Emissions Forecasting Approaches: Production Estimates and Statistical Models

In the oil and gas industry, emissions forecasting is crucial for operators to anticipate future emissions profiles. Emissions forecasting integrates analyses of several factors including projected production levels, various emission sources, regulatory frameworks, technological advancements, and economic variables.

This guide explores the methodologies for developing emissions inventories, presents two emissions forecasting approaches, and discusses the economic factors impacting oil and gas emissions. By understanding these approaches, operators can better anticipate future emissions profiles, ensuring compliance with regulatory frameworks and supporting sustainability goals.

Emissions Inventory Methods
An emission inventory identifies and quantifies the amount of pollutants released from various sources into the atmosphere within a defined area over a specific period of time. Emission inventory is a critical component in forecasting because it establishes the baseline for quantifying current and future emissions levels.

Bottom-Up Method

A bottom-up approach estimates emissions after compiling a detailed inventory of equipment, quantifying the associated emissions, and then classifying them according to the emission source category. This method uses emission factors and engineering models to allow for a detailed breakdown of emissions by source category, subcategory, and individual facilities and equipment. Bottom-up is valuable for providing detailed insights into the sources and drivers of emissions, allowing for targeted mitigation strategies and policy interventions. However, this approach may require significant data collection and modelling efforts, and the inventory also needs to be supplemented with measurement of emissions that are not as predictable, like fugitive emissions.

Top-Down Method

On the other hand, top-down estimations can be performed at a regional scale where it starts with aggregated data at a higher level, such as national or regional data, and then allocates emissions down to specific sectors or sources. These are then distributed among various sectors or sources based on specific allocation factors or models. Top-down also considers measurement methodologies, including satellite, plane, drone, truck or instrument measurement. One such methodology combines aircraft-measured carbon dioxide (CO₂) and nitrous oxide (NO_X) emission ratios with satellite-derived NO_X emissions from an Ozone Monitoring Instrument (OMI) and a TROPOspheric Ozone Monitoring Instrument (TROPOMI) [7]. This method combining aircraft and satellite observations was applied to the Athabasca Oil Sands Region in Alberta, Canada where CO₂ emissions were found to be 65% higher than bottom–up estimates from 2005 to 2020 [7]. This study shows the potential for high discrepancy between top-down and bottom-up estimates, which is often due to fugitive and/or “super emitters” not considered in bottom-up estimates, as well as other sources which may not be considered, such as methane slip in combustion and flaring.

The top-down method can be more efficient at developing the emissions profile, especially for large-scale analysis. However, it lacks detailed sector-specific information and may not capture variations within sectors and is not typically as effective as the bottom-up approach at identifying the root cause and therefore the mitigation efforts needed to reduce emissions.

Our recommendation: To learn more about bottom-up and top-down estimates, please refer to our article. 'Which is Better – Bottom-up or Top-down Emissions Estimates?'

The following table presents a summary of the two discussed emission inventory methods showing a general description, data inputs, and some advantages and limitations:

As outlined in the table above, the top-down approach offers a large-scale view of emissions based on satellite data or national level statistics. However, it lacks detail at facility-level emissions with high uncertainty associated with the applied emission factors. On the contrary, bottom-up approach estimates emissions from individual sources resulting in more granular and detailed emission inventories. It allows the use of site-specific emission factors but is more resource-intensive and time-consuming to collect required data.

Both approaches can complement each other to provide a more comprehensive understanding of emissions combining wide coverage with detailed insights. Besides, they both are suitable for conducting national or global inventories. The choice of each method depends on the scope of the inventory, availability of data and goals of the analysis.

Forecasting Approaches

Once the emissions inventory is well defined, the next step is to proceed with the emission forecasting. The next section presents two approaches to make emission projections: the first uses production estimates to calculate emissions based on facility-type emissions factors while the second uses a statistical model for several variables to forecast emissions.

1. Production Estimate Forecasting

This approach aggregates detailed data at the source level, such as individual facilities, industrial processes, or specific activities, to estimate total emissions [6]. Detailed data on each emission source are collected, often through site-specific measurements, industry surveys, or process modelling. These individual source emissions are then summed to calculate total emissions for a region or country.

To cite an example, Waxman et al., 2020 published a study where they show how this methodology can be applied to 327 midstream and downstream facilities located in the Gulf and Southwest regions of the United States.

The authors apply a facility-level emissions factor by facility type, applying specific factors (emissions per unit of production) for different facility types. Where there was not any data available, (1) for facilities of a given type, average CO₂equivalent (CO₂e) emissions were estimated, and the average was then assigned to any facility of that type, and (2) for facilities where there was not any information, emission factors were used from the literature.

Once the base emissions profile is determined, the next step is to predict the facility capacity over the next years. Oil production forecasting involves a combination of geological, engineering, and economic analyses. Here are some of the most important factors:

Decline curve analysis (DCA) analyzes historical production data to identify trends and patterns. DCA involves fitting decline curves to historical production data to estimate future production decline rates.
Reservoir simulation is another tool employed to simulate fluid flow within the reservoir. These models consider reservoir properties, well locations, and production strategies to predict future production rates.
Economic factors may also influence decisions on drilling new wells or implementing enhanced oil recovery (EOR) techniques.

Once the crude oil and gas production curves are determined, these production volumes are then multiplied by the base emission factors to determine the emission projections year by year.

This study shows how this approach was successfully implemented to quantify the aggregate greenhouse gas (GHG) emissions impact of oil and gas in the US Gulf region by decomposing it by value chain segment, facility type, operational activity, and expected facility production.

To cite another example, combustion emission is a significant contributor to the total GHG emissions of the Oil Sands Extraction and Processing. The stationary combustion emissions can be predicted per year using bitumen production forecasting and annual fuel gas demand for steam generation.

