3. Resource economics and development sustainability: scenarios
p. 231-242
Texte intégral
1Today, the available knowledge on mineral resources in French Polynesia, including polymetallic crusts, is too fragmentary to allow for the production of a proper economic feasibility study of their potential. Major research efforts will be necessary (see the following section). However, it is possible to construct scenarios that should help make and further refine political decisions on the exploration and exploitation of the deep-sea mineral resources in French Polynesia. These scenarios involve geological criteria (the depth, topography, content and thickness of the polymetallic crusts) and, from an economic point of view, two main categories of variable: resource valuation (price) and production costs (depending on the selected operating technology, cost of labour and inputs, and taxation). These scenarios also concern the technological and strategic options for extraction and/or whether processing will take place on site or outside the EEZ.
2To develop these scenarios we used the state of knowledge presented in the previous section and the potential economic opportunities on global markets for the main minerals in the polymetallic crusts. In the last section of this part, we will utilise these scenarios to address the issue of mining revenue distribution.
Economic opportunities: metal markets
3Using the data published by Martel-Jeantin et al. (2001), Bonneville (2002) and Hein et al. (2013) on the Kaukura region (Tuamotu) in French Polynesia, it is estimated that 89% of the value of its deep-sea mineral resources lies in only five metals, listed here in descending order by abundance: manganese (Mn), cobalt (Co), titanium (Ti), nickel (Ni) and platinum (Pt). The total in situ value of this type of ore is USD 1,026/tonne, an extremely high value roughly equivalent to copper ore containing 20% Cu; in other words, an extraordinary resource that is far more valuable than the ca. 1% copper ore currently being mined (see I-4).
Cobalt
4Cobalt is expected to account for 24% of the average economic value of the crusts, based on current information. The cobalt market is very dynamic, underpinned by the many uses for the metal (e.g. Li-ion batteries and superalloys). The demand for cobalt is expected to remain high in the years to come. And yet, there is little reason for alarm with regard to resource availability and little economic pressure to seek new cobalt resources in the ocean. One benefit -and drawback -of onshore mining is that cobalt is a by-product of copper and nickel mining, which reduces production costs. The grade of the largest copper-cobalt deposits (in Congo and Zambia) is also quite high. Nonetheless, operators may be tempted to develop deep-sea polymetallic crusts given the major geopolitical risks, from a Western perspective, currently associated with the resource (evident in price volatility). In fact, in 2013, 54% of production and 43% of refining took place in the Democratic Republic of the Congo and China respectively. Moreover, cobalt production in the Democratic Republic of the Congo could be impacted by Glencore’s decision in September 2015 to suspend operations in its copper mines until 2018. However, this decision has yet to slow down the drop in cobalt prices (a drop of approximately 10% in the fourth quarter of 2015).
5The market remains small, though it is growing rapidly. To avoid disrupting the market, a deep-sea cobalt mining project should aim to produce no more than 10,000 tonnes of cobalt metal per year (it is also necessary to demonstrate the competitiveness of mining the Polynesian crusts).
Manganese
6Manganese is expected to account for 50% of the average economic value of the Polynesian crusts. The manganese market is closely linked to global steel production, 91% of the manganese consumed being used to produce steel and superalloys. Global steel production is experiencing a profound crisis, with the development of excessive production capacity at a time when urbanisation in China has entered a structural slowdown, with little indication that demand will grow in other countries or regions of the world (particularly India and Africa). It is likely that within a few decades the manganese economy will be affected by a gradual decrease in the graded of terrestrial deposits, the geological probability of finding new rich deposits being low (though a high recycling rate -above 50% -should be noted). Moreover, half the world’s reserves are in two countries that pose major geopolitical risks: South Africa and Ukraine. This could stimulate investment in mining polymetallic nodules. Their manganese content (21.7%) is comparable to that of low-grade deposits on land, i.e. approximately half the content currently being mined on land. It is thus unlikely that exploration investments will be made to seek out manganese in the ocean any time soon. That said, if the value of cobalt and other metals justify mining operations, recovering manganese as a by-product could be profitable.
