G-Cubed Autonomous energy efficiency improvements
Table of contents
Autonomous Energy Efficiency Improvements (AEEI) are percentage improvements in energy efficiency of economic activities (production by sectors and consumption by households). The improvements reduce the energy required for production or consumption activities.
Over time, firms (and households) get more per unit of energy input, without needing a price signal or technology shock to make it happen. It is “autonomous” in the sense that it is exogenous to the model solution: it doesn’t depend on carbon prices, research and development, or behavioural changes.
The use of AEEI in G-Cubed models is described in Stegman, Pearce and McKibbin (2004).
In G-Cubed model projections, AEEI growth:
- Reduces baseline energy demand growth - Even without policy, energy use grows more slowly because you need fewer petajoules per unit of GDP.
- Lowers emissions before any climate policy is applied - Since CO₂ is tied to physical energy use (particularly fossil), higher AEEI means lower baseline emissions.
Faster AEEI means:
- Less need for explicit carbon policy to achieve emissions targets.
- More of the transition is driven by technological and efficiency drift.
On the other hand, slower AEEI means:
- Policies must do much more work.
- Energy prices and economic adjustment costs rise.
The projections in a G-Cubed model (and indeed simulation layers) can include adjustments to the default zero values for annual AEEI adjustments.
The AEEI adjustments are percentage improvements in energy efficiency for a sector, or for household consumtpion, in the specified region for a given year. Thus, a value of 1 is a 1% improvement in energy efficiency in that year.
The default name of the CSV file containing the AEEI adjustments is autonomous_energy_efficiency_improvements.csv. This file is located in the data folder. That file name can be overridden in the model configuration using the ModelConfiguration.AutonomousEnergyEfficiencyImprovementsFile setting.
The AEEI file records AEEI adjustments for all regions and all sectors within each region, through each of the projection years. It also captures these improvements for consumption.
An example layout for this CSV file is shown below for the 2 region/2 sector model:
| 2022 | 2023 | 20124 | …. | 2150 | |
|---|---|---|---|---|---|
| AEEI(a01,USA) | 0 | 0 | 0 | 0 | 0 |
| AEEI(a01,USA) | 0 | 0 | 0 | 0 | 0 |
| AEEI(a01,ROW) | 0 | 0 | 0 | 0 | 0 |
| AEEI(a01,ROW) | 0 | 0 | 0 | 0 | 0 |
| AEEIC(USA) | 0 | 0 | 0 | 0 | 0 |
| AEEIC(ROW) | 0 | 0 | 0 | 0 | 0 |
The first row contains the ordered years as column labels.
Note that the first year can be before the first projection year. However, the last year must be the last projection year.
The first column contains row labels. All row labels are made up from two components, in the following order:
- The prefix
AEEIfor production by sectors or the prefixAEEICfor consumption. - In brackets, the set combinations that are affected by the autonomous energy efficiency improvements. For sectors, the set combinations are a sector code followed by a region code. For consumption, the set combination is just a region code. For example, the row labelled
AEEI(a01,USA)is the AEEI projections for sector 1 for the United States.
The consumption rows must be the last rows in the file.
The sector rows must be in the SYM-defined sector order.
The rows must also be in the SYM-defined region order as you work down the file.
When a model is setup to be solved, if the baseline exogenous projections file does not exist, it will be regenerated from the file containin the raw AEEI adjustments.