CA1326916C

CA1326916C - Nuclear reactor cooling system

Info

Publication number: CA1326916C
Application number: CA000611551A
Authority: CA
Inventors: John S. Hewitt; Hani C. Ajus; Thomas C. Currie; John C. Atkinson; Bruce M. Pearson
Original assignee: ECS-POWER SYSTEMS INC.
Current assignee: METCALFE REALTY Co Ltd
Priority date: 1989-09-15
Filing date: 1989-09-15
Publication date: 1994-02-08
Anticipated expiration: 2011-02-08
Also published as: WO1991004560A1; EP0570360A1

Abstract

Abstract of the Disclosure A nuclear reactor has a cooling system which operates during impairment of the reactor's primary cooling circuit.
Ports are located in the inlet duct upstream of the reactor core and in the outlet duct downstream of the core which have exchange flow channels for conducting coolant between the inlet and outlet ducts and a reserve coolant tank. During impairment of the primary cooling circuit a natural convective flow of coolant is initiated and maintained without mechanical or operator intervention between the primary cooling circuit and the reserve cooling tank, through the ports.

Description

:

.~ NUCLEAR REACTOR COOLING ~Y~TEM

Bac~ground of the Invention This invention relates to a cooling system for particular class of water-cooled nuclear power reactor said 5 reactor having (a) a high-temperature (limited by boiling : considerations) primary heat transport circuit which dominates the reactor cooling process under normal operating conditions; (b) a large pool of cool water surrounding or otherwise in the vicinity of the reactor ~:. 10 core region and forming a component of the shutdown cooling circuit which may be said to be passive; which is to say that, in the post shutdown phase of reactor operation, adequate core cooling is assured for an extended period of -i time, independently of any sensing devices, externally 15 energized components, or operator intervention, and (c) a shutdown cooling circuit which is itself passive in the initiation of its operation; which is to say that the :i shutdown cooling circuit initiates its function without reliance on sensors, externally energized components, 20 mechanical actuation, or operator action.
More specifically, the invention relates to a cooling system for a nuclear reactor of the type in which the coolant liquid is normally circulated around the primary heat transport circuit which includes the reactor 25 core, various heat exchangers, and a primary circulation , pump. Located in this circuit and at points not far above .~ and below the core are the very important aspects of this invention, namely, three-way flow branching devices herein referred to as "hydrodynamic ports", or just "ports". The :~ 30 role of the ports is to provide free flow-through passage f of the coolant in the primary heat transport circuit as i long as normal reactor operating conditions prevail, and to 3 provide at all times for sizable branch-flow when the : passive shutdown cooling circuit, which includes the pool 35 liquid, comes into operation under various abnormal . :. ,.
: ~ .

13~691~

operating conditions.
It is a feature of the ports that they perform the above functions automatically, and do not involve any sensors, externally energized components, mechanical actuation, or operator action for either the maintenance of normal operation with high performance, or for the transition to, or sustaining of, the safe shutdown condition.
The ports, when functioning in support of normal ~ 10 operation, involve hydrodynamic principles applied in an -` integrated way to the overall thermal-hydraulic configuration, but require minimal complexity of associated equipment to sustain normal as well as shutdown operation.
Moreover it is characteristic of the invention that it is applicable tc a variety of reactor plants to be deployed under unusual combinations of constraints of space, weight, mobil~ity, etc.
Although the invention relates in the first instance to water cooled nuclear power reactors, the invention can be employea to advantage in other types of reactor, including those cooled by media other than water 'l and those which may be used predominantly for research ;~ purposes.
Prior art reactors have been described (though they have not yet been constructed or operated) in which ! passive shutdown cooling is available to coveE a range of potential accident situations, and in which the primary heat transport circuit, under normal operating conditions, is maintained in thermal-hydraulic isolation from the shutdown circuit. These reactors are known as the PIUS*
(described in Canadian Patent l,173,5~0 (Hannerz); the SECURE*(described in Canadian Patent l,070,860 (Blomstrand et al.); the TRIGA* power reactor, described in R.W.
Schleicher, "TRIGA Power System: A Passive Safe Co-Generation Unit for Electric Power and Low TemperatureHeat", in: Small Reactors for Low Temperature ~eat Applications, IAEA-TECDOC-463 (International Atomic Energy Agency, Vienna, 1988) pp. 45-55; and the GEYSER*heating * Trade Mark ..... ... .
. .. .

. ~ . .. .

1 32691 ~
reactor, described in G. Vecsey and P.G.K. Doroszlai, "GEYSER, A Simple, New Heating Reactor of High Inherent Safety", Nuclear Engineering and Design, 109 (1988) 141-~ 145. The PIUS, SECURE and TRIGA reactors rely on elaborate parameter-sensing and control systems to continuously maintain the integrity of normal reactor operations in the ., face of the inevitable adverse influences of the passive ;j cooling system on normal operations. Such active control ``~ measures detract from the overall reliability and the simplicity of operation of the reactor, and aggravate the space and weight requirements. The GEYSER reactor, while responding to passive cooling and normal operating requirements in ways that avoid the use of active systems, has a physical plant size which is consequently extremely large. Also, the system has very long response times in ; respect of load following and compensation of inadvertent changes in operating conditions which, along with the large j physical size, render the concept impractical in many application environments, including that of a submarine.
.' .
8ummary o the Invention The invention provides a cooling system for use in association with a nuclear reactor plant of the type having a primary cooling circuit comprising a reactor core, an inlet duct and inlet plenum for conducting coolant into the ! 25 core, an outlet plenum and an outlet duct for conducting ~, coolant out of the core, and means for circulating coolant through the primary cooling circuit. The cooling system ~' comprises: (a) a reserve coolant tank for containing coolant: ~b) a first port means in the inlet duct upstream of the core, having a primary flow channel for conducting , coolant through the inlet duct and having exchange flow , channel means for conducting coolant between the reserve ¦ coolant tank and the primary flow channel, the primary flow channel being open to said reserve coolant tank through the exchange flow channel means; and (c) a second port means in the outlet duct downstream of the core, and spaced generally above the first port means, having a primary flow ~, "

. ~ . ~ . , ., , ~

132691~

channel for conducting coolant through the outlet duct and having exchange flow channel means for conducting coolant between the primary flow channel and the reserve coolant tank, the primary flow channel being open to the reserve coolant tank through the exchange flow channel means.
During normal, unimpaired operation of the primary - cooling circuit, coolant circulates through the primary - cooling circuit and only minimal volumes of coolant flow - into the reserve coolant tank from the primary cooling .:
circuit or into said primary cooling circuit from the reserve coolant tank through said exchange flow channel means. During impairment of the primary cooling circuit, or other circumstance of mismatch between the energy generation and adequate dissipation rates of the nuclear reactor plant, a natural convective flow of coolant is initiated without mechanical or operator intervention, ' wherein coolant in the reactor core flows by convection ,, through the outlet plenum and outlet duct and through said exchange flow channel means in the second port means into the reserve coolant tank, and coolant in the reserve coolant tank flows into the exchange flow channel means in the first port means, through the inlet duct and inlet plenum and into the reactor core, whereby a natural '7 convective flow of coolant results.
, 25 The invention also provides a cooling system for ,! use in association with a nuclear reactor plant of the type ~ having a primary cooling circuit comprising a reactor core, ',7 a plurality of inlet ducts and an inlet plenum for conducting coolant into the core, an outlet plenum and a plurality of outlet ducts for conducting coolant from the core, the inlet ducts and the outlet ducts being configured with substantially radial symmetry about a vertical axis through the center of the core, and means for circulating coolant through the primary cooling circuit. The cooling system comprises: (a) a reserve coolant tank for containing coolant; (b) a first port means in each of the inlet ducts upstream of the core, each of the first port means having a primary flow channel for conducting coolant r 1 4 , :!, , . ' . : :
"' ~ . . ' ' ~ ..

132~91~
through the inlet duct and ha~ing exchange flow channel means for conducting coolant between the reserve coolant tank and the primary flow channel, the primary flow channel ;being open to the reserve coolant tank through the exchange flow channel means; (c) a second port means in each of the ;outlet ducts downstream of the core, and spaced generally at higher elevation than each of the first port means in the normal orientation of the reactor plant, each of the second port means having a primary flow channel for ` 10 conducting coolant through the outlet duct and having exchange flow channel means for conducting coolant between the primary flow channel and the reserve coolant tank, the ` primary flow channel being open to the reserve coolant tank through the exchange flow channel means. During normal, unimpaired operation of the primary cooling circuit, coolant circulates through the primary cooling circuit and only minimal volumes of coolant flow into the reserve ,~coolant tank from the primary cooling circuit or into the primary cooling circuit from the reserve coolant tank through the exchange flow channel means. During impairment of the primary cooling circuit, or other circumstance of mismatch between the energy generation and adequate dissipation rates of the nuclear reactor plant, a natural convective flow of coolant is initiated without mechanical or operator intervention, and irrespective of the orientation of the reactor plant with respect to gravity, wherein coolant in the reactor core flows by convection generally upward through at least some of the outlet ducts and/or inlet ducts positioned at a higher elevation than the reactor core and through the exchange flow channel means in the second and/or first port means in the ducts and into the reserve coolant tank, and coolant in the reserve coolant tank flows into the exchange flow channel means in the first and /or second port means in at least some of the inlet and/or outlet ducts positioned lower than the core, and into the reactor core, whereby a natural convective flow of coolant results.
An objective of the present invention is to allow ,: . . .
;.: . . : . : . .
. ~ ,,.. , . . . : .. .;.... . .
:r; ~, . ,..... ' ,' :. ' ' ' . '' ':

` :, ' ' ~ , , ' 132691~
the operation of a nuclear reactor having a primary circuit, a shutdown circuit, and, especially, a manner of coupling the two circuits such that:
A. The performance of the primary circuit under normal operating conditions is not significantly `~ compromised by the presence of the (passively coupled) shutdown circuit.
B. The shutdown circuit performc as indicated above, namely, it adequately cools the reactor core on a lo passive basis in the face of a variety of accident scenarios following self-initiation of the passive cooling function, such initiation being also on a passive basis.
C. The passive shutdown system, in both its initiation and sustained operation, is free of encumbrances 15 which could conceivably invalidate the assumption of -~ passive behaviour.
D. The primary circuit operates in the prescribed manner but with a bare minimum of active devices (e.g., sensors, pumps, actuators, valves, hydrodynamic shims, 20 feed-back circuits, central processing units, operator ; monitoring and intervention) being required to maintain operation at the prescribed performance level. This requirement is in the interests of: (a) minimizing the complexity and costs associated with the engineering, i 25 detailed design, construction, operation and maintenance of the plant; (b) maximizing reliability, safety, and licensability of the nuclear plant: and (c) freeing the basic passive safety design from additional complexities that would restrict the range of application of a reactor.
' 30 (Some applications, which may stand to gain the most from passive cooling, may also impose overriding requirements excluding a more complex form of passive safety.) E. The fundamental principles are universally adaptable to a sufficient degree that the basic reactor 35 design, including the approach to incorporating passively initiated passive cooling, can be applied to advantage in a range of circumstances, including the following, taken individually or in combination: (a) a range of power " .
i - 6 -~: ,' -.3 ,.

.'.` ' ' , ; ' ` , : : ' . ' ,: '' ~ '' , 1~26~6 output ratings; (b) a range of power reactor purposes, including those for space and process heating and those producing shaft power or electricity; (c) pool-type research reactors where the thermal-hydraulic configuration of the invention permits: (i) enhanced cooling, and thus higher sustained levels of power and, hence, neutron flux, for a given fixed reactivity margin, and (ii) effective localization of activation products within the primary transport circuit; (d) limited head space above the reactor ; 10 as in submarine applications of reactors; (e) varying orientation of reactor plant (which translates as variability of the gravitational g-vector), as in the case of a submarine-borne reactor during either normal operation ` at sea, or when capsized; (f) varying magnitude of the acceleration due to gravity during reactor operation, as for a submarine-borne submarine riding surface waves (bearing in mind that the acceleration due to gravity is a key parameter in all passive cooling designs); (g) a requirement for assured reactivity insertion accompanying initiation of shutdown cooling through high density neutron absorber dissolved in the pool water; and (h) a low margin `~ of auxiliary power available to operate the primary circulation pump.
In a reactor cooling system of the kind referred to in the present invention, the primary heat transport circuit which is otherwise of a more-or-less standard closed circuit design (i.e., with the coolant being circulated through the reactor loop core and the heat exchanging components by means of a pump) is fitted with branching devices called hydrodynamic ports in such a way ; that passive cooling is continuously available by process-inherent means to assure that the reactor is always adequately cooled, even following a range of events which would ordinarily lead to accident scenarios. It is a feature of the hydrodynamic ports and fundamental in the manner of their deployment that they permit reactors embracing a broad range of types, configurations and purposes to be accommodated in respect of the stated .` .~
i~
.. : : : .
", . . ., ~, . .
.''".
: . , .. ~: - :~ ,. ;

~;:

objectives of the patent, without significantly restricting or complicating plant design or operation.

