Methodological Documentation for the “Generator REal-Time emissions Assignment” (GRETA) module

Singularity Energy, Inc. December, 2024

1. GRETA Overview

The Generator REal-Time emissions Assignment (GRETA) module is a patent-pending emissions calculation engine that estimates the direct (scope 1) stack emissions of GHGs from an electric generator in real time, based only on data about the net generation output (MW) of a unit. This module meets a need for more accurate accounting of power sector emissions on a (sub-)hourly basis.

The status-quo approach of estimating a generator’s emissions using a single annual-average emission factor has two critical limitations:

1. For generators that fuel switch, annual factors misrepresent the specific emission rate each hour depending on the fuel being burned.
2. Generator heat rates can change hour to hour based on specific factors.

GRETA uses generator-specific modeled heat rate curves derived from historical operating data from Singularity’s Open Grid Emissions (OGE) dataset. This approach represents the most accurate way to estimate hourly emissions in near real time when compared to hourly CO₂ emissions measurements recorded at each plant by continuous emissions monitoring systems (CEMS).

Modeled heat rate curves are more accurate at estimating hourly emissions output because the heat rate varies based on multiple factors:

• Operating level / capacity factor: Generators are generally more efficient at full design load than at partial load.
• Operating state: Generators are less efficient during startup than during operations, and combined cycle units may have different heat rates at the same output level.
• Ambient air temperature: Generators are generally more efficient when temperatures are colder, allowing for more complete fuel combustion.
• Season: Emissions control limits may be stricter during “ozone season” (May-September), affecting generator efficiency.
• Fuel switching: Different heat rates may occur at the same output level for each fuel that generators burn.

2. GRETA API Inputs and outputs

2.1 API Inputs

The GRETA API accepts data for a single “generation event” which consists of generation data for one or more generators in a single operating interval. Specifically, the GRETA API requires three inputs: the EIA plant ID, the EIA generator ID, and the net MW of generation in the interval.

In addition to these required inputs, GRETA accepts several optional inputs:

• A manually-specified energy source code for known fuel switching generators.
• An adjustment factor for jointly-owned units (JOUs).

2.1.1 Client-specific configuration

For each client, several configuration options include:

• The length of each generation event in minutes.
• Whether the generation data represents net or gross generation.
• Data output adjustments for combined heat and power (CHP) plants.
• IPCC Assessment Report and horizon values for converting GHG emissions.

2.1.2 Data Pre-Processing

Generator-level data is assigned a subplant ID by the GRETA API to work with aggregated subplant data.

2.2 API Outputs

GRETA returns:

• Estimated fuel type for the interval.
• Total mmBTU of fuel consumption in that interval.
• Mass of CO₂, CH₄, N₂O, CO₂-eq, and non-biogenic CO2 emissions.
• Emissions rate (in lb/MWh) of CO₂, CH₄, N₂O, and CO₂-eq.

3. Detailed Methodology

3.1 Operating State

3.1.1 Modeling operating states

Operating states affect a generator’s heat rate and/or fuel type. GRETA uses data from CEMS to identify operating states based on fractions of an hour each unit was operating.

Identified states include:

• off: No operating time.
• on: Full operating time.
• partial-on: Some operating time.
• startup: Transitioning from off to on.
• shutdown: Transitioning from on to off.
• cycling: Rapid transitions between on and off.

3.1.2 Operating State Assignment

The GRETA API can currently identify three different operating states:

• off
• startup
• on

3.2 Fuel Type

For each interval, GRETA identifies the fuel being consumed. Categories include:

• Single-fuel generators: Only utilize a single fuel type.
• Fuel-switching generators: Switch between multiple fuels.
• Co-firing generators: Use multiple fuels simultaneously.
• Different startup fuel generators: Use a different fuel during startup.

3.2.1 Fuel type modeling

Fuel attribute categories are determined based on reported data from EIA-860 and EIA-923.

3.2.2 Fuel type assignment

API assigns fuel types through a multi-step process based on defined criteria and manual input.

3.3 Fuel consumption

Fuel consumption is calculated using heat rates which vary based on factors like capacity factor and operating state.

3.3.1 Heat Rate Curve Modeling

Heat rate curves are calculated using historical generation and fuel data, with adjustments made to ensure data accuracy.

3.3.2 Fuel consumption calculation

Calculated fuel consumption is adjusted for various factors, including jointly owned units and emissions rate thresholds. CHP adjustments are also made to account for fuel used for electricity production versus heat.

3.4 Emissions output calculation

Emissions calculations are based on combustion emission factors multiplied by fuel consumption and disaggregated back to individual generators.

4. Future work

GRETA aims to address known limitations, such as improving the temporal granularity of the fuel-switching model and refining CEMS data preparation.

5. Glossary / Definitions

Acronyms:

• API: Application Programming Interface
• CAMPD: Clean Air Markets Program Division
• CEMS: continuous emissions monitoring system
• CHP: Combined Heat and Power
• EAF: electric allocation factor
• EIA: U.S. Energy Information Administration
• GHG: greenhouse gas
• GRETA: Generator REal-Time emissions Assignment module
• HRC: heat rate curve
• mmBTU: million British thermal units
• MW(h): Megawatt(-hour)