Older Archives

How to measure security of supply?

September 9th, 2007 by Anne Neumann, Universität Potsdam and DIW Berlin

Everybody talks about security of supply these days, but there have been few attempts to put figures on this concept. In this contribution we develop an index to measure security of supply for the main fuels based on a measurement of diversity derived from biology. Our extended index takes into account the political stability of the exporting country, and the domestic energy production of the importing country or region. We apply the indicators and find that, in international comparison, Europe’s supply security is less critical than generally argued.

Besides competitive and clean energy supply, security is the third cornerstone of the EU energy policy. However, the debate is currently based on intuition rather than quantifiable indicators. We start by looking at an indicator of diversity known as the Shannon-Wiener-Index and apply it to energy. It is defined as where xi is the amount of fuel imported from a particular country, divided by the total imports of that fuel; the i are the countries from which the country is importing. Because the natural logarithm of a fraction is always negative, the minus sign at the beginning of the equation ensures that the index is always positive. The index increases as the number of different supply sources increases. Complete import dependence from one country yields an indicator of 0 (since ln1 = 0), whereas the indicator will generally not be higher than 2 (that would be seven equal shares of imports). To our knowledge, the Shannon-Wiener diversity index is the only indicator applied in energy so far (e.g., for the UK electricity system).

We now proceed with two extensions of the index:

First, political stability of the exporting country or region is an important determinant of supply security. This is often reflected by price discounts given to reliable countries. We therefore suggest to include an indicator of political stability in exporting countries into the index (let’s call it for the moment being the Shannon-Wiener-Neumann index, or SWN): , where bi, the index of political stability bi of a producing country or region, ranges from 0 to 100 where a high index indicates high stability. The SWN1 thus reduces the index for imports from unstable regions.

In a second step, let us consider indigenous production as well. In fact, the problem of the SWN1 is that it does not take account of production within a country. Imagine a country that only imports 5% of its energy consumption from one single country: it would have a SWN1 of 0, although its supply security is very high. Therefore, we define , where gi is the share of indigenous production of total consumption of a particular fuel. Countries with a high level of own production face less risk from supply disruptions, which is reflected by the SWN2. We are aware of the fact that the adjustment for indigenous production does not sufficiently capture all aspects and are open for better suggestions.

Figure 1: Shannon-Wiener index
Figure 1: Shannon-Wiener index

Source: Own calculations based on IEA (Coal Information and Natural Gas Information, 2006) and BP (2007)

Figure 1 shows the results of the traditional Shannon-Wiener index. The figures for natural gas show an evenly diversified portfolio for European imports, in particular for Spain and for France. Overall, natural gas trade has grown by 34% since 2001. LNG volumes represented 28% of 748 bcm totally traded natural gas in 2006 (26% in 2001) and increased the number of suppliers to OECD Europe to ten non-European countries (in addition to substantial supplies from Norway, the Netherlands, and the UK). Imports from the Russian Federation accounted for 24% in 2006. The US produced 85% at home and imported 15% of all natural gas per pipeline from Canada. Japan has a negligible production but and imports from ten countries, all of which was LNG. The biggest suppliers were Indonesia (22%) and Malaysia (21%). This leads to a high Shannon-Wiener index.

A similar pattern emerges for steam coal imports to Europe (we do not consider coke because this is a very specific market). Overall, only 16% of produced hard coal was internationally traded in 2005 (from 2000 to 2005, the amount of yearly traded coal more than tripled from 210 Mio. t to 775 Mio. T). In 2005, OECD Europe imported more than half of consumed steam coal from nine countries. South Africa and the Russian Federation accounted for 35% of total imports. In comparison, US consumption was covered by more than 98% indigenous production. Japan has no production and received more than half of all imports from Australia. Japan’s low level of diversification is also shown by its low SW-index.

Last but not least, Europe and the US accounted for 25% of total crude oil trade each and Japan for 10% (2005). The main exporting regions were the Middle East (38%) and the Former Soviet Union (13%). Imports to OECD Europe came from 18 countries and added to a high level of indigenous production. The derived SW for oil is particularly high for the US, Spain and France indicating a well diversified delivering structure. Recalling that an index above two indicates a stable portfolio we notice that for the majority of oil imports no concern seems to be justified.

