Planning for a Hawai'i Powered future

Explore the energy models that are part of our Integrated Grid Planning process.

Hawaii's energy future

Hawaii Powered is our vision for powering the grid with 100% local, renewable energy. It celebrates finding solutions for a clean energy future right here in Hawaii.

As we plan for a Hawaii Powered future, we're focused on:

Affordability: Stabilizing electricity bills by using more local, renewable energy sources, which will allow us to rely less on imported oil.
Reliability: Planning ensures we have enough power to keep the lights on, especially as demand for electricity continues to grow.
Resilience and sustainability: By using local, renewable resources, we support the state's goals for decarbonizing the economy, while also building a stronger, more resilient energy system that can better withstand storms and emergencies.

Hawaii's unique energy challenges and opportunities

Energy planning is critical everywhere, but planning in Hawaii comes with unique challenges and opportunities for our island way of life:

Island isolation: Because of our geographic isolation, we can't rely on neighboring states for backup power. That's why it's critical we plan carefully to ensure we have enough energy all year round.
Volatile import costs: Hawaiian Electric currently imports about 60% of our fuel for generators, making us vulnerable to fluctuating global fuel costs. Energy planning for more local, renewable resources helps reduce our reliance on imported fuels and stabilizes customer bills over time.
Achieving 100% clean energy goals by 2045: The world is watching as we innovate to scale up clean energy on our islands and do our part to address the changing climate.

Our data-driven approach

Hawaiian Electric applies a data-driven approach using analytical planning models to define and evaluate resource plans for the future electric grid that are affordable, reliable and resilient and can adapt to meet Hawaii's energy challenges and goals.

Presented below are the key planning models and concepts of the Hawaiian Electric Integrated Grid Planning (IGP) process.

Inside the planning process

Integrated Grid Planning

We approach energy planning through a process called Integrated Grid Planning, or “IGP.” This process brings many people together to build a resilient and reliable grid from diverse renewable resources.

In 2023, we filed the Integrated Grid Plan, which proposes actionable steps to decarbonize the electric grid on the state's timeline, with a flexible framework that can adapt to future technologies. The Integrated Grid Plan is the culmination of more than 5 years of technical analyses and partnership with stakeholders and community members across the islands.

The Integrated Grid Plan underscores the urgency of action needed to achieve a decarbonized future and shows that every industry and individual will need to play a role. It's everyone's kuleana to create a sustainable future for Hawaii.

Read the Integrated Grid Plan and the latest annual update.

What do you want to know?

Energy data deep dive

Our approach to modeling

Throughout our Integrated Grid Planning process, we use advanced modeling tools to consider a lot of data, evaluate complex trade-offs and analyze different possibilities.

Modeling helps us make informed decisions to meet our long-term goals, such as reaching the state's objective to use 100% renewable resources to power the electric grid by 2045 and achieve a net zero emissions clean economy by 2045.

Whenever we use a model, we take steps before and after modeling.

We ask key questions about resources, reliability, affordability and interconnection.
We gather data that serves as our inputs. This data includes estimates for resource costs, load forecasts and policy impacts.
We use modeling tools to provide different types of outputs and compare scenarios (for example, base, low and high bookends for energy demand).
We interpret and evaluate results to answer questions and create action plans for next steps.

Forecast future energy demands – Models help forecast how much energy people will use, and when, as more facets of daily life run on electricity. They also factor in renewable and emerging technologies like rooftop solar and electric vehicles.
Identify future transmission and distribution grid needs to host future demand and generation – Models help determine transmission and distribution grid infrastructure upgrades to host future demand and generation, such as increasing the size of circuits, adding more resources to maintain system stability and system upgrades to maintain effectiveness of protection systems. These upgrades enable customers to choose when and how they use electricity—for example, charging their electric vehicle or supplying electricity from their rooftop solar system. It also enables all customers access to low-cost renewable energy by interconnecting new renewable projects to the grid.
Find low-cost energy solutions and guide investments – Models help identify the lowest-cost mix of energy resources while meeting reliability and policy goals. This guides investment decisions about building new energy sources—like solar, wind, battery storage or firm generation—to meet future needs.
Evaluate different energy scenarios – Modeling tests how the grid would perform under various conditions, such as policy changes on renewable standards or emissions targets, shifts in demand due to changes in customer behavior and availability of new technologies.

Modeling cannot:

Predict the future with 100% certainty or guarantee specific outcomes – Models and simulations provide best estimates of how equipment will operate, but actual costs, policies and technologies can change. Results also depend on the data and assumptions used, which can evolve over time.
Account for every real-world factor – Models consider a lot of data and complex factors, but they cannot capture absolutely everything, such as all possible market shifts, daily use patterns, weather events or technological breakthroughs.
Make decisions on their own or replace the need for human judgment – Models don't make decisions—people do. Our staff, policymakers, partners and community input are essential in shaping final energy plans.
Solve every energy challenge instantly – Models are one tool in a continuous planning process that requires ongoing updates and adjustments.

Keep exploring for information about each individual model.

The planning process: many models working together

In energy planning, no single model tells the whole story. We use a series of models, each with a specific role, to help us understand what the future may hold and guide our decisions. The models work together, step by step.

Click each step to learn about the modeling process

Models: MetrixND and MetrixLT

We start by estimating how much energy customers will need in the future as we all increasingly rely on electricity to power our lives. The MetrixND and MetrixLT models help us forecast future demand on the grid.

Models: RESOLVE and PLEXOS

Next, we identify the energy resources needed to meet that future demand. We use RESOLVE to create a resource portfolio and then test its reliability using PLEXOS. If that portfolio doesn't meet reliability standards, we refine it—either manually or by re-running RESOLVE with more stringent reliability requirements —before testing again.

Transmission needs Models: PSSE/PSCAD

Large-scale resources must be connected to the grid. PSSE and PSCAD help us determine what upgrades are needed to interconnect new large-scale resources, carry larger loads and expand the transmission system.

Distribution needs Models: LoadSeer and Synergi

At the local level, we also assess how to connect customer-sited resources like rooftop solar. LoadSeer and Synergi help us identify where upgrades are needed to support growth on the distribution network.

Once we've used these models to identify future energy needs—from demand to generation to transmission and distribution—we combine the results into a unified, actionable plan. That's the Integrated Grid Plan: a roadmap outlining the steps we'll take to reach 100% renewable energy while continuing to deliver reliable power.

Meet the models

We start by asking questions—then use the right model to help find answers.

Get Started

Click a key question to learn more about each model

Key question:

How does Hawaiian Electric plan for increasing loads due to evolving customer choices and increasing electrification of daily life?

Key question:

What energy resources should Hawaiian Electric invest in, when should they be added and in what amounts?

Key question:

How does Hawaiian Electric ensure that the electric grid remains reliable given future expected changes in demand for electricity and resource mix?

Key question:

How does Hawaiian Electric plan for new transmission investments to interconnect new large-scale resources while ensuring the grid is stable during disturbances and equipment failure?

Key question:

How does Hawaiian Electric plan upgrades to the distribution system to connect new customer solar and batteries, and to meet growing electricity demand?

Key Questions

Future demand

Model: MetrixND/MetrixLT

Load forecast is an estimate of how much electricity our customers will use. It helps us understand and plan for how much power we'll need to deliver, especially during times of high demand.

The forecast takes into account things like weather, the economy and how quickly customers might adopt technologies like rooftop solar, batteries, electric vehicles and energy efficiency programs.

Tools like MetrixND and MetrixLT help model and combine all these factors into monthly and hourly load forecasts.

What do MetrixND and MetrixLT do?

MetrixND and MetrixLT are software tools that help predict how much electricity our customers will use in the future. These tools help us forecast the underlying load (amount of electricity customers use) before accounting for factors like rooftop solar, electric vehicles and energy efficiency.

The underlying load forecast is driven primarily by the economy, weather, electricity price and known changes to large customer loads. The forecast is informed by historical data, structural changes (like shifts in new technology, appliance adoption or major changes in customer behavior) and disruptions like COVID-19, volcanic eruptions and other natural disasters. MetrixND is primarily responsible for developing the underlying load forecast at the monthly level, then MetrixLT transforms monthly-level forecasts to hourly forecasts.

Additionally, these tools:

Offer a variety of statistical methods and models
Enable forecasters to quickly evaluate and select different models that can cater to each island's specific characteristics
Can generate forecasts at the monthly and hourly level
Are subject to garbage-in-garbage-out like any other software tool; input data quality plays an important role in developing reliable forecasts

Weather

Electricity use can be heavily dependent on the weather. Depending on variables like how hot and humid the weather is, electricity consumption can vary widely.

NOAA: Weather data primarily comes from the National Oceanic and Atmospheric Administration's (NOAA) National Center for Environmental Information (NCEI) office.

Economic indicators

Economic trends play an important role in forecasting future energy use. As these economic factors change over time, so does the demand for electricity. New construction and development, tourism fluctuations, the cost of imported fuels and the job market all shape how households and businesses use electricity.

UHERO: The Company partners with the University of Hawaii Economic Research Organization (UHERO) for historical and forecasted economic indicators such as personal income, population, jobs, etc.

