Discover
Discover
Project Types
Discover offers two types of projects, operational and synthetic. All projects have characteristics (i.e. latitude, longitude, capacity (MW), turbine model, panel model, etc.) that are used as inputs for our models.
Operational
Discover offers data on over 2,000 operational projects. Operational projects consist of utility level projects with capacity of 5MW or more only. Behind the meter projects are not included. Project characteristics are primarily sourced from the U.S. Energy Information Administration (EIA).
Operational wind projects include onshore wind farms in the contiguous United States.
Operational solar projects include solar photovoltaic plants in the contiguous United States.
Synthetic
Discover offers data on over 12,000 synthetic projects. Synthetic projects consist of hypothetical projects located at the center of each REzone. REzones are geographic areas delineated by REsurety that correspond to areas of similar wind or solar project value. There is only one type of synthetic wind project, however, there are two types of synthetic solar projects to account for tracking and fixed panels. Default project characteristics are defined by REsurety.
| Synthetic Wind Projects | Synthetic Tracking Solar Projects | Synthetic Fixed Solar Projects | |
|---|---|---|---|
| turbine count | 40 | ||
| turbine size (MW) | 2.5 | ||
| hub height (m) | 80 | ||
| capacity AC (MW) | 100 | 100 | |
| capacity DC (MW) | 130 | 130 | |
| tilt | 0° | 20° | |
| panel type | mono-crystalline silicon | mono-crystalline silicon | |
| mounting | single-axis (east-west), max 60° tilt | fixed | |
| orientation | NA | south facing |
Generation
Discover offers two types of generation data, historical and forecasted. Historical data can be either observed (solid line) or modeled (dashed line). Historical data is singular as it is based off a measured history. Forcasted data, however, is displayed as a distribution (line with confidence band) to account for potential variability. Both types of data are shown with monthly resolution below, but hourly resolution is also available.
Historical
There are 15 years of historical generation data made available in Discover. For historical generation data, observed data is displayed when available. When observed data is not available, modeled data is shown. Instances where observed data is not available and modeled data is shown:
- Synthetic Projects
- all historic data
- Operational Projects
- historic data
- prior to the project’s commercial operation date
- for recent history that falls into a reporting lag
- historic data
Observed Generation
Observed generation is also referred to as actual generation or metered generation. This data is only available for operational projects and is not available for synthetic projects.
This data is sourced from ERCOT and EIA. Inside ERCOT, observed generation is typically reported at the hourly level with a 60 day lag. Outside of ERCOT, monthly generation data from the EIA data is used and the reporting lag varies between 3 to 12 months depending on the project.
Modeled Generation
Modeled generation refers to generation data derived from weather data (i.e. wind speed, temperature, irradiance, etc.). Modeled generation consists of different data sources and methodologies depending on a project’s fuel type and whether it is operational or synthetic. Reasons being, (1) weather data sources have different advantages depending on whether a project is operational or synthetic, (2) modeled data can be trained to observed generation for operational projects in order to improve the accuracy of the model output–but this is not available for synthetic projects and (3) the model methodology will use project-specific characteristics if operational, whereas all synthetic projects use the same default characteristics.
Below is a summary of modeled generation components and methodology by fuel type and project type with more detailed descriptions following.
| Wind Operational Projects | Wind Synthetic Projects | Solar Operational Projects | Solar Synthetic Projects | |
|---|---|---|---|---|
| wind speed data source | MERRA-2 | HRRR | MERRA-2 | MERRA-2 |
| temperature data source | MERRA-2 | HRRR | MERRA-2 | MERRA-2 |
| irradiance data source | Also Energy | Also Energy | ||
| project characteristics | project-specific characteristics | default characteristics | project-specific characteristics | default characteristics |
| losses applied | wake, availability, electrical, performance | wake, availability, electrical, performance | electrical, performance | electrical, performance |
| training | If the project is located in ERCOT, the modeled output will be trained using ERCOT observed data otherwise it will be trained using EIA observed data. | If the project is located in ERCOT, the modeled output will be trained using ERCOT observed data. |
Wind Projects
- Synthetic: HRRR data is preferred here because it has a much higher spatial resolution, 3km, than MERRA-2, 50 km, and so is more well-suited to capturing regional changes in wind speed in areas where no observed data is available for training.
