EDF_REGISTRATION_DOCUMENT_2017

1.

PRESENTATION OF EDF GROUP Description of the Group's activities

Decommissioning costs EDF nuclear power plants

The estimate also includes, by caution, an assessment on risks and uncertainties. Third-party facilities: La Hague (AREVA-Orano) and Phénix (CEA) As the responsibility for the decommissioning of facilities is incumbent on their operator, EDF wished to free itself financially from these operations. As such, the agreements signed with AREVA-Orano in July 2010 and the CEA in late 2008 clarified the financial responsibilities of the parties. Following a cash payment, EDF was released from any obligation to finance the decommissioning of the Phénix facilities, which have been shut down, and the La Hague plant. Assets available to cover long-term 1.4.1.1.7 nuclear commitments (outside the operating cycle) Dedicated assets have been gradually established 1999 to cover long-term nuclear commitments (see section 6.1 “EDF’s “Consolidated financial statements”, note 47.2 “Content and evaluation of dedicated assets” to the consolidated financial statements for the year ended 31 December 2017). Article L. 594 of the French Environment Code and its implementing regulation specified which liabilities are not associated with the operating cycle and must therefore be covered by dedicated assets (see section 6.1 “Consolidated financial statements at 31 December 2017”, note 47.5 “Updated cost of long-term nuclear commitments” to the consolidated financial statements for the year ended 31 December 2017). New Nuclear projects 1.4.1.2 See also section 2.1.5 “Specific risks related to the Group’s nuclear activities – Construction of EPRs may encounter problems meeting the implementation schedule or the budgetary envelope or not be completed”. Organisation 1.4.1.2.1 EDF's DIPNN (New Nuclear Projects and Engineering Department) was reorganised in the second half of 2017 to clarify its engineering services. Today the DIPNN consists of three project departments (Flamanville 3, EPR 2 and Ingénierie Hinkley Point C), an engineering centre (CNEPE), an engineering firm (Edvance) and four operating departments overseeing development, industry, project support and digital transformation, and technology. Engineering activities relating to new nuclear islands fall under Edvance, a subsidiary of EDF and Framatome, under the supervision of the DIPNN. These entities provide consulting services, project ownership engineering support services and project management engineering services to the Group and its partners in the nuclear industry. Restructuring of the DIPNN is continuing in the interests of opening up to industrial partners and is speeding up in the area of digitization with a second restructuring cycle following on from the first. Update on the Flamanville EPR project 1.4.1.2.2 EDF SA is both the owner and manager of the Flamanville 3 EPR project (European Pressurized water Reactor) on the basis of its in-house engineering expertise. Commissioning schedule and budget The schedule announced in September 2015 and the interim dates announced in October 2017 were adhered to. The next milestones are: "hot functional” testing: July 2018 target. It consists of testing equipment under ■ similar temperature and pressure (300°C & 150 bars) as in operating conditions; fuel loading: target set at the end of the fourth quarter of 2018, then start-up of ■ the reactor. The connection of the Flamanville 3 EPR to the grid is then scheduled for the second quarter of 2019, and generation at full capacity, after a gradual ramping-up phase, for the fourth quarter of 2019.

Since the beginning of operations at its power plants, EDF has made provisions to cover decommissioning operations, engineering, monitoring and maintenance of facilities, and site security (see section 6.1 “Consolidated financial statements at 31 December 2017”, notes 29.1.3 and 29.1.5 to the consolidated financial statements for the year ended 31 December 2017). The aim of decommissioning is to return sites and land to a state suitable for industrial use. EDF continues to reinforce its analyses through international intercomparisons while making sure to take account of certain elements that could distort direct comparisons such as differences in the estimate scopes or national and regulatory contexts. The cost estimate for the dismantling of second-generation power plants (Gen II – PWR plants in operation) was revised in 2016 to take into account both the recommendations of the audit mandated by the DGEC (the French government's department for energy and climate change) on the decommissioning costs of PWRs, conducted between July 2014 and August 2015 on the basis of the “DA09” model, and the feedback gathered from the decommissioning of first-generation power plants (Gen I – in particular the Chooz A plant). Reviewing these cost estimates consisted in implementing a detailed analytical process, identifying all the engineering, work, operation and waste treatment costs linked to future decommissioning of units currently operated. It resulted in figures based on detailed plant decommissioning feedbacks. This implemented process allowed to deepen the assessment of the costs specific to leading models and of the series and pooling impacts, those costs and impacts being indeed inherent to the size and the design of the fleet. The financial impact is described in note 29.1 to the consolidated financial statements at 31 December 2017. The nature of main pooling and series effects selected when calculating the estimates are presented hereafter. There are different types of pooling effects: some of them are linked to the affectation of common buildings and facilities to ■ the operation of several reactors on a same location, and therefore these buildings and facilities will not have to be dismantled twice. Thus, structurally, the dismantling of a couple of reactors on a same location costs less than the dismantling of two single reactors on two different sites. In France, unlike in the other countries, there is no isolated reactor but locations with two, four and in one case six reactors; some costs are not higher if you dismantle 2 or 4 reactors on a same site. It is ■ usually the case for surveillance costs and costs incurred by keeping the site in safe operating conditions; waste processing in centralised facilities (for example for the cutting of major ■ components) is cheaper than multiple processing facilities on the dismantling sites. There are mainly two types of series effects: a first effect comes from the fact that on a fleet driven by a single technology, a ■ large amount of the studies does not need to be performed again each time; a second effect comes from the fact that on a fleet driven by a single technology, ■ robots and tooling can be largely reused from one site to another. Such series effects have the same nature than those observed during the construction of the fleet, in terms of studies or of components manufacturing plants. As an example, for the 900MW fleet, a series effect of approximately 20% is expected on an average 2 units reactor in comparison to a 2 units leading model. Due to series and pooling series effects in particular, a simple comparison of the average dismantling cost by reactor between the French fleet and other countries’ nuclear fleet is not relevant. However, figures include only very marginally the evolution of productivity and series effect. External audit conducted by the DGEC on the dismantling cost of the operating fleet had considered that series effect taken into account in the estimate was conservative.

28

DF I Reference Document 2017