Other indicators and flags

The models can benefit from additional data that identifies certain events and conditions by using dummy variables or scaling factors. These help the model consider how different conditions (such as day-of-the-week or major events, like the COVID-19 pandemic) may impact electricity usage.

Underlying load forecast

The underlying load forecast is the estimated electricity use by customers before considering the impacts of rooftop solar, electric vehicles and energy efficiency. MetrixND derives the underlying load forecast at the monthly level which is then transformed into hourly forecasts using MetrixLT.

Hourly forecasts

The final result from MetrixLT is an hourly forecast that originally started at the monthly level. MetrixLT is agnostic to the type of monthly forecast it starts with.

The following images depict outputs from MetrixND and MetrixLT.

Informing action

This forecast is the starting point for other modeling work, including PLEXOS and RESOLVE. We used MetrixND and MetrixLT outputs to inform our IGP action plan and preferred plans for each island we serve, including Oahu, Hawaii and Maui County.

Connect with the modeling team

Resource mix

Model: RESOLVE

What does RESOLVE do?

RESOLVE is an optimization model that recommends the lowest-cost renewable energy resources that meet a wide variety of constraints.

This model:

Helps us find the lowest-cost resource portfolio, selecting from a diverse set of potential resources including wind, solar, storage and firm generation.
Are subject to constraints like decarbonization goals, reliability targets, hourly demands and operational needs.
Considers a set of representative modeled days to make investment decisions.

RESOLVE allows grid planners to evaluate investments across many scenarios and system conditions such as:

Assess demand
- Baseline load growth and electrification impacts
- Hourly demand evolution
Achieve decarbonization
- Emissions reduction targets
- Renewable portfolio standard (RPS) goals
- Understanding new policies
Ensure reliability
- Resource adequacy balance
- Future system conditions
- Reserve requirements
Evaluate resources
- Renewable energy options
- Battery storage operations
- Firm and flexible resources

Loads

System (island-wide) demand is represented both annually and hourly to understand how daily and seasonal hourly load patterns can be satisfied in addition to total annual load. Examining the impacts of different demand layers (underlying, distributed energy resources, energy efficiency, electrification and time-of-use rates) over a long-time horizon on the resource portfolio is crucial to ensure the system is capable of handling future energy needs and changes in load patterns.

Underlying: Economic data from University of Hawaii Economic Research Organization (UHERO), weather data from National Oceanic and Atmospheric Administration (NOAA), historical Hawaiian Electric data on customer usage and count and various publicly available sources such as media and government agency reports.
Distributed energy resources: Current pace of incoming applications and executed agreements for existing programs, future costs of PV and battery systems from National Renewable Energy Laboratory (NREL) Annual Technology Baseline (NREL ATB), working group feedback on the size of the residential and commercial markets for new systems and Commission guidance on the assumed structure of future programs (including any applicable incentives at the State and Federal levels).
Energy efficiency: Projections developed by the Applied Energy Group (AEG) in their July 2020 State of Hawaii Market Potential Study.
Electrification: Light-duty vehicle registrations and other vehicle characteristics from government and industry agencies, such as Hawaii Department of Business Development & Tourism and the Federal Highway Administration.
Time-of-use rates: Program design discussed in the distributed energy resources docket as part of advanced rate design.

Resources

Existing, planned and candidate new resource options are represented in the model with specific operational characteristics for each. Existing and planned resources are dispatched to meet hourly and annual system demand. The model also builds new resources to address growing demand or to take advantage of favorable new resource economics.

Most of our resource data comes from the National Renewable Energy Laboratory (NREL) Annual Technology Baseline for resource costs and the NREL Wind Toolkit, National Solar Radiation Database and System Advisor Model for weather-dependent resource profiles. Historical production data is also used for existing resources. Fuel prices come from data developed by the U.S. Energy Information Administration and correlated to historical Hawaiian Electric fuel prices.

Policies and grid requirements

System requirements and policies—such as decarbonization goals, reliability targets, reserves or other grid needs—are also accounted for in the model. These constraints are considered when making investments in new resources or dispatching resources to meet hourly and annual demand.

Our policy and grid requirement data is informed by state law as well as Public Utilities Commission guidance in several key dockets such as IGP, IGP Request for Proposals, Distributed Energy Resources and Performance-Based Regulation. Policy and grid requirement data can also come from Hawaiian Electric's own standards and planning criteria.

The RESOLVE model offers a roadmap for energy planners. Given the complex and shifting nature of energy planning, we aren't done after RESOLVE. From there, we refine the plan with additional modeling to ensure every hour of the day can be met with the lowest-cost resource available.

Total costs

The cost of addressing the future system needs is a key metric that helps planners understand the impact on customer rates over time. Variable operational costs as well as new investment costs are reported to see the breakdown of where dollars are being spent.

Long-run marginal costs
(a valuation of avoided fixed and operating costs)

By comparing resource plans developed by RESOLVE, planners can identify the long-run avoided costs of planned resources. This is helpful for valuing resources that are part of the load forecast like rooftop solar or energy efficiency by comparing the resource build and total cost of plans with and without their adoption. The difference in total cost can then be used to inform what the program cost or incentives should be.

Timing, type and quantity of new resource investments

The amount of power from newly built resources are reported each year to help planners understand the timing, type and quantity of investments that need to be made. All resources have operational parameters reported, such as annual generation, capacity factor, fuel use and more, so that grid planners can understand the performance of its fleet of generators.

Satisfying system needs

Information about how constraints—such as reliability requirements, decarbonization targets and others—are satisfied. For example, annual emissions are reported, which indicates how a decarbonization target is achieved and to what extent (i.e., the target drives results or is easily met).

Informing action

We used RESOLVE outputs to inform our IGP action plan and preferred plans for each island we serve, including Oahu, Hawaii and Maui County. The resource plan developed in RESOLVE is then tested in the PLEXOS model to assess costs and system reliability in finer detail.

Download this presentation to the Stakeholder Technical Working Group (November 2022) for detailed information about how RESOLVE outputs informed our recommendations and decisions.

What's an example of when Hawaiian Electric used the RESOLVE model?

The following chart shows an example of what the RESOLVE model helped us understand throughout IGP. This chart shows how the Oahu resource mix changes through 2050 under different load scenarios.

Connect with the modeling team

Cost and reliability

Model: PLEXOS

What does PLEXOS do?

The PLEXOS model is focused on optimizing the data determined by RESOLVE. It finds the lowest cost for when to deploy various types of generation and helps us assess the reliability of the resource mix.

PLEXOS allows grid planners to evaluate the cost, operation and reliability of the various resource mixes developed by RESOLVE, such as:

Assessing cost of electricity production
- Cost of electricity production by hour and by year that includes fuel costs, operations and maintenance costs, and payments made to independent power producers.
Ensuring reliability
- Stress testing whether the resources generating power at any particular hour are going to meet reliability, given multiple weather scenarios and unit outage conditions.
Deciding which energy sources to use and when
- Grid planners use the data to help make decisions about which power generation source to dispatch at any given time, while also weighing cost considerations, to ensure energy demands can be met at all times.

Loads

System (island-wide) demand is represented hourly to understand how daily and seasonal hourly load patterns can be met in addition to total annual load. Examining the impacts of different layers (underlying, distributed energy resources, energy efficiency, electrification and time-of-use rates) over a long-time horizon on the resource portfolio is crucial to ensure the system is capable of handling future energy needs and changes in load patterns. The same load data used by RESOLVE is also used by PLEXOS.

Similar to RESOLVE, the PLEXOS model uses the same data sources to define the characteristics of each resource including the NREL Wind Toolkit, National Solar Radiation Database and System Advisor Model for weather-dependent resource profiles, historical production data for existing resources and fuel prices from data developed by the U.S. Energy Information Administration and correlated to historical Hawaiian Electric fuel prices.

Underlying: Economic data from University of Hawaii Economic Research Organization (UHERO), weather data from National Oceanic and Atmospheric Administration (NOAA), historical Hawaiian Electric data on customer usage and count and various publicly available sources such as media and government agency reports.
Distributed energy resources: Current pace of incoming applications and executed agreements for existing programs, future costs of PV and battery systems from National Renewable Energy Laboratory (NREL) Annual Technology Baseline (NREL ATB), working group feedback on the size of the residential and commercial markets for new systems and Commission guidance on the assumed structure of future programs (including any applicable incentives at the State and Federal levels).
Energy efficiency: Projections developed by the Applied Energy Group (AEG) in their July 2020 State of Hawaii Market Potential Study.
Electrification: Light-duty vehicle registrations and other vehicle characteristics from government and industry agencies, such as Hawaii Department of Business Development & Tourism and the Federal Highway Administration.
Time-of-use rates: Program design discussed in the distributed energy resources docket as part of advanced rate design.

Resources

Existing, planned and new resource options are represented in the model with specific operational characteristics for each and are dispatched to meet hourly and annual system demand.

The RESOLVE model determines the resource mix that is evaluated in PLEXOS. However, the same data to characterize how those resources operate are used by both models.

Policies and grid requirements

System requirements and policies—such as reliability targets, reserves or other grid needs—are also accounted for in the model. These constraints are considered when dispatching resources to meet hourly demand.

