Safety aspects and environmental impact of an inertial confinement fusion reactor

Nuclear Engineering and Design 100 (1987) 475-487 North-Holland, Amsterdam

SAFETY ASPECTS AND ENVIRONMENTAL OF AN INERTIAL CONFINEMENT FUSION G. KESSLER

475

IMPACT REACTOR

a n d A. B A Y E R

Kernforschungszentrum Karlsruhe, Postfach 3640, D-7500 Karlsruhe, Fed. Rep. Germar~v Received 27 October 1986

Releases into the environment of radioactive materials contained in heavy ion fusion (HIF) reactor plants must be prevented by similar safety design concepts as they are applied to present fission converter (e.g. LWR's) and breeder reactors (LMFBR's). This study identifies significant safety aspects of inertial confinement fusion power plant concepts and relates them to the more familiar basis of knowledge about the safety and the hazards of other advanced nuclear power reactor systems such as the LMFBR. Assessments of doses to be expected after the release of tritium from HIF reactor plants normally and accidentally - are performed and compared with dose limits and with doses resulting from facilities of the fission fuel cycle. Needs for safety related research and development specifically for inertial confinement fusion as well as for the modelling of the various exposure pathways due to released tritium are pointed out. 1. Introduction

Releases into the environment of radioactive materials contained in future commercial fusion breeder reactor plants must be prevented by similar safety design concepts as they are applied to fission converter (e.g. Light Water Reactors) and liquid metal cooled breeder reactors (LMFBR's). The safety concept of present commercial fission reactors is based on the concept of accident prevention by protection systems and in addition, on the concept of multiple containment barriers between the radioactive materials and the environment. Present safety regulations and siting criteria for nuclear power plants will certainly also have to be applied to future commercial size fusion breeder reactors. The amount of tritium and other radioactive materials contained in future commercial size fusion breeder reactors does not depend much upon their principle of plasma confinement. Reactor design studies with magnetically confined (Tokamak or Mirror principle) or inertially confined plasmas (Laser, Ion beams) show by and large the same amounts of radioactive inventory. This study concentrates primarily on the safety aspects of an inertial confinement fusion power plant concept and relates them to the more familiar basis of knowledge about the safety and the hazards of other advanced nuclear power reactor systems such as the liquid metal cooled fast breeder (LMFBR). Radioactive materials contained in an inertial con-

finement fusion breeder reactor plant are: - tritium in the fuel cycle of the plant, - activated debris material from burning pellets within the reactor cavity, - activated structural and shielding material of the reactor, - activated coolant in the cooling circuits, - activated structural material in beam channels and in the shielding of the focussing magnets. The H I B A L L fusion reactor design study is used as a reference for an inertial confinement fusion reactor system [1,2]. The results of I N T O R and N E T [3], are used for the comparison with magnetic confinement fusion. In many respects, a safety and environmental impact analysis of an inertial confinement fusion reactor study can only be very preliminary, as the necessary level of design detailing is by far not yet achieved. Hence, the study will concentrate more on inherent features and on the potential sources for release of radioactive and toxic materials as well as on the potential sources of destructive energy releases.

2. T h e H I B A L L plant design concept

The Heavy Ion Beam Inertial Confinement Fusion Reactor Study " H I B A L L " assumes four reactor chambers with a net electric power of 3.8 GW. The reactor

0 0 2 9 - 5 4 9 3 / 8 7 / $ 0 3 . 5 0 © E l s e v i e r S c i e n c e P u b l i s h e r s B.V. ( N o r t h - H o l l a n d Physics P u b l i s h i n g D i v i s i o n )

476

6. Kessler, A. BaFer : .~,q/etva,sl:ect.~a~M e:~:'~r:,/~mu~na:n W m < :

chambers are each driven at 5 Hz by pellet explosions initiated by 10 GeV 2°9Bi+ ion beams. The beams themselves are generated by a linear accelerator of 3 km length and a number of beam compressor and storage rings. The high electrical power is typical for heavy ion beam inertial confinement fusion reactors, since the pulse energy of about 5 MJ required to ignite a pellet by inertial confinement can only be generated by accelerator systems which have inherently high repetition rates of 20-30 Hz. As the reactor chambers are operated at 5 Hz, four to six such chambers can be driven by a single accelerator system. Twenty beam channels guide 2.5 kA ion beams of 20 ns duration into the reactor chamber. All 20 beams are focussed by strong magnets at the reactor chamber entrance onto a spot size of about 7 mm diameter. The fuel pellets are injected at 5 Hz frequency with a velocity of 200 m / s from the top of the reactor chamber. They are hit by the beams when they reach the center of the chamber. The pellets are spherical and consist of several layers of different materials. The innermost layer is cryogenic (4 K) D Tfuel, followed by a pusher ablator and a tamper shell of Pb~3 Li~7 and lead. The outer diameter of a pellet is 7.3 mm and roughly matches the focus size of the overlapping ion beams. REMOVABLE SHIELD SEGMENT

Fig. 1 shows a schematic view of a reactor chambc~ which has a diameter of 14 m and a height of 12 m The chamber tank is surrounded by a 40 cm thick lead--lithium cooled steel reflector and 350 cm of water cooled concrete. On the inner surface, the steel tank i., protected by the coolant breeder (Pb~3Li ~) streaming through hollow plate structures and pipes braided of SiC fiber. These so-called INPORT tubes [4] allow the bulk of the coolant to flow from the top to the bottom of the reactor cavity in a controlled way. At the same time, some of the coolant will penetrate and wet the outer surface of the tubes. The thickness of the liquid film on the outer tube surface is sufficient to absorb the X-rays and the debris from the exploding target, thus effectively protecting the structural components. The l0 m long INPORT tubes are arranged in two tube banks. The inner bank contains 3 cm diameter tubes ni a tight arrangement to remove the high heat flux caused by X-rays and ion debris (175 W / c t n z). The second mbc bank is made up of 10 cm diameter tubes and will remove most of the kinetic energy of the neutrons and provide the tritium breeding. The thick region of both tube banks reduces the displacement damage caused by the neutrons in the steel wall of the reactor chan~ber The reactor chamber is held at i0 ~ Torr vacuum m order to lower the interaction of the Bi ~ ion~ with the 12! /PELLET INJECIOR

PUMPS

SUPPLY~

UPPER BLANKET

REFLECTOR

J

,I L

L- -%~:-...... -T: "

BEAM

SiC TUBES SHIELD

2

[

4 -~ _....

