercially available, and it also highlights that they will be enabled with advanced communications and control technologies of the future. 3.1 TRADITIONAL ELECTRIC SYSTEM FLEXIBILITY/RELIABILITY FUNCTIONS Planners and operators are very familiar with variability and uncertainty as they are constantly balancing traditional demand and generation to maintain voltage and frequency. They adjust to demand and generation fluctuations, changes in transmission power flow, unexpected generation or large load outages and changing interconnection schedules with well defined flexibility functions. System flexibility is defined by NERC as the ability of supply side and demand side resources to respond to such system changes and is accomplished with specific electric system flexibility functions 19. Table 3 1 presents the traditional electric system flexibility/reliability functions supplied by traditional generators. It identifies the flexibility/reliability functions and benefits that are required, grouping them into ten categories. 19 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November

64 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Table 3 1 Flexibility/Reliability Functions 20 RELIABILITY FUNCTION RESPONSE TIME BENEFIT 1) Inertial Response Cycles to 1 2 seconds System stability 2) Primary Frequency Response Cycles to 5 10 seconds System stability 3) Regulation (a FERC defined ancillary service) 10 seconds to several minutes Minute to minute frequency corrections 4) 4. Load Several minutes to a few Slower response resources are following/ramping hours dispatched to follow system ramping requirements 5) Dispatchable Energy Focuses on energy consumption at times of peak capacity requirements and minimum load Obtained from sub hourly and hourly energy markets 6) Contingency Spinning Reserve Must respond with 10 to 15 minutes Enough contingency spinning and non spinning reserve must exist to deal with the largest failure that is anticipated 7) Contingency Nonspinning Reserve Response not required immediately but required within 10 minutes Meets largest failure anticipated in conjunction with Spinning Reserve 8) Replacement or Supplemental Reserve 9) Variable Generation Tail Event Reserve 30 to 60 minutes Backup reserves Ramps > 90 minutes Reserve for long duration wind events 10) Voltage Support Maintain voltage control to specified criteria 3.2 TRADITIONAL SOURCES OF FLEXIBILITY Traditionally, the existing technological solutions and mechanisms (flexible resources) listed below have provided the above electric system flexibility/reliability functions. However, advanced 20 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November

65 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII communication and control capabilities of the smart grid technologies can enhance the ability of traditional technologies and mechanisms to maintain grid reliability through increased flexibility. The following are technologies and mechanisms (flexible resources) that are currently used to accommodate variability in utility systems 21. Flexible Conventional Generation Demand Response Curtailment Sub hourly Scheduling Consolidation of Balancing Areas Enabling Transmission (Transmission cannot provide flexibility by itself;however it provides access to additional sources of needed flexible resources.) Energy Storage (Pumped storage plants are an example of a bulk energy storage technology where all ten of the flexibility/reliability functions identified in Table 3 1 have been proven.) These flexibility enabling resources are described in further detail in the following subsections Flexible Conventional Generation Manufacturers are already developing conventional generation units that have higher ramp rates (fast changes in generating output) and cycling capabilities (large movements up and down in generating output), while managing the potential maintenance costs that result from this cycling. Transitioning these older generation facilities to more flexible units will provide grid operators with additional capabilities to meet rapid changes in generation needs due to renewable generation disturbances. A more flexible fleet of generators will also require transitioning the traditional base load units to ones which can operate at lower minimum levels and have an increased ability to cycle. As an example, looking at the current situation at HELCO, a large percentage of its generation is not flexible. Many generators do not currently participate in load following or frequency management. These generators (run of river hydro (14MW), Wind (33MW), geothermal (30MW)), are operated as must take generation with attendant reliability impacts. The impacts associated with this inflexible generation include a reduced ability to optimize costs and more difficulty in forecasting fuel requirements. Hence, the fuel supply barge schedule has to be changed when as available energy exceeds forecasts, making fuel use less than forecast North American Electric Reliability Corporation, "Flexibility Requirements and Metrics for Variable Generation: Implications for System Planning Studies," NERC, Princeton, NJ, August HELCO IRP Advisory group, HELCO System Operations, Hawaii Electric Light Company, Inc., August

66 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII As a result, HELCO has identified the need to maintain fast starting units to provide offline reserves and the need to consider electric system flexibility functions through technology fast start, ramping, cycling capabilities for its future power generation additions. As part of these improvements, HECO has a project underway to retrofit eight of its fossil fuel units to increase flexibility. The project will increase the utility s ability to accept more energy from variable renewable resources by reducing the combined minimum load of the units from about 330 MW to 170 MW. The retrofit will involve new equipment and equipment redesign, although specific changes are yet to be determined. The project is scheduled for completion by Demand Response Like flexible conventional generation, demand response (DR) is another conventional flexibility. DR has been shown in some Balancing areas (BAs), such as the Electric Reliability Council of Texas (ERCOT) in Texas, to be a flexible mechanism for operators to use with wind generation. ERCOT uses DR for minute to minute regulation and to supply contingency reserves. ERCOT demonstrated the ability to call on 1,200 MW of DR to restore system frequency during an incident in Texas in February Effective DR can provide flexibility over short timeframes when an unpredictable change in intermittent renewable generation output occurs. A significant challenge in using DR to respond to intermittent renewable generation is that most DR programs are designed to operate only a limited number of times per year and thus are not necessarily matched to the high degree of variability of wind. DR is discussed more fully in Section Curtailment Curtailment of intermittent renewable generation output may be necessary if the amount of energy available at a specific time is more than the grid can reliably accept. For less flexible island power systems with small balancing areas dominated by thermal generation, wind curtailments could occur even at relatively low variable generation penetrations. Recent wind integration studies and operating experience demonstrate that at higher levels of penetration, wind generation may need to be curtailed during certain periods, unless suitable flexibility is designed into the bulk power system or measures are developed to store off peak wind power for use during higher demand periods. Wind generation can be designed to limit ramps and can be curtailed to provide reserves, i.e. a source of increased flexibility. Wind procurement contracts may need to be modified to accommodate curtailment or to require features that facilitate curtailment and utility control. 23 Western Governors' Association, "Meeting Renewable Energy Targets in the West at Least Cost: The Integration Challenge," Western Governors' Association, June 10, North American Electric Reliability Corporation, "Flexibility Requirements and Metrics for Variable Generation: Implications for System Planning Studies," NERC, Princeton, NJ, August S. Fink, et al, "Wind Energy Curtailment Studies, May 2008 May 2009,," NREL, Golden, CO., October North American Electric Reliability Corporation, "Flexibility Requirements and Metrics for Variable Generation: Implications for System Planning Studies," NERC, Princeton, NJ, August Kevin Porter, et al, PJM Renewable Integration Study, Task Report: Review of Industry Practice and Experience in the Integration of Wind and Solar Generation, Exeter Associates, Inc. and GE Energy, November

67 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Sub hourly Scheduling In many BAs, generation is scheduled on an hourly basis. Changes in load or generation occurring within the hour, must be met by generating units providing regulation and load following services. Scheduling generation on shorter time intervals can reduce the need for units to provide costly regulation services, freeing them up to support system flexibility requirements. If there are sufficient competitive resources in the market, sub hourly energy markets can reduce costs by providing flexibility more cost effectively. These markets can provide economic incentives that reduce dispatch that is out of economic merit order. Sub hourly scheduling also reduces the period of uncertainty around wind generation schedules and allows wind plant owners to adjust schedules more frequently Consolidation of Balancing Areas While each BA must continuously balance load and generation within its area, studies show that balancing area consolidation reduces ramping requirements for load, wind, and load with wind. If there is transmission capacity, increasing the size of a BA or collectively sharing the balancing obligation among BA s via dynamic transfers, formal energy imbalance markets or other means can provide flexibility to integrate variable energy resources (VER s). First, larger BAs provide access to more available generating resources and other sources of flexibility. Second, larger BAs can take advantage of the geographic diversity of wind resources. Results indicate the cost differential for operating independent balancing areas versus one consolidated BA increases significantly with significant wind penetration. Currently, each island can be considered as a unique balancing area. Consolidation will play a role when an inter island undersea transmission cable is in place. At that time, the interconnection of the separate grids will allow the consolidation of multiple island grids into a single BA with commensurate benefits. From a market structure and regulation standpoint, given that HECO controls the majority of the generation assets and load, especially on the islands that would be interconnected, HECO could act as the integrated BA authority Enabling Transmission Transmission (internal within a system and external to other systems) by itself does not provide flexibility. However, transmission provides access to geographically disperse renewable generation. Inclusion of a larger number of diverse locations of variable generation reduces the overall impact of changes in resource availability such as might result from changing cloud cover or gusts of wind. In order to include a wider array of sources, it is necessary to ensure adequate transmission between the diverse areas. Island systems are substantially limited in their ability to use transmission to integrate diverse geographic resources. Weather conditions may not vary significantly enough to provide the diversity required. However, the island interconnection cable will introduce some transmission 3 5

68 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII provided flexibility. Additionally, some opportunities will exist to facilitate stronger connections between generation and load centers to minimize the impact of transmission system outages or curtailments Energy Storage Energy storage is a technology that can be used to provide three different versions of electric system reliability functions: Load shifting service: The storage system charges in periods of surplus and discharges during periods of scarcity, Shorter term balancing service: Stored electricity is used to smooth variation of wind net load output thereby reducing the need for some spinning reserves, and Quick acting instantaneous service: The storage systems provide immediate frequency and regulation products, and Pumped hydro is an example of a proven traditional energy storage technique that is capable of providing any of the ten flexibility/reliability functions identified in Table Battery storage technologies are also a focus of significant development and investments to provide similar capabilities. 3.3 NEW EMERGING FLEXIBLE RESOURCES Historically, at HECO and most other utilities, system load was highly predictable within each day and from day to day. Previous days actual demand combined with near term weather forecasting provided system load predictability. However, this has changed with increased intermittent distributed generation (DG), mainly from solar PV, which has currently reached 5 42 percent of retail sales, depending on the island. Since DG appears as volatile load reduction, it creates load variability for the bulk system. Thus, today, the load shows greater daytime variation, based on temperature, cloud cover, humidity and other factors that affect PV generation. Also, the production at larger utility scale renewable generation sources varies significantly for the same reasons but is most often connected directly to the utility system as a discrete generation source and thus is easier to monitor. Utility systems, including those in Hawaii, traditionally use flexible resources that are designed into the utility system to adjust to demand and generation fluctuations, changes in load, changes in transmission power flow and unexpected outages. Wind and PV generation vary more than conventional generation due to their reliance on a natural energy source. Seasonal variation, diurnal variation, and weather based events can cause extreme voltage ramps and other power quality issues. In addition, the integration of power from numerous intermittent DG sources rather than large traditional generation sources, pose control and stability challenges for local distribution system operators that lack more sophisticated controls. 28 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November

69 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII New technologies, with increased flexibility will be required because operators must manage the additional variability and uncertainty that wind and PV introduce into the electric system. Intermittent renewable generation can reduce inertia and primary frequency response, cause largescale changes in generation, and cause stability issues. It can affect power quality, voltage control and reactive power management at the local level. For island systems without interconnections, such as those in Hawaii, relatively small mismatches between demand and generation result in a significant change in frequency. For HELCO, a 2 3 MW drop in energy production results in a 0.1 Hz drop in frequency, which on most systems would be considered a significant variation but is something accepted on an island grid. Larger balancing authorities with large amounts of generation require much larger changes in output to result in a similar frequency drop. As a result, additional flexibility will be required in the various time frames relevant to utility operations of milliseconds, seconds to minutes, minutes to hours, hours to days and beyond to many days to meet variability needs. This highlights the fact that additional flexibility will be necessary in Hawaii as compared to most utilities because smaller, isolated electric systems are more sensitive to variations in demand and generation than the larger and more integrated electric balancing areas that are found elsewhere in the U.S. This need for increased flexibility will affect long term transmission and resource adequacy planning and make it more costly to balance load and generation at all times. System reliability and stability and third party contract provisions may take precedence over economic dispatch based decisions. Strategies for addressing this can be seen in the utilities integrated resource plans (IRPs). Sophisticated communications and controls available from the Smart Grid will help provide the new flexible resources needed to accommodate high amounts of energy produced by variable renewable energy resources. New and emerging resources have more integrated electronic features than existing technologies. New emerging flexible resources include: Storage New bulk system central energy storage technologies, including flywheels, advanced batteries, Compressed Air Energy Storage (CAES), and others. These technologies are evaluated in another section of this HREGP report. Distributed stationary energy storage as demonstrated to date has not been economic on the mainland U.S. but may be economical in Hawaii considering the high cost of energy there. Distributed non stationary energy storage (e.g. electric vehicles) is being assessed under a separate section of this HREGP report. Demand Response 3 7

70 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII New types of DR, including direct load control, scheduled load management, customer load response, and rate design. Forecasting HECO has identified the benefits of improved forecasting and is working with industry forecasters to evaluate additional possibilities. Ancillary Services Reserves Management HECO will equip more existing conventional generators with automatic generation control (AGC) in the future. Hawaiian utilities have considered modest limits for ramp control of intermittent renewable generation and should use contingency reserves for extreme wind events. Proof of technical and commercial viability and some combination of changes to reliability standards, tariff and rate design and business rule changes will be necessary to adopt new technologies with increased capability to provide electric system flexibility functions. 3.4 A PROACTIVE APPROACH IN HAWAII A major effort is underway by the State of Hawaii to develop a proactive approach to achieve high levels of renewable generation. The Proactive Approach recommendation was developed in a multi party process by the Reliability Standards Working Group (RSWG), involving utilities, state agencies, and clean energy groups. The PUC convened the RSWG in September 2011 at HECO's suggestion to deal with reliability concerns arising from increasing levels of renewable energy entering the Hawaii utility grids. At the PUC's direction, the RSWG formally concluded its work in January 2013 by submitting an array of recommendations and reports, which included the development of a recommendation for a Proactive Approach (the label attributed to the recommendations). The RSWG supports and recommends the Proactive Approach as the next evolutionary step in renewable DG interconnection, towards the clean energy grid of the future. The recommendations included a set of transparent reliability standards for the operation and planning of the utility grid. When flexibility/reliability functions are sold in energy markets they are called ancillary services. Ancillary services are dependent on the characteristics of each and every power system. General Electric (GE) assisted the RSWG in defining a set of ancillary services that protect reliability, incent renewable generation, and minimize production costs. 29 They identified Hawaii specific differences relative to more standardized definitions. The following are the Hawaii specific RSWG recommendations regarding supply of ancillary services (flexibility/reliability functions). Smart Grid communication and control features will enable these flexibility/reliability functions, but in some cases will require additional demonstration, and cost justification before implementation. 29 HNEI, "Report on Business Case in Hawaii for Storage Options," US Department of Energy, Office of Electricity Delivery and Energy Reliability, Honolulu, HI, October

71 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Inertia The current and future combination of inertia and governor should be tracked and evaluated by the utilities to ensure all new generation fulfills frequency recovery needs and that needed response is provided at minimum cost. All power plants using rotating machines including those based on power electronics should be encouraged by utilities to provide maximum cost effective inertia to the power system. Emerging demand response and storage technologies that provide sufficiently fast response to assist with arresting frequency decay should be encouraged to provide maximum cost effective inertia like response to the power system. When the ability to provide adequate frequency recovery has been identified, utilities should endeavor to secure additional frequency recovery capability from new generating or storage plants or synthetic response from plants or storage systems where that is practicable. Engineering test procedures should be developed by utilities to perform field testing and measurement of synthetic or actual inertial capabilities for new or existing plants. Frequency Frequency response should be specified in terms of droop. The time frame of the response should be defined. All generators (thermal, hydro, geothermal, wind, solar, etc.) should be required by utilities to be capable of frequency response to the extent economically feasible. Storage and demand response should be encouraged to provide frequency response capability to the extent that they are physically capable of doing so. Utility system operators should utilize the least cost frequency response resources, which may vary from hour to hour. AGC All generators (thermal, hydro, geothermal, wind, solar, etc.) should be required to be capable of responding to AGC commands. Each utility's AGC system should be modified to be able to utilize the full range of responsive resources (conventional generators, variable renewable generators, storage, and demand response) unless detailed studies demonstrate that this is not economically justified. Storage and demand response should be allowed to provide AGC response to the extent that they are physically capable. If the device or generator interface and response capabilities support the necessary AGC response (i.e.; AGC interface is provided, the response is predictable, and within the appropriate time frame) the device or generator should participate in frequency regulation as well as economic dispatch under AGC control. Utility system operators should utilize the least cost AGC resources which will likely change from hour to hour during conditions when frequency is within specified tolerances. Spinning Reserves Utilities should require all generators (thermal, hydro, geothermal, wind, solar, etc.) to be designed with spinning reserve capability with the best response practicable. Storage and demand response should provide spinning reserve to the extent that they are physically and economically capable. System operators should utilize the least cost available resources for spinning reserves, which will likely change from hour to hour. Contingency Reserves Utilities generally calculate the amount of spinning reserve required by defining the largest single contingency (i.e.; largest online generator). Larger units or installations, while less expensive per kw for the initial installer, will increase contingency reserve requirements and costs for the BA. Any added reserve cost should be considered when evaluating any new generation project. Utilities should consider the need for non spinning and supplemental reserve in their routine generation planning. All generators, storage and demand response should be able to provide non spinning and supplemental reserve to the extent that they are physically and economically capable (i.e.; ability to come online quickly or operate in standby mode). Electric system operators should utilize the least cost replacement reserve resources, which will likely change from 3 9

72 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII hour to hour. The total cost for providing replacement reserves is typically dominated by the standby cost rather than by the activation cost. Reactive Power and Voltage Regulation All generators (thermal, hydro, geothermal, wind, solar, etc.) should be required by utilities to be capable of providing reactive power and controlling voltage in response to system operator commands. Storage should be encouraged to provide reactive power and voltage regulation. This may influence where storage is installed on the power system to make best use of the reactive power and voltage regulation capabilities. If storage must remain on line just to provide reactive power and voltage regulation, it should be compensated for doing so. Utilities should consider alternative sources of controllable reactive power to reduce operating cost where reactive power needs would otherwise require out of merit order or must run operation. Short Circuit Low short circuit current problems due to increased penetration of renewable resources should be monitored and appropriate remedial action taken by utilities. As more generators that provide little or no short circuit current are installed, problematic protection should be changed to a type that requires lower short circuit current. Motor starting or voltage flicker should be monitored at both the transmission and distribution levels for adverse impacts due to the increased penetration of renewables. When energy storage is considered for its many benefits, the short circuit contribution of pumped hydro systems should be taken into consideration. Planning HECO should evaluate and publicize excess energy expectations based on demand and generation forecasts for five to ten years into the future. Compensation To provide and obtain these ancillary services in the most cost effective way, utilities should compensate independent power producers (IPPs) for direct and opportunity costs of resources providing increased inertia or frequency regulation, spinning reserve, and replacement reserve. Utilities should develop energy pricing tariffs that reflect the true marginal cost of energy, especially during curtailment periods and tariffs for both storage and demand response to encourage investment in technologies that reduce or eliminate power system minimum demand problems.. In addition, the following are recommendations RSWG made with respect to demand side programs. The utilities should investigate and define options the use of demand response and energy storage to provide ancillary services that are technically possible and economically justified. They should investigate pricing programs and manual and automated demand response programs that will incentivize customers to change their consumption patterns in ways that are beneficial for stakeholders. Included in this investigation would be an analysis of the benefits of increasing demand during minimum load periods (i.e., examining the cost reductions that could be incurred and the impact on renewable energy purchases during the entire 24 hour day). Regulators should allow the utilities and other interested stakeholders to develop specific pricing and/or manual demand response programs, with expedited regulatory review and approval to get these programs in place as soon as possible. As the Commission reviews new DR programs, it should consider the appropriate role of third party agents and aggregators (i.e. curtailment service providers) to deliver demand response programs effectively and efficiently. 3 10

73 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII With regulatory approval utilities should ensure that demand response programs are considered in the Integrated Resource Planning process and direct the energy efficiency potential study contractor to perform specific load research data collection that will allow them to better estimate the demand response potential in Hawaii. State government should require that Hawaii Energy work with the utilities to identify those customers and loads that are most promising for demand response, and ensure that Hawaii Energy and the DR planners coordinate program plans and marketing to ensure that energy efficiency does not compromise promising DR opportunities (and vice versa). 3.5 SECTION SUMMARY Intermittent renewable generation increases the need to supply electric system flexibility/reliability functions with existing traditional and new emerging flexible resources. When flexibility/reliability functions are sold in energy markets they are defined as ancillary services. Hawaii s RSWG worked with GE to define a set of Hawaii specific ancillary services that could be used to protect reliability, incent renewable generation, and minimize production cost for each Hawaii island. 30 Existing traditional and new emerging flexible resources will be enabled and enhanced with the advanced communications and control capability of the Smart Grid. Examples of emerging flexible resources that will be enabled by Smart Grid communications and control capability include fixed and mobile energy storage, new types of demand response, improved forecasting of renewable energy resources, and new methods of reserves management. 30 HNEI, "Report on Business Case in Hawaii for Storage Options," US Department of Energy, Office of Electricity Delivery and Energy Reliability, Honolulu, HI, October

74 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 4.0 Energy Efficiency (EE) The Hawaii Clean Energy Initiative includes a significant energy efficiency (EE) goal: 30 percent of forecasted 2030 energy use, equivalent to 4,300 GWh reduced by For context, Hawaii s 2011 statewide energy efficiency program savings report indicates 130 GWh savings achieved that year. This single year savings is 3 percent of the 2030 goal. There is much to be done, in order to reach Hawaii s 2030 EE goal. At the time of writing of this report, because Hawaii has already achieved higher levels of renewable energy penetration in some parts of its grid than many other grids in the world, and because increased levels of EE reduce overall electric load, Hawaii, in particular will require significant Smart Grid advances, to manage the concurrent requirements of decreased load and increased penetration levels of intermittent renewable energy. The Energy Efficiency section of this report covers the following range of topics related to maximizing EE generated reductions through use of Smart Grid technologies. 1. Identifies the value and advantages of EE. 2. Identifies existing end use EE programs in Hawaii. 3. Identifies EE programs that improve the efficiency of the grid rather than the end use efficiency. 4. Explains how smart meters can improve end use and electric system efficiency. 5. Explains the effect of end use efficiency on electric system reliability and flexibility. 6. Identifies additional EE programs that can enhance the State s ability to meet its EE goals. 7. Identifies the difference between EE and DR since they can compete for the same loads. 4.1 THE VALUE AND ADVANTAGES OF ENERGY EFFICIENCY Energy efficiency is a critical component of Hawaii s Clean Energy Initiative, and it is an important part of each of the island utilities Integrated Resource Plans (IRPs). EE is typically considered the most cost effective means to meet future energy demand. EE reduces the need for new generation,, thus reducing overall system cost. Many States have recognized this and have provided EE performance standards and other incentives to induce utilities and customers to adopt it on a priority basis. Independent reports by both the Electric Power Research Institute and McKinsey and Company in indicated that energy efficiency presents a vast low cost energy source nationwide for the U.S. There is the potential, according to McKinsey, to reduce annual nationwide energy consumption significantly with an array of Net Present Value positive energy efficiency measures, resulting in net cost savings to ratepayers. In addition to economic advantages, energy efficiency provides an air quality emission free energy resource. Implementing energy efficiency 31 Note: the HCEI roadmap, dated 2011, indicates 4,300 MWh reduced by 2030 (pg 4). The HI 70% scenario analysis indicates that the statewide EE goal is 4,300 GWh reduced by 2030 (pg 43). This is 3 orders of magnitude different. I assume the HCEI Roadmap contains a typo, and the actual goal is 4,300 GWh, not MWh gy_efficiency_in_the_us_economy df 4 1

75 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII strategies can result in significant reductions in greenhouse gas emissions, and it can serve as an important bridging technology toward long term advanced low carbon technology scenarios. 4.2 EXISTING ENERGY EFFICIENCY PROGRAMS IN HAWAII Hawaii s 2011 Clean Energy Initiative Roadmap 33 contains a summary of statewide energy efficiency accomplishments to date, which include the following: Established quantitative 2030 energy efficiency savings goal. Decoupled HECO companies revenue stream from electricity usage. Established public benefits fund to help finance the retrofitting of the existing building stock. Deployed a State Lead By Example program. Transferred HECO building energy efficiency programs to public benefit fund (PBF) program administrator, and deployed monitoring and verification programs. Existing KIUC/PBFA programs are fully operational. Collected building stock data and identified needs. All counties have adopted new, more efficient building codes (IECC2006 for Maui, and Oahu and The Big Island). Pilots of highly efficient new homes have been conducted through Department of Defense and Department of Hawaiian Homelands. Conducted new construction efficiency savings potential analysis. Evaluated potential efficiencies from improving transmission/distribution infrastructure. Identified industrial and military electricity applications for future energy efficiency potential study. For end use energy efficiency in particular, Hawaii has established a public benefits fund for energy efficiency and demand side management. This fund was established in 2006, and its size is generally calculated as a fraction of the utilities annual revenues. For 2013, the fund is 2 percent of the utilities revenues including taxes. This funds The Hawaii Energy Conservation and Efficiency Program (Hawaii Energy), a program operated by SAIC, an independent third party contractor for the Hawaii Public Utilities Commission. The programs supported by the public benefits fund include residential programs, business programs, and Market Transformation programs (these include workforce development, energy audit and benchmarking tools and support, and education programs such as Energy Efficiency through Financial Literacy, and Residential Home Rating Analysis ). Energy efficiency technologies promoted by these programs include: Clothes Washers, Refrigerators, Ceiling Fans, Water Heaters, Lighting, Lighting Controls/Sensors, Chillers, Heat Pumps, Central Air Conditioners, Heat Recovery, Windows, Motors, Processing and Manufacturing

76 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Equipment, Custom/Others pending approval, LED Exit Signs, Pool Pumps, Commercial Refrigeration Equipment, Food Service Equipment, LEDs, and Heat Pump Water Heaters. 34 An overview of the latest results achieved by these energy efficiency programs can be seen in the statewide program measurement and verification reports, which are summarized below. 35 Table 4 1 PROGRAM 2011 Hawaiian Energy Efficiency Program Results DESCRIPTION 2011 VERIFIED SAVINGS [KWH] Business, Energy, Efficiency, Measures (BEEM). Custom, Business, Energy, Efficiency, Measures, (CBEEM). Business Service and Maintenance (BESM) Business Hard to Reach (BHTR). Provided prescriptive Incentives to business customers who purchased and installed energy efficiency measures. The program paid incentive rebates for lighting, air conditioning, water heating, water pumping, motors, building envelope improvements, energy awareness, measurement and control systems, and ENERGY STAR business equipment. Provided custom financial incentives based on calculated savings to commercial, institutional, governmental, and industrial sector customers. Provided incentives and direct installation of measures to business in addition to business design, audits, and commissioning. Provided equipment grants in addition to building owner, tenant and apartment/condo complex measures. 35,267,460 22,115,468 2,057,135 1,675,686 Non residential total 61,115,749 Residential Energy Efficiency Measures (REEM). Residential Energy Services and Maintenance (RESM) Provided prescriptive incentives to residential customers who purchased and installed energy efficiency measures. The program paid incentive rebates for prescriptive measures, including water heating, lighting, air conditioning, appliances, and awareness, measurement and control systems. Provided incentives to direct installations, design and audits, and system tune ups. 66,877,382 91,

77 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Residential Hard to Reach (RHTR) Provided equipment grants in addition to landlord, tenant and apartment/condo complex measures. 2,021,151 Residential Total 68,990,014 Program Overall 130,105,763 The energy efficiency portfolio design for PY12 yields an average measure life of 9.1 years across the overall portfolio and would result in lifetime program energy savings of $0.02 per kwh. 36 In 2011, the State invested 1.13 percent of annual utility revenues in electricity efficiency programs, and the State achieved savings of 1.15 percent of statewide retail electricity sales in For comparison, California (frequently perceived as a leader in EE) invested 3.35 percent of annual utility revenues in electricity efficiency programs in 2011, and achieved an average savings of 1.79 percent of statewide retail electricity sales in Much of the funding for energy efficiency in California is used to support natural gas savings, which are not included in our review of Hawaii s energy efficiency programs, and which are beyond the scope of this report. California has implemented several EE measures of note, which are not currently being implemented in Hawaii, but which may be worth consideration for adoption in Hawaii, as they have achieved great success and recognition in the EE professional community. These include the following, as noted by ACEEE: 38 Reward structures for successful energy efficiency programs: incentive earnings accrue only if the IOU energy efficiency portfolio of programs achieves a minimum percentage of the CPUC s goals. Utilities receive higher net benefits if they meet or exceed savings goals. Building codes: California s energy code is considered to be the most aggressive and best enforced energy code in the United States, and has been a powerful vehicle for advancing energy efficiency standards for building equipment. Many specifications are performancebased, offering flexibility for designers. The most recent code, effective January 1, 2010 is mandatory statewide and exceeds 2009 IECC standards for residential buildings and meets or exceeds ASHRAE/IESNA for commercial buildings. For comparison, Hawaii s current building codes are 2006 IECC with Amendments. According to ACEEE, Hawaii has not yet adopted IECC 2009, but adoption with amendments appears to have been in progress as of May Hawaii Energy s Program Year 2012 Annual Plan 37 sheet/2012 spending and savings tables.pdf 38 policy/california 4 4

78 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Appliance efficiency standards: Presently, California has adopted standards for ten products that are not covered by federal standards. Of particular note are the state s precedent setting efficiency standards for incandescent light bulbs, metal halide lamp fixtures, and televisions. Energy efficiency financing programs, which offer eligible customers 0% financing for qualifying energy efficient improvements. 4.3 TRANSMISSION AND DISTRIBUTION EFFICIENCY, VOLTAGE CONSERVATION AND VOLTAGE/VAR OPTIMIZATION In addition to end use efficiency, another type of energy efficiency involves increasing the efficiency of the power delivery system itself. Because Section 4 is focused on end use efficiency, we will simply note here that there are additional types of efficiency discussed elsewhere in this report. Please see Section 6 for a discussion of how the Smart Grid will allow more effective control of distribution system voltage and allow improvement of electric system efficiency. Utilities are implementing conservation voltage reduction (CVR) programs and volt/var optimization (VVO) programs, sometimes as part of wider ranging Advanced Distribution Management System (ADMS) initiatives, to decrease system losses. These programs have the potential to reduce demand by %, and to reduce energy by %. 39 With full implementation these programs will provide additional system flexibility, will provide additional control of local voltage and will allow greater penetration of distributed renewable energy while at the same time providing overall efficiency improvements. It is worth noting here that, while CVR programs tend to produce overall system efficiency improvements, there are cases where this rule does not apply. In particular for large industrial complexes with older motors, recent literature indicates significant value in optimizing for power factor rather than CVR SMART METERING INCREASES CUSTOMER DRIVEN EFFICIENCY Smart metering solutions, which will be discussed in depth in Section 6, are a Smart Grid feature that can enable electric system flexibility. However, they are also discussed here since they have great potential for raising consumer awareness and for driving energy efficiency. Smart meters can provide energy efficiency value due to the education effect they have on energy consumers. Traditional monthly billing provides baseline information but is generally confusing to customers and makes it challenging for consumers to reduce their day to day energy usage since they cannot see the impact of individual actions. Rather, consumers require a feedback loop with shorter delay to understand the implications of their usage patterns and behavioral changes. The European Smart Metering Industry Group (ESMIG) recently noted that unless people know how much energy they are consuming, when they are consuming it, and how much it costs, it will be almost Ibid. 4 5

79 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII impossible for them to conserve energy or shift consumption in any significant way. 41 Smart meters provide an effective avenue to address this gap and enable energy savings through better awareness of energy consumption. However, smart metering by itself does not result in smart consumers. Captured data must be coupled with end user services and devices to allow for interpretation of results. Data must be presented in a clear, attractive, and timely manner in order to affect consumer behavior. The tool for aggregating, managing and presenting the metering and associated data collected from AMI systems is the MDMS (Meter Data Management System). This application provides a platform on which on the data business rules and processes are applied to the collected data. The resultant processed information is then made available to other internal utility applications and users, and is structured for customers access. Figure 4 1 provides an example of a MDMS implementation architecture and data flows that illustrates the Smart Grid capabilities of an MDMS. Figure 4 1 MDMS Implementation Architecture The most basic feedback channels for this information are energy reporting via online portals or smart phone applications and indirect feedback via more informative billing statements which track individual progress and may also show comparisons to similar homes. More advanced 41 ESMIG, Position Paper on Smart Metering in the Energy Efficiency Directive, January

80 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII feedback involves real time data through in home energy displays and the introduction of smart products with automated controls which can react to utility signals, price changes, and consumer preferences. For customers that have internet access, consumer web portals have become very popular because they are relatively inexpensive, requiring no new specialized hardware (such as an in home display) and utilizing the metering data (commonly 15 minute or hourly) already available from the smart metering system. Figure 4 2 provides an example of a commercially available customer energy usage presentation application. Mobile applications, which can be viewed as an extension of web portals, are also gaining traction since they can be accessed from anywhere and since increasing percentages of consumers have access to the requisite mobile devices, such as smart phones and tablets. While not all customers have Internet access at home or possess smart phones, the overall percentage of access to these tools in increasing to where utility must focus on these customer engagement paths to reach the bulk of customers. Figure 4 2 Customer Energy Use Data Presentment A noteworthy point highlighted by a number of studies show that savings tend to improve when multiple methods are combined. This effect should be expected to grow as future consumers have more options and ways of receiving timely information. Simple and low cost consumption feedback systems could be widespread within the next decade, leading to some amount of energy savings purely through behavioral changes. The following figures show the range of energy conservation which can result from increasing consumer awareness through smart metering data. Real time displays: 3 12% (8.7% average) Web portals: 0 8% (5.1% average) 4 7