Because RESOLVE already builds a resource mix that meets state policies and grid requirements, PLEXOS doesn't check those rules again. Instead, PLEXOS makes sure the system can run at the lowest cost and at an hourly scale while still meeting those same requirements.

The PLEXOS outputs also provide an initial set of dispatch conditions that are used to assess transmission needs.

Core outputs:

Total costs

The cost of addressing the future system needs is a key metric that helps planners understand the impact on customer rates over time. Non-fixed costs can be broken down by cost category such as fuel, operations and maintenance and payments to independent power producers. New investment costs are also reported to see the breakdown of where dollars are being spent.

Short-run marginal costs
(a valuation of avoided operating costs)

Short-run marginal costs represent the cost of producing the next megawatt-hour of electricity. They include the costs of fuel, operations and maintenance and labor that are needed to produce electricity without requiring new investment in generation, transmission or distribution capacity. For example, hours with a high short-run marginal cost can occur when grid operators need to start up additional firm generation to serve the evening peak demand. Planners can use the short-run marginal cost as a way to evaluate new rooftop solar or energy efficiency programs that can target those hours with a high short-run marginal cost because they can avoid the start-up of firm generation.

Resource operations

Hourly dispatch and annual generation of existing and planned resources helps grid planners understand how the system responds to changes in demand shape and growth, seasonal changes in solar and wind production, and outages of firm generation.

Satisfying system needs

Stress testing the resource mix to see if it is reliable under different system conditions is especially important to understand how the system responds when different combinations of random and unexpected outages of firm generation occur, and when different weather patterns limit the production of solar and wind generation resources. The reliability of the system is then measured as the number of days per year that the resource mix was unable to serve the demand. For the system to be considered reliable, the number of days with unserved demand must be equal to or less than a standard threshold. The U.S. Mainland typically uses 0.1 days per year. The Commission has allowed the use of a standard of 0.1 days per year across the islands that may be modified in the future based on the recommendations of the Hawaii Electric Reliability Administrator.

Informing action

The dispatch of resources developed by PLEXOS are used as the starting point for the PSSE/PSCAD models to identify transmission needs. We used PLEXOS outputs to inform our IGP action plan and preferred plans for each island we serve, including Oahu, Hawaii and Maui County.

Download this presentation to the Stakeholder Technical Working Group (January 2023) for detailed information about how outputs informed our recommendations and decisions.

What's an example of when Hawaiian Electric used the PLEXOS model?

We used the PLEXOS model throughout the IGP process to examine the operations of the resource mix as it changes over time. The following charts show an example of what the PLEXOS model helped us understand throughout IGP as Oahu undergoes changes to its resource mix from 2030 to 2035. New offshore wind (dark blue) and biodiesel generation (pink) help to displace existing fossil fuel generation (black).

Connect with the modeling team

Transmission needs

Model: PSSE / PSCAD

What do PSSE and PSCAD do?

PSSE and PSCAD are simulation tools that help energy planners determine transmission grid needs. These tools model various sophisticated engineering and power system phenomena that may occur in the process of delivering electricity. The results of these models help planners make decisions about grid need upgrades that will allow the smooth and continuous delivery of electricity.

PSSE and PSCAD also allow grid planners to evaluate transmission grid capacity expansion and stability enhancement options such as:

Assessing transmission grid capacity
- Ensures the number and size of transmission circuits allow for proper power flow to maintain reliability, as overloading an electrical line will cause it to burn.
Evaluating renewable energy zones
- New renewable resources will need to be interconnected to achieve decarbonization goals. New resources in a renewable energy zone undergo a complex review of new generation siting and existing transmission to confirm if the transmission in that area is adequate to accommodate the new resources.
Evaluating system stability
- System frequency and voltages are measured to determine if the system is stable under its current configuration and under certain scenarios, such as periods where there is a high amount of generation from rooftop solar.

Transmission system design

Transmission system design defines how power plants, substations and transmission lines connect to form the electric grid. It includes where these components are located and technical details like how much electricity they can carry, how easily electricity flows through them and how well the system can respond to changes. The configuration of the transmission system and where resources are located affects how electricity moves across the grid, how reliable the system is and how well it can handle different conditions.

Transmission geographic information system (GIS) demand

Island-wide energy demand is represented both annually and hourly to evaluate transmission grid capacity and stability under various system conditions. Examining the impacts of different levels of system demand with various generation dispatches over a long-time horizon on the resource portfolio is crucial to ensure the transmission system is meeting planning criteria requirements.

The same load data used by RESOLVE is also used by PLEXOS.

