Implementation of the In-Vessel Retention Strategy for Rivne NPP Units 1,2

Analytical Research Bureau for Nuclear Power Plant Safety

ARB | Kyiv, Ukraine

Rivne Nuclear Power Plant, Reactor Division

RNPP-1,2 | Varash, Ukraine

R. Lishchuk

Email: lishchuk@arb-npps.com.ua

Y. Aleksyeyev

I. Goncharov

Overview of the problem

Introduction 1

Overview of existing IVR strategies 2

Selection of preferable design for RNPP-1,2 3

Problem definition of the IVR strategy for RNPP-1,2 4

Numerical analysis and results 5

Conclusion 6

2

Introduction

One of the key safety measures for Ukrainian utility NNEGC “EnergoAtom”, which have increased significance after Fukushima accident, is “in-vessel retention” of corium in case of severe accident. Integrity of reactor vessel gives a chance to significant reduction of radiological accident consequences. Other major aspects are a reduction of hydrogen generation and lower pressure in the containment.

IVR measure has been realized on most of WWER-440 reactors in Europe, more so large amount of borated water in Bubble Condenser Tower together with sufficient elevation difference allows to deliver passive system with high reliability and performance. Analysis of IVR strategy implementation on WWER in EU countries shows that general approach was almost the same for all units: Loviisa, Paks, etc.

The main objective of investigation was defined as follows, to check the applicability of common approach to the real geometry of Rivne Units 1,2 and to recommend a preliminary design of ex-vessel cooling system.

3

Overview of existing IVR strategies

• First technical solution is implemented in Loviisa NPP. • Main idea: the shield is moved down by pneumatic

driver to ensue enough free space for sufficient cooling.

1

2

• Second technical solution was realized on several NPPs in Europe (Paks NPP could be taken as a reference Unit).

• Main idea: save the initial design of the shield and to install float passive valve, which would opened by itself after flooding of reactor cavity. In this case, the coolant will circulate in the gap between reactor vessel and existing thermal shield.

Taisto Laato. “In-vessel melt retention at Loviisa NPP: past and present activities”. IAEA Technical Meeting on Phenomenology and Technologies Relevant to In-Vessel Melt Retention and Ex-Vessel Corium Cooling. Shanghai, China, 17-21 October, 2016.

J.Elter, P.Matejovic. “Proposal of in-vessel corium retention concept for Paks NPP”. Conference paper: MASCA2 Seminar 2007. At Cadarache, France, October 2007.

4

Selection of preferable design for RNPP-1,2

A-004

CYRC-CH1

CYRC-CH2

CYRC-CH3

BOX-SG2

CYRC-CH4

10

6

7

8

9

VENT-CH

5 4 31

BOX-SG

2

View A

02 0304

0506

0708

0910 11 12

1314

1516

1718

1920

2122

23

2425

26

27

28

29

30010203

04

05

06

07

08

09

10

11

12

13

CYRC-CH2

CYRC-CH1

01

Meshfilter

U-typeisolation siphon

Inletvalve

Reactor

Critical cross-sectionsof ex-vessel coolingcircuit

Inlet with float valvein thermal shield

Ventilation pipe

-4.1 m

+1.2 m

+6.0 m

+6.55 m+6.8 m+6.5 m

А

ATHLET nodalizationExternal cooling circuitschematic solution

Meshfilter

Second case is much easy, cheaper and do not lead to additional significant exposure of staff. Respectively, it was decided, to try adopting this design for Rivne Units 1 and 2.

The cooling circuit consists of: o inlet valve located in the

corridor connecting BCT and reactor unit compartments and provides water to the ventilation ducts in concrete;

o part of the ventilation ducts isolated by U-type siphon;

o reactor cavity compartment; o float valve installed on the

biological shield below the reactor bottom;

o a gap between the reactor bottom and the shield;

o the vertical heated part in the annular gap between the reactor vessel and the reactor cavity with biological shield;

o outlet between the reactor vessel and the bearing ring.

5

Problem definition of the IVR strategy for RNPP-1,2

Operability of the proposed circuit for RNPP-1,2 depends of availability of sufficient water inventory: • the minimal free volume, which should be filled to obtain stable natural

circulation is 863 m3; • is much less than a volume of water on trays in BCT (1463 m3).

Analysis of the key aspects of ex-vessel cooling, as follows: • departure from nucleate boiling ratio (DNBR) at the reactor bottom; • sufficiency of coolant flow rate and heat sink in the circuit for reliable reactor

cooling.

6

1 Calculation of in-vessel phase of accident with MELCOR computer code and comprehensive analytical model of Rivne-1: • the “Large break LOCA 2×Dn 500 with total station blackout” is the most challenging

scenario leading to the beginning to the earliest melting and formation of corium; • time of corium formation gives us a total decay heat power, which should have been

removed by ex-vessel cooling (maximum power of the decay heat is about 9 MW).

2 Definition of heat flux from corium to the coolant through the reactor vessel during in-vessel phase of severe accident was done using AIDA module of ATHLET-CD 3.1A: • the AIDA model considers a segregated melt pool consisting of a lower oxide layer with

decay power and an upper metal melt layer; • the power of decay heat was taken from MELCOR calculations; • at the outside surface boundary conditions are set:

the HTC for nucleate boiling is 10000 W/m2/K; the coolant temperature of the main flow is 120°C.

7 Numerical analysis and results

Numerical analysis and results (continuation)

3 Influence of AIDA model input parameters on HTС value was performed using Software for Uncertainty and Sensitivity Analyses (SUSA 4.0): • 34 input parameters of AIDA model with most potential influence on calculation results

were selected, including thermal and mechanical properties of materials of corium, zirconium oxidation ratio and others;

• sample ranking and preliminary analysis of the distribution of hypotheses; • 100 input decks for AIDA calculations were made by SUSA for combinations of

parameters; • to analyze the ranking of variable model parameters, the Spearman rank correlation

coefficient is used in SUSA.

8

Numerical analysis and results (continuation)

From the results of the analysis of sensitivity and uncertainty of maximum heat flux on the outer surface of RPV it is easy to see: • maximum density of heat flux is less than 750÷800 kW/m2 is in good agreement with

results of a probabilistic study of the melt behavior for the Loviisa NPP; • DNBR resulting from comparison of the calculated heat flux and critical heat flux, is

sufficient (used the most conservative values of the critical heat flux are obtained in experimental study on SBLB test facility).

9

4

Numerical analysis and results (continuation)

0

10000

20000

30000

40000

50000

700 750 800 850 900 950 1000

Hea

t tr

ansf

er c

oef

fici

ent

(W/m

2/K

)

Time (s)

Anode-05

node-10

node-15

node-20

node-25

node-30

node-35

node-40

node-45

node-50

-100

0

100

200

300

400

700 750 800 850 900 950 1000

Mas

s fl

ow

rat

e (k

g/s

)

Time (s)

B

Assessment of sufficiency of heat transferred from outside surface by natural circulation of ex-vessel cooling was checked with ATHLET 3.1A code: • the distribution of heat flux density at the RPV is set as a boundary condition from AIDA calculation; • the influence of a gap width between RPV and thermal shield is checked by variant calculations

with design gap and a gap reduced due to the possible thermal and mechanical deformation of RPV, the results of calculations confirm the reliability of heat removal in all cases.

10

As expected the behavior of a two-phase flow in the gap has an oscillating nature that intensified by the considerable unevenness of the heat load in the heated area. The process is a quasistationary. The average value of heat transfer coefficient in area with average heat load (nodes 30÷35) is set at approximate value of 9000÷12000 W/m2/K, maximum values of HTC were obtained for maximum heat flux areas (nodes 40÷50). The average value of coolant flow rate for two-phase circulations is equals to ≈150 kg/s

5

6

Conclusions

The applicability of the Paks NPP approach of in-vessel retention to the real geometry of Rivne units 1,2 was confirmed.

Numerical analysis of ex-vessel cooling system shows the sufficient DNBR and coolant flow is enough for reliable reactor cooling.

11

These are only preliminary results. Therefore, it is necessary to perform a more detailed analysis of all aspects of IVR implementation for the introduction of an ex-vessel cooling system.

Thank you for your attention!

12