-

j

~,

7 ~

'! , \

!

[

LOWER BASIN

J" /

Fig. 1. HIBALL reactor chamber: Cross-section of the layout.

FINAL FOCUSING OUADRUPOLE I

3 ~3i

G. Kessler, A. Bayer / Safety aspects and environmental impact

atmosphere of the reactor chamber and to allow the tritium diffusion out of the coolant. Vacuum pumps are provided to reduce the chamber pressure between target explosions below 10 4 Torr. The innermost INPORT tubes will withstand the radiation load for about two years. For their replacement the vessel cover can be rotated. It contains a removable plug through which both the ceiling structure and the wall elements can be removed and replaced individually. The final focussing magnets have to withstand the radiation damage and nuclear heating caused by fast neutrons. The magnet design in HIBALL uses either normal or superconducting magnets. Normal conducting coils can be used close to the beam pipe and to the reactor, whereas the power saving superconducting coils are arranged in better protected locations. The coolant is heated from 330 to 500°C in the reactor chamber. Four primary pumps take the hot coolant from the lower pool within the reactor chamber and push it through four intermediate heat exchangers (IHX) and back to the upper inlet ports of the reactor chamber. In the IHX the coolant (Pb83Li17) transfers its heat to a secondary sodium circuit with four secondary sodium pumps and four sodium heated H20 steam generators and superheaters. The steam drives a turbine/generator system. The overall thermal efficiency is about 41%. The beam lines, the primary and secondary coolant circuits axe arranged in steel clad concrete walls. To avoid coolant/air reactions in case of leaks in pipes or other circuit components, these concrete cells can be filled with nitrogen. The whole reactor chamber is surrounded by a double containment with a low leak rate to minimize tritium releases into the environment. The pellet factory contains the storage facilities for deuterium and tritium containers, the manufacturing line for the pellets, a tritium and deuterium filling station and a storage facility for the cryogenic DT pellets. Transport channels from the pellet factory lead to each reactor chamber. The pellets must be cooled up to the time when they are injected into the reactor chamber. The pellet injector is a pneumatic gun using deuterium as a propelling gas. Synchronisation of the pellet during its flight to the center and the accelerator beams is achieved by a laser light pellet tracking system which supplies the necessary signal to the driver rings releasing the ion beams.

477

3. Radioactive inventories of HIF power plants In the following are discussed the inventories of the different radioactive materials as they exist or are being generated in different components of the ICF reactor plant. For tritium, we assume a closed cycle with constant inventory. For the other radioactive species, activities at shut-down after two years of full-power operation are given. At the time, the more important nuclides will have saturated and radioactivity build-up will continue only at a slower rate. 3.1. Tritium inventory

Within the reactor cavities the thermal power is generated by igniting and burning multilayer spherical targets containing about 4 mg of D T. The shot frequency of a reactor cavity will be about 5 Hz for technical constraints explained in [1,2]. This leads to a required tritium supply of about 1 k g / d in one reactor or about 4.1 kg/(GWed ) for the total power plant consisting of 4 reactors (1 kg of tritium has an activity of 0.96 × 10 v Ci). Only a fraction of about 30% or 0.3 kg/(GWcd ) of the tritium can be burned during the fusion process within the pellet. Thus, about 0.7 kg/(GWed) of unburned tritium must be handled by the exhaust and vacuum pump systems of each reactor cavity together with about 0.87 kg/(GWcd ) which will be continuously bred in the blankets (breeding ratio of 1.25). The solubility of tritium in the coolant and breeder material determines how much of it will remain in the cavity atmosphere and how much will enter the coolant. Due to the low solubility of tritium and deuterium in Pb83Li17 used as liquid breeder material and coolant at operating temperatures of 330-500°C and at 10 -4 Torr vacuum conditions in the reactor cavity, most of the tritium generated by breeding will enter the cavity atmosphere. From there, it will flow to the exhaust processing system through large openings in the upper part of the reactor cavity together with the remaining deuterium, the 4He produced by fusion, 6He from the 6Li(n,p) reaction, and with some other impurities. The reactor cavity exhaust is pumped by compound cryopumps with a tritium inventory of about 0.1 kg/GWc. The compound cryopumps are regenerated so that helium is released first. Then, deuterium and tritium are released and sent to a clean-up unit where impurities are removed. The clean-up unit has a tritium inventory. of only about 10 g/GWe. The remaining deuterium and tritium is collected from all four reactor cavities and is finally separated in a cryogenic distillation unit (0.08 kg T inventory) to form a purified D - T stream for target

G. Kessler, A. Bc(ver/ SaJety aspects and ent,ironmental in pa( ~

478 Table 1 Tritium inventory per GW¢ ~) Plant component

Contributor

Blanket

LilTPbs3 (coolant and breeder material) Coolant guide tubes (SIC) in blanket section

(l.003-0.255

LilTPb83 Cryopumps Clean-up unit Isotope separation

0.119 0.1 0.01 0.02

Targets (one full-power power supply) Storage in uranium beds Targets in fabrication

