Appendix A – Additional data on PBMR-400 and PUMA simplified models

Appendix A – Additional data on PBMR-400 and PUMA simplified models

A.1 Introduction

This Appendix is mainly based on [1-1] written in the frame of PUMA project[I-5]. These data are reported in order to give a deeper description of the analyzed plant. There is also a brief analysis of one of most controversial plant component: the Cross Duct. That is the piping connecting the pressure vessel with the steam generator (in the case of a direct cycle plant) or with the turbine. Its break, that is generally linked to a P-LOFC (pressurized loss of forced circulation) or to a D-LOFC (depressurized loss of forced circulation), is considered a DBA Accident. Indeed it could be the initial event of the potential air ingress into the vessel, which is considered a severe accident for this kind of reactor ¹ . For this reason the cross-duct is one of the most analyzed components in HTR plant design. Nevertheless the response of these studies show that this type of reactor respects some goals (Generation IV) like the “major intrinsic safety”. Finally it is important to make a confrontation with this type of accident and the response of Safety Passive Systems[1-4].

A.2 Equilibrium specifications

The equilibrium core is defined as the reactor operational state achieved after a considerable time of operating at a specific set of conditions. For the reference problem the operating conditions are defined to be at full power and with the control rods inserted 2.0 m below the bottom of the top reflector (and therefore 1.5 m alongside the pebble-bed). Once equilibrium is reached no significant changes can be observed in the properties of the core.

For example the k eff , power profile, temperatures and isotopic concentration distribution do no longer change.

A.3 Boundary conditions

The following table A.1 shows some important data useful to calculate the

neutronic characteristics.

Table A.1: Neutronics boundary conditions[1-1]

Description Unit Value Neutronic model boundaries

Radial; outer boundary of the reactor barrel m 2.925 Top (beyond 150 cm of top reflector, i.e. 2 metres

above the core) m - 2.0

Bottom (beyond 150 cm of bottom reflector) m 12.5 Type of boundary conditions (on all boundaries) BLACK

A.4 Material specifications

As recolled in the previous paragraph, the Tables A.2 and A.3 shows other interesting characteristics connected with some material specifications.

Table A.2: Structural Material Specifications and densities[1-1]

Description Unit Value

The reflector graphite density. Central

column, top, bottom and side reflector g·cm ^-3 1.78 The reflector graphite density used in the

RSS, RCS, Riser channel and inlet / outlet plenums

g·cm ^-3 1.78

Density of RPV: Iron g·cm ^-3 7.8

Density of Core barrel: Iron g·cm ^-3 7.8

Table A.3: Control Rod and Shutdown Specifications[1-1]

Description Unit Value

Thickness of gray-curtain region

representing control m 0.115

Distance between core outer diameter and inner diameter of control rod grey-

curtain region

m 0.0795

Homogenised number density of B-10

representing control system barn ^-1 ·cm ^-1 6.0E-6 Density of graphite in the gray curtain

region g·cm ^-3 1.78

A.5 The Cross-Duct

The Cross-Duct is the biggest pipe of the PBMR-400 reactor. It is composed of two concentric pipes, of which the hotter is placed inside the colder. The wall of each pipe does not consist of a single layer, but of multiple layers.

This complex structure is justified by few principles:

• Reduction of the component numbers and simplification of the structures of the reactor

• Reduction of heat losses and increase in the performance of the power plant

• Reduction of He losses with multiple barriers

Containing He in the primary circuit is another important issue in PBMRs. The

unique behaviour of He, which infiltrates all fractures or gaps of the structural