Nickel
7Nickel is expected to account for approximately 6% of the average economic value of Polynesian crusts. Nickel is used for alloys, particularly with iron (66% of the nickel consumed is used for stainless steel, over 90% of which is recycled). Nickeliferous laterites and magmatic sulphides are the primary sources of nickel production, in Indonesia, the Philippines, New Caledonia, Brazil, Cuba, Canada, Russia, Australia, etc. The global nickel industry currently suffers from excess capacity and very low metal prices. If these economic conditions continue, profound restructuring of the nickel industry will be needed and the possible development of polymetallic crusts in French Polynesia would need considerable economic arguments in its favour to generate interest.
Copper
8Copper is expected to account for a small portion of the average economic value of the Polynesian crusts, around 1%. However, it was included in the economic litera-ture on the subject (see the section on scenarios below) and maximum recovery of metal content is clearly desirable, if it is technologically and economically feasible. Copper is one of the base metals whose consumption is strongly driven by emerging countries (especially India and China; the latter used 8 million of the 18 million tonnes produced worldwide in 2014; Jébrak, 2015: 11) because of its many industrial uses, generally in the form of alloys (telecommunications, construction, transport, energy and renewable energies). It should be noted that the recycling rate is very high (over 70% of the copper extracted from the Earth is still in use; ibid.: 12). Moreover, Chile’s central role in production (about 40%, with the United States in second place) means that the current geopolitical risk in terms of supply is quite low.
Titanium
9Based on the price adopted, titanium is expected to account for 8 to 20% of the theoretical value of Polynesian crusts. There are no pressures on titanium resources. They are very abundant worldwide and the presence of a certain percentage of titanium in French Polynesian crusts should not be considered a particular asset. With the current state of knowledge of economic conditions, the profitability of metallurgical recovery of this metal cannot be determined. It is impossible to know whether this will be technically feasible, under competitive economic conditions, while at the same time recovering cobalt, manganese and nickel. The above examples show that value is primarily generated through the metallurgical processes to produce cobalt.
Platinum
10Over 90% of platinum resources are located in high-risk countries (e.g. South Africa, Zimbabwe and Russia) but, as with rare earth elements, the values reported for the crusts are very low. The concentrations are less than 20% of the values for terrestrial ores, and in current conditions, a deep-sea mining operation is unlikely to be cost-effective.
Economic scenarios
11Given the development challenges linked to mining revenue, we need to put its potential into context in order to take the appropriate policy decisions. Economic rent is defined as the net present value excluding taxes (NPV21), as defined by the IMF (IMF, 2012). This definition has the advantage of being directly linked to profitability analysis for the projects. The rent depends on the prices of saleable metals and production costs. In a cost-benefit analysis, the issue of amortizing exploration costs and any spending on research and development (to develop a new process or a new type of equipment, for example) is a particularly thorny issue because investors will want to recoup that investment, which is not included when calculating the capital cost for the project.
12The fragmentary knowledge on the state of the resource (content, thickness, surface area and microtopography) and the lack of answers to important questions on the mineability and metallurgy of the polymetallic crusts rule out any economic feasibility studies (the studies reviewed in the literature cannot be considered as such either).
13To illustrate what is at stake economically with a polymetallic crusts project, we have put forward a simulation of its profitability and macroeconomic impact (see I-5). To conduct this simulation we discuss the hypotheses in two alternate scenarios: a baseline scenario and a favourable scenario. It should be noted that a scenario is not a prediction or forecast. It makes it possible to identify key variables, those that play a decisive role in project profitability. This type of exercise has the advantage of clarifying the orders of magnitude with regard to the French Polynesian economy.
14The baseline scenario is a direct adaptation of the assumptions used by Yamazaki, applying the average content of polymetallic crust samples taken from three areas of the French Polynesian EEZ (Kaukura, Tarava and north-west of Tuamotu) and correcting the production costs for the processed volumes.
15The second or favourable scenario considers a larger production volume, corresponding to 10% of the global cobalt market in 2014 or 10,000 tonnes (compared with 4,500 tonnes in the first scenario).
16These scenarios make it possible to specify the key hypotheses and identify the economic variables for the rent placing the project within the macro-economy of French Polynesia. The set of hypotheses is as follows.
17Metals. The metals of interest considered in the scenarios are cobalt, nickel and copper22. When reasoning in terms of in situ value, which also takes into account the value linked to manganese, titanium and platinum, the literature simulating the profitability of mining projects assumes that these metals will not be recovered (with the notable exception of Goto et al., 2010).
