Canada's Emission Trends 2014: annex 4


Annex 4: Methodology for Development of Emissions Scenarios

The scenarios developed to support Environment Canada’s GHG emissions projections derive from a series of plausible assumptions regarding, among others, population and economic growth, prices, demand and supply of energy, and the evolution of energy efficiency technologies. The projections also assume no further government actions to address GHG emissions beyond those already in place as of May 2014.

The emissions projections presented in this report cannot be viewed as a precise estimate of emissions at a future date. Rather, this report presents a simple projection of the current structure and policy context into the future, without attempting to account for the inevitable but as yet unknown changes that will occur in government policy, energy supply, demand and technology, or domestic and international economic and political events.

The emissions projections have been developed in line with generally recognized best practices. They incorporate IPCC standards for estimating GHG emissions across different fuels and processes, rely on outside expert views and the most up-to-date data available for key drivers such as economic growth, energy prices, and energy demand and supply, and apply an internationally recognized energy and macroeconomic modeling framework in the estimation of emissions and economic interactions. Projections and underlying assumptions in this year’s Canada’s Emission Trends report have been through a consultation process with key stakeholders, while a more detailed examination of methodological approaches is performed periodically by leading external experts on economic modeling and GHG emissions projections.

The approach to developing Canada’s GHG emissions projections involves two main features:

Up-to-date Data and Key Assumptions

Each year, Environment Canada updates its models using the most recent data available from Statistics Canada’s Report on Energy Supply and Demand in Canada and Canada’s NIR. For these projections, the most recent historical data available were for 2012. Environment Canada’s projections and historical data in the NIR are aligned, based on economic sector definitions.

In addition to the most recent historical information, the projections are based on expert-derived expectations of key drivers (e.g., world oil price). Projections are based on the latest energy and economic data, with key modeling assumptions aligned with Government of Canada views:

Even with the benefit of external expert assumptions, there is considerable uncertainty surrounding energy price and economic growth assumptions, particularly over the medium- to long-term. As such, a range of emissions is presented representing a series of sensitivity analyses. These cases were based on high and low GDP growth as well as high and low oil prices and production levels.

The Without Measures Scenario

In 2014, the “without measures” or no programs scenario has been fully remodeled to take into account all of the structural changes occurring within the model and also to update assumptions about key drivers. Moreover, a refined methodology is being used to ensure that the drivers are being reflected in accordance with the description of the scenario.

The “without measures” scenario is constructed by beginning the model’s forecasting mode in 2006, configured to exclude any government policies implemented after 2005. Historical macroeconomic data are used between 2006 and 2012, and wholesale energy prices throughout the entire projection period are kept the same as those used in the reference scenario. Changes in electricity-generation-sector energy use resulting from non-policy-driven factors, including nuclear refurbishment or historical weather-related fluctuations in hydroelectric dam capacities, are reflected in the “without measures” scenario. Exogenous agriculture emissions from livestock and crop production are maintained at reference scenario levels throughout the entire projection period. All other sectors belonging to transportation, oil and gas, buildings, emissions-intensive and trade-exposed industries, and waste and others are derived from the highest of either the 2005 or 2012 emissions intensity, subject to a limit of no greater than 30% more than the value in 2012.

Logarithmic-mean Divisia Approach Used in Box Entitled "Decomposition of Canada’s Energy-related GHG Emissions"

Since the late 1970s, researchers have been using decomposition analysis to help explain changes in energy and greenhouse gas emissions. The analysis is based on the identity presented in below:Footnote 25

E = A times (E over A) = sum from i of (A times (a sub i over A) times (e sub i over a sub i) = the sum from i of (A times S sub i times I)

Where:

E = Emissions

A = Total Activity

ai = Activity of sub-sector i

ei = Emissions of sub-sector i

Si = Structural Effect i

Ii = Intensity Effect i

The overall Activity Effect (A) is generally a measure of economic growth. The measures of activity vary from sector to sector. For example, for the residential sector, activity is measured in terms of floor space, while in the industrial sector, it is generally measured in terms of gross output. The Structural Effect (Si) measures the change in the composition of the economy, such as the shift from goods-producing to service-producing industries. The Intensity Effect (Ii) measures changes in emissions intensity at the subsector level. This variable will capture the impact of changes in fuel mix as well as energy efficiency improvements.

