DOCSEARCHbyAlgolia
Submit Docs

Regional Energy Deployment System (ReEDS)

repository·main·Indexed 41 days ago

https://github.com/natlabrockies/reeds-2.0

A capacity planning and dispatch model for the U.S. electricity system used for long-term power sector research. Version 2024.0.0 includes the Hourlize module for processing hourly resource and load data into model inputs, as well as tools for managing source documentation and converting dsgrid parquet files to compatible CSV formats.

Tokens
81.6K
Snippets
125
Records
340
Agent score
52%

What's inside ReEDS

  1. Overview of Hourlize

    main

    Hourlize is a module designed to process hourly resource and load data into ReEDS inputs, providing temporal and spatial flexibility.

    It is managed via the run_hourlize.py script, which acts as an orchestrator. Depending on the command provided, it executes either resource.py or load.py. The script is designed to support both direct sequential execution and job submission to High-Performance Computing (HPC) environments for parallel processing.

  2. Combine regional ReEDS runs into a national run

    main

    The combine_runs.py script is used to postprocess regional ReEDS outputs (such as county-level runs that are too computationally intensive to run nationally) into a single combined national run.

    It performs the following tasks:

    • Creates combined outputs.
    • Generates a combined bokehpivot report.
    • Generates combined transmission and VRE (Variable Renewable Energy) site maps.

    The script uses a configuration file named combinefiles.csv to determine which outputs are supported and how they should be combined (using methods like concatenate, sum, or keep first).

  3. Overview of the ReEDS mathematical programming model

    main

    ReEDS (Regional Energy Deployment System) is a mathematical programming model used to simulate the evolution and operation of the electric power sector. It models the deployment and operation of generation, storage, transmission, and production technologies (including hydrogen and carbon capture).

    Key modeling characteristics include:

    • System Representation: The system is divided into model regions with specific supply and demand sources, connected by a transmission network (including existing and endogenous new capacity).
    • Temporal Scale: Designed for decadal-scale modeling. Users can select the time horizon and the intervening solve years.
    • Operational Resolution: Uses selected representative periods (seasonal and diurnal patterns) to characterize supply and demand within a year. Full hourly modeling is an option if computational resources allow.
    • Reliability: Grid reliability is addressed via stress periods and linkages to the Probabilistic Resource Adequacy Suite (PRAS) model.

  4. Activate and understand Power System Water Use formulation

    main

    ReEDS includes a high-fidelity power system water use formulation that characterizes the existing fleet and new generation investments by both cooling technology and water source type.

    Note: This feature is inactive by default due to computational complexity. When activated, it allows for detailed modeling of water demand, constraints, and costs.

    Key Components:

    • Cooling Technologies: Differentiates thermal generation by once-through, recirculating, pond, and dry(air)-cooled. Technologies without cooling systems can still be assigned water use rates (e.g., for PV cleaning or hydro evaporation).
    • Water Source Types: Models six distinct sources:
      1. fresh surface water that is currently appropriated
      2. unassigned/unappropriated fresh surface water
      3. fresh groundwater
      4. brackish or saline groundwater
      5. saline surface water
      6. wastewater treatment facility effluent
    • Water Constraints: Water access is constrained by availability and cost. Requirements are calculated based on the annual volume needed to operate at 100% capacity factor (MW * water use rate * 8760 hours).

  5. Configure Storage technology assumptions

    main

    ReEDS includes Pumped Storage Hydro (PSH), utility-scale batteries, and hydrogen (long-duration storage).

    All storage options are modeled with the following capabilities:

    • Load shifting (arbitrage)
    • Providing planning and operating reserves
    • Reducing curtailment of variable renewable energy (VRE)

    To prevent the model from choosing storage simply to avoid renewable curtailment (degeneracy), a minimum Variable Operation and Maintenance (VOM) cost of $0.01/MWh (in 2004$) is applied to all storage technologies.

  7. Quantify uncertainty with Monte Carlo Sampling (MCS)

    main

    ReEDS includes a Monte Carlo Sampling (MCS) module to quantify how uncertainty in key model inputs affects capacity expansion outcomes.

    Capabilities:

    • Regionalized sampling: Perform sampling at a regional level.
    • Input treatment: Supports both grouped and independent treatment of input switches.
    • Distribution types: Supports multiple distribution types for user-defined probability distributions.

    Workflow:

    1. Define probability distributions for inputs.
    2. The module draws values, writes corresponding input files, and executes a set of ReEDS runs.
    3. Aggregate results to evaluate ranges, quantiles, and other metrics.

    For specific instructions on running MCS, required switches, and defining custom distribution groups, refer to the Monte Carlo Sampling (MCS) section in the User Guide.

  8. Analyze technology economic viability using Net Value metrics

    main

    ReEDS uses "Net Value" metrics to evaluate the economic competitiveness of generating technologies by comparing their revenues (values) against their costs. These metrics are related to the concept of "reduced cost" in linear programming.

    Key Metrics:

    MetricConceptual Expression (typical units)Description
    Net value of energy (NVOE)(Value – Cost)/Energy ($/MWh)Unit profit normalized by energy production.
    Net value of capacity (NVOC)(Value – Cost)/Capacity ($/kW-yr)Unit profit normalized by installed capacity.
    System profitability(Value/Cost) (unitless)A unitless ratio (e.g., profitability index or ROI) to compare technologies with different denominators.

    Important Considerations:

    • Values vs. Costs: "Values" reflect the economic benefit of displacing or avoiding the cost of other (marginal) assets. "Costs" aggregate all expenses needed to build and operate the plant.
    • Full Model Integration: In these calculations, "values" include all model constraints related to the investment decision. For example, if increased wind generation increases ancillary service requirements, that increase is counted as a negative value stream for wind.

  10. Translate ReEDS solutions to Production Cost Models (PCM) via R2X

    main

    Because ReEDS is a capacity expansion model, it cannot explicitly represent detailed power system operations. The R2X translation framework is used to implement ReEDS capacity expansion solutions in Production Cost Models (PCMs) for detailed operational study.

    Supported PCMs:

    • Sienna: An open-source NLR modeling tool. Sienna\Ops supports system scheduling (unit commitment, economic dispatch, AGC, and nonlinear optimal power flow).
    • PLEXOS: A commercial PCM tool capable of representing individual generating units and transmission nodes at hourly or subhourly resolution.

    Linkage Characteristics:

    • Spatial Resolution: Preserves ReEDS spatial resolution (zone or county) by operating the PCM as a zonal model.
    • Transmission: Uses ReEDS transmission interface capacity; reactance and resistance are derived from ReEDS transmission assumptions.
    • Capacity Disaggregation: Converts ReEDS aggregate capacity into individual units in the PCM using characteristic unit sizes for each technology.
    • Parameter Handling: Uses ReEDS cost and performance parameters where possible. If parameters are missing or inconsistent, the translation uses average values across the WECC for equivalent technologies.

    Detailed information is available in the R2X repository.

  11. Explore optional hydropower upgrade pathways

    main

    Users can enable optional hydropower upgrade pathways to explore increasing fleet flexibility. These features are unavailable by default due to high uncertainty.

    Available Upgrades:

    • Nondispatchable to Dispatchable: Upgrading existing run-of-river plants to provide dispatchable services.
    • Dispatchable to Pump-back: Upgrading existing dispatchable plants to include pumping capabilities (similar to PSH, but can use natural water inflows in addition to the grid).

    Users can make these options available by specifying a user-specified capital cost.

  12. Understand limitations of high spatial resolution (county-level) runs

    main

    While county-level resolution enables granular outputs for power plant siting, transmission expansion, and emission impacts, users should be aware of several technical challenges:

    • Input Data Gaps: Higher resolution can expose gaps, such as incomplete coverage for sub-transmission voltages (below 100 kV).
    • Downscaling Smoothing: Downscaled load data may appear smoother and less accurate than actual local load profiles.
    • Linear Modeling Constraints: The linear programming approach may underestimate costs for very small builds (e.g., a 3 MW plant) because input costs are typically based on larger plants.
    • Transmission Costs: ReEDS transmission costs are based on 500 kV lines. Since many counties connect via <100 kV lines, county-level transmission costs may be higher in reality than the model estimates.
    • Transmission Flow Accuracy: ReEDS uses a 'pipe-and-bubble' approach. The sum of individual county Interface Transfer Limits (ITLs) may overstate coincident flow limits compared to larger interfaces.
    • False Precision: Model solutions at high resolution should be treated with caution; small shifts in capacity between neighboring counties may have negligible impacts on total system cost.