Severe accident development modeling and evaluation for CANDU

Annals of Nuclear Energy 36 (2009) 1424–1430

Contents lists available at ScienceDirect

Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene

Severe accident development modeling and evaluation for CANDU Gheorghe Negut a,*, Alexandru Catana b, Ilie Prisecaru c, Daniel Dupleac c a b c

National Agency for Radioactive Waste, 1, Campului Str., 115400 Mioveni, Romania Institute for Nuclear Research Pitesti, 1, Campului Str., Mioveni P.O. Box 78, 0300 Pitesti, Romania Politehnica University Bucharest, 313, Splaiul Independentei, Sect. 6, 060042 Bucharest, Romania

a r t i c l e

i n f o

Article history: Received 22 January 2009 Received in revised form 10 June 2009 Accepted 11 June 2009 Available online 3 July 2009

a b s t r a c t Romania as UE member got new challenges for its nuclear industry. Romania operates since 1996 a CANDU nuclear power reactor and since 2007 the second CANDU unit. In EU are operated mainly PWR reactors, so, ours have to meet UE standards. Safety analysis guidelines require to model nuclear reactors severe accidents. Starting from previous studies, a CANDU degraded core thermal hydraulic model was developed. The initiating event is a LOCA, with simultaneous loss of moderator cooling and the loss of emergency core cooling system (ECCS). This type of accident is likely to modify the reactor geometry and will lead to a severe accident development. When the coolant temperature inside a pressure tube reaches 1000 °C, a contact between pressure tube and calandria tube occurs and the decay heat is transferred to the moderator. Due to the lack of cooling, the moderator, eventually, begins to boil and is expelled, through the calandria vessel relief ducts, into the containment. Therefore the calandria tubes (fuel channels) uncover, then disintegrate and fall down to the calandria vessel bottom. All the quantity of calandria moderator is vaporized and expelled, the debris will heat up and eventually boil. The heat accumulated in the molten debris will be transferred through the calandria vessel wall to the shield tank water, which surrounds the calandria vessel. The thermal hydraulics phenomena described above are modeled, analyzed and compared with the existing data. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction

2. CANDU severe accidents analysis

After the major events as Three Miles Island and, especially, Chernobyl, accidents beyond design basis, these types of accidents called severe accidents, which lead to a core degradation and reactor geometry modification are intensely scrutinized and analyzed. Computer codes which analyze severe accidents were developed. For LWRs: US MELCOR, MAAP4, RELAP5 SCDAP, German ATHLET and the European Union code ASTEC are available. For CANDU severe accidents only MAAP4 CANDU is available. Romania, as CANDU owner, is interested to have CANDU severe accidents analysis. ASTEC UE code should have a CANDU module in order to analyze PHWR nuclear reactors, besides the LWRs European reactors. The paper presents a simple model to analyze severe accident development, given the special features of the CANDU reactor, horizontal core, fuel channels in the pressure tubes, etc. The CANDU reactor has moderator calandria vessel as ultimate heat sink during severe accident development, which acts, also, as a core debris catcher.

2.1. CANDU severe accidents

* Corresponding author. Tel.: +40 742239633; fax: +40 248291400. E-mail address: [email protected] (G. Negut). 0306-4549/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.anucene.2009.06.009

Nuclear reactor severe accidents are important in terms of consequences: radioactive releases and the public perception. CANDU reactor is provided with safety features for severe accidents mitigation: two independent shutdown systems, moderator and the shield tank cooling systems to remove the reactor decay heat, in case of unavailability of other heat sinks. In terms of severe accident consequences is needed to define plant damage categories. AECL considers four CANDU reactor fuel damage categories (Mani Mathew, 2007):    

Local fuel damage. Widespread fuel damage. In-vessel core damage. Ex-vessel core damage.

Local fuel damage can happen in case of a LOCA with ECCS available, and only a limited part of fuel channels are affected. Widespread fuel damage can take place in case of a LOCA with ECCS unavailable. In this case more fuel channels are affected

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430

and an important quantity of fission products is released in the primary heat transport system. These fuel damage categories can be considered within the design base accident criteria. In-vessel core damage or the damage of all fuel channels due to core degraded cooling takes place in case of a LOCA, with ECCS and calandria moderator cooling unavailable. This LOCA accident develops in a severe accident with significant fission releases into containment and core geometry changing. In this case, primary heat system depressurizes, and the fuel channels cooling are reduced. The fuel heats up, deforms and sags onto pressure tube bottom. The pressure tubes will heat up until it reaches 1000 °C, when they get in contact with the calandria tubes. So, the most part of decay heat is transferred to the calandria moderator. Unless cooling, the moderator begins to heat up causing the pressure inside calandria vessel to increase, as the D2O vapor pressure increases and the calandria moderator begins to swell. This leads to compression of the calandria helium cover gas in the relief ducts. The rupture disks of the pressure relief ducts burst, eventually, and the D2O moderator overflows into the containment. The moderator is initially in a sub-cooled boiling regime and thereafter starts to boil. The moderator is expelled through the relief ducts outside the calandria vessel in the containment as a mixture of liquid and vapor. The moderator void fraction inside the calandria increases and the upper rows of the pressure tubes start to uncover. After calandria tube rows (including pressure tubes and fuel bundles) are uncovered, they overheat sag and calandria tubes may, finally, disintegrate and melt. Disintegrated and/or melted components of fuel channels fall into the remaining moderator at the calandria bottom as in Fig. 1 (Mani Mathew, 2007). Thus, the calandria vessel would act as an inherent ‘‘corecatcher”. A rapid vapor generation takes place due to high temperature debris which is quenched rapidly. This causes pressure surges and a great amount of two phase fluid discharges through relief ducts to the containment. The fission products resulted from degraded and partially melted fuel will escape from the calandria to the containment. Deuterium gas from Zircaloy–D2O reaction will escape into the containment atmosphere. Steam and deuterium released in the containment atmosphere will increase containment pressure. In case of D2 combustion the pressure increase might be more pronounced. As more fuel channels will disintegrate and accumulation of the debris will fall down at the calandria bottom. Due to the vaporizing caused by the fuel channels disintegration, eventually, the entire moderator will be lost from the calandria vessel. The debris accumulated at the calandria bottom is formed of the Zircaloy pressure tubes, calandria tubes, Zircaloy fuel clad and fuel

1425

pellets. Due to the concentrated residual heat the debris, that will heat up, melt and, eventually, reach excessively high temperatures as 4000 °C level, when the melted debris starts to boil. The molten pool is supported by the calandria wall which is cooled externally by the water from the shield tank. The heat from the molten debris pool will be transferred to the water in the shield as in Fig. 2 (Mani Mathew, 2007). Ex-vessel fuel core damage is in case of a LOCA with ECCS unavailable and calandria and shield tank loss of cooling. The severe accident scenario is similar with above scenario, completed with the calandria wall melt and break, shield tank water boil off, due to the molten debris bed escaped from calandria and, eventually, interaction between debris and concrete vault will take place. 2.2. CANDU severe accident calandria moderator thermal hydraulic model The thermal hydraulic model of the calandria boil off progression is based on a simple model that assess the mass and energy balance within the calandria vessel, (Rogers, 1962, 1979, 1984; Rogers and Currie, 1982). The calandria vessel is shared in 22 slices corresponding to the 22 rows of the CANDU 600 reactor, as in Fig. 3. The transient heat balance, calculated for each moderator slice, taking into account the heat source from the fuel channel, the heat transferred to the calandria walls, heat stored in the calandria walls and heat transferred to the shield tank water is done. Heat transfer coefficients from the moderator to the calandria walls and from the calandria wall to the shield tank water, using the correlations for turbulent natural convection on vertical surfaces, are predicted. For our case the surfaces are not strictly vertical, as can be seen from Fig. 3, but the natural convection circulation patterns near the walls are similar to those for vertical walls. In the beginning, decay heat transferred from the fuel channels determines calandria subcooled D2O to boil and pressurize up to the rupture disk burst pressure threshold. After that, the calandria pressure decreases to containment atmospheric pressure. Calandria moderator starts saturated boil that determines the fuel channels uncover. Calandria/pressure tubes sagging and, eventually, disintegration and fall down on the calandria bottom are in progress. During this process calandria moderator inventory is expelled through the relief ducts into the containment. Finally, all moderator inventories are expelled from the calandria vessel to the containment. The theory of this thermal hydraulic process was presented previously (Negut, 2006), based on the Rogers paper (1984). Al these phenomena are in three successive computer codes routine which form SEVMODE program (Negut, 2006). 2.3. CANDU severe accident debris bed thermal hydraulic model

Fig. 1. CANDU core during a severe accident.

After the complete disintegration of all the tube rows ends up, there is no more D2O moderator left in the calandria to cool down the debris, which is accumulated at the calandria bottom. The debris temperature begins to rise due to the heat generated by the remaining fission products. Debris temperature can reach the melting temperature of ZrO2 and UO2. Debris will then form a molten pool with the materials fallen down at the calandria bottom. This pool may, eventually, reach the boiling point. The final debris configuration modeled as subdivided in ten equal layers and their temperatures are predicted. At the same time it is examined the calandria wall melts down during the process of heat transfer to the shield tank water. It is assumed that the shield tank water cooling is still in operation (Rogers, 1984).

1426

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430

Fig. 2. CANDU core debris in the shield tank.

Fig. 3. Calandria vessel control volume (VMOD(I) – slice moderator volume, AI(I) – calandria slice wall inner surface, AM(I) – mean slice surface, AO(I) – outer slice wall surface, AF(I) – slice flow cross section, DHY(I) – slice hydraulic diameter).

In the situation of CANDU reactor the Zircaloy mass is greater due to the presence of pressure and calandria tube, representing 40% of fuel mass, compared with Zircaloy sheaths mass which is approximately 19% of fuel mass, allowing proportionally more ZrO2 production. ZrO2 is produced due to the exothermic reaction between Zircaloy mainly from the horizontal tubes and steam, in the process of fuel channels disintegration (Rogers, 1984). The disintegration of fuel channels under conditions of high temperatures and mechanical stress conditions, once uncover of the calandria tubes is a complex problem. The prediction of the physical and geometric characteristics of the debris is difficult. The constituents of debris are the fragments of UO2 pellets, pieces of fragmented oxidized Zircaloy with different forms and re-solidified particles of previously molten Zircaloy. In order to model this complex composition debris porous bed with a given porosity and average pore size is assumed. The solid portion of the debris is composed of UO2 and ZrO2 in proportion to the quantities of UO2 and Zircaloy originally found in the reactor core, allowing for the Zircaloy oxidation. The pores of the bed are assumed to be filled with heavy water vapor at atmospheric pressure. Lacking any other information, the bed was assumed to be uniform throughout, i.e., it consists of a uniform distribution of averaged-size pores in a uniform mixture of UO2 and ZrO2 (Rogers, 1984). The debris bed situation is presented in AECL study (Rogers, 1995; Meneley, 1996). The bed is assumed uniform throughout. The mechanism of heat transfer considered the conduction through the solid material and the conduction and radiation through the steam-filled pores. The depth of the one-dimensional debris bed is assumed to be equal to the distance from the surface of actual debris bed to the calandria wall along the vertical centreline of the calandria vessel and equal to the maximum thickness of actual bed. The bed depth is calculated from the calandria geome-