- For synthetic projects, all losses (wake, electrical, performance, and availability), as well as power curves, use a default value within the model (e.g., a 5% loss factor is applied to every synthetic project to account for availability and electrical losses).
- Operational: MERRA-2 data is preferred here because it has a much longer record length than HRRR and also has relatively few model changes. MERRA-2 is trained to observed hourly and monthly generation. The training method for wind projects includes seasonal and diurnal corrections to minimize model bias at the hourly level. Any periods of high availability losses are screened out to prevent them from biasing the model; these losses are incorporated later into the final energy output. Data from REsurety’s proprietary meteorological tower and project generation databases are also used to inform training parameters. Trained data is then used to predict the long term wind resource at the site. The model additionally uses the Park wake model to account for wake effects at each site.
- For operating projects where hourly observed energy is available, electrical and availability losses are inferred from reported generation, a default performance-driven loss of 5% is assumed based on historical performance across a variety of turbine types, and wake losses are calculated by the Park model.
- For operating projects where only monthly observed energy is available, a flat ~9% loss is used to account for electrical, performance, and small-scale availability losses, while large-scale availability losses are inferred from the monthly generation data.
- Synthetic: HRRR data is preferred here because it has a much higher spatial resolution, 3km, than MERRA-2, 50 km, and so is more well-suited to capturing regional changes in wind speed in areas where no observed data is available for training.
Solar Projects
- Synthetic: To model solar generation data for a particular project, project specifications including project location, DC and AC capacity, and panel and tracker type are received from EIA and ERCOT. Modeled solar irradiance data is then obtained from Also Energy, which provides three components of irradiance data — global horizontal, direct normal, and diffuse horizontal irradiance. Lastly, any additional meteorological data including temperature and wind speeds that affect the panel output are obtained from MERRA-2. Also Energy irradiance data is passed through PVLIB (an open-source PV simulation model). PVLIB also requires a DC and AC capacity, which is set at a DC to AC ratio of 1.3 if the information cannot be found. Additionally, all solar projects with tracking are modeled as single-axis trackers with a maximum tilt angle of 60 degrees and backtracking, unless otherwise specified.
- Operational: Methodology here is the same as Synthetic with the addition of training when projects are located in ERCOT and observed energy data is available.
Forecasted
There are 20 years of forecasted generation data made available in Discover. Each forecasted year of generation data is created from 40 underlying weather scenarios. These 40 weather scenarios reflect modeled weather data (i.e., wind speed, temperature, irradiance, etc) from the most recent 40 years of history. The weather from each of the 40 years is translated to generation using a similar methodology as historical modeled generation.
Once weather data has been translated to generation data, an average is calculated and displayed. To capture the variabilities of outcomes however, there is a distribution of outcomes returned as well, including P1, P50, P90 and P99 values. Every forecasted year is comprised of these 40 weather scenarios, so every forecasted year will have roughly the same generation values–there will be some small differences however due to peak/off-peak hour difference from calendar year to calendar year.
Price
Discover offers two types of price data, historical and forecasted. Historical data can be either observed (solid line) or modeled (dashed line). Historical data is singular as it is based off a measured history. Forecasted data, however, is displayed as a distribution (line with confidence band) to account for potential variability. Both types of data are shown with monthly resolution below, but hourly resolution is also available.
Historical and forecasted price data can be displayed in numerous configurations. The types of configuration available include:
| Historical Operational Projects | Historical Synthetic Projects | Forecasted Operational Projects | Forecasted Synthetic Projects | |
|---|---|---|---|---|
| Type: As-gen | x | x | x | x |
| Type: 12×2 | x | x | ||
| Type: 12×24 | x | x | ||
| Type: ATC | x | x | x | x |
| Market: RT | x | x | x | x |
| Market: DA | x | x | x | x |
| Settlement Point: Node | x | |||
| Settlement Point: Hub | x | x | x | x |
Historical
There are 15 years of historical price data made available in Discover. This data is primarily sourced from Yes Energy and is available for all ISOs.