81 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Monthly information & feedback: 2 9% (5.9% average) These data are based on a global study of over 450,000 residential energy consumers comparing 100 different utility projects. 42 Energy savings effects due to education and increased awareness may fit best under Hawaii s existing energy efficiency program structure, in the category of transformation programs activities include education, training and other similar transformational activities that may not result in immediate quantifiable energy savings, but are likely to contribute to energy savings over time. 4.5 GREEN BUTTON INITIATIVES PRESENT NEW OPPORTUNITIES TO INCREASE CUSTOMER DRIVEN EFFICIENCY In addition to smart metering, there is an information technology initiative underway nationwide in the U.S. that seeks to promote new methods for consumes to monitor and interact with their energy use. This Green Button initiative is a response to a 2011 White House call to action, requesting that the electricity industry provide electricity customers with easy access to their energy usage data in a consumer friendly and computer friendly format via a "Green Button" on electric utilities' website. Green Button is a common technical standard, which enables software programmers and developers to work with energy usage data, providing customers with apps and games to make informed energy decisions, to set energy reduction goals, to compete, and to track and share their progress toward these goals through social media ENERGY EFFICIENCY CREATES THE NEED FOR ADDITIONAL SYSTEM FLEXIBILITY Increases in EE bring reduction in load. As a result, the extant renewables therefore become a higher percentage of generation. This requires additional system flexibility. To maintain reliability, utilities are increasingly applying smart technologies such as Distribution Automation (DA) and Demand Response (DR,), as well as flexibility mechanisms that provide electric system reliability. These mechanisms are further discussed in Section 5. The grid stability issue related to renewables is a function of both the overall quantity of energy and also of the timing of energy delivery. While EE may represent the lowest cost and lowest impact activities, the overall grid stability must be kept in mind. Black & Veatch recommends that steps taken toward the State s EE and RE goals should be performed in sequence, without sacrificing grid reliability on the path to low energy consumption and high penetration renewable energy. Black & Veatch recommends an approach that balances EE and RE, in order to maximize system flexibility and reliability. 42 Jessica Stromback, Christophe Dramacque, Mazin Yassin; The potential of smart meter enabled programs to increase energy and systems efficiency: a mass pilot comparison, VassaETT,

82 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 4.7 OPPORTUNITIES TO ENHANCE HAWAII S ENERGY EFFICIENCY PROGRAMS, AS IDENTIFIED TO DATE IN STATE PLANNING DOCUMENTS This section reviews opportunities that have already been identified in state planning documents, for enhancing existing Hawaiian Energy Efficiency programs to meet State goals and objectives: The bullet list below summarizes lessons learned, reflections, and strategies for optimal future impact, from the Hawaii Energy Annual Plan : Minimize free ridership and eliminate underperforming incentive offerings, Focus incentives on high and persistent savings measures, Ensure hard to reach consumers are served well on all islands, Maximize direct customer benefit from each Public Benefits Fee (PBF) dollar collected, Make customers aware that Program incentives offer a great return on their PBF investment, Focus attention and budget on cultural behavior and other transformational energy changes, Use technology to increase consumer awareness of their own real time energy use, Make energy conservation and efficiency the go to first choice for every energy consumer, Use all available data to target best opportunities, identify outliers and track performance, Constantly review, modify and diversify portfolio offerings to ensure maximum performance, Explore best practices and new outside the box opportunities to keep strong momentum, Educate a broader base of customers/decision makers about energy conservation and efficiency, Provide critical leadership in EEPS, IRP, HCEI and other clean energy efforts, Give decision makers better tools to monitor and understand their own energy performance, and Reach deeper into consumer energy planning, operations, maintenance and energy use tracking

83 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Black & Veatch recommends the following strategies to reach these goals. Table 4 2 GOAL Strategy Recommendations to Meet Energy Efficiency Goals STRATEGY Minimize free ridership Maximize benefit for each dollar collected Increase customer awareness Best practices and out of thebox opportunities to keep strong momentum Free riders are defined as participating customers who would have bought a program sponsored product without the utility program intervention. For California EE program measurement and verification (M&V), the net to gross ratio is defined as the percentage of non free rider participants (i.e., the number of participants who purchased the product because of the program and its promised benefits) divided by total program participants (i.e., the number of people receiving a rebate). Hawaii s M&V methodology in the Hawaii Energy Efficiency Program Technical Resource Manual (TRM19) includes a version of this net to gross ratio the realization rate. The realization rate could be calculated in more precise detail based on updated surveys and field research. Enhance cost/benefit calculation methodologies to include demand response measures and enhanced monitoring and control measures, in particular, those described in Section 6 of this report. 1. Increase deployment of advanced metering. 2. Take advantage of nationwide Green Button initiatives to encourage customers to make informed energy decisions 45 Promote new and emerging EE technologies such as seawater air conditioning district cooling systems, solar air conditioning, heat pump water heating, ice storage, and the use of rejected heat from combined heat and power (CHP) systems. Increase stringency and enforcement of building codes and appliance efficiency standards Adopt reward structures for utilities exceeding energy efficiency goals Create energy efficiency financing programs, in order to increase accessibility of EE to the broader public. Continue to expand the high penetration of solar water heating already present on the islands

84 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 4.8 OPPORTUNITIES TO ENHANCE HAWAII S ENERGY EFFICIENCY PROGRAMS, AS IDENTIFIED BY THIRD PARTIES In the energy efficiency cost/benefit literature, there are several broad categories of typical EE measures which states implement in order to accelerate adoption of EE technologies and practices statewide. The American Council for an Energy Efficient Economy s EE scorecard 46 summarizes a representative sample of these measures. Scoring categories, and Hawaii s scores for each category, are provided in the table below. Table 4 3 ACEEE 2012 Energy Efficiency Scorecard for the State of Hawaii SCORING CONTENT AREA POINTS AWARDED TO HAWAII POSSIBLE POINTS PERCENT OF POSSIBLE POINTS ACHIEVED,% Utility & Public Benefits Programs and Policies Transportation Policies Building Energy Codes Combined Heat and Power State Government Initiatives Appliance Efficiency Standards Total Score Hawaii ranks 18 th of the 50 U.S. states, in ACEEE s 2012 Energy Efficiency Scorecard. Hawaii has demonstrated leadership in Building and Energy Codes, and in Utility & Public Benefits Programs and Policies, according to this scorecard. This scorecard also indicates that Hawaii has room for improvement in several areas. The areas with the greatest room for improvement are, in order of potential improvement by ACEEE standards: appliance efficiency standards, combined heat and power, state government initiatives, and transportation policies. Because the focus of this report is electricity only, Black & Veatch recommends prioritizing the two electricity measures in this list: appliance efficiency standards and state government initiatives. Recommendations in each of these areas are provided below

85 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Appliance efficiency standards The following recommendations apply to appliance efficiency standards. Adopt appliance efficiency standards for televisions and battery chargers, following California s best practice example. 47 Adopt efficiency standards for incandescent lamps that are more stringent than the federal standards, following Nevada s best practice example. 48 Adopt efficiency standards for toilets, urinals, faucet aerators, showerheads, commercial pre rinse spray valves, following the best practice example of California, Georgia, and Texas. These measures produce water savings as well as energy savings associated with the production, heating, treatment, pumping, and disposal of water and wastewater State government initiatives The following recommendations apply to appliance efficiency standards. Existing lead by example programs to improve the energy efficiency of state agencies, state owned buildings and fleets are widely recognized to be high achieving. Activities include promoting and executing energy savings performance contracts throughout the state, developing a pilot sustainable community with net zero energy homes, benchmarking, retrocommissioning, and LEED certifying a number of state buildings. Opportunities remain to enhance state funded research, development, and demonstration programs for energy efficiency technologies and practices. The state building code has been updated to IECC 2006 for most islands. This could be further updated to IECC 2009, as has been done in Kauai. 4.9 THE INTERACTION BETWEEN ENERGY EFFICIENCY & DEMAND RESPONSE EE and DR programs can target the same loads. Therefore it is useful to understand the difference between the two. EE seeks to provide a sustained reduction in the amount of energy consumed by an appliance or load device through the application of technology improvements or operating 47 Source: ACEEE 2012 State Energy Efficiency Scorecard, p. 86. Accessed online, June 2013 at the following URL: Excerpt: California, scoring the maximum 2 points, continues to take the lead on appliance efficiency standards, most recently adopting the first ever standards for televisions as well as standards for battery chargers. Not only has California adopted the greatest number of appliance and equipment standards, many other states standards are based on Califronia s, such as the television standards passed in Connecticut in Source: ACEEE 2012 State Energy Efficiency Scorecard, p. 86. Accessed online June 2013 at the following URL: Excerpt: In our 2011 Estate Energy Efficiency Scorecard, Nevada earned credit for adopting standards for general service incandescent lamps that are more stringent than existing federal standards. However, those standards are not yet being enforced and it is uncertain when they will begin to be enforced, so we have deducted those points indefinitely. 4 12

86 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII measures. This baseline energy use reduction lowers base load and can reduce plant cycling and the use of peaker plant units. It is common for utilities to capture and quantify both energy efficiency (kwh) and peak demand reduction (kw) benefits for each EE measure, in their EE program measurement and verification reports. Demand response is discussed further in Section SECTION SUMMARY Energy efficiency is a critical component of Hawaii s Clean Energy Initiative, and it is an important part of each of the island utilities Integrated Resource Plans (IRPs). EE is typically considered the most cost effective means to meet future energy demand. EE reduces the need for new generation,, thus reducing overall system cost. Many states have recognized this and have provided EE performance standards and other incentives to induce utilities and customers to adopt it on a priority basis. Independent reports by both the Electric Power Research Institute and McKinsey and Company in indicated that energy efficiency presents a vast low cost energy source nationwide for the U.S. By increasing the ratio of renewable supply to grid load, EE reinforces the need for Smart Grid technologies to keep supply and demand in balance. EE also overlaps in some cases with DR, and the result is a lean, reliable grid. Specific opportunities to strengthen energy efficiency policies and programs in Hawaii include minimizing free ridership, promoting new energy efficiency technologies, enhancing customer experience with smart meters, smart appliances, and green button tools, strengthening appliance efficiency standards, strengthening state government funded R&D initiatives, increasing stringency of building codes, adopting reward structures for utilities exceeding energy efficiency goals, and enhancing cost benefit accounting to include new programs and technologies. Black & Veatch recommends that the annual statewide measurement and verification reports, which currently address a number of energy efficiency programs, should be expanded to include cost benefit accounting for demand response, and for communication, monitoring and controls programs. This would enable capturing of additional savings, and ensure that co benefits of EE measures do not go unrecognized gy_efficiency_in_the_us_economy df 4 13

87 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 5.0 Demand Response Section 3 identified Demand Response (DR) as a flexible resource that can provide reliability functions necessary to integrate larger amounts of intermittent renewables into a grid. This section presents DR technologies in more detail, focusing on customer programs and engagement strategies which can optimize the use of renewable energy and help meet Hawaii s energy objectives. Existing and emerging DR programs and mechanisms will be reviewed as well as potential barriers to adoption and possible solutions to overcome those barriers. While DR programs have typically been economically justified by avoiding the high cost of building additional peaking generating capacity, they are equally relevant for integrating intermittent renewable generation. DR technologies that meet established performance criteria can provide power system balancing needs raised by the presence of renewable generation. By adjusting the timing of demand to enable renewable energy to be delivered and consumed during optimal hours, DR programs can address the intermittency of renewable energy generation. 5.1 WHAT IS DEMAND RESPONSE, WHAT ARE THE TYPES OF DEMAND RESPONSE Demand Response has been effectively used in the electric power industry for decades to induce lower electricity use during critical time periods such as times of system peaks, high wholesale market prices, limitations of power supply, or when system reliability is otherwise jeopardized. By their nature these critical periods are intermittent and random. In addition, DR programs were most effective in regions with large cooling loads or large amounts of electric heating (space and water). However, advances in communications and controls technologies are expanding the ability of all types of consumers to both respond to system operator directives and to respond to price signals, making DR applicable in almost any region. DR is not a single technology. Rather, DR is any technology that controls the rate of electricity consumption rather than the rate of generation and thus, there are numerous existing DR technologies. The Federal Energy Regulatory Commission (FERC) defines demand response as a reduction in the consumption of electric energy by customers from their expected consumption in response to an increase in the price of electric energy or to incentive payments designed to induce lower consumption of electric energy. 50 While DR traditionally referred to programs which reduced system peaks, FERC considers DR to include consumer actions that can change any part of the load profile of a utility or region, not just the period of peak usage. FERC goes on to recognize DR as including devices that can manage demand as needed (by, for example, responding 50 FERC, Wholesale Competition in Regions with Organized Electric Markets, Order No. 719, FERC Stats. & Regs. 5 1

88 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII automatically to near real time signals) to provide grid services such as regulation and reserves, and changing consumption for the smart integration of variable generation resources. All demand response technologies and programs fall into two broad families dispatchable and non dispatchable load. Dispatchable load provides utility system operators with direct control of devices or with specific actions which a customer has agreed to take upon request. Because it can be consistently relied upon, dispatchable load is also known as load as a resource. Nondispatchable load, on the other hand, encourages demand reductions or demand shifts indirectly by adjusting the price of electricity. Non dispatchable load, also known as price responsive demand, includes dynamic pricing programs such as Time of Use (TOU), Critical Peak Pricing (CPP), and Real Time Pricing (these programs are defined in detail below). While non dispatchable programs can be very effective if designed properly, the impact is generally more variable since customers choose whether or not to participate at any given time based solely on their level of price comfort. The following are types of DR programs and mechanisms: Direct Load Control, including control of individual appliances such as water heaters, air conditioners, and smart appliances (Dispatchable Load). Load aggregation and management by third party suppliers (Dispatchable Load). Loading Shifting & Scheduled Load Management (Dispatchable Load). Rate Design Programs Price based, including real time incentives for utility control of load (Non Dispatchable Load). Loading Shaping (Dispatchable Load) Direct Load Control Direct Load Control (DLC), or load shedding, is the most basic form of dispatchable load and has been in use for several decades to shut off devices during peak events. The most common solutions utilize load control devices which are designed to control up to a 30 Amp load per device using a relay switch. The devices are wired directly into the power supply for the targeted appliances and are reached using a variety of communication mechanisms, including power line based systems, dedicated radio networks, and cellular systems. Available from a wide variety of vendors, DLC switches have proven to be a versatile and reliable method for demand response since they can connect to common home loads and can be used to respond to DR signals from a utility, to local home energy management system settings, or to utility pricing signals. The combination of enabling technology with pricing signals has been shown to produce the greatest impact on customer loads. The most commonly controlled loads are electric water heaters and pool pumps since these appliances can be controlled with little to no end consumer impact/awareness. In addition to responding to basic on/off signals, most DLC devices today have been designed to provide appliance cycling over a requested period of time, mimicking the natural on/off cycle but with 5 2

89 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII shorter on durations. This makes DLC also effective for air conditioning control since compressors can be cycled on and off, minimizing the increase in room temperature to which the consumer would be sensitive, while still providing a decrease in overall load. One of the most successful DLC programs in the world has been the On Call Savings Program run by Florida Power & Light (FPL). Serving a population on a peninsula, much of which is concentrated in the southern end, FPL s sensitivity to grid reliability is similar in nature to Hawaii s in that there are limited ways for power to be transmitted into some portions of the territory. The On Call program attracts both residential and commercial customers and guarantees savings by providing participants with monthly credits, regardless of whether control events are called, which typically only occur 3 4 times per year. With more than 780,000 customers enrolled, the program reliably controls ~1,000 MW of power on a regular basis and can provide as much as 2,000 MW of dispatchable load in emergency situations. While providing a large reliability and emergency resource to the utility, the energy saved by the program is estimated to have deferred construction of three medium sized power plants since the program's inception. 51 The advent of the Home Area Network (HAN) has enabled a new generation of DLC, which creates small local networks to communicate between in home devices. Spurred largely by Advanced Metering Infrastructure (AMI) deployments, HAN s typically utilize a low power radio frequency (RF) link to connect a gateway (which could potentially be the smart meters) to in home appliances. A fully deployed Home Area Network (HAN) can include multiple end points, including DLC devices connected directly to large appliances, thermostats to manage heating and cooling systems, and in home displays (IHD) to provide near real time energy usage information. The DLC devices used in HAN solutions provide the same general functionality described above to control large home loads. The newer types of devices enabled by HAN solutions include programmable communicating thermostats (PCT), smart plugs, and smart appliances. PCT s, also known as smart thermostats, are compatible with a wide variety of residential heating and cooling systems. Using built in wireless chips, they can receive temperature set point and schedule changes initiated by the utility or by the user either from inside the home directly or remotely via web portals or smart phones. PCT programs have been a common focus of DR pilots the past several years since they utilize a device which is already common in many U.S. homes and since temporarily adjusting the set point of a thermostat during a DR event reduces the load but still allows the air conditioning system to cycle on/off, minimizing consumer awareness of the event. However, PCT programs have limited applicability for Hawaii s utilities since most homes in Hawaii do not have central air conditioning. Window air conditioning units are more common. Smart plugs are similar in functionality to DLC switches but plug directly into standard wall outlets, allowing them to turn almost any plug in load into a controllable one, including window air conditioning 51 Florida Power & Light Company,

90 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII units. Smart appliances, on the other hand, integrate the necessary communications hardware into traditional home appliances such as washers, dryers, and water heaters. Major appliance vendors such as GE have solutions, but options are limited and prices are high compared to standard appliances in a market which is extremely cost competitive. Further limiting penetration is that there is no real incentive for consumers to purchase such appliances save individuals own desires to do their part for the environment and grid reliability. Smart plugs may offer the best path to penetration since they are low cost and provide customers with easy entry, but, even in this case, there remains little tangible benefit for today s consumer Third Party Aggregation Third party aggregators bundle load for collective power purchasing and bundle DR for selling purposes back to utilities or grid operators. Electric load aggregation is the process by which large energy consumers band together to secure more competitive prices than they might otherwise receive working individually. Similarly, with DR aggregation a third party company pools together the dispatchable load from multiple commercial and industrial customers to provide what is essentially a standby virtual power plant. Through contracts with grid operators, third party aggregators guarantee specific amounts of dispatchable load by using software systems to remotely and automatically enact agreed upon energy reduction plans within specified response times, often through direct load control mechanisms. While most utilities have both residential and commercial facing DR programs, third party DR aggregation programs apply almost exclusively to commercial and industrial energy consumers since they represent much greater consumption. Whereas a residential demand response program may require hundreds or thousands of participants to provide any meaningful quantity of dispatchable load, industrial customers offer huge opportunities for DR because they are able to contribute extremely large amounts of load reduction. For example, a single brewery in Boston, MA provides independent system operator ISO New England with 350 kw in curtailment and a group of five shopping centers in Arizona provides Salt River Project (SRP) with ~1.5 MW in total ondemand reduction. 52 It would take approximately 75 and 330 residential consumers, respectively, on water heater load control to equal these levels of dispatchable load. 53 Third party DR providers offer commercial and industrial (C&I) customers a win win proposition by designing energy curtailment plans which reduce non essential energy use during critical periods of electric supplydemand imbalance while minimizing the impact on day to day operations. In return, these customers get paid for the energy saved. And, just like residential DR program participants, they also often get paid year round just for being on call. On the other end, utilities and grid operators are provided with virtual power plants from which they can draw when necessary and do so without the burden of managing an in house DR program since third party providers often support the entire delivery process: from marketing and customer recruitment to measurement and 52 EnerNoc, Inc.: resources/case studies 53 Calculations are rounded down and assume a standard 4500W heating element 5 4

91 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII verification. 54 A key factor, however, is that a critical mass of large industrial scale loads is required in order to make third party aggregation commercially viable. Therefore, Hawaii may not be a suitable choice for aggregators or have sufficient industrial load to support multiple aggregators. However, there may be opportunities for solutions targeted at commercial loads, such as office buildings and retail stores, rather than industrial loads found in other regions. Looking beyond commercial and industrial loads, there are discussions in the Smart Grid industry about offering similar third party programs to residential consumers. The concept has been put forward that DR aggregation could be offered to the masses by large retailers (e.g. electronics stores and home improvement chains) who would act as middlemen between consumers and utilities and grid operators selling the cumulative load savings back to the grid. 55 Like C&I aggregators, retailers would make direct payments to participants, making the benefits more visible than are electric utility bill credits. While this business model is far from becoming a reality, it suggests a potential path for how significant residential contributions can be achieved while providing a tangible consumer benefit for products such as smart appliances. Given Hawaii s relatively low number of large industrial and commercial loads, this model may have broader application in Hawaii than the traditional third party aggregation model Load Shifting and Scheduled Load Management The goal of most DR programs is only to reduce system peaks during critical situations. Events are thus initiated sparingly, perhaps only several times per year, to address these emergency situations, and overall energy consumption is not necessarily reduced. For example, temporarily turning off a water heater with direct load control will defer energy consumption but most of the saved energy will inevitably be needed to return the water to the desired temperature. This type of DR is known as load shifting and is very effective at moving demand away from system peaks. Scheduled load management takes this concept of moving demand from one part of the day to another and applies a regular schedule to it, providing permanent changes in demand profiles. Examples of shifting methods include thermal energy storage (TES) solutions such as residential electric heat storage and commercial ice storage. Residential electric heat storage switches around the typical winter heating cycle instead of consuming power during the cold late night and early morning hours, heat storage systems heat up thermal bricks during the day and then release that heat overnight to maintain a desirable building temperature. Conversely, commercial ice storage moves cooling loads from the day to the night. Using roof top refrigeration units, ice is built up over night and is then used to cool buildings during the day. As the name implies, ice storage solutions are currently limited to commercial and industrial settings such as office facilities, retail complexes 54 EnerNoc, Inc: utilities/program implementation 55 demand response Catch 22 and how to fix it

92 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII or hotels; TES, on the other hand, is a viable residential option, although it has limited applicability in the warm climate of the Hawaiian Islands. Scheduled load management solutions should be considered as one method for adapting demand to the inherent but often predicable intermittency of renewable energy resources. These forms of DR fit well into the renewable energy equation since they can be used to schedule large, predictable loads when renewable resources are plentiful. TES, for example, can be used very effectively in the winter months in combination with solar PV since the sun rises and falls on a very reliable schedule. Likewise, commercial ice storage complements off peak wind generation very well. By shifting air conditioning load from on peak to off peak periods, ice storage could be designed to consume offpeak wind generation during periods when wind is often curtailed today Rate Design Programs Another time tested method for shifting demand is the use of dynamic rate programs which are designed to encourage customers to shift consumption away from peaks and other periods when the grid is stressed. While they fall into the category of non dispatchable demand response since the utility has no direct control over loads, well structured rate design programs can very successfully change customer usage patterns, thereby lowering distribution and supply costs and increasing grid reliability. The overarching principal is to raise the price of electricity when demand is higher than desired and to make the price more attractive when supply is greater than demand. In essence, dynamic pricing mechanisms can better reflect the cost of supplying electricity than the normal flat rate pricing model, since the supply chain costs are themselves dynamic. There have been a lot of creative rate models developed and there is much debate over which ones are most effective, but there are four main types which have been implemented at scale or at a minimum widely piloted: time of use (TOU), real time pricing (RTP), critical peak pricing (CPP), and critical peak time rebate (CPR). All four of these design types have been repeatedly shown to be effective at changing customer usage patterns. The following descriptions are far from exhaustive, but rather, are intended to provide basic design information for each of these four key types. TOU plans assign different rates to different parts of the day, utilizing two or more rates. The most basic TOU plans simply split each day into on peak and off peak rate periods. More complex TOU plans have intermediate tiers, known as shoulder, mid peak, or partial peak rates (or other similarly descriptive names), with the day split into three or four periods and some programs breaking up the day even further. Mid day TOU rates are typically the highest, especially in the summer when the goal is reduce cooling loads. Also, since weekday work and school patterns often cause peaks in the morning and late afternoon hours, TOU plans tend to focus on weekdays, assigning lower rates to weekends and holidays. Finally, some TOU programs utilize year round schedules while others may have seasonal calendars (e.g. summer vs. winter). Regardless of design details, the goal of TOU programs is to get consumers to defer household activities such as laundry 5 6

93 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII and dishwashing to off peak periods with the tangible benefit of a lower electricity bill. TOU plans are popular because they are easy to understand and have a regular schedule on which consumers can rely, thus allowing the use of simple timer controls to shift load. TOU programs do not directly add flexibility to address the intermittent aspects of renewable generation but in as much as they can shift demand to periods where there may be high amounts of wind generation, they could provide realizable benefits. RTP programs adjust the price of electricity on an on going basis. Prices may change daily, hourly, every fifteen minutes, or even minute by minute in extreme cases. More than other pricing designs, these programs are intended to reflect the real fluctuations in electricity costs and the changes in the supply demand balance. While RTP programs are simple to understand conceptually, they can be the most challenging for consumers to manage since they must continuously remain informed of electricity rates in order to appropriately adjust their usage. Consequently, RTP programs are more susceptible to customer burnout and are more complex to implement since they require a mechanism to keep customers informed of rates on a timely basis. RTP also generally requires a mature competitive marketplace to produce a market price. In Hawaii, this does not currently exist and may not be practical given the construction and dynamics of the energy marketplace. CPP rate programs are the pricing equivalent of DLC since events occur infrequently and are designed to reduce load only during critical system peaks (for example on exceptionally hot summer days). These programs offer an off peak rate which is lower than the standard flat rate in exchange for extremely high peak rates during critical hours. For example, an off peak rate may be less than $0.10/kWh while the critical peak rate may be more than five times higher at $.50/kWh. However, like many DLC programs, the number of events is usually capped at several times per year in order to not put unreasonable burden on participants. Finally, customers are typically informed of peak days one day in advance so communication mechanisms such as or text notifications must be put in place. CPR rate plans are among the newest rate design and have the same narrow focus on critical peak periods as CPP plans. However, rather than penalizing customers for high usage during system peaks, CPR encourages energy savings by rewarding customers for the amount of energy consumption they reduce during critical periods. As with CPP events, customers are typically informed of peak days one day in advance, requiring communication mechanisms to be in place, and the number of events is usually capped at several times per year. The key difference for customers is that CPR programs have no downside since not taking action during peak events simply results in customers receiving their standard flat rate with no additional savings. These programs are, however, susceptible to free rider issues in that customers could get rebates for taking actions that they would have taken anyway. Some utility programs have utilized a combination of these primary types. For instance, TOU plans can be combined with either CPP or CPR events to encourage additional savings during peak system 5 7

94 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII periods. There have been at least one hundred well documented dynamic pricing pilots with each one demonstrating that dynamic pricing can be very effective at shifting demand. While individual program results can vary widely, the following figures provide a good comparison of the four key pricing program types and illustrate the magnitude of peak savings which can be achieved through dynamic pricing programs: 56 TOU 5% peak reduction RTP 12% peak reduction CPR 12% peak reduction CPP 16% peak reduction These figures are based on an exhaustive study by VaasaETT, an energy think tank organization, which examined the results of dozens of pilots conducted around the world. The study compared load reduction percentages during peak rate hours to those in standard rate periods, and in total included 158,000 participants across 340 sample groups, with over half of the sample groups coming from U.S. based pilots. While the figures clearly show peak load reduction in all cases, it is important to note that although TOU programs show the lowest reductions, they, along with RTP programs, have the advantage of providing daily reductions as opposed to CPP and CPR, which only produce reductions during critical period events. Therefore, the relative impact on total consumption and consumer bills could be quite different between the two groups and the peak reduction figures for TOU and RTP are not indicative of total energy saved. Furthermore, the results seen in the pilots have been largely independent of time of year, challenging the belief that impacts will be greatest in the winter in cold climates and greatest in the summer in warm climates. 57 In the U.S., the deregulated retail energy provider (REP) setting in Texas should be considered a future source of consumer behavior data since consumers there can pick from more than 250 real programs, not pilots: While studies are still on going, the programs, governed by ERCOT (the Energy Reliability Council of Texas), should provide a wealth of information on what type of programs are most attractive to customers. A recurring theme is that variable pricing is only effective if mechanisms are provided which give consumers time to react text messaging, , in home displays or other communication methods are needed to make consumers aware of rates (communications mechanisms are described in further detail in section 5.5). This is a major advantage for TOU programs since the repeating schedules do not need to be continuously communicated. In addition, studies have shown significant differences when pricing programs are combined with automation, such as appliances which are price/signal responsive. For example, a pilot conducted by Oklahoma Gas & 56 Jessica Stromback, Christophe Dramacque, Mazin Yassin; The potential of smart meter enabled programs to increase energy and systems efficiency: a mass pilot comparison, VassaETT, Ibid. 5 8

95 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Electric (OG&E) showed that peak reduction percentage roughly doubled when TOU pricing was combined with smart thermostats. 58 Along with other load shifting demand response solutions, variable rate design programs should be investigated for adjusting demand patterns to better accommodate the integration of renewable energy resources Load Shaping Load Shaping is the newest technique with the fewest commercially viable technology options. However, it is worthy of discussion because of its potentially large benefits for renewable energy integration. Load shaping is the fine tuning of system load curves by utilities. Using advanced sensors and software controls, supply and demand are dynamically balanced in real time by making adjustments to energy usage as conditions change (e.g. weather changes). Demand is thus kept in check by quickly responding to fluctuations in supply, such as those caused by variable energy resources. This next generation of DR, Demand Response 2.0, is the Smart Grid counterpart of static load following the practice of bringing on additional generation to meet moment to moment demand in the system, and/or keeping generating facilities informed of load requirements to ensure that generators are producing neither too little nor too much energy to supply the utility's customers. 59 However, load shaping occurs on a much more dynamic level. Using load following generation, supply is constantly chasing demand. With load shaping, demand is continuously adjusted to accommodate supply based on multiple variables, responding very quickly to utility signals, making it ideal for dampening the supply peaks and valleys of renewables such as wind and solar. So, although load shaping does not reduce overall consumption, it is very relevant for Hawaii and improving the conditions for renewables integration. Load shedding and load shifting are proven demand management techniques with many available technology and vendor options. Meanwhile, real time load shaping is a relatively new concept and its complexities require smart meters and advanced software and analytics to tie together weather conditions, energy supplies, grid conditions, power quality, and dispatchable loads to make realtime DR decisions. Because of this need for real time data processing and an interconnected network of intelligent electronic devices (IED s), loading shaping is in actuality an operational feature of an advanced Distribution Management System (DMS), covered in section 6. However, since it targets demand adjustments, load shaping is covered here for completeness. While the technical requirements and complexities raise the barrier to entry, load shaping could play a valuable role in maintaining system reliability and should become increasingly relevant as renewables reach higher levels of penetration, making up a larger percentage of total power OG&E, Mike Farrell, 2010 Demand Response Study Interim Results & Lessons Learned, DistribuTech EnergyVortex.com, Energy Dictionary 60 Smart Grid News, Smart Grid Demand Response: Why today s leaders are at risk as DR 2.0 emerges, April 27,

96 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 5.2 DEMAND RESPONSE FOR FLEXIBILITY Balancing the influx of new uncontrollable energy requires dispatchable resources on the system to ensure reliability. The electrical grid will need the ability to rapidly respond to a wide range of load and renewable energy production conditions. Flexible capacity can be provided from traditional generating resources or from demand responsive loads. However, to date, demand response development and the growth of variable generation capacity have largely proceeded in parallel but without much consideration of better connecting the two. The North American Electrical Reliability Corporation (NERC) and others have identified DR as an additional source of electric system flexibility that could aid in integrating variable generation. Effective DR programs can provide essential flexibility over relatively short timeframes when an unpredictable change in renewable generation output occurs. DR has been shown in some balancing areas to be a flexible tool for operators to use with wind generation. For example, in 2008, ERCOT, which operates the electric grid in Texas, used DR to restore system frequency and thereby demonstrated the effectiveness of using DR to enhance system flexibility. At high penetrations of solar and wind energy, Hawaii s grids will need to carry higher levels of operating reserves to account for the variability of these renewable resources. This provides a good opportunity for Hawaii s electric utilities to look at DR technologies as reserves to cover a portion of the operating reserves requirement. DR solutions may be more cost effective, as well as more environmentally friendly, than traditional reserve resources. While DR programs have been used to provide planning reserve margins, operators are beginning to realize the potential for dispatchable load DR programs to be used as operating reserves. Integrated with a DMS, which should be responsible for managing all generation and reserve types, DR becomes a valuable component of the total supply and demand equation. Response (i.e. ramp rates) can be quicker than generation for most responsive loads. The response can be automated as well, and different loads can be set to trip at different frequencies, thus providing a frequency droop curve that simulates generator governor response. Also, DR acts as a contingency reserve by always being ready, but with events happening infrequently and actual deployment being relatively short in duration, compared to building a new generating resource. In addition, system operators are now using DR more frequently to provide ancillary services, such as contingency reserves and supplemental reserves. While it should be noted that use of DR for ancillary services is a recent development and represents only a small amount of demand response, it is indicative of the potential of DR to address electric grid reliability issues. Overall usage of DR as a resource increased from 30 GW to 43 GW between 2010 and 2011 for all NERC regions combined. However, some regions consider DR for supplying reserves and some do not. ERCOT, for example, currently obtains half of its responsive (spinning) reserves (1,150 MW of the 2,300 MW total) through a Demand Response product called Loads Acting As Resources (LAARs). The program s customers have under frequency relays set to 59.7 Hz so that their 5 10