try and assumed fraction of fuel channels falling into the bottom of the calandria, allowing for the volume occupied by the vapor-filled pores. The heat transfer rates and resulting thermal behavior are not very sensitive to the parameters affecting heat flow rates in the solid fractions. When the solid core debris, due to the heat accumulated, heats up, the temperature increases and eventually the debris melts. The effects of the completely molten debris pool formed in the calandria vessel on the calandria wall, as depicted in Fig. 2 (Mani Mathew, 2007), are to be analyzed. As an assumption, all the debris at the calandria vessel bottom will eventually be molten. Due to heat transfer through the calandria wall to the shield tank water, molten pool forms a lower crust at the calandria bottom. The stainless steel wall of the calandria vessel may be partially molten. The interface temperature between molten pool and the crust is the pool freezing temperature and the interface temperature between solid and molten wall layer is the wall melting temperature which is lower than the former. These temperatures are assumed to be constant throughout the present analysis. In the upper part of the molten pool an upper crust is formed due to heat transfer from the molten pool to the vessel atmosphere above. The atmosphere is composed of a mixture of D2O and UO2 vapors. The heat transfer mechanisms from the crust are natural convection and radiation. Any molten pool would almost certainly be overlaid by a frozen crust of core debris. Nevertheless, the stability of such a crust is quite uncertain, so that some core material vapor would be present above the debris in the calandria. Vaporized material would be produced above the surface of the pool as droplets that form suspended aerosols. The larger droplets fall back into the boiling pool and some are condensing on the exposed upper calandria wall in the gas space. All these phenomena are modeled by the DEBRIS subroutine (Negut, 2006).

1427

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430

2.4. CANDU severe accident debris bed thermal hydraulic model RELAP-SCDAP COUPLE module

90 80 CANDU 600 decay heat El Wakil

3. Result and discussion 3.1. Fuel channels disintegration Severe accident progression evaluation, with the model presented above, needs as input the reactor decay heat power. Rogers (1984) used as input the curve presented in Fig. 5.

70

decay heat (MW)

For comparison reasons, a RELAP-SCDAP debris bed model for CANDU 600 transient was developed (Dupleac, 2007) using COUPLE module (Code Manual Volume II: Damage Progression Model, 1997). The COUPLE module is a part of SCDAP model developed in the RELAP-SCDAP code. The COUPLE code is a two-dimensional, finite element, steady-state and transient heat conduction code. The code was developed to solve both plane and axisymmetric type heat transfer problems with anisotropic thermal properties, subject to boundary conditions of the first kind (Dirichlet), second kind (Neumen), and third kind (combination of the first and second), and/or nonlinear boundary conditions such as radiation. COUPLE model behavior of the molten debris inside the bottom vessel of PWR. Due to the design similarities of the design the molten debris from the calandria bottom was modeled using COUPLE. This model represents behavior of the calandria vessel bottom where the debris bed is formed after all CANDU channels disintegration. Three volumes of calandria vessel and one volume for the shield tank were used to model the core catcher and the cooling means. The debris and the calandria wall were modeled using COUPLE module of RELAP5 SCDAP. The debris is shared in the finite elements mesh in order to perform the debris heat up and heat transfer to the shield tank water. The RELAP5 SCDAP COUPLE CANDU 600 calandria model is presented in Fig. 4. The COUPLE input data are the debris inventories and residual heat resulted from the CANDU fuel channels disintegration model.

CANDU severe accident decay heat

60

Bruce decay heat_Rogers

50 40 30 20 10 0 0

50

100

150

200

250

300

350

time (min) Fig. 5. CANDU severe accident decay power.

With input power Bruce reactor of 850 MWe power, for a severe accident progression, the calandria boils off in 45 min. In our analysis the decay heat power curve (El-Wakil, 1971), also, presented in Fig. 5 was used. For the start of the transient Rogers (1984), considers that up to the generalized saturated boiling, only a part of reactor decay heat is transferred to the calandria moderator. Thereafter, the reactor decay heat is fully transferred to the moderator. The reason for this is that in the fuel channels still are primary coolant that transfer decay heat in the reactor primary circuit. CANDU 600 analysis modified decay heat curve, shown in Fig. 5, was used. The LOCA assessment establishes that at 1000 °C the pressure tube and calandria tube get in full contact. This moment takes place after 6 min from the start development. In our prediction the rupture disks burst and contact between the calandria vessel and containment through the relief ducts takes place at 15 min compared with 9 min in the Rogers paper (1984). This is mainly due to lower energy transferred to the calandria moderator. This entire period time calandria moderator is in subcooled boiling mode.