2. Statistical Model Forecasting

This approach consists of an integrated global analysis of different variables that might impact the emissions projections. Usually, models are calibrated on historical data to make predictions. This approach involves making predictions about future emissions based on aggregated data at a higher level, such as national or regional levels.

Data Collection: Gather historical data on emissions at the desired level of aggregation. This data can come from various sources, including government reports, industry databases, and research studies. Historical trends for fugitive, flaring, combustion, and venting data can be used to come up with some predictive models.
Identify Key Variables: Determine the key variables that influence emissions at the aggregated level. These may include factors such as economic activity, population growth, energy consumption patterns, technological advancements, and government policies.
Develop a Forecasting Model: Build a statistical or mathematical model that relates the key variables to emissions. This model should capture the relationships between these variables and allow for the prediction of future emissions based on changes in the input factors.
Validate the Model: Validate the forecasting model using historical data to ensure that it accurately captures past trends and patterns in emissions. This may involve testing the model against a portion of the historical data that was not used in the model development phase.
Make Forecasts: Use the validated model to generate forecasts of future emissions based on different scenarios of economic growth, policy interventions, technological advancements, etc. These forecasts can provide insights into the potential trajectory of emissions and help policymakers and stakeholders make informed decisions.

Fugitive emissions are typically predicted using statistical models, empirical factors or correlations based on historical data since it is quite difficult to predict this emission source is challenging due to its dependency on multiple variables. In the absence of data, it is common practice to keep this value constant over time.

Overview of forecasting approaches

In summary, the following table shows the methodology and some steps used for each forecasting approach:

The Impact of Economic Factors on Emissions Forecasting

Economic factors must be considered while making an emission forecasting analysis since they strongly influence the future behaviour of emission patterns. The most important factors are listed below:

Crude Oil and Natural Gas Markets

Global crude oil and natural gas prices clearly have a significant impact on the local production of crude oil. Low

Carbon Technologies

New technologies will be put on the market in upcoming years including hydrogen, biofuels, and carbon capture, utilization, and storage (CCUS) among others. Their application will be dictated by technology maturity, costs, policy preferences and market conditions [3]. Since they will be big contributors to the emission reductions, it is paramount to consider their impact when making emission projections.

Carbon Policies

The strength of domestic climate policies is a key driver of emission reduction that can force companies to implement more aggressive strategies and in consequence change their trajectory [3]. It includes regulations, laws, and programs put in place by the government to reduce GHG emissions. International climate policies can also be important if domestic producers are exporting oil and gas to international markets.

Other Sources of Electricity Generation

Solar, wind, nuclear, natural gas with CCUS, and other electricity generation sources will impact emissions projections [3].

Given the uncertainty in these factors, it can be difficult to predict the trends. It can be useful to incorporate uncertainty analyses or consider different case studies to evaluate the impact of these factors on future trends.

Key Economic Factors Summary

Figure adapted from report “Canada’s Energy Future 2023 – Canada Energy Regulator”

Conclusions

In summary, while bottom-up approaches offer detailed insights into specific emission sources, top-down approaches provide a broader perspective on overall emissions. Choosing the appropriate approach depends on factors such as the scale of analysis, available data, and the objectives of emission assessment or management.
Based on the literature review, this article discussed two approaches for making projections of GHG emissions where each methodology and key variables are outlined with some advantages and limitations.
Production estimate forecasting employs a simpler methodology that requires just two components: the emission baseline and production curves. In contrast, statistical model forecasting involves collecting historical data, conducting a multivariate statistical analysis, and developing a predictive model. The choice of forecasting approach depends on factors such as data availability, data quality, process characteristics, cost-benefit analysis, and the required level of detail."
Including technology, economic, and policy factors in emission forecasting is imperative to reflect key analytical, regional, economic, and sectoral considerations.

Are you committed to reducing emissions in the oil and gas industry, improving sustainability, and advancing innovative forecasting methods?

At Process Ecology, we're equally passionate about helping you achieve these objectives. Our expertise is dedicated to providing customized solutions for your emission forecasting and sustainability goals.

Take the next step toward a cleaner, more efficient future. Contact us to discover how we can support your mission for a greener and more responsible industry. Together, we can drive real change.

Get in touch with us:

Contact us via email at info@processecology.com
Follow us on LinkedIn for updates: Process Ecology Inc.
Subscribe to our newsletter today!

References

Greenhouse Gas Emissions Forecasting: Learning from International Best Practices, July 2008. Appendix C: Discussion of Top-Down, Bottom-up, and Hybrid Approached to Energy-Economy Modelling.
Waxman Andrew, Khomaini Achmad, Leibowicz Benjamin, Olmstead Sheila. (2020). Emissions in the stream: estimating the greenhouse gas Impacts of an oil and gas boom. Environmental Research Letters 15.
Canada’s Energy Future 2023: Energy Supply and Demand Projections to 2050. Canada Energy Regulator, 2023.
Greenhouse gas emissions projections. Canadian environmental sustainability indicators. January 2023.
Dowlatabadi, H., D.R. Boyd, J. MacDonald (2004). Model, Model on the Screen, What’s the Cost of Going Green?. Resources for the Future, p. 10
The Greenhouse Gas Protocol. A Corporate Accounting and Reporting Standard (revised edition)World Resources Institute and World Business Council for Sustainable Development, 2004.
Sumi N Wren, Chris A McLinden, Debora Griffin, Shao-Meng Li, Stewart G Cober, Andrea Darlington, Katherine Hayden, Cristian Mihele, Richard L Mittermeier, Michael J Wheeler, Mengistu Wolde, John Liggio (2023). Aircraft and satellite observations reveal historical gap between top:–down and bottom–up CO2 emissions from Canadian oil sands.

About the authors: