Welcome to EnergyIntensityIndicators’s documentation!

EnergyIntensityIndicators provides a framework to quantify how energy efficiency may or may not be affecting energy use relative to other economic trends.

This model is based on decades of work by the Pacific Northwest National Laboratory (PNNL) for the Department of Energy’s Office of Renewable Efficiency and Renewable Energy (EERE). The original PNNL methodology documents these efforts, the results of which, along with related analyses are documented here.

The model enables the decomposition of energy use for economic sectors through the use of a Log Mean Divisia Index (LMDI) model. The LMDI methodology decomposes energy use into three main categories: activity, structure and intensity (i.e. energy efficiency). A user can thus explain changes to overall energy use in a sector through changes to sector output (activity), structural shifts within sectors (e.g. across transportation modes), and efficiency of energy use in that sector.

The LMDI model, developed by B.W. Ang and K. Choi, has several permutations featuring varying mathematical characteristics. LMDI-I is “consistent in aggregation” and “perfect in decomposition at the subcategory level,” while the LMDI-II model is preferred for index construction (Ang, 2015). In order to enable users to choose the LMDI version that best suits their needs, this EnergyIntensityIndicators library allows for choice between Additive (LMDI-I or LMDI-II) Multiplicative (LMDI-II) models. The LMDI model and its various versions are explained in further detail by Ang here , with additional materials linked in the Relevant Background Literature section below.

This project aims to add flexibility to the original model, which was implemented in Excel spreadsheets; the EII framework offers user choice in the following dimensions:

  • Additive (LMDI-I or LMDI-II) vs Multiplicative LMDI (LMDI-II)

  • Base Year (default 1985)

  • Economic Sector

The LMDI module outputs results in formatted csv for each sector and model appropriate visualizations (i.e. waterfall charts for additive, line charts for multiplicative) for each level of aggregation desired.

Indices and tables