Energy Policy 101 (2017) 683–691

Contents lists available at ScienceDirect

Energy Policy

journal homepage: www.elsevier.com/locate/enpol

China's nuclear export drive: Trojan Horse or Marshall Plan?$

Steve Thomas

Professor of Energy Policy, Public Services International Research Unit (PSIRU), Business School, University of Greenwich, 30 Park Row, London SE10 9LS, UK

HIGHLIGHTS

's nuclear industry expanded fast in the last decade and is targeting exports. Success in Europe would be a major boost to prospects elsewhere. National security concerns must be examined before investment decisions are taken. China's quality control and regulatory competence should be assessed. China's modern reactor designs are untested even in China. article info abstract

Article history: China's civil nuclear industry expanded strongly from 2008 onwards and nearly half of reactor construction Received 28 January 2016 starts worldwide since then are accounted for by the Chinese home market. Increasingly China is turning its Received in revised form attention to the export market using its own designs, which it claims emulate the safety standards of the 12 September 2016 latest designs of the established vendors. Its export efforts would be greatly strengthened if it Accepted 14 September 2016 were to win an order from an established user of and its best opportunity appears to be the Available online 29 September 2016 UK where it is at the early stages of negotiating the construction of nuclear reactors. The financial collapse of Keywords: the French nuclear company, Areva, gives it the opportunity to take a stake in the rescued companies giving it China access to important fuel cycle technologies and perhaps the large French reactor service market. Its other Nuclear power export prospects in Europe are in Romania and Turkey. There are a number of issues European governments Exports need to examine before committing to allow in Chinese nuclear companies. These include national security Europe National security concerns about dependence on China for key infrastructure, issues of quality control and regulatory com- petence and the lack of construction experience with China's modern reactor designs. & 2016 Elsevier Ltd. All rights reserved.

1. Introduction all other nuclear exporting countries, except Russia, through its ownership of the companies and its ability to marshal all aspects of After a period of rapid expansion of nuclear capacity in China in the economy especially the financial sector to support its industries. the past decade, China is looking to use its expertise to spear-head an To be successful in this export drive, it needs credibility and for export drive for its nuclear reactors. It has a particular advantage over many potential markets the ability to provide low-cost finance. The prestige gained from exporting to a country with a strong existing nuclear capability would be valuable. Exports to USA, Ja- ☆On September 15, 2016, the British government said that it had completed its pan and Russia appear implausible in the medium-term so Europe review of the Hinkley Point project announced in July 2016 and had decided that and, especially the UK, is a key market to crack. Importing reactors the project should go ahead. The result of the review was that the UK government would take a ‘special share’ (usually known as a golden share) in nuclear projects from China raises policy issues that must be addressed before any after Hinkley Point (it was said to be too late to require this for Hinkley Point). This commitment is given. These include national security concerns, would give them right of veto over changes in ownership of nuclear plants. The the safety of the designs and quality control of the components. Bradwell project would be covered by this policy but was not mentioned in the In this paper, we set out the history of the civil nuclear power announcement. A golden share appears irrelevant to the reported concerns behind the review of industrial espionage, leakage of military technologies and Chinese programme in China, we review the main nuclear companies in- control over major infrastructure. It seems unlikely that this requirement will be a cluding their technologies, their ordering history, their target major barrier to Chinese investment in the UK nuclear power sector. However, markets and their strategic alliances with European, Japanese and there remain several major hurdles to jump before construction of Hinkley can proceed. US nuclear companies. We then review the potential scale of the E-mail address: [email protected] European market and the strategies these companies are pursuing http://dx.doi.org/10.1016/j.enpol.2016.09.038 0301-4215/& 2016 Elsevier Ltd. All rights reserved. 684 S. Thomas / Energy Policy 101 (2017) 683–691 in the European markets and finally we examine the main policy Table 1 issues raised by China's attempts to export its China's nuclear power plants in service. to Europe. We pay particular attention to the UK and France as Source: IAEA PRIS database https://www.iaea.org/PRIS/home.aspx these are the markets which would carry most prestige and where Technology Supplier No of Capacity Construction First power efforts for Chinese companies to enter are most advanced. units (MW net) start