97 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII participating loads automatically trip off line during under frequency events. During emergency conditions, these loads will also disconnect upon receiving instructions from ERCOT. ERCOT also has an Emergency Interruptible Load Service (EILS), a separate program of loads that will separate from the system during emergency conditions upon receiving instructions from ERCOT. Hawaii s utilities should consider DR that supplies reserves similar to ERCOT. However, due to the ongoing variability of renewable resources, it is important to target loads which can be frequently interrupted with little to no customer impact to minimize participant burnout. 5.3 HAWAII S EXISTING DEMAND RESPONSE PROGRAMS This section will summarize residential and commercial DR programs currently in place in Hawaii and the most recent program developments. The Hawaii Energy Conservation and Efficiency Program (Hawaii Energy) is a program operated by SAIC, an independent third party contractor for the Hawaii Public Utilities Commission. The program includes four programs for the C&I sector and three for residential customers. 61 While this program is well organized and very well tracked (see section 4 for specific results), its focus is exclusively on energy efficiency and does not include any documented DR features. There are, however, meaningful DR pilot programs in progress on the islands by the Hawaiian Electric Company, which provides approximately 95 percent of electric power in the state. Through HECO, MECO, & HELCO, the program strategy is to identify residential, commercial and industrial customers who are able to reduce electricity use to help maintain grid reliability and increase the presence of renewable energy. The DR plan is being implemented in phases through a combination of pilot programs; research, development, and demonstration projects; and market studies. 62 For residential customers, the program, marketed as EnergyScout, offers traditional DLC (direct load control) for water heaters and air conditioning (AC). This program is an effective starting point given that a significant portion of residential loads can come from water heaters. Currently in the pilot stage, the program is fully subscribed but is looking to expand in Approximately 34,000 customers have been outfitted with water heater controls, providing operators with approximately 15 MW of controllable peak demand. In addition, controls on the AC units of approximately 4000 participants represent another approximately 2.5 MW, for a total of approximately 17.5 MW of controllable load. Hawaii Electric Company s pilot DR program offerings for commercial and industrial customers are more varied, which is typical given the greater cost/benefit which comes from C&I DR (see section regarding the relative impacts of C&I DR). Hawaiian Electric Company currently offers three

98 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII C&I programs AutoDR, Semi AutoDR, and Energy Scout for Business, and is awaiting approval of a fourth program, C&I Dynamic Pricing (CIDP) Pilot Program. Customers enrolled in AutoDR, also known as Automatic Fast DR, and Semi AutoDR, or Manual Fast DR, are engaged to proactively reduce their electrical consumption and are rewarded with financial incentives in return. For customers who qualify, Hawaiian Electric Company places automatic (AutoDR) or semi automatic (Semi AutoDR) controllers on non essential equipment. The key element is that the loads must respond in 10 minutes or less. These programs are extremely new and thus there are no published results, but several major customers have enrolled. Launched in February 2012; the first AutoDR customer, an office complex in Honolulu, started in the fall of When combined with a customer side EMS (energy management system), power consumption can be automatically and temporarily reduced using Automatic Fast DR. For example, the EMS could temporarily adjust building AC temperatures when triggered by a utility signal. While Fast DR requires some automation; customers always maintain complete control and can opt out of having their demand impacted at any time. Semi AutoDR is very similar to the AutoDR program but relies on manual controls. Under the programs, in addition to incentives paid during DR events, businesses automatically receive a bill credit of $3,000/year for every 50 KW of power they are willing and able to turn off within 10 minutes of being signaled. The EnergyScout for Business program is the commercial version of the EnergyScout residential DLC pilot. With approximately 43 customers enrolled, the program is providing approximately 18.2 MW of controllable load, eclipsing the 17.5 MW provided by the approximately 38,000 involved on the residential side. Like its residential counterpart, EnergyScout for Business is fully subscribed and newly interested customers are being moved to the new AutoDR programs. The application for the C&I Dynamic Pricing (CIDP) Pilot Program was submitted to the Hawaii Public Utilities Commission on December 29, If approved, the CIDP Pilot Program could start as early as The proposed rate design program is intended to evaluate dynamic pricing options offered to participating commercial and industrial customers for a two year period. As described in above, dynamic pricing allows customers to respond to the changing cost of electricity by adjusting their demand, especially during periods of limited supply. The proposed program would reduce the demand charge for program participants in return for their lowered energy use at certain times (and sometimes on short notice). As the Hawaiian electric companies move toward a grid composed of higher levels of intermittent renewable generation, these programs can provide an additional resource to reduce electricity demand until additional generating units are brought online. 5 12

99 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 5.4 OPPORTUNITIES TO ENHANCE HAWAII S DEMAND RESPONSE PROGRAMS The DR programs described in section 5.3 are very promising and indicate that with expanded participation the overall grid reliability impacts could be substantial. It is also clear that the curtailment benefits resulting from the commercial and industrial programs have a far greater incremental impact per customer than those of residential programs. Hawaii s existing DR programs should be evaluated for cost effectiveness to determine how best to prioritize future investments and resources. In parallel, load research data collection and market assessments are needed to allow utilities to better estimate the demand response potential in Hawaii and understand the overall role which DR could play. A good example of such research is the DSM Market Potential Assessment commissioned by Kauai Island Utility Cooperative (KIUC). Being fairly new, none of the assessment s recommendations have been implemented, but this study provides a good picture of the magnitude of peak demand savings which DR, combined with other demand side initiatives, could provide as well as estimated costs. 63 For DR to play a larger role, however, barriers in the regulatory environment must be reduced and business conditions must be favorable for all parties involved. Two specific areas which should be addressed are the pace of the regulatory process and the role of third party aggregators. Speeding up the regulatory process will provide a quicker first hand understanding of DR benefits and challenges. Utilities need to be able to further develop and test a range of value propositions to assess customer interest in direct load control and in pricing event strategies to support renewable generation. For example, the C&I Dynamic Pricing (CIDP) Pilot Program which was submitted to the Hawaii Public Utilities Commission on December 29, 2011 remains in the approval cycle. Expedited regulatory review and approval is needed to get such programs in place as soon as possible. In addition, improving the regulatory process could allow expansion of successful programs such as those already being piloted. As described in section and as is clear even with the limited sample size of the existing Hawaiian Electric Company DR pilots, C&I focused DR programs can be extremely effective at providing grid scale benefits. Therefore regulators and utilities should look for opportunities to expand these programs and to introduce similar ones. In parallel, regulators also need to consider the appropriate role of third party aggregators in delivering demand response programs to allow DR to compete on equal footing with supply side options. While Hawaii may lack the volume of large industrial scale loads needed to support a traditional third party aggregator market, policies pdf 5 13

100 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII and alternative technology solutions better suited for commercial loads, which are more common in Hawaii, should be pursued. The Demand Side Options Subgroup of the Hawaii PUC s Reliability Standards Working Group (RSWG) has made similar recommendations, including ensuring that DR programs are considered in the integrated resource planning process. 64 DR technology is widely available with proven, positive impacts, but regulators must act on recommendations in order to see progress Measurement and Verification of Demand Response Since DR is still relatively new as a resource, stakeholders need to measure its performance in order to gauge its benefits and impacts on reliability. Evaluating the performance of DR resources begins with and depends on the measurement and verification (M&V) methodologies for determining demand reduction quantities. For example, the methodology used to calculate and estimate the baseline will play a critical role in measuring performance. There are various baseline methodologies and some are more appropriate than others depending on the type of DR resource, the frequency and duration of events, and the timing of notification, as well as other factors. These differences have often made comparisons difficult since programs may not be compared in an apples to apples manner. From a practical perspective, data quality is critical regardless of the M&V methodology used. Smart Meter AMI systems in combination with Meter Data Management Systems (MDMS) offer the potential to use reliable 15 minute interval data to support M&V activities while also providing a robust and valid baseline. While 60 minute interval data is standard for residential smart metering, 15 minute data is standard for C&I metering and could be applied selectively on residential DR participants. By processing metering data after events occur, an MDMS can provide detailed DR event results. DMS, on the other hand, should be used for M&V analysis during events (e.g. tracking real time participation) since it has a view of potential resources. DMS, MDMS, and AMI systems are discussed in more detail in section 6. New standards for measurement and verification of results for DR programs are under development to simplify comparisons and assessments. In recent years, NERC, which has been at the forefront of developing performance evaluation criteria, has been developing a more systematic and uniform approach to collecting and quantifying DR performance. In addition to specifying a consistent and timely basis for counting and validating DR contributions, a major accomplishment thus far has been the establishment of the Demand Response Availability Data System (DADS). While it is still in the early stages, the goal of DADS is to collect DR enrollment and event 64 Reliability Standards Working Group, Hawaii PUC Docket No , Demand Response as a Flexible Operation Resource, RSWG Demand Side Options Subgroup 5 14

101 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII information to measure its actual performance. As the database becomes more populated, DADS will be able to provide the industry with a more reliable basis for projecting DR contributions. 65 Lawrence Berkley National Laboratory (LBNL) has been another key source of M&V methodology research. LBNL is seeking to develop lower cost, standardized M&V methods that are acceptable to system operators and load aggregators. 66 LBNL recently evaluated existing M&V processes for both settlement and impact estimation. 67 Settlement is the process used to calculate the reductions achieved by individual program or market participants, and to determine the corresponding payments or penalties owed to or from each participant. Settlement has major financial impacts since inaccurate processes can result in over or under payments, affecting program costs and participation levels. Impact estimation M&V is used to determine projected program level demand reductions. Impact estimation is immediately applicable to Hawaii since these methods are used to both valuate existing programs as well as to forecast impacts in planning processes. As part of their research, LBNL outlined the complex inter relationships between M&V methods, DR program design & rules, program operating conditions, load characteristics, and participant types: DR performance evaluation methods and results affect and are affected by many aspects of program planning, design, and operations. 68 For example, the M&V method used affects both M&V results and accuracy, but different participant types have different load characteristics, which in turn affect M&V accuracy. Thus, program design and M&V method are tightly intertwined. The key is that M&V cannot be an afterthought it must be part of the initial DR program design. To account for both load characteristics and program design, we recommend the consideration of several factors when designing and assessing DR M&V methods: 69 Weather sensitivity of loads, Seasonal sensitivity of loads, Customer type should be based on observable load characteristics and broad revenue class, rather than reported business category or customer segment, Variability of loads, especially for those which are highly variable, The effect of notification timing (e.g. day ahead notification), Gaming opportunities measures should be taken to limit baseline manipulation by participants, and Baseline methods should be assessed based on load simulation Overall, this is a very simplified explanation of the work by LBNL and others. They key point is that existing and future DR programs should carefully review M&V methodologies to use the most and verification protocols demand response programs 67 Goldberg, M, and Agnew, K., Measurement and Verification for Demand Response, LBNL, Goldberg, M, and Agnew, K., Measurement and Verification for Demand Response, LBNL, Ibid. 5 15

102 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII appropriate ones with respect to the specific conditions in Hawaii and to ensure industry best practices are being applied Evaluating the cost effectiveness of DR Programs Cost effectiveness is critical to DR program approval, program funding, and getting businesses to participate. So, using M&V calculations of load reductions, DR performance must utilize M&V to calculate the savings from load reductions and then measure that that against the cost of implementation to understand whether or not the benefits of the program outweigh its costs. As described above, the determination of cost effectiveness is very much affected by the type of M&V method used. 70 LBNL also provides a key source of guidance for evaluating DR program cost effectiveness. In a recently published work for the National Forum on the National Action Plan on Demand Response, LBNL provided a framework which should be used to evaluate the cost effectiveness of DR programs, detailing what key costs and benefits need to be accounted for in evaluations. 71 The basis of the LBNL framework is the California Standard Practice Manual which has become the industry standard for evaluating energy efficiency programs. The Manual defines five cost effectiveness tests and how to apply them. The five tests cover: societal costs & benefits, total resource costs & benefits, program administrator costs & benefits, participant costs & benefits, and the impact on utility rates. The LBNL DR framework adapts the successful EE cost effectiveness framework with key modification to account for the unique considerations of DR. The DR framework outlines eleven program cost categories and twelve program benefit categories which should be accounted for when evaluating cost effectiveness. While support for renewable energy is not listed as a specific benefit of DR, a number of the benefits do have direct relevance such as improved electrical system reliability, avoided ancillary service costs, and avoided environmental compliance costs. 72 As stated previously, DR programs which focus on commercial & industrial customers are likely to provide the largest incremental benefits. However, rather than high level calculations and anecdotal assumptions, the cost effectiveness of Hawaii s existing DR programs and programs under consideration should be critically evaluated using the latest cost effectiveness methodologies to determine next steps, to screen opportunities, and to prioritize future initiatives. 70 Ibid. 71 Woolf, T, Malone, E, Schwartz, L, and Shenot, J. A Framework for Evaluating the Cost effectiveness of Demand Response, Ibid. 5 16

103 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 5.5 CUSTOMER INTERACTIONS, COMMUNICATION TECHNIQUES AND ADVANCEMENTS IN LOAD CONTROL Getting consumers to actively participate can sometimes be the biggest challenge for Demand Response. DLC devices provide a reliable method to connect utilities with individual load sources, but it is also important to connect with consumers themselves to engage them in the energy process. At the same time, consumers do not want to be inundated with DR related messages; DR must be used at a level which does not seem burdensome. For example, DR could be initiated so often that people would drop out of the programs. In addition to being balanced with other Smart Grid reliability technologies, therefore, DR programs and communication strategies must be designed to yield long term consumer participation and engagement. Various methods have been used over the years to provide DR event notifications to program participants and to provide updated rate information for rate based DR programs. Tools for communicating DR related information have evolved from voice messaging, to and text messaging, and most recently to utility specific devices such as In Home Displays (IHD). IHD s provide customers with information on their energy usage and pricing and may also provide messaging capability for utilities to send alerts and notices. There are a significant number of IHD options on the market with varying levels of functionality, from simple text displays to full color touch screens with real time graphs. While they are potentially an effective means of communication, however, they are also relatively costly and utilities are thus trending away from dedicated devices in favor of web portals and mobile applications. In addition to being more cost efficient, web and mobile applications better suit today s consumer. The traditional relationship between utilities and consumers has been very static, with contact often limited to connecting and disconnecting service and bill paying. However, the modern consumer has changed significantly. Spurred by the proliferation of the internet, smart phones and applications such as Facebook and Twitter, today s consumers live in a world of real time information and instant gratification and are beginning to expect this type of interaction from everyone with whom they do business. Survey findings point to a growing segment of consumers who are interested in technology solutions to monitor and manage their usage, with approximately one third of consumers interested in seeing their personal usage in real time using a mobile application. 73 As a result, the electric industry is seeing a wave of efforts to increase customer access and standardize the format of energy usage information. This movement has fostered the rapid development of new applications to increase consumer engagement and awareness. One of the latest and most widely supported examples is the Green Button. The Green Button Initiative is a White House backed voluntary effort for utilities to provide 73 Accenture, Revealing the Values of the New Energy Consumer: Accenture end consumer observatory on electricity management,

104 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII retail electricity customers with easily accessible and up to date data on their electricity usage. With the Green Button, consumers throughout the country will be able to securely access their own energy usage information in a user friendly, easy to understand format. The goal is to provide consumers with an easy way to make more informed energy decisions. For example, consumers could use Green Button applications to optimize the size and cost effectiveness of solar panels modules for their home or verify that energy efficiency investments are performing as expected. Consumers can even use innovative applications to compete against Facebook friends to save energy and lower their carbon emissions. 74 Since its launch in January 2012, 34 utilities have implemented or committed to participate, and at least 48 technology companies have already released or committed to providing Green Button solutions including Web and smartphone applications and services. 75 Green Button aside, there are a multitude of utility specific web portals and smart phone applications on the market with a wide variety of capabilities and functionality. While some applications provide the same basic usage and pricing information as IHD s, others allow customers to respond to utility messages and events, opt out of a demand response event, or view their savings from participating in an event. Some applications even support device configuration and control (e.g. smart thermostat programming). Another great product to come out of the movement away from solution specific hardware devices such as IHD s is the USB plug in dongle. Similar in size to a USB memory stick, these inexpensive devices plug into a home computer and communicate directly with smart meters equipped with ZigBee Smart Energy wireless capability. Using built in software, the simple USB devices allow homeowners to see their energy consumption information in real time, on their own home computer. Beyond consumer engagement techniques, an area of development for DR technology is the method used to communicate to in home devices. Current Smart Grid implementations most commonly utilize smart meters via AMI networks to connect to the devices in the home. However, strong opportunities exist to connect utilities to the HAN without depending on an AMI system or Smart Meters. According to Internet World Stats, 82.6 percent of Hawaii s population was connected to the Internet as of June 2010, compared to a national average of 78.1% (as of June 2012). 76, 77 Consequently, there is a lot of interest in moving from Smart Meter based ZigBee HAN solutions to internet based WiFi solutions, driven by the fact that WiFi is common in homes and ZigBee is arguably a niche home automation solution. Rather than using the meter as an entry point, IP gateway solutions typically consist of a device connected to the consumer s broadband Internet or with an imbedded cellular modem. In these cases, the HAN gateway can still connect to the Smart Meter via ZigBee (or other commonly supported communication method), but the meter does not serve as the gateway into the home or as the HAN controller. In this configuration, the Smart Meter Internet World Stats, 77 At the time of writing, the most recent state level statistics were from June 2010; national level numbers were updated June

105 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII simply provides energy usage information to the HAN. All other data, including that for control of DR devices, travels directly to the HAN via the internet connection. This type of solution could apply to any of the DR methods outlined in section 5.1. In fact, while there has been a general decrease in the popularity of application specific hardware such as IHD s, interest in PCT s (and other DLC devices) remains popular since customers already need thermostats and appliances in their homes. Finally, a tremendous amount of growth is being seen in the area of electric battery storage. Since it supports grid reliability in general, with DR being just one of the instances where it may be called upon, energy storage will be discussed in further detail in section but is mentioned here for completeness. Energy storage does not directly reduce load but is viewed nonetheless as a key component in the renewables equation since it shifts demand patterns to better align with the generation patterns of variable resources such as PV solar and wind. Energy storage can also buffer against rapid drops of load or supply. Although energy storage is still relatively expensive compared to traditional DR methods and currently exists at low penetration levels, the combination of a significant number of energy storage providers and rapidly evolving battery technologies will lead to inevitable decreases in solution costs. Storage solutions should therefore be monitored closely when developing a comprehensive strategy for increasing renewable energy penetration. 5.6 SECTION SUMMARY Demand Response penetration is increasing significantly due to technology innovations and policy directives. These advances have made it both technically feasible and economically reasonable for consumers to respond to signals from a utility system operator, load serving entity, RTO/ISO, or other DR provider to make the grid more reliable. There is a key difference between DR for traditional peak shaving and DR for supporting renewable penetration. For peak shaving, based on day ahead or hours ahead forecasts, the decision to call DR events is made either via pricing signals or other utility signals; but for DR to respond to renewable generation variability, it needs to be faster acting and is therefore more akin to DR for reliability. This distinction increases the importance of integrating DR systems with technologies such as AMI, advanced DMS, and Distribution Automation (DA). Linked by an interconnected network of intelligent devices and utilizing real time data processing, utilities can take advantage of the synergies between these various technologies to optimize the role of DR. As outlined in this section, DR technologies come in many different flavors and DR can and should play an important role in the integration of the renewable energy resources. For example, such programs as time of use rates, which shift load to off peak periods when wind production is usually at its highest, could be used to avoid curtailment of variable generation or minimize cycling of large base load power plants. However, DR is not one size fits all and any demand response program should include a portfolio of options designed to match customer comfort zones to maximize adoption rates. Therefore, DR needs to be balanced and integrated with other Smart Grid 5 19

106 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII technologies to provide holistic and sustainable grid reliability solutions, and regulatory policies must accommodate the variety of viable options. 5 20

107 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 6.0 The Smart Grid Monitoring, Communications and Control Technologies This section will identify and assess existing and emerging monitoring and control technologies that can assist in the integration of renewable energy into the grid. Intermittent Distributed Generation (DG) can cause voltage and frequency irregularities, imbalances in reactive power, transient voltage variations, and harmonic distortions to power quality. As we continue to move from a world of centralized power generation and resource management to one of decentralized intermittent energy resources, enabling flexibility to address these impacts will be critical. Smart Grid technology provides system operators with real time continuous awareness of electrical generation, storage, transmission, distribution, and loads; this information is essential in order to maximize the integration of renewable resources onto the grid. According to the Energy Information Administration's (EIA) latest figures, as of the end of 2012, there were nine states where wind is generating at least 10 percent of the available capacity, with Iowa and South Dakota at well over 20 percent, and two others on the verge of breaking 10 percent. 78 Wind energy has also become the leading source of new generating capacity in the U.S., accounting for over 55 percent of new capacity in While this trend is positive in many respects, without proper planning and technology upgrades, the end result could be grid instability. Take the case of California, for example, which rather than having a capacity problem is on the verge of having a flexibility problem. 80 The state has very successfully ramped up renewable generation and is on pace to exceed its 2020 goal of 33 percent and overall has a capacity surplus. However, the rapid increase in renewable energy resources could cause significant reliability issues to the grid as a whole since such a great portion of the supply mix is intermittent, fluctuating with weather patterns. To date, California has not had problems but it is a significant issue of debate and the CPUC, California utilities and other stakeholders are voicing concern and attention to the risk. To prevent such issues and keep pace with the changing energy supply mix, California s utilities have had to undertake major Smart Grid infrastructure initiatives and will need to continue to do so to maintain reliability. Germany provides another valuable case study since their high penetration of wind energy is already causing notable grid instability. The German government is targeting 35 percent of its electricity to come from renewable resources by 2020, with an even more ambitious goal of 80 percent by Certainly, no traditional grid can handle this level of variable power supply 78 U.S. Energy Information Administration, Annual Electric Generator Report 79 power generation awea united states aspx 80 Rebecca Smith, Wall Street Journal, California Girds for Electricity Woes,

108 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII without significant reliability upgrades. Stakeholders are beginning to recognize the opportunity afforded by Smart Grid technology to automate electricity distribution by tying together electricity generation and consumers through additional sensors, instrumentation and control systems, and communication and data infrastructure. 81 Two valuable lessons learned by Germany which are relevant to Hawaii s renewable energy discussion are 1) the importance of the establishing an extensive communication network parallel to existing and planned energy grids, and 2) ensuring that new distributed generation resources (e.g. wind or PV) are approved as grid compatible before they can feed energy into the grid. As discussed further in this section, advanced communication networks are necessary to exchange information between various systems and devices and trigger appropriate response mechanisms to maintain grid reliability. On the administrative side, Germany requires DG owners to provide proof of compatibility before resources can go on line and utilizes accredited third party certification organizations to review design documentation. 6.1 HOW DOES THE SMART GRID ENABLE The deployment of Smart Grid technologies has surged to become one of the most significant areas of utility capital investment. Related projects include transmission upgrades, substation automation, distribution automation, information technology and smart metering. Rather than being a single definable end state, the Smart Grid represents a family of technologies and programs that increase reliability and provide flexibility functions. The Smart Grid vision is to remove barriers between transmission and distribution operations, improve communications, and enhance back office systems to create an integrated network that provides diagnosis and resolution of problems and enables real time information exchange throughout the utility and with its customers. Smart Grid technologies provide communications and automated control of devices associated with electrical generation, storage, transmission, distribution and loads. The result is an integrated network which provides utilities with visibility and control to improve the efficiency, reliability, and quality of the electrical energy provided. Smart Grid technology is evolving, even as utilities are implementing their initial demonstrations or projects. And it will continue to evolve as utilities reach higher levels of implementation. This is likely to be a continuous evolution with more technology and higher levels of integration occurring at each level of implementation and an increasing number of vendors entering the market. An important factor to note is that there is no one size fits all Smart Grid solution the terms smart and Smart Grid have different meanings around the world depending on local electric system conditions and starting points. However, while a significant milestone will be achieved when utilities have visibility of all generation sources and loads and are able to operate or cause operations that maintain a reliable system, many utilities around the globe view the first step as the installation of smart meters to enhance utility operations and to provide customers with an 81 Dieter Rosenwirth and Kai Strubbe, Renewable Energy World, Integrating Variable Renewables as Germany Expands Its Grid,

109 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII accurate picture of when and how power is being consumed. Future steps will involve use of realtime information about incoming loads, as well as automated systems which monitor and adjust system parameters to allow for more stable introduction of intermittent sources of power. There are a number of technology solutions which fall under the Smart Grid umbrella as follows. Smart Inverters Substation monitoring and control Voltage Management o Voltage monitoring & analytics o Integrated volt/var optimization (VVO) o Conservation voltage reduction (CVR) o Static VAr (SVAR) technology & distributed static VAr compensation (DSVC) Remote monitoring and control of distribution line devices o Capacitor banks o Reclosers Low voltage and low frequency ride through Ramp rate control Automatic load tap changing (LTC) Automated curtailment based on integrated forecasting Energy storage Distributed energy resource management systems (DERMS) Distribution management systems (DMS) Smart metering Smart appliances Energy & Building management systems (EMS & BMS) Outage management and restoration This list is not all inclusive but underscores the variety of technologies which are part of the Smart Grid. To put these concepts into more tangible terms, the following examples cover some, but not all, of the solutions listed to highlight some industry best practices followed by descriptions of a variety of renewable resource related initiatives being undertaken by leading utilities throughout the U.S. Smart Inverters As distributed generation and disruptive loads continue to increase in prevalence on the distribution grid, more granular and regular voltage regulation and VAr compensation will be required to maintain power quality throughout the distribution network. Most DG resources produce direct current (DC). To be integrated into the distribution network, this DC power is converted to AC power by inverters. Traditional inverters convert DG power from DC to AC current 6 3

110 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII with a power factor near unity to limit the effect of DG power on the voltage profile of the power traveling through the distribution grid. However, this equipment is installed and calibrated using model based calculations rather than any real time or empirical data. These devices were sufficient for maintaining voltage and power factor within limits when they were a small component of the power generation pool, but were by no means optimized on a system basis. Voltage conservation and volt/var optimization (VVO) are Smart Grid features which can be used to accommodate voltage fluctuations caused by intermittent renewables and thus maintain the desired voltage profile. A localized controller could activate and resolve a local imbalance, but it would do so without an understanding of what was happening elsewhere in the system with the consequence of creating suboptimal conditions locally or remotely. Smart Grid technologies allow big picture intelligence to be applied to such situations. With built in volt/var capabilities, smart inverters and transformers can ensure DC power from renewable resources is clean of harmonics and smart inverters of the future may provide many other smart advances as well. For example, smart inverters have the potential to allow PV generation to provide voltage support and improve response to under and over frequency events, but new inverter designs have even greater potential to reduce the impact of PV solar on the distribution grid. Recent EPRI reports estimate the cost of smart inverters at $ /Wac. 82 This is compared to typical market inverter prices around $0.20/Wac at the time of writing of this report. As there is significant active development in this area, it is likely that the cost effectiveness of static inverters will improve with time and field experience. Voltage Management Current generation capacitor banks, load tap changers, and voltage regulators were not designed to react on a minute by minute basis, as may be required to keep up with constant VER (variable energy resource) fluctuations. A new generation of power electronics based voltage control and reactive power compensation devices will be required to adjust voltages and reactive power compensation multiple times per minute on highly variable feeders. In addition to ensuring stable voltage conditions, protection systems are necessary to isolate DG in the event of faults. DG can impact the ability of the normal protection devices to function optimally. Utilities can leverage new low cost remote monitoring devices to detect reverse power flow, abnormal operating conditions and DG status. Distributed Energy Resource Management Systems 82 EPRI Estimating the Costs and Benefits of the Smart Grid: A Preliminary Estimate of the Investment Requirements and the Resultant Benefits of a Fully Functioning Smart Grid accessed online at the following URL on May 7, 2013: %20Estimating%20the%20Costs%20and%20Benefits%20of%20the%20Smart%20Grid.pdf 6 4

111 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII System planners and operators are very familiar with variability and uncertainty as they are constantly balancing traditional demand and generation fluctuations to maintain voltage and frequency. However, intermittent generation introduces additional uncertainty and variability which operators must manage. The variability of wind and PV generation requires special attention since conditions can change in the short term, from minute to minute based solely on wind speed or changing cloud cover, or from day to day, based on weather conditions. For years SCADA (supervisory control and data acquisition) has provided control & monitoring at the substation level but Smart Grid technology puts intelligence in the field, at individual devices. In addition, existing industry standards define advanced functions at the individual device level, lacking the more aggregated feeder level representations which are needed for enterprise level integration. Thus, tying all of these devices together in a coordinated fashion requires a strong communications backbone and advanced software technology. Without advanced communication and control, utilities cannot efficiently integrate all the pieces to make them work together as a holistic system. While Distribution Management Systems are discussed below in section and provide many Smart Grid advantages, the current generation of DMS does not specifically support integration of distributed energy resources (DER). In general, DER support is limited to monitoring the output of utility scale DG (i.e. greater than one megawatt), missing the smaller intermittent renewable generation resources which will be prevalent in residential settings. Consequently, the Electric Power Research Institute (EPRI) has been exploring various ways in which the DMS can utilize these distributed resources more effectively. 83 The conceptualized solution for managing all of the many diverse DER devices in a system is a Distributed Energy Resources Management System, or DERMS. A DERMS is an enterprise application which would provide the central control necessary to handle all of the smart DER devices throughout a distribution system by monitoring and managing the status and capabilities of each device and feeding that information into a DMS and other applications in a more useful form. Rather than having the DMS make sense of the individual device data, the DERMS could aggregate the information of individual devices, transforming their settings and effects so that they become attributes at the circuit, feeder, or segment level. 84 DERMS will thus improve the ability of operators to optimize the dispatch of distributed resources, DR, energy storage, and other Smart Grid features. For example, consider the existing standard functions of smart inverters, such as volt/var capability. With proper control and management, smart inverters and solid state transformers 83 Brian Seal, Robert Uluski, Integrating Smart Distributed Energy Resources with Distribution Management Systems, Electric Power Research Institute (EPRI), September Ibid. 6 5

112 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII could act as gatekeepers for DG resources by disconnecting the segment with local DG from the distribution grid during a fault. This would allow the DG supported island to continue to produce power for the loads on the customer side of that distribution secondary transformer. However, if the volt/var settings on a distribution circuit vary from one device to another, as they most certainly will in a diverse system, then managing voltage behavior becomes very complex. Without a DERMS, the complexities of such management fall directly to other applications such as DMS, which are not currently built for such tasks. In reality, DERMS functionality may or may not be dedicated software. DERMS functionality could be integrated as a module into a DMS or deployed as standalone application to work alongside a DMS. In either case, significant commercial development is still required, but DERMS concepts should nonetheless be considered in today s resource planning discussions. The following categories compiled by EPRI are examples of services which a DERMS could provide related to intermittent renewable generation: 85 Identify installed capability Report resource status Provide forecast/prediction of supply capacity Connect / disconnect DER from the Grid Provide volt/var regulation support Provide phase balancing Coordinate DER with circuit reconfigurations Provide maximum capacitive VAr support Provide support for conservation voltage reduction (CVR) Mode Outage Management & Restoration A related set of Smart Grid features which can directly facilitate higher levels of renewable energy deployment is outage management & restoration. Outage management systems (OMS) or the outage management module of Distribution Management Systems (DMS) and Fault Location Isolation and Service Restoration (FLISR) solutions are used to monitor, assess, and logically recover (automatically, when possible) from traditional power outages, such as those caused by storm damage. While these features alone make them very cost effective, OMS and/or FLISR can also help manage DG using the same self healing and proactive principles which are used to reconfigure a system during a traditional outage, restoring power to customers served by healthy sections of a feeder. 85 Ibid. 6 6

113 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII While DEMRS, and even DMS technology itself, is still in its infancy, the development and implementation of such systems and their combination with existing Smart Grid solutions such as OMS and FLISR will undoubtedly help overall system reliability and help optimize conditions for increased penetration intermittent renewable generation. SVAR & DSVC Static VAr (SVAR) compensation devices are common at the transmission level as an essential means to balance reactive power by dynamically producing reactive power in real time. At the distribution level, however, SVAR technology is still in development with a few pilot demonstrations. Combined with smart capacitor banks and inverters, smaller installations of distributed static VAr compensation (DSVC) can provide considerable increases in power quality and facilitate far greater injections of intermittent renewable power. Some examples of how utilities are beginning to use Smart Grid technology to enable higher penetrations of renewables are as follows. Arizona Public Service Company (APS) serves 1 million customers in Arizona. It is the second fastest growing electric utility in the country and has integrated about 6 GW of solar generation into its system. APS Smart Grid initiatives, and more specifically its distribution infrastructure upgrades, have made it one of the most reliable utilities in the U.S. APS has also implemented self isolating and self healing technologies on two of its main distribution lines in Flagstaff, AZ using an S&C Electric 900 MHz radio system. Additionally, the utility has integrated more than 200 MW of consumer owned photovoltaic (PV) systems into its distribution grid, one of the largest portfolios of PV capacity in North America. The company s Flagstaff micro grid project will incorporate an additional 1.5 MW of PV capacity, solar hot water heaters and micro wind turbines, as well as a 1.5 MWh lithium ion battery storage system designed to complement the utilities 500 kw Doney Park PV farm. Pacific Gas & Electric (PG&E) plans to install a voltage/var optimization (VVO) system between 2013 and This system would mitigate the voltage variations caused by distributed energy resources, maintain the desired voltage profile, and implement Conservation Voltage Reduction (CVR) year round to continuously reduce demand. PG&E also plans to expand the use of CVR from peak load only to full time application. In addition, automatic load tap changing is being implemented on transformers at PG&E to increase control, drive operational efficiency, and enable improved fault detection. To address rising penetration levels of distributed energy resources, San Diego Gas & Electric (SDG&E) is rolling out an advanced weather prediction and visualization system in 2013, followed by a distributed energy resource management system (DERMS) to be finished in These two systems will improve SDG&E s ability to predict the output of intermittent generation sources and optimize the dispatch of distributed resources, demand response, and PEV charging. 6 7