Underlying: Economic data from University of Hawaii Economic Research Organization (UHERO), weather data from National Oceanic and Atmospheric Administration (NOAA), historical Hawaiian Electric data on customer usage and count and various publicly available sources such as media and government agency reports.
Distributed energy resources: Current pace of incoming applications and executed agreements for existing programs, future costs of PV and battery systems from National Renewable Energy Laboratory (NREL) Annual Technology Baseline (NREL ATB), working group feedback on the size of the residential and commercial markets for new systems and Commission guidance on the assumed structure of future programs (including any applicable incentives at the State and Federal levels).
Energy efficiency: Projections developed by the Applied Energy Group (AEG) in their July 2020 State of Hawaii Market Potential Study.
Electrification: Light-duty vehicle registrations and other vehicle characteristics from government and industry agencies, such as Hawaii Department of Business Development & Tourism and the Federal Highway Administration.
Time-of-use rates: Program design discussed in the distributed energy resources docket as part of advanced rate design.

Resources

Existing, planned and new resource options are represented in the model with specific operational characteristics for each.

Resource dispatch data (how much and when resources are deployed and dispatched) is developed in PLEXOS. Information about how resources perform—like their maximum output and their ability to support voltage and stay connected during grid disturbances—comes from technical requirements and documents provided by the generation developers.

Policies and grid requirements

System requirements—such as having sufficient reserves that can respond quickly when parts of the grid are on outage—are also accounted for in the model.

Our policy and grid requirement data is informed by Hawaiian Electric's transmission planning criteria, Rule 14h technical requirements and grid-scale project performance standards.

Monitoring voltage to protect the grid

Voltage levels are monitored during normal conditions, as well as moments when parts of the system are on an outage (like a transmission line going out) to ensure that they can be kept within safe limits. It provides guidance to planners on where the system may struggle to deliver reliable power to customers.

Transmission line capacity

Transmission line capacities are studied to check whether the transmission lines are getting too close to their limits under normal and outage conditions. The study results are assessed against the planning rules to determine if new transmissions lines or other upgrades are needed.

Tracking grid stability in real time

Voltage, current and frequency are examined more closely during short emergency events (lasting 30 to 60 seconds) as a measure of the system's overall stability. These results help planners understand how stable the system is and how power sources respond during problematic conditions.

Informing action

We used PSSE and PSCAD outputs to inform our IGP action plan and preferred plans for each island we serve, including Oahu, Hawaii and Maui County.

Download this presentation to the Stakeholder Technical Working Group (2021) for detailed information about how outputs informed our recommendations and decisions.

What's an example of when Hawaiian Electric used the PLEXOS model?

PSSE:

One way we use the PSSE model is to study how new transmission lines can support future electricity needs and connect renewable energy sources.

In the following figure, dashed lines represent new transmission projects proposed in the IGP System Security Study for Oahu that would help meet the island's power demand by 2045. These new lines would keep the system operating within Hawaiian Electric's planning standards.

PSCAD:

One way we use the PSCAD model is to evaluate the stability of the grid with future resources. The figure to the rightbelow shows a PSCAD evaluation in action. The model starts with system frequency, power and voltage at rest. A short-circuit event is triggered in the model at 10 seconds and cleared shortly thereafter. The event briefly causes system frequency to spike and voltage to drop, with a reduction in output from new solar and storage (IBR) resources. We then monitor system frequency and voltages in response to the combined reactions from all connected resources. Through this analysis, we're making sure the system remains stable when outage events occur and there is no disruption in service.

Connect with the modeling team

Distribution needs

Model: LoadSEER / Synergi

What do LoadSEER and Synergi Electric do?

LoadSEER is a spatial forecasting tool for the distribution system that helps planners determine system grid needs (thermal capacity) to meet distribution planning requirements while considering the system demand, distributed energy resources (DER), electric vehicles (EV) and energy efficiency (EE) forecasts.

Synergi Electric is a power system simulation tool that analyzes several categories including voltage, current and power flow in the electrical distribution system. These models:

Help us build forecasts at the distribution/community level based on the system level forecasts and evaluate how the two work together
Are subject to constraints, including:
- Availability and quality of input data. For example, supervisory control and data acquisition (SCADA) measurement data is not available for every distribution feeder, so assumptions are made for feeders without this data available.
- Primary voltage level. The distribution planning analysis is focused on the thermal and voltage limits of the primary voltage system rather than the secondary voltage system.
- Customer usage patterns (load shapes). Historical customer usage patterns are used to forecast future customer behavior.
Both models allow grid planners to:

Evaluate thermal capacity overloads under various forecast scenarios and develop grid need profiles for these cases.
Estimate how much energy is needed on each distribution feeder based on the forecast for DER and use Synergi Electric to perform a hosting capacity analysis.
Determine mitigations needed to address grid needs due to the hosting capacity and location-based forecast analysis.

