.

Risk-Based Assessment of Energy Security

A Case from Europe

Boyko Nitzov*
The European Union's Agency for the
Cooperation of Energy Regulators (ACER)
Ljubljana, Slovenia
 

This paper suggests a risk-based approach to assessing energy security and the cost of its enhancement in the context of the European Union's framework. The paper highlights the inherent cost-benefit tradeoff of choosing a certain level of energy security which would satisfy the formal requirements of the regulatory framework and at the same time be cost-efficient. The suggested analytical framework is applied to the case of a Member State of the European Union.

1.   The regulatory framework

The European Union's regulatory framework related to its natural gas sector comprises several key elements, namely:

  • The so-called "third legislative package" contains inter alia two documents, namely Directive 2009/73/EC[1]and Regulation (EC) No 715/2009[2] that deal with gas. The package aims at establishing an integrated energy market by developing a reliable European energy network and investment in infrastructure. Two positive outcomes of market integration would be enhanced diversification and security of supply[3]. It is those two outcomes that we assess in terms of risked costs and benefits in this paper.
  • The Regulation on security of gas supply (Regulation 994)[4] strengthens the prevention and crisis response mechanisms and introduces formal requirements, in particular the so-called "N-1 rule", by virtue of which each regulatory authority has to assess the risks affecting the security of gas supply in its member state by using the rule and accounting for demand-side measures. In this paper, we use the N-1 rule to assess the minimum required capacity of new or upgraded gas supply infrastructure needed to achieve compliance to the Regulation, as well as to assess the magnitude of the current exposure of a country to interruptions of gas supply.
  • The TEN-E Regulation[5] sets out formal procedures and criteria for the adoption of lists of cross-border Projects of Common Interest (PCI) in electricity and natural gas, including cost-benefit analysis for assessing the proposed PCI. In this paper, we look at one possible way of introducing risk-based approach to assessing cross-border natural gas projects and their costs and benefits.

 

2.   Analytical framework

To arrive at the required risk-based assessment, a model using a four-step procedure is applied. First, the capacity needed for achieving compliance to the N-1 rule is assessed. Next, the cost of achieving compliance is evaluated. Next, the diversity of natural gas supply, the key element of security, is assessed by calculating the enhanced Shannon index (ESI, cf. Kaderjak et al., endnote 9). ESI is used as a proxy in assessing the magnitude of the loss incurred by a disruption in gas supply. Finally, the probability of interruptions in gas supply is evaluated based on actual events observed over a certain period of time, and the assessed probability is used in conjunction with the ESI as input to a Gambler's Ruin type of analysis (cf. Arps et al., endnote 12). Cost is determined on the basis of investment required to reach acceptable levels of the N-1 indicator and the ESI, and benefits are estimated as the avoided cost of natural gas supply interruptions.

The analytical frame is at sub-sector level, i.e. energy security is assessed at the level of a given kind of primary energy (in this instance, natural gas) rather than at the level of the entire energy sector. The approach allows to pinpoint the specific sources of potential threats to energy security and also to gauge the associated costs and benefits on a narrower range of values.

The output of the model allows an estimate of the maximum cost of enhancing energy security which would be economically justifiable under the specific conditions of a country or a region, i.e. an assessment of the breakeven point beyond which "buying" more security is not efficient. An example of such a cost-benefit assessment of energy security based on the actual case of a member state of the European Union is provided. For the purpose, a software utility is used (cf. Rebecca Byrd, L. et al., endnote 10).

 

3.   A model for assessing efficient levels of energy security

a.   Additional capacity needed to comply with N-1 criteria

Regulation 994 requires member states to maintain the value of the N-1 indicator at a level of at least 100% (Annex I, p. 4 of the Regulation):

Equation 1

where

  • EPm ’: technical capacity[6] of entry points, other than production, LNG and storage facilities covered by Pm, Sm and LNGm is the sum of the technical capacity of all border entry points capable of supplying gas to the area.
  • Pm’: maximum technical production capability, is the sum of the maximum technical daily production capability of all gas production facilities which can be delivered to the entry points in the area.
  • Sm ’: maximum technical storage deliverability, is the sum of the maximum technical daily withdrawal capacity of all storage facilities which can be delivered to the entry points of the area, taking into account their physical characteristics.
  • LNGm’: maximum technical LNG facility capacity, is the sum of the maximum technical daily send-out capacities at all LNG facilities in the area, taking into account offloading, ancillary services, temporary storage and re-gasification of LNG as well as technical send-out capacity to the system.
  • Im’ is the technical capacity of the single largest gas infrastructure with the highest capacity to supply the area.  When several gas infrastructures are connected to a common upstream or downstream gas infrastructure and cannot be separately operated, they are considered as one gas infrastructure.  
  • ‘Dmax’ is the total daily gas demand of the area during a day of exceptionally high gas demand occurring with a probability of once in 20 years. 
  • Deff’ is the part of Dmax that in case of a supply disruption can be sufficiently and timely covered with market-based demand-side measures in accordance with Article 5(1)(b) and Article 6(2) of Regulation 994/2010. 

