The National Energy Efficiency Strategy (NEES) released in 2005 set a national target for energy efficiency improvement of 12% by 2015 compared to a 2000 baseline.[i] The NEES also proposed the development of a monitoring and verification system to facilitate continuous energy efficiency measurement. The Draft Second National Energy Efficiency Strategy Review published for comment in July 2012 emphasised the importance of developing such a system to ensure that energy efficiency measures support rather than suppress economic growth.[ii]

Monitoring of energy efficiency performance should therefore be measured against a growth-adjusted 2000 baseline, and measurements should be done in accordance with the guidance of the relevant South African National Standards. This monitoring system however, is still under development, and there has been no monitoring of the NEES since it was released. The Energy Efficiency Target Monitoring System (EETMS), once completed, will rely on top-down (aggregate) data and decomposition analysis to calculate changes to energy efficiency within the South African economy.[iii]

Decomposition analysis is used to isolate the impact of different factors on energy use. Three factors are typically considered, namely changes in the level of economic activity, changes to the sectoral structure of an economy (the relative contribution of different sectors to economic output), and changes in energy intensity (energy used to produce a unit of output). A reduction in the energy intensity of a process indicates an increase in the efficiency with which energy is used within that process. Thus, when the joint influence of the first two factors is accounted for, any remaining changes in energy consumption can be directly attributed to changes in energy efficiency.[iv]

An increase in overall energy intensity as a result of an economy shifting to less energy intensive activities (structural shifts) thus does not indicate that energy is being used more efficiently, but simply that it is being used differently. Likewise, care must be taken not to attribute a reduction in energy use due to a decline in economic activity to increased energy efficiency.

Isolating true energy efficiency improvements is particularly important to avoid spurious results when assessing the effectiveness of interventions (especially costly ones) aimed at increasing energy efficiency. Such interventions can also lead to structural changes or reductions in economic activity as unintended consequences. Decomposition analysis will help to identify such cases and create opportunities for corrective measures to be undertaken.

The EETMS already allows decomposition analysis at an economy-wide level, and it is currently being refined to also allow the analysis at a sector and sub-sector level.[v] To the author’s acknowledge, however, the results of the economy-wide decomposition analysis are not yet publicly available. This makes it difficult to rigorously assess South Africa’s progress towards meeting the 2015 NEES energy efficiency targets. Consequently, this blog provides a non-rigorous look at what the available data can tell us about South Africa’s energy efficiency performance between 2000 and 2012 (the latest year for which source data from the International Energy Agency (IEA) is available).

Using data from the IEA, Figure 1 shows the energy and electricity intensity of the South African economy from 2000 to 2012. Over this period, energy intensity decreased by 13.2 percent and electricity intensity by an enormous 25.4 percent. This decline was relatively independent of business cycle impacts, since the annual economic growth rate varied widely between 5.6 percent to minus 1.5 percent over this period and was positive for all but one year (2009).

Sources: http://www.iea.org/ | https://www.resbank.co.za [Sectoral contribution to GDP]

* Series adjusted by factor of 0.5 to enhance visual clarity (actual values are thus twice those indicated on graph)

However, over the same period, the contribution of the tertiary sector (which is typically less energy-intensive than the primary and secondary sectors) to GDP increased by 5.1 percent (from 64.5 percent to 69.6 percent), while the contribution of the primary sector fell by almost a third (from 11.5 percent to 7.9 percent). At least part of the decrease in energy intensity is thus likely as a result of structural shifts within the economy rather than pure energy efficiency improvements.

While the decline in both the intensity variables started long before the global financial and local electricity supply crises of 2007-2008, it is interesting to note in Figure 3 that by 2012, neither overall electricity consumption nor overall energy use had returned to their pre-crisis levels in absolute terms despite three years of positive economic growth.

Sources: http://www.iea.org/

At the end of 2012 electricity consumption was 3.3 percent below its 2007 peak and energy use 4.8 percent below its 2008 peak, this despite the South African economy having grown by 11.6 percent between 2007 and 2012 and by 7.7 percent between 2008 and 2012. Since 2000 (which serves as the baseline for the 2005 NEES), the South African economy had grown by 50.1% in 2005 US dollar terms while electricity consumption had only increased by 11.9 percent, and energy use by 28.1 percent, by 2012.

What is therefore clear is that, for whatever reason, there has been a steady decline in the energy intensity of the South African economy since 2000, and this decline was relatively independent of business cycle impacts. What is not clear, however, is whether this is predominantly due to increased energy efficiency and decoupling between economic growth and energy use, or whether this is simply being driven by structural changes to the South African economy.

This simple analysis also confirms the importance of decomposition analysis. A naïve look at the data would suggest that South Africa had already met its aggregate 2015 energy efficiency improvement target in 2012, while in reality there probably was (and may yet be)some work left to be done to reach this target. The results from the DOE’s decomposition analysis of energy use trends in South Africa will thus make for interesting reading when released.