Demand

The forecasted profiles are developed using data on equipment usage, customer requests for future power needs and overall system demand projections.

The same load data used by RESOLVE is also used by PLEXOS.

Equipment loading: Company supervisory control and data acquisition (SCADA) measurements and aggregated advanced metering infrastructure (AMI) data provide historical profiles.
Service requests for future load: Customer requests from Customer Engineering identify size and location of future services.
System forecasts: Island-wide forecasts for underlying load, DER, energy efficiency and electric vehicle adoption.

Spatial factors

LoadSEER considers various spatial factors, such as rooftop solar, to determine the likelihood of new load growth and DER adoption in certain areas when growing the distribution system up to the forecasts.

Locations of existing services and DER systems: Internal company databases like the Customer Information System, Locational Value Maps and Customer Interconnection Tool
Substation and feeder topology from GIS and engineering drawings provide equipment information
Hawaii solar radiation maps
Census tract/American Community Survey: impactful categories include median income, owner occupied (%), renter occupied (%) median home value
Weather: We use historical weather data to examine how weather affects customer demand seasonally (and is shaped customer-sited solar)

Grid requirements

Both models also consider grid requirements, such as how much generation the equipment can handle at any one time and voltage criteria, to determine when system projects are required.

Grid requirements come from Hawaiian Electric's own standards and planning criteria including Distribution Planning Methodology.

Distribution feeder level DER forecasts

The LoadSEER generates a forecast that illustrates a scenario for how the system may look in the future depending on adoption of rooftop solar and batteries, electric vehicles and energy efficiency within a certain community at any one time, and how Hawaiian Electric must plan for their adoption. The feeder DER hosting capacity limits are determined in Synergi Electric.

10-year transformer and feeder profiles

LoadSEER builds 10-year hourly load profiles based on new demand coming onto the system, spatial allocation of the system level forecasts and different weather scenarios.

Grid needs

To ensure we meet future energy demands, we compare existing capabilities against future demand.

Mitigation upgrades

Synergi Electric simulates voltage and power flow conditions under different scenarios to help identify the most effective upgrades for the system.

Informing action

We used LoadSEER and Synergi outputs to inform our IGP action plan and preferred plans for each island we serve, including Oahu, Hawaii and Maui County.

Download this presentation to the Stakeholder Technical Working Group (January 2023) for detailed information about how outputs informed our recommendations and decisions.

What are examples of when Hawaiian Electric used the LoadSEER and Synergi models?

LoadSEER

Hawaiian Electric has used LoadSEER to determine when and by how much is a local line getting overloaded. The modeling builds us a map to illustrate where the overloading is happening, and it pulls in spatial factors that show where the onsite solar, electric vehicles and energy efficiency are at the customer level. We then aggregate the data for each individual feeder line and transformer.

The image to the rightbelow shows how LoadSEER uses the system-level forecasts for DER, electric vehicles and energy to develop circuit level forecasts. This diagram shows the individual technology layers for a specific feeder and how they are all combined to develop an overall forecasted load profile for that feeder. This is a 24-hour snapshot, but LoadSEER creates this profile for every day of the year for multiple years.

The charts below are snapshots from the NWA Opportunity Evaluation Methodology. The top diagram is the annual peak load for a specific transformer, and it shows an overload in 2027. The bottom diagram shows hours of overload on the peak load day in 2027.

The following graph shows the magnitude and duration of the overload for an example grid need.

Synergi

We have used Synergi Electric to determine the capacity for new rooftop solar and battery additions on a specific feeder line. The snapshot below shows the physical equipment layout of example feeders. Synergi Electric uses information from our GIS and engineering drawings to build these models.

Specific load cases (example: daytime minimum) can be simulated on feeders to see how much electricity is being used across the entire circuit. This figure is colored by percent loading.

Synergi Electric also simulates voltage conditions. The diagram to the leftabove shows simulated voltage across a feeder for a specific load case (example: daytime minimum load). This diagram is colored by voltage. Green means that the voltages are well within the planning criteria limits. As the color gradient shifts towards red, the voltages increase where red is outside of the planning criteria and needs to be mitigated.

Connect with the modeling team

Diving Deeper into Data

Frequently asked questions

Yes, we use separate models for each island—for example, we run five separate RESOLVE models in total (Oahu, Moloka'i, Maui, Lana'i and Hawaii).