Brief De~cription of the Drawings In drawings which illustrate embodiments of the invention, : Figure 1 is a sectional elevation of one embodiment of the invention;
Figure 2a is an axial view of a hydrodynamic port;
,;. Figure 2b is a cross-section along the line AA of Figure 2a;
Figure 3a is an axial view of a hydrodynamic port having bypass ducts;
` Figure 3b is a sectional view along the line AA of :, Figure 3a;
Figure 3c is a sectional view along the line BB of , FIgure 3a;
Figure 4a is an axial view of a hydrodynamic port with an anti-convective shroud fitted thereon;
Figure 4b is a sectional view along the line CC of Figure 4a;
Figure 5 is a sectional elevation of a second embodiment of the invention; and Figure 6 is a sectional elevation of a third embodiment of the invention.
.~ 25 The numerals of the drawings identify various elements and components pertinent to the invention according to the following list:
1 - reactor core assembly 2 - inlet plenum 3 - outlet plenum 4 - inlet duct . 5 - outlet duct 6 - diffuser . 7 - inlet hydrodynamic port 8 - outlet hydrodynamic port 9 - thermal insulating layers 3 lo - outlet conduit $,~

1326~1 ~
11 - inlet conduit : 12 - pressurizer . 13 - liquid-vapour interface :. 14 - reserve coolant tank 15 - reserve tank access hatch 16 - radial gamma shield 17 - axial gamma shield 18 - fuel elements ` 19 - radial reflector J 10 20 - circumferential exchange flow slot : 21 - flow area defining duct : 22 - flow area defining duct (complement) 23 - flange 24 - slot-defining plate 25 - spacer . 26 - bypass passage 27 - bypass tube 28 - bypass slot 29 - ringed flange 30 - anti-convective shroud . 31 - stratification grid vanes 32 - isothermal plane 33 - qua`si-stagnant zone 34 - circulated coolant 35 - reserve coolant 36 - inlet manifold 37 - outlet manifold - 38 - "borated" water shielding tank 39 - rod drive mechanism housings 40 - reactor core access tubes 41 - baffle and shield 42 - regulating rod guide tube 43 - shutoff rod guide tube 44 - regulating rod ;~. 35 45 - shutoff rod ; 46 - reactor vessel ,~;
,1 ,~ .
,...
_ g _ ., ~' r.. , ~, .

~r~ : . :: -~ . . . : .

132~916 Detailed Description of the Preferred Embo~iments In the reactor plant shown in Figure 1, the primary heat transport circuit comprises the equipment components normally found in a typical reactor cooling system of the 5 closed circuit variety. Included are the reactor core 1 and the inlet and outlet plena, 2 and 3, respectively, all located inside the reactor vessel 46. Also included are the inlet and outlet ducts, 4 and 5, respectively, leading by means of inlet and outlet conduits 11 and 10, 10 respectively, through penetrations in the walls of the reserve coolant tank 14 to connect with external components, principally the main circulating pump and s combinations of heat exchanging components, such as ~ preheaters, evaporators, steam generators, or simple heat 15 exchangers. The part of the primary circuit that embraces these latter components could be located within the reserve coolant tank along with the reactor vessel and connecting ducts and conduits, if required by the safety and space restrictions of a particular design. However, the form, 20 specification, and location of the pump and heat exchanging components, as a matter of particular note, are not , critical to the functioning of the subject invention, and ' have not been included in the drawings.
over the range of anticipated operating and 25 shutdown conditions of the reactor of Figure 1, the coolant ~, circulating within the primary circuit will be at temperatures considerably higher (typically by 80 - 250 C) than the reserve coolant maintained in the reserve coolant tank 14. Thermally insulating layers 9 of a design 30 suitable for such applications are attached to most of the primary circuit boundaries lying within the reserve coolant tank. Such insulation enhances the energy delivery efficiency of the reactor and facilitates the maintaining of the reserve coolant at a standby temperature 35 sufficiently low to support safety objectives. The role of ~3 the insulation in enhancing the actual operation of the ! passive cooling circuit is not necessarily of importance, however, in meeting most specific reactor safety ';~ '.' ~ - 10 -!j ,j :

'." " " ' '' ', ; . ' ' ~. '; ' '` " ' ` ' ' '"' , , . ' ','.' , , : ' ' ' ', ;: ' ~ ~. ' .
~'." ~ ' ` '' ' : . . ,. ' ' ",~ . ; .,. . ' . i . ' ' "'' i :' ' requirements.
The reserve coolant is maintained at suitable standby temperatures with the aid of reserve tank coolers (not shown in the drawings) which transport, to the external environment, residual heat transferred from the primary circuit to the reserve tank during normal operations, and also the decay heat transferred to the reserve tank by the passive cooling circuit following emergency shutdown. The reserve tank coolers may operate on either active or passive principles but, for consistency with the passive safety principles of the present invention, the capacity of the passive cooling component of the reserve tank coolers should be adequate for safe decay heat removal over an indefinitely long period following reactor shutdown. The specific design of the reserve tank coolers may be carried out by persons skilled in the art of heat transfer as required to suit the particular reactor and the circumstances of its application.
The elements of the primary heat transport circuit not ordinarily found in a reactor cooling system, but which are a key feature of the present invention, are the inlet and outlet hydrodynamic ports, 7 and 8, located in the circuit just upstream from the inlet duct 4 and downstream from the outlet duct 5, respectively. The hydrodynamic ports, exemplified by the basic version shown in Figure 2, exhibit specific hydrodynamic propsrties when correctly incorporated in the design of a reactor primary heat transport circuit. The most important of these properties are the ports' capabilities of: (i) supporting to a high degree the continuity of primary circuit flow between the ~ducts 21 and 22 in normal reactor operating conditions, :!(ii) offering large resistance against a tendency toward both combining and dividing branah flows through the exchange flow slots 20, relative to port axial flows, when ~!35 the system temperatures, primary flow or reactor orientation deviate inadvertently from normal, and (iii) offering relatively small resistance to large branch flows of either the combining or dividing kind in severely ~;', ', ' ' '`.' ~ '. ' . ' 132~9~6 abnormal or accident conditions, such as when the main circulating pump has stopped.
- Under these latter conditions, core cooling relies solely on naturally convective coolant circulation, established and maintained by inherent processes. The operative convective circuit in such conditions is between the core 1 and the reserve coolant tank 14 and is referred to herein as the passive cooling circuit. A principal flow pattern for natural convection is depicted by the broken arrows in Figure 1. The solid arrows show the flow pattern in the primary cooling circuit under normal operating conditions with the circulating pump on.
The above-mentioned properties of the hydrodynamic ports are manifest in the first embodiment of the present invention as it applies to the reactor plant of Figure 1.
j A detailed description of the operation of the cooling - system of this plant follows.
As implied above, under normal operating conditions a steady flow is maintained through the primary circuit by the continuous operation of the circulating pump. The pumping head equals the algebraic sum of the pressure changes across individual circuit components, including (i) -9 the heat exchanger components located externally to the , reserve cooling tank as referred to above, and (ii) the components forming that part of the circuit delineated by the solid arrows and shown residing inside the reserve coolant tank 14, namely (in flow sequence), the inlet conduit 11, the inlet hydrodynamic port 7, the diffuser 6, the inlet leg 4, the inlet plenum 2, the reactor core 1, ;~ 30 the outlet plenum 3, the outlet duct 5, the outlet hydrodynamic port 8, and the outlet conduit 10.
~ For purposes of providing the constant availability'1l of passive cooling by process-inherent means during normal `, rector operation, the hydrodynamic ports, 7 and 8, are ~ 35 designed (as clarified in Figure 2) to present no physical ; barrier against the flow of coolant between the primary cooling circuit and the reserve coolant tank 14. However, to avoid the undesirable transport of either thermal energy .~3!
:, ,........ . .

~, .. . , . ,: . .
.j, . . , , , , -,,., , ~ .

1326~
or radioisotopes from the primary circuit into the reserve tank during normal operation, it is important to reduce to acceptable levels the tendency for such exchange flow, and also the magnitude of such flow if it should occur. (In some reactor variations, in which safety shutdown is induced by the onset of passive cooling, the limiting of exchange flow in normal operations is essential also to avoid the premature transport of dissolved neutron absorbing nuclides from the reserve tank into the primary circuit and the resulting reactor shutdown.) The tendency for such exchange is reduced to acceptable proportions, in the first instance, by the incorporation of the port concept in the design of the primary circuit in accordance with a specific design requirement applying only to that part lying within the reserve coolant tank 14 of Figure 1. The requirement is stated as follows: At normal operating conditions, the net pressure from all effects accumulative around the passive cooling circuit delineated by the broken arrows in Figure 1, and tending to support net flow around this circuit, must be substantially zero.
Basic parameters of the primary circuit that must be taken into account in designing for zero net accumulative pressure are the primary circuit mass flow rate, the coolant temperature profile in the primary circuit, the bulk temperature of the coolant of the reserve coolant tank, and the resistive pressure drop due to the primary circuit mass flow through the core and other 1 components in the flow path from port 7 to port 8. All `~ 30 such parameters will have been determined as the normal ~ operating design values selected to yield optimal plant ;, energy production, efficiency and safety, as appropriate to ~! the intended application.
More specifically, the criterion for meeting the ~ 35 requirement of zero net accumulative pressure may be ;~ expressed in terms of the values of Al and A2 which are ~l shown in Figure 1 as the axial flow areas associated with hydrodynamic ports 7 and 8, respectively. These flow areas ,................................................. .

,~, - : , . .- :. -:

.: . ~ . .: ~ ' , . . . .. . ~ . ,: ., . . . ~ :,. . ..

~ 132691~
can be chosen by those skilled in the science of hydrodynamic design so that the change in dynamic pressure (velocity head change) associated with a constant mass flow passing from the vicinity of the inlet port 7, of flow area Al, to the vicinity of the outlet port 8, of flow area A2, is equal to the deficit by which the net elevation pressure, assessed accumulatively around the passive cooling circuit, fails to match the resistive pressure losses associated with the same primary coolant mass flow through the core. The correct assessment of the net elevation pressure is based on the vertical port separation L as indicated in Figure 1, and the between-port coolant temperature profiles (and hence the density profiles) along paths through both the reactor vessel and the reserve tank.
The resistive losses identified above are due to the hydrodynamic resistance offered by the core itself, and other components between the inlet port 7 and the outlet port 8.
Within the requirement of zero net accumulative pressure, as spelled out in the stated criterion for meeting the requirement, it is clear that a great deal of - flexibility is possible in the thermal-hydraulic design of the reactor depicted in Figure 1. For example, if the temperatures of the profile extending through the core and between the two ports are considerably higher than those ~il that are characteristic of the bulk of the reserve coolant, as will certainly be the case when a relatively high ~ thermodynamic efficiency is needed in the conversion of :~ reactor thermal power output to mechanical or electric forms, and when the temperature of the reserve coolant is ; kept relatively low in the interests of safety, it is possible to meet the requirement of zero net accumulative pressure requirement without prescribing any change in flow area in going from the vicinity of one port to that of the other, i.e., with Al equal to A2. As the net change in dynamic pressure from the vicinity of port 7 to port 8 is ~i zero in this case, the design criterion is simply that the net accumulative elevation pressure be equal to the , :. : :
: :

;, ~ . , - . , :
.

