The PFBR under construction at Kalpakkam Nuclear Complex, Tamil Nadu, 2015. Photo: IAEA Imagebank/Petr Pavlicek.
New Delhi: On March 5, in response to a question in Parliament, the Union minister of state for atomic energy Jitendra Singh said that India’s first prototype fast breeder reactor (PFBR) will be “commissioned and operationalised” in December 2021.
The PFBR is a nuclear power reactor currently under construction at the Madras Atomic Power Station in Kalpakkam, Tamil Nadu. Fast breeder reactors, or FBRs, in general produce more fissile material than they consume. The PFBR in Kalpakkam will use a mixed oxide of plutonium-239 – derived from reprocessed spent fuel from the thermal pressurised heavy water reactors – and uranium-238 as fuel to generate energy in a nuclear reaction. This reaction will also produce – or ‘breed’ – more plutonium-239. This is possible because the reaction converts both uranium-238 in the fuel mix as well as a blanket of depleted uranium surrounding the core into plutonium.
This plutonium will then be processed and used as nuclear fuel in a chain of commercial FBRs that constitutes stage II of the nuclear programme. The stage will also include FBRs that will use thorium-232, mined in India, as a blanket. Thorium will get converted to uranium-233, which will serve as the fuel for advanced reactors in stage III. Ultimately, these reactors will burn uranium-233 and convert thorium-232 to more uranium-233, creating a self-sustaining cycle of nuclear power generation.
Singh’s words mark yet another assurance in a long list of similar statements that ministers and officials have issued in the last decade, only to be subsequently put off for one reason or another. Most of them have never been properly spelt out either, even in Parliament, leave alone to the people.
In its March 2020 report on the Department of Atomic Energy’s (DAE) demand for grants for 2020-2021 (no. 326), the Parliamentary Standing Committee on Science and Technology, Environment, Forests and Climate Change expressed the hope that the PFBR would be commissioned by end-2021. So this new deadline to commission the PFBR, shared with the minister as well as the committee, must have come from the DAE.
The standing committee had said, “Even though it would have taken almost two decades [of work on PFBR] when commissioning takes place, this is a pioneering initiative of which India can be justifiably proud. (sic) It will transform our nuclear energy programme.”
True enough, this would indeed be a great achievement – signalling the start of stage II of India’s three-stage nuclear power programme, a culmination of 50 years of work on fast reactors by the country’s nuclear scientists and engineers. This project was heralded in the 1970s with the design and construction of the fast breeder test reactor (FBTR), modeled on the French test reactor RAPSODIE. The FBTR became the learning and testing ground for the complex technologies that ultimately went into the design of the PFBR, in particular the use of liquid sodium as coolant, to transfer heat from the reactor core to the steam generator.
However, should we actually believe in the new deadline? Should we be assured that scientists and engineers of the Indira Gandhi Centre for Atomic Research (IGCAR), Kalpakkam, who have designed this first-of-a-kind reactor, and the Bharatiya Nabhikiya Vidyut Nigam Ltd. (BHAVINI), the special purpose vehicle set up under the DAE to execute stage II projects, have really sorted out the issues that caused the repeated delays?
The CAG audit
The government sanctioned the PFBR project in September 2003 at an estimated cost of Rs 3,492 crore. It was to be completed in seven years’ time – by September 2010. But since the project was running significantly behind schedule and cost overruns were only to be expected, the government in April 2012 sanctioned a revised estimate of Rs 5,677 crore (which the DAE had pitched for in April 2009 itself), extended the reactor completion schedule to September 2014 and set March 31, 2015, as the deadline for commercial operations. So if the reactor really goes into commercial mode in October 2022, as stated in Parliament on March 19, the total delay would be seven years after the revised schedule.
Emails to both IGCAR director A.K. Bhaduri and the BHAVINI chairman and managing director Kallol Roy for a realistic assessment of the present situation and their confidence in achieving criticality on the new target date elicited no response.
During its 2014 audit, the Comptroller and Auditor General (CAG) noted that BHAVINI had attributed the delays and cost overruns (in report no. 21, 2015) to “delay in obtaining government sanctions, damages due to tsunami [of December 2004], significant increases in prices of raw materials and labour rates, changes in designs and specifications, impact of taxes and duties, etc.” (emphasis added). The nature and the extent of the italicised part, and their impact on the project’s subsequent progress, have never been discussed openly either by IGCAR or BHAVINI.
However, these reasons, courtesy BHAVINI, didn’t include what the CAG’s audit actually found.
The CAG audit revealed significant deficiencies in the existing procurement procedure, which had delayed the placement of purchase orders for certain high-value items by up to three years. Since December 2003, when it was a new organisation and lacked the experience to procure nuclear-related equipment and systems, BHAVINI has outsourced the entire procurement process to the Contracts and Materials Management division of the Nuclear Power Corporation of India Ltd. (NPCIL), a DAE enterprise that had been performing these tasks for many years for all other nuclear power plants in the country.
This in itself is quite understandable and par for the course. However, according to the CAG, the procedural chain established was far from efficient, precipitating more delays. Though BHAVINI had set up its own Contracts and Materials Management division in May 2004, it had failed to develop the necessary in-house expertise to independently undertake procurement activities.
“[By] entrusting NPCIL with the entire gamut of activities relating to procurement such as processing of indents, tendering, evaluation of bids, price negotiations, placement of orders, etc., BHAVINI virtually transferred full control to NPCIL and decision-making by BHAVINI in procurement related matters became a mere formality,” the CAG’s report observed.
In its audit of purchase orders placed between 2012 and 2014, the CAG observed that though BHAVINI had placed most orders on its own, in terms of value they made up less than 20%, leaving BHAVINI dependent on NPCIL for the high-value items.
Out of 131 such high-value purchase orders that the CAG audited, the report noted that 100 purchase orders (76%) were delayed from one day to 1,092 days (nearly three years), with a median delay of 158 days. The audit also found that, subsequent to placement, the actual receipt of materials didn’t conform to the specified delivery schedules. A test check of 25 orders revealed that the delays ranged from five to 55 months.
These issues of material management apart, the most significant causes of delay were technical in nature – especially relating to the primary and secondary sodium coolant circuits.
The sodium circuits
The PFBR reactor core produces heat by nuclear fission; the heat is carried away by liquid sodium circulated using two primary pumps. (DAE scientists and engineers developed these pumps – including designing, testing and manufacturing – in collaboration with Kirloskar Brothers Ltd., Pune.) The hot and radioactive primary sodium coolant transfers the heat to secondary sodium coolant via four intermediate heat exchangers (IHXs). The non-radioactive secondary sodium is circulated through two independent secondary loops, each with two IHXs, a secondary sodium pump and four steam generators. The generators are linked to steam turbines to generate electric power.
The sodium coolant system in this final PFBR design is significantly different from the original, drawn up in the 1980s. The original design had provided for four primary sodium pumps, eight IHXs, four secondary loops and 36 steam generators. Officials gave up on this after a design review by a peer group between 1994 and 2002, undertaken at the suggestion of P.K. Iyengar, the then chairman of the Atomic Energy Commission. The new design required higher capacities for the pumps, loops and related components. It was finalised in 2002 and the project was approved in 2003.
According to R.D. Kale, a former IGCAR scientist-engineer and a key person responsible for the development of the sodium coolant technology at IGCAR, the aforementioned design change – which had been examined by French and Russian experts as well – is by itself unlikely to be the cause of the current problems with the sodium circuits.
In the current design, the reactor can’t be operated if even one of the two primary pumps isn’t operational. In a January 2020 article entitled ‘India’s Fast Reactor Programme – A Review and Critical Assessment’ in the journal Progress in Nuclear Energy, Kale wrote:
“The pool type [fast] power reactor[s] operated so far [have] had three primary pumps and three or more secondary coolant loops so that even in case of unavailability of one pump or loop, the plant could be operated at a reduced power level generating electricity.”
This is an important shortcoming. Kale also stated that from 1994 to 2002, most of the senior personnel involved in designing, constructing and commissioning the FBTR had retired or were about to. “This loss of technical expertise needed for a large project [like PFBR] was the second and more important shortcoming,” he added.
The two 600 MWe fast breeder reactors (FBR-1 and FBR-2), which the government has already approved as follow-ups to the PFBR, have been designed with three primary coolant pumps, instead of PFBR’s two. This is in line with the general practice around the world and the fact that no fast reactor built to date has had fewer than three primary coolant circuits.
There also seems to be a rethink these days on the power capacity of these two FBRs. They were originally rated 500 MWe before being increased to 600 MWe. However, a statement by BHAVINI’s Kallol Roy in the company’s 2018-2019 Annual Report read: “… based on the ongoing difficulties and experience generated during the entire ongoing commission phase of PFBR, it is being deliberated whether, for the purpose of standardisation, it may be prudent to retain them as 500 MWe units.”
But what were these problems? Kale, who was involved with the PFBR after his retirement as a consultant of sorts, said he is familiar with some of these problems, particularly with the sodium circuits. In fact, in his article, he recalled the difficulties with commissioning the sodium circuits, particularly the secondary loops.
Noting that it is essential to commission at least one secondary loop before attempting to fill the main primary vessel, Kale wrote, “During filling of the no. 2 secondary loop difficulties were encountered with [the] electromagnetic pump, which malfunctioned and failed due mostly to entrapped argon in the loop.”
The secondary loop is purged with inert argon gas to remove all traces of air and preheated to 150º C, by passing hot nitrogen gas, before sodium is filled. The electromagnetic pump is used to fill and drain the sodium in the secondary circuit as well as to maintain sodium levels during reactor operation, but it does not circulate sodium in the loops. The circulation is done by the main secondary sodium pumps, of which there is one for each secondary loop.
“The no. 1 secondary loop had no difficulty with its electromagnetic pump, and after filling the loop, the secondary main pump was put in operation,” Kale continued. “However, the sodium flow in the secondary circuits has been oscillating even with pump speed held constant. Entrapped argon gas in the main circuit could be a possible cause. The problem needs to be solved early. More recently the main pump in secondary loop-2 has failed. This has resulted in further delay in commissioning sodium circuits.”
According to Kale, there was actually no problem with the loop-2’s electromagnetic pump itself because an identical pump in loop-1 had been working without any problem for years. “Loop-2 pump had worked in the laboratory for nearly 20,000 hours without a hitch,” he told The Wire Science. “There probably was some engineering defect in loop-2 that was putting a higher load on the pump, which the project group did not agree with. Anyway the problem now seems to have been solved by changing the pump with one [with] a wider duct to bear the load.”
Kale also highlighted a procedural fault in the commissioning of the primary sodium circuit. According to him, the personnel haven’t adhered to the prescribed guidelines vis-à-vis the sequence of preheating the main vessel and preheating the IHX (to 150º C) plus commissioning the secondary loop. This caused the main vessel to preheat much slower than at the calculated rate of 10º C/day which in turn caused the vessel preheating to take as many as 12 months instead of the predicted one month.
However, the group apparently attributed this delay to limitations in the nitrogen gas feeding arrangement and changed it “arbitrarily” – against the recommendations of the design peer group and the heat transport specialist group.
“The tasks related to [main vessel] sodium filling, purification of [main vessel] sodium, commissioning of primary pumps, decay heat removal circuits, fuel handling machine checks are to be taken up before the fuel loading work can start. At the time of writing, it appears that these tasks will take about a year or more for completion,” Kale wrote. This assessment also points to a commissioning date near the close of 2021.
Need for a commissioning authority
In the absence of any responses from either BHAVINI’s Roy or the IGCAR director to queries from The Wire Science, we must go by statements on BHAVINI’s website and by Roy at annual general meetings (as reproduced in the company’s annual reports) to understand the project group’s perspective vis-à-vis the PFBR’s progress towards commissioning and approach to first criticality.
In BHAVINI’s September 2019 annual general meeting, Roy said that in March that year, both secondary loops had been restarted and that they were operating continuously. “The failure zones of the electromagnetic pump,” Roy continued then, “have been identified and dynamically adjusted based on vibration parameters, so as to ensure that their operations matched the operating regimes of the secondary loops. Further, the performance of both the SSPs has been characterised during the past six months of loop operation.”
He also said the transfer arm, a part of the reactor’s fuelling machinery – modifications of which had become necessary due to observed deviations in its performance – had since been completed and the sub-system had been reinstalled.
According to an undated statement (but likely to have been issued in late 2019 or early 2020, judging from the context) on the BHAVINI website, the present situation is the following:
“After erection and integration of all the systems both in Nuclear Island and in Power Island, commissioning of the individual systems and the integrated commissioning is in progress … Presently both the secondary loops are operational and performing satisfactorily. Main vessel is in preheated condition and the crown region of [the vessel] is being maintained at an average temperature of 140º C.
During commissioning, [the] large rotatable plug (LRP), which is used for positioning and approaching various fuelling positions, has developed a snag and the same is being attended [to] in consultation with the design team … Preparations are underway for commissioning the primary system, which will be taken up after completion of LRP bearing rectification work.”
Roy had noted in his September 2019 speech that many of these failures were happening as the equipment components were being commissioned, essentially due to their first-of-a-kind design. “Consequently,” Roy said, “all these equipment failures have been a setback to the earlier targets and it is estimated that, typically, two more years may be required to make the PFBR operational.” This also sort of ties up with the criticality target of December 2021.
However, it’s notable that Roy had specified new deadlines in previous annual statements, and which couldn’t be met for different reasons every time. In September 2017, he said, “For the present, while we are still hopeful of achieving criticality by December 2017, it could as well shift by a month or two and thereby achieving full-power operation and commercialisation should take place by June 2018.” In September 2018, noting “a significant shift in the project progress”, he said all efforts were being made to achieve criticality by mid-2019.
We now hope that IGCAR and BHAVINI scientists and engineers will succeed this time around and kick off the Indian nuclear programme’s stage II, and put it on course to use thorium as a nuclear fuel in stage III.
It’s pertinent to quote Kale’s article again here. An important reason for the significant delay with the PFBR, he wrote, is the “limited experience of both project and design personnel from BHAVINI and IGCAR respectively … While BHAVINI has employed a few ‘consultants’ here and there, it is not adequate. (sic) The department must try and utilise the expertise of retired personnel to support BHAVINI and IGCAR designers in the commissioning of PFBR. In addition it should seek active support from the personnel of the fast reactor technology group of IGCAR, [which possesses] considerable field experience in sodium systems.”
One of Kale’s recommendations “for effective and timely commissioning” is to set up a ‘commissioning authority’ to oversee the concerned activities. “This authority,” Kale wrote, “should be headed by a convenor, a senior officer from [fast breeder test reactor] with requisite background, assisted by several consultants chosen from experienced retired personnel with reactor/sodium systems background.”
Whether this advice has any takers within DAE and/or BHAVINI remains to be seen. In the meantime, we also wait for the PFBR to be commissioned next year – unless fresh problems crop up.
R. Ramachandran is a science writer.