material because of its high diffusion coefficient, tends to increase its own

concentration in environments around the pressure vessel. For this reason it

is difficult to set up the control system to check the concentration of He in

the surrounding air and to find out breaks of the primary system. Hence, in

order to decrease this diffusion capability of He, a complex geometry may be

an important design parameter. Both pipes of the cross-duct are designed on

the basis of these principles, along with the usual mechanical and safety

A.6 Definitions and Analyses of LOFC

A.6.1 Introduction

For the analysis of the accident sequences, the Graphite Reactor Severe Accident Code (GRSAC) is used. The development, use, and validation exercises of this code began over 25 years ago starting also from several predecessor codes[A-4]. Current interest in GRSAC[A-2] involves the simulation of accident scenarios and of benchmark transients run on the HTTR (Japan) and HTR-10 (China). GRSAC employs a detailed (~3000 nodes) 3-D thermal-hydraulics model for the core, plus models for the reactor vessel, shutdown cooling system (SCS), and shield or reactor cavity cooling systems (RCCS). There are options to include anticipated transients without scram (ATWS) accidents and to model air ingress accidents, simulating the oxidation of graphite and other core materials.

The spectrum of accidents covered ranges from what are normally classified as design basis accidents (DBAs) to accidents well-beyond DBA with extremely low probabilities. Typically the initial event is assumed to be a loss of forced circulation (LOFC), which may or may not be followed by a scram or start up of SCS. If the primary system maintains pressure, the event is termed P-LOFC (pressurized LOFC). The LOFC may be accompanied by primary system depressurization (D-LOFC). The D-LOFC can include air ingress and graphite oxidation, where air circulation is driven either by buoyancy (chimney) effects from single break or double breaks, or by forced circulation.

A.6.2 P-LOFC

The reference case P-LOFC assumes a flow coastdown and scram at time zero, with the passive RCCS operational for all the accident duration[A-2].

The natural circulation of the pressurized helium coolant within the core

tends to make core temperatures more uniform, therefore lowering the peak

temperatures, than would be the case for a depressurized core, where the

buoyancy forces would not establish significant recirculation flows. The

chimney effect in P-LOFC events also tends to make the core (and vessel)

temperatures higher near the top. The peak fuel temperature of 1266°C

occurs at ~37 hours, with a maximum reactor vessel temperature of 501°C at

77 hr. Sensitivities to variations in the emissivities of the vessel and RCCS are

nearly identical to those for the GT-MHR.

A.6.3 D-LOFC

In the D-LOFC reference case “conduction/heat-up” accident, T fuel max at 1517°C occurs at 77 hr after the accident begin, and for this configuration, maximum temperatures for the reactor vessel (SA 508) and core barrel (316 SS) are not of concern.

The PBMR on-line refuelling results in a mixing of pebbles with various burn- up and irradiation histories, and the effective core conductivity is usually considered to be primarily due to the radiant heat transfer between pebbles, and so it is only a function of temperature. The reference conductivity correlation is derived from a combination of the Zehner-Schlunder and Robold correlations (Fig. A-1).

Variations on this “reference case” show the sensitivity of peak fuel temperature for changes as follows:

1) 25% decrease in core conductivity: 165°C increase in T _fuel ^max

2) Use of the THERMIX code default core conductivity correlation: 64°C increase in T fuel max

3) Use of the core conductivity correlation derived from SANA tests at KFA by H. F. Niessen (see Fig. A-1): 103°C decrease in T _fuel ^max

4) 15% increase in afterheat: 121°C increase in T fuel max

5) 20% increase in maximum radial peaking factor: 17°C increase in T fuel max

Figure A-1: Comparison of heat transfer calculated with Zehner Bauer Schlünder and measured heat transfer[A-5]

A.6.4 D-LOFC with air ingress

These accidents assume that the D-LOFC is followed by ingress of ambient

air into the primary system, either just after the depressurization is complete

Key factors are the net air flow rate into the reactor vessel and core, and ultimately the “availability” of fresh air over the course of the accident. The net air flow through the core is mainly dependent on the buoyancy forces due to differential temperatures and the flow resistances in the core. For a single “break” or opening in the primary system, calculations and experiments have shown that it may take a long time (~days) before a sustained, significant net air inflow is established. For the much less likely case of a double break in the vessel that allows access to both the top and bottom of the core, a chimney-like configuration could promote a higher net flow more quickly. Since the reactor cavity is below ground and to some extent sealed-off, even for a confinement (vs. “leak-tight” containment), at some point in the accident there would not be oxygen-rich air available to sustain significant graphite oxidation rates. Air availability limitation models are currently not incorporated in GRSAC. In the first case it is assumed that a single break occurs and it takes 2 days to establish the net air ingress flow.