Now we look at the adjusted SWN1 incorporating an adjustment for political risk in exporting countries and show the difference to the traditional index.

Figure 2: Difference to SW–index incorporating political risk (SW-SWN1)
Figure 2: Difference to SW–index incorporating political risk (SW-SWN1)
Source: Own calculations based on IEA (Natural Gas Information, Coal Information, Oil Information, 2006) and BP (2007)

In general, introducing risk measurement of exporting countries reduces the original index most significantly in the case of oil. For natural gas, Germany and the Netherlands are the least affected, revealing an import portfolio from a majority of stable countries (Note that the risk index for Algeria is 46 and for the Russian Federation 62.5 (PERC Ltd.)). The decrease of the Spanish index is mainly driven by LNG imports from the Middle East and Africa. For steam coal, the risk adjustment results in a decline evenly distributed amongst countries. This mirrors the importance of a small number of big steam coal suppliers worldwide. The dependence on imports from OPEC-countries with a low stability index is known to all. Interestingly, Spain and the US are the most affected by this adjustment. One could argue that the energy sector in less stable countries should be rated less risky than the overall stability, but we leave this to political scientists.

To finish, we take a look at the final results of SWN2 in Figure 3 taking into account indigenous production gi.

Figure 3: Shannon-Weiner-Neumann index (SWN2, political risk and indigenous production)
Figure 3: Shannon-Weiner-Neumann index (SWN2, political risk and indigenous production)
Source: Own calculations based on IEA (Natural Gas Information, Coal Information, Oil Information, 2006) and BP (2007)

In general, the index adjusts again substantially into the direction of the traditional Shannon-Wiener-Index but does not reach its original level.
For natural gas the index increases substantially for producing countries (Netherlands, the UK, and the US). However, taking into account decreasing production in the years to come should be taken into account. Nevertheless, the discussion of natural gas supply security in OECD Europe seems to be at threat less severe than argued in many places. Countries raising their security of coal supply through domestic production are Germany, Spain, the UK and Japan making it a secure energy resource for Europe and Japan. Moreover, all indices for oil imports, except for Turkey exhibit a comforting security level. Turkey imports about half of its oil from Iran and Iraq and does therefore depend on only few political stable countries represented in a lower index. The indices for European countries’ oil security do not differ significantly mainly due to similar importing portfolios (Norway, Former Soviet Union).

In this blog contribution we have tried to develop an indicator for supply security in the European energy market. We use an accepted measure of diversity and adjust it accordingly for an application to fossil fuels. The results indicate a rather strong position of European countries importing natural gas, oil, and steam coal in case of severe and prolonging supply disruptions. However, the weaknesses of the used index have been pointed out and leaves room for further discussion.

Anne Neumann

Send this to a friend

5 Responses to “How to measure security of supply?”

  1. Jon Stern Says:

    A very interesting and useful piece but it only covers one of the two main sources of supply insecurity. In practice, almost all electricity and gas supply interruptions experienced by consumers have arisen from network failures or accidents. These have been primarily in transmission/transport and to a lesser extent in distribution

    Is it possible to extend the work reported to measure security of supply from these issues? That would be extremely valuable. If doable, it could usefully be combined with the energy import index.

  2. Lars J Nilsson Says:

    I agree, an interesting piece! In addition to the risk of network failures, the analysis could also be expanded to include the security and portfolios of exporting countries. It should be noted that as much as some countries are depending on imports, other countries depend on exports and presumably have an interest in the long term well-being of their customers.

    It would also be interesting to expand the analysis to the security of food supply and biofuels. For the latter, it seems like supply security in the EU is generally interpreted as domestic production. A study of the balance between food supply security and energy supply security could lead to the conclusion that Europe should import less energy and more food (and use more agricultural land for bioenergy production)?