CNP-300 CNNC 1 288 1985 1991 M310 / 4 3760 19871997 19932003 2. Literature review CGN CNP-600 CNNC 6 3160 19962010 20022016 There are major problems in writing about China's nuclear in- CANDU AECL 2 1354 1998 20022003 dustry because of the lack of up to date independent analysis. (Canada) Most of the detailed articles in journals and books, for example, AES-91 Rosatom 2 1980 1999 2000 2006 2007 (Russia) Ramana and Saikawa (2011), Zhou et al. (2011), Zhou and Zhang CPR-1000 CGN 15 15229 20052010 20102016 (2010) and Yi-chong (2010) were written before the Fukushima CPR-1000 CNNC 5 6000 20082010 20142016 disaster, which marked a major policy change in China and do not Total 35 31771 reflect the major technological and corporate changes to the Chi- nese nuclear industry that have occurred from 2013 onwards. King Table 2 and Ramana (2015) provides useful and up to date analysis. This China's nuclear power plants under construction. article draws heavily on the trade press, especially nuclear news- Source: IAEA PRIS database https://www.iaea.org/PRIS/home.aspx letters with correspondents based in China. China's nuclear export ambitions mean that senior officials in government and companies Technology Supplier No of Capacity (MW Construction start feel the need to provide reliable and honest answers to questions units net) from such correspondents because the audience for such pub- AP1000 West/SPIC 4 4000 20092010 lications includes their potential customers. EPR Areva/CGN 2 3320 20092010 CPR-1000 CNNC 3 3000 201216 CPR-1000 CGN 1 1000 2012 HTR-PM Huaneng 1 200 2012 3. China's civil nuclear programme and its industry structure ACPR-1000 CGN 4 4000 20132015 HPR-1000 CNNC 2 2000 2015 3.1. History of China's nuclear power programme HPR-1000 CGN 1 1000 2015 Total 18 18520 China carried out its first nuclear weapons test in 1964, but its exploitation of civil nuclear power came later.1 Tables 1–3 Table 3 summarise China's reactor programme. It was not until 1985 China's next new-build projects. that construction work began on the first reactor, a small Source: Nuclear Intelligence Weekly ‘Nine projects top priority list’ May 16, 2016, p 5 (300 MW) indigenous design of Pressurised Water Reactor (PWR)2 designed by the Shanghai Research Site/province Capacity Design Owner Approval expected and Design Institute (SNERDI), which was established in 1970. This was supplied by China National Nuclear Corporation Rongcheng/ 2 1500 CAP1400 SPI 2016 (CNNC), a long established company with expertise in weapons andsubmarinesaswellaspowerreactors.3 In 1978, China Haiyang/Shandong 2 1250 AP1000 SPI 2016/17 reached agreement with France to buy two reactors for the Daya Sanmen/ 2 1250 AP1000 CNNC 2016/17 Lufeng/ 2 1250 AP1000 CGN 2016/17 Bay site but it was not till 1987 that their construction started. Xudapu/ 2 1250 AP1000 CNNC/Datang 2016/17 The reactors, each with an output of about 950 MW, were sup- Ningde/ 2 1150 HPR-1000 CGN/Datang 2016/17 plied by Framatome (renamed Areva NP in 2002) using its M310 Zhangzhou/Fujian 2 1087 HPR-1000 CNNC/ 2017/18 design,withconstructionmanagedbytheFrenchnationaluti- Guodian Changjiang/ 2 1087 HPR-1000 CNNC/ 2018 lity, Electricité de France (EDF). The Chinese partners became a Huaneng 4 new state-owned company, China Guangdong Nuclear (CGN) in Haixing/ 2 1087 ? CNNC 2018 1994. Reactor ordering was slow in the period 1987–2007. CNNC produced a larger indigenous PWR design, the CNP-600, first beginning construction (four CGN and two CNNC) based on the construction start in 1996, and imported two heavy water reactors CPR-1000 design. In 2009 and 2010, ten more reactors of this of Canadian design in 1998. Two more imported French reactors design started construction. were built in 1997 in collaboration with CGN. However, both CNNC The M310 design dates back to the 1960s, having been li- and CGN were building up their design expertise and in 2005, censed by Framatome in the early 1970s from Westinghouse, construction was started on a Chinese version of the M310, the and by 2006, the Chinese authorities acknowledged more CPR-1000.5 In 2008, construction work took off with six reactors modern designs were needed. Its plan was to select one of the advanced designs on offer importing a few reactors initially but progressively transferring the technology so it could be supplied 1 Data on construction and operation of reactors are all taken from the Inter- by Chinese companies. Two technologies were seriously con- national Atomic Energy Agency's PRIS data base. https://www.iaea.org/PRIS/home. aspx (accessed 12.08.16.). sidered, the EPR supplied by Areva and the AP1000 supplied by 6 2 The PWR is the most widely used type of reactor worldwide accounting for Westinghouse, which was owned by . In 2007 the about two thirds of the world's operating reactors. 3 http://www.cnnc.com.cn/tabid/643/Default.aspx (accessed 15.12.15.). 4 http://www.cgnpc.com.cn/n1500/index.html (accessed 15.12.15.). 6 Westinghouse's nuclear division was sold to the state-owned UK company, 5 Both the CGN and CNNC versions of the M310 are designated CPR-1000 al- BNFL in 1999 for $1.1bn who in turn sold it to Toshiba in 2006 for $5.4bn. http:// though there are differences between the two. The CNNC version of the M310 is www.toshiba.co.uk/innovation/NEWSARCHIVE/archived_news_article.jsp? also sometimes known as the M310þ. ID¼0000006709 (accessed 14.07.16.). S. Thomas / Energy Policy 101 (2017) 683–691 685

AP1000 was chosen with four reactors ordered and a new disaster, the Chinese government prohibited construction of re- company created, State Nuclear Power Technology Company7 actors at inland sites. The main concerns were the lack of sufficient (SNPTC), re-named State Power Investment Corporation (SPIC) water to mitigate the effects of a serious accident and the devas- in 2015 (see below), to indigenise the technology. SNERDI be- tation to land and to inland waterways an accident would cause came a subsidiary of SNPTC giving it established expertise. In compared to a coastal site where much of the radiation would land 2008 two orders were placed for EPRs with CGN partnering in the sea. In 2015, the ban on developing coastal sites was under Areva and EDF for this project. review (Yu, 2015a) but by August 2016, no decision had been taken. The Chinese vendors began to produce their own advanced It had become clear soon after construction start that the EPR designs using imported technology from their existing partners as and the AP1000 were probably too expensive to form the basis of the basis: CGN developing the ACPR-1000, CNNC the ACP-1000 Chinese orders. In 2013, the Chinese government required CNNC and SPIC the CAP1400. The ACPR-1000 and the ACP-1000 are de- and CGN to ‘merge’ their advanced designs (ACP-1000 and ACPR- velopments of the French M310 design while the CAP1400 is a 1000) to produce the design (HPR-1000). Four re- scaled up AP1000. The technology license agreement for the actors of the CGN ACPR-1000 started construction and CNNC an- AP1000 is reported to stipulate that any derivatives of the AP1000 nounced it had won an order for two reactors using the ACP-1000 with output larger than 1350 MW would be SPIC's intellectual design for , although the design is now generally de- property.8 scribed as HPR-1000. Construction work started on this in August The ordering spurt from 2008 was achieved with relatively 2015. Even though they were required to merge their designs to little public debate and participation (Guizhen et al., 2014). This form a single one, it is now clear that CGN and CNNC have their lack of scope for public participation should not be seen as a lack own versions of the HPR-1000 with their own supply chain. It of public opposition. For example, in July 2016, the government of appears the two versions of HPR-1000 are little more than re- Lianyungang, a city near the coast of Province, said that it named versions of the ACP-1000 and the ACPR-1000 rather than had halted any plans to build a reprocessing plant any significant attempt to merge the designs and there remain 9 there after days of strong public protest. significant differences between the fuel assemblies and between The Fukushima disaster coincided with a dramatic slowing of the safety systems (Yu, 2016). ordering with construction starts on only four reactors in the next In December 2015, CNNC and CGN announced the creation of a four years. How far this was due to a reassessment of technology joint venture, Hualong International Nuclear Power Technology Co as a result of Fukushima and how far it was down to other factors, (Yu, 2016) This would be: “a powerful force to the integration and such as overstretching of resources is difficult to tell. In 2015, development and market expansion of HPR 1000 and boost the construction started on six reactors all of advanced design but it landing in more countries and regions of the technology [sic]”.Itis fi was not until September 2016, that the rst construction start on a not clear how far this move will lead to the merger of the two new reactor took place that year. separate HPR-1000 capabilities or whether it will simply be a In 2012, China set itself a target to have 58 GW of nuclear ca- figurehead company with CGN and CNNC continuing to pursue pacity in operation by 2020 with a further 30 GW under their own separate markets. Attempts to decide which version 10 construction. This target does not appear achievable given that would be the one chosen for a unified design were stalled in Au- in August 2016, it had only 30.4 GW of nuclear capacity in op- gust 2016.13 eration with 20.5 GW under construction and constructions starts China has also been developing high temperature gas-cooled appeared to have largely stalled again with only 17 GW of capacity reactors since the mid-80s via a technology license for German short-listed for construction start by 2018 (see Table 3). A potential pebble bed technology. This work has been led by Tsinghua Uni- barrier to expansion of nuclear capacity beyond 2020 is that the versity which completed a 10 MW prototype plant (HTR-10) in Chinese State Council has approved only CGN, CNNC and SPIC as 2000. It set up a joint venture with China Nuclear Engineering nuclear developers. The vast majority of generating capacity in Corporation (CNEC) and Chinergy, and is building twin 105 MW fi 11 China is owned by ve other large generation companies but reactors at the Shidao Bay site, which in 2016 were expected to be they were allowed only to take a minority stake in nuclear completed in late 2017. Whether this will result in a commercially 12 projects. These rules were under review in 2015. In 2015, SNPTC viable technology remains to be seen. fi merged with the smallest of the big ve Chinese generation By end 2013, the three Chinese reactor vendors had target ex- companies, China Power Investment Corporation (CPIC) to form port markets: CGN was competing in UK, Romania and Kenya; fi the State Power Investment Corporation (SPIC) increasing its - CNNC was competing in Argentina, Algeria and Sudan; and SPIC nancial strength for exports and giving it a potentially larger home was competing in Turkey and South Africa. This export drive is market. Despite several announcements forecasting the imminent backed by Chinese financial institutions. The China Development fi construction start for the rst CAP1400, by August 2016, this had Bank (CDB) and the Export and Import Bank of China are sup- not happened. porting state-backed companies, with CDB offering government- Another barrier to expansion is that all the orders by September to-government low interest loans to Argentina and Algeria for 2016 and all the short-listed projects were for coastal sites. If nu- their nuclear programmes as well as loans to CGN for the UK's clear power is to expand dramatically in China, inland sites will Hinkley Point project. The Industrial and Commercial Bank of have to be developed and there is significant opposition to this China has agreed to offer loans of €10 billion to support CGN's happening (King and Ramana, 2015). Following the Fukushima nuclear project in Romania (Yu, 2015b).

7 http://www.snptc.com.cn/en/(accessed 18.12.15.). 3.2. The Chinese nuclear companies 8 Nuclear Engineering International ‘Power plant design - China - Develop- ’ ments in Chinese reactor technology. March 2016, p 30. 3.2.1. CNNC 9 International New York Times ‘Chinese city withdraws nuclear plan after protests’ August 12, 2016. The Chinese National Nuclear Corporation is the longest es- 10 Nuclear Engineering International ‘China – News update - Steady and safe’ tablished and most broadly based of the Chinese nuclear December 2012. 11 State Power Investment Company, Huaneng, Datan, Huadian and Guodian. 12 For more details, see Nuclear Intelligence Weekly ‘Beijing Mulls Widening 13 Nuclear Intelligence Weekly ‘CNNC and CGN Stalemated Over Hualong-One Circle of Majority NPP Owners’ December 12, 2015, p 3. Design’ August 12, 2016, p 6. 686 S. Thomas / Energy Policy 101 (2017) 683–691

Table 4 3.2.3. SPIC Sales of CNNC reactors.17 The State Nuclear Power Technology Company (SNPTC) was Source: Author's research created in 2007 incorporating the long-established design ex-

Design Units sold Output MW Construction Commercial Partner pertise of SNERDI to be the partner to Toshiba for indigenising start power Toshiba's AP1000 technology. It merged with a large utility, China Power Investment Corporation (CPIC) to form the State Power – CNP-300 5 310 340 1985 2011 1994 - Investment Corporation (SPIC) in 2015.19 CNP-600 6 650660 19962010 2002 - – CANDU-6 2 728 1998 20022003 AECL When the AP1000 was chosen in 2007, it was forecast that the CPR-1000 8 1080 20082016 2014 - – AP1000 would quickly take over from the CPR-1000 as the basis HPR-1000 2 1087 2015 –– for orders for China.20 Like the Taishan project, the AP1000 pro- ACP-1000 1 1080 2015 –– jects have gone badly and are 3–4 years late (see Table 6) but many of the problems seem to be caused by Toshiba. However, experi- companies. It was originally the Second Ministry of Machine ence with construction of the has led to serious doubts Building, then the Ministry of Nuclear Industry and reorganised to that this design could be built cheaply enough to form the basis of Chinese ordering. Despite repeated claims that construction would become CNNC in 1988. It makes no secret of its military roots and start soon on the first CAP-1400, by August 2016, this had not continued military connections. Its web-site states14: “Historically, happened leaving SPIC well behind CGN and CNNC with its ad- CNNC successfully developed the atomic bomb, hydrogen bomb vanced reactor design. Of the nine reactor projects shortlisted for and nuclear submarines and built the first nuclear plant in the development from 2016 to 2018, five are for AP1000s and only one main land of China. CNNC is the main body of the national nuclear is for CAP1400s (see Table 3). In May 2016, an official of SPIC said: technology industry, the core of the national strategic nuclear “It's no longer clear” which design will be selected for “the future” deterrence and the main force of the national nuclear power de- of the Chinese fleet and a decision over whether the CAP1400 will velopment and nuclear power construction, and shoulders the be exclusively for exports also “is in flux.”21 Export customers duel [sic] historical responsibilities for building of national defence would be concerned if there were no home orders about buying a force, increasing the value of state assets and developing the technology not demonstrated in its home market. It would gain society.”. little strategically from CGN winning a reactor order in the UK Despite this strong historic position, CNNC seemed to lose out other than that it would demonstrate that a Chinese company to its rival, CGN, in the burst of orders from 2008 to 2010 could satisfy an experienced and rigorous regulatory body. (See Table 4). It only supplied six of the 22 CPR-1000 reactors, CGN was the chosen partner for Areva for the Taishan project and by end 2015, CNNC had started construction on only three reactors 4. China's nuclear export strategy using its advanced designs compared to five for CGN. Of the nine reactor projects (each for two reactors) shortlisted by China for Despite its limited experience, China has a number of im- development from 2016 to 18, only three projects are for HPR- portant expected advantages over the established vendors based 22 1000s, two of which are expected to be supplied by CNNC (see in Japan, France and Russia : Table 3).15 In May 2016, CGN stated that it had agreed not to compete with CNNC for business in export markets with CGN fo- 1. There was an assumption that Chinese reactors could meet cusing on Europe.16 standards required by experienced, developed country markets at lower cost than its competitors; 3.2.2. CGN 2. The high rate of ordering for its home market gave it scale China Guangdong Nuclear was formed in 1994 and renamed economies and a large skilled workforce that its competitors could not rival; China General Nuclear in 2013. This allowed it to reflect its as- 3. The huge financial reserves of the Chinese government meant pirations as a global company whilst retaining its established ac- that its vendors would be able to provide finance as well as ronym. As argued above, by 2015 it seems to have achieved a equipment, a big advantage in most potential markets; somewhat stronger position than CNNC (see Table 5) despite its 4. Its competitors, Areva and Toshiba/Westinghouse, had serious more recent origins and as a provincial rather than as a national financial problems while international sanctions and the col- 18 company. How far this position is down to superior technical lapse of the Rouble meant that Russia was unlikely to be able to expertise, its early experience with the Daya Bay project and the provide the financial support it had been expected to give its location of many of the nuclear projects in regions of China, the vendor, Rosatom. South East, where CGN is strong is difficult to determine. However, of the nine reactor projects (each for two reactors) shortlisted for The first and second assumptions are hard to test because of development from 2016 to 18, only one is for CGN HPR-1000s (see the difficulty of getting reliable cost data from China, the lack of Table 3). The Taishan project, in which it was the Chinese partner experience with modern designs and the lack of much operating to Areva, to build two EPRs has gone badly and by 2016, it was at experience. In Appendix 1, we review the available data least three years late. In August 2016, China was forecasting first (see Tables 7 and 8). We conclude that the record of China in terms power from unit 1 in early 2017 and in unit 2 by end 2017. of construction and operation is good if not outstanding. The third advantage relies on the existence of a centralised

14 http://www.cnnc.com.cn/tabid/643/Default.aspx (accessed 18.12.15.). 15 Nuclear Intelligence Weekly ‘Nine projects top priority list’ May 16, 2016, p 5. 19 http://www.snptc.com.cn/en/index.php?optionid¼911 (accessed 19.08.16.). 16 Nuclear Engineering International ‘China's CGN and CNNC agree not to 20 http://www.snptc.com.cn/en/(accessed 18.12.15.). compete’ June 2016. http://www.neimagazine.com/news/newschinas-cgn-and- 21 Nuclear Intelligence Weekly ‘Weekly Round-up’ May 20, 2016, p 1. cnnc-agree-not-to-compete-4907664 (accessed 28.08.16.). 22 The two longest established reactor vendors were originally US companies, 17 Tables 1–3 only include reactors on which construction had started by Jan- Westinghouse and GE, but the Westinghouse reactor division is now owned by uary 2016 Toshiba (Japan), while GE's reactor business outside the USA is run by the Hitachi- 18 http://www.cgnpc.com.cn/n1500/index.html (accessed 15.12.15.). GE joint venture which is 80% controlled by Hitachi (Japan). S. Thomas / Energy Policy 101 (2017) 683–691 687

Table 5 Sales of CGN reactors. Source: Author's research

Design Units sold Output MW Construction start Commercial power Partner

M310 4 984 19871997 19942003 Framatome CPR-1000 16 1080 20052012 2010 - – EPR 2 1750 20092010 – Areva ACPR-1000 4 1087 20132015 –– HPR-1000 1 1087 2015 ––

Table 6 Table 8 Sales of SPI reactors. Construction time and operating performance of CNP-1000 reactors by year. Source: Author's research Source: IAEA PRIS database. https://www.iaea.org/PRIS/home.aspx (accessed 15.08.16.) Design Units sold Output MW Construction Commercial Partner start power Year construction No of Mean construction Lifetime load start units time (months) factor (%) AP1000 4 1250 20092010 – Toshiba 2005 1 57 84.1 (1) 2006 1 62 87.5 (1) 2007 1 70 79.2 (1) Table 7 2008 6 68 88.3 (3) – Construction time and operating performance by vendor and model. 2009 5 75 – Source: IAEA PRIS database. https://www.iaea.org/PRIS/home.aspx (accessed 2010 6 67 15.08.16.)

Vendor/model No of Mean construction Lifetime load fac- weak state of its economy means it is unlikely it will have the units time (months) tor (%) (no of financial capability to support more than a small fraction of these units) orders. Most OECD countries would not try to exert the degree of CNNC/CNP-300 1 109 82.3 (1) control China has. France is often seen as a country with a com- CNNC/CNP-600 6 56 87.5 (4) mitment to mobilise all French resources to support its nuclear Fram/M310 4 68 86.7 (4) industry. It controls the utility, EDF, the reactor vendor and fuel CNNC/CNP-1000 5 71 – CGN/CNP-1000 15 68 83.4 (6) cycle company, Areva and has, in the past granted export credit CANDU 6 2 55 91.5 (2) guarantees on a small scale for reactor exports. However, the fi- Russia V-428 2 87 87.2 (2) nancial weakness of Areva and EDF and financial measures France AP1000 4 96 – EPR 2 96 – had to impose following the 2008 world financial crisis mean it ‘ ’ Other post-2000 does not have the scope to offer the one-stop shop capability for programmes reactor orders that China appears capable of providing. Japan was Korea 6 66 78.7 (4) considering a more coordinated approach to reactor exports in India 6 80 74.5 (6) Russia 5 100 – 2010 offering export credit guarantees and coordinating its ven- dors but the Fukushima disaster may mean this will not happen. Notes By 2015, Chinese vendors were competing in markets in almost every continent. Historically, sales in countries with relatively weak 1. For China, includes all reactors in service by August 2016 except for AP1000 and economies are seldom realised, in part because of the difficulty of EPR. Construction time for AP1000 and EPR are based on the estimates in Au- fi gust 2016. raising the huge amount of nance needed. To put the scale of in- vestment in perspective, it is worth noting the plan to build just two 2. Lifetime load factors include only reactors that have completed at least two Areva EPRs in the UK would make it the most expensive civil pro- calendar years of commercial operation. ject ever carried out in Britain.23 China's prospects, especially those of CGN, would be considerably enhanced by the prestige winning 3. For Korea, India and Russia, Table 4 includes reactors on which construction started in or after 2000. Only one of the five reactors from Russia was in service an order in a large developed country in Europe with a long history by August 2016 and the construction times are based on the estimates in August of nuclear power exploitation would give. The most likely such 2016. market is the UK, where, in 2015, one of its three reactor vendors, CGN, announced preliminary agreement to build a nuclear power fi 4. Construction time is calculated as time from pouring rst structural concrete to plant in the UK. Elsewhere in Europe, Chinese reactor vendors are commercial operation competing for business in Romania (CGN) and Turkey (SPIC). The fi 5. Load factor is calculated at the output of the plant in kWh as a percentage of the nancial collapse and the proposed break-up of the French nuclear output it would have produced had it operated uninterrupted at full power group, Areva, gives Chinese companies the opportunity to take output. equity stakes in the resulting companies with long established ex- pertise in areas China would like to build up, such as reactor design, planned economy in which the government has strong control enrichment and spent fuel reprocessing. over all aspects of the economy and with a high degree of public ownership. Of the major economies, only China and Russia have 23 The Telegraph ‘Hinkley: a truly major national scandal’ September 26, 2015. this capability. Russia has ambitious nuclear export plans with http://www.telegraph.co.uk/comment/11893698/Hinkley-a-truly-major-national- more than 30 claimed orders for new nuclear reactors but the scandal.html (accessed 12.01.16.). 688 S. Thomas / Energy Policy 101 (2017) 683–691

4.1. Opportunities in Europe produce a complete, detailed design and the ONR says it expects to receive the design in 2016 for review. However, the HPR-1000 design In 2016, there were about ten countries in Europe seriously con- still seems to be evolving as CGN's and CNNC's designs are merged so sidering new reactor orders (for a review of European market pro- it may take a couple of years before the design is stable and detailed spects, see Schneider and Froggatt (2016). However, of these, six are in enough to be submitted. The project is therefore at a very early stage Eastern Europe, (Czech Republic, Slovakia, Hungary, Romania, Bulgaria and the CGN press release does not specify any target dates or even and Poland) a region where orders have long been discussed but have how many reactors are planned for Bradwell. In May 2016, an EDF consistently been postponed, usually due to problems of finance. Of spokesman said it was “very early days” in the development of the these, only Poland has a large enough electrical system to be able to proposed power plant29 and by August 2016, the HPR-1000 design accommodate more than one or two reactors. There appears to be the had not been submitted to ONR. There were also serious doubts as to political will for new reactor orders in France and Finland but nuclear whether the Hinkley Point project would proceed amongst other power already meets such a large proportion of electricity demand in things because of the financial collapse of Areva and opposition to the both countries, there is little scope for further orders. Only in the UK deal within EDF to the deal because of the financial strain it would put and Turkey is there both political backing and scope for a significant on EDF (Thomas, 2016). The completion date of the Hinkley project number of reactors. has continually slipped. In 2010, EDF was forecasting first power in Until the Chinese companies had designs that they could claim 2017, but by 2016, first power was not expected before 2026. were their own Intellectual Property (IP), they could not enter the After six years of claiming a Final Investment Decision (FID) for export market because the owners of the IP, Toshiba/Westinghouse Hinkley was imminent, EDF took an FID on July 2830 with the and Areva, were not willing to allow China to export these designs. expectation that an Investment Agreement with the UK govern- The market for China's small designs, CN-300 and CNP-600 is minimal. ment, committing the UK to the project would be signed the fol- The Chinese vendors claim they own the IP for their new designs, lowing day. However, only hours after the EDF Board took the FID, HPR-1000 and CAP1400, which became available from about 2013 the UK government unexpectedly announced that the project was onwards allowing China to enter the export market. There are few under review, at the instigation of the newly appointed UK Prime credible markets for nuclear power in the world so they are not in a Minister, Theresa May. The government was not explicit about the position to pick and choose which markets they compete in. The three rationale for the review, but press reports centred on concerns companies do not compete against each other in the same export about national security.31 If the Hinkley deal does collapse, the market and their export activities are coordinated, but not allocated by future of the whole UK nuclear programme including Bradwell the China Atomic Energy Authority (CAEA) and the National Devel- will be in doubt. The GDA process for any reactor design cannot opment and Reform Commission. start without explicit UK government approval for a reactor ven- dor to submit its design to ONR so if the UK government decides 4.1.1. The UK not to go ahead with Hinkley it might not approve the submission The most strategically important opportunity in Europe is the of the HPR-1000. However, if CGN and CNNC could get a Design proposal to build HPR-1000 technology at the Bradwell site. In October Acceptance Certificate from the ONR for a unified HPR-1000 de- 2013, a deal was announced by EDF to build two EPRs at the Hinkley sign, even if Bradwell does not proceed, this would be a huge Point site and potentially two further EPRs at the Sizewell site.24 The advantage in markets with less experienced regulatory bodies. consortium to own the plants was not finalised then but Chinese in- volvement was confirmed with CGN and CNNC expected to take up to 4.1.2. France 40 per cent of the company between them. In October 2015, more Areva was created in 2002, primarily from the reactor vendor details and some modifications to the deal were announced with Framatome and the nuclear fuel cycle company, Cogema and these CNNC no longer mentioned and CGN's stake 33.5 per cent.25 Subse- two companies formed the main divisions of Areva, Areva NP and quently, CNNC said it still expected to be a partner with CGN in the UK Areva NC respectively. It is about 87 per cent owned by the French butitwasimplieditwouldprovideonlyfinance and was leaving state. In March 2015, Areva announced losses for the fifth con- negotiations to CGN.26 The Chinese stake in the Sizewell project was secutive year, this time of €4.8bn and it was clear the group could reduced to only 20 per cent but a new element was that EDF would not continue without major restructuring. The French government release its Bradwell site to CGN for the construction of its HPR-1000 has been trying to broker a rescue which would involve essentially technology with it taking 66.5 per cent of the consortium and EDF the splitting it up into its constituent parts. Areva NP is responsible for rest.27 The first step will be for CGN to submit its HPR-1000 design to most of the losses and there are also huge potential liabilities, the UK safety regulator, the Office of Nuclear Regulation (ONR), to perhaps in excess of €10bn, from the four EPR reactors under undergo Generic Design Assessment (GDA). This is an exhaustive construction, two in China, one in Finland and one in France. Given process taking four years or more aimed at ensuring all generic design that Areva NP is valued at less than €3bn, any rescue plan will issues are resolved for any project, leaving only site-specificissuesfor inevitably require new investors to be insulated from these li- individual projects.28 Before this process can start, CGN will have to abilities, which will have to be borne by French taxpayers. Nevertheless, the opportunity to take an equity stake in the res- cued companies represents a major opportunity for the Chinese 24 https://www.edf.fr/sites/default/files/contrib/groupe-edf/espaces-dedies/ espace-finance-en/investors-analysts/events/special-announcements/agreement_ companies to gain expertise, technologies and access to markets. reached_on_commercial_terms_for_the_planned_hinkley_point_c_nuclear_power_ The French government is requiring EDF (about 85 per cent station.pdf (accessed 21.12.15.). 25 http://www.edfenergy.com/energy/nuclear-new-build-projects/hinkley- point-c/news-views/agreements-in-place (accessed 21.12.15.). 29 In January 2016, a spokesman for CGN said the design would be submitted to 26 The Telegraph ‘Second Chinese company poised to invest in Hinkley Point’ the UK regulator in June 2016. Nucleonics Week ‘Bradwell B remains at early stage: May 7, 2016 http://www.telegraph.co.uk/business/2016/05/07/second-chinese- EDF, regulator’ May 12, 2106, p 1. company-poised-to-invest-in-hinkley-point/(accessed 20.07.16.). 30 https://www.edfenergy.com/energy/nuclear-new-build-projects/hinkley- 27 http://en.cgnpc.com.cn/n1017152/n1017227/c1141640/content.html (ac- point-c/news-views/FID (accessed 19.08.16.). cessed 21.12.15.). 31 May's joint chief of staff, Nick Timothy had written in 2015: Rational con- 28 The EPR received generic design approval in 2012 after five years of review. cerns about national security are being swept to one side because of the desperate The AP1000 was submitted in 2007. The process was interrupted by the vendor in desire for Chinese trade and investment. http://www.conservativehome.com/the 2011, but re-opened in 2014. In August 2016, completion was not expected before columnists/2015/10/nick-timothy-the-government-is-selling-our-national-secur 2017. See http://www.onr.org.uk/new-reactors/assessment.htm (accessed 28.08.16.). ity-to-china.html (accessed 19.08.16.). S. Thomas / Energy Policy 101 (2017) 683–691 689 state-owned) to take a 75–80 per cent stake in a new reactor Canada Limited (AECL), to supply five reactors for Cernavoda. The vendor company with a reformed Areva parent company holding first unit was completed in 1996, the second in 2007 and since no more than 25 per cent. EDF would look to sell on some of its then Romania has been trying to fund construction of a third and a stake leaving it with at least 51 per cent of the shares. The main fourth unit. In 2008, an international consortium involving five candidates for these share are CGN and the Japanese company large European utilities37 reached agreement with the state con- Mitsubishi Heavy Industries (MHI). In June 2016, a memorandum trolled Romanian nuclear generation company, Nuclearelectrica, to of understanding was signed by MHI and EDF to strengthen co- build these two units. However, by 2013, all five of these utilities “ operation between the two companies foreseeing: the potential had withdrawn leaving Nuclearelectrica with all the shares in the participation of MHI as a partner in the French nuclear landscape project. reorganization with the acquisition of a minority equity interest in In 2013, Nuclearelectrica announced agreement with CGN to ”32 AREVA NP. ; Areva NP has been developing a reactor design, complete two CANDU reactors at Cernavoda at an estimated cost One, through a joint venture with MHI, Atmea SAS, since of €4bn.38 The project company will be a joint venture in which 2007 and this has competed in a number of markets with the best CGN will own at least 51% of the share capital. Construction was prospect of success in Turkey and Vietnam. Under the terms of the expected to start in three years with completion expected a fur- joint venture, both parties have agreed not to compete against ther six year later. However, by June 2016, the joint venture be- Atmea One with their own mid-sized (about 1000 MW) reactor tween Nuclearelectrica and CGN had yet to be set and construction design. It seems likely that if MHI does take a stake in Areva, this start was still 2–3 years away. will preclude a stake in Areva by CGN and if EDF/Areva chooses This project is more advanced than Bradwell and the fact that Atmea One as its preferred mid-size reactor design, the Bradwell project and the scope for cooperation between EDF/Areva and CGN reactors of this design are already in operation at the site should would be limited.33 reduce the risk of delay from the regulatory process. However, the fi The deal to rescue Areva's reactor division will be complex with fact that ve large European utilities all walked away from this fi EDF carrying out a rigorous due diligence and negotiating hard to project does suggest it is still nancially risky. CGN is also in- minimise its exposure to losses. In July 2016, the European Com- experienced with CANDU technology, the two CANDUs in service mission announced it would be carrying out an in-depth in- in China being built and operated by CNNC. vestigation to determine whether the rescue breaks EU state-aid legislation.34 It gave no time-table for completion of this in- 4.1.4. Turkey vestigation but it may mean completion of the rescue will not be Turkey has attempted several times over the past 35 years to possible, if at all, before 2018. launch a nuclear power programme but all attempts came to A more immediate prospect is that CNNC will take a stake in nothing until 2010 when a deal was agreed with Russia to buy four Areva's fuel cycle business, Areva NC, and in November 2015, a reactors for the Akkuyu site. This was followed up with a deal to memorandum of understanding was signed by the French and buy four reactors for the Sinop site from MHI in 2013 using Atmea Chinese presidents likely to lead to CNNC taking a minority stake One technology. In November 2014, Toshiba/Westinghouse, SPIC in Areva's fuel cycle business.35 China has long had aspirations to and the Turkish state-owned generation company, EÜAS an- complete the fuel cycle by acquiring reprocessing technology using nounced they were in exclusive talks to buy four reactors, two the recovered in fast reactors (Bunn et al., 2016). These using the Toshiba/Westinghouse AP1000 design and two using plans met a hostile public response in the Lianyungang, a city near SPIC's development of this technology, the CAP1400, to be built at the coast of Jiangsu Province where a reprocessing plant was an as yet unidentified site.39 Construction start was forecast for planned, and the plans were blocked by the city. 2018/19 (O’Byrne, 2014). However, the talks have not progressed France's nuclear power plants start to reach their fortieth well with apparent tensions between Toshiba/Westinghouse and birthday, the end of their design life, in large numbers from 2017 SPIC as to who should lead the talks, concerns in Turkey that it onwards. If EDF is able to life-extend these plants, a stake in Areva would be buying a ‘first-of-a-kind’ with the CAP1400 and difficulty fi NP could give CGN access to a signi cant and prestigious market assembling the financial package Turkey would need.40 By July carrying out the upgrades required to life-extend them. The French 2016, it was reported that the talks were no longer exclusive with Court of Auditors estimated that life-extension and upgrades re- Turkey open to offers from other vendors. The deal with MHI quired as a result of Fukushima would cost about €74bn between appears to be just as problematic. The Russian plant was expected 2016 and 2030.36 A stake for CNNC in Areva NC would give it ac- to start construction in 2011 when the deal was announced but by cess to additional technologies, uranium reserves and markets. 2016, construction start was still said to be a couple of years away. 4.1.3. Romania While the Bradwell project is much more strategically im- portant for CGN, the project to build a reactor using the Canadian 5. Issues raised by Chinese nuclear investment in Europe CANDU design in Romania is a more immediate prospect. The Cernavoda project dates back to an agreement in 1979 between Despite its huge recent order book, China's nuclear industry Romania and the Canadian reactor vendor, Atomic Energy of remains an unknown quantity. This is because China's nuclear experience is essentially only in China and it is difficult to get

32 https://www.edf.fr/en/the-edf-group/dedicated-sections/press/all-press-re reliable, independent information from China. The concerns are leases/mitsubishi-heavy-industries-and-edf-sign-memorandum-of-understanding- therefore inevitably somewhat speculative. on-collaboration-in-civil-nuclear-power-businesses (accessed 20.07.16.). 33 Nuclear Intelligence Weekly ‘EDF's Balancing Act Between MHI and CGN’ July 8, 2016, p 4. 37 GDF Suez (France), CEZ (Czech Republic), RWE (Germany), ENEL (Italy) and 34 http://europa.eu/rapid/press-release_IP-16-2587_en.htm (accessed Iberdrola (Spain). 20.07.16.). 38 http://www.world-nuclear-news.org/NN-Nuclearelectrica-owners-approve- 35 http://www.areva.com/EN/news-10655/areva-and-cnnc-considering-in China-agreement-26101501.html (accessed 22.12.15.). depth-partnership-with-capitalistic-and-industrial-components.html (accessed 39 Business Wire ‘Westinghouse Inks Multi-Party Agreement to Develop Nu- 22.12.15.). clear Power in Turkey’ November 24, 2014. 36 Power in Europe ‘CdC ups EDF investment forecast’ February 15, 2016. 40 Nuclear Intelligence Weekly ‘Wither Westinghouse and SPI? ’ July 22, 2016, p 4. 690 S. Thomas / Energy Policy 101 (2017) 683–691

5.1. Regulation country that is viewed with suspicion. In October 2015, The Times (O’Neill et al., 2015) reported: The Chinese safety regulator faces a workload unprecedented “Senior military and intelligence figures have warned ministers anywhere in the world. As of August 2016, it has to oversee the that plans to give China a big stake in Britain's nuclear power in- operation of 35 reactors, the construction of 20 reactors and the dustry pose a threat to national security.” There is particular con- commissioning of the first units of two new designs (EPR and cern that if CNNC is involved in the consortium, this will give it AP1000). More significantly for the export market, it is carrying access to some of UK's technologies with military applications, 41 out ‘first-of-a-kind’ (FOAK) reviews on five new reactor designs. such as reprocessing spent fuel to extract plutonium and enriching A regulator carrying out a FOAK has a particular responsibility as it uranium to increase the proportion of material that could be the cannot rely on the work of other reviewers and other regulators basis for a . will usually build on the work carried out in the FOAK review. In November 2015, the UK Commons Defence Committee ex- Many of China's target markets outside Europe have no history of pressed concerns about whether letting Chinese interests to control building and operating power reactors and their regulatory cap- a sensitive piece of infrastructure was wise. It concluded: “Whilst ability is undeveloped. Such countries will have to rely heavily on foreign investment in capital projects has a positive impact on the analysis by the Chinese safety regulator and if this analysis proves UK economy, such investments may open up vulnerabilities in our inadequate and, in a worst case, leads to a significant accident, the infrastructure. The resilience of our critical national infrastructure is credibility of the Chinese nuclear industry in export markets in- vital for UK defence and security and we expect the SDSR [Strategic cluding Europe will be seriously damaged (see Wübbeke and Ting Defence and Security Review] robustly to address this potential (2016)). danger” (House of Commons Defence Committee, 2015). There are concerns that the China's National Nuclear Safety Professor Helm (2015), a well-known economist who has served Administration (NNSA) is struggling to meet its workload. In 2014, as a member on various UK government bodies stated: “Add in the ’ Stephane Pailler, head of international relations at France s Auto- military and security issues of letting Chinese state-owned com- “ ’ rité de Sûreté Nucléaire (ASN) said: We don t have a regular re- panies into the heart of the British nuclear industry, and it seems lationship with the Chinese on EPR control like we have with the positively perverse to prefer Chinese government money to British ”42 ’ fi Finnish. ; Philippe Jamet, one of the French regulator s ve government money in so sensitive a national project.”. fi governing commissioners, testi ed before French Parliament in It seems clear worries within government about these issues “ ’ February 2014: Unfortunately, collaboration [with China] isn tata were over-ruled by Prime Minister Cameron and the Chancellor of level we would wish it to be. One of the explanations for the the Exchequer, George Osborne. Both left office in July 2016 after difficulties in our relations is that the Chinese safety authorities the referendum on the UK's membership of the European Union lack means. They are overwhelmed.”.43 (EU) came out in favour of leaving the EU. Cameron's successor, Theresa May, was reported as long having had reservations about 5.2. Quality the Hinkley deal and in July 2016, as discussed above, placed a hold on signing the Investment agreement pending a review of the ’ In March 2014, EDF s internal safety inspector Jean Tandonnet project. The government did not say what issues it was particu- published his annual report to the utility’s chief executive that larly concerned about but the media appeared to have been detailed a mid-2013 visit to the Taishan building site. He wrote briefed that it was the involvement of China that was behind the that “the state of conservation” of large components like pumps review. Her advisor, Nick Timothy, had warned in October 2015 and steam generators at Taishan “was not at an adequate level” and that “Security experts – reportedly inside as well as outside gov- was “far” from the standards of the two other EPR plants in Finland ernment – are worried that the Chinese could use their role to and France.44 In 2015, referring to the prospects for Chinese nu- build weaknesses into computer systems which will allow them to clear power plant exports, a senior expert at China’s SPIC said: shut down Britain's energy production at will”. “Our fatal weakness is our management standards are not high The more specific threat seems to be about industrial espionage enough. There is a big gap with international standards.”45 As with in general and would apply in equal measure to any major infra- regulatory capacity, countries with little nuclear experience will structure project. The former Chancellor, George Osborne was re- have to rely more heavily on China's quality control than on their portedly keen to get Chinese involvement in the UK's high speed own capability. rail proposal (HS2), a project of comparable scale to Hinkley Point. Whether Chinese involvement in this project will be subject to 5.3. National security and connections to military technology review under the new regime remains to be seen. The issue of industrial espionage was given particular point by the charging in 5.3.1. The UK the USA of one of CGN's advisers with conspiring to help the Concerns about national security and military connections are Chinese government to develop in a manner that seldom clearly articulated because of the political sensitivities they breaches the law - a euphemism for spying.46 raise. They are a particular factor for the UK market and France Whether the threat posed by Chinese ownership of strategic because of the capabilities with military connections these coun- infrastructure is realistic is not clear – Bradwell would account for tries possess. There is also a more general concern about owner- only a small fraction of the UK's generating capacity. Equally, it is ship of key infrastructure, such as power stations, in the hands of a not clear whether the UK's dual civil/military technologies of po- tential interest to China such as reprocessing are of any value. The 41 HPR-1000, CAP1400, ACP-1000, ACPR-1000 and a High Temperature Gas- UK's reprocessing plants are old, have been technologically pro- Cooled Reactor design. blematic and are expected to close by 2019, before CGN would be 42 http://www.bloomberg.com/news/articles/2014-06-18/french-nuclear-reg ulator-says-china-cooperation-lacking (accessed 12.01.16.). in a position to make an investment decision for Bradwell. The 43 Ibid. issue therefore seems to be one of industrial espionage. 44 Energy Monitor Worldwide ‘China regulators ‘overwhelmed’ as reactor building steams ahead’ June 21, 2014. 45 http://www.firstpost.com/fwire/made-in-china-nuclear-reactors-a-tough- 46 The Times ‘Chinese spies are biggest threat to nuclear power deal’ August 13, sell-in-global-market-2140127.html (accessed 6.03.15.). 2016. S. Thomas / Energy Policy 101 (2017) 683–691 691

5.3.2. France For China, its nuclear industry is reaching the point where it The issues of national security would appear much more ser- can attempt to enter export markets as an independent player. It ious for France than for the UK. If Chinese companies were to in- has formidable advantages in the strength of the Chinese economy vest in re-launched parts of Areva, this would give China direct and the degree of control ownership of the companies gives it. access to most of France's civil and military nuclear technologies. However, its nuclear industry is still young and its ability to pro- Through servicing, it would also give China a role in maintaining duce its own innovative designs that will satisfy the demands of and servicing France's 58 reactors, which supply about three experienced users rather than just copy ex- quarters of France's electricity. However, by August 2016, there isting designs remains unproven. Its ability to build reactors was no sign that concerns about Chinese companies’ role in res- quickly, cheaply and operate them reliably is also unproven out- cuing Areva were playing an important role in the debate about side China. More fundamentally, it is far from clear that the market Areva's future. It may be that, as in the UK, when the proposals are for nuclear reactors in Europe and beyond that would be open to more concrete or when there is a different political regime, these Chinese exports will be large enough to be worth pursuing. concerns will come to the fore. If the world market for nuclear reactors does not pick up, China may judge its efforts are better spent in less problematic sectors. Despite the scale of China's nuclear programme, nuclear power 6. Conclusions and policy recommendations only accounts for 3 per cent of its electricity, so while it may be that the world nuclear industry needs China for it to have a future, The nuclear industries of Europe and China are at a critical it is not clear whether China needs and can afford nuclear power. point in their evolution. For Europe, the number of countries ac- tively looking to pursue new nuclear programmes is confined to the UK, Finland, Turkey and France plus a few Eastern European Appendix A. Supporting information countries and it may that only a handful of orders will be placed in the next decade or two. Of these potential markets, only the UK, Supplementary data associated with this article can be found in Turkey and Romania are seriously considering China as supplier. the online version at http://dx.doi.org/10.1016/j.enpol.2016.09.038. The future of the reactor market in Europe may be near the point of no return and if orders are not placed soon, the nuclear option will be abandoned. Specifically, if the economics of nuclear power References do not improve dramatically, more countries are likely to follow Germany's example and replace their nuclear capacity with ’ cheaper renewables and energy efficiency measures. Bunn, M., Zhang, H., Kang, L., 2016. The cost of reprocessing in China Belfer Center for Science & International Affairs, Harvard Kennedy School. http://belfercenter. The UK is the key reactor market in Europe for China and a sale ksg.harvard.edu/files/The%20Cost%20of%20Reprocessing.pdf (accessed to the UK could be the catalyst for other sales to Europe. However, 20.07.16.). there are many serious obstacles to overcome before the Bradwell Guizhen, H., Mol, A., Zhang, L., Lu, Y., 2014. after Fukushima: understanding public knowledge, attitudes, and trust. J. Risk Res. 17 (4). project can proceed. The opportunity presented by the failure of Helm, D., 2015. ‘British energy policy – what happens next?’ Energy Futures. Netw. the French nuclear company, Areva, also gives China the prospect Pap. No, 11. of strengthening its technical capability and perhaps playing an House of Commons Defence Committee, 2015. Flexible response? An SDSR checklist of potential threats and vulnerabilities’ House of Commons. http://www.pub important role when and if France decides to replace or life-extend lications.parliament.uk/pa/cm201516/cmselect/cmdfence/493/49302.htm (ac- some of its existing nuclear fleet which begins to reach the end of cessed 7.01.15.). its design life from 2017 onwards. King, A., Ramana, M.V., 2015. The China syndrome? Nuclear power growth and safety after Fukushima. Asian Perspect. 39, 607–636. From a European policy point of view, China is potentially an O’Byrne, D., 2014. Turkey to ratify agreement for second plant by end 2014: min- attractive supplier of nuclear plants because of the expectation of ister’ 55, 48. low costs, the large amount of recent experience of construction O’Neill, S, Haynes, D, Pagnamenta, R., 2015. Nuclear deal with China is threat to UK and operation that exists in China and because of the likely security. The Times, October 16, 2015. http://www.thetimes.co.uk/tto/news/ politics/article4587446.ece (accessed 7.01.15.). availability of finance from China. Ironically, the degree of co- Ramana, M.V., Saikawa, E., 2011. Choosing a standard reactor: International com- ordination and support the Chinese government is able to give to petition and domestic politics in Chinese nuclear policy. Energy 36 (12), – its nuclear industry would be totally contrary to European Union 6779 6789. Schneider, M., Froggatt, A. 2016. World Nuclear Industry Status Report 2016, Mycle competition rules. Schneider Consulting, Paris. (http://www.worldnuclearreport.org/IMG/pdf/ There are concerns about quality, the rigour of the Chinese 20160713MSC-WNISR2016V2-HR.pdf) (accessed 15.08.16.). regulatory system and the national security issues raised by al- Thomas, S., 2016. The Hinkley Point decision: an analysis of the policy process. Energy Policy 96, 421–431. lowing China to control such a sensitive piece of infrastructure. For Wübbeke, J., Ting, G., 2016. ‘China's Nuclear Industry Goes Global’ The Diplomat, the UK, the issue of national security seems to be the dominant February 11, 2016. http://thediplomat.com/2016/02/chinas-nuclear-industry- one in its review of the Hinkley project. The judgement that will goes-global/ (accessed 28.08.16.). have to be made is whether allowing China in would be a good Yi-chong, X., 2010. The Politics of Nuclear Energy in China. Palgrave, Basingstoke. Yu, C.F., 2015Aa. Inland nuclear developers await policy change. Nucl. Intell. Wkly. way to provide much needed technology and capital to modernise IX (1), 3–4. key national infrastructure or a threat to national interests through Yu, C.F., 2015Bb. HPR1000 going global, what next? Nucl. Intell. Wkly. IX (44), 4–5. the opportunity for China to control key assets and the opportu- Yu, C.F., 2015Cc. The IPO strategies of China's nuclear majors. Nucl. Intell. Wkly. IX (40), 4–5. nities for industrial espionage it would open up. These three is- Yu, C.F., 2016. Is the Hualong one overtaking the AP1000? Nucl. Intell. Wkly. X (1), sues, national security, quality and safety culture are ones that 4–5. every government considering allowing Chinese companies into Zhou, S., Zhang, X., 2010. Nuclear in China: a study of op- portunities and challenges. Energy 35 (11), 4282–4288. its home market needs to consider carefully before any commit- Zhou, Y., Rengifo, C., Chen, P., Hinze, J., 2011. Is China ready for its nuclear expan- ment is made. sion? 39, 2, 771–781.