114 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII As part of their Circuit of the Future initiative, Southern California Edison (SCE) is piloting distribution level rapid voltage support technologies to lower the impact of increasingly large concentrations of VER on individual circuits. The utility plans to install a small number of DSVC units in a small group of substations that have a high penetration of distributed PV generation. Static VAr compensators are typically placed near high variant voltage sources or at transmission substations to compensate for rapid shifts in voltage and automatically adjust the reactive power supply. This pilot project will attempt to cost effectively regulate rapid voltage shifts that accompany distribution circuits with high PV penetration when solar radiation intensity shifts rapidly (e.g. from passing clouds). The project will run from Tennessee Valley Authority (TVA) is piloting a newly developed technology by Smart Wire Grid, Inc. which clamps onto transmission lines to control the flow of power, turning ordinary transmission lines into smart wires. The Smart Wire devices are powered using line current and do not require communications among devices since they are designed to autonomously switch on at pre specified current levels. With such tools, system operators can limit the current flow on the line to maintain quality and reliability. 6.2 HAWAII S EXISTING SMART GRID, COMMUNICATIONS AND CONTROL PROGRAMS Smart Grid implementation is in the infancy stages in Hawaii. Smart meter solutions have been piloted but have not found support for full scale implementation. At the end of 2012, FERC reported that only 0.2 percent of the state s approximately 485,000 total meters met FERC s current definition of AMI smart meters, with only 0.1 percent of residential services connected to a smart meter. 86,87 This penetration level is clearly very low, especially when compared to the 2012 national average of 23 percent. This is an area of significant opportunity which should be investigated to enable the State to meet its aggressive renewable energy goals. On the island of Maui, a project is in progress to test the feasibility of home based Smart Grid technologies. The primary focus of the project is the deployment of a smart meter network in the Maui Meadows neighborhood in South Kihei. 88 In addition to allowing both customers and the Maui Electric Company (MECO) to more precisely monitor home electricity usage in real time, project goals include fostering increased integration of renewable energy resources. In cooperation with multiple project partners, the initiative is being managed by HNEI and MECO and is funded by the U.S. Department of Energy (DOE) through the American Recovery and Reinvestment Act (ARRA). The AMI network was installed in 2012, laying the communication system foundation for a Smart Grid. Smart meters were subsequently installed at the homes of volunteer customers and 86 Federal Energy Regulation Commission Staff Report, David Kathan, et.al; 2012 Assessment of Demand Response and Advanced Metering, December Previously reported figures were as high as 2% but were comprised of AMR/1 way technology meters

115 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII participants can track their energy consumption on line through a personalized energy website. Program participants may also choose to have optional in home devices installed as part of the project. The in home device options include in home displays (IHD), smart thermostats (PCT), and smart water heater control switches. The project is currently in the data collection phase. Surveys and home visits are being conducted periodically to gather feedback from program participants. After a minimum of one year data collection, a report is due to the DOE evaluating the technologies and impacts on energy usage. The project will be used to support decision making on future smart grid initiatives in Hawaii. Another promising initiative is occurring in eastern Oahu and Waikiki, where HECO is starting development of a self healing grid. In partnership with Siemens Corporation, the East Oahu Transmission Project (EOTP) is intended to increase electricity reliability in East Oahu by automating HECO s 46 kv sub transmission system using self healing technologies to modernize existing energy distribution infrastructure. The first completed phase included installation of new transformers and the second phase aims to automate high load distribution circuits. 89 In parallel, the East Oahu Switching Project (EOSP) aims to provide automated maintenance switching, fault isolation, and restoration capabilities managed through HECO s control center. This project includes the installation of intelligent substation controllers, automated switches, and automated reclosers to quickly isolate and restore power. The broader for HECO is to provide a working model for how to satisfy Hawaii s growing energy needs without expensive construction and expansion in Hawaii s delicate environment. 90 Efforts such as these ones by HECO and MECO will be critical in improving the grid landscape for renewable energy penetration. 6.3 OPPORTUNITIES TO ENHANCE HAWAII S SMART GRID PROGRAMS Available Smart Grid solutions cover a range of applications. Utilities must evaluate which mixture of solutions would be optimum for voltage and frequency management under varying scenarios of intermittent renewables penetration. Hawaii s utilities will gain advantage by including Smart Grid considerations in their renewable energy integration efforts and resource planning processes. Smart Grid implementations are improving the efficiency and reliability of the grid and allowing direct two way communication with customers, allowing the potential for local voltage control and the ability to turn on or off resources and load at the source to assist in managing system events. By integrating Smart Grid efforts and renewable energy efforts, Hawaii s utilities will be able to make better technology decisions, maximizing the value of future investments while reducing the risk of one effort outpacing the other Ibid. 6 9

116 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Smart Metering as an enabler Many utilities have identified smart meters as the first step toward a Smart Grid, connecting them to their customers by integrating the delivery and consumption of electricity using two way wireless communications. Figure 6 1 below illustrates the growth in smart metering penetration between 2006 and Figure 6 1 Source: Federal Energy Regulatory Commission 2012 Assessment of Demand Response and Advanced Metering; Staff Report; December 2012 Estimated Advanced Metering Penetration Nationwide Although there are differing definitions, the following are generally the minimum characteristics of a Smart Metering AMI system. At least hourly reading frequency, with most new systems supporting 15 minute increments At least daily retrieval of readings Scheduled remote reading Possibility for real time display Voltage profile and exception reporting Internal disconnect switch on single phase meters 2 way fixed network communication Ability to receive and execute on demand commands (e.g. for demand response) While their primary purpose is tracking the usage of individual power consumers, modern solid state meters provide a wealth of information which can be remotely harvested by modern metering systems such as AMI networks. Utilizing advanced electronics, smart meters provide valuable sensors distributed throughout a service territory. These sensors are capable of monitoring voltage, current, rudimentary power quality, tamper and theft activity, and service outages. Deploying AMI can therefore significantly improve utility knowledge of circuit specific conditions, 91 Federal Energy Regulation Commission Staff Report, David Kathan, et.al; 2012 Assessment of Demand Response and Advanced Metering, December

117 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII reducing operating costs and improving reliability. In addition, AMI systems provide a communications network which can be utilized for much more than just automating the meter reading process and harvesting smart meter data. In addition to the operational benefits of smart meters, there are also a host of consumer benefits such as those described in section 4.4. A good example relevant to the integration of distributed generation is use of net metering. Net metering is the measurement of energy flow in both directions, into a meter from the grid, and out of the meter, as would occur when an individual is producing more energy than they are consuming, thus putting energy back onto the grid. With analytics, the advanced meter data can be used to proactively identify customers who may have installed PV generation but who have not registered with the utility as such and are therefore not being net metered Smart Meters and Planning With integrated advanced electronics, the always on sensing capability of smart meters makes them ideal for identifying unbalanced and overloaded circuits, conditions which are increasingly caused by high PV/DG penetration. Using GIS (geographic information system) and distribution planning and modeling tools, meter data can be overlaid onto geospatial models and single line diagrams in order to visually identify areas of clustering where grid challenges exist. Also, new planning and modeling tools consider not only steady state loading during normal conditions but also more complicated modeling of intermittent renewables fluctuations or coldload pick up conditions of DR events where customer loads may radically change over a very short duration (for example, at the end of a DR event where controlled loads simultaneously turn on). While these conditions may seem radical or extreme today, they will become increasingly normal as renewable energy resources grow as a percentage of total supply capacity and smart meter data can provide a valuable method to plan for them Smart Meters and Voltage Control Voltage regulation is a core function of all electric utilities. Most electric equipment is designed to operate at a consistent voltage, with excessive fluctuations causing wear and tear, leading to premature equipment failure and unhappy customers. Thus, industry standards require utilities to maintain acceptable voltage (120 volts +/ 5 percent) at the service entrance of all customers served under all possible operating conditions. Increased intermittent renewables will make the job of maintaining consistent voltage significantly more challenging. Traditional grids were designed to transport power in one direction, away from the substation, with voltage controlled at the distribution level and declining as the current traveled farther from the substation. Renewable energy resources, however, increase the voltage at the 6 11

118 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII point of interconnection and also fluctuate as the fuel sources (i.e. wind or solar) fluctuate. 92 For grid scale renewable energy resources utilities must implement solutions which incorporate low voltage and low frequency ride through and can take advantage of regulating capacity if available. However, dealing with voltage fluctuations caused by residential renewable energy sources, especially solar, is a much larger issue for utilities. Electric meters provide the ability to measure actual voltage at individual customer locations. By being connected to an intelligent communications network, smart meters enable a continuous feedback loop for measurement and verification of local voltage, allowing voltage regulator and capacitor bank operations to be optimized around voltage management. Smart meters are also capable of logging voltage levels, sending report by exception voltage alarms, and with analytics and event management solutions, these new telemetry sources can be used to maintain system voltages for all customers closer to the nominal 120 volt service level. For example, voltage analysis using advanced analytics on smart meter voltage data can be used to proactively address poor voltage quality issues caused by malfunctioning equipment. AMI voltage data can also improve CVR (conservation voltage reduction) solutions. CVR intentionally lowers the voltage on distribution feeders to a minimum while ensuring it is still acceptable at the end of line (within the lower limit requirements). By lowering voltage to the lowest acceptable voltage level, electric utilities reduce overall demand and reduce energy consumption without any adverse impacts on customers. While CVR is not new, emerging technologies better enable utilities to optimize the voltage towards the lower end of the spec range (114 volts), while still complying with customer equipment performance requirements. Finally, smart meter voltage data can play a critical role in VVO (volt/var optimization). Traditional feeder voltage regulators and switched capacitor banks are operated as completely independent (stand alone) devices, with no direct coordination between the individual controllers. This can be problematic since changes in one device could cause an opposite reaction in another, negating the intended system operator s goals. The coordinated control of voltage and reactive power via modern software systems is needed to determine and execute volt/var control actions that are truly optimal. Stand alone VVO solutions added a level of intelligence and central coordination, but the advent of the Smart Grid has spawned Smart VVO. By integrating smart meter data feeds and other telemetry sources in near real time, advanced VVO systems can now determine the optimal set of control actions for voltage regulating and VAr control devices. VVO strategies should be deployed in order to make the grid more energy efficient. Although energy savings will differ by circuit based on various types of loads, the benefits are clear, as is the role which AMI data can play. Implementation of advanced VVO techniques are not necessarily needed 92 Sandia National Laboratories, "Solar Energy Grid Integration Systems (SEGIS) Program Concept Paper," DOE EERE/SNL, October

119 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII system wide in the near future, but could very much justify their costs on feeders with high VER concentrations where voltage & VAr fluctuations occur frequently Smart Meters and Reactive Power Management While the implementation of more capacitor banks can reduce line losses when functioning properly, malfunctioning timers and blown fuses reduce the effectiveness of these devices, often unbeknownst to grid operators. Technology vendors have developed solutions which can directly monitor and even control regulating devices such as capacitor banks but these solutions typically require dedicated telemetry systems. However, in recent years, AMI system vendors have started to integrate these types of monitoring and control solutions directly into AMI systems, further increasing the value of AMI beyond simply reading meters. In addition, reactive power measurements recorded by smart meters can be used indirectly to verify that power management assets are functioning properly and achieving the intended benefits in supporting system voltage. Smart meters can also help grid operators more strategically deploy the assets used in the management of reactive power. Using smart analytics, AMI meter data can be used to identify customer loads which generate the most reactive power flow and those circuits which may benefit from grid assets such as capacitor banks, resulting in better prioritization of limited utility resources Control systems as an enabler Historically distribution operations have been characterized by: The manual operations of devices by field personnel, Local operation of devices by control panels configured largely to operate independently of one another, Operational processes based on past practices and a fixed circuit topology, Operating parameters based on calculations and approximations focused on peak loading, and Data spread across multiple databases using various storage media including computerbased storage as well as hard copy. While these approaches have served the electric utility industry well for over a century, there are inherent limitations and software based control systems are well suited to address them with more efficient, automated solutions. Monitoring and control is required for load devices, DG devices, distribution system devices, transmission system devices and generation devices, often using separate systems. An Advanced Operating System can integrate all of these communications and controls. A major component of an AOS is the development of Distribution Management Systems (DMS), which allow utilities to take advantage of the unprecedented data and technologic advances that make up the Smart Grid. Performing long term generation planning, substation and distribution system upgrade design and routine load flow analysis requires insight into all available DG resources. This in turn requires that 6 13

120 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII these sources provide real time information that can be fed into the new generation of DMS power flow applications for both planning and system operations. This provides the information that utilities need to plan and operate the grid safely and reliably. Figure 6 2 illustrates how DMS solutions are integrated into the broader utility operational and control landscape. Figure 6 2 DMS Example Integration Scenario The heart of a DMS is the data model of the distribution system, which requires both operational and static data. Operational data deals with changes to connectivity and equipment settings. This typically includes device status changes, such as switch status or the position of a capacitor bank. Static data, on the other hand, describes the distribution system over an extended time frame. These data include information from the GIS for the distribution connectivity model, system configurations and land base to represent the distribution lines, as well as substation internal connectivity to model substation connectivity configurations. Additionally, relay settings and system impedance data are required along with component and facilities ratings, equipment impedance and ratings, and device settings from reclosers, capacitors and regulator control panels. Furthermore, the DMS could not run near real time load flows without customer information regarding customer count, load data and load schedules data which are readily available through AMI systems. Finally, when combined with actual data from AMI enabled smart meters and data from SCADA devices, the DMS data model, using static and operational data, provides a platform from which to perform real time power flow calculations and analysis that represent current conditions on the system. These calculations can enable advance functions like voltage and VAr control, fault identification and service restoration. 6 14

121 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Another key example of Smart Grid control software is a Demand Response Management System. As discussed in section 5.4.1, it is critical to accurately measure and track the effectiveness of DR programs. Relatively speaking, today we are primarily in the realm of ManualDR, where operators trigger events manually or utilize basic scheduling software. DRMS software offers a path to AutoDR and is used to: Manage the overall Demand Response (DR) program, Manage customer enrollment and registration, Support device provisioning, Monitor (& quantify) available DR resources and interruptible demand, Forecast DR resources and interruptible demand, Initiate load control/dr events > Command & Control, Track event participation and actual demand reduction > Validate & Verify, and Manage DR related devices. DRMS solutions are especially useful for residential programs with large enrollment numbers, where the administrative aspects of program management alone can be a major challenge. In parallel, the OpenADR Alliance industry group has led the development of standards for AutoDR to move the industry away from proprietary solutions. Focused on commercial and industrial loads, the goal of OpenADR is to provide standardized processes and protocols to automate the response of commercial and industrial buildings to utility signals. For example, price signals and reliability signals could feed directly into customer side EMS/BMS to execute pre programmed responses. With a DRMS on the utility side, the entire DR process is thereby automated from event initiation to event participation to M&V and settlement. A real world example of an advanced control system initiative is the Sacramento Municipal Utility District (SMUD), which expects to implement an Advanced Operating System (AOS) in its service territory. SMUD is installing automated equipment on select distribution lines and building a wireless communication network to support the AOS. The system will tie together traditional separate solutions including CVR, VVO, and Automatic Sectionalizing and Restoration. Furthermore, SMUD is proposing a DMS with combined functionalities of OMS, and SCADA/EMS that leverages GIS capabilities and integrates into other SMUD operational technologies. The system will enable operators to view and control the distribution system and optimize Smart Grid resources with real time analytics for an efficient and reliable grid. Hawaii s utilities need to strongly consider similar solutions as part of the planning process in order to help meet the aggressive renewable energy goals set by the State Communication systems as an enabler No single communications system is capable of communicating with all the load devices, generation devices, and distribution and transmission system devices. Nevertheless, a key enabler for effectively integrating, monitoring and managing renewable energy resources on the grid lies with 6 15

122 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII the utility's ability to ingest, route, process and act upon the increased levels of instrumentation data from DG sources and the grid as a whole. Smart Grid technology can tie together new and existing communications systems in a coherent and effective manner. With the increasing deployment of information technology (IT) solutions such as enterprise service buses (ESBs) and stream processing tools, the time between "sensory" and "actionable" information can be dramatically compressed, thereby reducing the time between grid stimulus and effective utility response. For example, when combined with advanced analytical software solutions, Smart Grid communications systems can intelligently connect energy storage systems with the grid, leading to an interconnected and orchestrated fleet of communicating batteries 93 which would enable increased levels of renewable energy penetration while providing stable power quality. Up to now, most customer side renewable energy systems have not generated sufficient energy to put electricity back onto the grid. Monitoring these small and dispersed systems was generally therefore not necessary. For wholesale DG systems, this type of monitoring is already typically present and relied upon by utility operators. 94 However, many utilities now feel that visibility into smaller systems, particularly those larger than 250 kw, is needed as well and will allow them to factor small DG resources into their transmission capacity margin calculations. In the future, with a Smart Grid dependent upon new standard communication protocols, it will be more feasible to create visibility and control for these many customer side systems and all of the larger systems. Europeans have already started implementing new communications and control features and strategies. In addition to the existing lack of monitoring and control of customer side energy resources, reliable means of forecasting the near term output from renewable resource systems is only now emerging. Forecasting of DG, and PV in particular, is necessary for better integration and optimization of these resources. For example, forecasting moving cloud cover and other weather changes will allow for improved control and function of the distribution system. And, when combined with advanced communication and control technologies, better forecasting will allow grid operators to maximize the use of available renewable resources or otherwise control distributed generation as necessary. Understandably, utilities want to run multiple applications over their communications networks to maximize the value of their investments and simplify operations. For example, AMI networks 93 companys approach to innovation inelectricity storage focus on the software/ 94 Currently, telemetering requirements are imposed only on DG systems 1 MW or larger. Smaller units do not have this requirement since at lower penetrations the expectation was that the impacts would be minimal, the cost to collect the data would be relatively high, and processing of this additional data may not be necessary. This assumption should be revisited as penetration increases or the operational requirements for DG change. For example, California utilities have recently proposed telemetering requirements for certain wholesale DG systems smaller than 1 MW that will export power to be aggregated and scheduled into CAISO s market. As of this writing, the CPUC has not yet evaluated those proposals. 6 16

123 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII provide significant foundational platforms for communication with intelligent devices, enabling initiatives such as large scale Demand Response programs. Smart Grid communications systems can also address utility concerns with the cost of communication and control of devices, such as smart inverters, by eliminating the need for dedicated infrastructure. With Smart Grid technologies, utilities can leverage the same communication & control features used for automated substations and also integrate them with SCADA. Utilities should therefore evaluate the cost of implementing promising device options which integrate well with Smart Grid communications strategies Enabling DR As discussed in section 5 above, Demand Response solutions provide a host of benefits to increasing renewable energy penetration. Sophisticated software, enhanced communications techniques, and load control features and devices are required to enable DR on any significant scale. Smart Grid technology provides a foundation with appropriate functionality to allow utility control and encourage customer response to utility supplied signals and information. However, DR is not needed everywhere in a service territory. As was recommended by RSWG, Hawaii s utilities should work to identify those customers and loads that are most promising for demand response. 95 Smart meters identify how much energy is consumed, where it is consumed, and when it is consumed. Using this smart meter data, utilities can take advantage of targeted DR, prioritizing where DR investments are required. For example, customers can be targeted on individual substations, or even individual feeders, which are more susceptible to variability due to the higher levels of renewable energy resources. In addition to helping determine where to best target deployment of DR devices, an analytics platform can use smart meter information to identify and prioritize load curtailment requirements for active DR programs. Also, smart meter data can enable customer segmentation and surgical load curtailment, which can exclude critical facilities such as hospitals, nursing homes, and police and fire stations from DR events. BGE (Baltimore Gas & Electric) has already used their AMI data to perform these exact types of analytics for targeted demand response. Armed with smart meter data, BGE has used advanced analytics to help proactively identify the right customers to target for BGE's peak rewards program. 96 Already seeing clear benefits, BGE is continuing to refine their algorithms and to expand the scope of their smart meter data analytics efforts. Hawaii s utilities should look at programs such as BGE s as models of how smart grid data can enable programs which have direct benefits to renewable energy integration. 95 Reliability Standards Working Group, Hawaii PUC Docket No , Demand Response as a Flexible Operation Resource, RSWG Demand Side Options Subgroup 96 Grid Demand Response Today s Leaders at Risk as DR 2 0 Emerges 2252.html 6 17

124 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Enabling Energy Storage Energy Storage (ES) is a technology that can provide electric system reliability functions. Smart Grid communications and control functions can enable this technology. While energy storage does not directly control usage (as DR) nor reduce overall consumption (as EE), it has a lot of potential benefits for integrating renewable generation by working as part of the Smart Grid to smooth out the supply curve (as opposed to demand curve). Different storage technologies have different characteristics (such as storage duration and cost) that make them more or less suited for particular applications. Below are some potential energy storage solutions: Short Duration (flywheel, ultra capacitor), Medium Duration (batteries), and Long Duration (pumped hydro storage, compressed air storage). Rather than shifting demand via DR, energy storage with Smart Grid control capabilities provides a method to shift supply, particularly in conjunction with renewable generation. For example, residential demand typically spikes in the early mornings and early evenings and is low during the middle of the day when solar energy would be at its peak. Through Smart Grid control capabilities, solar energy would be stored during the day and then be available in the evening when consumers return to their homes. Energy storage devices do not necessarily need to be co located with renewable generation. They could be widely distributed across the grid. With Smart Grid controls, distributed energy storage devices may meet multiple needs in different applications at various scales from large centralized storage devices at substations, to community energy storage devices on distribution feeders, to small devices at individual homes (see Error! Reference source not found. overlaid with Smart Grid controls). 6 18

125 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Figure 6 3 Potential Applications of Energy Storage With two levers at their disposable DR to shift demand and Energy Storage to shift supply grid operators have two powerful and complementary methods to tailor the supply demand balance, providing better conditions for increasing renewable energy penetration. Energy storage can likely meet all needs for regulation, spinning reserve and other ancillary services required for variable renewable generation; however, it is costly. The relative economics of these ES and DR options should be analyzed in relation to other approaches to integrating renewables while still maintaining power quality (such as voltage and VAr control technologies smart inverters, forecasting, automated substations, etc.) under various grid scenarios. Effectively integrating energy storage as a flexibility tool will require Smart Grid technologies to ensure that storage is used optimally. A specific storage method with strong potential for Hawaii is commercial ice storage, specifically for large hotels. Commercial ice storage shifts cooling loads from the day to the night. Using roof top refrigeration units, ice is built up over night and is then used to cool buildings during the day. Thus, ice storage is very complementary to wind generation since wind energy typically peaks during offpeak periods, periods during which wind is often curtailed. In regions with high concentrations of large hotels (or other significant commercial buildings), cooling through ice storage could significantly improve the use of renewable energy and help improve grid reliability Enabling Generation and Storage and Reserves Management Smart Grid technologies can enable the management of generation, energy storage and reserves in an effective way that can limit the challenges with intermittent renewable integration. Recent wind turbine technology improvements, specifically from GE, are moving towards smart (what GE is 6 19

126 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII calling brilliant ) turbines. These turbines communicate with each other and such other devices as energy storage to maintain grid power quality. 97 As an additional mitigation technique, conventional generation not already equipped with automatic generation control (AGC) can be modified to include AGC and appropriate SCADA communications to integrate with the greater Smart Grid. Intermittent generation can also be equipped with AGC capabilities that communicate with the Smart Grid. And in general, AGC operations can be modified to enhance the integration of intermittent resources; HELCO has already modified AGC control for intermittent generation including allowing a greater range of frequencies as compared to conventional generation technologies. 98 The use and management of reserves can be modified with the Smart Grid to enhance intermittent renewables integration. It is considered good practice to use contingency reserves for extreme wind events causing extended ramps. 99 Such reserves will receive signals from the AOS. If contracts require, wind farms can be equipped with features that allow modest limits for ramp rate control and AGC can be provided for down regulation. Communication is required between the AOS, the wind generators and the wind meteorological stations involved in wind forecasting Smart Grid technology will be the enabler for this communication. Using the Smart Grid to dynamically calculate regulation reserves and load following reserves can mitigate net load variations and can help to reduce overall reserve requirements. Reserve sharing between electric systems will average out fluctuations. Reserves sharing will be enabled when the undersea transmission cable comes on line and through Smart Grid communication and control capabilities Enabling Forecasting and Curtailment Improved forecasting, possibly including ramp forecasts, helps grid operators monitor system conditions, schedule or de commit fuel supplies and power plants in anticipation of changes or ramps in wind and solar generation, and prepare for extreme high and low events. Some studies show that forecasts can reduce costs by decreasing fuel consumption. 100 For less flexible power systems with small balancing areas dominated by thermal generation such as Hawaii, curtailments of generation may be required, even at low variable generation penetrations new brilliant wind turbine putto the test 98 HELCO IRP Advisory Group Oriented Session, HELCO Systems Operations, August 22, Kevin Porter, et. al; PJM Renewable Integration Study, Task Report: Review of Industry Practice and Experience in the Integration of Wind and Solar Generation; Prepared by Exeter Associates, Inc. and GE Energy. 100 Kevin Porter, et al; PJM Renewable Integration Study, Task Report: Review of Industry Practices and Experience, in the Integration of Wind and Solar Generation, Prepared by Exeter Associates, Inc. and GE Energy. 6 20

127 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Demonstrations are occurring to better understand the technology and cost tradeoffs necessary to optimize forecast accuracy. While better accuracy is always desired, there are limitations on load forecasting accuracy 101. Overall, there is a point of diminishing returns on forecasting accuracy that needs to be identified. Key barriers to better forecasting are deficiencies in forecast accuracy; time required to implement, process and collect data; more need to incorporate VER into day ahead dispatch and lack of real time dispatch schedule updates. Curtailment of VER output may be necessary if the amount available at a specific time is more than the grid can reliably deliver while maintaining stability. Recent wind integration studies and operating experience demonstrate that at higher levels of penetration, wind generation may need to be curtailed during certain periods, unless suitable flexibility is designed into the bulk power system. HELCO has already implemented curtailment to reduce variability and excess generation 102. From a technology standpoint, wind generation can be designed with equipment to limit or flatten ramps and wind generation can be curtailed to provide reserves i.e. a source of increased flexibility. And smart inverters may provide curtailment capabilities for solar PV facilities as well. These features may require additional communications and utility control via the Smart Grid to maximize their benefit. To optimize the use of renewable energy resources and minimize curtailment, forecasting tools need to be combined with Smart Grid technologies, including advanced DMS, AMI, and DRMS. Using these combined systems, electric grid operators will have the tools to monitor both energy supply and energy demand and react in real time with appropriate system responses. Emerging wind turbine technologies such as the brilliant GE turbines mentioned above may also aid in limiting or flattening ramps. There is a demonstration project on Lanai where a 1.2 MW solar PV system is equipped with a Satcon inverter that has a remote utility control system. This control system, in conjunction with a 12 to 24 hour solar resource forecast enables MECO to predict the solar PV generation and effectively integrate it into the grid How microgrids help A microgrid is a technical solution that may have the ability to assist renewable integration at the local or distribution level. Microgrids typically integrate storage and generation into a seamless system that integrates renewables into reliable local grid solutions. On an island with no interconnections, several microgrids integrated into a total island grid may provide increased system reliability as compared to one central grid solution. 101 Bri Mathias Hodge, Anthony Florita, Kristen Orwig, Debra Lew, and Michael Milligan, A Comparison of Wind Power and Load Forecasting Error Distributions, National Renewable Energy Laboratory, Presented at the 2012 World Renewable Energy Forum, Denver, July HELCO IRP Advisory Group Oriented Session, HELCO Systems Operations, August 22, Bower, W; et al, Solar Energy Grid Integration Systems: Final Report of the Florida Solar Energy Center Team, SAND , March

128 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII SMUD defines a microgrid as a smart grid concept that is an Integrated energy system consisting of interconnected loads and distributed energy resources which as an integrated system can operate in parallel with the grid or in an intentional island mode. Microgrids can vary in scope, size, and ownership. 104 The CEC says, Microgrids are basically self contained electrical ecosystems. Power is produced, transmitted, consumed, monitored, and managed all on a local scale. In many cases, they can be integrated into larger, central grids, but their defining characteristic is that they can operate independently if disconnected from the whole. This single point of common coupling with the macrogrid can be disconnected. The microgrid can then function autonomously. Microgrid generation resources can include several distributed generators, including fuel cells, wind, solar, or other energy sources. The multiple dispersed generation sources and ability to isolate the microgrid from a larger network can provide highly reliable electric power. As SMUD demonstrates, smart grid features need to be included in modern microgrids to allow monitoring, operation, and automated connection or disconnection from the macrogrid, to optimize storage, if included, and to eliminate the potential for unsafe islanding (operating unsafely after disconnection from the macrogrid). Interoperability with demand response may also be included. Renewable resources connected to the microgrid may include real time forecasting for optimum automated operation. Microgrids are also used by campuses, military bases, communities or other loads to integrate distributed renewables and increase their reliability above and beyond a grid only connection. The military is very interested in microgrids to allow military bases, including those in Hawaii, to operate independently when there is a grid outage due to equipment failures, weather events or terrorism. A microgrid allows itself to seamlessly and automatically operate as an island when there is a grid outage. A key component is the electronic switch that provides this capability. It monitors grid voltage and automatically disconnects the microgrid from the grid when a grid outage occurs. In addition, a microgrid will likely have a controller to optimize operation of the distributed generation and the storage as well as the fossil fuel fired DG if provided. A battery, an internal combustion engine and a wind turbine integrated with a load comprise a simple microgrid when integrated with the controller and electronic switching. The growing frequency of extreme weather events increases the appeal of the reliability benefits attached to a microgrid. For this reason university campuses are also interested. The University of California, San Diego has a well recognized microgrid that has operated independently for months while San Onofre Nuclear outages decreased grid reliability. The University of Santa Clara has 104 Mark Rawson, Microgrid and Smart Grid Activities at SMUD, Sacramento Municipal Utilities District, Presented at the 2010 MicroGrid Symposium, July

129 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII adopted renewable energy and pursued microgrid development as a result of continued outages which prevented operation of schools in New Orleans after Hurricane Katrina. Universities in Hawaii could be interested. SDG&E is developing a microgrid at Borrego Springs in the high desert, looking to tap excellent solar energy resources owned by its customers to help sustain operations at a remote utility substation. The State of Hawaii is involved in federal microgrid demonstration projects and has a number of private microgrid projects, including the one described in the following excerpt: A separate project from this SEGIS effort, but with particular relevance, involves a PV generation site on the Hawaiian island of Lana i, where a moderate PV plant (1.2MW) is integrated into a small diesel engine generator grid (4MW). The 1.2 megawatt (MW) installation is the first solar photovoltaic power plant to be controlled remotely by a utility, Maui Electric Company, Ltd. (MECO). MECO has to manage the grid remotely from the island of Maui, thirty miles and an ocean channel away, by ramping up and down different generation sources. They also have to be able to handle intermittent power output from the PV plant due to cloud cover during the day and lack of solar power at night. Given a hour weather forecast and knowledge of expected solar irradiance levels at all times during the day, MECO is able to predict when the solar plant is available to generate power and when spinning reserve diesel generators need to be used to meet demand. Satcon developed a remote utility control system embedded in the inverters to be used at Lana i. With this advanced system control technology, the PV plant is seamlessly integrated into the utility grid, with the inverter serving as the central control point. The intelligence contained in the inverter means that MECO can dynamically switch from solar to diesel energy sources and back, thus supplying uninterrupted power to the island. Additionally, MECO can use the inverter to generate reactive power from the PV plant, stabilizing the grid when it comes under stress. The penetration of 30 percent Solar PV coupled with the small grid highlights many of the same concerns of large scale integration into the larger utility grid. This project was undertaken at the request of the PV Integrator (SunPower), the PV plant owner (Lana i Sustainability Research a division of Castle and Cooke), the DOE, and the local utility (MECO, HECO), to incorporate inverter control features that would enable the high PV penetration, address and control intermittency, give the utility control over the plant from a remote control center (on the island of Maui), and upgrade overall grid performance. Incorporating these inverter and PV plant control features, integrated with the utility SCADA system on 6 23