~ 1 326916 resistive pressure losses along the flow path through the core and between the ports. This criterion is met by the adequate protraction of the vertical separation L between the ports, and seeing that the segment of the primary circuit including the core and lying between the ports, itself now considerably lengthened as a consequence of the ~-increased port separation, is of a design consistent with good hydrodynamic practice.
In a different example, space limitations are assumed to impose a restriction on the reactor plant and ~-hence on the vertical separation L between the ports in the reactor of Figure 1. For a prescribed disposition of normal operating temperatures throughout the system, this reduction can be permitted, while preserving the zero net accumulative pressure requirement, by making A2sufficiently larger than A1 to create a net dynamic pressure decrement in the coolant passing from port 7 to port ~. This decrement constitutes additional driving pressure to compensate for the reduction in the net accumulative elevation pressure associated with the shortening of both the hot and the cold hydrostatic columns operative around the passive cooling circuit. As in the previous case good thermal-hydraulic design applied to that part of the primary cooling circuit lying between the ports and including the core is essential i25 and may be carried out by those skilled in the science of hydrodynamic design. The specification of such components as the diffuser 6 and its location within the circuit must ,jibe in accordance with established design practice for such components.
In a third example relating to the reactor depicted ,.7 in Figure 1, sufficient vertical space is postulated so that, for the prescribed disposition of normal operating temperatures throughout the system, the distance L between the two ports can be made such that the net static head becomes more than sufficient to compensate the resistive pressure losses associated with the passage of the primary coolant at design mass flow rate through the core. In this case, the zero net accumulative pressure requirement may be :~
~ ~ - 15 -,~
, \

~ ~ ;'', : .: ~ , .... .
!, ` . ` .
'; ' . , ' ''' ;' , ' ' 132691~

met by making A2 smaller than Al so that the dynamic pressure increment cancels the exces~ net elevation pressure over the resistive pressure loss across the core.
Also, the excess net elevation pressure contributes, along with the circulating pump if one is present, to overcoming the resistive losses in those primary circuit components ~ implied as lying outside the reserve coolant tank. Again, - the careful application of hydraulic design practice is required.
10A principal objective of the invention is to -~provide in a reactor cooling system, of otherwise ordinary but efficient design, a passive cooling system capability which is constantly available for deployment by process-inherent means during normal reactor operation, and which 15 is capable of mitigating satisfactorily a set of reactor accident initiating events. In the cooling system of the reactor depicted in Figure 1, as an example, the hydrodynamic ports, 7 and 8, are designed to present no physical barrier against the flow of coolant between the 20 primary cooling circuit and the reserve coolant tank 14.
-~ The principal means of avoiding the potentially deleterious effects of such an arrangement on the normal reactor operation is an important element of this invention. The operation of the cooling system in mitigating accidents 25 effects is now described.
In the event that pumped flow through the primary $J circuit of the reactor cooling system shown in Figure 1 suddenly ceases during the course of normal reactor ~^ operation (due to a loss of pumping power, for example), 30 the resulting drop in axial coolant flow through the hydrodynamic ports leads to a loss of the dynamic pressure change that is maintained during normal operation. The net accumulative pressure around the passive cooling circuit becomes non-zero and dominated by elevation pressures 35 differences which are enhanced momentarily by the rising temperature of coolant within the core. In these circumstances, the reactor core becomes a component of the now operative passive cooling circuit and, following some ,., ;~ - 16 -~ 1, :, :, . - . . . :

.,. ., . - .

~J . ` ' "
1' ~' ' ~ ' ' ,, ,:' , ~ ' ": ' 1326~16 reactivity-related transient oscillations, caused in part by the sudden engagement through the lower port of the relatively cool reserve coolant with the passive circuit, a relatively steady flow supported entirely by natural 5 convection becomes established. The coolant column of approximate height ~ and temperature equal to the reactor outlet temperature serves as the hot leg, and the column of height L partially cancelled by the column of height L1, both columns being at the reserve coolant temperature, 10 serves as the cold leg of the operative thermosyphon.
In the event of loss of capacity of the heat sink of the primary circuit, the resultant return of the circulating primary coolant at significantly elevated temperatures raises the temperature of the thermosyphon hot 15 leg (now of length L) to above-normal values, forcing the passive circuit to operate in parallel with the pumped circuit, thereby ameliorating the core coolant condition due to increased core flow and, more importantly, the combining of cool reserve coolant with primary circuit flow 20 at the inlet port 7. A similar chain of events results, in the near term, following a loss of regulation of reactor power event in which for a time the reactor undergoes a ~j significant inadvertent increase in reactor power.
~g Before describing the behaviour of the reactor ~ 25 cooling system in the event of the breakage of a component,$. which has a role in defining the pressure boundary of the primary heat transport system, it is important to describe ~i further the disposition of some of the system components, as the nature of the consequences depends greatly on the 30 location of the break within the system, and on the i specific arrangements whereby the system is maintained at ;3 atmospheric or elevated pressures during normal operation.
i The primary heat transport system pressure boundary includes the reserve coolant tank and those components of 35 the primary circuit indicated with respect to Figure 1 to 3 lie outside the boundary of the reserve coolant tank, unless of course they too are located inside the reserve tank in the alternative arrangement mentioned earlier. The , ~j - 17 -.
... . . .

~; .

1326~1 ~
reactor of Figure 1 is depicted as being maintained at an ; elevated operating pressure by means of the pressurizer 12 connected to the reserve coolant tank. A liquid-vapour interface 13 is shown to be maintained in the pressurizer.
If the reactor is to be operated nominally at atmospheric pressure, either the pressurizer, or the reserve tank itself, is left open to the atmosphere in the design, or is in some manner vented to atmosphere. The so-called (open) pool-type reactors fall into this latter category.
Because substantially zero pressure differences must exist across the barrier-free interfaces that occur - between the reserve tank and the primary circuit under normal operating conditions, and the operating pressures of these two parts of the system are therefore intimately related, the degree of pressurization of the primary system ; is largely unimportant in a description of the basic normal functioning of cooling system according to the invention.
By similar reasoning, the actual point at which the pressurizer is connected to the system is relatively unimportant to the objectives of the present invention.
Although the pressurizer of the reactor cooling system of Figure 1 is shown connected to the reserve coolant tank, it may be preferable in certain circumstances to connect the pressurizer to the primary circuit instead. In some cases, it may be advantageous to connect a single pressurizer to the reserve tank and the primary circuit in parallel.
These observations apply so long as the degree of pressurization and other key operating parameters do not stray too far from design values, in either the normal operating or the passive cooling modes, as described previously.
The principal objective of pressurizing the reactor coolant in any reactor plant is to permit the transport of heat at elevated temperatures, in the interest of enhancing the thermodynamic efficiency of the plant, while retaining the advantages of the coolant in its liquid state. In the reactor plant of Figure 1, only the coolant circulating in the primary circuit has a requirement for elevated :, ~ 18 -:., . ~; . - ~ , , .

13~6~

temperatures in consideration of plant thermodynamics.
Therefore, the reserve coolant may be maintained at L, relatively low temperatures; safety considerations require that it must be maintained by means of the tank coolers (not shown) at temperatures much lower than the saturation temperature corresponding to atmospheric pressure. This requirement is entirely consistent with the separate safety requirement that the reserve coolant temperature remain low in consideration of enhancing the passive cooling system's potential in mitigating accidents conditions in the manner already described.
Because of the way in which the pressures of the primary circuit and the reserve coolant tank are linked in the maintenance of the zero net accumulative pressure requirement, a requirement for pressurization of the primary circuit results also in the pressurization of the reserve coolant tank. Thus, in a pressurized system, the nominal operating pressure of the entire heat transport system is required to be conservatively in excess of that required to prevent boiling at the point of the hottest ,Jcoolant within the primary circuit. Conversely, in the special case of the reactor not being pressurized to higher than the ambient pressure, the reactor is constrained to operate with core temperatures not exceeding the saturation temperature at atmospheric pressure. The reactor shown in Figure 1 is representative of either the pressuxized or non-pressurized case.
The arrangement exemplified by the reactor in Figure 1, in which the primary circuit and the reserve ~!30 coolant tank are required to operate at identically regulated pressures, enhances the safety performance of the reactor in the event of a primary circuit pipe break. This follows since only the relatively small fraction of the .total primary inventory which lies within the primary circuit itself is at a temperature exceeding the saturation temperature at the ambient pressure (normally one atmosphere) and is therefore subject to rapid vaporization and accompanying energy release at the instant of the ,~, - 19 -., ,. . :
-, ;~ . -.: .. ~ , :., -,. ; ;,. .: :~
., , :, . .
:. .
.

132691~

break. ~he bulk of the reserve coolant inventory remains as cool liquid immediately following the break and serves to continue the core cooling function after entering the core region through the ports and eventually establishing a steady passive cooling flow.
- By suitable design, in which, for example, the inlet and outlet conduits penetrate the wall of the reserve coolant tank at levels higher than the locations of the upper hydrodynamic port (as shown in Figure 1), the passive - 10 cooling system will continue to serve the reactor core cooling function for a considerable time following the pipe break. The design may also call for a syphon break arrangement (not shown) between the two conduits in order to limit the extent of removal of reserve coolant by syphoning action. The potential for loss of coolant : through rupture of the reserve coolant tank itself is minimized by care and redundancy in design.
~- In the special case of the reactor which is non '~ pressurized relative to atmospheric pressure, and the maximum coolant temperature does not exceed the saturation i temperature at this pressure, a pipe break is followed by a very orderly progression of events (due to the absence of vaporization of liquid) culminating in the steady operation of the passive cooling circuit in the dissipation of residual reactor power production.
It was previously emphasized that current practice in the art of thermal-hydraulic design would be required in designing the conventional aspects of the primary heat transport circuit which support normal operation. Similar ; 30 practice is to be applied in achieving a satisfactory passive cooling design in which the core resistance for a ~ broad range of thermal-hydraulic conditions, the lengths ;J and cross sections of the inlet and outlet legs, the standby temperature of the reserve coolant, and the reserve tank cooler capacity must be specified. It is also a requirement of the design that the resistances of the hydrodynamic ports to large scale branch flows do not adversely impede the natural convection of the passive .

, ... . .
:......... . ;

..... . . .
,;,~ - . : .

132691~
cooling circuit in any depicted accident scenario, including ones in which reactor shutdown is not effected immediately. Further consideration of this aspect of the detailed design of the hydrodynamic ports is given in the paragraphs that follow.
In the selected illustrations of the response to accident conditions of a reactor cooling system made in accordance with the invention, no particular reference was made to either reactivity reductions or reactor shutdown action instigated early in the events. It is clear in all examples given, however, that the provision for early reactivity reduction will a~sist in arriving at a detailed thermal-hydraulic design which will maintain the safety of the reactor following all credible initiating events.
Furthermore, the provision of an early shutdown will clearly extend the period of time that may be allowed to pass before human intervention is required to assure long term safety, following an accident. Such reactivity-~31 limiting mechanisms are not prescribed as a part of the present invention, but it is clear that if the reactivity reduction or shutdown action is inherent in the processes ,j~ which are intimately related to the accident condition, then reactivity-limiting mechanisms can be seen to be consistent with inherent safety aspects of the cooling system.
Two reactivity-limiting mechanisms, both of which are generally understood by persons skilled in the art of reactor design, are mentioned here briefly, however. One is based on negative reactivity power coefficients associated with either the expansion of the coolant or the effect of fuel heating on either resonance absorption or neutron thermalization. The other is less direct and is ~' initiated by the onset of passive cooling which, in turn, is triggered in response to the initiation and development of the accident scenario. The first of these mechanisms may be adopted in the detailed design of a reactor such as the one in Figure 1, in the selection of the fuel type and associated core design.

~ ?~ -- 21 . `.................................. . .

.~" , ' , , : " :
':''' - ': : - ~ -, ,,: ~: -.

~ 1326916 The second mechanism may be implemented in the detailed design of a reactor such as that of Figure 1 by arranging for the permanent addition of heavily concentrated neutron absorber in solution with the reserve coolant, and placing neutron absorber only in mild solution with the primary coolant. The reactor, in this case, is regulated by actively varying the concentration of absorber - in the primary circuit. Under normal operating conditions the two levels of solution are kept from mixing with one another as a byproduct of the arrangement described earlier as keeping the same two categories of coolant from mixing ~- for reasons of both thermal efficiency and prudent radiological management. The additional requirement of maintaining the two levels of solution in fairly strict isolation from each other may in turn require an elaboration of operating protocol or, preferably, the - adoption of a somewhat more sophisticated arrangement of the hydrodynamic ports such as the ones discussed below in connection with the Figures 2, 3, and 4, and with the reactors depicted in Figures 5 and 6.
In terms of strictly thermal-hydraulic considerations, the reactor cooling system of Figure 1 does not place any formidable restrictions on the operating protocol of the reactor plant. In a typical start up scenario, the primary circuit pump may first be turned on, in which case there will be a significant bypass exchange flow in which some circulating primary coolant enters the reserve tank through the inlet port 7, while an equivalent flow of reserve coolant passes into the primary circuit ~, 30 through the outlet port 8. As the reactor power is increased after being brought to criticality, the primary circuit warms more quickly than does the reserve coolant and, by the time the primary circuit reaches the normal operating temperature, it will have ceased altogether to exchange with the reserve coolant. A less likely alternative start up procedure is to first bring the ` reactor to an operating power level while relying on the ! convective flow of the passive system for core cooling. As .
s - 22 -. .
,j, .