  3. Jacques de Jong Says:

    Triggered by both Anneneumann’s intersting contribution on measuring supply security and Jonathan Stern’s comment and questions, i would like t refer to the study CIEP and ECN made last year and its 2007 update on EU Standards for Energy Security of Supply (De Jong, Jacques, Hans Maters, Martin Scheepers en Ad Seebregts (2007): EU Standards for Energy Security of Supply (update), CIEP/ECN, April 2007)

    The model developed by CIEP and ECN focuses on a process that is based on a common and objective framework for reviewing and assessing energy supply security on the basis of pre-agreed criteria. The model therefore has to be seen as a policy tool and not as a regulatory tool. It uses two quantitative indicators and includes some qualitative considerations. The first quantitative indicator deals with the risk of sudden unforeseen short-term supply interruptions and the capability to manage them. This is the Crisis Capability or CC Index. The second indicator is covering full energy supply and demand balances, both present and future ones. This is the Supply/Demand Index or S/D Index. Finally, qualitative considerations concern the multilateral measures for securing overall producer/consumer relations and safeguarding vulnerable transport routes for oil and gas.

    The Crisis Capability Index (CC Index) combines the risk of a country to be confronted with sudden supply interruptions and its potential impact (the Risk Assessment, RA) and the capability of that country to manage and mitigate these impacts (the Mitigation Assessment, MA). Each country is invited to make its own RA and MA on the basis of checklists with some simple scoring values. If the RA is higher than the MA value, the CC Index gets a value of less then 100. The CC Index model in itself is complete and ready for use, but concrete examples of the outcome for individual MS or the EU have not been developed, among other reasons due to a lack of confidential data.

    The S/D index aims at review and assessment of energy security of supply in the medium and longer run. The S/D Index covers final energy demand, energy conversion and transport and primary energy supply. It uses four types of inputs, two objective types and two types of a more subjective nature. The objective inputs concern the shares of different supply and demand types (i.e. for demand: industrial use, residential use, tertiary use and transport use; for supply: oil, gas, coal, nuclear, RES and other) and the values characterizing capacity and reliability in conversion and transport based on the secondary energy carriers (electricity, gas, heat and transport fuels). The subjective inputs concern the weights that determine the relative contribution of the different components in the Index (such as the relation between supply and demand outputs in the Index, or the relation between EU imports and non-EU imports) and the scoring rules for determining various Index values reflecting different degrees of perceived vulnerabilities.

    The use of the S/D Index is illustrated in the study with examples for the EU-27 and its Member States for the years 2005 and 2020. The average value of the S/D index in 2005 is about 65, ranging from 82 (Denmark) to 25 (Cyprus). MS with high import dependencies for oil and gas, combined with high shares of these imports originating from outside the EU/Norway, have a relatively low score as expected (Cyprus, Luxembourg, Malta, Latvia, Greece, Lithuania and Portugal). On the other hand, MS that are net exporting for gas and/or oil have a relatively high score (Denmark and the UK). As most of the largest MS (Germany, France, UK) are part of the best scoring MS, the EU27 aggregate score is also relatively high (65). It is interesting however to note future developments, that were calculated based on the 2020 updates from the DGTREN baseline scenario. On average, the S/D Index decreases by almost 3 points compared to 2005 (from 65 to 62), with Ireland and the UK observing the largest decrease. These decreases are generally caused by higher import dependencies in 2020.

    The qualitative factor is dealing with multilateral policies and makes a distinction between the willingness of a MS to participate in multilateral approaches and programmes and its capabilities to follow through on these intentions. Here again, the Member State is invited to develop a policy document with these two elements, in which all kinds of energy diplomacy dealing with producer/consumer relations is discussed, together with participations in joint projects supporting these relations. A more controversial element might be the actions of a more military nature, securing vulnerable transport routes of oil and gas.

    The example of a set of standards for EU energy supply security in a wider sense then for instance “just the gas market” is used in illustrating the idea of a more integral approach to energy supply security at EU levels. When energy markets in the EU and even around the EU are increasingly integrating, mutual dependencies are further expanding, as would be the outreach of national and EU energy policy decision making. Policies aiming at enhancing gas supply security in a strict sense will increasingly loose therefore their effectiveness, due to their impacts on the whole energy supply and demand equation. That does not mean to say that a further analysis of the gas value chain and the impact that the internal market designs are having would be inappropriate. To the contrary. But this analysis would the more so have to take into account as well that the internal market concept for gas is not solely impacting the gas value chain, but has wider impacts on the whole energy balance as well.

  4. Pierre Noel Says:

    This is not a very useful index. The idea — central to these calculations — that supply security is an import problem is wrong for oil and coal and disputable for gas.

    1) Security of oil supply for any country is not linked to the diversity of import sources as this is a global, fully integrated market. It is of no value to import most of your oil from a variety of “stable” countries, or to produce a large share of the oil you consume domestically, if the global market itself is highly dependent on one or a few highly unstable regions. For this reason, it is most likely that the level of oil security is roughly the same for all OECD countries (and actually most countries in the world), irrespective of their degree of “import dependence” and where they source their oil from.

    2) The US has only one (meaningful) supplier of imported gas, Canada. This lack of diversity translates into a low index of security (even after taking into account Canada’s stability). But the reality is that the US and Canada are close to being a single gas market with a degree of self-sufficiency of roughly 100%. Still, one of the main sources of domestic US gas production, the Gulf of Mexico, is exposed to a serious and growing risk of disruptions from hurricanes, which is not reflected in the index: political instability is not the only potential source of supply disruption. Finaly, gas security is largely determined by market institutions and the US, because of its flexible competitive gas market, is much better equiped than Europe to cope with gas supply disruptions, whatever their origin.

    3) The index suggests that if, say, the UK could produce 100% of the coal it consumes it would enjoy a higher degree of energy security. But the biggest energy supply disruption in the UK’s history was the result of the coal miners’ strike in the early 1980s. Similarly, the (less known) French miners’ strike in 1963 had a major impact on the country’s energy supply. Finaly, such an index would fail to take into account factors like bottlenecks in the Chinese railway system which transports Chinese coal from producing regions to consumption centers.

    The author’s conclusion that the level of energy security in Europe is higher than most people think might well be right, but the index proposed here (whether the simple or the sophisticated versions) is not useful to make this point and the comparisons between countries are not reliable.

  5. Jon Stern Says:

    I agree with Pierre Noel that indices placing high weight on domestic supply can easily provide very poor measures of energy security. The question is how to measure genuine diversity, as would be done by financial economists in measuring portfolio diversification based on a balance of risk-reward criteria. Unless this is done, the supposed ‘security of supply’ indices provide high marks to domestic supply – including domestic monopoly suppliers – and long-term supply contracts rathe than to genuine security from diversity.

    But, what about networks and measuring network security? In practice, OECD countries have never since 1960 or earlier experienced energy supply cuts from interruptions from oil, gas or coal import cuts. There have been major price spikes (eg 1973-4, 1979 and in recent years but no cut in supply volumes that has led to supply rationing. In contrast, neither Russia nor any Middle Eastern oil or gas supplier has deliberately cut supplies below contract for reasons of economic or political muscle-flexing – for good reason given the prospective impact on their long-run reputation as suppliers and the large sunk investments.

    We have, though, had several supply interruptions because of interconnector and domestic transmission/transport failures as well as regular supply interruptions at the distribution level (eg 10-30 minutes per year with some longer ones in response to storms, high winds, network failures, etc.)

    In the UK this summer, large parts of South West England were without power (and mains water) for over 1 week because of flooded sub-stations. Indeed, had flood water levels risen another 25 cms, over 1 million people would have been without power for several weeks as a major regional substation would have been taken out of action. The British government were seriously planning for major evacuation of very large numbers of people

    So, can we please focus on – and try to measure – genuine security of supply issues rather than those relevant for fighting major World Wars 1 and 2.

    Jon Stern

    PS To avoid any misunderstanding, I am Jon Stern the regulatory economist and NOT Jonathan P Stern the leadining gas expert. We often agree on issues like this but, he may not. I would not want my statements to be attributed to him and similarly on his views being ascribed to me.

Leave a Reply

You must be logged in to post a comment.