130 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII the Lana i grid, required detailed software development, extensive simulation work, and laboratory and field testing, which ultimately enabled the site owner to meet the requirements of the Power Purchase Agreement with MECO, the local grid operator. While this Lana i project is separate from the SEGIS effort and has distinct differences, it did provide an opportunity to provide concept checks on SEGIS and to work on high penetration at 30 percent of the grid load during peak solar hours, but also 10 percent of the annual electrical energy demand, which today is all produced from diesel fuel. 6.4 SECTION SUMMARY In preparing themselves for the increased penetration of renewable DGs, utilities need to investigate and apply smart grid solutions such as smart metering, smart communications solutions, distributed monitoring and control, and DMS applications. The Smart Grid enables electrical system flexibility to meet the challenges of integrating intermittent renewable energy resources. The level of challenges will depend on the amount, size, location, and voltage levels of the interconnected resources; the capabilities of those resources for dispatch and communication (and similar Smart Grid capabilities); and the ability of the grid to accept generation at multiple distributed sites and flow power as needed in the opposite direction as initially designed. In addition, these solutions have far reaching value beyond supporting DG. Smart Grid technology provides system operators with real time awareness of electrical generation, storage, transmission, distribution, and loads, resulting in improved system performance and reliability. The development and implementation of such technologies and their combination will enables real time information exchange throughout the utility and with its customers. Therefore, while Smart Grid solutions should be considered necessary for providing flexibility functions to optimize the use of renewable energy resources, the greater impact is an integrated electrical grid which provides utilities with visibility and control to improve the efficiency, reliability, and quality of the electrical energy provided. 6 24

131 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 7.0 Barriers to Adoption of Smart Grid Technologies in Hawaii This section discusses the following barriers to the adoption and deployment of the technologies and programs which are valuable in facilitating the increased penetration of renewables: Lack of technical and commercial readiness Technologies must demonstrate technical viability and economic viability, their performance must be properly valued by regulators, and costs must be properly allocated. Business rules, regulatory rules, and contracts must be adapted to deal with these implications. An example is a technology to limit ramping of wind turbines which would supply a flexibility/reliability function that would be paid for by utilities after appropriate business rules and contracts are put in place. Lack of utility preparedness, Cost Implementation costs: Incentive levels Critical mass to support 3 rd party participation Unrecognized benefits of technology or enabling solution to grid, ratepayers, and State Cost recovery Lack of customer awareness Customer s lack of access to their own energy information Policy, regulatory, statutory barriers Permitting requirements/procedures 7.1 BARRIERS Lack of technical and commercial readiness Energy efficiency and some demand response programs are well understood, and performance and cost data is well documented. A mature smart grid does not exist at scale in any utility system. However, pilots and demonstrations are being implemented globally. Most utilities begin smart grid implementation with smart meter implementation and they are working hard to utilize the data and communications available. However, full implementation of smart meter capability does not yet exist either. Additional demonstrations are necessary to understand the implementation of all smart meter features Lack of Utility Preparedness Smart Grid solutions are only one category of investments and initiatives that utilities are undertaking in today s rapidly changing utility environment. The Smart Grid solutions discussed 7 1

132 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII herein must compete with other projects and investments underway at the HECO Companies. And it is not just a matter of sufficient financial inducements or capabilities but the ability of a utility, or any organization, to withstand significant amounts of change. The enabling technologies and advanced applications require new employee skills, redefined job classifications, modified organizational structures, new operating processes in general a fundamental transformation of how HECO operates the electric grid. The HECO Companies also have very different current configurations and level of technology advancement that must be assessed. The amount of distribution automation, substation automation, advanced applications, GIS capabilities, and other foundational systems varies significantly across the island. Therefore the starting point to a smarter grid and the enablement of increased grid flexibility is island specific Cost Costs are well understood for existing EE technologies. This is also partially true for commercial DR, but the role of large aggregators will be dependent on the level of incentives offered because of the high cost of establishing operations in Hawaii, relative to the size and number of customers with large, dispatchable loads. Costs for new advanced DR programs that depend on smart grid communication and control features are not completely understood. Additionally, a method to allocate costs based on actual achievable benefits is required. The costs and benefits of smart grids and nearly all smart grid features necessary to enable system flexibility are still being defined. Comprehensive studies which measure costs and benefits of various flexibilities and enabling smart grid features are necessary for each unique electric system Cost Recovery Cost recovery is an issue that needs to be addressed differently for different technologies. For energy storage new tariffs and other rule changes that allow benefits to be monetized are only now beginning to be considered. Interconnection costs for wholesale DG systems are often an issue because of rules for allocation of costs to the DG owner versus the utility. There are also issues of cross subsidies and tariff structure for net metered systems. And in the future, requirements for smart grid features will need to be allocated between the utility and the owners of the generation. Finally, there are also issues with respect to rate structures and the availability of price signals that modify consumer behavior, especially for customer driven DR programs. High existing rates and customer resistance to assuming the burden of higher rates may limit the amount of cost that can be allocated to the ratepayer Lack of customer awareness Well designed behavior based EE and DR programs require informed and educated customers with adequate awareness, education, and access to data. The lack of a feedback process and ability of customers to connect actions and results are a significant problem with some EE and DR programs. Customers must be able to understand the impact of such programs on their lives and homes; 7 2

133 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII concern over how such programs will affect them results in a portion of customers deciding not to participate. Conversely, obtaining information on how programs positively impact them, either financially or through contribution to societal benefits have been shown to further stimulate customer responsiveness Customer s lack of access to their own energy information Customers require information to make informed decisions that can contribute to utility objectives. A mechanism to get more usable information to the customer is the smart meter. In home displays or web pages, programmable appliances, thermostats, and power strips will assist with displaying or automatically using energy information. Rate structures also need to be modified to deliver proper price signals, especially to enable functionality of smart meters and smart appliances Policy, Regulatory, Statutory barriers Policy, regulatory, statutory, and contractual or arbitrary business rules are all barriers. New contracts may be required to modify behavior and designs. Codes and standards are needed to allow smart inverters and customers may need to be paid for supplying benefits to the electric system. Regulatory changes will be necessary to allow new types of reserves, monetizing of benefits or modification of reliability standards, methods of calculation, protection of data privacy and other factors that impact how we design and operate electric systems with new automated communication and control technologies. Larger DG projects need policy mechanisms to ensure they are sited to provide utility system benefits. Regulatory actions are also required to allow or cause widespread smart grid implementation Permitting requirements/procedures Permitting requirements are of more significance to utility scale projects than net metered DG. Mechanisms should be put in place to identify preferred locations for siting utility scale projects in areas of minimal environmental impact while achieving maximum beneficial impact to the utility system. Similarly procedures should be put in place for larger DG projects to ensure they are located strategically to provide distribution system benefits. 7.2 SECTION SUMMARY The biggest barrier to attaining high penetrations of energy (kwh) from variable renewables is lack of information that informs utilities about the costs and benefits of options to provide reliability functions and lack of information to enable customer based DR programs. As more demonstrations are conducted for DR technologies and smart communications and control technologies that enable various flexibilities, there will be a need for comprehensive studies that identify cost benefit tradeoffs so utilities can identify and rank alternatives which allow higher penetrations of intermittent renewables to be achieved on each Hawaiian island. 7 3

134 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Another barrier exists with respect to the rule changes to UL 1741 and IEEE 1547 safety standards that will be necessary to implement smart inverters and the business case development for regulatory actions necessary to achieve full smart grid implementation. 7 4

135 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 8.0 Solutions to Barriers to Adoption of Smart Grid Technologies Increased Penetration of Renewables in Hawaii This section addresses the barriers identified in Section 7 and presents solutions designed to overcome these barriers. Recommended focus areas include policy, customer engagement and technology readiness assessment. 8.1 INCENTIVES, TAX CREDITS AND REBATES THAT CHANGE PERCEIVED COST Beyond existing federal, state and utility incentives for renewable energy projects and energy efficiency, Hawaii utilities should evaluate the potential for using different types of incentives that would place new DG in locations on the grid that maximize system benefits. 8.2 MODIFICATION OF COST RECOVERY MECHANISMS Hawaii utilities should investigate options that would allow utility ownership of DG. Hawaii utilities should investigate alternatives for evaluating benefits and costs of DG. The goal is to get DG placed in locations that help utilities achieve specific beneficial objectives. Much of the benefit ascribed to DG is not currently realized since utilities have yet to adopt planning processes designed to achieve real documented savings realizable from strategically sited DG. 8.3 MECHANISMS THAT IMPROVE TECHNICAL AND COMMERCIAL READINESS Utilities around the country are performing demonstrations focused on improving technical readiness. These demonstrations prove feasibility and provide technical information necessary for changes to standards, rules and regulations. However, very few of these demonstrations are focused on defining implementation costs and the next steps necessary to move from technical success to commercial scale implementation. Also, the results of these demonstrations will need to be analyzed to make the results applicable to Hawaii electric systems. Demonstration programs which should be defined and implemented include those to achieve the RSWG objectives above: Inertia All demonstrations of new technology should include consideration of inertia or synthetic inertia. Providing inertia like response from technologies which have inverters has not previously been a priority. The demonstrations need to consider the cost of providing these ancillary services, the mechanisms to pay for these services and the rule changes necessary to require them in all future installations. Frequency Similarly, frequency response must be considered in technology demonstrations and the smart grid features need to be included to achieve this frequency response automatically. 8 1

136 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII AGC AGC and associated Smart Grid communications and control will need to be demonstrated on all generators, storage and demand response. Spinning Reserves Demonstrations will be required to provide command and control functionality into variable generation that will provide spinning reserve functionality. Contingency Reserves Demonstration of new types or combinations of reserves Reactive Power and Voltage Regulation Smart inverter technology should be investigated through demonstration to determine its usefulness for different system sizes. Smart grid technology will be necessary to manage information from the large number of DG inverters anticipated. It may not be necessary to require all sizes of projects to be included, but the need for more flexibility on the Hawaii Island electric systems means a greater need for smart inverter capability. Smart grid technology which manages and controls voltage on distribution systems will need to be demonstrated. Short Circuit Smart Grid features that provide the recommended short circuit functionality will need to be demonstrated. These features can be coupled with advanced protection schemes that leverage the status data available from remote devices as well as a communications system to provide real time information on fault conditions and locations. Planning Planning should seek to use probabilistic techniques to establish priorities for technology implementation and geographic implementation that minimizes costs. New techniques for integrating transmission and distribution planning may need to be demonstrated. Compensation Mechanisms to compensate developers of flexible resource projects will need to be developed once the technology demonstrations are completed, as part of the early commercialization process. 8.4 EDUCATION TO IMPROVE CUSTOMER AWARENESS As mentioned previously, barriers to efficiency and renewable energy can be overcome by providing information and enabling customers to make informed energy decisions. Progress toward state goals can also be accelerated through enhanced marketing efforts, to improve accessibility of existing EE, DR, RE and smart grid incentive programs. Particular tools and technologies that can support customer education are smart meters, energy data visualization tools including web portals, and green button enabled programs. These technologies are discussed in Section RULE CHANGES Business rules will need to be created and possible regulatory changes will be necessary to allow ancillary services defined by RSWG to be bought and sold in Hawaii. 8 2

137 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Wind procurement contracts may need to be modified to accommodate curtailment or to require features that facilitate curtailment and utility control. To better manage generation reserves, HECO will need to equip more of its existing conventional generators with automatic generation controls. Hawaiian utilities should consider modest limits for ramp control of VERs and should use contingency reserves for extreme wind events. Another barrier exists with respect to the rule changes (UL 1741 and IEEE 1547 safety standards) necessary to implement smart inverters and the business case development to justify regulatory actions necessary to achieve full smart grid implementation. 8.6 SECTION SUMMARY HECO needs to demonstrate smart grid enabled technologies which provide electric system grid flexibility. The costs and benefits identified from these technology demonstrations will need to be analyzed and compared to each other. In some cases, cost recovery mechanisms will need to be adjusted to achieve implementation and in some cases incentives will need to be identified to achieve commercialization. Also, in some cases rule changes may be required and it may be necessary to provide education to improve awareness of customers, utilities and independent power producers. Key technology demonstrations will be related to smart inverters, distribution system voltage control, provision of AGC to enable regulation, frequency response, reserves management, and overall system integration. Today, the greatest unknowns preventing the development of a set of implementation priorities are the costs for individual technologies and mechanisms and the information necessary to compare those costs. Once the information is available to establish priorities, the emphasis will switch to understanding and modifying cost recovery mechanisms, establishing appropriate incentives and identifying appropriate rule changes and education mechanisms. A typical or representative sequence of barrier removal which applies to all technologies and mechanisms used to achieve electric system flexibility are the following. Conduct technological demonstration Evaluate and assess costs Determine cost recovery Establish incentives Identify and implement rule changes Provide necessary education mechanisms 8 3

138 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 9.0 Integration of Enabling Technologies 9.1 HOW MANY FLEXIBLE RESOURCES DO WE NEED? This white paper identifies flexible resources that can help to integrate VERs. It assumes that the implementation of storage options, interconnection of islands via the undersea cable and electric vehicle impacts are covered elsewhere. The discussion here relates primarily to EE, DR and the communications and control technologies required to implement EE, DR and other flexible resources. Since smart grid and renewable energy integration demonstrations are still underway throughout the country, and Hawaii utilities have very unique utility systems with limited existing flexibility, it is unclear at this time which portfolio of flexibility technologies and mechanisms will work best, at least cost in Hawaii; but it is possible to generate an initial sequence of implementation that can serve as a starting point for future planning discussions and further investigations. This portfolio of flexible technologies and mechanisms will evolve over time and the sequence of implementation can change as new and additional Hawaii specific information becomes available. 9.2 HOW DO WE SELECT THEM AND WILL THEY BE READY IN TIME? While items can be eliminated or reprioritized in the future based on the results of demonstrations and economic analyses, it is acknowledged that most smart grid capability will be justified by reliability criteria, and economic savings potential achieved from energy efficiency, not on the criteria of VER integration. However, once that smart grid capability exists, it will be useful for VER integration (to enable the flexibility/reliability functions of new technologies and mechanisms). 9.3 A STRAW MAN ORDER OF GRID FLEXIBILITY OPTIONS An implementation sequence is suggested below based on current knowledge about technology status and relative implementation costs. Items not particularly relevant to Hawaii, such as increased interconnection and expanded balancing areas are not included. This straw man sequence can serve as a starting point for planning discussions and further investigations: Use conventional fossil additions for regulation and contingency with smart grid enablers, Implement community and residential storage with smart grid enablers, Implement smart grid to allow distribution system operation and voltage/var control at the substations, Implement smart grid and smart inverter implementation to allow voltage/var control at the load and DG, Implement smart grid with DR and smart appliance control, Integrate advanced forecasting and curtailment with smart inverter implementation and DR with smart grid enablers, Implement utility scale storage with smart grid enablers, and 9 1

139 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Deploy AMI to enable smart meter data gathering and provide a Smart Grid communication platform. Energy efficiency programs can be inserted into the straw man at any point If demonstrations necessary to create technology implementation timelines and measure cost effectiveness are not fully successful, items in this list can be moved up or down in priority. The first step would be getting experts to develop a consensus of the order of implementation and to develop a process for moving items on this list up or down in priority. Modeling could be conducted at appropriate times to assist in re prioritizing the list. In our straw man list above, conventional fossil generation, pumped hydro storage and the undersea cable are the likely bookends in terms of lead time, cost, and mitigation effectiveness. Other options should be modeled/compared and their relative positions adjusted as sufficient data becomes available. At current penetrations, it is still cost effective to provide reserves with conventional resources/technologies. It is hypothesized that at scenarios with very high penetrations, storage and the undersea cable will be required. It is further hypothesized that implementation of other items listed above can progressively delay the need for storage and the undersea cable by pushing the time they are needed to higher and higher penetrations. It is not known if other techniques can totally eliminate the need for storage and the undersea cable; that will need to be determined by future studies of new techniques as they are further developed. 9 2

140 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 10.0 Proposed Implementation Roadmap 10.1 WHAT ARE THE STEPS TO MAKE THESE FLEXIBILITIES AVAILABLE The generalized steps required to assess the potential impact and implications of any particular smart grid enabling technology would follow an implementation roadmap: Demonstrate technologies capabilities and effectiveness, Develop inter utility coordination, Evaluate the costs and benefits, Determine cost allocation and recovery mechanism, Obtain regulatory approval, Establish incentives (customer, utility and other stakeholders), Identify and implement rule changes, Implement deployment planning, Implementation project to integrate solution into grid, and Provide necessary education mechanisms. This sequence applies, with necessary modification, to all technologies and mechanisms used to achieve the electric system flexibility that is discussed in the report. In addition to the assessment of the individual technologies, it is critical to understand and validate the synergies that result from the integration of the different Smart Grid solutions into a more holistic solution set. The integration and interaction of the individual solutions result in improved information, status and control capabilities across the utility and marketplace. The roadmap therefore must take into account how technologies interact and identify which are foundational to others. While there is flexibility in how solutions are implemented, there is a natural progression to some of the technologies and advanced applications. In general, those enabling technologies and mechanisms that are the easiest to implement, have the lowest initial cost and are the most mature and proven will be implemented first and others will follow. The variation on this order occurs when alternative sequencing is necessitated by the interaction of multiple technologies or dependencies therein. The key is to identify the end state desired and to establish a sequence of intermediate states (See Figure 10 1) that provide for an orderly progression from the current state to the envisioned end state. The progress and benefits of this migration must be validated and adjustments made accordingly. By taking the approach of breaking the entire list of solutions into segments, it also allows HECO, customers, regulators and other stakeholders to be in a better position to understand, absorb and respond to the significant impacts of some of the Smart Grid technologies and advanced applications, and other enabling mechanisms. 10 1

141 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Figure 10 1 Strategic Plan Path Forward The underlying solutions and mechanisms that comprise the straw man sequence above will undoubtedly change as demonstrations are successfully completed and their implementation costs defined. An important consideration will be how much smart grid implementation is justified for reliability purposes on the existing grid and how those costs are allocated against renewable energy integration and ultimately developers and owners of the generation resources. At the moment, it appears that the option with the greatest lead time and cost is the undersea transmission cable. Other options will be available sooner at lower cost and should be demonstrated and evaluated to adjust the straw man sequence. The list below captures the Smart Grid technologies and advanced applications as well as other mechanisms that are recommended for evaluation to enable the levels of renewable energy desired in the Hawaiian Islands. Smart Grid Enabling Technologies: Demand Response programs that combine proven load control programs for commercial and residential customers with more advanced load shaping solutions. Load shifting solutions such as Ice Storage solutions can better match energy requirements with off peak wind generation resources. Installation of AGC on existing conventional generation fleet to provide enhanced low power operation and improved ramping control. 10 2

142 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Distribution automation solutions to provide more monitoring and control of the distribution grid. Enhanced SCADA capabilities to provide more premise information on status of grid operating conditions and power flows. Smart Meter solutions that leverage AMI technology and MDMS applications to collect and make available meter and customer quality of service information. Static VAr (SVAR) technology & distributed static VAr compensation (DSVC) solutions that provide faster acting response to voltage disturbances Smart inverter technology on the renewable generation resources. Advanced power electronics based distribution line voltage regulation that provides finer resolution and faster response capabilities. Island wide communications networks to support the collection of distributed data and control capabilities. Multiple networks will be required to support the wide variation of service requirements. Microgrid integrated technology solutions. Smart Grid Advanced Applications: ADMS solution in the HECO control centers (initially multiple solutions on each island but eventually coordinated solutions when balancing areas may be combined) The ADMS must be integrated into the existing HECO EMS solutions to ensure seamless and complete grid management. The ADMS brings curtailment capability on the distribution grid, renewable and loan forecasting, real time loan flow and other analytic and planning capabilities. FLISR implementation demonstrating the ability to improve reliability and resiliency of grid and to allow improved availability of distributed resources. VVO application to respond to voltage fluctuations due to a more distributed generation profile on the distribution grid utilizing smart meter voltage data, DA information, SCADA information and analytic tools. DERMS module that integrates DER into the ADMS control paradigm. Foundational technologies and solutions that HECO re may require to support migration from the current grid operations to a DMS Programs Customer Engagement solutions that provide a mechanism to provide information and notification to customers based on their desired configuration. M&V analysis of EE solutions through use of AMI data Enhanced rate programs that encourage load shifting to better match energy consumption with renewable resource generation 10 3

143 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Regulatory and Policy Allow demand response and demand shaping solutions to substitute for traditional reserve requirements Determine the acceptable level of cost allocation and impact due to the anticipated costs to implement and integrate fully the smart grid solutions. Integrate renewable policy required SG solutions with HECO plan The following figure presents a sequenced view of these recommendations, indicating the interdependencies involved, level of technological maturity or anticipated cost and complexity of implementation. The following chart provides a straw man sequencing of the key Smart Grid technologies and advanced applications and other enabling mechanisms discussed in this report as being critical to Hawaii meeting its twin goals of the Hawaii Clean Energy Initiative and achieving 1 GW of additional renewable energy generation capacity. Figure 10 2 Recommended Smart Grid Solution and Enabling Mechanisms Sequencing The information illustrates the interdependencies of many of the technology solutions and while the actual start dates and length of implementation can be flexible and variable depending on level of funding and effort, the sequencing is the important association to obtain from this chart. 10 4

144 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 10.2 SOLUTION COSTS Evaluating the costs for the recommendations included in this report will be vital to understanding the cost/benefit relationship of the individual technologies and measures as well as the potential synergies amongst the solutions. This information is also necessary for prioritizing investment resources and assembling a detailed implementation strategy. Implementation details, and thus solution costs, are utility specific and have many dependencies. These dependencies include: The existing technology landscape in the utility and in the case of Hawaii, each HECO operating company has a different current configuration. Organizational design and change management the ability of a utility to fully leverage the new smart grid technologies and advanced applications is directly related to the ability and willingness to modify current utility approach, job design and classifications, and work practices. IT policies and standards related to the implementation and management of the advanced applications The level of field force automation in existence or planned that would integrate with the proposed technologies. One of the first tasks identified in this white paper is to work closely with HECO to determine that relevant technology initiatives are underway or planned that would impact the value or implementation ability of the proposed solutions. This is also a critical component of understanding what unique conditions could exist that would impact costs and benefits of the proposed solutions. While the quantification of detailed implementation costs and specific benefits of the various recommendations is outside the scope of this paper, a high level understanding of some solution costs can be developed using industry standard costs and estimates. Based on prior industry experience, order or magnitude estimates for some of the Smart Grid technologies identified herein are provided below: Table 10 1 Smart Grid Solution Reference Cost Estimates Cost Smart Grid Technology Estimate Smart Meter mass deployment $100M DMS initial implementation, including foundational systems work $20M MDMS implementation $8M DMS advanced applications $10M Distribution Automation $50M 10 5

145 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Costs associated with Substation SCADA, Distribution Automation and other technologies are difficult to estimate with any accuracy without taking into account unique conditions, design criteria and plans of HECO. The existing level of automation of the electric grid on the islands varies significantly with Maui and Hawaii having significantly more existing field automation than Oahu. Oahu however has more experience with the complexities of advanced applications and will be more prepared to implement these solutions without substantial changes in organization and resources. The Smart Grid technologies and advanced applications identified herein also have many system reliability, operational, power quality and customer services benefits beyond those related to renewable energy support and penetration. These benefits range from those well defined and proven for smart metering and distribution automation to those still being validation such as VVO and DMS. It is not appropriate to estimate the benefits to HECO and the State of Hawaii without looking at specific HECO conditions and opportunities. A recent study generated by GE for Hawaii s RSWG provides cost per kilowatt for a wide array of resource types.105 In the overall spectrum of energy resource costs, the extremes are represented by energy storage and traditional generation. Storage enables energy to be consumed at a different time than it was generated. Since storage does not create energy, its cost is above and beyond the cost to generate the energy which is stored, resulting in one of the highest cost options. At the opposite end of the spectrum are traditional generation resources, which can provide low cost alternatives by operating existing resources more flexibly. Such flexibility could be changing the speed of ramping, or reducing the minimum operating load or similar operational changes. Depending on the existing generation type, age, original design, etc. such changes can potentially have a very minor initial cost and a very small change in O&M cost. Or, the costs may be larger but still relatively low compared to other approaches to adding system flexibility SECTION SUMMARY The individual Smart Grid solutions do not operate in isolation but work cooperatively to create a more reliable and flexible electricity grid. These benefits and synergies are obtained by integrating the solutions together to share information and leverage capabilities. The relative sequencing of the solutions is important as the benefits envisioned will not flow if required foundational solutions or necessary precursors are not implemented in advance. Given the length of time required to fully implement many of the recommended solutions, it is recommended that the actual demonstration activities, regulatory and financial analysis and project planning be performed in parallel even if actual project implementation may be sequenced. Given the uncertainties involved with many of the advanced technologies, it is essential that appropriate project integration and schedule risk be addressed. The recommendations as to roadmap content and schedule are provided based on best current understanding of the challenges and barriers to be overcome as well as assessment of the benefits of the various initiatives and level of complexity and readiness. The assumptions underlying these recommendations must be validated and where changes are necessary, modifications of the roadmap will be required. 105 General Electric, Ancillary Services Definitions and Capability study, Part 1, Tasks 1 2 Final Report, December

146 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 11.0 Summary of Overall Recommendations or Proposed Next Steps Based on the assessment of existing and emerging technologies and programs, the barriers to their adoption, and possible solutions to overcome those barriers, this report provides recommendations that will help Hawaii realize the benefits of these technologies and programs to integrate high levels of as available energy into the Hawaiian Islands electric grids. The recommendations include: Identification of Smart Grid solutions that can enable renewable generation increases Changes to policies, regulations, programs and initiatives the support renewable generation A prioritization of technologies and programs having the greatest potential to help integrate large amounts of renewable energy into the grid, and A high level road map for implementation of the recommendations. The next steps required to promote these recommendations include the review of the recommendations in the broader context of research and recommendations undertaken on associated work efforts. A wider prioritization must be performed to identify the highest potential impacts and while Smart Grid solutions provide real enabling benefits, they must compete with other technologies and programs not covered in this report. It is critical that the recommendations contained herein be shared with HECO and a dialogue developed to identify which Smart Grid technologies and advanced applications may already be part of technology roadmaps underway or envisioned at the HECO Companies and which are not being considered. For those solutions that already form part of HECO s Smart Grid roadmap and plans, the value to increased penetration of renewables can be derived and used to generate additional support for HECO s plans. Where there are gaps between what HECO is investigating and what is recommended herein, it will be necessary to bridge that difference. The impact of the recommended Smart Grid technologies on HECO s initiatives is a critical factor in implementation success and given the time required to integrate these technology opportunities into other HECO initiatives, the timeline provided in this report for eventual implementation of these solutions is very likely to be lengthened. 11 1

147 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII 12.0 Conclusion This report presents the barriers and challenges to reaching the energy efficiency and renewable penetration goals of the Hawaii Clean Energy Initiative and the goal of attaining 1 GW or renewable generation on the Oahu grid. The need for mechanisms to provide increased flexibility and enhanced reliability of the electric grid are presented as well as the benefits that these options provide. The potential of Smart Grid technologies and advanced applications to enable the desired renewable energy levels is discussed along with a characterization of the commercial maturity and availability of the various solutions. Many of the Smart Grid technologies presented have reached a level of technical maturity but have not attained a similar level of commercial maturity, at least as they relate to the costs and benefits associated with using the solutions for the purpose of supporting renewable generation. There is a need to work with HECO to build a broader business justification and benefit analysis for the Smart Grid solutions identified. Some of the technologies require further demonstration of their effectiveness or robustness and it s important to keep in mind that technology is rapidly changing. Additionally, no utility has implemented the full range of Smart Grid solutions envisioned herein and consequently there is no real world example demonstrating an optimal mix of technologies on any particular grid. The complexity and organizational impact of integrating all the enabling technologies cannot be stressed enough. In addition to the technical issues of the solutions, therefore, the ability of HECO to implement the significant changes required and the ability and willingness of HECO s customers to pay for it must be determined. 12 1

148 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Appendix A. Glossary of Abbreviations and Terms AC air conditioner ACCEE American council for an Energy Efficient Economy AMI Advanced Metering Infrastructure Automatic Generator Control(AGC) AGC systems control generators to increase or decrease electrical output to match changes in load. Changes in load are typically not predicted or scheduled and must be followed by generation reserve capacity that is online and grid synchronized 106. Ancillary Services Broadly, ancillary services are the regulation and reserves utilities supply to reliably balance generation and load. However, ancillary services are controversial because they are defined and used differently in each regional market. They are dependent on the characteristics of the system. In fact, no one had heard of ancillary services before 1995 when FERC defined them to enable new electricity markets following utility deregulation. FERC defined ancillary services as those necessary to support the transmission of electric power from seller to purchaser given the obligations of control areas and transmitting utilities within those control areas to maintain reliable operations of the interconnected transmission system. Others describe ancillary services as the functions performed by the equipment and people that generate, control, transmit, and distribute electricity to support the basic services of generating capacity, energy supply, and power delivery. 107 In 2012, HNEI sponsored, with support and guidance from the Hawaii RSWG, a study by General Electric to provide a standardized set of ancillary services, along with their associated definitions, to reflect the operation needs of the Hawaii system. 108 AC Alternating current ACEEE American Council for Energy Efficient Economy AOS Advanced Operating System AMI Advanced Metering Infrastructure APS Arizona Public Service ARRA American Recovery and Reinvestment Act ASHRAE American Society of Heating, Refrigeration, and Air Conditioning Engineers AZ Arizona BA Balancing Authority BAPV Building Applied Photovoltaics 106 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November Eric Hirst and Brendan Kirby, "Electric power Ancillary Services," Oak Ridge National Laboratory (ORNL/CON 426), Oak Ridge, Tennessee, February HNEI, "Report on Business Case in Hawaii for Storage Options," US Department of Energy, Office of Electricity Delivery and Energy Reliability, Honolulu, HI, October A 1

149 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII BEEM Business, Energy Efficiency, Measures BESM Business Service and Maintenance BGE Baltimore Gas & Electric BHTR Business Hard to Reach BMS Building Management Systems CAES Compressed air energy storage CA ISO California Independent System Operator, the independent system operator for the CA electric grid Capacitor an energy storage device used on electric distribution systems that provides voltage support to the grid by injecting reactive power during voltage fluctuations CBEEM Customer, business energy efficiency measures CHP Combined heat and power CIDP Commercial & Industrial Dynamic Pricing Program CPP Critical Peak Pricing CPR critical peak time rebate CPUC California Public Utility Commission CVR Conservation voltage reduction or regulation C &I Commercial and Industrial DA Distribution Automation DADs Demand Response Availability Data System DC Direct Current DG Distributed generation DMS Distribution management system Distributed Generation(DG) What constitutes distributed generation is not perfectly defined in the industry; however, it is generally understood to consist of electric power resources connected to the distribution system and frequently limited in size, for example not larger than 20MW. Demand Response (DR) Systems that react to signals from utilities or third parties designed to shed load under predefined conditions, for specific periods of time and in response to either pricing or reliability disruptions Demand Response 2.0 Next generation demand response A 2

150 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII DER Distributed Energy Resource DERMS Distributed Energy Resource Management System, Provides an integrated management application for distributed energy resources on the distribution grid. It may be a stand alone application or a module in a DMS solution. DLC Direct Load Control DOE Department of Energy DSM Demand Side Management DMS Distribution Management Systems DSVC Distributed Static VAr Compensation EE Energy efficiency EEPS Energy Efficiency Portfolio Standard EOSP East Oahu Switching Project EOTP East Oahu Transmission Project EIA Energy Information Administration EILS Emergency Interruptible Load Service EMS Energy management system EPRI Electric Power Research Institute ES energy storage ESMIG European Smart Metering Industry Group ESB Enterprise service buses EV Electric vehicle ERCOT Electricity Reliability Council of Texas FERC Federal Energy Regulatory Commission Flexibility/ Reliability Functions Traditionally utility system operators have obtained most of the system flexibility and stability support from the performance capabilities of and specific A 3

151 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII services provide by traditional generators. When these flexibility/reliability functions are provided within a regional market, they may be designated as ancillary services. 109 Flexible Resources Sources of system flexibility, examples of newly emerging sources of system flexibility are DR, energy storage and electric vehicles. 110 FLISR Fault Location Isolation and Service Restoration, a module in the new evolution of DMS solutions FPL Florida Power & Light GIS Geographic Information System GW Gigawatt HAN Home Area Network HCEI Hawaii Clean Energy Initiative HECO Hawaii Electric Company HELCO Hawaii Electric Light Company HNEI Hawaii Natural Energy Institute, School of Ocean and Earth Science and Technology, University of Hawaii. ICE internal combustion engine Ice Storage A load shifting solution that makes ice during off peak periods and uses the stored cooling capability for air conditioning during peak periods. IED Intelligent electronic devices_ IECC International Energy Conservation Code IESNA Illumination Engineering Society of North America IHD In home display IEEE Institute of Electrical and Electronics Engineers IED intelligent electronic device IHD In home display IOU Investor owned utility 109 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November 2010 A 4

152 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII IPP Independent Power Producer IRP Integrated Resource Plan ISO Independent System Operator IT Information technology KIUC Kauai Island Utility Cooperative KW Kilowatt LAARs Loads Acting As Resources LBNL Lawrence Berkeley National Laboratory LED Light Emitting Diode LTC Automatic Load tap changing transformer Market Transformation The term market transformation is the strategic process of intervening in a market to create lasting change in market behavior by removing identified barriers or exploiting opportunities to accelerate the adoption of all cost effective energy efficiency as a matter of standard practice (ACEEE). 111 MECO Maui Electric Company MW Megawatt M&V Measurement and verification MDMS Meter data management system NERC North American Electric Reliability Council NPV Net Present Value Net Present Value Positive for energy efficiency The sum of the present values of the expected annual energy savings are greater than the initial energy investment, the NPV of the project will be positive and should be undertaken OG&E Oklahoma Gas and Electric OpenADR industry standards for Auto Demand Response OMS Outage management system PCT Programmable Communicating Thermostat 111 American Council for an Energy Efficiency Economy ( transformation) A 5

