Extension of the high ILUC-risk concept

As noted in section 4.2, The RED II introduced the concept of high ILUC-risk biofuels. The idea of the high ILUC-risk mechanism is that while ILUC assessment in general is plagued by uncertainty, it is reasonable to take steps to reduce to support for the use of crops that are directly associated with loss of high carbon stock land. The RED II requires that any biofuel feedstocks for which more than 10% of global expansion³³ is identified as occurring at the expense of high carbon stock areas shall be characterised as ‘high ILUC-risk’ and that those fuels should be made ineligible to count towards targets and for support under Member State schemes by 2030. Based on the first ILUC-risk assessment by the European Commission, only palm oil is currently identified as high ILUC-risk (European Commission, 2019). The results of this assessment are shown in Table 4.

Table 4 Share of production of biofuel crops identified by the European Commission as coming at the expense of high carbon stock land

Crop Share of

expansion on forested land

Share of expansion on peatland

Productivity factor

ILUC risk score (productivity and carbon adjusted share of expansion on high carbon stock land)

Wheat 1% 1 1%

Corn 4% 1.7 2%

Sugar cane 5% 2.2 2%

Sugar beet 0% 3.2 0%

Rapeseed 1% 1 1%

Palm oil 45% 23% 2.5 42%

Soybean 8% 1 8%

Sunflower 1% 1 1%

Source: European Union (2019)

It can be seen that palm oil has by some distance the highest ‘ILUC risk score’ at 42%. This is well above the 10% threshold value. The European Commission is expected to periodically review the assessment, but it would take a dramatic and demonstrable

33 The score is in fact adjusted for productivity, so that as a more productive crop the threshold for palm oil is effectively set at 25% expansion into forestland. Peatland is also counted 2.6 times in the calculation in recognition of its higher carbon stocks compared to forest areas. We henceforth refer to the productivity and peat-carbon adjusted share of expansion into high carbon stock areas as the ‘ILUC risk score’. There is also a minimum requirement on total rate of expansion before a crop can be considered high ILUC-risk.

Options for addressing ILUC in Danish regulation

reduction in the rate of deforestation and peat loss associated with palm oil for the classification to be revised. The second highest ILUC risk score is soy oil, 8%. This falls below the 10% threshold, but it is plausible that a reassessment might result in the classification of soy oil being changed if the data shows an uptick in deforestation rates (cf. Malins, 2020).

The other crops assessed all have relatively low ILUC risk scores of 2% or less, and it would seem less likely that any of those would be reassessed into the high ILUC-risk category unless very significant changes in expansion patterns are observed.

Denmark could consider building on the high ILUC-risk principle by imposing additional restrictions on biofuel feedstocks with an intermediate ILUC-risk score based on the current ILUC risk assessment. Setting an intermediate category in the range 5-10% would currently affect only soy oil, though this could change with the Commission’s next report. Setting the intermediate threshold lower, for example at 2%, could bring additional feedstock into the category (corn and sugar cane based on the current assessment). Doing so could be controversial, as whereas for palm and soy oil the results of the high ILUC-risk assessment are consistent with the results of ILUC modelling corn and sugar cane tend to have relatively low ILUC estimates in modelling. An ILUC risk score of 2% could also be seen as reflecting a fairly low risk of deforestation – for those feedstocks land use change emissions from non-forest land conversions might be expected to be an equal or greater emissions source, making it less justifiable to focus only on high carbon stock land. While the RED II does not make any requirement that the high ILUC-risk assessments should align with ILUC model results, we would suggest that the ILUC-risk approach is on firmer ground when the resulting regulatory action is also supported by best evidence from ILUC models.

The most obvious ways to restrict support to ‘intermediate ILUC-risk’ fuels would be to impose a limit on total supply volume (to be phased in on a similar schedule to the restriction on high ILUC-risk biofuels) or to create a category of lower value biotickets, for example by counting intermediate ILUC-risk fuels as half of their physical energy content.