The N-1 formula without demand-side measures omits Deff from the equation, but is otherwise identical to the N-1 formula with demand-side measures.

Denoting z for N-1 and Sadd=EPm+Pm+Sm+LNGm,, in case N-1 is less than 100%, the required additional new cross-border supply capacity Sadd (which may be added by any combination of increasing EPm, Pm, Sm, and/or LNGm) is given by

Equation 2, required additional capacity (mcm/d)

 

b.   Cost of required additional capacity needed to comply with N-1 criteria

The cost of new capacity, measured in total cost and in cost of service per capacity unit, varies by type of infrastructure. For natural gas pipelines, it is given by the required horsepower of the compressor stations (installed and used), the fuel efficiency of the compressors, the capital cost of constructing the line, the mode of operation (pressure differentials, etc.), the condition of the pipe ("roughness"), ambient pressure and temperature, gas composition, and other factors. In a generalized form, the total cost function of transporting gas via pipeline is[7]:                                                                       

 

Equation 3

where c1 is the cost of operation per HP/day, c2 is the construction cost per inch/mile, c3 is the construction cost per installed horsepower and p(r,t) is the imputed daily cost of the pipeline as a function of the interest rate r and project life t. Similarly, the total cost function can be derived for any natural gas infrastructure which requires a certain throughput per given period of time (capacity), for example, daily send-out capacity for LNG terminals or withdrawal rates for underground gas storage. In this paper, we use the above approach to assess the cost of a pipeline or its equivalent in terms of capacity needed to achieve compliance to the N-1 criteria.

c.    Import disruption impact indicator[8]

Diversity in meeting the fuel demand of a country or a country group, including imports, is the principal element of supply security. The Shannon-Wiener index (‘Shannon index’) measures the diversity of meeting fuel use. The general form of the Shannon index is as follows:

Equation 4

where pi is the share of fuel type i in gross inland fuel consumption and n is the number of different fuels used.

Beyond diversity, import dependence is another major determinant of supply security. To measure the impact of import dependence, an enhanced version of the Shannon index (ESI) is used, as proposed by Hirshhausen and Jansen:

Equation 5

where ci is a correction factor for each primary energy source.

The correction factor takes into account the share of net imports in total consumption of a given source of energy and the rate of diversification of the import sources of energy. Typically, the five major primary energy sources are considered (oil, natural gas, coal, nuclear, and hydro). For coal, oil, and nuclear fuel, a number of supply sources and routes are usually available, and the correction factor is set at unit. However, the value of the correction factor for these primary sources of energy must be individually assessed in each specific case. For natural gas, the correction factor's value exhibits greater "scatter" and may be anywhere between close to zero and one. For natural gas, the correction factor is calculated as:

Equation 6

where mgas is the share of imports in gas consumption, Sgas is the Shannon index for gas, Smax is the maximum of the Shannon index, and

Equation 7

where mgasj is the share of import gas from source j in the total imported gas for the given country.

The ESI is used in this paper as a proxy in the calculation of the magnitude of the impact of a disruption of imported gas supply, whereby a level below unit indicates the proportion of consumers who will be impacted by the effects of the disruption. For example, an ESI of 0.6 would mean that 40% of consumers would not be getting any gas for the duration of the gas supply disruption.

ESI is a convenient tool for evaluating the magnitude of the gas supply disruption impact, as it may be used to assess the loss due to lack of gas supply. The loss would be comprised of product which has not been produced because of operation interruptions due to lack of natural gas (the unavailable services of natural gas as an input to production and consumption), plus the lost revenue from gas sales. Loss does not include implied cost, such as loss of comfort by customers who use gas for heating and cooking and other similar loss. Equation 8 is used for evaluations of the negative impact of gas supply interruptions.)

Equation 8

where Igsi is the value of the negative impact in currency units, GDP is the gross domestic product during the year, dd is the duration of the disruption in days, Gpfis the share of gas in primary energy consumption during the year, and SIeis the ESI.

d.   Probability of natural gas import supply disruption

The assessment of the probability of natural gas import supply disruption is purely statistical and requires a sufficiently long period of time. Supply disruption "events" are considered to have occurred if lasting for more than 24 hours, regardless of the reason for the disruption. Only supply-side disruptions are taken into consideration, i.e. unexpected surges in demand are not considered as a reason for gas supply shortfall. Over the last decade, scarcely a member state of the European Union has not experienced at least several natural gas supply disruptions, which means that the probability of supply disruptions over a period of time typically needed for the construction of new gas infrastructure is very high. The issue therefore is not whether such disruptions would occur (they will), but whether their impact merits the construction of new supply infrastructure. Put simply, one has to compare the cost of constructing new gas supply infrastructure with the benefit provided by the infrastructure in terms of ability to avert the negative consequences of gas supply disruptions, over a time frame of sufficient duration.

e.   Risk and reward: applying Gambler's Ruin analysis to investment decisions on projects for natural gas supply security enhancement

For assessing whether the risk to energy security from potential gas supply disruption is of such a magnitude as to merit the construction of new supply infrastructure, we use Gambler's Ruin analysis as described by BDM Petroleum Technologies.[9] An important concern of investors[10] is the possibility of going broke through continuous failure. This risk is similar to gambling and is called “gambler’s ruin.”

Investment risk analysis involves these basic parameters:

  • Risk money (X) = the money loss for each failure
  • Potential value of the reward (R) = the commercial value of each success
  • Amount of available capital (C) = total investment or available funding for the adventure
  • Probability of success of the venture (P) = estimated success rate
  • Acceptable risk level (αm) = the acceptable level of risk (i.e., the probability of going broke through continuous failure).

Based on the Gambler’s Ruin theory presented by Arps and Arps[11], the minimum acceptable probability (Pm) of success for the venture breaking even in the long run is:

Equation 9

and the chance (α) of an investor going broke through a continuous string of failures is:

Equation 10

The investment analysis will inform the following:

  • Minimum acceptable probability of success
  • Chance of going broke
  • Minimum acceptable probability of success in order to meet the acceptable risk level.

In the case of disruptions of import natural gas supply, we use the following:

  • Risk money X is the cost of each cut-off of energy supplies. X includes direct loss (value of gas not delivered) and indirect loss (the one incurred by the users of gas due to interruptions of operations), with the latter being the product lost due to unavailability of the service of gas as input to production (i.e. the annual GDP prorated by the duration of the disruption and the share of natural gas in total primary energy consumption, corrected by ESI).
  • Potential value of reward R for each success is the imputed value of averted gas supply interruptions over the lifetime of a gas infrastructure project until decommissioning (in the case, assumed to be 30 years), discounted to present day values by using the same rate of discount as in Equation 3 (in the instance, 5%)[12].
  • Available capital C is the value of the most expensive alternative import gas supply project proposed by project promoters, for which a final investment decision has been taken.
  • Estimated probability of success P is the reverse of observed probability of gas disruptions per year, as actually observed in the country, over a five year period of time, which is the time needed from start of construction to commissioning of the new supply infrastructure. For example, if disruptions have occurred twice over ten years, the probability of “success” in a given year if the status quo is maintained is 80% (doing nothing to diversify import gas supply means that in four years out of five there will likely be no interruption of supply), and P is 0.2.
  • Acceptable probability of Gambler’s Ruin (acceptable risk level for total gas supply interruption) is one event per 30 years (3.3%), the assumed lifetime of the new supply infrastructure. We note again that EU regulations require gas supply security to be at least 100% over 20 years by using the N-1 formula.

 

4.   Facing ruin in the gas business: The case of a EU member state

The four elements of the model described above enable the assessment of the levels of risks to natural gas supply security which a given country faces, as well as making the right choice in terms of investing in upgrades of gas import infrastructure. The following analysis refers to one of the member states of the European Union and uses actual data.  As the example is for illustration only, we will refer here to that country as "Country Y".

a.   Calculation of the N-1 indicator for Country Y and determination of required new import capacity to achieve compliance to regulation 994/2010

The following actual data is used to calculate the N-1 indicator for Country Y:

  • EPm, the technical capacity of entry points other than production, LNG and storage facilities, is about 21 mcm/d, all of which at one entry point.
  • Pm, domestic gas production, is 1 mcm/d.
  • Sm, gas storage withdrawal capacity, is about 4 mcm/d, sustainable for a short period of time (about 10 days) and decreasing exponentially thereafter.
  • LNGm is zero (there is no LNG terminal in Country Y and no connections to any LNG terminal).
  • Im, the technical capacity of the single largest gas infrastructure, is identical to EPm, the only import gas entry point. Im is 21 mcm/d.
  • Dmax, the total daily gas demand during a day of exceptionally high gas demand occurring once in 20 years, is about 18 mcm/d over 1991-2011.
  • Deff, the part of Dmax that in case of a supply disruption can be sufficiently and timely covered with market-based demand-side measures such as fuel substitution and increasing efficiency of gas use, is some 3 mcm/d, the bulk of that by substitution of other fuels for gas.

With the above parameters, the N-1 indicator for Country Y calculated by using Equation 1 is 28% without demand-side measures and 33% with demand-side measures. The required additional gas import capacity calculated by using Equation 2 is about 13 mcm/d, which could be added by constructing one or more new pipelines, LNG, and/or storage infrastructure, or increasing production. Regarding these options, Country Y' gas sector features are the following:

  • By substituting other fuels for gas Country Y can reduce gas demand by up to 3 mcm/d;
  • No discoveries of gas fields have been made public that would support a significant increase of gas production over current levels of about 1 mcm/d;
  • At this time, there are no known sites in Country Y that can assure the required gas storage active gas volume and withdrawal rates.

The available options for achieving satisfactory levels of energy security, after exhausting the potential of demand-side measures and increasing domestic production, are therefore limited in the instance to either constructing new import gas pipelines or an import LNG terminal. The new infrastructure should be able to deliver 13 mcm/d without demand-side measures or 10 mcm/d with demand-side measures. In a nutshell, the country needs at least one new 28 inch pipeline or a LNG terminal of similar capacity, from a source, route, and supplier different from the current ones.

b.   Sizing and costing of the required new gas import infrastructure

By using the determined minimum of required import gas capacity, and optimizing for pipeline diameter and horsepower at various pressure differentials, total length of pipeline depending on source of gas, distances between compressor stations and other parameters, the required new gas import infrastructure for Country Y is sized and costed.

c.    Determining the impact of gas supply disruptions

By using Equations 4-7, the correction factor and the ESI are calculated for Country Y. ESI for 2009 is 1.13 overall and 0.7 for fossil fuels only, well below EU-15, where the index is about 1.32. For natural gas, accounting for the complete lack of diversification apart from very modest domestic production, ESI is a dismal 0.02: an interruption in import gas supplies will affect 49 customers out of 50.

Country Y experienced a total cut-off of gas imports in 2009 for a period of 14 days. During the year, the country's GDP was $46 billion, the share of gas in primary energy supply was 13%. Given the above data and using Equation 8, the value of the negative impact of the gas disruption is assessed at about $225 million.

d.   Probability of natural gas import supply disruption

Between 2001 and 2011, Country Y experienced two complete cut-offs of imported gas supply, which means that in any given year there is a 20% chance for a major import gas supply disruption.

e.   Gambler's Ruin: Playing dice with energy security

With the results of steps 5a through 5d above, Gambler's Ruin analysis yields a chance of going broke of over 41% and a minimum acceptable probability of success with the acceptable probability of ruin of 78.5%. This means that the current level of energy security in Country Y is much lower than required and, furthermore, that import gas supplies should be diversified within about five years if Country Y is to avoid a loss equal or greater than the available capital for executing gas supply diversification projects. In the case of Country Y, the benefits of diversifying import gas supply clearly overweigh the cost of diversifying.

 

5.   Discourse and conclusions

As far as Country Y is concerned, results demonstrate that it cannot achieve compliance to Regulation (EU) No 994/2010 without diversifying gas supply, and that in order to assure compliance, the minimum new gas supply capacity should be about 13 mcm/d without demand-side measures and about 10 mcm/d with demand-side measures, from a source and via a route which is different from the current import ones. With an ESI of only 0.02, Country Y carries an unacceptably high risk of experiencing a large loss from cut-off of gas supply. In monetary terms, given the history of gas supply interruptions, the size of the economy and its gas business, carrying such risk is equivalent to losing about $70 million every year. Country Y has a fair chance (just under 80%) of "going broke" in five years time if it continues playing dice with energy security. Failing to diversify gas supply and comply to EU regulations within this time horizon is likely to result in spending money on investment in diversification of gas supply later on and still carrying excessive risk of complete cut-off of gas supply at least once in the meantime, i.e. becoming a "ruined" gambler.

The results of the analysis also allow projects to be ranked by using multi-criteria analysis. For example, in Country Y any economically justifiable diversification project in the light of Regulation (EU) No 994/2010 must meet the following criteria:

  • Minimum capacity of 4.7 bcm/y (3.6 bcm/y with demand-side measures);
  • Maximum cost of $500 million;
  • Become operational within five years;
  • Be independent from current supply by source, route, and supplier.

Country Y is now considering at least 16 cross-border gas supply and UGS infrastructure projects, only one of which satisfies all criteria. Regretfully, it is not a priority on the government's list.

One of the features of the results of our analysis is that they are quite time-dependent. For example, if the same analysis is to be carried out in several years time, many of the variables in the model will have different values. It would also be possible to use a different period of time for determining the values instead of the one used in the case (2001-2011). There are also similar considerations related to the choice of rates used for discounting, a topic which is beyond the scope of this paper. 

What is probably most important to understand is that energy security is not an absolute value, but one which changes over time. Besides, there are clearly cases where the cost of bringing up energy security to match a certain regulatory requirement may be greater than the benefit provided by achieving improved energy security. For these reasons, discussing energy security without taking into account the specific circumstances of a country or a region in terms of patterns of supply and use of energy, prices, physical deliverability limitations, and other fine-print type of data is prone to making costly mistakes.

One may argue, of course, that buying "excess security" by investing in infrastructure that would diversify supply beyond what is actually reasonable in economic terms is still better than under-investing, which may expose the country or region to bullying by suppliers and dangers beyond the economic remit. Nevertheless, a risk-based assessment of the economic costs and benefits associated with investing in new natural gas infrastructure, for the purpose of contributing to energy security, is a must in an informed decision making process.

 

6.   Notes and References

Boyko Nitzov is TSO Cooperation Officer in the gas department of the European Union's Agency for the Cooperation of Energy Regulators (ACER). The views expressed in this paper are views of the author and do not necessarily reflect the official position of ACER or any of its Boards.

[1] Directive 2009/73/EC of the European Parliament and of the Council of 13 July 2009 concerning common rules for the internal market in natural gas and repealing Directive 2003/55/EC (Text with EEA relevance).

[2] Regulation (EC) No 715/2009 of the European Parliament and of the Council of 13 July 2009 on conditions for access to the natural gas transmission networks and repealing Regulation (EC) No 1775/2005 (Text with EEA relevance).

[3] Cf., for example, a summary of the third legislative package available at http://europa.eu/legislation_summaries/energy/internal_energy_market/index_en.htm. 

[4] Regulation (EU) No 994/2010 of the European Parliament and of the Council of 20 October 2010 concerning measures to safeguard security of gas supply and repealing Council Directive 2004/67/EC. Text with EEA relevance. The Regulation is available online at http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32010R0994:EN:NOT.

[5] Regulation (EU) 347/2013 of the European Parliament and of the Council on guidelines for trans-European energy infrastructure and repealing Decision No 1364/2006/EC.

[6] All capacity is in million cubic meter per day (mcm/d) except where otherwise indicated.

[7] The derivation of the production cost function of a gas pipeline is published by the author in: Hetland, Jens; Gochitashvili, Teimuraz (Eds., 2004). Security of Natural Gas Supply through Transit Countries. Springer.

[8] The method for calculating the enhanced Shannon Index is quoted from:  Péter Kaderják, Peter Cameron & András István Tóth (2007). "Unilateral natural gas import dependence: a new supply security issue for Europe".European Review of Energy Markets, volume 2, issue 2, pp. 13-15. 

[9] Cf. L. Rebecca Byrd and Frank T-H. Chung (1998). Risk Analysis and Decision Making Software Package (32-Bit Version) User Manual. BDM Petroleum Technologies Under Contract to BDM-Oklahoma, Inc., p. 11.

[10] Even when money comes from public funds - B.N.

[11] Arps, J. J, and J. L. Arps. 1974. Prudent Risk-Taking. Journal of Petroleum Technology 27(7): 711–716, as quoted by L Rebecca Byrd et al., op. cit., p. 11.

[12] The importance of time horizons and rates used for assessments are discussed in Section 5.

 

Click to view a printable version of this article.