153 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII PHEV Plug in hybrid electric vehicle PJM Pennsylvania, New Jersey, Maryland energy market PBF Public Benefit Fund PEV Plug in Electric Vehicle PHEV Plug in Hybrid Electric Vehicle PG&E Pacific Gas & Electric PUC Public Utility Commission PV Photovoltaic Ramping large changes in electricity generation, for example solar or wind generation Recloser also known as an autorecloser, is a circuit breaker which can automatically close the breaker after it has opened due to a fault; used to interrupt momentary faults on overhead distribution lines. REEM Residential Energy Efficiency Measures Regulator a device used in substations or on the distribution grid that provides voltage regulation capability by adjusting taps on an internal transformer to change the transformer ratio between high and low sides of the regulator. Regulation Continuous response of reserves under Automatic Generation Control (AGC) that are deployed to correct minute to minute deviations in system frequency or to return system frequency to desired range following system disturbance. Regulation is a FERC defined service, most frequently obtained from energy markets. Reliability Standards Working Group (RSWG) In 2010, the Hawaii Public Utilities Commission indicated its intent to establish a RSWG; that group began working together in July Since that time the RSWG has been meeting and working diligently to produce policy and technical recommendations that will facilitate the increased use of renewable energy in the islands without compromising grid reliability. As of mid December 2012, the RSWG members had made a significant amount of progress at developing a shared, sophisticated technical understanding of the relevant issues and developing specific work products that can help the state's utilities, generators, and stakeholders operate the grid with high reliability and high levels of renewable generation. REP Retail Energy Provider, term used in Texas RF Radio Frequency RESM Residential Energy services and Maintenance RHTR Residential Hard to Reach A 6

154 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII RTO Regional Transmission Operator RTP Real Time Pricing SEGIS Solar Energy Grid Integration System SCE Southern California Edison SDG&E San Diego Gas & Electric SRP Salt River Project SAI SAIC S&C South Carolina SCADA System Control & Data Acquisition SMUD Sacramento Municipal Utility District System Flexibility System Flexibility is defined as the ability of supply side and demand side resources to respond to system changes and uncertainties. 112 Smart Grid a term referring to the collection of enabling technologies and advanced applications that are based on digital technologies and provide a two way flow of information to better monitor and manage the electric grid. SVAR Static VAr technology TES Thermal Energy Storage TOU Time of use TRM Technical Resource Manual TVA Tennessee Valley Authority UL Underwriters Laboratories USB Universal Serial Bus VaasaETT an energy think tank and consultancy focusing on issues relating to customer behavior and psychology in energy and utilities markets. Variable Energy Resource (VER) Renewable energy generation resources such as wind and PV use weather based fuel, the availability of which can vary over time. VERs are generators that vary output as the local renewable energy resource increases or decreases. 112 North American Electric Reliability Corporation (NERC), "Potential Reliability Impacts of Emerging Flexible Resources," NERC, Princeton, NJ, November 2010 A 7

155 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII VAr volt ampere reactive; also written as Var, VAR, or var. VVO Volt/VAr optimization WiFi a wireless local area network (LAN) open standard technology based on the IEEE standard ZigBee ZigBee is a wireless open standard technology for low cost, low power wireless networks. The ZigBee standard operates on the IEEE physical radio specification and operates in unlicensed bands including 2.4 GHz, 900 MHz and 868 MHz. A 8

156 Hawaii Natural Energy Institute TOOLS TO ENABLE RENEWABLE ENERGY IN HAWAII Appendix B. Power Generation Resource Mixes ISLAND OF ISLAND OF ISLAND OF ISLAND OF HAWAII MAUI OAHU KAUAI Utility Name HELCO MECO HECO KIUC Firm Capacity 287 MW 262MW 1,756 MW 125 MW Units Puna Geothermal Ventures (27 MW off peak, 30 MW on peak)(no droop response, off AGC) Three steam units: Hill 5 14 MW, Puna 15 MW, and Hill 6 20 MW (AGC) 35.5MW One combined cycle unit: HEP 28.5 MW (AGC)Two small steam units (Shipman 3&4, 7.5 MW each) operated 2 shifts (AGC) The second train of HEP (in this configuration, produces up to 60 MW) (AGC) Three simple cycle gas turbines (CT3, CT4, CT5 20 MW each) (AGC) Kahului Power Plant 4 steam units 37.6 MW Hana Substation 2 DG Diesel Engines 2 MW Maalaea Power plant 15 Diesel Units 2 Dual train Combined Cycle Units MW IPP Puunene Mill 3 steam turbines, firm baseload 12 MW Honolulu oil 113 Waiau oil 499 Kahe oil 651 CIP oil 120 IPPs H Power waste to energy 46 MW Kalaeloa Partners oil 208 MW AES Hawaii coal 180 MW Port Allen Diesel D1 & D2: 4 MW Port Allen Steam Turbine S1: 10 MW Port Allen D3 5: 8.2 MW Port Allen Combustion Turbine GT1: 19.2 MW Port Allen Combustion Turbine GT2: 23.7 MW Port Allen Diesel D6 & D7: 15.7 MW Peaking/emergency units 14 small diesels ( nine 2.5 MW, four 1 MW, one 2 MW) (off AGC, remote start) Two simple cycle gas turbines (CT1 11 MW, CT2 14 MW) (AGC) Port Allen Diesel D8 & D9: 15.7 MW Kapaia Steam Injected Gas Turbine: 27.5 MW Waihi Hydro: 0.5 MW B 1

157 ELECTRICITY STORAGE FOR HAWAI I Prepared for the: State of Hawai i Department of Business, Economic Development and Tourism Under Supplemental Contract No. 1 for Contract #60514 for the NETL Grant Project Task 1, Subtask 5, Energy Storage Element Prepared by: Abbas Akhil Renewable Energy Ventures, LLC Submitted by the: Hawai i Natural Energy Institute July 2013

158 June 2013 Electricity Storage for Hawaii: Technologies and Issues to Facilitate High Penetration of Renewables Prepared for Hawaii Natural Energy Institute by Abbas Akhil Renewable Energy Ventures, LLC

159 Electricity Storage for Hawaii: Technologies and Issues to Facilitate High Penetration of Renewables Contents Introduction... 4 Executive Summary... 4 Section I: Need for storage with Renewables... 6 Section II: How Storage Acts... 9 Short duration Storage 1 second to 1 minute... 9 Mid duration Storage 1 minute to 30 minutes Long duration Storage 1 hour to 6 hours Section III: Storage Technologies: Description, Commercial Status Short and Mid Duration Storage Technologies Advanced Lead acid (Adv. Pb acid) Batteries Lithium Family of Batteries Flywheel Energy Storage Systems Long Duration Storage Technologies Pumped Hydro Sodium Sulfur Batteries Sodium Nickel Chloride Flow Batteries Zinc Bromine Vanadium Redox Batteries Compressed Air Energy Storage (CAES) Section IV: Storage Technologies: Cost of Systems Section V: Energy Storage Integration in Hawaii Grids Types of storage deployments Technology Diversification: Utility side Storage System Deployments Distributed Storage Systems Deployment Section VI: Barriers for Energy Storage Deployment in Hawaii Economics/Markets Regulatory and Policy Drivers

160 Section VII: Technical Barriers Environmental constraints Public Perception/Cultural Section VIII: Risks and Uncertainties Section IX: Lessons Learnt and Performance Assessment of Existing Storage Systems Section X: Conclusions List of Figures Figure 1: Ramp Event At Big Island Wind Farm... 7 Figure 2: Simplified Storage System Response In The Lanai Grid Figure 3: The Sequential Actions Of Frequency Controls Following The Sudden Loss Of Generation Figure 4: Xtreme Power Advanced Lead Acid System At The Kaheawa, Maui, Wind Farm 15 Figure 5: Current And Planned Li Ion System Demonstrations Figure 7: Beacon s 1 Mw Smart Energy Matrix Plant Figure 9: Okinawa Seawater Pumped Storage Figure 10: Nas Battery Module Components Figure 11: Fiamm s 220 Kwh Pre Production Unit At Duke Energy Site Figure 13: Trailer Mounted 200 Kw/400 Kwh Zinc Bromine Battery System Figure 14: Premium Power Transflow 2000 System Being Unloaded At Kotzebue Figure 15: Prudent Energy 600 Kw/3,600 Kwh Vrb System Figure 16: Schematic Of A Caes Plant Figure 17: Value Estimates For Various Storage Services List of Tables Table 1: Cost Estimates For Select Storage Technologies Table 2: Regulatory Agency Rules And Their Impact On Storage Deployment

161 Introduction This report reviews the characteristics, commercial status and cost of various storage technologies and how they act, or could act in Hawaii electric grids to facilitate high penetration of renewables. It is intended to inform policy and decision makers the potential barriers, uncertainties, and regulatory initiatives that may hinder or facilitate even higher penetration of renewables as the Hawaii grids adapt to their operating environment. Executive Summary This report is organized in nine major sections to review the storage technologies and discuss various issues related to the use of energy storage to facilitate a high penetration of variable renewable resources. Sections I and II discuss the need for storage and how storage acts to mitigate the effects of variability of renewable generation output on the host grid. While there are different ways of discussing the technical aspects of the use of storage to achieve the desired effect, this report attempts to explain this in terms of three time domains that storage needs to act in: short and mid duration and long duration. This allows the various storage technologies to be classified in the time domain that best matches their performance characteristics. The short and mid duration storage technologies include Advanced Lead acid, Li family of batteries, and Flywheels. The long duration technologies include a wider selection: Pumped hydro, high temperature batteries: Sodium Sulfur, Sodium Nickel Chloride; flow batteries: Zinc Bromine, Vanadium Redox and Compressed Air Energy Storage. Section III discusses the storage technologies listed above in greater detail including their commercial readiness and identifies key vendors who are supplying these storage systems. It is to be noted that traditional pumped hydro plants are generally too large and pose significant environmental concerns of land and water use and are deemed unsuitable for Hawaii. However, a variant of this technology using seawater as the storage medium is discussed because these can be built in smaller sizes without significant land requirements and zero impact on fresh and potable water. Sections IV include storage system cost information and explains some of the difficulties encountered in generalizing storage costs because of the unique nature of storage systems and how they are used in the electric grid. The cost information presented in this section is summarized from a soon to released Handbook compiled by DOE and EPRI and is the most recent information available for the storage technologies of interest. The costs are not absolute values, but ranges expressed in $/kw and $/kwh. Sections V through IX discuss integration issues, barriers to deployment, risks and uncertainties, and lessons learnt recommendations for Hawaii specific applications. Three regulatory and policy drivers that could facilitate larger deployments of storage systems 4

162 are discussed in Section IV. These include establishing requirements for spinning and regulating reserves and allowing storage to provide these services; modifying feed intariffs to recover addition of storage to renewable projects; reviewing reliability metrics and using storage to relax limits on renewable penetration; and modifying the Renewable Energy Infrastructure Program to allow recovery of storage related studies and projects undertaken to support renewable projects. 5

163 Section I: Need for storage with Renewables Normal electric grid operations are based on the fundamental assumption that the generation sources within its boundaries have a predictable certainty of providing the constant power requirements determined by the system operator. Abnormal operational conditions exist when there is an unexpected loss of a generation unit or other system component that requires rapid addition of reserve power to meet the load requirements and maintain stable operation of the grid. Similar conditions are created when a large penetration of renewable resources occurs in the grid, because their power output is not constant and depends on the steady availability of wind to support wind generation or clear sky conditions in the case of photovoltaic (PV) generation. The output from both wind and PV fluctuates rapidly when the wind shifts or there are intermittent clouds shadowing the PV array. These rapid and sudden fluctuations or ramps require other generation sources in the grid to modulate their power up or down in the opposite direction to the renewable source to maintain a stable grid. The secondary impact of large penetration of renewables in the grid is the inability of the fossil generators to operate at or below their minimum power generation limits to maximize the intake of all renewable energy available to the grid at any given instant. Maximizing the renewable energy intake while maintaining fossil based generation at their minimum or optimal output is a delicate balancing act orchestrated in real time by the grid operator. Grid operators balance these constraints to keep the voltage and frequency within normal ranges. When the renewable generation exceeds the fossil generation on the grid, the operator has limited choices either spill or curtail the renewable energy production, or turn down the power generation of the fossil units sufficiently to absorb all the available renewable generation to maintain a stable state. In some instances, especially in the smaller size electric grids for islands, this condition is more critical, because the fossil units can only be turn downed to fixed minimum limits and cannot be operated below these set points. These conditions are illustrated in Figure 1 which shows a typical ramp event at the Tawhiri wind farm in the Big Island and its effect on the rest of the HELCO system 1. The green color trace in the lower part of the plot shows the wind farm output that is steady at 10 MW until there is a sharp down ramp and the output dips to almost zero megawatts. There is a consequent dip in the system frequency from the nominal 60 Hz to below 59.4 Hz and the diesels ramp up their generation to stabilize the grid frequency to its nominal 60 Hz. 1 Source: Presentation to HCEI, October 20, 2010, by Dora Nakafuji, HECO, Forecasting Activities: Improving WindSENSE and Solar Forecasting Capabilities for Control Rooms 6

164 Figure 1: Ramp Event at Big Island Wind Farm The second condition, where fossil units are turned down to absorb more wind and reduce curtailment can be described by using the MECO example in Maui when future additions to the existing wind farms are expected to total 72 MW and will exceed the 70 MW s projected minimum system load. In this scenario, MECO has to substantially curtail wind generation and run the minimum required fossil generation at low and inefficient set points. A recent study 2 by Sandia National Laboratories examining the various options for MECO found that storage would reduce both wind curtailment and the annual cost of producing energy due to the more efficient operation of MECO s diesel units, which would have otherwise operated at less efficient set points. Electricity storage is being effectively utilized to mitigate or eliminate these conditions and provide other services to the grid such as time shifting of and ancillary services. Electric storage applications can be characterized as acting over three time domains in island grids to mitigate the effects of high penetration of renewables and provide other operational 2 Maui Energy Storage Study by James Ellison, Dhruv Bhatnagar, and Benjamin Karlson, Sandia National Laboratories, December

165 needs of the grid to enhance its stability and reliability. These three time domains can be characterized as short mid and long duration as described below: Short duration storage acts to mitigate the effects of rapid up or down ramps in power output as shown in Figure 1 and provides stability in the 1 second to 1 minute time window when the renewable generation is varying rapidly due to solar or wind conditions. Mid duration storage acts in the 1 minute 3 to 30 minute time window and mitigates the effects of variations in momentary changes in the grid frequency due to changing loads and as rapid reserve to compensate the sudden failure of a generator. Long duration storage acts in the 1 to 6 hour window to store energy from wind or PV plants that would otherwise be curtailed or spilled, and, where applicable, time shift energy for operational or economic advantage. This report discusses these three uses of storage in the Hawaii grids to facilitate the increased penetration of renewable resources including wind and PV. The report includes a discussion of the storage technologies, their commercial status, barriers and challenges for deployment and recommendations to support wider deployment in Hawaii. 3 The operational capability of storage technologies that are available today spans both the short- and mid-duration time domains. The characterization of time domains in this report is mostly to clarify their functional use in the grid. 8

166 Section II: How Storage Acts Short duration Storage 1 second to 1 minute The plot in Figure 2 is a hypothetical representation illustrating how the fast response of an energy storage system compensates for the sudden drop of power of a PV system when clouds shadow the array. This simplified example uses the La Ola PV plant in Lanai as an example with a hypothetical 375 kw battery storage system to control its ramps. In Figure 2, it is assumed that the Lanai system load is 5.5 MW at T = 0 seconds. At that instant, the PV array is producing 1.2 MW with the remaining 4.3 MW being supplied by diesel generation and the battery storage system is in idle mode. At T = 10 seconds, the PV output drops suddenly due to cloud cover producing a shortage of power in the grid. The diesels do not increase their output as rapidly as the drop in PV output occurs. However, the storage system detects this grid condition and responds by instantaneously discharging its rated capacity of 375 kw for the next 30 seconds; this allows sufficient time for the diesels to gradually increase their output to make up for the lost generation from the PV plant, and the storage system goes to a standby mode again at the T = 40 second mark. In the real life implementation, the battery storage system installed at La Ola PV plant is much larger with a rating of is MW 4. However, its functionality is similar to this illustrative example and its ability to smooth the variability in the PV output is working as designed. Similar short duration storage systems that mitigate the variability of the PV and wind output are successfully operating on Maui, Oahu, Kauai and Hawaii. 4 The 375 kw size of the storage system shown in a the example is larger than the 250 kw battery storage size originally specified in the Power Purchase Contract between MECO and Lanai Sustainability Research, LLC, the owner and operator of the La Ola PV station in Lanai. The 250 kw battery size was later increased to its present MW rating as a result of studies performed by Sandia National Laboratories, National Renwable Energy Laboratory and Sunpower that showed that the larger battery size was necessary to control the ramp rate of the 1.2 MW PV plant within MECO s specified limits. 9

167 Figure 2: Simplified representation of Storage System Response in the Lanai Grid Mid duration Storage 1 minute to 30 minutes The sudden loss of a generating unit triggers both a drop in frequency and an immediate need to replace the power lost with other reserves. The rate of frequency decay following the loss of generation is proportional to the mechanical inertia of the units that remain online and is faster when the electric grid is made up of smaller generating units. The typical system response as soon as a generation unit is lost till it stabilizes minutes later is shown in Figure 3, in which the upper portion represents the frequency excursion following the outage. The initial dip triggers a governor response shown in the lower plot. This is followed by the Automatic Generation Control (AGC) response to stabilize and restore the frequency to the desired 60 Hz. The sustained recovery of the grid then relies on operating reserves that replace the lost generation and return the grid to a normal state. Energy storage systems have been effectively used both in mainland and island electric grids to provide the reserve power necessary to restore the electric grid to normalcy 10

168 following the loss of a generating unit. Their fast response characteristic faster than the response of any conventional generator is perfectly matched to provide the energy in the first second and subsequent minutes that are critical to stabilize the frequency and maintain continuity of supply to the customer. Studies have also shown that the fast response of batteries and flywheels is at least twice as effective as a comparable conventional generator 5, which means that a 20 MW storage system can stabilize the system as effectively as a 40 MW fossil fired unit following a contingency. Battery and flywheel energy storage systems designed for reserve power can also provide frequency regulation on a 24/7 basis. The combination of these two grid services into a single storage system makes optimal use of the inherent properties of these technologies. The frequency regulation duty requires a continuous shallow discharge/recharge operation of the storage system and has minimal life limiting impact on the battery and even less so on the flywheel. It is the reserve power duty which requires a deep discharge that impacts the operational life of the battery system more than the shallow frequency regulation duty. 5 Energy Storage a Cheaper, Faster and Cleaner Alternative to Conventional Frequency Regulation. Whitepaper by the California Energy Storage Alliance (CESA) ( _regulation.pdf 11

169 Figure 3: The Sequential Actions of Primary, Secondary, and Tertiary Frequency Controls Following the Sudden Loss of Generation and Their Impacts on System Frequency 6 Long duration Storage 1 hour to 6 hours Long duration storage spanning from one to six hours can capture energy from renewables that would otherwise be spilled or curtailed due to low system load or other operational constraint on the utility side. Unlike the short and mid duration storage use described in the prior section, long duration storage is designed for sustained discharges over longer durations and the power (MW) to energy (MWh) ratio of such storage systems is generally 1:5 or greater. 6 Use of Frequency Response Metrics to Assess the Planning and Operating Requirements for Reliable Integration of Variable Renewable Generation. Joseph H. Eto, Principal Investigator Lawrence Berkeley National Laboratory, et al. ( act/reliability/frequencyresponsemetrics report.pdf ) 03/25/

170 Battery storage systems are operating on Maui and Oahu to mitigate the curtailment of wind farm output. These are early battery storage projects in Hawaii and have lower power to energy ratios than indicated above, but are still mitigating curtailments to a certain degree. Recent studies by NREL 7, HNEI 8 and Sandia Labs 9 have shown that more powerful and longer duration storage systems can prevent even more curtailments as the renewable generation component increases in the future. Other reports 10,11 have noted a wide range of services that energy storage can provide to the grid and have classified these services according to their location within the grid as generation, transmission or distribution services. The time duration based classification used in this report differs from that approach and is specifically used to characterize how storage could promote the penetration of renewable resources in Hawaii. However, the wider range of storage services for other grid applications described in those reports could be of value to Hawaii as Smart Grid and other adaptive changes occur to Hawaii s electric grids. Section III: Storage Technologies: Description, Commercial Status Each of the three storage types described above has unique attributes and utilizes technologies that have correspondingly unique characteristics. The short and midduration storage is best supplied by Lead acid (Pb acid) batteries, the lithium family of batteries, and flywheels. The long duration storage requirements are best met by pumped hydro, compressed air energy storage (CAES) or high temperature batteries such as Sodium/Sulfur (NaS) 12 and Sodium Nickel Chloride (NaNiCl) or flow batteries such as Zinc Bromine (ZnBr) and Vanadium Redox (VR). The technology characteristics, maturity and experience as well as the system vendors working in these technology types are described in the following section. 7 Hawaii Solar Integration Study: Final Technical Report for Maui, December Oahu EV Charging Study, September Maui Energy Storage Study, December Battery Energy Storage for Utility Applications: Phase I Opportunities Analysis, SAND , and, Energy Storage for the Electricity Grid: Benefits and Market Potential Assessment Guide, SAND DOE/EPRI Electricity Storage Handbook in Collaboration with NRECA ; due for release in June/July NaS is a generic term used to refer to sodium-sulfur batteries based on atomic symbols Na and S. Whereas the capitalized NAS is a registered trademark for NGK s sodium-sulfur battery systems. 13

171 Short and Mid Duration Storage Technologies Advanced Lead acid (Adv. Pb acid) Batteries Lead acid batteries have been used in the electric grid in one form or the other for over 100 years to energize switchgear in distribution substations or as back up to auxiliary systems in power stations. Starting in the mid 1980 s, large lead acid based storage systems were built to provide grid support for load leveling, spinning reserve and frequency regulation to electric utility grids in North Carolina, California, and Puerto Rico and in some overseas grids as well. The lead acid batteries that were used at that time have been further refined in recent years into advanced, robust versions with improved cycle life, and higher power and energy capabilities. These modern versions of lead acid batteries are being referred to as advanced lead acid technologies because of these recent innovations. Their chemistries have been modified with carbon or other additives and some changes to the design of internal components contribute to their improved performance. Some Adv. Pb acid batteries incorporate capacitor like internal components to meet high power application requirements, especially suited for short and mid duration ramping duty to support renewable technologies. Since lead acid is a widely used battery type, its end of life disposal and recycling processes are also well established. While the other, non lead acid batteries are still in the process of establishing a reliable recycling chain, lead acid batteries, including the advanced variants, have a reliable and proven track record of recycling all their used product from commercial applications. Procurement contracts for utility scale stationary storage systems using lead acid batteries have contractual clauses that ensure that the batteries are sent to a regional recycling center at the end of their operational life. Adv. Pb acid turnkey storage systems for stationary electricity storage applications are available from Xtreme Power, Ecoult/EastPenn, and Axion Power International. Each supplier has a different implementation of the advanced lead acid chemistry and each developer is targeting a specific niche market. Xtreme Power systems have found a strong foothold in Hawaii with an early installation in Maui, followed by systems in Lanai, Kauai and Oahu. Ecoult/East Penn has demonstrations in the mainland with promising results and is seeking opportunities in Hawaii. Their mainland demonstrations are showing promising results and their system design and performance is backed by East Penn s long history of in house R&D and manufacturing expertise in traditional lead acid batteries. 14

172 Figure 4: Xtreme Power Advanced Lead acid System at the Kaheawa, Maui, Wind Farm Lithium Family of Batteries Within the past few years, there has been a strong interest in Lithium ion (Li ion) batteries for stationary applications due to their high energy density, high power and robust cycle life. The Li based battery chemistries are already commercially mature for consumer electronic applications, and Li ion is positioned to be the preferred technology for plug in hybrid electric vehicle (PHEV) and electric vehicles (EV), which will use larger format cells and packs with capacities of 15 to 20 kwh for PHEVs and 50 to 85 kwh for EVs. The more common type of liquid Li ion cells are cylindrical or prismatic cells found in the battery packs of laptops, notebook computers, digital cameras and other consumer electronic products. A variation using prismatic polymer Li ion cells is also used for small portable products such as cellular phones and MP3 players. On aggregate basis Li ion batteries used in consumer electronic products make up the bulk of the worldwide production volume of 10 to 12 GWh per year. This production volume is projected to grow to 30 GWh by 2015 and this growth could result in potentially lower cost battery systems for stationary utility scale storage system deployments. Many utility scale systems have already been deployed in the field and are providing valuable experience in siting, grid integration and operation of Li based systems. Figure 5 shows some of the Li ion energy storage system deployments in the past two years. The stars represent the most significant projects; several additional Li ion projects are underway in other locations. The major manufacturing base for Li batteries is in China, Korea and Japan. From a commercial status perspective, the US manufacturer/system supplier base for Li batteries for stationary applications is currently in a fluid state. International Batteries, A123 and Altairnano are the three prominent US suppliers, and each has either been through 15

173 bankruptcy or is re organizing their corporate structure in some form or the other. This uncertainty is impacting further deployments of Li based stationary storage systems and will remain so in the near future. Figure 5: Current and Planned Li ion System Demonstrations. Flywheel Energy Storage Systems Flywheels store energy in the form of the angular momentum of a spinning mass, called a rotor. The work done to spin the rotor is stored in the form of kinetic energy and during discharge this energy is converted into alternating current (ac) power via use of controls and power conversion systems. Rotational speeds in the order 16,000+ revolutions per minute are needed to generate sufficient power for electric grid applications. Historic flywheel designs were flat, pancake shaped wheels and required even higher rotational speeds to deliver useful power. The design of flywheels for grid applications has since evolved into a tall cylindrical configuration that is typically seven feet tall and about four feet in diameter as shown in Figure 6. This wheel has a safer failure mode and has proved to be more stable in practical applications. Flywheel systems available today have containment vessels for safety and performance enhancement purposes. This is a thick steel vessel surrounding the rotor, motor generator, and other rotational components of the flywheel. If the wheel fractures while spinning, the 16

174 delaminated layers rub against the containment vessel which slows them down and stops the fragments preventing damage to surrounding equipment and injury to personnel. Containment also serve as vacuum chambers or filled with a low friction gas such as helium to reduce the effect of friction on the rotor and further enhance the performance of the flywheel. Figure 6. A Flywheel System Cutaway Diagram (Courtesy Beacon Power) Flywheels have excellent cycle life in comparison to battery based energy storage systems. Most flywheel developers estimate cycle life in excess of 100,000 full charge discharge cycles that makes this an attractive technology option for renewable support and frequency regulation service. Flywheels also have a very fast response times of four milliseconds or less which also makes these systems more suitable for renewable ramp support and frequency regulation services. But, flywheel storage systems can support only short duration discharge requirements and are not useful for the long duration applications that require discharge times greater than 30 minutes. For example, the flywheel storage system that is available from Beacon Power, is rated for a 15 minute discharge time. Beacon Power is the sole manufacturer and supplier of large scale flywheel storage systems in the US. The building block of their storage system is a 100 KW/25 kwh flywheel and the larger systems are built by aggregating multiple building blocks. Beacon has successfully built and operated 10 to 20 MW flywheel systems based on this aggregated configuration to provide regulation services to ISO s in NY and California. 17

175 Beacon s original business model was to build flywheels and own merchant plants to provide ancillary services to independent grid operators. This business model initially failed to generate sufficient revenue to support Beacon s operations due to lower than projected income from market operations. Consequently, Beacon Power declared bankruptcy in October 2011 and was acquired by Rockland Capital in early The acquisition gave Rockland Capital access to Beacon s flywheel intellectual property and the revenue generated by the 20 MW flywheel storage plant in Stephentown, NY. Figure 7 shows a 1 MW system installed at Beacon s headquarters in Tyngsboro, MA. Rockland Capital also formed Spindle Grid Regulation, LLC, to develop and own merchant plants that provide ancillary services to various grid operators. They are currently building a 20 MW flywheel plant in Hazle Township, PA, to provide fast regulation services to PJM, which is the independent system operator for the Pennsylvania, New Jersey and Maryland region. Figure 7: Beacon s 1 MW Smart Energy Matrix Plant with Ten 100 kw Flywheels Housed in the Blue Silos (Photo courtesy: Beacon Power) 18

176 Long Duration Storage Technologies Pumped Hydro Pumped hydro (PH) storage works on the simple principle of pumping and discharging water between an upper and lower reservoir. The water is pumped into the upper reservoir during periods when there is excess or low cost energy in the grid and the stored energy is recovered by allowing the stored water to run through a turbine to a lower reservoir when there is a need for energy in the grid, or when the price of electricity is greater. This is illustrated in Figure 8, which shows the upper reservoir, lower discharge, the connection between the two and the prime mover that functions both as the pump when the upper reservoir is being filled and as a turbine when the stored water is discharged. Figure 8: Schematic of a Pumped Storage Plant (Source: PH is the most widely deployed form of electricity storage in the US mainland grid, and 24,000 MW of pumped storage systems are in operation for 50 years or more and continue to provide reliable service. While the principle of PH operation is simple, the practical implementation is much more complex, because the economies of scale dictate that these plants are economically viable only in the hundreds of megawatt size with storage capacities of ten hours or more. The construction of reservoirs to fit this size requires massive upfront capital investment. In addition, the engineering challenges of building such large PH plants is compounded by stiff opposition from environmental groups due to significant impacts on land and water use. Consequently, PH projects have not been built in the US since the last 1,000 MW Rocky Mountain pumped storage plant was commissioned in Georgia in

177 The environmental impacts and scale of this technology make it unsuitable for Hawaii, where the island grids are smaller and land and fresh water requirements will conflict with other uses. However, a variation of the original technology using seawater to store only a few tens of megawatts with only one upper reservoir is being successfully used in Japan and Ireland. These smaller plants do not require large land areas and do not impact fresh water, because only sea water is used as the storage medium. The water stored in the upper reservoir is pumped up and discharged back into the sea with negligible losses and it is at the same temperature since there is no thermal process involved in the generation of electricity. These plants are located close to the coastline utilizing existing cliff shorelines to provide the gradient. Figure 9 is a photo of a sea water PH plant in Okinawa, Japan. As shown in Figure 9, only the upper reservoir needs to be constructed since the sea serves as the infinite lower reservoir. Figure 9: Okinawa Seawater Pumped Storage In a smaller, MW configuration, the scale of the plant is orders of magnitude smaller than its conventional full size, two reservoir counterparts. For example, an upper reservoir located 300 feet above sea level, with a diameter of 400 feet and 20 foot depth, will provide 20 MWh of storage capacity, or 10 MW for 2 hours at constant discharge. An upper reservoir for such a plant could be built on coastal cliffs in any of the islands without a significant impact on land use and zero net impact on fresh water sources. In Hawaii, seawater PH offers the ability to build larger storage systems that are other wise impractical with battery and flywheel systems. Seawater PH could effectively bridge the gap between short and mid duration and the long duration storage needs as Hawaii heads into higher penetration of renewable resources. This assertion is further supported by the 20

178 fact that seawater PH uses existing off the shelf hardware and is based on operating principles that electric utilities are already familiar with. Sodium Sulfur Batteries Next to lead acid, sodium sulfur (NaS) batteries have found the most prevalent use in electric utility storage service. NaS batteries have been used for distribution grid support, wind power integration, and other high value grid services because they are capable of long discharges of up to 6 hours and have longer cycle life than comparable lead acid batteries. Like many other storage technologies, they are capable of prompt, precise responses to the grid s needs such as mitigation of renewable generation variability and spinning reserve events. NaS batteries are high temperature batteries with a normal operating range of 330 C to 350 C. Therefore, they are assembled in air tight, double walled stainless steel metal enclosures that contain the series parallel arrays of NaS cells. As shown in Figure 10, each cell is hermetically sealed and surrounded with sand to both anchor the cells and mitigate fire. Other safety features include fused electrical isolation and a battery management system that monitors cell block voltages and temperature. The sodium, sulfur, betaalumina ceramic electrolyte, and sulfur polysulfide components of the battery are disposed of, or recycled at the end of the NaS battery life by routine industrial processes. Figure 10: NAS Battery Module Components (Courtesy NGK Insulators, Ltd.) NGK Insulators, Ltd., and Tokyo Electric Power Co. (TEPCO) jointly developed NaS battery technology over the past 25 years and NGK is the sole supplier of NaS battery systems worldwide. Standard building blocks used in NGK s system contain five 50 kw NaS modules that include a control unit, heater, heater controller, and voltage and current measurement sensors. NGK s NAS storage systems are only available in multiples of 1 MW/6 MWh units with installations typically in the range of 2 to 10 MW. NAS installations providing the functional equivalent of about 160 MW of pumped hydro storage are currently deployed within Tokyo. The largest single installation is the 34 MW 21

