The reactor failed four times in the final commissioning test. Numerous emergency shutdowns and four maintenance outages have kept the reactor off-grid for days, and it has supplied far less electricity than it was supposed to. Yet, plans to import a dozen more reactors of the same vintage from the same vendor are in progress.
V.T. Padmanabhan, Paul Dorfman & Anisur Rahman, EPW
We are very grateful to the anonymous peer reviewer for valuable comments. V T Padmanabhan fondly remembers R Ramesh, V Pugal, J Makolil, Raminder Kaur and K Sahadevan, colleagues in the Koodankulam Risk Audit Group.
We’re entering an era when a few individuals could, via error or terror, trigger societal breakdown.
—Rees M, Science, 8 March 20131
The Nuclear Power Corporation of India Limited (NPCIL), a public sector undertaking under the Department of Atomic Energy (DAE), has been operating a 1,000 megawatt electric (MWe) water–water–energy–reactor (WWER), the Russian version of the pressurised water reactor (PWR), since 22 October 2013 at the Koodankulam Nuclear Power Plant (KKNPP) in Thirunelveli district of Tamil Nadu, 98 km north-east of Thiruvananthapuram, Kerala’s capital. The NPCIL’s contractors constructed the plant with equipment supplied by the Atomstroyexport (ASE) of Russia. The reactor could not pass the “initial start up” (ISU) tests, which alone can ensure that it “will operate in accordance with design and is capable of responding to anticipated transients and postulated accidents” (United States Nuclear Regulatory Commission 2013: 7).
Within a decade, the KKNPP campus is expected to have six such reactors. A detailed history of this trans-boundary, trans-generational health risk for millions of people has been published elsewhere.2 We narrate the history of commissioning the reactor in haste in violation of safety guidelines, its transfer for warranty period operation, and analyse its performance since the synchronisation with the electricity network from 22 October 2013 till 30 April 2016.
Sources of Data
This study is based on the daily reports of power generation and station outages published by the Southern Regional Load Dispatch Centre (SRLDC), Bengaluru on its website (www.srldc.org). The data includes generation at the peak hour (19:30 hr); non-peak hour (03:00 hr); maximum generation of the day; and the electricity received by the grid after deducting internal consumption. The data on outage include the time and cause of outage, and the expected and actual dates of revival.
According to the official narrative, the initial fuel loading (IFL) of the reactor was completed on 2 October 2012. Though the first few fuel assemblies were loaded during the third week of September 2012, the process was stopped due to flaws in the neutron flux monitoring equipment (NFME). Safety guidelines of IAEA and AERB say that fuel loading should be stopped if the NFME is not working properly. An AERB presentation at the VVER Regulators’ Forum meeting at Kanyakumari, India in December 2013 reveals that NFME defect was noticed during the fuel loading.3 A report in the newspaper the Hindu on 18 October 2012 said that “the loading of enriched uranium was almost complete” and “a 10-member team from the AERB comprising experts in various sub-fields of Atomic Energy technology, began inspection of the plant.” On 26 January 2013, another report said that “the plant was almost on the verge of being commissioned last month, but got stopped after the authorities decided to do some maintenance work when some of the parameters were found to be not falling within the prescribed norms in toto”.4 On 3 June 2013, a Russian scientist at KKNPP noted in a Russian discussion forum that: “4 of intricate valve the size of a suitcase holding a block thousands, in a unique state where fuel is half loaded, and load characteristics have not been confirmed.”5 The reactor attained the first act of criticality (FAC) in July 2013, and was connected to the electricity grid on 22 October 2013 for power ascension (C-Phase commissioning) tests.