1 1 1

Primary coolant circuits Tritium cycle Target fabrication

Tritium inventory (kg/GW~) 0.004

") 1 kg corresponds to an activity of 0.96 × 107 Ci

manufacturing and a pure D 2 stream for the target injection system. The amount of tritium to be stored in the target manufacturing station depends essentially upon the target manufacturing process and upon whether the filling of the targets will have to be done batchwise or in a continuous process. The number of targets to be produced will be about 3.2 x 10 5 targets/(GWed ). For the H I B A L L plant with 4 cavities, this amounts to 1.3 × 10 6 per day (5.2 kg D - T ) . A continuous fill and fabrication process is assumed, as it leeds to the lowest requirement for D - T inventory. In this process, cryogenic spherical targets will be transported past sputtering guns which apply the consecutive spherical layers onto the spheres. At the same time they are cooled by a cold H e h u m gas jet [5]. Table 1 summarizes the tritium inventories within a reactor plant like HIBALL. By far the highest tritium inventory will occur in the target fabrication facility, where a one day supply of cryogenic targets to fuel the reactor cavities and a oneday tritium supply in uranium beds (prior to target tilling) have been assumed. The tritium inventory within the liquid blanket and breeder material Pbs3Li17 will be relatively small (see table 1). Due to the low solubility of only 7 × 10 2 w p p m T in the coolant the tritium inventory amounts only to about 4 g / G W e or 4 × 1 0 4 C i / G W e. A similar inventory of 3 g / G W c was assumed for the coolant guide structures ( I N P O R T units) in the H I B A L L study.

Table 2 Radioactivity inventory in a reactor cavity due t~ neutron activation of Pb Isotope

Activity (el/OWe)

Half life

2°3Hg 2°SHg 2°4T1 a°3pb -~°sPb 2m Po ~)

1.2 × 106 1.1 × 106 0.4 × 105 0.6 × l0 n 0.5×102

47 d 5.2 min 3.78 y 52 h 3× I0' y 138 d

Total

6.2 × 10 v

1.6 × 10 2

~ From 40 atom-ppm Bi impurity.

However, more recent estimates uncovered larger uncertainties of the tritium solubility in the SiC material of the I N P O R T structures. These estimates extend to as much as 255 g / G W e. The tritium inventory in the primary coolant circuits is estimated here to be 119 g/GW~.

3.2. Other radioactive inventories of the reactor cavity Neutron induced activation of Pb83Lil7 leads the radioactivity inventories as shown by table 2. The highest activity with 0.6 × 10 8 C i / G W ~ is contributed by 2°3pb which has a half life of 52 h. Bismuth, a c o m m o n natural impurity in lead, would be activated to 2t0 Pb. In addition, radioactive corrosion products would have to be taken into account as radioactive materials. N o data are presently available for these. The additional bismuth brought in by the ion beams can be neglected. In addition, 6He and SLi are produced from lithium. They contribute to the radioactive inventory of the operating plant but not to the release hazard because of their short half-lives of less than a second [1,2].

3.3. Radioactivity from the burning pellet The H I B A L L pellet contains, besides D - T , only materials that are also present in the coolant (Li and Pb) and thus produces the same radioactive nuclides. Quantitatively, the pellet radioactivity adds a negligible amount to the coolant radioactivity.

3.4. Radioactivity of structural design parts Radioactivity will also be induced by neutron capture in the structural material of the blanket, the first

G. Kessler, A. Bayer / Safety aspects and environmental impact steel wall, the reflector, and shielding of the reactor. Calculations show that the buildup of radioactivity due to short living isotopes will saturate after about 2 years operation. The radioactivity level at shutdown was estimated to be 0.6 Ci//~Vth or 1.5 × 109 C i / G W e. The bulk of this activity is due to neutron activation of steel structures. It decreases relatively slowly requiting about three weeks to be reduced by a factor of 10, and two years to be reduced by a factor of 100. The radioactivity in the shield is significantly lower than the radioactivity in the reflector. However, its value of 6.3 × 10 3 Ci/tVVthor 1.6 × 107 C i / G W e is still significant.

4. The importance of different radioactive materials The confinement of tritium within the reactor plant is one of the most difficult technical problems of fusion breeder plants. Tritium permeates relatively easily through steel walls to the environment. Fortunately, the high inventory of about 12 kg T in the target factory of the HIBALL plant will be stored either at cryogenic temperatures within the pellets or safely absorbed in uranium beds. Only about 0.5 kg or 5 x 106 Ci are in the coolant circuits or in the fuel cycle from where a fraction can permeate to the environment. Special design measures must be taken to preven excessive tritium releases. The 2°3pb activity of 2.4 × 108 Ci in the coolant of the whole HIBALL plant must be taken into consideration primarily in the accident analysis. A certain percentage of this radioactivity could reach the environment as aerosols. It can hardly be envisioned how many major portions of the radioactivity accumulated in the blanket structures, the first wall, and the reflector structures could be released into the environment even in an accident. However, this radioactivity level will be important for considerations on remote handling and maintenance operations of blanket sections and for waste disposal questions. 4.1. Permeability of tritium into the steam cycle The permeation of tritium from the Pbs3Li17 through the walls of the steam generators presents a difficult technical problem, as the permeability of tritium through steel and the heat transfer areas of s t e m generators are large enough so that considerable quantities of tritium can reach the steam cycle and the environment. If one assumes tritium permeabilities as given in the literature [1,2,6,7] and a heat transfer area of about 1 0 4 m2 for a

479

Table 3 Tritium released from the HIBALL plant Ci/(GW~d) Intermediate coolant circuit (sodium) Double walled steam generators Water coolant circuits of shield Cryopump system Fuel clean-up unit Isotope separation unit Target factory Buildings and tritium recovery systems Total release