There are a number of methodologies that can be used to decompose the identity, each with pros and cons. The methodology that has been used here is the logarithmic mean Divisia index method that is used by Natural Resources Canada as well as other international experts. The main benefits of this approach as described by Ang (2004)Footnote 26 are:

Using the LMDI method, the decomposition of the change in GHG emissions over time or from two cases can be defined by the following equations:

Equation set showing that the decomposition of the change of GHG emissions over time is the summation of the Activity Effect, the Structure Effect and the Pure Intensity Effect

Energy, Emissions and Economy Model for Canada

The projections presented in this annex were generated from Environment Canada’s Energy, Emissions and Economy Model for Canada (E3MC).

E3MC has two components: Energy 2020, which incorporates Canada’s energy supply and demand structure, and the in-house macroeconomic model of the Canadian economy.

Energy 2020 is an integrated, multi-region, multi-sector North American model that simulates the supply of, price of, and demand for all fuels. The model can determine energy output and prices for each sector, both in regulated and unregulated markets. It simulates how such factors as energy prices and government measures affect the choices that consumers and businesses make when they buy and use energy. The model’s outputs include changes in energy use, energy prices, GHG emissions, investment costs and possible cost savings from measures, in order to identify the direct effects stemming from GHG reduction measures. The resulting savings and investments from Energy 2020 are then used as inputs into the macroeconomic model.

The in-house macroeconomic model is used to examine consumption, investment, production and trade decisions in the whole economy. It captures the interaction among industries, as well as the implications for changes in producer prices, relative final prices and income. It also factors in government fiscal balances, monetary flows, and interest and exchange rates. More specifically, the macroeconomic model incorporates 133 industries at a provincial and territorial level. It also has an international component to account for exports and imports, covering about 100 commodities. The macroeconomic model projects the direct impacts on the economy's final demand, output, employment, price formation and sectoral income that result from various policy choices. These, in turn, permit an estimation of the effect of climate change policy and related impacts on the national economy.

E3MC develops projections using a market-based approach to energy analysis. For each fuel and consuming sector, the model balances energy supply and demand, accounting for economic competition among the various energy sources. This ensures consistent results among the sectors and regions. The model can be operated in a forecasting mode or an analytical mode. In forecasting mode, the model generates an annual energy and emissions outlook to 2050. In analytical mode, it assesses broad policy options, specific programs or regulations, new technologies, or other assumptions.

The model’s primary outputs are tables showing energy consumption, production and prices by fuel type, year and region. The model also identifies many of the key macroeconomic indicators (e.g., GDP or unemployment) and produces a coherent set of all GHG emissions (such as CO2, CH3 and N2O) by sector and by province.

Figure A.5 shows the general structure of E3MC. The component modules of E3MC represent the individual supply, demand and conversion sectors of domestic energy markets, and also include the macroeconomic module. In general, the modules interact through values representing the prices of the energy delivered to the consuming sectors and the quantities of end-use energy consumption. 

Figure A.5: Energy, Emissions and Economy Model for Canada

Figure A5 (see description below)
Text description of Figure A.5

The diagram provides a visual description of the Energy, Emissions and Economy Model for Canada. The diagram explains how the Macroeconomic Model feeds the following data into the Energy 2020 model: Gross Output by Industry Jurisdictions, Personal Income, Inflation, Tax Rates, and Exchange Rates. Energy 2020 then balances demand with supply incorporating prices. The output from Energy 2020 is then fed into the macroeconomic model creating a circular path. Output includes: (i) changes to investments in energy using equipment and structures by sector and industry; (ii) changes to energy intensity (energy input per unit of output) by sector, by industry and fuel; (iii) changes in energy prices.

To develop this projection of energy use and related emissions, it was necessary to provide a view of the Canadian economy to 2020. The level and composition of energy supply and demand, and the resulting GHG emissions, are determined based on many assumptions that influence the overall size and growth rate of the economy.

Treatment of Interaction Effects

Estimates of the net impact of government measures incorporated into the modeling scenarios need to take into account major interaction and behavioural affects. The analytical approach permitted by E3MC addresses these key modeling challenges:

Additionality

This issue relates to the question of what would have happened without the initiative in question. Problems of additionality arise when the stated emissions reductions do not reflect the difference in emissions between equivalent scenarios with and without the initiative in question. This will be the case if stated emissions reductions from an initiative have already been included in the reference case: emissions reductions will effectively be double-counted in the absence of appropriate adjustments. The E3MC model controls for additionality by basing its structure on incremental or marginal decision making. The E3MC model assumes a specific energy efficiency or emission intensity profile at the sector and end-use point (e.g., space heating, lighting or auxiliary power). Under the E3MC modeling philosophy, if the initiative in question were to increase the efficiency of a furnace, for example, only the efficiency of a new furnace would be changed. The efficiency of older furnaces would not change unless those furnaces are retired and replaced with higher-efficiency ones. As such, any change in the model is incremental to what is reflected in the business-as-usual assumptions.