C-Phase commissioning has 45 tests in three sub-phases of C-1, C-2, and C-3, during which the reactor power is raised to 50% of full power (FP), 75% of FP, and 100% of FP respectively. The final test in C-3 is non-stop operation of the reactor for 100 days at 100% FP. The C-1 and C-2 tests take about 10 days, while the C-3 tests need 120 days. Add 30–40 days for regulatory deliberations and the reactor should have been commissioned within 160–170 days after the grid connection. The initial plan to commission the reactor before 22 April 2014 had to be shelved due to unusually high shutdowns due to trips (scrams)6 and maintenances.7
The C-1, C-2, and C-3 sub-phases of the KKNPP-1 began on 22 October 2013, 4 January 2014 and 3 May 2014, respectively. The reactor was in the C-1 sub-phase for 74 days and in the C-2 sub-phase for 119 days, a total of 193 days against the 180 earmarked for the entire phase. Technically the reactor is still in the C-3 sub-phase as four attempts to clear the final test failed, and the latest attempt began on 3 May 2016. Details of the failed FP tests are given in Table 1.
Clearing all the commissioning tests is the precondition for crossing four main milestones. They are (i) the award of a licence for regular operation of the reactor from the Atomic Energy Regulatory Board (AERB); (ii) commercial commissioning; (iii) transfer for warranty operation by the equipment supplier to the operator; and (iv) final transfer of ownership after successful completion of the warranty period operation. The reactor has crossed the first three milestones without clearing the final test.
(i) Provisional transfer: According to an official Russian blog, KKNPP-1 was “put into warranty operation, which was preceded by signing the power unit’s provisional acceptance certificate by the contractor, ASE, and the customer, NPCIL on 27 December 2014” (power-m.ru 2015). News agency TASS (2015) reported that “a year-long period of operation on warranty has begun” and “upon the results of these 12 months, the power unit will be fully transferred to the Indian side.”
(ii) Commercial commissioning: Three days after the provisional acceptance of the reactor from the ASE, at the 00:00 hour on 31 December 2014, NPCIL declared the commercial commissioning of KKNPP-1. According to the initial plan, this was supposed to be a star-studded event with the Russian President and the Indian Prime Minister jointly dedicating the generating station to the nation. Since the site received the order for this major milestone from the NPCIL head office only six hours before it took place, there was no celebration or celebrities at the site.
(iii) Licence to operate: During the first six months of commercial operation, the reactor tripped two times and was shut down on 24 June 2015 for a six-month long maintenance. A fortnight later, on 8 July 2015, the AERB granted a licence for its regular operation even though it did not clear the stipulated final commissioning test.
Warranty Period Performance
During the warranty period of one year, the reactor worked for 166 days and was on full power for 102 days, in two instalments. On 27 December 2015, the last day of the warranty period, the machine was still under maintenance and was connected to the grid on 31 January 2016. On 2 February 2016, the ASE extended the warranty period till the second quarter of 2016 “to carry out the final transmission,” after “the end of the warranty period of operation and confirmation of all installed power performance” (TASS 2016). During the 121 days of the extended warranty period, the reactor was offline for 49 days; 31 days on maintenance; and 18 days on three trips. Of the 72 online days, it was on full power for 18 days only. The rate of trips per year increased from 3.5 during 2015 to 12.2 during 2016. As the latest attempt for the final test began on 22 June 2016, it could not clear the test within the extended warranty period.
According to the US Nuclear Regulatory Commission (NRC), the power ascension test phase “should be completed in an orderly and expeditious manner,” and failure to complete it “within a reasonable period of time may indicate inadequacies in the applicant’s operating and maintenance capabilities, or may result from basic design problems.” In India, 17 reactors have been commissioned so far without any outside support, and a 540 MWe heavy water reactor at Tarapur near Mumbai was commissioned in 2006 in a record time of 148 days of grid connection. Only insiders will be able to judge whether the non-completion of power ascension tests and unusually high number of trips and shutdowns are due to operator incompetence, basic design problems, or defective equipment.
Violation of Court Orders
On the question of commissioning the reactor, the AERB said that “the safety guide on consenting process does not identify any stage of commercial commissioning” and the “declaration of NPP as commercial is the prerogative of the utility”.8 The NPCIL also said that “declaration of commercial operation is the purview of NPCIL”.9 Though the AERB guide does not mention anything about commercial commissioning (AERB 2014), the Madras High Court and the Supreme Court assigned a central role to the AERB in the commissioning of KKNPP-1.
83.2. These two stages (siting, construction) have been completed in respect of KKNPP. The third stage of ‘commissioning’ involves regulatory clearances at several intermediate stages/phases starting from hot commissioning stage to raising reactor power up to 100% rated power. (Madras High Court 2012)10
The plant should not be made operational unless AERB, NPCIL, DAE accord final clearance for commissioning of the plant ensuring the quality of various components and systems because their reliability is of vital importance. (Supreme Court 2013)11
The hurried commissioning of the reactor, its transfer for warranty operation, and the award of an operating licence without clearing the commissioning test were dictated by extraneous reasons. A couple of hours after the AERB board decided to award the licence, during a press conference after the bilateral meeting of the Indian premier and the Russian president at Ufa in Russia, the Indian ambassador said,
The Kudankulam 3 and 4 contracts have already been signed. There is what is called the Long Cycle Supply of Equipment which is a contract that has been signed. So, it is in the process of implementation. That is what the (Indian) Prime Minister and President Putin noted with satisfaction as progress in the nuclear energy cooperation.12
Incidentally, the AERB officially announced its 8 July 2015 decision to grant a licence to operate the KKNPP-1 in a press release a week later. Unlike all the earlier AERB consents that used to mention the Advisory Committee for Project Safety Review (ACPSR), which has been overseeing the KKNPP project from the beginning, there is no reference to the ACPSR in this last consent.
In short, all the three vital events in the life of the reactor were dictated by politics and diplomacy, and not based on safety science and regulatory norms. With the midnight commissioning and the award of the operator’s licence for a reactor, which was in an intensive care unit of sorts, India has entered a brave new world of regulation-free fission technology.
The KKNPP-1 is the world’s first PWR certified as Generation-III by the European Utility Requirement (EUR) Club, a consortium of power-generating companies based in Europe (Ermakov and Rousselot 2007). During the proceedings for the EUR certification of the WWER reactor in 2006, the Russian designers had shown the Koodankulam reactors, which were then under construction, as prototypes. Gen-III reactors are supposed to be more reliable, safe, and economical.
Since the reactor attained FAC during the middle of July 2013 and was grid-connected in October 2013, we now have its operational data for two and a half years, which represents about 6% of its design life of 40 years. Being the only Gen-III certified Russian reactor operating in the world, the designer and the global nuclear industry might be keeping a close vigil on its performance. As they have not published any report so far, we present this performance analysis.
In the Words of Designers
KKNPP houses version V-392 (also known as AES-92) WWER-1,000 reactors, which according to the Russian designer,
has better economical parameters than those of a V-320 plant. Specific capital investments in the construction are 1.6 times less and the electricity generation costs are about 1.5 times less than for existing serial plants with WWER-1000/V-320 design. Economical efficiency of NPP operation is determined by quantity of the produced power which depends on reliability of power production. The plant availability goal for WWER-1,000/V-392 design was established as 0.85 around the year. The factor of power generation losses due to repair and maintenance must not exceed 0.1 (that is, the duration of the scheduled outage must not exceed 15–37 days per year). The factor of the power generation losses due to system failures must not exceed 0.05 per year. (IAEA 2004)
Reactor in Real World
In the 922 days between its grid connection and 30 April 2016, four maintenance outages and 22 scrams kept the reactor off-grid for 470 days. Year-wise data is given in Table 2. The reactor was supposed to be on 100% full power about a month after grid connection. We assume that KKNPP-1 reached this milestone on 1 January 2014. As per the preliminary safety analysis report (PSAR), the duration for refuelling and annual maintenance is 30 days, and it is assumed that the reactor worked for 335 days in a full year. Its maximum contribution to the grid was 22.8 million units (MU) on 7 January 2015, the average during the full power days was 22.5 MU, and this is considered the norm (reference energy supply) for the unit since 1 January 2014 (row 11 of Table 2). The SRLDC provides data on the net supply of electricity after deducting in-house consumption, which is about 1.5 MU/day when the reactor is working. When the generator is off-grid, the unit consumes about 1 MU/day. The estimated electricity drawn by the unit from the grid during its non-working days is in row 13. The net amount of electricity received by the grid from KKNPP-1 after deducting the withdrawal during non-working days is given in row 14. The data for the period of commercial operation is given in column (g).
(i) During 2014–16, the overall contribution of the reactor was on 40% of its design potential.
(ii) The contribution during the warranty period operation was 42.6%, which is less than half the quantity promised by the designer.
(iii) The total loss of revenue to the NPCIL due to the lost productivity of KKNPP-1 since January 2014 till 30 April 2016 was `4,245.5 crore. Lost revenue accounts for about 20% of total revenue from sales and 81% of the net profit of the NPCIL.
(iv) The rate of trips per 7,000 hours at KKNPP-1 and workdays lost per trip are 14.1/yr and 5.1 days/trip at KKNPP as against 0·5/yr and 1·5 days for all the reactors in the world respectively (World Nuclear Association 2014). The trip rate at KKNPP-1 is 28 times higher than that of all the commercial reactors in the world.
Comparison with Other Reactors
The International Atomic Energy Agency (IAEA) publishes the performance statistics of all commercial reactors in the world. Three indicators (operation factor, energy availability factor, and load factor) for 5 WWER-1000 reactors and one heavy water reactor commissioned during this century are given in Table 3. Of these, KKNPP-1 is the only reactor certified as Generation-III, while all others belong to Generation-II. These data are for the first full calendar year after commissioning the reactor. KKNPP-1 is vying for the bottom most position in all the indicators. TAPS-2 on row 7, one of the oldest boiling water reactors in the world, commissioned some four decades ago, is far more efficient.
The dismal performance of KKNPP-1 during the 30 months of its grid connection raises doubts about the Gen-III certification process. As the certification was done 10 years ago, when the reactor was in the early stages of construction, the certifier (the EUR Club) should take cognisance of the performance of the real reactor in the two and half years of its operation.
In spite of the repeated failures in clearing the commissioning tests and its position as one of the worst performing commercial reactors in the world, the NPCIL, the owner of KKNPP-1 is busy negotiating more deals for reactors of the same vintage from the same vendor. The people, Parliament, and governments have a right to know what is happening behind the high walls of the nuclear campus. Most of the legislators and decision-makers believe that projects such as KKNPP are important for national defence, which was true till the Indo–US nuclear deal. In reality, all the 14 operating reactors, including KKNPP, which are under IAEA safeguard, are simply commercial ventures. The NPCIL being a public sector undertaking, its accounts ought to be scrutinised by Parliament and the Comptroller and Auditor General of India.
After the meltdown of five reactors, we are now better informed about the catastrophic risks of fission technology. All the five meltdowns from Three Mile Island to Fukushima were accidents in the real sense because nobody had foreseen them. In the case of Koodankulam, several well-informed persons who had held important positions in academia and in civil, military and public services in India and abroad have raised concerns about the safety of the reactors.
The plant’s continuous existence without a proper safety audit poses an unacceptable risk, not only for the people of India, but also for the people of South Asia and wherever the wind will blow.
V T Padmanabhan (vtmaaiom) is Member, Nuclear Consulting Group; Paul Dorfman (rfmcl.k) is with the Nuclear Consulting Group and UCL Energy Institute, University College London; and Anisur Rahman (nsrrha1bicm) is a former director of Nuclear Safety Consultants, UK.
1 M Rees, “Denial of Catastrophic Risks,” Editorial. Science, 8 March 2013: DOI: 10.1126/science.1236756
2 V T Padmanabhan, Paul Dorfman, A Rahman, 2015 doi: 10.13140/RG.2.1.3239.8561.
3 K J Vakharwala, 2013, Safety Review of Kudankulam NPP KK NPP -1, 2, AERB, 20th Meeting of WWER Regulator’s Forum – 11–13 December 2013, Kanyakumari, India.
4 The Hindu, New Delhi, 26 January 2013 AERB permits repeat of hydro tests at Kudankulam http://www.thehindu.com/news/national/tamil-nadu/aerb-permits-repeat-of- kudankulam/article4345280.ece.
5 http://forum.atominfo.ru/index.php?showtopic =716.
6 Definition of Trip/SCRAM: US NRC defines scram (trip) as follows: “Key operating parameters of a nuclear power plant are continuously monitored to detect conditions that could lead to exceeding the plant’s known safe operating limits, which might damage the reactor core and release radiation to the environment. If any of these limits is exceeded, then the reactor is automatically shut down, in order to prevent core damage. The automatic shut-down of a nuclear reactor is called a reactor trip or scram. A reactor trip causes all the control rods to insert into the reactor core in about three seconds and release of boric acid into the coolant” (http://public-blog.nrc-gateway.gov/2012/12/28/what-is-a-reactor-trip-and).
7 Central Electricity Regulatory Commission New Delhi Petition No 72/MP/2014, Date of Order: 12 May 2014, http://www.cercind.gov.in/2014/orders/SO72.pdf.
8 AERB letter No AERB/RST/RTI Appln 587/2015/4008 dated 23 June 2015.
9 NPCIL letter No NPCIL/CPIL/MUMBAI-HQ/327/2015/873/421 dated 7 Jul 2015.
10 Justice P Jyothimani and the Justice M Duraiswamy dated 31 August 2012 (WP No 24770).
11 Justice K S Radhakrihnan and Justice Dipak Misra, 2013 Judgment in Civil Appeal No 4440 of 2013 (Arising out of SLP (C) No 27335 of 2012, p 242).
12 Ministry of External Affairs press briefing, http://www.mea.gov.in/bilateral-documents.htm?dtl/25443/Transcript+of+Me.
AERB (2014): “Commissioning of Pressurised Water Reactor Based Nuclear Power Plants,” Safety Guide No AERB/NPP-PWR/SG/O-4 C, October, http://www.aerb.gov.in/AERBPortal/pages/English/t/publications/CODESGUID.
Ermakov and Rousselot (2007): “European Utility Requirements for LWR Nuclear Power Plants for LWR Nuclear Power Plants: EUR volume 3 AES 92 subset,” EUR Seminar, Nice, France, 15 May.
IAEA (2004): Status of Advanced Light Water Reactor Designs, IAEA-TECDOC-1391, pp 358–59, http://www-pub.iaea.org/MTCD/publications/PDF/te_1391_web.pdf.
Power-m.ru (2015): “Partner Company Highly Appreciates Participation of Power Machines Ojsc in Installation and Commissioning of First Power Unit at Kudankulam NPP in India,” power-m.ru, 19 March, http://www.power-m.ru/en/press-center/news/partner-company-highly-apprec.
TASS (2015): “Kudankulam-1 Transferred to India for Warranty-Period Operation,” Russia and India Report, 2 January, http://in.rbth.com/economics/2015/01/02/kudankulam-1_transferred_to_indi.
— (2016): “Final Transfer of the First Unit of Kudankulam NPP in India Is Planned in the Second Quarter – Rosatom,” atominfo.ru, 2 February, http://atominfo.ru/newsm/t0624.htm.
United States Nuclear Regulatory Commission (2013): “Regulatory Guide 1.68: Initial Test Programs for Water-Cooled Nuclear Power Plants,” June, http://pbadupws.nrc.gov/docs/ML1305/ML13051A027.pdf.
World Nuclear Association (2014): Optimized Capacity: Global Trends and Issues 2014 Edition, World Nuclear Assocation, http://www.world-nuclear.org/uploadedFiles/org/WNA/Publications/Working_.
See more at: http://www.epw.in/journal/2016/37/commentary/koodankulams-untameable-atomic-reactor.html#sthash.5H4F558b.dpuf