For historical price data, observed data is displayed when available. When observed data is not available, modeled data is shown. ATC observed data is always available for any point in the past. However, As-gen, 12×2 and 12×24 prices are dependent on corresponding generation data and therefore observed data is not always available. If only modeled generation is available, then price data will be considered modeled as well.
Instances where observed data is not available and modeled data is shown:
- Synthetic Projects
- all historic data (when viewing As-gen, 12×2 or 12×24 prices)
- Operational Projects
- historic data
- prior to the project’s commercial operation date (when viewing As-gen, 12×2 or 12×24 prices)
- for recent history that falls into a reporting lag (when viewing As-gen, 12×2 or 12×24 prices)
- historic data
Market Forecast
Forward Valuation estimates provided are based on the recent historical observed and/or modeled relationship between the value of as-generated power and the value of fixed volume power (such difference is referred to as the “Shape” of generation). Shape is then projected onto market based forward curves for fixed volume power as of the last business day of the previous month. The Forward Valuation estimates provided do not reflect REsurety’s proprietary view of if and how the future Shape relationship will differ from the recent past. Note: The “12×2” weighted price used in forward valuation is based on the ISO-defined peak/offpeak (rather than the 12×2 day/night definition used elsewhere in Discover). Market Forecasts are in nominal terms.
Fundamentals Forecasting
Our fundamentals forecast produces a 20-year projection of monthly price datafor select hubs throughout the following ISOs in the United States: CAISO, ERCOT, ISO-NE, MISO, NYISO, PJM and SPP.
As part of our methodology, our Power Markets team builds grid models for all supported regions, while setting forward looking assumptions that pertain to significant impact drivers that will affect energy prices such as (but not limited to): demand growth, transmission buildout, battery deployment, policy changes, and geopolitical events. These assumptions are researched, tested, and validated throughout the year which means that they are updated bi-annually in May and November, after which they are deployed throughout CleanSight. Fundamentals forecasts are in nominal terms.
Due to the volatility of gas prices and their direct effect on power grids, we offer three scenarios that users can switch between to see their impacts on the performance of their portfolio:
| Scenario | Description |
|---|---|
| Baseline | Our primary view of the evolution of power markets, including generator additions and retirements, weather variability, electrification of load, etc. |
| High Gas Prices | We rerun the Baseline scenario with the forward gas prices materially increased to understand how sensitive the market is to higher gas prices, holding all else constant |
| Low Gas Prices | We rerun the Baseline scenario with the forward gas prices materially decreased to understand how sensitive the market is to lower gas prices, holding all else constant |
Each scenario is designed to be plausible and are grounded in authoritative sources of data that often originate from ISOs/RTOs.
Type
Around the Clock (ATC)
Around the clock price is an unweighted average.
As-generated (As-gen)
As-generated price is a weighted average, weighted by hourly generation at the project selected
12×2
12×2 price is a weighted average, weighted by peak/off-peak generation at the project selected. To calculate 12×2 price, the following formula is used:
12×24
12×24 price is a weighted average, weighted by monthly hour generation at the project selected.
Market
Day Ahead (DA)
The Day Ahead market happens the day before the delivery date, where electricity generators bid tomorrow’s power amount and prices to the ISO. The ISO decides which generators it needs, sets the price, and ensures that the grid can support this particular layout and load.
Real Time (RT)
The Real Time market is happening right now, typically receiving data in ~5 minute intervals. RT prices fluctuate much more than the DA prices, resulting in the likelihood of seeing large price spikes and dips. This is because the RT market responds to unpredicted events like outages or inaccurate forecasts. If a generator does not meet their power production bid from the day before, they will need to buy power at the price reflected in the RT market.
Settlement Point
Node
The node is typically where a project is actually located and is meant to represent the point of interconnection (POC). Locational marginal prices (LMPs) are defined at the nodal level.
Hub
The hub is a collection of nodes intended to represent an un-congested price for electric energy, facilitate electric energy trading, and enhance transparency and liquidity in the marketplace. Since hub prices are an average of nodal prices, they are generally less volatile.
All projects are mapped to a default hub. For most metrics, there is also a “Custom Hub” option where this default can be overwritten.
Compare
Comparisons are defined as the difference between two prices. Discover offers comparisons along the three main parameters of price (1) Type (2) Market and (3) Settlement Point. There are a few predefined differences as well as a custom option.
The availability of these comparisons varies by Project Type and Analysis Type:
| Historical Operational Projects | Historical Synthetic Projects | Forecasted Operational Projects | Forecasted Synthetic Projects | |
|---|---|---|---|---|
| Shape | x | x | x | x |
| DART | x | x | x | x |
| Basis | x | |||
| Custom | x | x | x | x |
Comparison calculations make certain assumptions (i.e. which market to use or which settlement point to use). Users can change these assumptions by defining a custom comparison.
ISO Definition
ISOs as shown in Discover are used for defining ISOs in the API. In Discover the regions covered by NWPP, WECC, and SERC are labelled as ISOs to conveniently compare regions.
ISO Peak and Off-Peak Definitions
ISOs define on-peak and off-peak periods differently, this table summarizes on-peak and off-peak definitions by ISO:
| ISO | Time Zone | On-Peak Periods | Off-Peak Periods |
|---|---|---|---|
| ISONE | EPT | 8:00 to 23:00, Monday-Friday | 24:00 to 7:00, Monday-Friday; All-day Saturday and Sunday |
| NYISO | EPT | 8:00 to 23:00, Monday-Friday | 24:00 to 7:00, Monday-Friday; All-day Saturday and Sunday |
| PJM | EPT | 8:00 to 23:00, Monday-Friday | 24:00 to 7:00, Monday-Friday; All-day Saturday and Sunday |
| MISO | EST | 7:00 to 22:00, Monday-Friday | 23:00 to 6:00, Monday-Friday; All-day Saturday and Sunday |
| ERCOT | CPT | 7:00 to 22:00, Every Day | 23:00 to 6:00, Every day |
| SPP | CPT | 7:00 to 22:00, Monday-Friday | 23:00 to 6:00, Monday-Friday; All-day Saturday and Sunday |
| CAISO | PPT | 7:00 to 22:00, Monday-Saturday | 23:00 to 6:00, Monday-Saturday; All-day Sunday |
Some ISO’s consider NERC holidays to be entirely off-peak. NERC holidays are New Year’s Day, Memorial Day, Independence Day, Labor Day, Thanksgiving Day, and Christmas Day. These are off-peak periods for most ISOs; ERCOT treats holidays as weekend days with a weekend on-peak and off-peak period.
Third Party Data Sources
- EIA
- Energy Information Administration, US government agency
- ERCOT
- Electric Reliability Council of Texas, US Corporation
- HRRR
- High Resolution Rapid Refresh
- Provided by NOAA (National Oceanic & Atmospheric Administration), US government agency
- Spatial resolution 3km
- MERRA-2
- Modern Era Retrospective Analysis for Research and Applications
- Provided by NASA (National Aeronautics and Space Administration), US government agency
- Spatial resolution 50km
- ERA5
- ECMWF RE-Analysis (European Centre for Medium-Range Weather Forecasts), Intergovernmental
- Spatial resolution ~30km
- Also Energy
- Yes Energy
- OCTGH
- Over-the-counter commodities, Global Holdings
Third Party Data Source Uses
| Data Source | Data Type | Data | Wind Projects | Solar Projects |
|---|---|---|---|---|
| EIA | project | project characteristics | x | |
| EIA | energy | plant generation | x | |
| ERCOT | energy | plant generation | x | x |
| HRRR | weather | wind speed | x | x |
| HRRR | weather | temperature | x | x |
| MERRA-2 | weather | wind speed | x | x |
| MERRA-2 | weather | temperature | x | x |
| ERA5 | weather | irradiance | x | |
| Also Energy | weather | irradiance | x | |
| Yes Energy | price | $/MWh | x | x |
| OTCGH | price | $/MWh | x | x |