' 1~2691~

the pumped cooling circuit is brought on line, the initial ingress flow of the passive cooling is gradually reduced to zero and the normal operating conditions are reached.
During the course of extended normal operations of the reactor depicted in Figure 1, a finite amount of exchange flow may occur inadvertently due to fluctuations in the operating paxameters which lead to momentary departures from the zero net accumulative pressure requirement. It is an advantage of the cooling system - 10 constructed according to the invention that there is a continuity in the behaviour of the exchange flow deviation from the null point as the values of the operating parameters vary about the precise set of values that provide null exchange. Such advantage is manifest in the amenability of such a functionally "well-behaved" system to various remedial measures of confining exchange flows to arbitrarily small values by either active control of the operating parameters or by arrangements for passive compensation.
Exchange flows, including those which occur during either the start up scenarios described or inadvertent lapses in meeting the zero net accumulative pressure requirement, contribute accumulatively to the loss of some -~ useful energy, to the loading of the reserve tank coolers, and to the build up of what may be radiologically significant quantities of radioisotopes in the reserve coolant. In many applications, significant levels of these effects may be acceptable, depending on the circumstances of the particular application. If, however, such effects must be strictly controlled, even to the point of being eliminated entirely over a fairly broad range of operating parameters, remedial measures in the form of either the precise control of the operating parameters during operation, or as preferred, through special designs of the ports and the cooling system as a whole, may be implemented . ., - in such a way that the system is self-compensating by inherent processes over a meaningful range of parameter variations. Such special designs are embodied in the '.
s ;~ - 23 -", ~ .......
,; "~

:", -. ~ , . ..

. , .
, , . , .. .
., 1 3 2 6 9 ~ ~

different versions of the hydrodynamic ports discussed below with reference to Figures 2, 3, and 4, and with the reactors depicted in Figures 5 and 6. Such special designs would be necessary, for example, to maintain an adequate operational separation between the reserve and the primary coolants if reactor shutdown relies on two distinct levels - of dissolved neutron absorber. Also, a somewhat more elaborate start-up protocol than those just described would be necessary in reactors involving dissolved absorber in this way. Such protocols may be readily devised by those skilled in the art of reactor dynamics and control.
Beyond the imposition of the zero net accumulative pressure criterion on the primary circuit design through the proper selection of the flow areas A1 and A2 as described previously, additional factors which limit the magnitude of exchange flow between the primary circuit and the reserve tank during normal operations are vested in the design of the hydrodynamic ports themselves. A detailed representation of a basic version of the hydrodynamic port, of a kind suitable for implementation in the cooling system of the reactor of Figure 1, is shown in Figure 2. This port constitutes an approximation of a continuous pipe made up of two flow area defining ducts 21 and 22 with attached flanges 23, separated axially by the branch opening composed of a series of several slot-defining plates 24 , alternating with spacers 25 which give rise to the circumferential exchange flow slots 20.
During normal operation, when the requirement of zero net accumulative pressure is nominally satisfied, essentially all flow through the port is axial and the flow magnitude is substantially due to primary pump operation.
Under these conditions, branch flow tends to be discouraged , on the basis that the particles of coolant involved in any such branch flow would experience high momentum changes, whether combining with or dividing from the main axial flow, by passage through the slots. The purpose of the slot and plate structure of the branch openings is to require the largest possible momentum change for the ., ~ - 24 -"

"`: ' ' ,. ~ ' -~x, ~
.

132691~
"typical" particle of coolant involved in branch flow.
Thus, the port structure shown in Figure 2 is seen to be supportive of the first two of the previously mentioned important properties of the hydrodynamic ports, namely, (i) ; 5 the support of the continuity of primary circuit flow between the flow area defining ducts 21 and 22 in normal reactor operating conditions, and (ii~ the offering of large resistances to both combining and dividing branch flows, when the tendency for such is created due to the `` 10 inadvertent deviation of principal operating parameters, such as coolant temperatures and primary flow rates, from nominal values.
The third of the previously mentioned important properties of the hydrodynamic ports, namely, the offering of relatively small resistances to large branch flows, of .either kind, in severely abnormal or accident conditions, is characteristic of the hydrodynamic port shown in Figure 2 when properly incorporated in the reactor cooling system, ;for the following reason. In abnormal circumstances, such `i20 as during an impairment of the primary pump, the required ,~mass flow through the core for adequate cooling is considerably reduced due to (i) the inlet temperature being now determined by the reserve coolant temperature which is maintained at values greatly depressed relative to operating core inlet temperature, and (ii) the expected reduction in reactor power due to a suitable provision for reactor shutdown in accident conditions. The relatively small mass flow requirement, and the resulting acceptability of relatively low passive cooling flows under shutdown conditions, makes tolerable the residual resistance which branch flows experience in negotiating the slot-defining plates of the ports.
The detailed design of the hydrodynamic ports based on the principles of operation just outlined may be carried out by persons skilled in the art of hydrodynamic design.
Thus, the variables such as the axial flow area, the number and thickness of the slots, and the sizes of the plates and spacers may be selected to suit the capacity and the -`~: d~

, . , , - .

, ~
: ~s, ~3~91~
performance specifications for a reactor designed for a particular application. It is an advantage of the port concept and the manner of its implementation according to the invention that the ports are simple of structure and, therefore, afford great flexibility in adaptation to a variety of reactor configurations. Nevertheless, the ~, important properties of the ports discussed above can be enhanced through any number of refinements which may be ~ implemented by those skilled in the art of hydraulic -~. 10 design. Such refinements may include adaptations in the : form of the shaping of the edges of the slot-defining plates, the minor progressing of their inside diameters, '".! the introduction of an axial rod of cylindrical, conical or other shape, or the "dishing" of the plates themselves.
Two specific enhancements of the important properties of the hydrodynamic ports as refining elements of the invention will be described in some detail. The first enhancement is represented in Figure 3 as a modification of the basic hydrodynamic port depicted in Figure 2. The objective of this modification is to enhance, relative to the port depicted in Figure 2, the previously-mentioned second important property, namely, the offering of large resistances to both combining and dividing branch flows when the normal operating parameters inadvertently deviate somewhat from nominal values, while leaving essentially unaltered the branch flow area available to the shutdown passive cooling mode. The second enhancement is represented in Figure 4 as an anti-convective shroud to be attached to each hydrodynamic port to not only contain local exchange flow in the port, but also to provide for the self-acting hydrostatic compensation of pressure imbalances tending to produce exchange flow, as discussed below.
The hydrodynamic port depicted in Figure 3 may be described as the basic structure of the hydrodynamic port depicted in Figure 2 modified in the following ways.
First, the port of Figure 3 has two extra slot-defining plates 24 in the series which, with the flanges 29 which . , ... i.

~,.' ' ' ' ,, ~ .
:: .. .. .
.: ,.,: : . ... ...
.... : - :-::

.,. . ;.. ~ .

are now ringed, form blind bypass slots 28, one at each end of the series of regular slots 20. Second, the bypass slots so-formed at either end of the series of regular slots are connected by means of several longitudinal bypass tubes 27 which are located symmetrically about the `principal axis, providing a plurality of bypass passages 26 between the bypass slots and, thereby, providing passages between the two flow area defining ducts 21 and 22.
The role of the bypass passages so defined may be ;10 stated as follows: Under ideal normal reactor operation conditions, the ports carry only axial flow. Under such conditions the flow experiences very little resistance in passing through, and between, the regions of the area defining ducts 21 and 22. In the absence of significant levels of such resistance, a negligible pressure difference exists between the entrances to the two bypass slots, and negligible bypass flow is present. Therefore, under the ideal operating conditions characterized by axial flow only, the modified port of Figure 3 behaves within the circuit essentially as does the basic port of Figure 2.
In the event of the reactor operation departing from normal, the zero net accumulative pressure requirement ceases to be satisfied and there arises in either the basic or the modified port the tendency for branch flow, i.e., combining branch flow in one port and the corresponding dividing branch flow in the other. In the case of the modified port, however, even small branch flows of either kind create sufficient longitudinal pressure differences between the entrances to the bypass slots ~due to the momentum changes inflicted on the axial flows by the branch flows) to cause significant bypass flows through the bypass passages.
In the case of combining branch flow, the pressure drop and the induced bypass flow are in the general direction of the main flow. Therefore, the bypass tubes may be seen as assuming a share of the port's normal axial mass flow, thereby reducing the axial flow velocity within the port. The corresponding increase in the internal .
` A 27 ~, ~
.~ , ~: .. . . . .

. : , ~ .... .
, : .. ~ .... . .
. : . :: . .

.~ `
static pressure in the vicinity of the port opposes the incoming branch flow and therefore tends to correct the initial cause, namely, the departure from zero net accumulative pressure in the passive cooling circuit.
In the case of the dividing branch flow simultaneously in the other port for the same causative departure from zero net accumulative pressure, the pressure drop and the induced bypass flow are in the general reverse -~ direction to the main flow. Therefore, the bypass tubes f lo may be seen as recirculating a portion of the port~s normal axial mass flow, thereby increasing the axial flow velocity within this port. The corresponding decrease in the ~ internal static pressure in the vicinity of the port - restrains the outgoing branch flow and therefore tends, as s 15 for the port experiencing combining branch, to correct the initiating departure of the system from the required zero net accumulative pressure in the passive cooling circuit.
Since the basic port of Figure 2 does not have the benefit of the bypass channels and, hence, the associated 20 counteracting actions described above, a greater rate of exchange flow may be expected in a system incorporating such ports, in comparison with a system incorporating the ports of Figure 3 which are the equivalent except for the ; bypass channels. In fact, it has been demonstrated 25 experimentally for branch flows in the region of greatest s interest for present purposes (i.e., for branch flows of less than 5 percent of the axial flow), that the branching loss coefficients in general increase linearly with the magnitude of the branch flow for both the ba~ic and the 30 modified ports, and that the addition of the bypass channel can increase the linearity coefficient from zero to about 2, in the case of dividing flow, and from 4 to about 8, in j the case of combining flow.
As discussed above with respect to the basic port, 35 the details of the structures for the modified hydrodynamic ports can be specified according to the invention by persons skilled in the art of hydraulic design. Included ~j in this work is the selection of the capacities and the ,, ~,.......... , , ~

:., ~ . . ~, ~

: ., . . -. ~. .~ . .

132~gl6 configurations of the bypass channels so that they provide significant enhancement of the second important property as discussed, while not impairing significantly the third important property, namely, the offering of relatively small resistances to large branch flows in severely abnormal or accident conditions.
It is an important feature of the modified ` hydrodynamic port depicted in Figure 3, that the effects at work in opposing the departures from the zero net accumulative pressure requirement (giving minimal exchange flow between the primary circuit and the reserve coolant tank) are present immediately, i.e., in the order of the through-port transit time. Therefore, the time required for the corrective processes within the hydrodynamic ports furnished with bypass tubes does not limit the speed of the response to the inadvertent variations of normal operating conditions. The immediacy of response is of particular importance in counteracting the influences of periodic inertial forces that are commonly imposed on mobile reactors, in which case the effective acceleration due to gravity, which determines the elevation pressures operating within the system, may be subject to severe time variation.
The discussion to this point has focused on two principal features of the invention which provide for the restraint of exchange flow between the primary circuit and the reserve tank in the absence of a physical barrier ! separating the two. These features depend on (i) thecorrect specification of the hydrodynamic ports within the system to which they are applied~ in order to fulfil the zero net accumulative pressure requirement in the virtual elimination of exchange flow driving pressures when system parameters are at their normal operating values, and (ii) the appropriate detailed design of the hydrodynamic ports which, in either the basic or modified version, tend to counteract exchange flow arising when the system parameters deviate from normal. According to the invention, these features may be embodied in a variety of reactors, of which one example has been described with reference to Figure 1.

132`69~
- For many applications of such reactors, these two features alone may be sufficient It is recognized, however, that without further measures to limit exchange flow, residual levels of such flow may persist at levels unacceptable in some applications.
The principal reasons for the persistence of exchange flows, in spite of the two features mentioned, are two-fold. The first is port related and results from the fact that a finite amount of local exchange flow will ; 10 always be present in ports defined in the manner of either Figure 2 or Figure 3. The local exchange flow is the result of small internal pressure differentials caused by both the local dynamic and elevation pressure effects.
Such differentials may be experienced in progressing longitudinally from one slot to the next, or in sampling pressures while progressing vertically across the port particularly when the axis is oriented more or less in line with the horizontal. ~hese effects give rise to local circulation of coolant into and out of the slots, even when ` 20 the net branch flow in the port may have been cancelled due to the above-described design measures. The effects are , aggravated by the inevitable hydrodynamic roughness near the slots and plates as viewed from the port interior, and the severe temperature differential in the vicinity of the ~, 25 inescapable thermal interface between the circulating coolant and the reserve coolant.
The second reason for persistent exchange flow is system related. Even when the earlier-described design measures succeed in either minimizing the pressure differentials of the system that drive exchange flows, in the first instance, or counteracting such pressure ;; differentials as they may arise, in the second, such measures cannot in principle eliminate the residual ;~ exchange flows absolutely. Therefore, in reactor applications which require that exchange flows be restricted to very small values, or indeed virtually eliminated (as required when shutdown absorber is present in the reserve coolant), a supplementary mechanism to ,, ... . : ~, ..