3 5 1 3 2 1 5 2 22

rector plant like HIBALL, one obtains a leakage rate of about 105 Ci/(GWed). This is by a factor of about 103 higher than can be allowed. Therefore, a diffusion bartier of about 103 to 10 4 for tritium is needed, which can only be achieved by either double walled steam generators as they are investigated presently also for LMFBR applications or by an intermediate liquid metal circuit as presently already applied in LMFBR's. Estimates summarized in [2,5] have shown that double walled steam generators with a diffusion barrier concept of a quality of 103 could be sufficient to keep the permeation of tritium into the steam cycle below about 5 Ci/(GWcd). However, experimental results would be needed to prove this concept. If an intermediate liquid-metal loop according to standard LMFBR technique is employed instead of the Duplex steam generator technique, again diffusion bartiers of a quality of 103 are required. A design study [8] showed that. this can be met with eight intermediate heat exchangers and 24 sodium/water steam generators. The total losses of tritium from the intermediate circuits (with sodium as intermediate coolant) would then be about 2 Ci/d. The primary coolant circuits consisting of pumps, valves, pipings, flow meters, storage or hold-up tanks, purification systems etc. will have to be designed with aluminium sleeves to provide a secondary containment barrier against tritium permeation. Pumps, valves or similar components where leakages may occur will need additional jacketing or glove boxes to keep tritium leakages low. Tritium leakages from the cryopumps, clean-up and isotope separation systems are difficult to estimate. If one uses similar release rates as estimated for the Tritium Systems Test Assembly (TSTA) at Oak Ridge [9] and if one applies an adequate scaling up to a power plant like HIBALL, one obtains tritium release rate of 5 Ci/(GWed).

480

(7. Kessler, A. Bayer / Safe(v aspects and environmental impa~ t

The buildings in which tritium is handled will be equipped with a tritium recovery unit. In this unit, the tritium will be catalytically oxidized and then be absorbed on molecular sieves. Tritium losses due to building leakage are, therefore, expected to be small and in the range of 2 C i / ( G W e d ). The tritium leak rate from the target factory is difficult to estimate as no operating target factory exists. A loss rate of about 5 C i / ( G W c d ) of tritium from the target factory appears to be permissible. Table 3 summarizes the tritium losses from a fusion reactor plant like HIBALL. The total losses are estimated to be 22 Ci/(GW~d). Of these total losses, 80% are expected to go into the atmosphere and 20% into water.

5. Chemical reactions of the coolant Pbs3Li 17 and sodium with air and water

The L i - P b in the first wail and the p r i m a l coolanl system as well as sodium in the intermediary circuits are a potential source for energy release in two respects: - chemical reactions with air, water and concrete are possible, - intimate mixing with water may generate high pressures due to rapid steam generation. It is considered to be one of the great advantages of Pba3Livl over other liquid metals that it is chemically much less agressive, e.g., than sodium or hthium [10]. The principai technique to avoid the potential of large

Fig. 2. HIBALL reactor building (940 MWe) and beam lines. 1 Reactor chamber 2 INPORT blanket 3 Final focusing quadmpole magnets 4 Rotatable top shield 5 Coolant exit 6 Primary pump 7 Heat exchanger 8 Secondary coolant pump 9 Steam generator 10 Water intake

11 Steam exit 12 High pressure turbine 13 Low pressure turbine 14 Electricity generators 15 Condenser and water preheater 16 Beam lines 17 Target transport line 18 Target factory 19 Reactor containment 20 Machine building

G. Kessler, A. Bayer/ Safety aspects and em~ironmentalimpact scale thermal reactions between Li-Pb and water is the provision of the two walls between the two liquids. The double walled steam generator can account for this requirement in the coolant cycle. If intermediary sodium circuits would be used direct contact of LiPb with water after failure of steam generator tubes would also be avoided. Safety design provisions will have to be applied in both cases in order to master pressure peaks as a consequence of sodium-water or LiPb-water reactions. For sodium-water reactions these design solutions were developed already for the LMFBR technology.

6. Containment philosophy and requirements

The general safety philosophy applied to nuclear reactor design is to protect the environment from the radioactivity contained within the reactor chamber by several leak-tight barriers. As the tritium inventory in fusion reactors is by almost three orders of magnitude higher compared with fission converter and breeder reactors, the requirements for leak-tightness of the containment barrier must be high. This inner containment (fig. 1) consisting of concrete cells with steel lining must be surrounded by an outer containment which may be air filled but has to be designed to guarantee high leak-tightness, typically 0.25 vol%/day leak-rate as is the present standard technical requirement for LWR's [11]. This outer containment must be kept at a somewhat lower pressure against the environment so that air can only leak from the outside into the containment. The exhaust air of the outer containment is carried via filter systems into a stack and to the environment. Filter systems become necessary to retain aerosols which could develop in the case of chemical reactions of the liquid metal coolant (PbLi or sodium) with the oxygen in the containments. The outer containment must also be equipped with a tritium removal system which consists of several catalytic recombiners, where tritium contained in the air is recombined with oxygen. The resulting T20 can be extracted after cooling down in a heat exchanger, while cooler air is returned to the containment. Fig. 2 shows the main components of such a containment system.

7. External events

7.1. External missile protection External missiles may be generated from aircraft crashes, chemical explosions at or near the reactor site,

481

or may result from other mechanically accelerated objects. Containment buildings are designed to accomodate such loadings and to avoid penetration of the containment barrier. This is standard practice for all nuclear reactors. HIBALL does not present any particular problems in this respect, 7.2, Natural disasters Natural disasters, such as earthquakes, floods and extreme wind loads, need similar consideration and analysis for inertial confinement fusion reactors as for other fusion and fission reactors. A safe shut-down of the plant and a reliable functioning of all afterheat removal systems must be guaranteed. One area in operational safety and reliability of the plant possibly requiring special treatment is the driver system. Because of its large spatial dimensions, preventive measures against flooding with water and the guaranty of proper geometrical alignment after earthquake loading will have to be investigated in more detail. The long beam channels and their connections at the reactor cavity are particularly vulnerable to earthquakes.