179 Rokkasho wind stabilization project in Northern Japan that has been operational since August 1, Currently, over 400 MW of NAS installations have been deployed globally at over 230 sites, representing 2,000 MWh. Sodium Nickel Chloride Sodium nickel chloride batteries are high temperature battery devices like NaS but operate between 270 C and 350 C. These batteries are also assembled in hermetically sealed modules that are typically 20 kwh each. General Electric (GE) and FIAMM are the two leading suppliers and pre production units in the 200 kwh size range have been fielded at Duke Energy and other mainland US utilities. GE, with name recognition in the utility industry, and established products dedicated to electric utilities, is best positioned to deploy their stationary battery systems. However, GE has a parallel interest in deploying this battery technology in hybrid locomotives that it also manufactures and markets to a worldwide customer base. The hybrid locomotive market initially represents a much larger market opportunity than the stationary market, and its higher volume manufacturing could provide GE an advantage in developing a lower cost stationary storage systems by capturing cost reductions on the electromotive side of their business. Figure 11: FIAMM s 220 kwh Pre production Unit at Duke Energy Site Flow Batteries Unlike other battery types, flow batteries have aqueous electrolyte that is circulated through a stack where an electrical potential is produced through their reaction, hence the 22

180 label flow batteries. These batteries were originally developed for long duration grid and EV applications with development work starting in the late 1960 s, because they can be configured for 3 to 6 hours discharge duration, and scaled up to megawatt size systems. A simplified schematic is shown in Figure 12. Figure 12: Schematic of Zinc/Bromine Flow Battery Development work in flow batteries branched into several chemistries with limited success; recently, attention has focused on two types Zn/Br and Vanadium Redox which will be the likely candidates for near term use in the electric grid. Kilowatt size systems have been deployed in small numbers in utility applications, but, these technologies have not shown reliable performance in the larger megawatt size range. Zinc Bromine Zinc Bromine batteries are currently under development by ZBB Corporation (ZBB) and Premium Power. ZBB Energy Corporation designed and manufactured a 50 kwh battery module in the late 1990s that served as a building block for larger systems. A trailermounted 200 kw, 400 kwh battery shown in Figure 10 was built for demonstration at two sites in the Detroit Edison service area in

181 Figure 13: Trailer mounted 200 kw/400 kwh Demonstration Battery System by ZBB Corporation Subsequent systems have been fielded at other demonstration sites. However, scaling up to the megawatt size systems has encountered technical issues and the ZBB systems remain in the 200 kw to 500 kw size range. Another flow battery is made by Premium Power (Premium) which announced plans for the production of TransFlow This was projected to be a low cost storage system, with a rating of 500kW/ 2.8MWh in a single 53 enclosure and was scheduled to be available in 2009/2010. Several US utilities, including MECO, entered into agreements with Premium to purchase the TransFlow storage system. However, production plans were delayed and most were nullified. One pre production unit was shipped to Kotzebue Electric, Alaska, in 2011, to support the integration of wind turbines into the town s electric grid. The unit did not meet its operational requirements and was shipped back to Premium in late Premium is under new management and future development of the Premium storage systems remains somewhat uncertain at this time. 24

182 Figure 14: Premium Power TransFlow 2000 System Being Unloaded at Kotzebue Vanadium Redox Batteries The vanadium redox flow battery is based on redox (reduction and oxidation) reactions of different ionic forms of vanadium and is somewhat similar to the Zinc Bromine battery in its construction. The cells are separated by a proton exchange membrane, which allows for the flow of ionic charge to complete the electrical circuit; the electrolytes are stored in external tanks and pumped as needed. As with other flow batteries, the vanadium redox systems tend to be physically large due to the large volumes of electrolyte required when sized for utility scale (megawatt hour) projects. Unlike many other battery technologies, the cycle life of vanadium redox systems is not dependent on depth of discharge. Systems are rated at 10,000 cycles, although some accelerated testing performed by Sumitomo Electric Industries, Ltd., produced a battery system with one 20 kw stack for cycle testing that continued for more than 13,000 cycles in about two years time. Vanadium redox systems made by VRB, now Prudent Energy, have been demonstrated in two projects in the US. The first deployment was in Utah followed by a larger 600 kw/3,600 kwh at an onion processing facility in California, see Figure

183 Figure 15: Prudent Energy 600 kw/3,600 kwh VRB System at Gills Onions, Oxnard, California. Compressed Air Energy Storage (CAES) CAES is the only other long duration energy storage option available besides pumped hydro. CAES works by storing compressed air in an underground reservoir or above ground in pressure vessels. During discharge, electricity is produced by allowing the compressed air to expand through an expander or conventional turbine generator while adding heat. Figure 16 is a schematic of a CAES plant with compressed air storage in an underground cavern created in a salt dome. 26

184 Figure 16: Schematic of a CAES plant with Underground Compressed Air Storage (Source: Electric Power Research Institute) So far only two CAES systems have been built: one in Germany and one in Alabama. The plant in Germany has been operating since It is rated at 290 MW with four hours of storage and uses two man made, solution mined salt caverns to store the compressed air. The plant in Alabama is located in McIntosh, AL, and is operated by PowerSouth Energy Cooperative (formerly Alabama Electric Cooperative). It was commissioned in 1991 and has a rating of 110 MW and has a 14 hour storage capacity and stores compressed air in underground caverns at 1,000+ psi. Creating a cavern in existing salt domes or finding suitable geologic formations to store the compressed air reliably over a twenty or thirty year operating life has been challenging and additional CAES plants with underground storage have not been built. Second generation CAES concepts with above ground storage in pressure vessels has been proposed and could hold potential for future applications in Hawaii. These would be smaller in size, MW with storage capabilities in the 2 to 6 hour range. The further development of above ground CAES technology should be tracked closely by HNEI or DBEDT, but, any prototype demonstrations should be tested at mainland sites and re located to Hawaii only after specific performance metrics are achieved. Other variations of CAES utilizing air storage vessels immersed in deep sea locations, especially in channels near several Hawaii locations have also been proposed. The designs for these types of CAES is still in the concept stage, and its practical implementation could present presents even greater challenges than the above ground CAES because of pressure and temperature gradients and the survivability of the storage vessels in a deep sea environment. It is also very unlikely that the built cost of the deep sea CAES systems would offer any advantages over the above ground CAES systems, because of the obvious ease of construction, maintenance and repair of hardware above ground as compared to hardware in a deep sea environment. 27

185 Section IV: Storage Technologies Cost of Systems Unlike other grid resources such as combustion turbines or diesel generators, the cost of a storage system is difficult to bracket due to the various technologies and applications of storage. And, unlike a pure generation resource such as a combustion turbine, a storage system has a power, kw or MW, and energy, kwh or MWh, cost component that can vary in value as the storage system size changes. For example, a 2 MW, 7 hour battery system such as the NGK NAS system will have a total cost that is different from the same 2 MW battery with a shorter discharge capacity of 2 hours. Estimating the cost of the storage systems is further complicated by the various maturity levels of the storage technologies and vendor products. Cost estimates from vendors who have longer experience backgrounds and those that have successfully fielded operating systems are more accurate than from vendors whose systems are in earlier development stages and have not reached full commercial maturity. Several studies have presented cost estimates, but, the most recent cost information is available from the soon to be released DOE/EPRI Electricity Storage Handbook and the Storage Technology Assessment report of the Congressional Research Service 13. The cost data presented in these two reports for the storage technologies and applications of interest in Hawaii is summarized in the Table 1. The Congressional Service report presents cost information on only three storage technologies of interest. These include Sodium Sulfur, Zinc Bromine and Vanadium Redox and their variations highlight the difficulty in comparing and generalizing cost estimates obtained from different sources and for different applications. Table 1: Cost Estimates for Select Storage Technologies Technology $/KW $/kwh Adv. Lead acid 4,000 8,000 1,000 2,000 Li family 2,500 12,000 2,500 4,500 Flywheels 4,000 5,000 20,000 22,000 Pumped Hydro 5,000 7,000 2,000 3,000 Sodium Sulfur 5,000 7,000 1,000 2,500 2,970 3,450 Sodium Nickel Chloride 5,000 10,000 2,000 5,000 Zinc Bromine 2,500 6,000 1,500 3,000 3,400 Vanadium Redox 3,000 7,000 1,500 3,000 2,450 Compressed Air 4,000 5,500 1,000 3, Energy Storage for Power Grids and ElectricTransportation: A Technology Assessment by Paul W. Parfomak for the Congressional Research Service, March 27,

186 Note: The estimates reported in the Congressional Research Service report for select technologies are blue italics in the table above. Section V: Energy Storage Integration in Hawaii Grids Types of storage deployments Future storage system deployments in Hawaii present opportunities to adopt new and nontraditional approaches to the operation and ownership of such systems while promoting the increased penetration of renewables in the grid. A few of these are discussed in this section: Technology Diversification: The storage systems deployed in Hawaii to date are all battery based systems. Other, nonbattery storage technologies, such as flywheels and sea water pumped storage offer viable alternatives and technical advantages as well. For example, flywheel systems have orders of magnitude longer cycle lives compared to batteries, and this technical characteristic makes flywheel storage a superior alternative to battery based storage systems where continuous shallow depth of discharge cycles are needed. As stated elsewhere in this report, flywheel systems are already providing regulation services to grids in the mainland and should be included in future assessments to verify their applicability in Hawaii specific cases. Similarly, Sodium Sulfur or Sodium Nickel Chloride battery storage systems available today can meet long duration discharge requirements of 4 to 6 hours. In this discharge duration range, sea water pumped storage systems can offer matching discharge times, but, without the cycle life limitations of batteries. Hawaii grids could greatly benefit from sea water pumped storage, that extends the discharge time into the 4 to 6 hour range with power in the 10+ MW size and without the cycle life limitations, especially as penetration of wind generation increases and curtailment becomes a significant issue. Utility side Storage System Deployments The storage systems deployed in Hawaii to date are predominantly on the developer side of the meter. This electrical placement and ownership of the storage system limits it to serve the needs of the renewable project only, and it cannot be used for other services that it could provide to the grid. Moving the storage to the utility side would enable the storage to provide other services such as regulation, spinning reserve or load leveling, while still damping the effects of the renewable generation ramps. However, this placement also 29

187 requires that the cost of the storage be equitably split and shared between the utility and the renewable project developer. However, agreement of how the cost should be split and other considerations of warranty coverage by the storage system supplier may pose some hurdles, but, suitable guidelines by the HPUC to allow such dual use and ownership could open broader opportunities for future storage deployments. Distributed Storage Systems Deployment The storage systems deployed in Hawaii to date are predominantly larger, megawatt size systems. A distributed approach combined with larger storage systems, for future deployments would offer greater operating advantages to each island s grid. The smaller size systems would be greater in number, but would be in the kw size range, with one to three hours of storage, located deeper in the distribution network and in proximity to residential or commercial/industrial customer loads. In sufficiently large numbers, storage systems deployed in this manner could offer multiple benefits to the grid. They would significantly off load the substation transformer during peak times, increase distribution system reliability, and hence, customer reliability and provide local voltage support. The distributed deployment also does not require a single, large capital expense that the mega watt scale systems require. Instead, it requires smaller investments spread out over a longer period of time to acquire individual systems or a block purchase of ten or more units. Such a staged deployment not only offers an easier financial planning scenario, but allows the acquisition to be made as and where needed almost a Just in Time approach to storage. An additional advantage of the distributed, staged deployment is that it does not commit the owning utility to one specific technology or vendor. Different technologies and vendors can be selected at any stage. Distributed, smaller size storage systems can be integrated in any Smart Grid communication and control scheme so that they can be controlled individually or collectively dispatched to support the grid needs at even the transmission level. American Electric Power has proposed a similar approach through their Community Energy Storage systems and have written a specification for such a storage system that is publicly available at: ev2_1.pdf 30

188 Section VI: Barriers for Energy Storage Deployment in Hawaii Until recently, pumped hydro was the only viable storage technology for electric utility service and, as a result, 24 GW of pumped storage systems are operating in the national grid. On the other hand, battery energy storage systems have found only a niche foothold in the electric grid with early projects that were mostly demonstrations to validate performance and provide hands on experience with large systems. However, improvements in battery technology, a regulatory environment that recognizes the beneficial role of storage in the grid, and the need to mitigate the variability of increasing penetration levels of renewables has provided greater opportunities for deployment of battery storage systems. The electric grids in Hawaii will soon have a rapidly increasing portfolio of wind and PV generation and Smart Grid options. These future grid configurations could greatly benefit from the services that energy storage can offer to enhance this imminent reality. The storage systems that have been deployed recently have demonstrated important, but limited, roles in the Hawaii grids. As described in this report, storage can offer an expanded portfolio of grid services as the Hawaii grids evolve. However, some barriers and challenges need to be overcome before such widespread deployment of storage can occur. Some if these are discussed below. Economics/Markets In considering storage, the prevailing opinion is that the total capital cost of storage systems, especially battery and flywheel systems, are higher in comparison to other utility options such as substation upgrades or addition of new peaking capacity. However, the counter argument to that is that the sum of benefits 14 these storage systems offer to a grid is not fully quantified or, more frequently, cannot be monetized due to restrictive or nonexistent regulatory structures to allow their recovery. This leads to the misconception by utility planners and commission regulators that storage is expensive and is justifiably excluded as a viable technology option in most utility planning scenarios. The regulatory aspect will be discussed in further detail in the Political/Regulatory subsection below, and issues directly related to cost will be addressed here. The cost of a storage system is composed of three components. These are: the energy component, which is the battery or flywheel, the power conversion system (PCS) component that performs the bi directional conversion between ac and dc, and balance ofplant component that includes controls/communications, housing, switchgear and 14 A benefit is generally quantified in terms of the monetary or financial value, whereas a grid service or application is a use whereas benefit equates to a value. 31

189 miscellaneous equipment to connect the load. On a percentage basis, these three components have been almost equal contributors to the total cost of the storage system over the last 20+ years. The energy component costs are storage technology specific and their future trend is influenced by complex factors. For example, storage systems using advanced lead acid batteries are unlikely to see any great reductions in price. Because, it s close cousins the starter battery for cars, trucks, recreational vehicles, boats, etc. are already produced in very high volumes in existing factories and their raw material supply chains and other related infrastructure is already mature. The overriding cost contributor for the advanced lead acid batteries will be the recovery of the R&D investments made in developing this variation of technology, and not in building their manufacturing infrastructure. Similarly, Li ion batteries for stationary, utility scale systems are close cousins to the Li ion batteries used for consumer appliances and battery packs for the growing number of EV s and PHEV s. The R&D and manufacturing investments made to produce those batteries should lead to cost reductions in Li ion batteries needed for electric utility stationary applications. NaS batteries have a unique situation because NGK Insulators, Japan, is the sole worldwide supplier/producer of NaS storage systems for electric utility applications. NGK NAS storage systems have a strong market demand, and they currently have a lead time of over six months for new orders. With over 400 MWs of installed storage systems in overseas markets and the continuing high demand for their product line, it is unlikely that there will be significant price reductions in NaS storage systems in the foreseeable future. Flow batteries have shown great promise as a low cost storage option for electric utility use, but their production volume is very limited and at least one supplier, Premium Power, has a containerized system that needs extensive re engineering to bring it to market. The other suppliers are producing their systems in limited quantities and the low volumes cannot benefit from the cost reductions that can be expected from high production volumes of the other storage technologies discussed above. The power conversion subsystem (PCS) costs have seen reductions in recent years, driven mainly by lower costs of its silicon based switches, improvements in overall engineering design, and production efficiencies. Further cost reductions may occur only if sales volumes of energy storage systems increase significantly. While the high initial cost of the storage systems has been a barrier, the valuation of the services that storage provides to the grid is also difficult to quantify in most cases and becomes another barrier. For example, in Hawaii s specific case, the value of providing spinning reserve with storage is difficult to quantify, because this service is already rolled into the total cost of operations for each utility and cannot be easily broken out. Similarly, 32

190 the value of providing ramp support or reducing curtailment is also location and resource specific, and generally difficult to quantify. Several reports have estimated the value for various grid services and the Congressional Service report summarizes these estimates as shown Figure 17. Figure 17: Value Estimates for Various Storage Services 15 Regulatory and Policy Drivers Policies set forth by the State and Federal regulatory bodies do not affect the capital cost of storage systems, but they provide the framework for utilities and third party investors to monetize the benefits of storage and recover their investment in a predictable manner. Recent regulatory actions have recognized the services that storage can offer to the grid and policies are in place by Federal Energy Regulatory Commission (FERC) and some State Public Utility Commissions (PUCs), notably the California PUC (CPUC), to allow recovery of investments in storage assets as a grid resource. Table 2 summarizes recent rules and the impact they have had in creating direct and indirect opportunities for energy storage deployment. 15 Energy Storage for Power Grids and ElectricTransportation: A Technology Assessment by Paul W. Parfomak for the Congressional Research Service, March 27,

191 Table 2: Regulatory Agency Rules and Their Impacts on Energy Storage System Deployment AGENCY RULE/ACTION INTENT OF THE RULE OR ACTION IMPACT FERC Rule 755 Directs to ISOs compensate suppliers of frequency regulation resources based on the actual service provided, including a capacity payment for the marginal unit s opportunity costs and a payment for performance that reflects the quantity of frequency regulation service provided. This benefits energy storage because storage suppliers are compensated for their ability to supply power to full capacity instantly. Rule 719 Directs ISOs/regional transmission organizations (RTOs) to accept bids from demand response resources for certain ancillary services on a basis comparable to other resources. Opens up the possibility for meeting commercial and industrial customers critical loads using storage that enables frequent demand response participation. FERC Rule 745 FERC's Market Based Demand Response (DR) Compensation Rule establishes that electric utilities and retail market operators will be required to pay for resources at the market price for energy, known as the LMP, when load reductions are able to balance the grid's supply and demand as an alternative to a generation resource. May allow utilities or demand response aggregators to include energy storage in their portfolio of assets for DR. Rule 1000 In an effort to address deficiencies in regional and interregional transmission planning and cost allocations, FERC Order 1000 requires Public Utility Transmission providers to participate in transmission planning at the regional level. These plans must include a comprehensive evaluation of transmission solutions in coordination with transmission providers from neighboring regions to ensure cost effectiveness and must account for public policy requirements. Second, the order requires that the costs of transmission facilities be allocated fairly to estimated beneficiaries. Finally, the order identifies non incumbent developer requirements. Removed resource type restrictions and reduced minimum rated capacity to 500 kw from 1 MW to provide certain Because Order 1000 requires alternatives in transmission planning, non transmission alternatives such as energy storage could potentially see an increased deployment and in some instances may provide a more cost effective solution than other transmission investments. California Independent System Operator Modified Rules to Allow Non Generation Allows energy storage resources, such as batteries and flywheels, to provide regulation service by fully 34

192 AGENCY RULE/ACTION INTENT OF THE RULE OR ACTION IMPACT (CAISO) Resources ancillary services. Reduced minimum continuous energy requirement from 2 hours to: Day Ahead Regulation Up/Down: 60 minutes; Real Time Regulation Up/Down: 30 minutes; Spin and Non Spin: 30 minutes. Will allow minimum continuous energy measured from the period that the resource reaches the awarded energy output. Measurement starts once resource reaches awarded energy, not end of 10 minute ramp requirement. utilizing their fast response, fast ramping capabilities. Allows new storage technologies to provide regulation energy over a continued sustained period that does not have seemingly inexhaustible energy like fossil fuel resources. CAISO Flexible Capacity Procurement to Integrate Renewable CAISO is considering various electricity capacity sources to help manage the steep ups and downs due to wind and solar coming on line under the Renewable Portfolio Standards (RPS) mandate. CAISO defines the characteristics of the acceptable resources to manage steep & sudden ramps. If superior abilities of energy storage to ramp up quickly in response to needs and reach full capacity are included in the characteristics required, energy storage systems can participate in this market. CPUC Energy Storage rule making for AB2514 AB2514 sets up a framework for identifying storage services to the grid, evaluating their cost effectiveness, and identifying barriers to implementation of storage projects and possibly setting storage deployment targets if deemed necessary. Requires California utilities to procure energy storage to a set target provided costeffectiveness criteria are met. CPUC Self Generation Incentive Program (SGIP) rules Self Generation Incentive Program (SGIP) offers incentives to customers who produce electricity with wind turbines and fuel cells. Recent revision has made advanced energy storage systems eligible for rebate. Either as stand alone or combined with other renewable, eligible technologies, owners of energy storage systems receive rebates of $2/watt. The FERC and State policies and rules shown in Table 1 are an indication of the trends on the mainland that are favorable for the widespread deployment of storage. In Hawaii, regulatory initiatives, derived from or related to some of the policies listed above could have a significant impact on the widespread use of storage. Some of these are outlined below: 35

193 1. Establish requirements for spinning and operating reserves based on specific land needs: At present there are no policies for utilities to provide such reserves and each utility adjusts generation unit output to provide appropriate levels of regulation leading to nonoptimal dispatch of generating units. This mode of operation would become more inefficient as the proportion of variable renewable generation increases. A policy that sets a requirement for spinning reserves and regulation, and allows recovery of these assets through the rate base, would create an opportunity for using storage in lieu of providing these services with fossil generation. But, the present cost of storage systems may be too high to make this an economically viable option, so it may be necessary to make provisions to allow the resources used for reserve duty to concurrently provide other grid service(s) such as renewable ramp support. FERC Rule 755 and CAISO Flexible Capacity Procurement rules can provide guidance to frame Hawaii specific regulations for storage in these services. 2. Modify the existing feed in tariffs (FIT) to recover the additional cost of energy storage if it is integral to the proposed renewable energy project: At present, the FIT does not differentiate between renewable projects with or without energy storage to mitigate their ramp characteristics and the burden of the additional cost of the storage component is borne by the developer. The addition of storage to a renewable project, especially a photovoltaic system, provides a more grid friendly resource with system wide services. Thus, creating a framework to recover the additional cost of storage through a rider to the FIT could incentivize the addition of storage in future renewable energy projects. The PUC rules would have to provide guidance for the storage system size for a specific renewable system or allow the utility and developer to mutually agree on the storage size best suited for that particular renewable project and location. The regulations listed in the table above do not address this issue and it is unlikely that FERC or mainland PUC s will be considering this. However, a Hawaii specific rule may incentivize future storage applications. 3. Establish electric system reliability metrics and re examine the penetration limits of renewable systems integrated with energy storage on feeders: The Hawaii PUC recently received a report from the Independent Facilitator (IF) 16 summarizing the findings and recommending future actions for the Reliability Standards Working Group (RSWG) including a recommendation for the formation of a Hawaii Electricity Reliability Administrator (HERA). Among other issues, the IF report will impact the feeder penetration levels of PV and the minimum operating levels of fossil generators and could open new opportunities for the use of energy storage. It is recommended that 16 Reliability Standards Working Group Independent Facilitators Submittal Final Report, March 13,

194 the recently submitted IF report be reviewed to identify opportunities for energy storage and how it can be integrated into future studies and activities of HERA. 4. Modify the Renewable Energy Infrastructure Program (REIP) to allow recovery of future energy storage studies and projects: The HPUC recently approved a request by HECO to recover the deferred cost of Big Wind Implementation Studies through REIP 17. But, REIP, as it currently stands, does not allow the recovery of storage studies or the full or partial cost of storage project costs, even when it is recognized that energy storage is an enabling technology for greater penetration of renewables. A HPUC initiative is needed to modify REIP to allow the recovery of costs of future storage integration studies and implementation of large or distributed energy storage projects that support or facilitate renewable projects. 17 HPUC Decision and Order Number in Docket Number : HECO Tariff Sheets for REIP Surcharge: 01A13D08B12531E A13D08B12531E

195 Section VII: Technical Barriers Various reports have cited technical barriers that inhibit the widespread use of storage. Most notably, the Congressional Research Service report 18 listed these three barriers that are widely applicable: Unquantified and uncaptured benefits Regulatory and market uncertainty and risk Lack of incentives for customer sited storage While these three barriers are more applicable in the mainland, there is equal applicability in Hawaii as well. The unquantified and uncaptured benefits barrier is evident in the fact that Hawaiian utilities do not have a regulatory requirement for reserves as discussed in the earlier Regulatory and Policy Drivers section. Similarly, customer sited, but utility owned storage is not presently incentivized or valued. This falls under the distributed approach to storage deployment and is presently a barrier which if removed could spur further deployment of storage in Hawaii. The sizing of storage systems also emerges as a technical barrier. At present, there is not a set, commonly accepted methodology for sizing of a storage system, especially in combination with renewable generation. Thus, sizing the storage system is presently a project and location specific exercise mostly based on assumptions of the size of the renewable project and ramp rates specified by the utility. Experience from existing storage systems in Hawaii is showing that a single ramp rate specification may be insufficient guidance for sizing the storage system, because it ignores the possibility of multiple ramps occurring in rapid succession and the possibility that the battery has insufficient capacity to handle them. Another barrier that is being addressed by a DOE and EPRI initiative is defining test protocols to qualify storage systems for specific applications such as frequency regulation or time shifting renewable energy. At present, there is not a standard specification that states required performance parameters such as the number of cycles or depth ofdischarge that a storage system has to meet to perform these services. The lack of such a standard creates a barrier and an uncertainty for the utility in procuring a storage system with the confidence that it will meet certain performance goals in the short and long term. At present, performance warranties are presently bridging this uncertainty gap and a specification standard or test protocol to verify the vendor claims would greatly increase the confidence in storage system performance. 18 Energy Storage for Power Grids and ElectricTransportation: A Technology Assessment by Paul W. Parfomak for the Congressional Research Service, March 27,

196 Environmental constraints The environmental constraints for siting electrical storage plants vary by technology. The footprint of the complete battery system plays an important role when it is to located inside existing substation or other utility owned installations such as a switching station. The gravimetric, Wh/kg, or volumetric, Wh/L, densities reported in literature or provided by battery manufacturers for various battery types are somewhat misleading metrics from this practical perspective. These metrics not only discount the space requirements of the balance of system components, but also the fact that most battery systems are packaged in containers that are the standard shipping size or similar. Thus, regardless of battery chemistry, sub megawatt or 1 MW size systems are all packaged in containers with a nominal footprint of 12 x 40. Therefore, the footprint of the entire system, including all the system components, not just the battery stack, is of greater practical importance than the stated energy densities of the battery alone. Battery and flywheel storage systems are relatively easy to site due to nominal land area requirements. By contrast, pumped hydro facilities require significantly larger land area for their reservoirs and land use is an environmental constraint that needs to be addressed at the earliest stages of project planning. All the storage technologies described in this report are non emitters of pollution at the project site. However, storage systems derive their charging energy from external sources such as a fossil fired generation unit or a wind farm or PV plant. If the charging energy is derived only from a fossil fueled source, then source emissions could increase depending on when and how the unit is used as the charging energy source. Public Perception/Cultural The recent fire at the Xtreme Power storage system at Kahuku, Oahu, will likely create a heightened sensitivity to locating future storage facilities in populated settings. This constraint, if it is needed, should be addressed in the early planning phases of the project to address specific concerns in siting the storage facility. Electric utilities that have implemented storage projects have simultaneously provided training and awareness education to emergency first responders, including local fire departments, to equip them to handle the unique needs of each battery technology. The equipment layout of the storage project also takes into consideration the access needs of emergency personnel and equipment so that there is sufficient clearance between buildings and system components to accommodate their needs. This requires an additional interaction between the electric utility and the storage system vendor to ensure that this critical detail is addressed satisfactorily to comply with all the requirements. 39

197 Traditional pumped hydro storage projects that require two reservoirs and use fresh water have justifiably been rejected in the past due to the inevitable conflict over land and water use issues. However, the seawater pumped hydro systems discussed in this report require only one reservoir, use only sea water that is fully recycled and do not produce a thermal plume at the sea water discharge flumes. Their land use impact is much less and there is zero impact on fresh water use. These systems offer a viable long duration storage option that is superior to battery based long duration storage systems since their operational life is not constrained by life cycle limitations. The use of such sea water pumped hydro systems could be revisited through town hall meetings with the community to explain their merits over previously held notions about traditional pumped hydro systems. 40

198 Section VIII: Risks and Uncertainties Some of the battery technologies deployed or to be deployed in Hawaii have limited operational histories with full scale systems in the field. There is an element of risk in how these battery technologies will continue to perform in the latter phases of their operational life, especially if several storage systems use the same battery type and are supplied by the same vendor. The lack of proper training and familiarity of utility operations and maintenance personnel who oversee the storage system is a related risk that lies on the host utility side. Storage systems have very different physical and operational characteristics than the traditional utility resources such as generators and transformers. Proper training and familiarity of the all personnel with the storage system reduces this risk factor. Similarly, as indicated earlier in the storage technologies section of this report, some of the storage system vendors are recently formed companies or recent startups. Most of these companies rely heavily on venture capital to sustain their operations, including warranties on their product. The venture capital funding introduces another element of business risk, whereby the longevity and ability to provide field maintenance and repair services could become questionable or cease altogether. The use of storage to smooth and shift renewable generation is still in relative infancy. The algorithms developed by system vendors have been mostly developed using solar irradiance and wind data often recorded at other sites. Which introduces an uncertainty about the optimal performance of the algorithms. There has not been a critical assessment of these algorithms and questions remain about their complexity and efficiency. Similarly, the correct sizing of the storage system, both in power and energy rating, is still not a standardized methodology given the same parameters of the renewable system that it is meant to support. Another uncertainty is the effect that other technologies that are likely to be introduced into Hawaii in the near future and their impact on the storage systems that are already in the field at that time. For example, a high penetration of vehicle to grid storage could introduce an uncertainty on the effectiveness and operational impact of distributed, community storage systems, if these are already in place. 41

199 Section IX: Lessons Learnt and Performance Assessment of Existing Storage Systems There are over twenty storage systems deployed in Hawaii, and each system has some unique lessons learnt through all stages from its initial planning, technology and vendor selection, procurement and field installation and startup. These experiences need to be captured and shared and their lessons learnt to be used to simplify future storage projects. There also has to be a coordinated effort to assess the predicted and actual performance for those storage systems whose data can be publicly shared. Such a comparative evaluation could lead to several improvements for future systems, including sizing, technology selection and monitoring. Section X: Conclusions Energy storage can play a key role in facilitating high penetration of variable renewable energy sources in Hawaii. Over twenty systems have already been installed and are providing valuable operational experience to the host utilities. While a large number of novel storage technologies are being investigated, only a handful of storage technologies have achieved commercial maturity and are deployable. However, the deployments of storage system thus far have been narrowly limited to two or three battery technologies. It is recommended that other technologies such as flywheels and seawater pumped storage be included in future technology selection opportunities. Total system cost of storage systems has always been difficult to present on a generalized basis without having application, technology and system size specific information. Nevertheless, some recent cost estimates for the technologies included in this report have been shown. Some regulatory initiatives have also been outlined that could facilitate the wider deployment of storage systems. And, it is also recommended that a distributed, smallersize deployment approach be combined with larger sized systems to capture benefits at the distribution and customer service level. Several technical barriers such as valuation of storage services and sizing of storage systems exist and need to be addressed to facilitate future deployments. Finally, it is recommended that performance data and lessons learnt from the existing systems be gathered and made available to facilitate the deployment of future system. 42

200 UNDERSEA CABLE Prepared for the: State of Hawai i Department of Business, Economic Development and Tourism Under Supplemental Contract No. 1 for Contract #60514 for the NETL Grant Project Task 1, Subtask 5, Undersea Cable Element Submitted by the: Hawai i Natural Energy Institute July 2013

201 UNDERSEA CABLE Prepared for the: Hawai i Renewable Energy Grid Project Prepared by the: State of Hawai i Department of Business, Economic Development and Tourism June 2013

202 Table of Contents Contents Executive Summary... 1 Hawaii Energy Geography... 2 Resources... 2 Demand... 2 Reconciling Demand and Economic Viability... 3 Summary of Findings from Cable Studies... 4 Navigant Cable Study... 4 AECOM HIREP... 5 Air Quality and Climate Change... 6 Cultural Resources... 6 Land Transportation... 8 Land Use... 9 Marine Transportation Marine Biological Resources Natural Hazards, Hazardous Materials and Munitions and Explosives of Concern Noise and Vibration Public Health and Safety Public Services and Infrastructure Recreation SOEST Hawaii Deep Water Cable Study Case for Cable Capacity to Deliver Energy Load Balancing Opportunities & Efficiencies Economics Construction Phase i

203 Analytical Assumptions Estimated Economic Impacts Converter Station Operations and Maintenance Analytical Assumptions Estimated Economic Impacts Other Considerations Energy Security Environmental Trade offs (Alignment with HCEI) Emission Reductions Reduced Dependence on Oil Net Energy Savings Construction Damage Description of Cable Background on Cables Currently in Operation (State of Commercialization) Sizing and Capacity Routing Financial Structure Agreement Terms Risk Risks and Uncertainties ii