. . ;. :: : ~-~. . . ~ . -.

` 132691~

complete the suppression of exchange flows within the ` reactor system is necessary. An active mechanism, that regulates one or more of the main system parameters to - eliminate instrumentally detected or anticipated residual ; 5 exchange flow within a cooling system constructed in accordance with the invention as described to this point, may be devised by those skilled in the art of thermal-hydraulic control. For reasons of reliability, .! effectiveness, universal applicability, and harmony with the basic objectives of the invention, however, a passive mechanism is preferred to the active one.
According to the invention, the phenomena of local exchange flow associated with each hydrodynamic port of the system, and the exchange flows residual in the reactor cooling system as a whole, are addressed through the introduction of a single attachment applied to each hydrodynamic port of the system. The attachment, an example of which is shown in Figure 4, is referred to ; herein as the anti-convective shroud. The anti-convective shroud is a passive device with the capacity to suppress the local exchange flows in the port to which it is attached, and to accumulate incipient residual exchange .~ flow within the system flow in such a manner that the pressure differentials tending to drive the residual flows 25 are completely compensated.
Referring to Figure 4, the anti-convective shroud 30 is exemplified as an attachment to the hydrodynamic port made up of the components 21, 22, 23, 24, and 25. These components were previously identified in the description of 30 an identical example of a hydrodynamic port, namely, the basic version shown in Figure 2, but without an anti-convective shroud. Such an attachment may be fastened, with similar expected advantage, to any one of a variety of versions of hydrodynamic port, including, for example, the l 35 modified version incorporating bypass elements as shown in 3 Figure 3, also described above.
;~ The anti-convective shroud 30 is essentially a hood which opens downward and encloses the entire structure of .~
~i , .
: - 31 -` .
,,.;,', ` ' . - , ' , .:., , .. . . : .. .. . .. ....

132~

:`
the hydrodynamic port to which it is attached. Apart from the downward opening, and possible vent hole (not shown) through the uppermost point in the shroud surface to eliminate any gases collected, the enclosing surface of the shroud is complete and terminates in sealed joints at the flow defining ducts 21 and 22. The anti-convective shroud has sufficient internal clearance from the port to avoid ; directly interfering with the flow of coolant through the exchange flow slots 20.
10Under normal operating conditions, the shroud provides for a largely isothermal, quasi-stagnant zone of coolant generally enveloping the port. This zone of coolant continually adopts the approximate temperature of the coolant flowing axially within the port, by virtue of the stratification qrid made up of vanes 31 placed vertically in the downward opening and the thermally insulating material (not shown) that may be placed on the remaining surfaces of the shroud. Since the temperature of the circulating primary coolant, and hence the temperature of the quasi-stagnant zone, is considerably hotter than that of the reserve coolant, a large thermal gradient must exist in the vicinity of the downward opening. The purpose of the stratification grid is to inhibit the formation of local convective circuits in the vicinity of the opening, thereby facilitating the orderly formation and maintenance of thermally stratified layers of coolant ranging from the nearly primary circuit temperatures at the top of the grid, to typically reserve coolant temperatures at the bottom of the grid. Regardless of the orientation of the port axis i 30 in a particular reactor application, the specific configuration of the anti-convective shroud must be designed so that the alignment of the stratification grid is nominally horizontal, in order to best support the thermal stratification of which one isothermal plane is identified by the numeral 32 in Figure 4.
It was implied above that, in the absence of special measures such as the anti-convective shrouds, local exchange flow through the slots of a given hydrodynamic "~

.,~, : , ~ : .. ,;,. . .

, ~26~

port would exist as the result of small internal pressure differentials caused by both the dynamic and the elevation pressure effects. In the presence of the quasi-static zone of coolant around the port, however, significant elevation pressure differentials cannot exist, since the circulating primary coolant inside the port and the coolant of the quasi-static zone outside and along the port are now essentially of the same temperature. Moreover, while the dynamical pressure differences occurring within the port may result in some local exchange flow through the slots of the port, such exchange generally involves coolant of the single temperature and is contained within the shroud.
Only if the local exchange is of such vigour as to induce disruptive secondary exchanges through the stratification grid will a net exchange of primary coolant with the reserve tank occur. Such contingencies may be guarded against in the proper detailed design of the anti-convective shroud, carried out by persons skilled in the art of hydraulic design. Thus, the anti-convective shroud can be viewed as a passive device with the capacity to suppress the local exchange flows caused by dynamic and elevation pressure differentials in the port to which it is attached.
It was also implied above that, in the absence of special measures such as active primary parameter controls, or the passive compensating effects of the anti-convective shrouds, some residual exchange flows would persist in the reactor cooling system as a whole. Key points of behaviour of the anti-convective shroud which enable it to play a key role in limiting the extent of such residual exchange flow are described in the paragraphs immediately following.
JjConsider the vertical temperature profile of the coolant located in the stratification grid 31 of the anti-convective shroud exemplified in Figure 4. A transition in i~35 coolant temperature will be observed in passing from the top of the grid where the temperature is that of the local primary coolant, to the bottom of the grid where the temperature is nearly that of the reserve coolant. In a .. .

:,........ . - . ,. :, . .
, ... , .. :

13269~
- somewhat idealized example in which a reactor cooling system incorporating such a port is operating with absolutely zero net exchange flow passing through the port, the temperature profile in the grid may be characterized by the location of the median temperature plane located, say, midway between the upper and lower extremes of the grid and corresponding to the isothermal plane 32 identified in Figure 4. If, however, the sy~tem should experience changes in operating parameters such that a small branching flow of primary coolant enters the anti-convective shroud, then a net downward displacement of the median temperature ' plane in effect occurs. This shift in the elevation of the median temperature plane amounts to a lengthening of the column of hot coolant extending below the port. As this ` 15 column constitutes a component of the passive cooling circuit, the lengthening of this column contributes to the change in the net elevation pressure evaluated around the passive cooling circuit. Thus, the branch flow initiated by a disturbance in the operating parameters contributes to the creation of a component of driving pressure in the passive cooling circuit.
The remarkable feature of the anti-convective shroud in the configuration indicated is that, when such a shroud is added to each of hydrodynamic ports used in the cooling system arrangement exemplified in Figure 1, the induced change in the net elevation pressure induced in the , manner described above is always in the sense which tends ij to cancel the initiated branch flow, regardless of its cause. This observation applies regardless of whether it is the inlet port or the outlet port that is involved, or whether the initiating disturbance gives rise to an inflow from the primary circuit into the shroud, as in the above , example, or to an outflow from within the shroud into the primary circuit, in which case the median temperature plane shifts up rather than down in effecting the correct compensation.
It will be obvious that the maximum disturbance in the primary system that can be fully compensated by the ., .
~ ~ ~s - 34 -.

:

,, : .
: !
`~
.~. , : -,.. ';' : -i326916 interaction of coolant with the anti-convective shroud will depend, to a large extent, on the depth of the stratification grids for both the inlet and outlet ports, (although the performance of the system is not necessarily limited by the depth of the shallower grid of the system).
It may be clear that, if the median temperature plane in responding to primary disturbances should be moved vertically within the grid to a level where, for practical purposes, it approaches the upper or lower limit of a given stratification grid, further compensation may not be forthcoming from that port assembly. However, the partial compensation already achieved at that point will be sustained as long as the temperatures of the coolant residing over the full extent of the grids are sustained, as they surely will be due to the small, but persisting residual exchange flows passing through the quasi-static ; zones and the grids. For a given magnitude of initiating i disturbance, the persisting exchange flows will be much .!~ smaller in magnitude than would exist in the absence of the compensating behaviour of the described anti-convective shrouds and incorporated stratification grids.
~ In summary, moderate disturbances in the primary ; system operating parameters cause the anti-convective shrouds to accumulate residual exchange flows within their quasi-stagnant zones in such a way that the resulting elevation pressure differentials cancel the original tendency for the exchange flow. Thus, for a wide range of normal operating scenarios, residual exchange flows may be virtually eliminated. For more extreme disturbances, the induced exchange flows may persist, but at flow rates that have been reduced, by the presence of the anti-convective shrouds, to persisting values that may be acceptable in many reactor applications.
The particular design of the anti-convective shroud enabling the cooling system to meet the requirements of a given reactor application may be specified by those skilled in the art of thermal-hydraulic design. In addition to ' considering the adequacy of the anti-convective shrouds in ; ~ - 35 -;

~: .
. - . .
, ~ . - , , ,`, ~ ` ' " ~ `' ' .; . ' ' . ' limiting both local and systemic residual exchange flows, the design process must take account of the impact of the presence of shrouds on the performance of the passive cooling system. In each case, both the anti-convective shroud and the stratification grid must be dtasigned to neither significantly impede the operation of the passive cooling system operating in the shutdown mode, nor adversely affect the interaction of the passive circuit with the cooling system as a whole, in accident scenarios.
A second embodiment of the invention is shown in ',J the sectional elevation of the reactor plant of Figure 5.
;~ The cooling system of this plant has all of the essential attributes, and performs all of the basic functions, ascribed to the reactor plant of Figure 1. In addition, the embodiment shown in Figure 5 includes two anti-;~ convective shrouds 30 attached to the inlet and outlet hydrodynamic ports, 7 and 8, respectively. These shrouds, and their associated stratification grids 31, enhance the basic performance of hydrodynamic ports in their prescribed '20 roles according to the invention, in the manner described ;~in some detail with reference to Figure 4. Therefore, iduring normal operations, the reactor plant depicted in Figure 5 may be expected, even in the presence of significant operating disturbances and in the absence of actively compensating control equipment, to exhibit verylittle exchange flow between the primary circuit, as previously defined, and the reserve coolant tank 14.
Building upon the basic functions of the anti-convective shroud and the stratification grid, as described with reference to Figure 4, the compensating actions of these attachments on a reactor system as a whole, in response to inadvertent variations in the normal operating parameters, may now be readily understood in terms of the example provided in Figure 5.
Suppose the reactor plant of Figure 5 is operating initially under a set of nominal operating parameters which satisfy the zero net accumulative pressure requirement for zero exchange flow. The criterion for meeting this ~t~

... .
., .

, :. - . . .
.,~,...... . .

.. -:: .... , ,-13269~
requirement is that the change in the dynamic pressure, associated with a constant mass flow passing from port 7 to port 8, must equal the deficit by which the net elevation - pressure, assessed accumulatively around the passive cooling circuit, fails to match the resistive pressure losses associated with the primary mass flow through the core 1. The important point to be noted here is that, in assessing the net elevation pressure in terms of essentially the weight difference between a predominantly hot and a predominantly cold column, the height of both such columns is now effectively the vertical distance Lg between the median temperature planes falling at approximately the mid-elevation points in the stratification grids of the two ports, 7 and 8. ~g is now the operative hydrostatic height in satisfying the zero accumulative pressure requirement, as opposed to the height L which, in the absence of shrouds and grids as for Figure ;, 1, was taken as the operative height. A further important point is that the magnitude of Lg can vary in response to the receipt of small exchange flows within the shrouds, and such variation is the basis of the mechanism for compensating inadvertent departures of the cooling system from the zero net accumulative pressure requirement, which give rise to these flows initially.
Now suppose that a tendency develops within the operating system to produce ingress exchange flow, in which case an incipient combining branch flow occurs in port 7 and a corresponding dividing branch flow occurs in port 8.
In the embodiment of the invention depicted in Figure 5, 30 such a tendency may result, for example, from an : inadvertent reduction in pump speed, an increase in circulating primary coolant temperature, or an effective :~ (inertial) increase in the effective gravitational constant ' due to upward acceleration of the reactor. Aq the slight ingress exchange flow progresses, the median temperature plane in the grid of the inlet port rises and the median ~. temperature plane in the grid of the outlet port falls, in i; the manner described with reference to Figure 4. The ~ ~ - 37 -., ,.
~:
!.' :.