18Duration. The duration traditionally used for the production phase in the literature is 20 years. In addition to this operational period, there are the construction and ramp-up periods. In accordance with the parameters used by Yamazaki, a period of 5 years with no production is expected with production at 50% in the sixth year.
19Production. We have chosen a production goal of 4,500 tonnes of cobalt per year for the baseline scenario and 10,000 tonnes of cobalt per year (corresponding to 10% of global output) for the favourable scenario.
20Location. Samples were taken in two EEZ areas with polymetallic crusts with high cobalt content: Kaukura (Tuamotus), Tarava (Society Islands). While no samples have been taken north-east of the Tuamotus, the geological and geographic context also appears promising.
21Content. Cobalt: 1.06%; nickel: 0.61%; copper: 0.11%.
22Industrial processes and recovery rate. There are two major possibilities: local metallurgical processing (on land or offshore on a floating platform) or a process that takes place outside the EEZ entirely. For the profitability simulation, here we have chosen a process outside the EEZ entirely (similar to the one used by Yamazaki). Given the constraints in terms of infrastructure and costs (energy and labour), developing an onshore process through the construction of an ore or metallurgical processing plant in French Polynesia seems highly unlikely at this time. The idea does merit more study, however.
23Financing. In the scenarios, the required investment is funded via 30% equity and 70% loans.
24Price. To perform a simulation of gross revenues, we used the average price for the three main metals of interest (cobalt, nickel and copper) over the past thirty years, in line with the super-cycle hypothesis, given that demand from India and Africa seems to be taking over for that of China.
Baseline scenario
25In the baseline scenario, the profitability simulation is in line with Yamazaki’s work and adapted to the geological conditions of the three areas of interest in the Polynesian EEZ. This initial simulation yields an internal rate of return (IRR23) of 4%, which is too low24 to consider starting a mining project. In order for a polymetallic crust project to achieve a minimum internal rate of return of 10%, the operators’ minimum threshold, price levels need to be 60% higher than the average for the 1984-2013 period. In the baseline scenario, the prospect of mining polymetallic crusts in the French Polynesian EEZ appears very unlikely in the short and medium term.
26If we calculate net present value (NPV)25 in the baseline scenario based on an assumption that the cobalt price will be 60% higher than the average level over 1984-2013 and a discount rate26 of 10%, corresponding to a low point, we obtain the equivalent of 1% of French Polynesia’s GDP in 2011 (note that NPV is measured as a stock concept, whereas GDP is measured as a flow concept). This simulation assumes a 30% tax on profits.
27Excluding taxes, the weight of NPV represents 6% of the 2011 GDP, a more significant amount, still relatively low but comparable to the value of the fishing or pearl industry.
28When the average effective tax rate (AETR) is taken as an indicator of rent sharing (IMF, 2012) with a 30% corporate tax rate, the yields underperform the minimum 10% threshold (initial assumption of a 10% discount rate).
29In the end, the baseline +60 scenario we used to conduct simulations of the economic profitability of a polymetallic crusts project and its economic possibilities bears out the idea that the conditions required to undertake such a project are not currently met. This echoes a finding that is well-represented in the literature (Ecorys, 2014; all Yamazaki studies).
Favourable scenario
30An increase in the volume of cobalt produced, taking the total up to 10,000 tonnes (10% of the global market), profoundly changes the profitability of a polymetallic crust mining project. This level of output would position French Polynesia as a major player on the global cobalt market while avoiding a depressive effect on prices. However, as in the baseline scenario, this assumes a buoyant global cobalt market27. For a project to reach 20% profitability, price levels must be 60% higher than the average for the 1984-2013 period.
31NPV excluding taxes is then equal to 50% of 2011 GDP, a very significant amount. In this scenario, the macroeconomic impact of the project becomes considerable.
32The average effective tax rate (AETR) as an indicator of rent sharing (IMF, 2012) is 18.6% in this scenario. On average, AETRs calculated by the IMF fall in the 40-60% range (IMF, 2012). In other words, in this scenario, there would be real room for manoeuvre when it comes to implementing a tax on mining in addition to the corporate income tax, allowing the state to lay claim to a significant share of the economic rent. Regardless of the scenario however, this share would be reduced once exploration and risk compensation costs were taken into account.