8. Accident analysis

A first step in the assessment of safety and environmental effects for fusion power reactors is the identification of potential sources of hazard to the public's health and safety. 8.1. Primary coolant system accidents The pressure in the blanket (i.e. in the SiC-tubes of the so-called INPORT units) is low since the coolant is essentially falling under gravity through the woven SiC tubes. A preliminary investigation of the coolant system failmes must concentrate on events which will occur in the reactor chamber as a consequence of such failures as:

- loss of coolant flow, pressure pulses, - coolant flow blockages, - leaks and ruptures. -

8.1.1. Loss of coolant flow In the event of total and instantaneous loss of primary coolant flow from a HIBALL reactor coolant would not flow anymore through the INPORT tubes. The evaporation-condensation heat transfer mechanism which both protects the SiC-tubes from excessive heat

482

G. Kessler, A. Bayer / Safety aspects and ent,ironmental impact

loads and provides heat removal from the cavity, would become inoperable. The liquid Pb83Lilt would no longer protect the steel wall and shield from the radiation and energy release from the target.

8.1.2. Loss of seconda~ or tertiary coolant flow Loss of coolant flow in the secondary or tertiary coolant circuits would cause a loss of heat sink incident. If the accelerator and the pellet delivery would not be shutdown immediately the energy still being dumped into the primary coolant and blanket would produce a rapid temperature increase. Two effects then have to be considered: subsequent loss of coolant from the blanket due to vaporization, - build-up of a Pb83Li17 vapor atmosphere in the cavity due to the lack of recondensation capability on the blanket tubes. -

8.1.3. Pressure pulses Potential sources of pressure pulses are: - shock waves produced by closure of isolation valves, liquid-metal coolant/water interaction in the steam generator. Experimental experience from LMFBR safety investigations shows that pressure spikes of 10 MPa occur in liquid metal/water reactions, especially at contact of sodium with water. Reactions of P b - L i with water will be smaller but will have to be investigated in large scale experiments as it has been done for sodium/water steam generators. -

8.1.4. Coolant flow blockages Coolant flow blockages are a point of concern. Partial blockages would cause the operating temperatures to be exceeded and coolant evaporator as well as two-phase flow to be generated. The stability of such a flow needs more investigation. Total blockages would cause local vaporization, subsequent bum-out and likely failure of the corresponding INPORT unit. Monitoring the coolant flow (outlet temperature) of each coolant flow path individually is strongly recommended. 8.1.5. Leaks and ruptures In the discussion of coolant system ruptures, we must distinguish between those leaks which cause loss of liquid from the coolant system, and others which cause undesired by-passes but without integral loss of material from the system. Leaks out of the system present a potential radioactivity release mechanism and may cause coolant freezing in areas where it may have considerable effect on later repair operations. For pre-

venting and monitoring such leaks, the same techniques as in LMFBRs apply, up to and including double wall design in inaccessible areas. Leaks within the coolant system do not immediately present a potential for radioactivity release. However, they cause the coolant flow to concentrate in areas where it is not expected, and cause reduction of cooling capability in areas where such reduction may lead to damage. An example of this type of leak is the failure of an INPORT unit inside the reactor cavity. Significant difficulties in detecting such failures are envisaged. As a precaution, monitoring the outlet temperature (if not the flow rate itself) of each INPORT tube individually will have to be considered.

8.2. Cryogenic system failures Failures of the cryogenic system have to be considered with respect to the following potential consequences: 1. Failure of cooling of the superconducting magnets with subsequent release of the magnetic energy resuiting in mechanical forces able to perform destructive work. This chain of events is much less severe in HIBALL than in a tokamak-type reactor, e.g., because HIBALL has much smaller superconducting coils. 2. Evaporation of the cryogenic liquids (He and N2) leading to pressure build-up in the containment buildings. Again, this problem is less severe than in a tokamak reactor of similar size because of the smaller amounts of cryogenic fluids available. 3. Subcooling of structural components inducing thermal stress while at the same time causing embrittlement of the material. Such sequences of events would have to be studied in more detail, but no fundamental difficulties in dealing with this problem are envisaged. 4. Explosive vaporization of cryogenic liquid upon contact with a hotter fluid like water or lead-litl-fium. Here, more design details are required to analyze this possibility properly on the basis of plausible mechanisms, for such an event. Also for tokamak type reactors, this possibility should be investigated more intensively.

8.3. Hydrogen fires Most of the deuterium and tritium will be stored in the target fabrication area. Leaks in storage facilities and pressurized containers could lead to air-deuterium or air-tritium mixtures having the potential of explo-

G. Kessler, A. Bayer / Safety aspects and environmental impact sions if hydrogen concentrations above 4.1% are attained. Application of inert gas atmospheres will decrease the potential of explosions somewhat by increasing the lower limit for the critical concentration. More effective means will be double containment for pressurized containers and storage facilities as well as the application of hydrogen recombination devices in areas where leaks might develop. 8.4. External events (to the facility) External events (to the facility) fall into three categories: 1. natural disasters (earth-quakes, strong winds and floods), 2. aircraft crashes, 3. gas cloud explosions and 4. events caused by intentional human intervention. It would be much too early to undertake a quantitative risk assessment. In such a later assessment, two aspects would have to be considered in more detail: 1. the effect of combining the power plant together with the reprocessing and fuel pellet manufacturing plant on one site, and 2. the large area of the site (about 2 by 2 km, with an appendix of about 3 km by 500 m for the linear accelerator) which makes it perhaps more susceptible to any kind of external event. The reactor cavities are operated at 10 _4 Torr, whereas the beam channels and storage rings are at 10 9 Torr. Any leaks in the beam channels caused by channel rupture, e.g. by earthquakes or intentional human intervention would open also the reactor cavity to air. Air or water ingress to the reactor cavity and to the cryopump system through the beam channels could give rise to coolant/air or coolant/water reactions and fires. Air craft crashes and gas cloud explosions will be mastered by strong outer containments as it is done in present fission reactor plants.

9.

E n v i r o n m e n t a l

a n a l y s i s

9.1. Radiological exposure and exposure pathways Radioisotopes normally or accidentally released from nuclear facilities into the atmosphere are, due to transport and transfer processes, present in the various geophysical and biological media. Via breathing of air and consumption of food they are also incorporated by man. The radiation of the decaying radionuclides which are present in these various media can reach man and

483

his organs and tissues from outside whereas the radiation from the incorporated radionuclides a~fects the organs and tissues directly from inside. The various ways via which radiation may reach man are called exposure pathways. In general, the following exposure pathways have to be taken into consideration with regard to the release of radioactive material with the exhaust air: external exposure by beta radiation inside the exhaust air plume (beta submersion); external exposure by gamma radiation from the exhaust air plume (gamma submersion); - external exposure by gamma radiation from contaminated ground (ground radiation); internal exhaust by radionuchdes as a result of breathing of air (inhalation); internal exposure by radionuclides as a result of consumption of food (ingestion). To assess the doses which result from the release of radionuclides mathematical models are established for the various exposure pathways which describe the transport and transfer processes of the radionuclides and the associated transport and absorption process of radiation. The main steps of the models can be described as follows: D=g.U.T.x.S

,

where D = dose, g = dose factor, U = rate of interaction with man, T = transfer factor for transfer from one media into the other, X = dispersion factor of the transport media, and S = source strength. The actual computational procedures for the various pathways are partially very detailed and complicated and a lot of information is required about the environment and man. 9.2. Exposure pathways relevant to tritium As tritium is a pure r-emitter and the r-energy is very low, only the internal exposure pathways (inhalation and ingestion) have to be considered. The calculations performed follow mainly the procedure recommended by the German Advisory Committee on Radiation Protection [12-14] for the conservative assessment of the radiological impact due to nuclear power plant, completed by some parameters taken from [15]. Some additional peculiarities will be pointed out below: For the "vegetable" product or "vegetable" intermediate product (e.g. grass) the intake of tritium via the air (humidity of the air) as well as via the ground (rainwater) is considered [14]. In the models applied it is postulated that tritium will be absorbed by plants, in relation to natural

484

G. Kessler, A. Bayer / Safe(v aspects and em;ironmental impa~ l

hydrogen, at the ratio prevailing in the air or the rainwater resp. at the place under consideration [12,141. - For the assessment the two age groups "adults" and "infants" are considered [12,13]. - In the case of inhalation not only the inhalation via the lungs is considered but also the inhalation via the skin which is approximately one half that by the lungs [!6]. In the case of normal operation of the facility the consumption of green vegetables, other plant products, milk, and meat are considered for the age group "adults". For the age group "infants" the consumption of milk is considered [12,13]. In the case of accidental release the consumption of meat is neglected due to the relative short residence time in animals. Other food is considered as well [14]. - As tritium is distributed in the human body nearly homogeneously the dose factors for the various organs are identical [17,18]. -

9. 3. Tritium release characteristics Most of the tritium released from a fusion plant will be discharged as HTO. It is assumed here that all tritium is released as HTO; this is a conservative assumption from the radiological point of view. In the case of normal operation the daily release is shown in table 3. Assuming a load factor of 80% the annual release with the exhaust air amounts to approx. 5000 C i / G W e a = 1.9 × 1014 Bq/GWea. It is assumed that this amount is released alternatively via a stack of 100 m and a stack of 200 m, respectively. In the case of an accident hypothetically a release of 0.5 kg tritium is assumed. This amounts to approx. 5 × 106 Ci = 1.9 × 1017 Bq. Furthermore, it is assumed that the tritium witl be released from the top of the building of the facility, the height of which might be 50 m. The release duration is assumed to be one hour. 9.4. Weather condition characteristics The atmospheric dispersion of the released tritium depends on the prevailing weather conditions. For the assessment of the radiological impact due to the release during normal operation a hypothetical site with an annual average windspeed of 2.5 m / s (at a height of 10 m) was assumed. This average wind speed is representative for Southern Germany. For Northern Germany this value has to be doubled. The doubling results in doses which are one half of the computed doses. Furthermore a uniform wind rose was assumed.

Details about the atmosphere dispersion categories are in table 4. For the assessment of the radiological impact after an accidental release weather situations as recommended by the German Advisor, Committee on Radiation Protection [13] for the conservative assessment of the radiological impact were applied. The characteristic data of these situations are shown in table 5. 9.5. Radiological exposure On the basis of the models and data mentioned in the paragraphs above, doses were calculated for the reference persons "adult" and "infant". The obtained doses for the case of release during normal operation are shown in figs. 3a and 3b and those for accidental release in figs. 4a and 4b. From figs, 3a and 3b (normal release) two general statements can be derived the exposure pathway "ingestion" contributes the overwhelming portion to the total dose, the doses for "adult" and "infant" do not differ v e ~ much. The doses at a reference distance from the source of 1 km (distance: stack of the facility - fence of the plant) amount to adult: stack height H = 100 m: 0.23/zSv/a = 0.023 m r em / a, H = 200 m: 0.19/~Sv/a = 0.019 mrem/a. infant: stack height H = 100 m: 0.25 # S v / a = 0.025 torero/a, H = 200 m: 0.21 ~ S v / a = 0.021 mrem/a. These doses are well below the limit of 30 m r e m / a imposed by the German Radiation Protection Ordinance [19] for release during normal operation. Beside the comparison with the legal limit it makes sense to compare the calculated exposure also with the exposure resulting from the releases of facilities of the fission fuel cycle. This is done again on a 1 GWe basis for a Pressurized Water Reactor (PWR) and the respective reprocessing capacity. The whole body doses assessed for adults [20] are shown m fig. 3a, too. The doses assessed for the fission facilities are numerically a little bit higher. But in making such comparisons, however, it should be recalled that the results for a PWR and its respective reprocessing capacity are based on realistic i.e. measured release data, whereas the estimates for a ICF plant depend on the

G. Kessler, A. Bayer / Safety aspects and environmental impact Table 4 Annual weather data [15] Dispersion category i)

unstable

D neutral stable

485

Table 5 Selected weather data [13]

Probability (%)

Average wind speed ~) (m/s)

Dispersion category

2 8 16

2.1 2.8 3.O

i}

42

2.9

D neutral

15

1.7

E F}

Wind speed ~) (m/z)

Precipitation intensity (mm/h)

0.9 1.3 1.7

unstable

2.0 stable

1.2 0.4

D, rainfall

2.0

a) At a height of 10 m. Rain duration: 10% of the year, precipation intensity: 1 ram/h, dispersion category: D.

~) At a height of t0 m.

assumptions made for the diffusion barriers in the steam generator tubes and other permeation rates of tritium in different parts of the plants.

Figs. 4a and 4b (accidental release) show the doses below the plume centerline. Here the same general statements concerning the contribution of the exposure

10

,

'

I

f

_ _~_.~

~

= I ~ , age group : adult exposure pathway: inhalation, I H ingestion, IG D~v. (H=10Om)

I

~o ,~ 10_~ E

/ D~G(H: 100m)

-D =G(H=lOOm)

> U3

E

.10-1

-2

o~ O O

D'G(H:200m)

Om(H=20Om)

I 10-2

D~(H=100rn)

<.=100mll.:200m110 D J

L

lo-'1-

10 2

1

age group : infant exposure pathway: inhalation, IH ingestion. IG

i0 -2

10"~L

I

- 1(]-I

/

i

T

I

(H:20Om) Ln :~ 10-~1 O~ m

10

11

/

~

I

I I(] 3

/,

OlH

10-~I

I

I /~ 10 ~

~

1105 10 s

Distance, m

Fig. 3a. Distance dependent doses for the age group "adult" due to normal release of tritium for the stack heights 100 m and 200 in.

10"41 I 10 2

I

;

~

110-~

I I 10 ~

=

~ kllO-S 105

[ /= 10 ~ Distance, m

Fig. 3b. Distance dependent doses for the age group "infant" due to normal release of tritium of the stack heights 100 m and 200 m.

-

486

G. Kessler, A. Bayer / Safe(), aspects and ent,ironmental m~pact

10

1000

10

-~!QO0 i:

100

10"

D[ ~

i

+ loo

1

lo

m-

I

10

E

f

o~

o rq

\,

10<

1'0 2

1'0 3

1'0 4 Distance,

1~ ~

10 2

i:

-10 i

10"3-

m

m

10 3

1'0 ~ Distance,m

10 s

Fig. 4a. Distance dependent doses for the age group "adult" due to accidental release of tritium for various atmospheric dispersion categories.

Fig. 4b. Distance dependent doses for the age group "infan['" due to accidental release of tritium for various atmospheric dispersion categories.

p a t h w a y s to the total dose a n d the difference between doses for adults a n d infants can b e derived, too. The m a x i m u m doses b e y o n d the reference distance from the source of 1 k m a m o u n t to adult: 2.4 Sv = 240 rem, infant: 3.1 Sv = 310 rem. W h e n evaluating these doses for accidental release the following two facts should b e considered: • The a s s u m p t i o n s a b o u t the source term are not b a s e d o n a deterministic analysis b u t r a t h e r represent a postulated conservative u p p e r b o u n d for radioactivity entering the e n v i r o n m e n t . M o r e realistic analysis m a y lead to substantially lower releases. • The dose assessments are m a d e o n the basis of a conservative specific activity model. D u e to the fact that the periode of the assumed puff release is short c o m p a r e d with the exchange times for H T O between air a n d p l a n t n o t the full equilibrium activity will actually be reached in the p l a n t b u t only a few tenths of it.

-

-

-

T h e a s s u m p t i o n " i m m e d i a t e harvesting after passing of the cloud" would be fulfilled at the most by the p r o d u c t " g r e e n vegetables". Taking into account only this product, the m a x i m u m doses due to ingestion would be reduced to adult: 0.12 Sv = 12 rem, infant: 0 Sv = 0 rem. Moreover, a reasonable and simple countermeasure n a m e l y " a b a n d o n m e n t of harvesting for one day" would reduce the exposure p a t h w a y " i n g e s t i o n " to u n i m p o r t a n c e . T h e n only the exposure p a t h w a y " i n h a l a t i o n " remains. F o r that exposure p a t h w a y the m a x i m u m doses are adult: 22 mSv = 2.2 rem infant: 11 mSv = 1.2 rem. These investigations show the need of a more realistic modelling of the exposure p a t h w a y " i n g e s t i o n " connected with the release - especially of accidental release - of tritium.

G. Kessler, A. Bayer/ Safety aspects and environmental impact Summary This study identified significant safety aspects of inertial c o n f i n e m e n t fusion power p l a n t concepts a n d related them to the more familiar basis of knowledge a b o u t the safety a n d the hazards of other advanced nuclear power reactor systems such as the L M F B R . Assessments of doses to be expected after the release of tritium from H I F reactor plants - normally a n d accidentally - were p e r f o r m e d a n d c o m p a r e d with dose limits and with doses resulting from facilities of the fission fuel cycle. Needs for safety related research and d e v e l o p m e n t specifically for inertial c o n f i n e m e n t fusion as well as for the modelling of the various exposure p a t h w a y s due to released tritium were pointed out.

Acknowledgement T h e authors wish to express their sincere appreciation for the support they received from Dr. E.G. Schlechtendahl, co-author of the precursor report " G . Kessler, E.G. Schlechtendahl, Safety Aspects of a n Inertial C o n f i n e m e n t F u s i o n Reactor, KfK-3834 (1985)" a n d Dr. U. v o n M~511endorff; they also t h a n k Mr. J. B r a u n for establishing the p r o g r a m for the c o m p u t a t i o n of the relevant doses and p r o d u c i n g the numerical resuit; all are from the K e r n f o r s c h u n g s z e n t r u m Karlsruhe.

References [1] B. Badger et al., HIBALL - A Conceptual Heavy Ion Beam Driven Fusion Reactor Study, Volume 1, KfK 3202/1, UWFDM-450, Kernforschungszentrum Karlsruhe (1981). [2] B. Badger et al., HIBALL - A Conceptual Heavy Ion Beam Driven Fusion Reactor Study, Volume 2, KfK 3202/2, UWFDM-450, Kernforschungszentrum Karlsruhe (1981). [3] International Tokamak Reactor (INTOR): Phase One, IAEA, Vienna, 1982, STI/PUB/619; ISBN 92-0-131082X, Karlsruhe (1981). [4] G.L. Kulcinski et al., The INPORT Concept - An improved method to protect ICF reactor first walls, UWFDM-426 (August 1981). [5} B. Badger et al., HIBALL-II, An Improved Conceptual Heavy Ion Beam Driven Fusion Reactor Study, KfK 3840, Kernforschungszentrum Karlsruhe (1985). [6] T.A. Renner and D.J. Rave, Tritium permeation through Fe-214 C r - l M o steam generator material, Nucl. Technol. 4 (1979) 312-319. [7] J.T. Bell et al., Tritium permeability of structural mated-

487

als and surface effects on permeation rates, Proc. on Tritium Technology in Fission, Fusion and Isotopic Applications, ANS topical Meeting, US-DOE Report CONF-800427 (1980) 48-53. [8] M.E. Sawan, W.F. Vogelsang and D.K. Sze, Trans. Am. Soc. 39 (1981) 777. [9] J.L. Anderson and R.H. Sherman, Tritium Systems Test Assembly. LA-6855-P (1977). [10] L. Muhlestein et al., Summary of HEDL fusion reactor safety support studies, HEDL-SA-2360 (1984). [11] W. Braun et al., The reactor containment of German PWR's of standard design, Int. Conf. on Containment Design, June 17-20, 1984, Toronto, Ontario. [12] Allgemeine Berechnungsgrundlage for die Strahlenexposition bei radioaktiven Ableitungen mit der Abluft oder in Oberfl~ichengewi~sser (Richtlinie zu §45 StrlSchV), Gemeinsames Ministerialblatt 30 (1979) 371-436. English translation: General principle of Calculation for the Radiation Exposure Resulting from Radioactive Effluents in Exhaust Air and in Surface Waters (Guideline under Sec. 45 of the Radiological Protection Ordinance), GRS Translations, Edition 11/80. [13] St~Srfallberechnungsgrundlagen for die Leitlinien des BMI zur Beurteilung der Auslegung von Kernkraftwerken mit DWR gem~iss §28 Abs. 3 StrlSchV, Bundesanzeiger 35, Nr. 245a (1983) 19-24. English Translation: Incident Calculation Bases for the Guidelines Issued by the Federal Minister of the Interior (BMI) for the Assessment of the Design of PWR Nuclear Power Plants pursuant to Sec. 28, para (3) of the Radiological Protection Ordinance (StrlSchV), GRS Translations, Edition 7/83. [14] Modifizierung bestehender Berechnungsgrundlagen zur Anwendung bei der Ermittlung der Strahlenexposition for Anlagen des Brennstoffkreislaufs, Empfehlung der Strahlenschutzkommission (1986), in press. [15] A. Bayer, Die radiologische Belastung der BevNkerung der Rhein-Maas-Region, Bericht der Kommission der Europ. Gemeinschaften V/2475/81D (1981). English translation: The Radiological Exposure of the Population of the Rhine-Meuse Region, Report of the Commission of the Europ. Communities V/2475/81 EN (1982). [16] ICRP, Limits for Intakes of Radionuclides by Workers, Publication 30 (Pergamon Press, Oxford, 1979). [17] D. Noske, B. Gerich und S. Langner, Dosisfaktoren for Inhalation oder Ingestion von Radionuklidverbindungen - Erwachsene, Bericht ISH-63 (1985). [18] K. Henrichs, U. Elsasser, Ch. Schotola und A. Kaul, Dosisfaktoren for Inhalation oder Ingestion von Radionuklidverbindungen - Altersklasse 1 Jahr, Bericht ISH-78 (1985). [19] Verordnung iaber den Schutz vor Sch~iden durch ionisierende Strahlen (Strahlenschutzverordnung), Bundesgesetzblatt I, 28 (1976) 2905-2995. English translation: Ordinance on Protection Against Damage and Injuries Caused by Ionizing Radiation (Radiological Protection Ordinance), GRS Translations, Edition 8/77. [20] E. Lessmann, Personal communication (1985).