Fig. 4. RELAP5 SCDAP couple model used for CANDU 600 debris heat up resulted from a severe accident.

1428

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430 0.3

87

0.25

86 85.5

0.2

85

moderator temperature(C)

84.5

moderator pressure(MPa)

84

rupture disk pressure threshold(MPa) saturation pressure (MPa)

83.5

0.15

0.1

83

0.05

moderator pressure (MPa)

moderator temperature (C)

86.5

82.5 0

82 6

7

8

9

10

11

12

13

14

15

16

time (min)

and uncovering process continues up to complete moderator expulsion from calandria vessel. The rows uncovering, rows disintegration and correlated with moderator level during boil off process in Fig. 8 are presented. When fuel channels rows starts to uncover calandria tubes temperature rise sharply to 1500 °C. Fuel channel starts to sag reach the lower ones and, eventually, disintegrate and fall dawn on calandria bottom as in Fig. 1. The debris is formed by disintegrated fuel channels material which falls down on the calandria bottom. The progression of material quantities accumulated as debris on the calandria bottom and debris volume are presented in Fig. 9. Our prediction is that the calandria moderator boils off finishes at 278 min, approximately 4 1/2 h. This time interval is longer that Rogers value (Rogers, 1984) for Bruce reactor of 45 min. Rogers (1995), Meneley et al. (1996) predicted for a CANDU6 severe acci-

Fig. 6. Sub-cooled calandria moderator main parameters evolution during beginning of a CANDU severe accident. 18

1200 1000 800

14

10 60 8 6

40

3

12

80

volume (m )

1400

16

tube mass*103(kg) fuel mass*103(kg) debris vol(m3)

100

1600

mass*103 (kg)

expelled moderator flowrate (kg/sec)

120

4

600

20 2

400

0 300

0 0

200

50

150

100

200

250

time (min)

0 0

50

100

150

200

250

300

time (min)

Fig. 9. Disintegrated Zircaloy and fuel mass rates on the calandria bottom during CANDU severe accident progression.

Fig. 7. Expelled moderator flow rates during boil off in a CANDU severe accident.

debris temperature (C)

3000

From this moment important quantities of moderator are expelled through the relief ducts to the containment. The moderator pressure and temperature evolution are presented in Fig. 6. At 24 min the transition from moderator sub-cooled boiling to the full saturated boiling mode starts. At 38 min all calandria inventory is in full saturated boiling. Rogers’s paper shows that saturated boiling starts at 16 min and at 17 min all calandria moderator is in full saturated boiling. Important vapor quantities are produced in calandria. Moderator expelled flow rates are presented in Fig. 7. The highest flow rate is reached before moderator generalized full saturated boiling is reached. The first three rows uncover at 43 min

2500 2000

t_uppercrust_DEBRIS t_midpool_DEBRIS t_lowercrust_DEBRIS t_midpool_Meneley [8] t_uppercrust_Meneley[8] t_lowercrust_Meneley [8] t_midpool_COUPLE

1500 1000 500 0 270

290

310

330

350

370

390

t (min) 25

6 5

15 4 10

3 2

5 1 0

0 6.1 43 48 53 63 83 98 118 143 148 168 183 209 218 243 244 278

time (min) Fig. 8. Uncovered fuel channels rows and calandria level during severe accident (0 level = maximum D2O level in calandria vessel, 7.6 m = minimum D2O level in calandria).

calandria wall temperature (C)

fuel channels rows

20

Fig. 10. Heat-up debris core temperatures in a CANDU severe accident. 7

moderator level (m)

uncovered disintegrated level

400 lower wall inner temperature_DEBRIS

350 300

lower wall outer temperature_DEBRIS twall_inner_Meneley [8] lower outer wall temp_Meneley[8]

250 200 150 100 50 270

290

310

330

350

370

time (min) Fig. 11. Calandria wall temperatures during debris bed heat-up.

390

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430

1429

Fig. 12. RELAP5 SCDAP COUPLE debris temperature distribution for a CANDU 600 severe accident.

dent calandria moderator complete boil off a time of 300 min, 5 h. The present time prediction of 4 1/2 h closed with 5 h values predicted by the other studies. Another AECL study (Meneley, 1996) gives complete boil off prediction for 2–2 1/2 h, which is shorter than predicted by this paper. For a better comparison same time span was used.

100 °C, due to shield tank water cooling. So, the color1 is blue. Same situation is with the upper crust and debris near the wall. Inside the molten debris the temperature is much higher, so, the color is red.

3.2. Debris behavior

Finally, we will summarize the important milestones of our CANDU6 severe accident evaluation. The pressure tubes reach 1000 °C and contact the calandria tubes. Therefore, a significant amount of decay heat is transferred to the calandria moderator. The initial heat transfer mode is sub-cooled boiling. After 6 min the calandria tube and the pressure are in contacts and an important quantity of decay heat is transferred to the calandria moderator, the moderator starts sub-cooled boiling. The pressure inside the calandria increases and causes the rupture of pressure relief ducts disks break at 15 min. After 24 min transition from sub-cooled boiling to full saturated boiling starts and significant quantities of calandria moderator is expelled out into the containment through the relief ducts. Full saturated boiling in calandria moderator starts after 38 min. The calandria moderator boils off uncovering fuel cannels. At 43 min the first three rows of the pressure tubes is uncovered. The uncovered pressure tubes starts to heat up, starts pressure tubes sagging and eventually disintegrate and fall down onto calandria bottom after 53 min. After 278 min all the remaining rows disintegrate and fall down on the calandria bottom. The calandria bottom as debris receiver and a retention volume for the fallen material can be regarded as a core catcher. Entire moderator is now expelled out into the containment. At 278 min the debris begins to heat up. The middle of the molten pool reaches over 3300 °C and boils off. Only a part of the calandria wall melts, due to the cooling of the shield tank, and the molten debris pool remains confined inside the calandria vessel. The proposed model of a CANDU severe accident seems to give satisfactory results, in comparison with AECL estimations. The data obtained during our evaluation is in good agreement with results from other severe accident assessments for the CANDU 600 (Meneley, 1996; El-Wakil, 1971).

The solid debris bed begins to heat up, after the moderator is expelled completely from the calandria. Predicted upper debris crust temperature in contact with calandria atmosphere of moderator and fuel and Zircaloy vapors, middle debris pool temperature, and lower debris crust temperature, in contact with calandria wall bottom are compared in Fig. 10 with temperatures from Meneley (1996). Middle debris pool temperature is lower than predicted by other studies (Rogers, 1995; Meneley, 1996), but has the same trend. It is possible that the decay heat and Zircaloy exothermal contribution used in that studies to be greater than in our model. Middle debris pool temperature is lower than predicted by other studies (Rogers, 1995; Meneley, 1996) but has the same trend. It is possible that the decay heat and Zircaloy exothermal contribution used in that studies to be greater than in our model. The debris upper crust temperature is greater than in our prediction. RELAP5 SCDAP COUPLE middle debris temperature, as a lumped value, is quite similar with that obtained with DEBRIS routine as it is observed from Fig. 10. Anyway the debris lower crust temperature is quite similar, given the same shield tank temperatures. The calandria wall temperatures during debris heat-up period are presented in Fig. 11. These temperatures obtained by DEBRIS subroutine, Meneley (1996) results were compared with results from RELAP5 SCDAP COUPLE model (Dupleac, 2007), lumped wall temperature. Comparable results are obtained, also, for the calandria wall temperatures as it is observed from Fig. 11. The COUPLE model input data are the fuel and Zircaloy masses resulted from our fuel channels disintegration model used in the above assessment. COUPLE decay heat input curve was same as in the disintegration model. Also, COUPLE, using its finite elements mesh, gives all the temperatures during transient. The COUPLE temperature distribution in the debris and the calandria wall temperatures are presented in Fig. 12. It can be observed that temperatures of the calandria wall are lower around