33Ultimately, both profitability scenarios presented here lead to the conclusion that, while the development prospects of a project involving the crusts are limited in the short and medium term, it is possible that the economic contribution of such a project could be considerable in the longer term.
34Some works (Loudat et al., 1995) studied methods for mining polymetallic crusts (cobalt, manganese and nickel; no copper) in the laboratory and produced more attractive economic scenarios. As the facts currently stand, the only conclusion that can be drawn from the literature review and economic scenarios based on it is that the prospects for mining polymetallic crusts depend heavily on technological developments for extracting and recovering the metal they contain. Technological and geopolitical priorities pertaining to cobalt, a metal that is more resilient than many others in the face of the current depression in commodity markets, make the case for continuing and intensifying efforts in data acquisition and innovation.
35As a final precaution for the figures produced by the profitability and feasibility calculations, the values employed are monetary and rely primarily on direct use values. If we include the costs associated with restoration and compensation, especially for local populations, some of the other values (option, quasi-option and heritage) can be partially integrated. These values are more uncertain; having no markets to determine prices, their monetary valuation fluctuates wildly, ultimately depending on the balance of power between different stakeholders. Here, the importance of representations comes into play; likewise the importance of creating arenas for discussion and building shared knowledge and representations of the different dimensions of the value of a natural resource.
Infrastructure and industry options: two scenarios
36The possibility of a deep-sea mining industry must be reviewed with regard to three aspects: (1) exploration (public or public-private partnership), (2) operations (industry) and (3) ore processing and operating system maintenance (industrial activity on land).
37The discussion must take into account the fact that in the operating projections in the various scenarios, we assumed that around one million tonnes of polymetallic crusts would be extracted per year28, equivalent in volume to four bulk carriers a year, which is a relatively small volume of material to transport. The question then is what would happen to the material extracted. This has strong implications for port infrastructure development. This question may seem premature, as exploration is necessary first of all, but this review is primarily a long-term discussion of a project that may one day involve large investments in infrastructure. Several solutions are possible for processing, which here includes the production of a concentrate, but could also go as far as metal extraction through hydrometallurgy:
Scenario 1: processing takes place entirely in Polynesia (on land or offshore)
38Option 1. The ore is transported from the offshore site to the onshore port facilities, which would require modifications to the existing port infrastructure. Transporting ore requires the availability of either a factory ship that can be disconnected from the riser (a tubing system connecting the seabed to the boat, used to bring up the ore), the solution chosen by Nautilus in Papua, or a sea platform for storing the mining products and serving as a port for bulk carriers.
39Option 2. A platform would make it possible to carry out the initial mineral processing at sea. This structure should be designed to limit any environmental impact. It could even be a “floating island” capable of supporting any type of infrastructure (a hotel or even a marina or ferry terminal), including a power plant to power the collection system and surface operations performed on the collected material. This solution would result in a mobile plant that could be reused and moved as required by the mining sites on the seafloor.
Scenario 2: processing takes place outside French Polynesia
40If processing is performed outside the EEZ, bulk carriers will be used for transport, which may also necessitate the construction of offshore port infrastructure in the form of a platform for storing the ore prior to transport via bulk carrier.
41In both scenarios, it will be necessary to supply energy to the entire mining system, especially if metallurgical processing takes place. In the absence of specific references concerning the processing of polymetallic crusts, we used the known requirements for processing nickel ore (which can sometimes involve cobalt as a co-product) for informational purposes. We thus estimate that the entire mining system would require several dozen megawatts of power. In the event of major offshore infrastructure, renewable marine energy sources seem particularly well-suited to the task (solar, wind or ocean thermal energy).
42Whatever the chosen scenario, local residents will be given priority when recruiting. Training should be planned in various fields:
- scientific (geology/volcanology/tectonics, geophysics and mapping, geochemistry and metallogeny, biology and microbiology, and mining economics);
- administrative and legal (mining law, mining services, and environmental management);
- technical (engineering, materials handling, and ore processing).
The issues of rent and sustainability
43Under what conditions could the mining of deep-sea mineral resources be regarded as sustainable for the economy of French Polynesia? Based on the definition proposed in the Brundtland Report – “Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs” –, we find the usual dichotomy between “weak sustainability” and “strong sustainability” (Daly, 1990) (see contribution I-6).
44According to one interpretation of weak sustainability, the destruction of part of the natural capital could thus be offset by investment in economic capital or intangible capital. In French Polynesia, the criterion of weak sustainability would be satisfied today through public transfers (high gross savings, limited degradation of natural capital and major investments in human capital). In contrast, advocates of strong sustainability consider that mining a non-renewable resource is a form a destruction of natural capital that cannot be offset by investment in other forms of capital. The notion that is central to the debate is whether it is possible to substitute one form of capital for another (critical natural capital), which itself depends on technical progress.
45In anticipation of the possible development of deep-sea mineral resources, the difficulty lies in assessing the damage to natural capital and possible harm to the functioning of the marine ecosystem. Current available knowledge on mineral resources and on the potential impact of mining them is still too limited to estimate the extent of that impact. However, it is known that polymetallic crusts form on deep-sea mounts over very long periods of time. The long lifespans of the organisms that are the architects of these habitats make these communities potentially less resilient to disturbances.
46Even taking a weak sustainability approach, the rent from mining is not necessarily sufficient to ensure transmission of the same level of overall wealth, if we do not take into account other sources of wealth, including public transfers29. This factor is also linked to the specific arrangements for macroeconomic funding in small island economies (SIE), which are highly dependent on access to various sources of revenue whose fluctuations raise questions of sustainability. For a small island economy, subject to strong economic constraints (transport costs and limited markets), the conditions for successful substitution between economic capital and natural capital – i.e. converting mining income into diversified, lasting economic growth – are all the more difficult to attain. It is necessary to target opportunities for efficient specialisation, which are severely limited in many small island economies.
47The conditions for achieving sustainability through investments in human capital, in a configuration marked by migration within several SIEs, depend on social capital, ensuring that migrants send a portion of their compensation back to their homeland.
48As Poirine (1994) shows, this type of income may not last, in which case migration would eventually cease to produce financial returns. In the case of French Polynesia, outbound migration flows are limited and do not constitute a significant source of revenue. A strategy for building human capital would then have mainly positive effects on revenue by increasing labour productivity in activities carried out in the territory. However, the range of economic activities is limited by the particular characteristics of island economies (remoteness and small size).
49Strategies for accumulating economic and human capital rely heavily on developments in social and cultural capital, i.e. perceptions of economic activity, and striking a balance between accumulation and redistribution, or between the needs of individuals and groups (Geronimi, 2015). Here again, we are faced with the issue of cultural representations and conditions for successfully developing mineral resources. Thus, we could consider that the World Bank weak sustainability approach30 leaves out a key aspect of small island economies: the heritage character of a fundamental part of their wealth. While we can indeed accept that not all heritage values can be expressed monetarily (particularly for the values of legacy and existence), a weak sustainability evaluation (genuine savings) rules out all cultural and heritage aspects, only some of which can be measured in monetary terms (Throsby, 2002). These cultural and heritage aspects are nevertheless sources of comparative benefits, and revenue, for a significant portion of the SIEs. Indeed, the evaluation of total wealth (natural capital plus physical capital) leaves over 70% of island economy revenue sources unaccounted for, labelled as “intangible capital” (Couharde et al., 2011). This covers all non-monetarily assessed values, including those relating to geostrategic position and historical relations, which form the basis for the majority of the revenue collected by French Polynesia today31 .
50Finally, a weak sustainability interpretation means threshold effects are ignored (Couharde et al. 2010); their inclusion is essential when assessing the sustainability of economic trajectories. Advocates of strong sustainability (Daly, 1990) also put forth the concept of critical natural capital, representing a threshold below which substitution with human or economic capital is no longer possible. In this regard, the “elastics model” presented by Giraud and Loyer (2006) has the advantage of highlighting the possibility that relations involving substitution between forms of capital (e.g. between natural capital and human capital) can leave room for complementary relations. In their example, environmental degradation led to large-scale migration (diminished natural capital resulting in a loss of human capital, with no possibility of substituting these two dimensions) and hence the collapse of sustainability. If we accept that cultural capital is the “glue” that holds the members of a community together, reflecting a common history and a collective accumulation of knowledge, creativity and values, then accounting for heritage and cultural aspects will lead favour strong sustainability, focusing on thresholds and complementarities between the different aspects of total wealth, especially for SIEs. It is important to assess the complementarities and thresholds that may emerge in French Polynesia in the context of exploiting deep-sea mineral resources.
51In French Polynesia, sustainability issues relating to the potential development of deep-sea mineral resources indicate the need for strategic choices in terms of development and growth trajectories.
Notes de bas de page
21 For further details, see the Glossary.
22 The recovery of other metals, representing less than 5% of the value of the ore, will of course be reviewed in future studies of the metallurgical processing of the crusts.
23 Consult the glossary for further details.
24 An IRR of 10% is considered the minimum needed to encourage private agents to invest in a project (IMF, World Bank). For mining projects, the average IRR is often much higher (around 30%), because of the specific risks posed by these projects.
25 To calculate NPV, a discount rate for future income from operations must be chosen. In the context of the simulations conducted here, we have chosen a rate of 10%, equivalent to a minimum level of profitability.
26 Consult the glossary for further details.
27 Note that the current price for cobalt (USD 27,740 per tonne on the LME as of 5 October 2015), is below the average level for the 1984-2013 period (USD 40,514 calculated in 2006 dollars). At this price, the mining project would not be profitable.
28 According to the baseline scenario approximately 1 million tonnes of polymetallic crusts need to be extracted to recover 4,500 tonnes of cobalt. In the favourable scenario, it would be necessary to extract about 2 million tonnes (for 10,000 tonnes of cobalt). Here again, these estimates are highly dependent on geological and technical parameters (density, recovery rates, etc.) that remain uncertain.
29 In reality, mining a mineral resource means degrading this resource; this must be offset if one wishes to maintain the same level of total wealth (economic, natural and human) in an economy. In New Caledonia, for example, investing in human capital and economic capital is expected to offset the destruction of natural capital (mining nickel ore), but this investment relies in part on public transfers. Without public transfers, sustainability (maintaining the same level of total wealth) would not be given.
30 This approach is based on the assumption that one can freely substitute investments in human or economic capital, or even the genuine savings indicator, for natural capital destroyed by extraction.
31 Note that public transfers to French Polynesia, which provide macroeconomic funding for the foreign trade deficit, are akin to compensation for geostrategic and historic capital.
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Substances naturelles en Polynésie française
Stratégies de valorisation
Jean Guezennec, Christian Moretti et Jean-Christophe Simon (dir.)
2006
L’énergie dans le développement de la Nouvelle-Calédonie
Yves Le Bars, Elsa Faugère, Philippe Menanteau et al. (dir.)
2010
La lutte antivectorielle en France
Didier Fontenille, Christophe Lagneau, Sylvie Lecollinet et al. (dir.)
2009
Le mercure en Amazonie
Rôle de l’homme et de l’environnement, risques sanitaires
Jean-Pierre Carmouze, Marc Lucotte et Alain Boudou (dir.)
2001
Diasporas scientifiques
Comment les pays en développement peuvent-ils tirer parti de leurs chercheurs et de leurs ingénieurs expatriés ?
Rémi Barré, Valeria Hernández, Jean-Baptiste Meyer et al. (dir.)
2003
La dengue dans les départements français d’Amérique
Comment optimiser la lutte contre cette maladie ?
Raymond Corriveau, Bernard Philippon et André Yébakima (dir.)
2003
Agriculture biologique en Martinique
Quelles perspectives de développement ?
Martine François, Roland Moreau et Bertil Sylvander (dir.)
2005
Lutte contre le trachome en Afrique subsaharienne
Anne-Marie Moulin, Jeanne Orfila, Doulaye Sacko et al. (dir.)
2006
Les espèces envahissantes dans l’archipel néo-calédonien
Un risque environnemental et économique majeur
Marie-Laure Beauvais, Alain Coléno et Hervé Jourdan (dir.)
2006
Les ressources minérales profondes en Polynésie française / Deep-sea mineral resources in French Polynesia
Pierre-Yves Le Meur, Pierre Cochonat, Carine David et al. (dir.)
2016
Le développement du lac Tchad / Development of Lake Chad
Situation actuelle et futurs possibles / Current Situation and Possible Outcomes
Jacques Lemoalle et Géraud Magrin (dir.)
2014