' ' ~ ' '~ ~ ' ' ' " ` ` ~ ' " . '.' , . ' " : `

1326~1~
consequent contraction in ~9 corresponds to the reduction in the elevation head that is necessary to cancel the cause of the initiating tendency for ingress exchange flow.
Suppose, on the other hand, that a tendency develops within the operating system to produce bypass exchange flow, in which case an incipient dividing branch ; flow occurs in port 7 and a corresponding combining branch flow occurs in port 8. In the reactor plant depicted in Figure 5, such a tendency may result from inadvertent 10 parameter variations, such as an increase in pump speed, an increase in reserve coolant temperature, or an effective decrease in the vertical component of Lg, if the reactor takes on a slightly inclined orientation. As the slight bypass exchange flow progresses, the median temperature 15 plane in the grid of the inlet port falls and the median temperature plane in the grid of the outlet port rises, again in the manner described with reference to Figure 4.
~ The consequent protraction of Lg corresponds to the increase 3 in the elevation head that is necessary to cancel the cause 20 of the initiating tendency for bypass exchange flow.
The foregoing operational scenario shows, by example, how the cooling system of the reactor depicted in Figure 5 provides, by virtue of the arrangement of the hydrodynamic ports, anti-convective shrouds and 25 stratification grids in accordance with the invention, the process-inherent means for regulating the cooling system during normal reactor operations in such a way that the potentially undesirable side-effects of the passive shutdown system on normal operations are automatically and 30 continuously curtailed. This situation contrasts with that :
for other reactors having passive shutdown cooling systems ~ which, as in the present invention, are deployed by ;~ process-inherent means. Such other systems depend on sophisticated sensors and active control systems to avoid ` 35 undesirable side-effects.
In accordance with the invention, the specification ^ of details relating to the capacities of the various component parts of the cooling system shown in Figure 5, h,~ -- 38 .. .. .
: ., :, . .. ,. ,. -.~ :, :, ~ . -- .. : -.... .. . .

~` 13269~6 when adapted to a particular reactor design and application, may be determined by those skilled in the art and science of thermal hydraulics. Since no coolant ` pressurizer is shown as part of the cooling system in Figure 5, it may be assumed that pressure and inventory control equipment, in this particular embodiment, are a part of the primary cooling circuit lying outside the reserve coolant tank and, hence, not shown. The location and design of such equipment is in the nature of elements to be specified by a skilled designer to suit a particular application.
The embodiment shown in Figure 5 also exemplifies the adaptation of the invention to applications which place severe restrictions on the size of the plant, while ~; 15 requiring passive cooling to become available as needed through inherent processes. In addressing such requirements, the vertical inter-port distance L (shown in Figure 5) has been made shorter than that which would be required to support the normal through-core flow rate by natural convection alone. Consequently, the zero net accumulative pressure requirement, as stated with reference to the passive cooling circuit, is met by making the axial flow area A2 f the outlet port 8 somewhat larger than the area Al of the inlet port 7. Moreover, in the interest of lessening the height and width requirements of the plant, the inlet port and the inlet duct are arranged horizontally, while the outlet port is oriented vertically - with its anti-convective shroud and stratification grid -. arranged symmetrically about the port axis. It should be noted that the individual components depicted in Figure 5 are not drawn to scale and no relative size or configurational relationships among the various components should be inferred literally from the drawing. The detailed design of such components, including the sizing for a particular reactor application, may be carried out by those skilled in the art of thermal-hydraulic design. The drawing is intended to convey, however, representative examples of how the basic principles of the ports and their f ~
~ ~ - 39 -., . ~ , :: , - , ~,.

132691~
attached shrouds offer flexibility in the design of reactors for which there is a stipulated requirement of passive shutdown cooling that becomes available, as needed, by process-inherent means.
5The passive cooling circuit of the reactor of Figure 5 is readily seen to include the anti-convective shrouds 30 and the stratification grids 31 among its ' components. Accordingly, the thermosyphon hot leg, operative during convective core coolant exchange with the lo reserve coolant, consists of the upper plenum 3 and the outlet duct 5 leading to the outlet port 8, plus the upper shroud 30 and grid 31. The effective vertical height of the hot leg is therefore approximately equal to ~, less the vertical distance from the middle of the upper port down to the median temperature plane of the upper grid.
Correspondingly, the effective vertical height of the cold leg is approximately equal to Lg, less the su~ of Ll and the vertical distance from the ~iddle of the lower port down to the median temperature plane of the lower grid. The layers of insulation 9 shown attached to the shrouds, as well as to the other components of the primary circuit, are consistent with the roles of the shrouds and grids as ~components of the passive cooling circuit. The idetermination of the optimal dimensions to be adopted in achieving adequate thermosyphon head, while conforming to imposed space limitations in a particular reactor design and application, may be performed by those skilled in the art of thermal-hydraulic design.
,In addition to accounting for both the potentially adverse effects of the inadvertent departures of operating parameters from their nominal value~, and the limited space available for the reactor plant in some applications, as ~ust described, the embodiment of the invention shown in Figure 5 also exemplifies applications in which vertical access to the reactor core must remain available for jpurposes of shielding placement, control and shutoff mechanism deployment, experimentation (as in the case of a research reactor), and refuelling and general maintenance.

.i - : .. :

- , :

. ~ : , . ~ . ; :

132691~
Access to the top of the reactor core is retained in the -design of the subject reactor primarily by offsetting the outlet duct 5 immediately as it leads away from the outlet plenum 3. This arrangement is readily accommodated by the - 5 flexibility of design yielded by the reactor cooling system in accordance with the invention, and leaves almost the entire space above the core open for the placement of equipment, instrumentation, or experimental apparatus, at the discretion of the designer.
10The reactor plant as depicted in Figure 5 shows typical utilization of the reactor access thus made available. For example, the reactor access tubes 40 may be used as control or shutdown rod guide tubes, or for in-core flux monitoring devices, in specific reactors requiring such facilities. The access tubes may also be used in research reactors for the insertion of irradiation target samples into the reactor core, or as beam tubes for the formation and extraction of neutron or gamma-ray beams.
The reactor vessel 46 of the plant in Figure 5 is shown in an abbreviated form. In this example, the component 41 -serves as a demountable, thermally insulated baffle which is essential during normal operation to prevent the primary coolant from mixing with the reserve coolant, but is obviously not exposed to very large pressure differentials.
In some applications, combining the baffle with reactor core shielding may be advantageous.
i~Access for servicing such facilities, as well as for performing refuelling operations, is obtained through the hatch 15 in the case of a closed or pressurized reactor, or simply through the open surface of the coolant in a pool-type reactor. For either type of reactor, it may be advantageous in the design to abandon component 41 and to extend the rim of the reactor vessel 46 to a much higher level than shown, or even to the level where it joins with the hatch (in a closed reactor), or breaks the coolant surface (in a pool-type reactor). Such arrangements may provide for continual access to the core while the reactor is operating, without interfering with the operation of the ` ~ - 41 -, . .. .

":. . . . .

.: . : ,. : . ~
, :: . . . : ; . . . .

6 9 1 ~
cooling system.
A third embodiment of the invention is shown in the sectional elevation of the reactor plant of Figure 6. The cooling system of this plant has all of the essential attributes, and performs all of the basic functions, which were generally ascribed to the reactor plant of Figure 5.
In addition, the embodiment shown in Figure 6 includes a multiplicity of inlet and outlet hydrodynamic ports, 7 and 8, each coupled hydraulically to a corresponding inlet or outlet plenum, 2 or 3, by means of a corresponding inlet or ;~outlet duct, 4 or 5. Leading in directions generally away from the reactor core, the inlet and outlet ports couple also to their respective inlet and outlet manifolds, 36 and 37. The manifolds, in turn, couple to the inlet and outlet conduits, 11 and 10, which connect with the remaining components of the primary circuit which, as in the first and second embodiments, lie outside the reserve coolant tank 14.
The arrangement of components in the third embodiment exhibits general s y etry about the central axis shown in Figure 6. When the reactor is at rest in its normal physical orientation, this axis is aligned with the vertical. Thus, the several inlet ports are normally found at a common elevation, as are the inlet ports. The arrows show the flow pattern of the primary coolant during normal operation, in which case the total upstream flow to the core is shared, more or less equally, among the several ` inlet ducts, and the downstream flow, correspondingly, by the several outlet ducts.
30Each of the hydrodynamic ports shown in Figure 6 is fitted with an anti-convective shroud 30 and a stratification grid 31, which are of designs similar to, , but not necessarily identical to, those of the second embodiment. Also shown in Figure 6 are such reactor components as fuel elements 18, a neutron reflector 19, gamma shielding, 16 and 17, a neutron shielding tank 38, and various components, 42, 43, 44, 45 and 39, relating to reactivity control. These components do not have a direct ,,~, , ................................ . . . .

. :.:~ . -. . .. .
~:.".
.,, ~ . .
: ~ . . .. .. .
; .~ ~ . :
.. :. - ~ ~ . : .:
..... . . .

1 3 2 ~
role in the functioning of the reactor cooling system ` shown, but are included in the embodiment to demonstrate how various reactor components and a reactor cooling system made according to the invention may co-exist in an actual reactor.
The plant shown in Figure 6 typifies applications in which the reactor cooling system, as well as performing in the manner described with reference to the plants of Figures 1 and 5, also meets the re~uirements of (i) passive shutdown cooling being always available, regardless of the orientation of the reactor with respect to gravity, and (ii) the normal operation of the reactor being tolerant of various types of dynamic motion and net displacement, both rotational and translational, imposed within specified limits on the reactor plant as a whole. The basic operational details whereby these requirements may be met, according to the invention, are now given with reference to the reactor plant depicted in Figure 6 and defined in the preceding paragraphs. In proceeding, an understanding of the operation of the first and second embodiments of the invention is presumed.
As already indicated, the primary coolant of the ; reactor in Figure 6 is circulated during normal operation according to the pattern depicted by the arrows. The flow to the core 1 is delivered in more or less equal shareæ by the several inlet ducts 4, and each share is received by a duct as an axial flow transmitted by the corresponding inlet port 7. Similarly, the flow out of the core is received in more or less equal shares by the several outlet ducts 5, and each share is transmitted by a duct to become an axial flow in the corresponding outlet port 8.
Normal operating conditions, in the application environments for which the embodiment shown in Figure 6 is eminently suitable, include an alignment of the central 3S axis of the reactor to coincide more or less with the , vertical, and an absence of motion of the platform on which ,3 the reactor is mounted. These stipulations for normal ~ operations are in addition to those relating to the key ., - ~ - 43 -.
., .
,,. , , : : , . . , :

.: : - : , ..

13269~ ~
design parameters, including the mass flow rate, the circulating coolant temperature profiles, and the reserve coolant temperature, which are chosen, as for the first two embodiments, in the optimization of the plant's basic thermal efficiency.
The operation of the reactor of Figure 6, under normal conditions as defined above, may be described in terms essentially the same as those used previously in describing the reactors of Figures 1 and 5. For example, any tendency towards exchange flow between the circulating primary coolant and the reserve coolant is minimized for all three reactors because the axial flow areas of the hydrodynamic ports are chosen so that the zero net - accumulative pressure requirement is fulfilled. This requirement may be restated more aptly, however, for the specific case of the multiple port configuration of Figure 6: At normal operating conditions, the net pressure from all effects accumulative around each of the many identifiable passive cooling circuits, and tending to support net flow around any such circuit, must be substantially zero. A passive circuit, in this context, can be defined as any closed path in the system which `includes the reserve coolant tank, and the two ducts connecting with any arbitrarily chosen pair of hydrodynamic ports, including pairs of inlet ports, pairs of outlet ports, and pairs formed of any combination of one of each.
As in the case of the second embodiment shown in Figure 5, the sizing of the port areas for the third embodiment shown in Figure 6 is based on Lg, the operative hydrostatic height now defined as the vertical distance between the inlet and '!
outlet groups of median temperature planes 32, as well as the nominal values of the key operating parameters.
Either of the two versions of hydrodynamic port lpreviously described in reference to Figures 2 and 3, or ;~35 indeed a further version, may be chosen for application in the reactor of Figure 6, depending on the degree of resistance needed against the tendency toward exchange flow. In the meanwhile, the anti-convective shrouds 30 . , :~, ,".
~l- 44 -:'~
. ~
".~..
, . ..

~: . :

13269~
eliminate or isolate the effects of local exchange flows in the multiplicity of ports, and the self-adjusting capability of the median temperature planes 32, occurring within the various stratification grids 31, serve to automatically eliminate, or limit, incipient exchange flows. As already described in considerable detail with reference to the reactor of ~igure 5, such exchange flows are brought about by inadvertent departures of the normal operating parameters from normal. Additional sources of tendency toward exchange flow arise in the configuration of the passive cooling system of Figure 6, however. These may be due to the presence of minor non-symmetries in the manufacture of the various hydrodynamic components relating to the multiplicity of ports and associated flow paths.
The disposition of this source of exchange flow tendency will be addressed later, along with the effects of various types of dynamic motion and net displacement imposed on the reactor plant by its operating environment.
The multiplicity of passive cooling circuits addressed in connection with the maintenance of ideal operating conditions under normal (upright) conditions, are also the key to the main feature of the third embodiment of the invention, namely, the continuous availability of the passive cooling function, regardless of the orientation which the reactor may assume with respect to gravity. The operation of the provisions for passive cooling may be readily understood with reference to the plant depicted in Figure 6. It is assumed for the present discussion that the axial flow areas of the ports either remain the same or increase, in passing from the inlet to the outlet port.
Consider the operation of the passive cooling system, following the failure of the circulating pump and j the general stoppage of flow in the part of the primary circuit external to the reserve coolant tank, while the reactor, for the time being, remains upright. Under these conditions, the mass flow through the core reduces to the level which can be supported by natural convection in the passive cooling circuits. In this, more or less equal i" . i . ,. ,: , .: . , 1326~
flows from the reserve coolant tank enter the inlet ports through the shrouds of the inlet ports, pass through the core, and emerge into the reserve tank as equal flows through the shrouds of the outlet ports. Apart from the multiplicity of flow paths, the general behaviour of the system under these conditions is essentially the same as for the reactors of only two ports as in the second embodiment of Figure 5. As for that reactor, the rate of core cooling remains adequate, in spite of the reduced flow, because the core inlet temperature is now determined largely by the relatively low reserve coolant temperature.
If, in the meantime, the reactor becomes nominally shutdown, the demands on the passive cooling system are further reduced and the reactor remains adequately cooled for an indefinitely long period.
Consider now the case in which the reactor's physical orientation departs to an arbitrary degree from the upright position. It may be inferred from Figure 6 that, regardless of any such orientation, a variety of viable passive cooling circuits is always available for natural convection. Such circuits were defined, in the ~ discussion on the requirements for zero exchange flow under i normal operating conditions, as including the reserve tank, ' and two of the ducts connecting with any pair out of the multiplicity of hydrodynamic ports. Depending on the degree of disorientation from the normal, the various passive cooling paths may operate in concert with each other and carry different components of the total flow 3~ through the core. Certain aspects of the detailed design of the cooling system require special attention in respect of passive cooling being available at all reactor orientations. These include the fact that the majority of the passive coolant flow in the core will be, for a range of reactor orientations, in directions other than parallel to the general orientation of the fuel elements, assuming j a standard fuel arrangement. Another aspect requiring j attention is the fact that the anti-convective shrouds, which are designed in the first instance to optimize the . .
: ~ .
..... ~ . . ...

...:.
.. - ~

.2~
;~ resistance of the cooling system to disturbances occurring ` during normal operation, give rise to non-symmetries among the various passive cooling circuits. It may be noted, however, that the elongation of the shrouds and the associated grids, a desirable measure for enhancing normal operation, does not reduce the convective driving head for ~any of the passive cooling circuits more severely than it ;does in the (worst) case of the reactor in the upright position. These and similar aspects of design may be dealt with appropriately by persons skilled in the art of thermal-hydraulic design.
As just described with reference to Figure 6, the main feature of the reactor plant, showing the third embodiment of the invention, is the availability of effective passive cooling at all times and in all circumstances, including all physical orientations of the reactor plant. Such facility for passive cooling would be of limited value, however, if the presence of such facility were to degrade siqnificantly the efficiency of normal operations, or if the measures required to avoid such degradation were to compromise the operability of the plant. Moreover, the operating environment that requires a passive cooling capability at all physical orientations is likely to be a dynamic environment ~as on board ship) for which the effects on normal operation have to be .accommodated, in addition to the effects of inadvertent variations in the normal operating parameters as already discussed for all three embodiments. The manner in which the normal operation of the reactor of Figure 6 is tolerant of the various types of dynamic motions and net displacements, is now described.
Consider first an incremental rotational displacement of the reactor in Figure 6, operating 'jinitially with an upright orientation under ideal normal l35 conditions of no exchange flow. Assuming that the rotation ,.is clockwise about an axis corresponding to a normal to the plane of the paper of Figure 6, the distortion of the original flow symmetries due to the changes in elevation ;~

~' 132691~
heads results in a dividing branch flow at the left-most outlet port, and a combining branch flow at the right-most inlet port. The resulting movements of the median temperature planes in the grids of the two ports, due to the accumulation within the shrouds of minor amounts of exchange flow, causes the value of Lg, as it pertains to the two ports in question, to tend to shorten toward its original value, even though an increased vertical separation has occurred between these two ports.
;10 Similarly, the occurrence of combining branch flow in the right-most outlet port, and dividing branch flow in the left-most inlet port, tends to lengthen the value of Lg, as ,it pertains to these two ports, to its original value, even thouqh the two ports now physically have less vertical separation. Thus it may be seen for this simple example of rotational displacement that, by the interaction of incipient residual exchange flows with the anti-convective shrouds and stratification grids, the parameter Lg self-:adjusts within the incrementally rotated system to satisfy the hydraulic requirements for zero exchange flow with respect to all four ports simultaneously. On extending 'this line of reasoning to more general circumstances, it may be concluded that the parameter Lg will automatically adjust to the value satisfying the zero flow requirement with respect to all hydrodynamic ports in the systemsimultaneously, including those not visible in Figure 6, even when the plant is subjected to arbitrarily complex rotational displacements.
At certain limits of reactor inclination, it is apparent that the range of useful self-adjustment of Lg in the inhibition of exchange flow becomes exhausted. Such a limit occurs when the plant inclination becomes so great that the stratification grids of two diametrically opposite ports can on longer both intersect a single horizontal plane. It is apparent, also, that the said limit of inclination in radians is approximately equal to the ratio of the (vertical) grid length to the diametrical separation of the said ports. ~herefore, for a given basic reactor ,:~
.~

" .
., .. . , . . , .: .

~. .. I

13~`6~

confiquration, the range of inclination within which the reactor plant can remain relatively free of exchange flow is proportional to the length of the stratification grid in the anti-convective shroud of each hydrodynamic port.
While lengthening the grids in the design of a reactor for a given mobile application would increase the reactor's immunity to exchange flow, the indiscriminate use of such measures could compromise the effectiveness of the reactor's passive cooling system. The reduced effectiveness, caused by the effective shortening of the thermosyphon hot leg, would be the most severe for a nearly upright reactor, as suggested earlier. The task of optimising the reactor design for a particular application - environment, in which an optimal balance between immunity to exchange flow and the effectiveness of the passive cooling is achieved, may be performed by those skilled in the art of hydrodynamic design.
It was stated earlier that any tendency towards exchange flow between the circulating primary coolant and the reserve coolant may be minimized for all three embodiments, because the axial flow areas of the hydrodynamic ports are chosen so that the zero net , accumulative pressure requirement is fulfilled. This ~, statement implies that, in a most likely configuration of `1 25 the third embodiment of the invention, i.e., one in which 3 the multiplicity of ducts and ports are arranged i symmetrically about the principal axis, a certain quality of manufacture is achievable. The quality of manufacture , would have to be such that the total mass flow would be 30 shared identically by all ducts, and the specified axial flow areas of the hydrodynamic ports would precisely ~A materialize in the manufacturing process, to the extent that the zero net accumulative pressure requirement would be truly satisfied simultaneously for all passive cooling 35 paths to be found in the system. Since it may be impractical, however, to manufacture the hydraulic ,~ equipment to such precision, it may appear that some form of post-assembly adjustment might be in order. Mechanisms ~ .
. ~ .

: : :.... . .
. :~ . . .
:: : ,., ~ , .. .
:,, . : ~ . . :: . -132~9~6 for such adjustment might be suggested by those skilled in both thermal-hydraulic and mechanical design. However, it should be pointed out that, through the line of reasoning used previously to show that the anti-convective shrouds and stratification grids automatically adjust system elevation pressures to inhibit exchange flow in the face of disturbances in the primary parameters and changes in reactor orientation, it may be shown that tendencies toward exchange flow arising from the practical limitations in ;10 component manufacturing are likewise compensated, within reasonable limits, by the self-adjusting properties of the shroud-equipped hydrodynamic ports. It follows, as a corollary, that any minor changes in the geometry of the hydraulic equipment, due to corrosion, deposition, or deformation associated with aging, are likewise compensated automatically.
In consideration of the effects of dynamic motions on the integrity of normal reactor operation, two types of motion are important in the operating environment ~20 anticipated for the reactor of Figure 6, depicting the -~jthird embodiment of the invention. The types of motion are (i) rotational oscillations of limited amplitude about horizontal axes, and (ii) vertical translational oscillations of limited magnitude.
Consider first the rotational oscillations. Here we consider only rotations about horizontal axes located approximately at the level of the reactor core. (The 1 effects of oscillations about axes at other levels may be 3 considered approximately in terms of a superposition of a translational component of motion on a rotational component about a horizontal axis through the core.) Such rotational : oscillations, typified as a rocking motion about an axis ;, through the core and normal to the paper in the representation in Figure 6, can be considered to potentially create exchange flow in two ways, i.e., through inertial (accelerative) effects and through the displacement effects, which will be dealt with in turn.
Since the rotations are defined to be about the , .
. .
.- -- ~ , . - ~ .
~`.'.-' ' ' ~ , ;
. ,:

,- .;.

1~26~1~

core, the forces of centrifugal acceleration on the coolant in the various ducts, 4 and 5, will tend to cancel.
However, the scooping effect, as the opening of an anti-convective shroud 30 accelerates into the reserve coolant 35, tends to produce an exchange flow of the "inlet port to inlet port" variety, and similarly for the outlet portæ.
While such an inertial effect persists, an incipient exchange flow actually occurs and results in the displacements of the median temperature planes 32 in directions that tend to cancel the said inertial effect.
Since the inertial effect reverses direction for the second half of the oscillation period, the displacement of the median temperature planes 32 will be reversed. If this reversal takes place before the ranges of travel in the stratification grids 31 become exhausted, as will be the case ideally, no net exchange flow will be experienced as a result of the scooping effect. The design factors which tend to minimize the probability of exchange flow due to the scooping effect, as just analyzed, are (i) high resistance of the ports to branch flow at nominal operating conditions, by method such as those that have been described earlier in accordance with the invention, and ~(ii) a reasonable capacity of the anti-convective shrouds -~to accumulate incipient exchange flow~ Such factors may be -'25 appropriately addressed in the detailed design process by ,those skilled in the art of thermal-hydraulic analysis.
The strictly displacement effects of rotational oscillations may be analyzed using a similar line of reasoning to the one used previously in addressing the effects of net (static) displacements. In the case of very slow oscillations, the very same criterion for inhibiting exchange flow applies as for static displacements, namely, the maximum rotational displacement from the upright orientation of the reactor in radians should not exceed the static limit, namely, the ratio of the length of the stratification to the diametrical separation of the outlet or the inlet ports. As the frequency of the oscillations ;~ speeds up for a specific reactor configuration, however, , ~ ~
.....

;
,1 1",'', ~ , .~, . : , . , ~ .

~32~6 ..~
there will be an increase in the maximum amplitude of oscillation that can be allowed before a net exchange flow occurs, provided that the mid-position of the oscillations coincides with the reactor upright position. The i 5 permissible amplitude increases because the incipient exchange flow rate is limited by the resistance of the ports to branch flow at the nominal operating conditions, and the reverse swing in each oscillation causes a reversal of the incipient exchange flow, before the available range travel of the median temperature plane in each port becomes exhausted. It follows, therefore, that the period of oscillation at which dynamic amplitudes may be allowed to exceed the maximum permissible static displacement corresponds approximately to twice the time required for the incipient exchange flow to displace the median temperature plane by half the length of the associated grid, when the inclination of the reactor is at the static ~'~ limit. For shorter periods of oscillation, the permissible amplitude increases more or less inversely with the period ;3 20 in a manner that depends on the resistance of the ports to branching flows at nominal operating conditions and on the capacity of the anti-convective shrouds.
Finally, the effects of vertical translational oscillations on the performance of the third embodiment of the invention are considered. With reference to the reactor of Figure 6, suppose the reactor operating normally is subjected to vertical oscillation as may arise if the $. reactor is mounted on board a ship being subjected to surface wave motion. The effect of such motion on reactor ` 30 operation may be expressed in terms of a periodic modulation~of the net elevation pressure evaluated around ;~ any one of the passive cooling circuits which were defined ` previously. It may be clear, therefore, that, for that `Jl part of the oscillatory cycle corresponding to the ship being on the crest of the wave, there will be a tendency for bypass exchange flow in the reactor cooling system. On ~ the other hand, for that part of the cycle corresponding to ,.''~J a depression, there will be a tendency for ingress exchange ... .