204 Executive Summary In an effort to diversify the renewable energy capacity in the State of Hawaii, and with the support of the U.S. Department of Energy ( USDOE ), the primary stakeholders in Hawaii s energy future, the State of Hawaii Department of Business, Economic Development and Tourism ( DBEDT ) and the Hawaiian Electric entities, decided that there was a need to investigate the possibility of connecting the islands through an inter island cable system, which would allow the development of larger scale renewable projects to meet Oahu s demand as well as increase grid stability for all connected islands. Several key reports have been commissioned, including a overarching report by Navigant, an environmental impact study by AECOM; technical reports by the University of Hawaii School of Ocean and Earth Science and Technology ( SOEST ); and the 1989 Hawaii Deep Water Cable Study ( HDWCS ) which assessed the feasibility of transmitting geothermal electricity via a submarine cable system from the Island of Hawaii to Oahu and Maui and contains valuable information that can be used for the Cable Project. The AECOM review was prepared at a programmatic level, meaning that it reflects the environmental consequences attendant upon a wide ranging program. The report is a comprehensive study of the project s potential environmental impacts vis à vis several resource areas, including: air quality and climate change; cultural resources; geology and soils; land transportation; land use; marine transportation; marine/benthic species and habitat; natural hazards, hazardous waste, and munitions and explosives of concern; noise and vibration; public health and safety; public services/infrastructure; recreation; socioeconomics and environmental justice; terrestrial/coastal biological resources, species and habitat; 1

205 visual resources; and water resources. The report identifies all assumptions relied upon when determining the potential impacts, as well as measures that can be taken to mitigate any negative effects. The SOEST reports discuss the results of ocean floor surveys conducted by SOEST for the purpose of ascertaining the feasibility of laying a power transmission cable on the ocean floor to transport renewable energy generated on Lanai and Molokai to Oahu. Hawaii Energy Geography Resources Hawaii s economy is not energy intensive. The state is geographically isolated from major natural gas and coal resources, yet depends heavily on fossil fuels to meet energy demand. Indeed, it is the most petroleum dependent state in the nation. 1 In 2009, close to ninetenths of Hawaii s energy came from imported petroleum. 2 In 2010, Hawaii imported 94 percent of its energy. 3 While Hawaii s location presents challenges in supplying the fossil fuels needed to meet all of its energy demands, the state is also uniquely situated for innovation in the renewable energy sector. Generally, renewable energy is not cost competitive in other parts of the country where energy produced from natural gas and coal is less expensive. However, because of the geographic isolation of Hawaii, and the high price of importing fossil fuels, Hawaii is uniquely positioned to produce renewable energy at a competitive cost with energy from fossil fuels. Moreover, Hawaii s unique geography makes the natural renewable resources available in Hawaii, particularly wind, solar, and geothermal resources readily available in abundance. Demand Among the islands, Oahu is by far the dominant consumer. In June 2012, Honolulu consumed 571,895,959 kilowatt hours of electricity. The same month, the County of Maui (including the islands of Molokai and Lanai) consumed 92,635,449 kilowatt hours, Hawaii County consumed 87,471,902 kilowatt hours, and the County of Kauai consumed 37,416,623 kilowatt hours of electricity. 1 U.S. Army Corps of Engineers, Hydroelectric Power Assessment State of Hawaii, Asia Pacific Clean Energy Summit (created Sept. 14, 2011) < 2 U.S. Energy Information Administration, Hawaii State Profile and Energy Estimates: Analysis (last modified Oct. 2009) < 3 U.S. Energy Information Administration, Hawaii State Profile and Energy Estimates: Overview (last modified July 2012) < 2

206 5% 11% 12% Honolulu Maui Hawaii Kauai 72% Figure 1: Data from Hawaii Electric Company, Inc., Hawaii Electric Light Co., Inc., Maui Electric Company, Ltd., and Kauai Island Utility Cooperative, Monthly Financial Reports Oahu s energy demands are almost three times the demand of the other counties combined. Honolulu is also the most densely populated county. Thus, there is limited availability of land upon which renewable energy projects may be constructed to meet the county s needs. Reconciling Demand and Economic Viability The availability of land and demand for energy on Oahu leads to the unique problem facing the State and the Hawaiian Electric entities Hawaiian Electric Company ( HECO ), Maui Electric Company ( MECO ), and Hawaii Electric Light Company ( HELCO ) in meeting the RPS. The vast majority of energy demand is on Oahu, but the existing population density and higher cost of land and development results in Oahu not being the optimal location to develop all the commercial scale renewable energy projects needed to meet Oahu s RPS requirements. Meanwhile, the comparatively larger undeveloped areas available on the neighbor islands allow for such development of a larger scale project. However, as the grids of the respective islands are not currently connected, there is currently no ability to develop such a large scale project on another island for the purpose of assisting in meeting Oahu s energy demand and thus reducing its dependence on fossil fuels. Along with the support of the U.S. Department of Energy ( USDOE ), therefore, the primary stakeholders in Hawaii s energy future, the State of Hawaii Department of Business, Economic Development and Tourism 3

207 ( DBEDT ) and the Hawaiian Electric entities, decided that there was a need to investigate the possibility of connecting the islands through an inter island cable system, which would allow the development of larger scale renewable projects to meet Oahu s demand as well as increase grid stability for all connected islands. Summary of Findings from Cable Studies Navigant Cable Study DBEDT retained Navigant Consulting Inc. ( Navigant ) to act as a subject matter expert to assist the state with matters relating to the undersea cable, including development, design, regulation, financing, ownership, procurement, and analyses related to an inter island undersea power cable. The Navigant Report focused on the optimal business structure for the project, with an emphasis on burden sharing and economic risk taking. Under Navigant s recommended structure, the project should be developed by a private developer selected through a competitive procurement process. Navigant found that neither the State nor HECO is wellsituated to accept project development risk, particularly in light of the downgrade of HECO s credit rating to BBB. Under the recommended structure, the private developer would develop, own, and maintain the project for a period of at least ten years. Thereafter, HECO would have the option to purchase the cable. If the option to purchase is exercised, HECO would seek a rate recovery of its purchase price through a Hawaii Public Utility Commission ( PUC ) order. The report further discusses legislation Navigant developed for implementation of the proposed business structure. Briefly, the legislation would authorize the PUC (i) to approve the creation of a transmission utility (ii) to approve a transmission tariff and mechanism for a surcharge to be collected by the existing electric utility for transfer to the certified cable company; and (iii) to approve investments made by the existing electric utility in on island infrastructure. Navigant determined that the enactment of such authorizing legislation was a necessary pre condition for further material progress on the project. 4 Navigant ultimately concluded that under the proper business structure, the project is both technically and economically feasible. Moreover, it will produce savings for retail ratepayers as of the date of first commercial operation and continuously thereafter. 4 The legislation proposed by Navigant was, with modifications, signed into law by Governor Neil Abercrombie on June 27, 2012 (Act 165, Session Laws of Hawaii 2012). 4

208 AECOM HIREP The State retained the consulting firm AECOM to prepare an environmental review of the potential impacts resulting from the undersea cable project. AECOM s review was prepared at a programmatic level, meaning that it reflects the environmental consequences attendant upon a wide ranging program. Programmatic analyses look at broad issues relevant to the overarching program design. 5 As a result of the programmatic nature of AECOM s analysis, the analysis itself will continue to provide valuable insight on the project s impacts even if particular segments of the project are altered in the future. Segment specific analyses will still need to be conducted at a later date. The report is a comprehensive study of the project s potential environmental impacts vis àvis several resource areas, including: air quality and climate change; cultural resources; geology and soils; land transportation; land use; marine transportation; marine/benthic species and habitat; natural hazards, hazardous waste, and munitions and explosives of concern; noise and vibration; public health and safety; public services/infrastructure; recreation; socioeconomics and environmental justice; terrestrial/coastal biological resources, species and habitat; visual resources; and water resources. The report identifies all assumptions relied upon when determining the potential impacts, as well as measures that can be taken to mitigate any negative effects. 5 See e.g. Publication of Memorandum to Agencies Containing Answers to 40 Most Asked Questions on NEPA Regulations, 46 Fed. Reg. 55 (Mar. 23, 1981),

209 The following is a brief summary of AECOM s analysis of each of the 17 resource areas considered: Air Quality and Climate Change The undersea cable project will likely have temporary, construction related impacts and long term, operations related impacts on air quality. Construction of the converter stations will potentially produce fugitive dust and construction vehicle emissions, while at sea the operation of the cable laying vessels and related machinery will produce similar emissions. To minimize these adverse effects on air quality, the study recommends that conservation and construction measures ( CCMs ) be implemented during construction. Dust control measures, such as a dust control (watering) program and covering of soil stockpiles during transport or storage, will be developed and implemented by the construction contractor. Construction vehicles will either remain on site or be scheduled to arrive and depart the landing site area during nonpeak traffic hours, to reduce vehicle emissions. Over the long term, the converter stations could result in emissions both from the converter equipment and from the vehicle travel required to operate and maintain the stations. However, by their nature, converter stations emit little to no air emissions, and the stations themselves would not be considered a major source as defined by Haw. Rev. Stat. 342B 1. 6 It is anticipated that emissions during construction will be temporary and relatively minor, can be minimized through implementation of CCMs, and will not significantly affect air quality. Further, the use of a cable system to transmit renewable energy for use in lieu of fossil fuels will result in the reduction of greenhouse gas emissions. Cultural Resources Many of the proposed landing sites lie in areas which are largely undeveloped, and it is expected that these areas will have archaeological significance upon further subsurface testing. The particular impacts will be largely dependent upon the landing sites ultimately selected and the further investigation of the locations. The impacts will arise almost entirely during the construction phase of the project, rather than during operations. 6 The state s air pollution control laws, set forth in Haw. Rev. Stat. Chapter 342B, define Major source as "any stationary source, or any group of stationary sources that are located on one or more contiguous properties, and are under common control, belonging to a single major industrial grouping and that emits or has the potential to emit, considering controls: (1) Any hazardous air pollutant, except radionuclides, in the aggregate of ten tons per year or more, twentyfive tons per year or more of any combination, or such lesser quantity as the director may establish by rule; (2) One hundred tons per year or more of any regulated air pollutant, including fugitive emissions of any such regulated air pollutant as the director may establish by rule; and (3) For radionuclides, "major source" shall have the meaning specified by the director by rule. 6

210 Although cultural resources are not recoverable, certain measures can be taken to ensure that areas of cultural significance are left intact as much as possible. These include selecting lands which have already been developed or otherwise disturbed, using existing roads and minimizing the creation of new roads, having cultural and archaeological personnel on hand during construction when there is a likelihood of encountering cultural resources (especially Native Hawaiian burials), and ensuring compliance with the state s historic preservation laws (Haw. Rev. Stat. Chapter 6E). The report also recommends the creation of cultural resources management plans to address the discovery of cultural resources at a site or where there is a high potential of such discovery. Such plans would address mitigation activities to be taken for cultural resources found at particular sites. While avoidance is always the preferred mitigation option, the plans will also identify the steps to be taken in the case that avoidance is not possible. Such steps will include the establishment of monitoring programs, identification of measures to prevent looting, vandalism and the impacts of erosion, education of workers and the public, and the proper handling of discovered artifacts. Among the conservation measures identified by AECOM, many can be implemented early on in the planning process. For instance, it is noted that developers must consult with the State Historic Preservation Division ( SHPD ) and Native Hawaiian organizations to identify culturally sensitive sites and resources a process which is required by the National Historic Preservation Act. The report also recommends that developers conduct a comprehensive review of archival archeological literature to identify culturally sensitive sites and potential lineal and cultural descendants from the proposed landing site area. Geology and Soils The sites on every island will almost certainly require clearing, grading, excavating, engineering and re contouring of soils, both on the land upon which the converter stations will be constructed as well as for the access roads to the sites. This activity will primarily impact the soils. Construction activities may result in the clearing of vegetation, leaving the underlying soils prone to erosion. The threat of erosion will be greater at sites with greater slopes. However, most of the contemplated landing sites are located in flat, coastal plains which will be less vulnerable to erosion. AECOM s report sets forth seven recommendations for measures which can be taken in order to minimize the impact of the project on geology and soils. The starting point of the recommendations, which are summarized below, is the avoidance of sites with particular geologic traits. The recommendations further provide guidelines as to what measures should be taken, if any, in the event that avoidance is impossible or not practicable. 7

211 1) Avoidance of steeply sloping landing sites. Where avoidance is not possible, a geotechnical survey should be prepared to provide design and construction recommendations to ensure stability. 2) Dewatering of coastal sites. 3) Avoidance of lands designated as prime under the Agricultural Lands of Importance to the State of Hawaii rating system. 7 If avoidance of such lands is impracticable or impossible, the developer will have to petition the county and the state department of agriculture for reclassification or rezoning of the site pursuant to Haw. Rev. Stat ) Avoidance of landing sites on or near wetlands or surface water/drainage ways. If avoidance is not possible, development must comply with the requirements of the Clean Water Act (33 U.S.C et seq.), Section 10 of the Rivers and Harbors Act of 1899 (33 USC 401), the Safe Water Drinking Act (42 U.S.C. 300f et seq.), the federal Coastal Zone Management Act (16 U.S.C ) and the state Coastal Zone Management Program (Haw. Rev. Stat. Chapter 205A, the Hawaii Water Plan (Haw. Rev. Stat. Chapter 174C), and county water use and development plans. 5) If more than one acre of land is disturbed by the project, obtaining a Water Quality Certification pursuant to Section 401 of the Clean Water Act (33 USC 1341) and, potentially, a grading/grubbing permit from the county in which the development occurs. The developer may further be required to undertake permanent erosion control measures upon completion of construction. 6) Control of potential impacts to water quality and quantity during grading and construction by obtaining a National Pollutant Discharge Elimination System permit pursuant to Section 403 of the Clean Water Act(33 USC 1343) and a grading/grubbing permit from the pertinent county. Land Transportation Impacts to land transportation from project activities likely will be unavoidable. The use of heavy load trucks to deliver equipment and materials will probably result in the deterioration of roadway conditions, the alteration of existing roadways (either by widening narrow roads to accommodate oversized vehicles or the surfacing of unpaved 7 Haw. Rev. Stat through provide for the enactment of state and county agricultural policies, tax policies, land use plans, ordinances, and roles to promote the preservation of lands designated as important agricultural lands. There are three classes of important agricultural lands prime, unique and other important agricultural lands. Prime land refers to land whose soil has, among other characteristics, proper moisture, temperature, and acidity. See Agricultural Lands of Importance to the State of Hawaii, Classification System, available at (last visited December 14, 2012). 8

212 roadways), and increased traffic congestion from heavy load trucks moving at slow speeds. Further, new roads may need to be constructed in order to reach certain landing sites. However, these impacts will occur primarily during the construction phase of the project, with only minimal post development impacts resulting from ongoing maintenance and operations. To minimize the potential impacts, AECOM s report recommends the identification of efficient routes which avoid highly congested roadways and/or roadways that do not have the capacity to handle oversize and overweight vehicles. In conjunction with this recommendation, a transportation management plan should be developed to ensure the safe flow of traffic during construction operations. Land Use The report discusses impacts to land use, defined as the allowable uses of land as prescribed by the governing entity with jurisdiction over the land. Land use encompasses zoning, which is used to regulate activities permissible on land and implement land use plans. Pursuant to Haw. Rev. Stat. Chapter 205, all state lands are placed in one of four land use districts: urban, rural, agricultural, and conservation. 8 Jurisdiction varies by district. Urban districts lie primarily within the jurisdiction of their respective counties. Jurisdiction over rural districts is shared by the Land Use Commission and county governments. Agricultural lands are further divided into sub categories depending upon their productivity. The use of land within the highest categories of productivity is governed by statute, while the Land Use Commission establishes the permissible uses of lowerproductivity agricultural lands. Finally, conservation districts fall within the jurisdiction of the Hawaii Board of Land and Natural Resources. The report recommends that development is sited to minimize the displacement of existing uses. To the extent that avoidance of such displacement is not possible, the developer will be required to obtain the requisite permits and variances and/or apply for recategorization of the land in compliance with the applicable law. In addition to the restrictions placed on land use based upon a land s categorization under Haw. Rev. Stat. Chapter 205, lands located within the shoreline must follow the guidelines and permitting procedures set forth in the Coastal Zone Management Act (Haw. Rev. Stat. Chapter 205A). The act establishes minimal guidelines and permitting procedures for the counties to follow, but the counties are responsible for designating the lands which shall constitute special management areas subject to the act. All prospective landing sites fall 8 Haw. Rev. Stat established these four major land use districts. Descriptions of the districts are available on the State of Hawaii Land Use Commission s website. See land use districts/ (last visited December 17, 2012). 9

213 within special management areas, and developers will necessarily have to comply with the Coastal Zone Management Act s guidelines and permitting processes. Marine Transportation The resource of marine transportation refers to the transportation of people or goods by watercraft. As a remote island state, Hawaii is dependent upon marine transportation for the bulk of its goods. Use of watercraft for the transportation of people, however, is minimal. After the suspension of the Hawaii Superferry services in 2009, the only regularly scheduled interisland passenger ferry services run between Maui and Lanai and Lanai and Molokai. 9 The transportation of food, fuel, building materials, and other consumer products to the islands will be impacted by the cable project, particularly during the construction phase. However, as the in flow of goods must remain uninterrupted, it will be necessary for the cable developer to work with cargo carriers to ensure that transportation corridors remain open throughout construction. This may be achieved by publishing notice of cable laying in the USCG District 14 Local Notice to Mariners one week prior to commencement. Likewise, the cargo ships must be able to dock and unload at the various harbors and facilities. Notice should also be provided to harbormasters when cable routes terminate at or near a state harbor. This section of AECOM s report also covers the activities of the United States Coast Guard and Navy. The developer must consult with the Navy to ensure that defense facilities, movements, or security procedures are not impacted or compromised by the construction or operation of the cable facilities. Marine Biological Resources AECOM s report on the potential impacts to marine biological resources is based on data collected by the University of Hawaii School of Ocean and Earth Science Technology ( SOEST ). SOEST focused primarily on the southern route from Pearl Harbor on Oahu to northwest Molokai and the northern route between Kaneohe and Molokai. Numerous mustard gas bombs were detected along the southern route, as documented in SOEST s report. Accordingly, SOEST was unable to recommend a safe passage for the cable along the southern route. While the northern route did not have similar hazards, it is likely that there are several areas of sensitive habitats containing deep water coral along the route. The ocean floor was studied using a towed camera sled ( TowCam ) or remotely operated vehicles (ROVs). The ROV coverage was good, but only covered 10 percent of routes surveyed. The TowCam 9 Christopher Pala, A Hawaii Ferry Ends its Choppy Ride, New York Times (March 21, 2009), available at (last visited December 17, 2012). 10

214 was inconsistent, and produced only vague images that give an idea of seafloor topography and bottom type. It occasionally caught fish or mega fauna that remained in the field of view for a sufficient amount of time. The TowCam footage s primary value was its indication of areas where sensitive coral habitat communities may be present, thus warranting further study. The ROVs observed a large number of fish and mega fauna. 10 Among those most frequently observed by the ROV are sponges, anemone, coral, spider craps, shrimp, starfish, urchin, dogfish shark and butterfly fish. The waters around Hawaii also have a large number of endangered or threatened species protected by the Endangered Species Act. These include turtles, humpback whales, sperm whales, blue fin whales, sei whales, and monk seals. Further, large swaths of the ocean between Oahu, Molokai, Lanai and Maui are within the National Marine Whale Sanctuary, which falls under NOAA s jurisdiction. 10 A detailed list of the fish and megafauna observed in the ROV videos is contained in the report. 11

215 Figure 2 The project will impact marine biological resources during construction and operations. The trenching required to bury the cable will bury and destroy such adjacent organisms as coral. Sediment plumes (i.e. sediment disturbed from the ground floor during construction) will affect benthic communities and particularly sensitive filter feeding organisms. These plumes will affect not only those organisms adjacent to construction activities, but could potentially affect other organisms at great distances depending on currents. The report recommends avoidance of reef habitats that would be irreversibly damaged by laying cable. Further surveys should be taken in areas where there is the potential for sensitive habitats as determined from the ROV footage. The electromagnetic field emitted by the cable will interfere with life functions of organisms particularly marine mammals and turtles when the organisms are within sensory range of the cable. Marine mammals orientation, homing, navigation, predation, social and reproductive behaviors may be impacted. However, based on a study conducted 12

216 on the effect of DC cables on bottlenose dolphin orientation, it is expected that marine mammals will correct their orientation once outside the cable s electromagnetic field. And, while the cable could have severe impacts on sea turtles at a critical life stage, such as adults and hatchlings traversing shallow waters at natal beaches, the likelihood of exposure is likely low assuming the cables are sited carefully. The report recommends that observers be employed to watch for marine mammals and turtles during cable laying activities, and the development of a plan in the event that marine mammals are encountered. Sensitive deep water coral habitats must be identified before construction so that they can be avoided. Locating such habitats will require further study with improved visual recording. Avoidance of sensitive habitats is the primary objective. Cable routes must also be analyzed to determine the composition of the sediment which will be disturbed during the cable laying process. Natural Hazards, Hazardous Materials and Munitions and Explosives of Concern While the cable will not make natural disasters more likely, natural hazards could damage the cable itself, thereby affecting the human environment due to electrical disruptions. Natural hazards affecting the marine environment in the areas of analysis would be associated with damage to a potential undersea cable from tsunami, off shore earthquakes, undersea landslides, or underwater volcanic activity. The greatest potential natural hazard pertaining to the marine environment would be seismic events and associated tsunami. The undersea power cable system could be impacted by the shaking force, underwater landslides, or wave energy generated by these two types of events. Potential damage from natural disasters would depend on the location of the cables; the type of benthic and geological surface upon which cables are laid; terrestrial connection of cables via directional drilling; and cable design and engineering. Hawaii s flood dangers are primarily related to flash flooding of streams and flooding related to storm generated coastal surges. Those potential landing areas most affected by flooding are those areas located in an official FEMA flood zone. Tsunamis, which can affect all shorelines in Hawaii, are typically generated in the waters off South America, Japan, Alaska, and the west coast of the United States. The Hawaiian Islands are seasonally affected by Pacific hurricanes and tropical storms from the late summer to early winter months when ocean waters around the state are at their warmest. These storms generally travel toward the islands from a southerly or southeasterly direction and can deposit large amounts of rain with high winds on all the islands. The storms generally exacerbate localized stream flooding and coastal storm surges. Landing area sites in tsunami evacuation zones or official FEMA flood zones would 13

217 be most impacted by these events, although areas outside these zones could be impacted based on severity of the event, elevation, shoreline topography, and near shore geology. In Hawaii, most earthquakes are related to volcanic activity. Each year, thousands of earthquakes occur due to volcanic activity. Most of these events are so small as to be detectible by only the most sensitive of seismic instruments. Because Oahu is an older Hawaiian Island, it is not considered particularly prone to major seismic activity. Generally, the risk and intensity of earthquakes increases moving east along the Hawaiian Island chain, toward younger islands. Violent earthquakes can occur anywhere in the state, along with associated liquefaction events on low lying and fill areas. Smaller landslides pose a continual concern in Hawaii. These landslides are activated by storms, earthquakes, volcanic eruption, and fires. Related to landslides, debris, and mud flows are rivers of rock, earth, and other debris saturated with water. They develop when water rapidly accumulates in the ground during heavy rains. While frequently the result of natural events, landslides also can be the result of land mismanagement or dam failure particularly in mountain, canyon, and coastal regions. Wildfire occurs on all of the major Hawaiian Islands, with human activity as the primary cause. Because Hawaii s native ecosystems are not adaptive to wildfires, wildfires can result in extinction of native species and increased coverage of nonnative, invasive species. Other results of wildfires include soil erosion, increased runoff, and decreased water quality. Wildfires can occur on all the islands. Generally, dry, windier leeward sides of islands are more prone to risk of wildfire. Haleakala volcano on the island of Maui is considered dormant, since it last erupted sometime between 1500 and 1800, but it is not yet considered extinct. The islands of Oahu, Lanai, and Molokai do not have any active volcanoes. There are numerous active volcanoes on the Big Island, including Mauna Loa, Mauna Kea, and the offshore underwater volcano known as Loihi. These volcanoes are associated with seismic activity that can, in turn, affect the site study area. However, none of the landing site areas are located on islands with active volcanoes, so there would be no risk from lava flows. When the exact locations of land based facilities within the landing site areas are determined, it may be necessary to perform detailed investigations, including a site specific Environmental Site Assessment in conformance with the American Society for Testing and Materials Method E , which would provide for review of accessible public records and require site inspections to identify classes of land uses and/or hazardous materials that may be present at the project specific locations and along possible cable alignments. 14

218 Since the early 1900s, the state has been home to military bases for all branches of the U.S. military. The military has disposed of munitions and explosives of concern ( MECs ) on land and in the surrounding waters. As noted in the summary of the SOEST report in Section 3(b)(ii)(3) herein, some of the routes originally proposed specifically, those to the south of Oahu intersect with a four mile swath of disposed munitions. AECOM recommended the following general criteria and measures that should be taken to minimize risks in the event of a natural disaster. Project locations shall be outside of FEMA flood zones, where possible and practicable. Projects located on Oahu shall be required to comply with Revised Ordinances of Honolulu Section ( Flood hazard districts ). Projects located in Maui County shall be required to comply with Maui County Code, Title 19 Chapter ( Flood hazard districts ). 11 For any project seeking Federal Emergency Management Agency ( FEMA ) flood insurance, the project shall also be required to comply with provisions of FEMA s National Flood Insurance Program (NFIP) administered by the City and County of Honolulu Department of Planning and Permitting or the Maui County Department of Planning and in conjunction with DLNR under provisions of Haw. Rev. Stat. Chapter 179. For military sites, future project development shall comply with location, siting, and construction requirements as outlined by the Department of Defense. Projects shall be located outside of designated tsunami inundation/evacuation zones where possible and practicable. Projects shall be designed to comply with seismic regulations applicable to the relevant zones of the Uniform Building Code and any other applicable county seismic regulations. Landing site areas shall be developed in areas with minimal slope and no known geological hazards. Projects located on Oahu shall comply with ROH Sections and (relating to grading). Maui County does not have ordinances that specifically address the siting. The project developer shall comply with all design, fire safety, and emergency response requirements of respective counties, the Department of Defense or military fire departments having jurisdiction over the specific landing site area. 11 Both the Revised Ordinances of Honolulu and the Maui County Code relate to the management of development in areas designated as flood hazard districts. 15

219 Upon completion of construction, developers shall plant native, fire resistive vegetation; remove all dry grass, brush, and trees within at least 100 feet from buildings; and locate fuel tanks at least 30 feet from any structure. Developer shall avoid placement of an undersea cable in or near marine areas of known geologic instability, including seismic faults. Developer shall conduct an environmental assessment prior to obtaining property for the converter stations in order to assess the presence of any environmental contamination, hazardous substances, or MECs on the property. Developer shall create a disposal and recycling plan to deal with construction and demolition waste. Developer shall ensure that hazardous materials used during construction are handled and stored properly, and there are procedures in place to handle inadvertent spills. Site construction and demolition shall be performed in accordance with a sitespecific safety and health plan. The plan shall identify safe working conditions for construction areas. Safety measures shall include proper techniques for personal protective equipment, use of allowable tools, and mechanical measures as appropriate. Developer shall avoid siting the cable across areas known to contain hazardous materials. Developer shall require the contractor to repair site specific mitigation measures to minimize the potential for accidental release of hazardous materials into the environment, including procedures for hazardous material storage, handling, and staging; spill prevention and response and waste disposal; and good housekeeping. Although military munitions could be present on the seafloor throughout the Hawaiian Islands, of the most concern are areas of dense concentrations where it would be difficult for cable installers and maintenance workers to completely avoid the munitions. The main construction and mitigation measure for the MEC areas shall be avoidance. The SOEST report suggests that the greatest concentration of MEC is present off the south shore of Oahu. Noise and Vibration Elevated noise levels from project related construction activities are inevitable, but are expected to be minor and short term, i.e. limited to the construction phase. Noises generated from the construction equipment would need to comply with Hawaii Occupational Safety and Health Division regulations for temporary construction. Once the facility is operational, noise levels at the property lines would need to comply with HRS Chapter 342F, HAR Title 11 Chapter 46, and HAR Title 12 Chapter

220 Potential noise impacts associated with undersea cables would occur during the construction phase and in connection with maintenance or repair. Noise may arise from operation of vessels or machinery during the cable laying process, but the impacts would be temporary. If High Voltage Alternate Current (HVAC) cables are used as part of future developments, possible long term vibration and noise emission from the cable would be taken into account. Potential impacts from HVAC cables on marine habitats and species are summarized above in the section entitled Marine Biological Resources. To minimize impacts, AECOM s report recommends that the public be given timely notice of upcoming construction activities. Newer equipment should be used as much as possible to reduce noise (as compared to older equipment), and all equipment with internal combustible engines should be equipped with a properly maintained muffler. Construction equipment and the equipment of the facilities themselves should be operated according to manufacturer s instructions and routinely maintained. Public Health and Safety Public health and safety may be impacted during the construction phase (from construction and material transportation related accidents) and during operations. The presence of electrical generating equipment, cables and various oils create a potential for fire. The project may also generate electromagnetic fields. Numerous reviews of epidemiological and biological research studies have generally concluded that there is no scientific basis to support a finding of adverse human health effects from electromagnetic frequencies ( EMF ) although others have found that there may be an association between EMF and certain diseases. To minimize the potential adverse effects to public health, AECOM s report recommends the development of a health and safety program during construction, operation, and decommissioning of a project. The program should identify all applicable federal and state occupational safety standards; establish safe work practices for each task; and establish fire safety evacuation procedures; and define safety performance standards. All Occupational Safety and Health Administration ( OSHA ) laws and standards should be followed. Electrical systems should be designed to meet all applicable safety standards (e.g., National Electrical Code and International Electrical Code). Moreover, the project should be planned to minimize electromagnetic interference (such as interference with radar, microwave, television, and radio transmissions) and comply with Federal Communications Commission regulations. Signal strength studies should be conducted when proposed locations have the potential to impact communications transmission particularly where a converter station and transmission lines would travel on or near a military base. To minimize potential impacts from EMF, the submarine cable should be designed with the proper electrical shielding and burial (as appropriate) in the ocean floor. 17

221 Public Services and Infrastructure During construction and operation, there may be interruption of various public services, including electrical services, potable water supply, wastewater and solid waste disposal, and gas supply. In addition, the project may require police, fire and emergency medical response service, and thus detract from the availability of these services to respond to other calls. To avoid and mitigate these impacts, the project developer should consult with government agencies (county, state and federal) in regard to safety and emergency supports. In addition, a public facilities assessment should be conducted before construction commences to determine the adequacy of police, fire control, hospital, and emergency medical response. The developer should also consult with public and private utility entities to minimize and avoid conflicts with existing utility infrastructures. Recreation Popular recreational activities in the state include water related activities (boating, kayaking, diving, surfing, swimming, and fishing), land based activities (hiking, camping, hunting), and sports (baseball/softball, volleyball, soccer, football). Generally, impacts to these activities would be short term and limited to the construction phase of the project. To minimize impacts, it is recommended that there is adequate continuous access to significant natural resources in public ownership, and that public beaches be protected for public use and recreation. SOEST Between November 2009 and November 2010, the University of Hawaii School of Ocean and Earth Science and Technology ( SOEST ) prepared four technical reports. The reports discuss the results of ocean floor surveys conducted by SOEST for the purpose of ascertaining the feasibility of laying a power transmission cable on the ocean floor to transport renewable energy generated on Lanai and Molokai to Oahu. SOEST s studies uncovered certain physical obstacles along the proposed cable routes. In particular, SOEST identified the need to modify proposed routes in order to avoid an extensive disposed munitions field along one four mile section of the route south of Oahu. 18

222 Video camera footage recorded by SOEST revealed a number of small aerial bombs, believed to be mustard gas bombs disposed of south of Oahu in SOEST was also tasked with identifying technically feasible routes given legal restrictions and the desires of major stakeholders, including HECO. Specifically, DBEDT asked SOEST to identify potential routes which do not enter federal waters of the Hawaiian Islands Humpback Whale National Marine Sanctuary. 13 SOEST was further informed that HECO favored a south Oahu landing site near Honolulu Harbor and the Iwilei sub station instead of Pearl Harbor. SOEST identified a number of potential routes for the cable, including routes between: (i) Pearl Harbor/Honolulu Harbor and Lanai; (ii) Kaneohe to northwest Molokai; (iii) Northwest Molokai to Lanai; (iv) Oahu to Molokai and (v) Lanai to Maui. Figure 3: Preferred (black) and alternate (dashed) cable routes, other routes surveyed (white dashed), on sunlit bathymetry (gray) and seafloor acoustic imagery (red = strong, green = weak). Pink = existing telecom cable, red dot = observed cable crossing, blue box = bottom fish refuge, red box/circle = dump areas, ruled area = humpback whale sanctuary SOEST s report exclusively addresses the physical feasibility of the interisland cable given the seafloor bathymetry and existence of physical obstacles (both manmade and natural). 12 The Hawaii Undersea Military Munitions Assessment Project assisted in identification of these obstacles. Additional information is available at (last visited October 30, 2012). 13 Under the National Marine Sanctuaries Act, codified in 16 U.S.C , the Secretary of Commerce has the authority to designate certain marine areas designated as sanctuaries. It is unlawful for any person to destroy, cause the loss of, or injure any sanctuary resource. See 16 U.S.C. 1436(a). However, the Secretary of Commerce may issue special use permits authorizing the conduct of specific activities in a national marine sanctuary if the Secretary determines such authorization is necessary to establish conditions of access to and use of any sanctuary resource. See 16 U.S.C. 1441(c)(3). The Secretary of Commerce has delegated responsibility for overseeing the national marine sanctuaries program to the National Oceanic and Atmospheric Administration ( NOAA ). NOAA designated the Whale Sanctuary as National Marine Sanctuary on November 4, See 15 CFR