8.4.1 Market mediated ILUC-risk

The high ILUC-risk framework is intended as a way to use a direct impact metric to inform regulatory decisions in order to reduce expected indirect impacts. It therefore differs from ILUC analysis in that it ignores market linkages between feedstocks. This is particularly important among vegetable oil feedstocks, as the lack of market linkages results in rapeseed and sunflower oils being given very low ILUC-risk scores even though they are linked by the vegetable oil market to high ILUC-risk palm oil. It would be possible to reintroduce an element of consequential thinking to the high ILUC-risk approach by considering that demand for one vegetable oil may trigger supply of another. Figure 20 and Figure 21 show results from MIRAGE and GLOBIOM respectively for the change in vegetable oil supply associated with increase in demand for biodiesel from a single feedstock. On the left of each chart, we see that both models assume that when palm oil demand increases this is overwhelmingly met by increased palm oil production. In contrast, the third bar of each chart shows that a large fraction of an increase in soy oil demand is expected to be met with increased palm oil production. This difference is explained by the fact that soy oil represents less than half of the value of the soy crop whereas palm oil is most of the value of the palm crop.

We therefore expect palm oil supply to be more sensitive to vegetable oil demand than soy oil supply is.

Figure 20 Increases in vegetable oil production in response to demand for different biodiesel feedstocks in MIRAGE

Source: Laborde (2011)

Figure 21 Increases in vegetable oil production in response to demand for different biodiesel feedstocks in GLOBIOM

Source: Valin et al. (2015)

By combining the ILUC-risk scores from the European Commission analysis with the ILUC model outputs on the amount of additional consumption met by increased supply of each oil, it is possible to calculate a new ‘market mediated ILUC-risk’ metric.

0%

20%

40%

60%

80%

100%

Palm oil biodiesel Rapeseed oil biodiesel Soy oil biodiesel Sunflower oil biodiesel Palm oil Rapeseed oil Soy oil Sunflower oil

0%

20%

40%

60%

80%

100%

Palm oil biodiesel Rapeseed oil biodiesel Soy oil biodiesel Sunflower oil biodiesel Palm oil Rapeseed oil Soy oil Sunflower oil

Options for addressing ILUC in Danish regulation

Table 5 ILUC-risk scores obtained by cross referencing feedstock ILUC-risk assessments with model data on combination of oils expected to meet additional demand

‘Market mediated' ILUC-risk score Palm oil biodiesel 37.2%

Rapeseed oil

biodiesel 10.5%

Soy oil biodiesel 22.3%

Sunflower oil

biodiesel 10.0%

Source: Own calculation based on (European Commission, 2019; Laborde, 2011; Valin et al., 2015) The resulting metric shows the hierarchy that one would expect – the strongest deforestation link is for palm oil, then soy oil and then rapeseed and sunflower oil. Both rapeseed and sunflower have scores on this new metric of 10% or more. Some form of limitation could be imposed on feedstocks that have a market mediated ILUC risk above 10% or some other threshold, which would provide an alternative (or complementary) justification for limits on these other vegetable oils.

The obvious criticism of such a market-mediated ILUC-risk metric would be that if one goes so far as combining outputs from economic models with data on deforestation and peat loss risk, why not go the full distance and take the ILUC results from the model as the main regulatory metric? There is no simple answer to that challenge, and such a compound metric would be likely to be perceived as complex (perhaps over-complex) by stakeholders. Nevertheless, results such as this provide a useful reminder that the high ILUC-risk framework as it stands fails to capture the full risk of ILUC emissions associated with the other vegetable oils.

8.4.2 Compatibility with the RED II

The RED II defines only the high ILUC-risk category, and therefore adding a

‘intermediate ILUC-risk’ category would need to be justified under Article 26(1) as a form of additional limit on biofuel supply. In practice, adding a new intermediate ILUC-risk category would be comparable in legal terms to imposing additional limits on a single feedstock as discussed in section 8.1.1, and we believe it would be defensible in the terms of Article 26(1).

The market-mediated ILUC-risk calculation presented above is novel, and combines elements of the high ILUC-risk assessment with ILUC modelling. Given that the intention of the high ILUC-risk assessment in the RED II is to allow measures to reduce ILUC to be defined without relying on ILUC modelling results it might be expected that the European Commission would be reluctant to endorse such an approach.

8.4.3 Administrative burden

Creating an additional category of intermediate ILUC-risk crops would require some initial setup within the Danish system, but (much as discussed in section 8.1.2) we would not anticipate a significant longer-term increase in administrative burden.