I ~ - 52 -.
..... ~ 7 ,, ' ' " ' ' , ~ . ~ ' ' ' ` ' ' 1326~1~
flow in the system. Whether these tendencies lead to actual periodic exchange flows between the primary circuit and the reserve coolant depends on a number of factors in common with those that were considered in connection with the rotational oscillations. The resistance of the hydrodynamic ports, 7 and 8, to branch flow, at close to nominal operating conditions, tends to limit the maximum rate of accumulation of incipient exchange flow in the anti-convective shrouds 30 during each cycle. The capacity of each anti-convective shroud 30 determines the fraction of the range of the stratification grid 31 that is traversed by the advancing median temperature plane 32 during the first half of each cycle, before it recedes in the second half. If the ports, shrouds and grids are ` 15 specified so that, for the entire spectrum of anticipated oscillation frequencies, wave amplitudes do not exceed the values which push the median temperature planes beyond the -~ range provided by the grids in a single cycle, then such motions need not result in any actual exchange of primary coolant with the reserve tank. The specification of the components to meet these conditions may be accomplished by persons skilled in the art of hydraulic design.
It is in the nature of the hydraulic processes at j work, in both the cause and the compensation of the various influences that tend towards exchange flow, that the corrections generated in opposition to the various causes, `! according to the invention, will be superimposed appropriately within the system, and the many normal ~ operating points of concern are thus diminished by the `;i 30 described passive corrective actions operating in parallel.
The identification of the limiting sources of perturbation on the normal operating system that dominate the tendency i for exchange flow, and to make the necessary design choices to adequately control the effects in a particular application, may be carried out by those skilled in the art of hydrodynamic design.

, ~ - 53 -"

.:: , . . . .. , ~ . . .. . . . .

' ' , , , ' `' .
, ' ' ~ ,,' ' ., ' , . '

Claims

1. A cooling system for use within a nuclear reactor plant having a primary cooling circuit comprising a reactor core, an inlet duct and inlet plenum for conducting coolant into said core, an outlet plenum and an outlet duct for conducting coolant out of said core and means for circulating coolant through said primary cooling circuit, said cooling system comprising (a) a reserve coolant tank for containing coolant; (b) a first port means in said inlet duct upstream of said core, having a primary flow channel for conducting coolant through said inlet duct and having exchange flow channel means for conducting coolant between said reserve coolant tank and said primary flow channel, said primary flow channel being open to said reserve coolant tank through said exchange flow channel means; (c) a second port means in said outlet duct downstream of said core, and spaced generally above said first port means, having a primary flow channel for conducting coolant through said outlet duct and having exchange flow channel means for conducting coolant between said primary flow channel and said reserve coolant tank, said primary flow channel being open to said reserve coolant tank through said exchange flow channel means; (d) wherein the ratio of the area of flow in said primary flow channel of said inlet duct to the area of flow in said primary flow channel of said outlet duct is selected so as to minimize exchange flow through said exchange flow channel means in said first and second port means during normal operation of said primary cooling circuit; (e) whereby during normal, unimpaired operation of said primary cooling circuit, coolant circulates through said primary cooling circuit and only minimal volumes of coolant flow into said reserve coolant tank from said primary cooling circuit or into said primary cooling circuit from said reserve coolant tank through said exchange flow channel means; (f) whereby during impairment of said primary cooling circuit, or other circumstance of mismatch between the energy generation and dissipation rates of said nuclear reactor plant, a convective flow of coolant materializes without mechanical or operator intervention, wherein coolant in said reactor core flows by convection through said outlet plenum and outlet duct and through said exchange flow channel means in said second port means into said reserve coolant tank, and coolant in said reserve coolant tank flows into said exchange flow channel means in said first port means, through said inlet duct and inlet plenum and into said reactor core, forming an exchange flow cooling circuit, whereby a natural convective flow of coolant results.

2. A cooling system according to claim 1 wherein said ratio of flow areas is selected to provide net zero pressure change around said exchange flow cooling circuit to minimize exchange flow through said exchange flow channel means during normal operation of said primary cooling circuit.

3. A cooling system according to claim 1 wherein said flow area of each exchange flow channel means is selected so as to minimize exchange flow through said exchange flow channel means during normal operation of said primary cooling circuit while maximizing exchange flow during impairment of said primary cooling circuit.

4. A cooling system according to claim 1, 2 or 3 wherein under conditions of departure in either direction from the condition of zero net exchange flow, namely during net inflow of reserve coolant into the primary cooling circuit at either the first or the second port means, during normal operation, stable flow conditions persist at all stages of transition from exchange flow in one direction to exchange flow in the other.

5. A cooling system according to claim 1 wherein said first and second port means are positioned within said reserve tank.

6. A cooling system according to claim 1 wherein that said reactor core and said first and second port means are positioned within said reserve coolant tank.

7. A cooling system according to claim 1, 2 or 6 wherein said first and second port means are so configured that particles of coolant passing from said primary flow channel into said exchange flow channel means, and particles of coolant passing from said exchange flow channel means into primary flow channel in said first port means and second port means, undergo large momentum changes.

8. A cooling system according to claim 7 wherein said exchange flow channel means comprises a flow channel positioned at about 90° to said primary flow channel.

9. A cooling system according to claim 1, 2, 3 or 6 wherein said first and second port means includes a plurality of substantially parallel annular plates, spaced from each other and defining slots therebetween, said plates being mounted concentrically with the adjacent duct segments leading to and from said port means, the bore through said plates defining said primary flow channel, said slots defining said exchange flow channel means.

10. A cooling system according to claim 9 wherein the two slots defined by the two pairs of outermost plates which are nearest said adjacent duct segments are closed at the outer circumference of said plates, thus defining blind slots between each pair of said outermost plates at the adjacent ducts, and a plurality of bypass ducts is provided between said blind slots parallel to and spaced generally symmetrically about said bore through said plates.

11. A cooling system according to claim 1, 5 or 6 further comprising an anti-convective shroud operatively associated with each of said first port means and said second port means, each of said shrouds comprising a hood enclosing said port means and spaced from said exchange flow channel means, said hood having walls defining a downward opening into said reserve coolant tank.

12. A cooling system according to claim 11 wherein said anti-convective shrouds further comprises means in said downward opening for maintaining stratified layers of coolant in said shrouds.

13. A cooling system according to claim 12 wherein said means for maintaining stratified layers of coolant comprises a stratification grid, said grid comprising a first array of substantially vertically-oriented parallel plates and a second array of substantially vertically-oriented parallel plates positioned at an angle to said first array.

14. A cooling system according to claim 13 wherein said stratification grid is of sufficient depth as to accommodate changes in vertical temperature profiles within said anti-convective shrouds in a manner such that internally-generated adjustments to static pressure heads cancel externally-induced tendencies toward unwanted exchange flows during normal operation of said primary cooling circuit.

15. A cooling system in accordance with claim 1, 2, 3 or 5 wherein said inlet duct and/or said outlet duct is offset from the axis of said core to facilitate access to said core.

16. A cooling system for use within a nuclear reactor plant having a primary cooling circuit comprising a reactor core, a plurality of inlet ducts and inlet plenum for conducting coolant in to said core, an outlet plenum and a plurality of outlet ducts for conducting coolant from said core, said inlet ducts and said outlet ducts being configured with substantially radial symmetry about a vertical axis through the center of said core, and means for circulating coolant through said primary cooling circuit, said cooling system comprising: (a) a reserve coolant tank for containing coolant; (b) a first port means in each of said inlet ducts upstream of said core, each of said first port means having a primary flow channel for conducting coolant through said inlet duct and having exchange flow channel means for conducting coolant between said reserve coolant tank and said primary flow channel, said primary flow channel being open to said reserve coolant tank through said exchange flow channel means; (c) a second port means in each of said outlet ducts downstream of said core, and spaced generally at higher elevation than each of said first port means in the normal orientation of said reactor plant, each of said second port means having a primary flow channel for conducting coolant through said outlet duct and having exchange flow channel means for conducting coolant between said primary flow channel and said reserve coolant tank, said primary flow channel being open to said reserve coolant tank through said exchange flow channel means; (d) wherein the ratio of the area of flow in each of said primary flow channels of said inlet and outlet ducts to the area of flow in all other of said primary flow channels of said inlet and outlet ducts is selected so as to minimize exchange flow through said exchange flow channel means in said first and second port means during normal operation of said primary cooling circuit; (e) whereby during normal, unimpaired operation of said primary cooling circuit, coolant circulates through said primary cooling circuit and only minimal volumes of coolant flow into said reserve coolant tank from said primary cooling circuit or into said primary cooling circuit from said reserve coolant tank through said exchange flow channel means; and (f) whereby during impairment of said primary cooling circuit, or other circumstance of mismatch between the energy generation and dissipation rates of said nuclear reactor plant, a convective flow of coolant is initiated without mechanical or operator intervention, and irrespective of the orientation of said reactor plant with respect to gravity, wherein coolant in said reactor core flows by convection generally upward through at least some of said outlet ducts and/or inlet ducts positioned at a higher elevation than said reactor core and through said exchange flow channel means in said second and/or first port means in said ducts and into said reserve coolant tank, and coolant in said reserve coolant tank flows into said exchange flow channel means in said first and/or second port means in at least some of said inlet and/or outlet ducts positioned lower than said core and into said reactor core, forming exchange flow cooling circuits, whereby a natural convective flow of coolant results.

17. A cooling system according to claim 16 wherein said flow area of each exchange flow channel means is selected so as to minimize exchange flow through said exchange flow channel means during normal operation of said primary cooling circuits while maximizing exchange flow during impairment of said primary cooling circuit.

18. A cooling system according to claim 16 wherein said ratio of flow areas is selected to provide for zero net pressure change around said exchange flow cooling circuits, to minimize exchange flow through said exchange flow channel means during normal operation of said primary cooling circuit.

19. A cooling system according to claim 16, 17 or 18 wherein under any mode of conditions of departure from the condition of zero net exchange flow during normal operation, stable flow conditions persist at all stages of transition from exchange flow in any given mode to exchange flow in any other mode.

20. A cooling system according to claim 16 wherein said reactor core and said first and second port means are positioned within said reserve coolant tank.

21. A cooling system according to claim 16 wherein said first and second port means are configured that particles of coolant passing from said primary flow channel into said exchange flow channel means, and particles of coolant passing from said exchange flow channel means into said primary flow channel in said first port means and second port means, undergo large momentum change.

22. A cooling system according to claim 21 wherein said exchange flow channel means comprises a flow channel positioned at about 90° to said primary flow channel.

23. A cooling system according to claim 16, 17, 18 or 20 wherein said port means includes a plurality of substantially parallel annular plates, spaced from each other and defining slots therebetween, said plates being mounted concentrically with the adjacent duct segments leading to an from said port means, the bore through said plates defining said primary flow channel, said slots defining said exchange flow channel means.

24. A cooling system according to claim 23 wherein the two slots defined by the two pairs of outermost plates which are nearest said adjacent duct segments are closed at the outer circumference of said plates, defining blind slots between each pair of said outermost plates at the adjacent ducts, and a plurality of bypass ducts is provided between said blind slots parallel to and spaced generally symmetrically about said bore through said plates.

25. A cooling system according to claim 20 further comprising an anti-convective shroud operatively associated with each of said first port means and said second port means, each of said shrouds comprising a hood enclosing said port means and spaced from said exchange flow channel means, said hood having walls defining a downward opening into said reserve cooling tank.

26. A cooling system according to claim 25 wherein said anti-convective shroud further comprises means in said downward opening for maintaining stratified layers of coolant in said shroud.

27. A cooling system according to claim 26 wherein said means for maintaining stratified layers of coolant comprises a stratification grid, said grid comprising a first array of substantially vertically-oriented parallel plates and a second array of substantially vertically-oriented parallel plates positioned at an angle to said first array.

28. A cooling system according to claim 27 wherein said stratification grid is of sufficient depth as to accommodate changes in vertical temperature profiles within said anti-convective shrouds in a manner such that internally-generated adjustments to static pressure heads cancel externally-induced tendencies toward unwanted exchange flows during normal operations of said primary cooling circuit, that include the reactor platform being subjected to routine motions.

29. A cooling system in accordance with claims 16, 17, 18 or 20 wherein said inlet ducts and/or said outlet ducts are offset from the axis of said core to facilitate access to said core.