A statistic in the MPs’ ‘form letter’ (see my previous post) intrigued me. Quoting from the Carbon Plan, the letter said ‘industrial emissions have fallen by 46 per cent’. This post is about that statistic.
[Update: Feb 2013 In early 2013, DECC’s website moved to http://www.gov.uk, and many of the links in this post now don’t work. Some of the documents are now here:
DECC’s Carbon Plan 2011 says (para 7):
‘Since 1990 industrial output has grown at an average of 1 % a year while emissions have fallen by 46 %. Industry has become more energy efficient and the UK’s industrial base has shifted towards higher value, more knowledge-intensive sectors.’
Greg Barker (Minister of State for Climate Change) also mentioned this statistic in evidence to the Energy and Climate Change Committee in January 2012:
‘It might be helpful if the Committee were to look at industrial output, which since 1990 has averaged about 1 % increase per year, while emissions from the industrial sector have fallen by 46 %. In buildings, emissions have fallen by 18 % despite the growth in population and housing. Obviously, regulation and changes in practices have contributed to that in both sectors.
Q150 Sir Robert Smith: Would the [fall in industrial emissions] partly be the outsourcing of the supply chain?
Gregory Barker: I am sure there are many reasons.’
The Carbon Plan says UK greenhouse gas (GHG) emissions are down overall by 25 per cent since 1990. So how did industrial sector emissions fall by 46 per cent when other sectors showed much smaller changes (e.g. buildings (18 per cent fall since 1990) and transport (no change)). Have we closed down almost half of our industry? Has manufacturing become much more efficient than any other sector?
The Carbon Plan does not explain how the 46 % fall in industrial emissions was calculated or give a reference to its source (oddly, for a document that has over 100 footnotes). I assumed I would find the answer in the DECC’s Greenhouse Gas Inventory but there’s no obvious fall of 46 per cent in any of the industrial emission figures given there (some fall more, most fall less, but nothing corresponds exactly with the Carbon Plan figures). Finally I found a note on the Methodology of the Carbon Plan (not a very user-friendly document). With its aid and the figures in DECC’s inventory, I have been able to reconstruct the emissions figures shown in the Carbon Plan (to within a few percent).
It turns out that the Carbon Plan’s ‘industrial emissions’ are the sum of ‘Industrial Process’ and ‘Industrial Combustion’ (the latter can be calculated from ‘Business’ emissions shown in the DECC inventory by subtracting ‘miscellaneous combustion’ (but not ‘other’!) and part of refrigeration – the subtracted amounts appear in ‘Buildings’ in the Carbon Plan).
Industrial emissions accounted for about a third of the UK’s annual GHG emissions in 1990 and now about a quarter, having fallen by 112 MtCO2e (megatonnes CO2-equivalent) since 1990. 46 MtCO2e of this fall was from industrial process (which represents an 80 % fall in those emissions) and 66 MtCO2e from industrial combustion (a 35 % fall, of which about a third occurred between 2008 and 2009).
Industrial Process was clearly an important factor in the reduction. Industrial Process emissions are the GHG emissions from chemical processes (both CO2 and non-CO2 GHG). The following graph shows that most of the fall in Industrial Process emissions was due to a fall in non-CO2 greenhouse gases. Those gases, such as methane, nitrous oxide and halogenated gases are more powerful greenhouse gases than CO2, and they are reported in terms of the amount of CO2 that would have an equivalent GHG potential.Putting together the information from the different tables in DECC’s inventory [Update: link to 2010 data tables here], I found that the main falls in industrial process emissions were in nitrous oxide (N2O) and ‘F-gases’ (fluorinated gases used in refrigeration etc.). According to a DECC factsheet on industrial process emissions, this N2O was mostly produced as a by-product of adipic acid production (for nylon manufacture). The UK’s only adipic acid factory was fitted with N2O removal equipment in 1998 (and closed in 2009). Similarly, ‘F-gas’ emissions fell in the late 1990s, due to removal equipment being fitted in the factories where they were made.
The following graph shows the falls in the different gases, each converted to its CO2 equivalent.
The DECC factsheet shows that little or no further reduction is expected:
To sum up, I reconstructed – with some difficulty – the Carbon Plan’s 46 % fall in industrial GHG emissions from the figures in the DECC inventory. I found that a third of this fall comes from the almost complete elimination of F-gases and N2O from industrial process emissions. The fall in CO2 emissions (35 % since 1990) is impressive compared to other sectors but it is not clear why over a third of this fall apparently occurred between 2008 and 2009. The fall in CO2 since 1990 can mainly be attributed to the ‘dash for gas’ and ongoing improvements in energy efficiency as well as a general reduction in energy-intensive industry. The following graph shows the changes in industrial emissions since 1990.
I find these figures disconcerting for a few reasons.
- The Carbon Plan (quoted above) follows the statistic on industrial emissions with the statement that industry has become more energy efficient and has shifted towards knowledge-based industries. To me, that implies that the 46% fall was due to a reduction in CO2 emissions. Greg Barker’s exchange with a member of the Energy and Climate Change Committee quoted above did nothing to dispel this impression. The same impression is given by the DECC factsheet on industrial process emissions. It says ‘CO2 is the dominant GHG emitted by the industrial sector’ (which is of course true, but the sheet is about industrial process emissions, which before 1998 were dominated by other gases). It does not say how much of the overall fall was due to N2O and F-gases.
- The changes to N2O and F-gas emissions were mostly achieved by one-off technological changes to a relatively few factories. They do not represent a change to ‘industry’ as a whole, and no further reductions are available from those sources.
- The 46 % figure is given without any indication of uncertainty. Non-CO2 greenhouse gases are subject to much greater uncertainty than CO2 both in terms of historical emission estimates and the factors used for estimating equivalent global warming potential. Table 10 in the DECC inventory shows that N2O emissions in particular are subject to enormous uncertainties. The uncertainties quoted in the Carbon Plan and the DECC factsheet either refer to CO2 (where the uncertainty is low) or to 2009 (when N2O and F-gas emissions were low) and so mask the high historical uncertainties in those gases.
Highlighting the ‘46 % fall in industrial emissions’ without making any of these issues clear makes it all look too easy. Did the MPs who quoted that statistic know how it had been achieved?