Project Partners Hawai‘i Natural Energy Institute (HNEI) of University of Hawaii Project Manager & Principal Investigator; Maui Electric Company (MECO) Host utility; Project design, system operator interface, SCADA integration; Cofunding with substation/feeder construction, BESS Hawaiian Electric Company (HECO) Co-host utility; Power systems engineering, cyber security Maui Economic Development Board (MEDB) Education and outreach to community Maui County Community outreach; demand response for county pumping loads Sustainable living Institute of Maui (SLIM) at University of Hawaii Maui College Energy audits and training; participant support SRA International (SRA/Sentech) Requirements definition; system integration; test protocol Silver Spring Networks (SSN) Vendor for communication system, AMI, demand response; Customer outreach support Alstom Vendor for Distribution Management System; SCADA integration support


Presentation Overview • • • • • •

Project Overview Project Objectives System Design Analysis Conclusions and Accomplishments Lessons Learned

Project Overview Maui




1518 Basic MECO System Facts: System peak load ≈ 200MW Fossil fuel and biofuel capacity = about 250 MW

Project will use 2 circuits @ Wailea Substation circuits 1517 and 1518

Kaheawa Wind plant = 30 MW

Maui Meadows ≈ 900 homes Other circuit with resorts & commercial

Up to 90 MW of proposed renewables

Project Objectives Project’s challenge is to determine whether the capabilities of the Smart Grid can improve utility operations and customer service, and if so, at what cost and benefit. • Distribution Level – D-1: Reduce a distribution system’s peak grid energy consumption. – D-2: Improve voltage regulation and power quality on the selected distribution feeder. – D-3: Demonstrate that the architecture of the demonstration project is compatible with additional distribution management system functions, customer functions, and legacy systems. – D-4: Develop and demonstrate solutions to significant increases in distributed solar (photovoltaic systems) technologies.

• By making the distribution system dispatchable, provide Transmission-level benefits of

– T-1: Management of short-timescale intermittency from resources elsewhere in the grid, such as wind energy, solar energy, or load intermittency. – T-2: Management of spinning reserve or load-following regulation. – T-3: Reduction of transmission congestion (through peak curtailment).

Needs and Project Targets Objective Integrate Renewables

Inform Consumer Energy Use Improve Service Quality Reduce Peak Load by 15%

Key Utility Needs

Smart Grid Project Targets

Grid stability with variable resources

DER as substitute for conventional reserves

Limited visibility of distributed RE

Use AMI to monitor distributed PV and voltages

Increasing curtailment

Use BESS and AMI to capture curtailed energy

Monthly bill primary source of information

New displays provide near real time data

New rates being implemented (tiered and TOU)

Help inform consumers on pricing information

Limited outage and voltage violation detection in distribution system

Use AMI to detect outages and improve response

Peaking units very expensive on Maui

Use AMI to detect voltage violations DMS volt-Var control and load flow to reduce losses BESS and DR provide new resources

Limited load research data

AMI/HAN data measure appliance load profiles and help develop DR strategies

Coordinating use of DER

DMS helps manage DER with other system resources

Key Maui-specific Issues

• Improve visibility into the distribution system; evaluate methods to acquire, transmit, process and display the information; data resolution and latency requirements. Data on customer voltages, resulting in better power quality • Understand impacts of distributed photovoltaic (PV) systems on service voltages • How much PV energy is being supplied • Use Demand Response (DR) to reduce peak load and mitigate renewable energy variability • Specification, installation and operation of a Battery Energy Storage System (BESS), including smoothing variability from renewable energy and loads • Identify “Smart Grid” functions, especially “smart meter” functionality, of most value to MECO customers (before system-wide smart meter rollout) • Improved volt/var management • Determine MECO training and staffing requirements for smart grid implementation and operation (meter shop, installers, system operators, etc.) • Integration of AMI, DR and Distribution Management System (DMS) together with MECO’s SCADA/EMS • Experience specifying, procuring and testing smart grid systems

Primary Functions Implemented • Advanced Metering Infrastructure (AMI)

– “Smart” meters - household energy and voltage in 15 minute increments. – Web pages for customer to access web energy use information messages

• Photovoltaic (PV) Metering using additional meter • Demand Response (DR)

– Electric water heaters (WH) turned off – Central air conditioners (A/C) thermostat setpoint raised

• In-Home Display (IHD) of energy use, cost of house and appliances • Battery Energy Storage System (BESS)

– 1 MW / 1.2 MWh battery installed on feeder 1517 close to substation. – Charge and discharge by schedule or MECO command

• Distribution Management System (DMS) – Distribution load flow and volt/var control – Distribution voltage/current monitoring

Technical Approach Main Project Tasks Planning/ Initial Design

Detailed Design / Testing

• Set project objectives • Select Substation / Feeder • Obtain Baseline Data

Implementation / Installation

• Functional Requirements • Contracting/agreements • System Architecture & Data • Equipment Installation Flows o AMI, HAN, BESS, • Vendor Selection DMS • Technical Review • System integration Installation / • Factory Acceptance Test



Data Collection, Analysis, Evaluation • Baseline Data • Metrics / data collection • DOE Reporting • Equipment (decommissioning/ extension)

Outreach & Engagement (Customers, MECO Employees, HI Stakeholders, U.S. Energy Sector) Volunteer/Community Outreach

Develop and approve Outreach Plan

• • • • • • •

Letters to local residents Community meetings Local media stories Project website Project newsletter Recruit and enroll volunteers Ongoing outreach & support

MECO Employees • •

Employee training Inform staff on project and goals

DOE & other Stakeholders • Press releases/ media on project • DOE Reporting • Presentations to Hawaii stakeholders (e.g., PUC) • Spoke at Asia Pacific Economic Cooperation (APEC) Summit

System Design AMI and Communication 20 Energate Load Control Switches (Monitoring only) for home solar

15 Energate PCTs

15 EnergyAware InHome-Displays

195 GE i210+c meters with SSN communication module with HAN (30 providing Voltage Monitoring near Sentient AMPs) 5 GE kV2c meters with SSN Communication module

SSN Data Center Internet

MECO Data Center

Mobile Data Backhaul

2 Cellular Access Points

Wailea Sub Station Maui Meadows SSN Mesh

Voltag e Monit or

Alstom EMS

MECO Backhaul 50 Energate Load Control Switches


10 x 3 Sentient MM2 (one per primary phase, 10 locations)

Master eBridge communicating via DNP3 to Sentient MM2 and connected to the MECO RTU

AMI: IPv6 AMI and HAN traffic DA: IPv4 DNP3 Web Services: AMM to MS Web Services: Voltage Monitoring to DMS Web Services: DRM to MS

System Design Distribution Management System

System Design Cyber Security • Approach – Designed as part of the system – not an add-on – Defense in Depth Architecture - multiple layers of security controls placed throughout the system • Cyber Standards: PCI Version 3.0, ISO 27001/02, NIST 800-53, NERCCIP, NIST-7628, FIPS 140-2

• Cross domain segregation – Feeder Current Transformer to MECO RTU – MECO SCADA System interconnection with DMS – DMS / MECO SCADA and MECO IT

• Protect customer information Security plans were reviewed and approved by MECO and HECO cyber Security

Communication and Outreach •

• •

Customer Recruitment Lead: Maui Economic All participants were Development Board offered an energy audit All participants were volunteers Ideal scenario - mix of customers with smart meters, PV, in-home devices such as programmable thermostats, monitored appliances, and web portal access Development of several communication tools (website, mailers / trifolds, fact sheets, etc.)

Communication and Outreach Outreach Activities

All participants were volunteers • Initial meetings with Maui Meadows Homeowners Associations and residents • Community information events • Meetings with local government, community and environmental groups • Media interviews • Website • Home visits • Customer surveys • Newsletters • Monthly coffee groups

ESSENTIAL: Continuing feedback, participant meetings, fast response to problems

Workforce Development • SLIM at UHMC – Provided energy audits to home volunteers – Trained local students in audits and clean energy technologies

• SLIM Program Goals – Train students in home energy management skills – Provide students with real-world experience – Train a local workforce in energy management – Prepare students to gain employment in the energy management field – Offer free home energy audits to volunteers

Workforce Development

Workforce Development

Analysis and Evaluation • • • •

Residential customer load and voltage profiles Customer energy use Estimating PV output Use of customer energy use Web portals and in-home displays (IHD) • Demand Response (DR) • Battery Energy Storage System (BESS) • Distribution system visibility (voltage support)

Analysis and Evaluation Residential Customer Load and Voltage Profiles

• Observed Maui Meadows load shapes similar to MECO’s estimates for residential class load profile • Observed instances of high and low voltage (out of acceptable range) – High voltages may be caused by PV power injection – MECO adjusted transformer tap settings to bring voltages within acceptable range – showed value of increased visibility – Future MECO distribution transformers purchased will have the ability tot adjust voltage down, not just raise it

Analysis and Evaluation Customer Energy Use • MSG volunteers reduced energy use 23% after smart meters were installed • Audited homes reduced their energy use by 37% • Smart meters (and associated Web portal) gave participants the ability to reduce their energy consumption significantly • Energy audits discovered setpoint errors, equipment malfunctions, and other home issues. • Audits were another opportunity to answer questions and raise energy awareness. The smart meters demonstrated their ability to help customers save energy

Analysis and Evaluation Estimating PV Output

• In the aggregate, the feeder’s inverters behaved similarly • Due to the effects of PV (and its variability) voltages, a much faster sampling rate than 15 minutes is needed • An irradiance sensor in the substation can be used effectively to estimate a PV generation on a feeder, and its data can be read every 4 seconds through the SCADA system Determining the actual amount of electricity being generated by PV was a key project objective

Analysis and Evaluation Customer Use of Energy Web Portals and In-Home Displays (IHD) • Customer access of the Web portal declined as the project progressed • Most customers just glanced quickly at the portal when they accessed it • A few customers used the portal to analyze their energy consumption in depth

• Customers with higher electricity use tended to be more likely to request IHDs • Customers with IHDs did not appear to save more energy than those with Web portal only • Some customers really liked their IHDs IHDs did not appear to offer significant additional value to the portal alone

Analysis and Evaluation Demand Response (DR) • Raise A/C thermostat 3⁰ F 3- 4 PM • Turn off WH 7 – 8 PM • May extend WH control into the night to raise MECO minimum loads, reduce wind curtailment • These DR strategies seem to be acceptable to customers • DR appears to be effective in reducing system (A/C) or feeder (WH) peak

• DR can compensate for a sudden drop in renewable energy output • DR can reduce MECO system and feeder (residential class) loads • May be possible to defer WH use until late night, raising minimum load and recuing wind curtailment • More tests, load research needed Successfully demonstrated residential DR in Maui

Analysis and Evaluation Battery Energy Storage System (BESS) • 1 MW / 1.23 MWh (1 MWh effective capacity) battery installed on Feeder 1517 near the substation • MECO’s first experience procuring, installing, operating a large battery

• Developed procedures and standards for battery installation, safety and fire protection • Refined specifications, factory and site acceptance test procedures for future BESS procurements • Able to test battery operating modes, obtain performance data • MECO is better prepared to specify, procure, install, commission a BESS • MECO successfully integrated BESS management into SCADA

Analysis and Evaluation Battery Energy Storage System (BESS) • • • •

Successful peak shaving Successful load following Voltage and reactive power support Reduce transmission loading and congestion

• Reduced feeder peak by over 20% • Load following – can mitigate renewable energy variability • Feeder voltage support –active power injection • Transmission congestion relief – supply reactive power • Charge at night, in part utilizing wind power BESS proved effective at reducing peak loads and transmission congestion, as well as enabling MECO to support more as-available renewable energy

Analysis and Evaluation Distribution Management System • Implemented pilot DMS • Validated feeder load model • Study mode – to determine best LTC and capacitor settings for voltage support • Integrated with SCADA / EMS

• Improved customer service – helped maintain feeder voltages within limits • No opportunity to test Outage Management System: there we no outages Identified additional modeling and analysis capabilities needed for high penetrations of PV

Summary of Key Findings • Smart meters (with Web portal) help customers understand and reduce their energy consumption ‒ In-home displays did not result in more energy reduction than Web portal • Increased visibility into the distribution system (with smart meters and DMS) can detect unexpected voltage excursions and help eliminate them, improving service quality • Irradiance sensors in substations can provide necessary fast-updated estimates of PV generation • Demand Response/Load Control can help mitigate sudden drops in renewable energy production ‒ It may be possible to defer WH until low load periods at night • BESS is effective for: ‒ Peak reduction (feeder/substation and system) ‒ Increasing minimum system loads (charging at night reduces curtailment of wind) ‒ Mitigating variability of renewable energy (load following mode) ‒ Supporting voltage on the feeder ‒ Reducing transmission loading through peak reduction and by supplying reactive power

SUMMARY OF RESULTS • Battery Energy Storage System – BESS is effective in load following mode, to “smooth” variations in loads and/or renewable energy production. – The load following control is effective for minimizing peak. – When located on the feeder, BESS charging and discharging does not markedly affect substation voltage. Setting feeder voltage is best done using transformer tap changers, switched capacitors, or other means. – BESS is most effective at supplying active power on the feeder, to reduce substation transformer and transmission system load . – BESS can supply reactive power without significantly affecting state of charge. Having BESS supply reactive power is effective in reducing transmission losses and reduces transmission congestion

Objectives Met!

• D-1: Reduce a distribution system’s peak grid energy consumption: BESS reduced Feeder peak load by over 20%. Proof of concept of two DR programs. • D-2: Improve voltage regulation and power quality within the selected distribution feeder: AMI and DMS detected out of voltage occurrences and helped mitigate them. • D-3: Demonstrate that the architecture of the demonstration project is compatible with additional distribution management system functions, customer functions, and legacy systems: The platform developed in the project supported AMI, DR, BESS, IVVC, and improved system visibility. These were integrated with legacy SCADA and transformer tap changer control systems. Lessons learned have already been applied to new MECO / HECO projects. • D-4: Develop and demonstrate solutions to significant increases in distributed PV: AMI & DMS provided a process to estimate PV output using irradiance sensors. BESS was effective in smoothing variations in PV. • T-1: Provision for management of short-timescale intermittency from resources elsewhere in the grid, such as wind energy, solar energy, or load intermittency: BESS was dispatched to mitigate short time scale intermittency. The monitoring of PV output and analysis of its variability showed the need for faster monitoring of PV status. • T-2: Provision for management of spinning reserve or load-following regulation: BESS was operated successfully in load following/regulation mode. • T-3: Reduction of transmission congestion (through curtailment of peak load): BESS was used to supply real and reactive power on Feeder 1517, reducing transmission congestion.

Maui Smart Grid Project Goals Achieved Distributed Resources for Transmission-level Support • Reduce distribution circuit peak loading by >15% • By demand response, switching peak loads to energy storage, and supporting more renewable energy • Improve service quality • By improved visibility, voltage monitoring, and volt/var control study mode • Enable consumers to manage their energy use to minimize electric bills • By using AMI “smart meters” with customer portals • Support grid stability • Through controllable loads, storage, and improved voltage/current information • Enable greater utilization of as-available renewable energy sources • By providing measurement and estimation of distributed PV to the utility operator • By mitigating PV variability through BESS • By increasing minimum system load (BESS, DR), thus reducing wind curtailment 30

Lessons Learned • Smart Grid system design and technology options – – – –

Appropriate requirements for system-wide smart grid functions Vendor Selection and Procurement Volt/Var Control (VVC) methods Integration With Other Utility Systems, Especially Legacy Systems – Data Management and Sampling Intervals – Managing security and access to systems and data

• Customer Interface and Education – outreach, education, communication are critically important • MECO staff training ‒ Familiarize all utility staff with the project ‒ As a result of this project, MECO can better estimate training and staff support requirements for smart grid functions Going forward, MECO and HECO are better prepared to design, implement and operate advanced technologies consistent with providing reliable, affordable and environmentally compatible electricity to its customers.