RESOLVE and other models consider a wide range of resources, which generally fall into these four categories:

Variable: Variable resources, such as wind and solar, have intermittent energy production, driven by weather patterns. They can often be curtailed if supply exceeds demand.
Firm: Firm resources typically combust a fuel to spin a turbine and then convert that mechanical energy into electricity. These resources are dispatchable and have output driven by fuel use and heat rate.
Energy-limited: Resources such as battery energy storage or pumped storage hydro typically charge from the grid at times of oversupply or low energy prices and discharge when demand is high.
Other: The model also considers energy efficiency and distributed energy resources (DER). Their adoption levels are driven by the energy demand forecast.

There are many factors that can change the cost of renewable resources over time. Technological improvement and innovations can drive down costs of resources like wind and solar. Resource costs can also change based on changes in state and federal tax credits and the expected cost of project financing.

We generally consider near-term as the next 1-5 years from now, mid-term as 5-10 years from now and long-term as 10-30 years from now. Modeling helps us look at each of these different timescales. We use these different timescales to help focus our planning and actions, understanding that we need to take immediate steps while ensuring we can invest in long-term upgrades and adapt to unforeseen changes.

A reliable and resilient grid takes a diverse portfolio of renewables that includes large and small-scale generation and firm generation. Firm generation fills the gaps from weather-dependent resources like solar and wind, as it is available any time for as long as needed. Examples include geothermal or conventional generators that use zero-carbon fuels.

Modeling for the Integrated Grid Plan helped us understand how much firm generation we would need by analyzing different portfolios of firm and variable generation to see how much of each resource type could meet reliability goals. While large amounts of low-cost variable renewables and storage helped to reduce the total cost of Hawaiian Electric's resource portfolio, a small amount of firm generation was still needed as a back stop to ensure the reliability of the system, particularly under adverse conditions where the weather was poor and existing generators were on outage.

We're often asked, “Can we reach a 100% clean energy future if everyone installs solar panels on their rooftops?” As much as we love rooftop solar, the short answer is no. We'll need a mix of small-scale and grid-scale (large generation facilities and transmission infrastructure) for a reliable, resilient, renewable-powered grid. There are few reasons why:

Solar isn't right for every rooftop. It's unlikely that we could install solar on every roof in Hawaii—not all roofs have the right conditions to generate solar energy, and some customers will be unable or unwilling to install panels.
A diverse grid is resilient. Generating electricity from a mix of resources and scales benefits our overall energy resilience and customer bills. Diversifying our energy generation to include small- and grid-scale projects keeps us from depending on any one source for all our electricity. This helps us bounce back faster from disasters and shields us from fluctuating costs of resources. For customers, this means a lower risk of outages and more stable bills.
Even if it could work, it wouldn't power the whole grid. Forecasts show that private generation alone would not supply enough electricity to meet future energy needs. To replace just one fossil fuel generator, we expect the islands will need new wind and solar projects with a collective footprint 29 times the size of Aloha Stadium. Grid-scale renewables are essential to supply this much energy.
Rooftop solar is more costly than large-scale renewable projects that can realize economies of scale. The IGP modeling showed a preference for maximizing large-scale resources to achieve lower overall costs of the plan. Only in the land-constrained case on Oahu where the development of new large-scale resources was limited did the models seek to expand rooftop solar.

Stay involved

Sign up for updates, share your thoughts and stay informed as we plan for a more resilient, affordable and sustainable energy system for our islands.

Email us at igp@hawaiianelectric.com or use our online form.

Planning for a Hawai'i Powered future

Hawaii's energy future

Hawaii's unique energy challenges and opportunities

Our data-driven approach

Inside the planning process

Integrated Grid Planning

What do you want to know?

Intro to energy modeling

Our approach to modeling

What is a modeling tool?

What can modeling do (and not do)?

What's a scenario?

The planning process: many models working together

Step 1: Forecasting future demand

Step 2: Building a reliable resource portfolio

Step 3: Planning for interconnection

Step 4: Bringing it all together

Meet the models

Get Started

Key question:

Key question:

Key question:

Key question:

Key question:

Key Questions

Future demand

What do MetrixND and MetrixLT do?

What inputs does MetrixND consider?

Weather

Where does the data come from?

Economic indicators

Where does the data come from?

Other indicators and flags

What inputs does MetrixLT consider?

Monthly energy

Energy patterns

What outputs do MetrixND and MetrixLT provide?

Underlying load forecast

Hourly forecasts

Informing action

Diving Deeper into Data

Frequently asked questions

Do you use separate models for each island?

What resources do the models consider?

What factors impact cost of renewable resources?

What do you mean when you say near-term, mid-term and long-term?

Why is firm generation part of the Integrated Grid Plan? Did that come from modeling?

Why is there not more rooftop solar in the Integrated Grid Plan? How did modeling inform the amount of rooftop solar?

Stay involved