At that time, oxidation occurs in the lower part of the core, in the bottom reflector, but the oxygen is depleted before the “air” reaches the active core area. Later in the transient, however, oxidation occurs in the lower part of the active core, since the lower reflector cooled sufficiently and no longer oxidizes. For this case, the maximum (initial) oxidation power is ~350 kW, and T _fuel ^max is about the same as in the D-LOFC case with no air ingress.

For a single vessel break with two days to establish the net air ingress flow, and assuming unlimited fresh air availability at the break, after 7 days ~5 % of the total core graphite is oxidized, and T _fuel ^max is about the same as with no air ingress. For the “chimney” case (double vessel break that allows air access to both the bottom and top of the core), and assuming a 2-meter high chimney “appears” above the vessel, air ingress flow is assumed to begin upon depressurization. The higher oxidation rate than in the previous case, after ~4 days, penetrates further up the core as the lower reflector and support structure are cooled to the point that little oxidation occurs there.

T fuel max is lower than in the reference (no air ingress) case, but the maximum vessel temperature is higher, 453°C at 168 hr. With unlimited fresh air available, after 7 days ~10 % of the core graphite is oxidized. Some mitigating actions (to limit the air supply) are necessary.

A.6.5 P-LOFC with ATWS

The early part of the transient is very similar to the P-LOFC with scram since

the negative temperature-reactivity coefficient is quite strong and reduces

the power quickly as the nuclear average temperature increases and the Xe

poison builds up. In this PBMR design, re-criticality occurs at about 28 hours,

and T _fuel ^max reaches 2127°C at 103 hr. Maximum vessel temperatures are also

higher, 711°C at 145 hr. Fuel failure after 7 days was 57%. Variations in this

accident are sensitive to fuel and moderator temperature-reactivity feedback coefficients. If after re-criticality the SCS is started (with still no scram), peak fuel temperatures would exceed limits even more due to the selective undercooling.

A.6.6 D-LOFC with ATWS

Recriticality occurs at 31 hr. In this case, T _fuel ^max is 2166°C at 137 hr, and the maximum vessel temperature (496°C at time = 168 hr) was still rising slowly after a week. Fuel failure at the end of the week was 59%.

A.7 Conclusions

The data presented in this appendix lead to some interesting and important conclusions. The PBMR is a “Major Intrinsic Safety” reactor, because it allows longer time than LWRs to improve some safety procedures. Typically to have a significant fraction release from the pebble during an accident transient few days (typically a week) are needed. The CFD model shows that in the case of double break with D-LOFC there is not a significant air ingress for 96 hours.

The analyses evidence that the most danger behaviour happens with a P- LOFC (in presence of small breaks), special in the case of ATWS. Failures percentage of CPs arrives at 57% for high temperature inside the core but this phenomenon requires about a week. That happens even if inside the core He natural circulation starts among the pebbles, because without a heat sink, the He temperature continues to rise up (specially in the central volume near the graphite column). The D-LOFC with ATWS is the worst transient with the CPs failure of 59%. Fortunately for the long time extent of these transients, we have more times than LWR to use the ECCS. For example, in the case of P-LOFC, the depressurization of the reactor leads subsequently at the beginning of the natural circulation in the rooms of containment building.

This phenomenon together with the start of safety systems guarantee the decay heat transfer and the integrity of the CPs.

Really these events have a very low frequency and the risk connected for

workers and people is quite low. Only in the case of dust explosion in a small

room (with consequently release of He with few FPs) we can have a

significant dose. But for this reason there are safety systems designed

specifically for severe accidents. The analyses suggest that the double break

with D-LOFC is not so dramatic accident as people could think. Only the

breaks in the presence of ATWS are dangerous.