Free Ridership

A related problem, free ridership, arises when stated reductions include the results of behaviour that would occur regardless of the policy. This can occur when subsidies are paid to all purchasers of an item (e.g., a high-efficiency furnace), regardless of whether they purchased the item because of the subsidy. Those who would have purchased the product regardless are termed free riders. In the E3MC model, the behaviour of free riders has already been accounted for in the reference case. Thus, their emissions are not counted toward the impact of the policy. Instead, the E3MC model counts only the incremental take-up of the emissions-reducing technology.

The Rebound Effect

This describes the increased use of a more efficient product resulting from the implied decrease in the price of its use. For example, a more efficient car is cheaper to drive and so people may drive more. Emissions reductions will generally be overestimated by between 5% and 20% unless estimates account for increased consumption because of the rebound effect. Within the model, we have mechanisms for fuel choice, process efficiency, device efficiency, short-term budget constraints and cogeneration, which all react to changes in energy and emissions costs in different time frames.Footnote 27 All of these structures work to simulate the rebound effect. In the example above, the impact of extra kilometres that may be driven as a result of improved fuel efficiency is automatically netted out of the associated emissions-reduction estimates.

Policy Interaction Effects

This describes impacts on the overall effectiveness of Canada’s emissions-reduction measures when they interact with each other. A policy package containing more than one measure or policy would ideally take into account these impacts in order to understand the true contribution that the policy package is making (in this case, to emission reductions).

E3MC is a comprehensive and integrated model focusing on the interactions between sectors and policies. In the demand sectors, the fuel choice, process efficiency, device efficiency and level of self-generation are all integrally combined in a consistent manner. The model includes detailed equations to ensure that all the interactions between these structures are simulated with no loss of energy or efficiency. For example, the electric generation sector responds to the demand for electricity from the energy demand sectors, meaning that any policy to reduce electricity demand in the consumer sectors will impact the electricity generation sector. The model accounts for emissions in the electricity generation sector as well as for emissions in the consumer demand sectors. As the electricity sector reduces its emissions intensity, policies designed to reduce electricity demand in the consumer sectors will cause less of an emissions reduction. The natural gas and oil supply sectors similarly respond to the demands from the consumer sectors, including the demands for refined petroleum products for transportation. The model also simulates the export of products by supply sectors.

Taken as a whole, the E3MC model provides a detailed representation of technologies that produce goods and services throughout the economy, and can simulate, in a realistic way, capital stock turnover and choices among technologies. The model also includes a representation of equilibrium feedbacks, such that supply and demand for goods and services adjust to reflect policy. Given its comprehensiveness, E3MC covers all the GHG emissions sources, including those unrelated to energy use.

Simulation of Capital Stock Turnover

As a technology vintage model, E3MC tracks the evolution of capital stocks over time through retirements, retrofits and new purchases, in which consumers and businesses make sequential acquisitions with limited foresight about the future. This is particularly important for understanding the implications of alternative time paths for emissions reductions. 

The model calculates energy costs (and emissions) for each energy service in the economy, such as heated commercial floor space or person-kilometres traveled. In each period, capital stocks are retired according to an age-dependent function (although the retrofitting of unretired stocks is possible, if warranted by changing economic conditions). Demand for new stocks grows or declines depending on the initial exogenous forecast of economic output (i.e., a forecast that is external to the model and not explained by it) and the subsequent interplay of energy supply-demand with the macroeconomic module. A model simulation iterates between energy supply-demand and the macroeconomic module until there is a convergence. The global convergence criterion is set at 0.1% between iterations. This convergence procedure is repeated for each year over the simulation period.

The E3MC model simulates the competition of technologies at each energy service node in the economy, based on a comparison of their cost and some technology-specific controls, such as a maximum market share limit in cases where a technology is constrained by physical, technical or regulatory means from capturing all of a market. The technology choice simulation reflects the financial costs as well as the consumer and business preferences, revealed by real-world technology acquisition behaviour.

Model Limitations

While E3MC is a sophisticated analytical tool, no model can fully capture the complicated interactions associated with given policy measures between and within markets or between firms and consumers. Unlike computable general equilibrium models, however, the E3MC model does not fully equilibrate government budgets and the markets for employment and investment. That is, the modeling results reflect rigidities such as unemployment and government surpluses and deficits. Furthermore, the model, as used by Environment Canada, does not generate changes in nominal interest rates and exchange rates, as would occur under a monetary policy response to a major economic event.

Page details

Date modified: