Pebble bed reactor C safety in perspective by yurtgc548


									FUEL & FUEL CYCLE

Pebble bed reactor –
safety in perspective
Dr Albert Koster, PBMR (Pty) senior consultant,
nuclear safety, replies directly to criticisms of
the PBMR reactor in NEI’s March 2009 issue.

                           n 5 February 2009 PBMR           from coal, also stimulated
                           issued a news bulletin stat-     designers to look at HTRs
                           ing that the company will        for the production of
                           be focusing on the design of     process heat for a variety of
           a plant to service both the electricity and      uses. But politics (reunification
           process heat markets. In the March issue         of Germany in 1989 scup-
           of this magazine [pp22-3] Prof. Stephen          pered a project to replace coal
           Thomas insinuates that the change was            in East Germany) and the col-
           motivated not by commercial considera-           lapse of the oil price kept such
           tions but because PBMR was aware of              plans on the drawing board.
           problems with pebble bed performance                In 1988, Eskom was considering           once again. At the same time, the glob-
           at high temperatures, as alleged by Dr.          future expansion of the nuclear option      al energy community was exploring
           Rainer Moormann in a second article              and together with the South African         other uses of nuclear energy under the
           [pp16-20]. In order to set the record            Atomic Energy Corporation looked at         umbrella of the Generation IV Inter-
           straight it is necessary to chart a short his-   various options, one of which was an        national Forum. The US Department
           tory of the development of HTRs and              HTR for which the AEC took the lead.        of Energy identified the VHTR, a ver-
           PBMR in particular.                              When some money was made available          sion of a high temperature gas reactor
              The backbone of the High Temper-              from a different source in 1989, a new      as their primary selection for the USA
           ature Reactor (HTR) is the coated par-           study led to the selection of a direct      NGNP programme that is based on
           ticle fuel that was initially developed in       cycle HTR electricity plant as a possi-     process heat applications with the ulti-
           the 1950s for the DRAGON reactor in              ble alternative or supplement to future     mate goal of hydrogen production.
           the UK as part of a European/USA                 LWRs for the Eskom grid. The selec-         PBMR (Pty) Ltd. has been actively
           project. The original particle had a             tion was made for the HTR, which is         involved in this programme since 2002.
           high-density pyrolytic carbon layer              designed to eliminate or withstand             This brings us to the allegations of
           and a porous buffer layer and was                catastrophic nuclear accidents like a       hiding facts and supposed safety prob-
           called the BISO particle. This was               core melt accident. The direct cycle        lems in pebble bed reactors as so
           used in the early HTRs in either peb-            promised higher efficiencies, but           stridently described in the March issue
           bles or prismatic graphite blocks                requires high temperatures. Again, this     of NEI. To this end, only the safety
           (AVR, THTR, Peach Bottom). Later                 project only had electricity production     concerns raised by Moormann need to
           fuel development included an interlay-           in mind. In 1999, when PBMR (Pty)           be addressed as Thomas based his
           er of SiC within the high-density pyro-          Ltd. was finally established with a staff   argument on the premise that Moor-
           carbon layer. All these layers effectively       of around 30 people, the oil price was      mann is correct: that PBMR knew
           block gaseous fission products and               still close to the lowest level since       about safety problems all the time and
           most, but not all, metallic fission prod-        WWII and, hence, process heat appli-        opted to keep quiet about it.
           ucts. This fuel, which was developed             cations were still in the future. But          In his articles, Moormann presents a
           mainly in the USA and Germany, was               times have changed drastically since        number of different arguments but by
           exhaustively tested in both material             then. The sharp rise in oil price (which    some analysis, these can be boiled
           test reactors and in the latter half of          peaked at almost $150 a barrel in July      down to two major issues. The first is
           the AVR operation. This fuel was also            2008) made such ventures as the             that both the AVR and THTR were
           used for the fissile particles in the Fort       Canadian oil sands viable. The global       shut down because of safety concerns.
           St Vrain plant in the USA and forms              warming issue, barely visible in 1998,      The second is that the AVR was highly
           the basis for all new HTR projects               has become a major concern.                 contaminated and this was due to high
           since then.                                         Whereas the business case for            fuel temperatures that caused excessive
              Initially, the HTR was conceived for          PBMR (Pty) Ltd. in 2002 was still firm-     release of caesium and strontium from
           electricity production, as were all other        ly focused on electricity, changes in       the fuel. He then advances reasons
           reactors. However, the first oil shock of        global energy supply security, cost and     why the fuel temperatures might have
           1973 which inter alia gave a temporary           environmental impacts have made the         been high and draws conclusions
           boost to the development of synfuels             process heat market very promising          about the safety of future pebble bed

22                                                                                      NUCLEAR ENGINEERING INTERNATIONAL MAY 2009
                                                                                                                             FUEL & FUEL CYCLE

                                                                   ed to the plant itself.”[3]   to absorb gaseous fission products, was
                                                                     Moormann also claims        much denser than prescribed. This
                                                                  that the THTR was shut         fuel batch is generally accepted to have
                                                                  down because of techni-        caused excess failed particles with
                                                                  cal problems, but the          resultant release of metallic fission
                                                                 technical problems experi-      products [6]. The release of strontium
                                                                 enced by the THTR were          followed the introduction of a specific
                                                                teething troubles that had       UC fuel in 1974. It is thus stretching
                                                                been anticipated and bud-        the argument to compare this to fuel
                                                                geted for in the risk-sharing    used in Peach Bottom. The detection
                                                               agreement between the util-       of caesium and strontium in the gas
                                                               ities and the government.         stream was preceded by the detection
                                                              The unplanned increased            of noble gases, which was the main
                                                              cost in 1989 was due to the        reason the particular fuel was identi-
                                                              updated estimates of the           fied as being the culprit.
                                                             decommissioning costs and              It is unfortunate that Moormann
                                                             the potential delays caused by      has chosen to include so many unsub-
reactors based on his speculations. The          having to re-license the fuel plant or          stantiated (by references) statements
major contentions are addressed                  have no fuel for more than two years.           such as “however, in THTR300, peb-
below; others have been covered at               Although the government agreed to               ble flow was virtually restricted to the
the HTR 2008 conference or in prior              the increased share in the risk, the state      core axis, and hardly occurred at all in
published articles.                              of Nord-Rhein Westphalen did not.               the outer core zones.” PBMR staff
                                                 The utility threatened to stop opera-           consulted extensively with engineers
WHY THE AVR WAS SHUT DOWN                        tions and in the dispute that followed          and scientists involved in the design
Moormann makes a few statements on               the decision was taken that enough              and operation of the THTR. PBMR
the closure of both the AVR and                  had been learned and the State did not          also has access to all design and test
THTR that are not supported by the               want any further expenditure. This is           documents from the THTR and no
literature and personal recollections of         substantiated by a citation (translated)        reference to non-existent pebble flow
the people involved. In effect, the AVR          from a paper by Prof. Knizia, the               can be traced. However, based on the
had come to a natural end of life for a          then chairman of the board of                   removal rate of spheres broken in the
research reactor. Further research               VEW (whose subsidiary HKG operat-               initial testing of the in-core control
required a new project called the AVR            ed the THTR) and Dr. Baumer, who                rods, it was deduced that the flow
II, which was not approved by the Ger-           at the time was station manager for the         velocity of pebbles against the side
man government. As a result, in 1985             THTR:                                           reflector was about a factor of six
the AVR researchers defined 21 final                “…It was not technical and especially        lower than those in the middle. This
experiments that they still considered              not safety related technical problems in     was partly due to the large core diame-
worthwhile [1]. Of these, all were                  the plant, but external economical factors   ter (5.6m) but also due a to high
performed but one. A blowdown                       that caused risks that were outside the      temperature difference between the
experiment to investigate the mobility              influence of the operator, together with a   side and the centre of the core, which
of the dust in the primary circuit was              lack of commitment from the political        was later adjusted. For modular pebble
not done, even though it was approved               sides to further support the project that    bed reactors like PBMR the long and
by the regulator. If safety problems                caused the eventual early closure of the     narrow core causes less of a radial gas
existed at the AVR, continued opera-                project THTR...”[4]                          temperature difference and also, due
tion would not have been allowed, nor            Moormann also states “a study on                to the geometry, less variation in peb-
would significant experiments be                 the operational safety of the PBR               ble flow. In any case, the pebble flow at
approved by a regulator. There is no             prototype, the AVR reactor, was per-            most causes a shift in power density,
known document that cites safety                 formed at FZJ and published in June             but has no effect on fission product
reasons for the shutdown, but there are          2008.” The author therewith implies             release and only a second-order effect
statements on the AVR closure of                 that this was a study initiated by the          on the maximum expected accumulat-
which we quote, the first by Prof.               research centre. However it is entirely         ed burn up. Note that cycling the fuel is
Theenhaus, a member of the FZJ                   his own initiative and the references           a means to reduce the maximum to
board and published in [2]. The trans-           are to his own report and not indepen-          average flux distribution, but whether
lated quotation reads:                           dent corroboration of his opinions.             a pebble is cycled 3 times or 10 times
   “In more than 20 years of operation the          He also states “the same fuel type           makes no difference to the integrity of
   advantages and positive characteristics of    that in the AVR (BISO coating, UC2-             the fuel.
   this type of reactor have convincingly        kernel) has released a major fraction of           Moormann states correctly that at
   been demonstrated. Many experiments           Sr-90 was also used in core 2 of the US         the end of life of the AVR, it was
   have been performed, with particular          Peach Bottom block-type HTR with-               found that the gas temperatures in the
   emphasis on safety research…This              out any significant Sr-90 release”….            core were much higher than anticipat-
   demonstration reactor and in a certain        “there was almost no correlation                ed or derived. He advances many
   sense research reactor completely fulfilled   between the release of metallic fission         reasons why this could be the case. He
   its mission.”                                 products and noble gas release that             states inter alia: “If external flows
Dr. Marnett, technical director of the           indicates inadequate fuel quality.”             reduce the core cooling, it is reason-
AVR GmbH, stated: “The AVR-                         The fuel that failed in the AVR and          able to expect a homogeneous core
Experimental Power Station has oper-             led to caesium releases had a much              temperature increase. This was not
ated for 21 years...and was taken out of         higher fuel density than the other fuel         found, so external bypass flows cannot
operation in 1988 for reasons unrelat-           pebbles. The buffer layer that is porous        explain the high AVR temperatures.”

WWW.NEIMAGAZINE.COM MAY 2009                                                                                                                 23

                    At the HTR-2008 conference, there            high caesium and strontium release            conditions. Afterwards the irradiated
                    was a contribution by PBMR (Pty) Ltd.        from modern fuel and he rejects the           fuel will be subjected to post-irradia-
                    analysts who did a complete review of        accepted explanation given in [6].            tion heat-up testing to simulate
                    the AVR design and operation [5]             From this again, he calculates that a         design-based accident events where
                    using modern analytical tools. One of        potential existed for high doses if all       measurements of all significant
                    the challenges are that there were no        the contaminated dust were released to        nuclides will be made.
                    fixed in-core and only limited ex-core       the environment and that the same will
                    (above the core top ceiling structure)       occur in newer pebble bed reactors.           Full understanding and reliable
                    gas temperature measurements, thus           Using the German dose criteria on             modelling of core temperature
                    the statement by Moormann based              advanced PBMRs leads him to the               behaviour, and of pebble bed
                    only on these measurements quite far         conclusion that PBMRs need a                  mechanics, including pebble
                    from the top of the pebble bed is puz-       gastight containment.                         rupture.
                    zling. The analysis shows clearly that          Moormann apparently assumes a              The publications presented at HTR-
                    the main contributor to the increased        complete lack of mitigating defence-          2008 and the final model results show
                    average temperature was the neglect of       in-depth design solutions. Design             that this is already achieved. The
                    core bypass flows. Final 3D analysis         solutions are tailored to cover all possi-    terminology ‘pebble rupture’ is mis-
                    now being processed that uses the            ble accident progression scenarios.           leading; pebbles do not rupture. A very
                    actual fuel load history and therefore       The initial release of fission products       small percentage may fail due to
                    the correct spatial power profile shows      for HTRs is small, despite what he            mechanical handling and movement
                    very good agreement with the AVR             maintains, but delayed releases may           and the faulty pebbles are automatical-
                    measured data and the resultant              occur if active cooling in depressurized      ly removed from the core when they
                    high gas temperatures. The so-called         conditions fails. If the reactor were in a    exit at the bottom of the core.
                    ‘unpredictable hot gas currents’ can be      containment, this would lead to an
                    explained by detailed analysis taking        increase in containment activity that         CONCLUSION
                    the specific design features of the AVR      cannot be filtered due to excessively         Re-interpreting old data always pre-
                    (upward gas flow, fuel load and pebble       high temperatures and pressures. By           sents a problem, as it is impossible to
                    flow, noses protruding the core, bypass      first relieving the initial pressure burst    go back and perform more experi-
                    streams in the control borings, etc) into    and then closing the containment, this        ments. It thus requires a concerted
                    account, and these are not relevant for      problem is solved. This solution can          analytical effort that PBMR has
                    other pebble bed designs. It therefore       variously be called a confinement or a        undertaken with promising results.
                    shows that the other reasons advanced        low-pressure containment and has              Moormann, without the benefit of
                    by Moormann are at the most sec-             been the preferred solution for all           such analysis, has very recently decid-
                    ondary, or do not fit at all. He then uses   modular HTRs and is also as such              ed that the previous interpretation of
                    these high gas temperatures to postu-        described in an IAEA report [7]. The          the reason for spikes in fission product
                    late that this gave rise to unacceptably     initial release can, as is foreseen for the   release in the AVR is wrong and
                                                                 PBMR, be filtered to remove the dust.         advanced his own view. This view is
References                                                          Moormann proposes areas where he           not shared by anyone else directly
                                                                 feels more research is needed, some of        involved with the operation of the
[1] KIRCH-IVENS-1990, in AVR-High-Temperature                    which are addressed below.                    AVR and the analysis provided in [6] is
Experimental Reactor, 21 Years of Successful Operation,                                                        considered a valid interpretation of
VDI-1990; page 95.                                               Full evaluation of the opera-                 the data, especially when combined
[2] R Theenhaus, Final Report on AVR Operation, 1997,            tional experience and problems                with the many irradiation experiments
Introduction, page 4.                                            of AVR and THTR300.                           performed over the years. It is faulty
                                                                 PBMR (Pty) Ltd. has been in the               reasoning to conclude that pebble bed
[3] Marnet, Chrysanth; Wimmers, Manfred; Ziermann,
                                                                 process of evaluating the AVR for the         reactors are unsafe, based on this
Egon: Das Versuchskernkraftwerk der AVR – wichtige
                                                                 last two years. The starting point was        one-sided conclusion and a flawed
Ergebnisse eines anderen Reaktorkonzepts, (The
                                                                 to collect all the design drawings and        understanding of what mitigating
Experimental Nuclear Power Station of the AVR GmbH -
                                                                 descriptions to (for the first time)          design solutions have been included in
Important Results of a Different Reactor Concept),
                                                                 enable the AVR to be modelled in              the PBMR design for a considerable
Schulten-70-1993, pages 275-284.
                                                                 detail. The latest results were present-      number of years.
[4] Klaus Knizia, Ruediger Baeumer: Bau, Betrieb und             ed at HTR-2008 [5]. Additional results          PBMR is not about to ignore any
Stillegung des THTR 300, Erfahrungen und ihre Bedeutung          explaining the fuel temperatures will         valid safety concern or new opinions
für die weitere kerntechnische Entwicklung; Schulten-70-         appear soon in print.                         about earlier experiments. But
1993, pages 285-292.                                                Components from the AVR contin-            Moormann’s views are old news and
[5] Viljoen C, Sen S, Ubbink O , Reitsma F, Pohl P, Barnert H;   ue to be examined, at the request of          not supported by new, advanced
The re-evaluation of the AVR Melt-wire Experiments using         PBMR (Pty) Ltd., for dust characteri-         work or the preponderance of con-
Modern Methods with Specific Focus on Bounding the               zation, concentration and dust adher-         temporary evidence, analysis and
Bypass Flow Effects. HTR-2008-58115.                             ence to better understand the mobility        expert opinion established during
                                                                 of agglomerated dust, or the lack of it.      the AVR and THTR operating peri-
[6] Nabielek H, Verfondern K, Kania; Fuel and Fission                                                          ods and more recently. Further, mod-
Products in the Julich AVR Pebble Bed Reactor, paper             Experiments on iodine release                 ern tools and techniques are highly
HTR-2008-58337.                                                  from fuel elements in core heat-              reliable tools only conceived of just a
[7] Considerations in the development of safety                  up accidents.                                 short time ago that make the ability
requirements for innovative reactors: Application to             This is part of the planned PBMR              to accurately predict system and
modular high temperature gas cooled reactors. IAEA               fuel qualification tests where fuel will      component performance in ways
TecDoc 1366, August 2003.                                        be placed in test reactors and subject-       unimagined when earlier designs
                                                                 ed to expected operating temperature          were developed.

24                                                                                            NUCLEAR ENGINEERING INTERNATIONAL MAY 2009

To top