4. Conclusions

1 For interpretation of color in Fig. 12, the reader is referred to the web version of this article.

1430

G. Negut et al. / Annals of Nuclear Energy 36 (2009) 1424–1430

The present CANDU severe accident model is still in development and we hope consider all the relevant severe accident issues associated with the fuel channels: sagging and disintegration mechanism, debris molten pool interaction with the basement concrete, hydrogen production and explosion, etc. Some of these issues are discussed in more recent papers dedicated to the problem (El-Wakil, 1971). These models presented are a contribution for CANDU severe accidents benchmark. CANDU severe accidents is still an issue to debate, given lack of experimental data on CANDU fuel channels behavior, their heat up and sagging during calandria moderator boil off and the calandria debris interaction. This analysis is a step to develop an alternate model to the MAAP4. The model developed is available and can be integrated in the severe accidents code such as ASTEC code, potentially an EU standard code for severe accident analysis. The reason is that such code developed for LWR should be used, also, for CANDU type reactors. Within SARNET network and SARNET code as an open source cooperation group there are such preoccupations to develop PHWR model integrated in ASTEC modular code that should be available for severe accident analysis. In the future we hope that the model proposed should be integrated in the ASTTEC. This kind of analysis can be an important contribution to the European reactor safety analysis.

References Daniel Dupleac, 2007. RELA-SCDAP application for PHWR severe accident assessment. IAEA Meeting on Preparatory Review of Accident Management, Bucharest, Romania, 12–16 February 2007. El-Wakil, M.M., 1971. Nuclear Heat Transport. Scranton International. Gheorghe Negut, Alexandru Catana, Ilie Prisecaru, 2006. A CANDU Severe accident analysis. In: Proceedings of the 2006 International Congress on Advances in Nuclear Power Plants (ICAPP ‘06), Reno, Nevada, June 4–8, 2006, ANS Order #: 700324. ISBN: 0-89448-698-5. Mani Mathew, P., 2007. Severe accident phenomena. IAEA Meeting on Preparatory Review of Accident Management, Bucharest, Romania, 12–16 February 2007. Meneley, D.A., Blahnik, C., Rogers, J.T., Snell, V.G., Nijhawan, S., 1996. Coolability of severely degrade CANDU cores, AECL-11110, January 1996. Rogers, J.T., 1962. Heat Transfer Course, 1961–62, Part1, Conduction. Canadian General Electric Co. Ltd.. Rogers, J.T., 1979. CANDU moderator provides the ultimate heat sink in a LOCA. Nuclear Engineering International 24 (280), 38. Rogers, J.T., 1984. Thermal and hydraulic behavior of CANDU cores, under severe accident conditions. AECB, INFO-1036-3, June 1984. Rogers, J.T., 1995. Heat and Mass Transfer in Severe Nuclear Reactors Accidents. Begell House. Rogers, J.T., Currie, T.C., 1982. Analysis of transient dry-patch behavior on CANDU reactor calandria tubes in a LOCA with late stagnation an impaired ECI. In: Proceedings of International Meeting on Thermal Nuclear Research Safety, Chicago, Ill, September 1982. SCDAP/RELAP5/MOD3.2 Code Manual, 1997. Damage Progression Model, Theory, NUREG/CR-6150, INEL-96/0422, Rev. 1, vol. II, October 1997, Idaho National Engineering and Environmental Laboratory, Lockheed, Martin Idaho Technologies, Idaho Falls, Idaho.