SIMULATION AND CALCULATION THERMAL LOAD ON VVER1000

LOWER HEAD VESSEL UNDER SEVERE ACCIDENT

Doan Manh Long
Nuclear Training Center
Email: longdoanmanh28@gmail.com

Abstract: Incase of severe accident, reactor core structures melt and relocate to
lower plenum, possibly establish a molten pool in lower plenum and threaten to the
integrity of lower head vessel unless there is any sufficient cooling measures. The
formation and configuration of molten pool in lower plenum will determine mode of
thermal load on lower head wall. Research and calculation on accident progression have a
significant role in studying the In-Vessel Retention (IVR) strategy.
In this paper, the MELCOR 1.8.6 code will be used to study in-vessel severe
accident progression and evaluate thermal load on lower head vessel of VVER1000/V320
in case of simultaneous happening Large Break Loss of Coolant Accident (LBLOCA)
and Station Blackout (SBO) accompanied with implementing IVR strategy. The scenario
would be divided into two sub-cases with additional goal to evaluate the role of passive
core cooling system (four hydro-accumulators) in in-vessel accident progression: 1)
LBLOCA plus SBO with the failure of passive core cooling systems; 2) LBLOCA plus
SBO with the availability of four hydro-accumulators.
The results showed that four accumulators had significant impact to accident
progression in terms of oxidation, relocation of core debris to lower plenum and mass of
components of molten pool which decided the thermal load on lower head vessel. The
results showed that heat fluxes on external surface of lower head vessel estimated in two
sub-cases are 0.59 MW/m2 and 0.62 MW/m2, respectively. Both the two estimated heat
fluxes were much smaller than that obtained in bench mark calculation for
VVER1000/V320 [7]. The results also showed that there was a failure of lower head
reactor vessel in Case 1 and Case 2, happened at 15192 seconds and 14000 seconds,
respectively.
Key words: VVER1000, IVR, MELCOR1.8.6, severe accident.

1. Introduction
In-Vessel Retention through External Reactor Vessel Cooling (IVR/ERVC) was proposed and
studied by T.G. Theofanous and his colleagues [1] and was adapted as severe accident management
strategy for low power reactor such as VVER440 [2] and AP600 [3, 4] have proved the feasibility of
IVR/ERVC strategy, and these study showed that there was a high safety margin of IVR/ERVC for
low power reactor.
Thank for successful adoption for low power reactor of IVR, the strategy has been preliminarily
studied for higher power reactor such as AP1000 [5], VVER1000 [6, 7]. The studies indicated the
potential adaption of IVR/ERVC for high power reactor, but there were many uncertainties and the
results strongly depended on specific accident scenario.
The major challenges of IVR study for high power reactor are to evaluate the thermal load of hot
molten pool to lower head vessel and increase the critical heat flux of external cooling ambient. Beside
the experiments has been done, the computer codes have been also being used to evaluate the thermal
load. One of the solutions to reduce the uncertainties is the combination between severe accident
analysis codes, likely SCDAP/RELAP5, MELCOR, with mechanic-thermal interaction analysis codes
likely CFD (FLUENT), ASTEC, PECM, such as the combination between MELCOR code with
PECM which has been performing at Institute for Nuclear and Energy Technologies [8].
The work done in the paper is the first step in the long term IVR/ERVC study with efforts to
combine MELCOR code with PECM (Phase-changed Effective Convectivity Model). The MELCOR
1.8.6 code will be used to evaluate the thermal load inserting to VVER1000/V320 lower head vessel in
case of severe accident. The selected scenario is the Large Break Loss of Coolant Accident (due to
1
end-double rupture of cold leg: 2x850mm) simultaneously happening with complete Station Blackout.
There are two sub-cases will be studied: first case (Case 1) accompanied with the failure of passive
safety injection from four hydro-accumulators; the second (Case 2) with availability of passive safety
injection. This is aim to see how important the hydro-accumulators are in term of in-vessel severe
accident progression.

2. MELCOR 1.8.6 input model for VVER1000/V320

MELCOR code [9], a fully integrated system computer code developed by Sandia National
Laboratories, can simulate a broad spectrum of severe accident phenomena in light water nuclear
power plants such as thermal-hydraulic response in the reactor coolant system, reactor cavity,
containment and confinement buildings; core heat-up, degradation and relocation; core-concrete
interaction; hydrogen production, transport, and combustion; fission product release and transport
behavior. The version MELCOR 1.8.6 is a recent version of MELCOR code. It has some new features
to improve the capability of MELCOR code for simulating and calculating severe accident, especially
are the improvement in molten pool model in lower head and new nodalization model for lower vessel
which improve the capability of MELCOR for calculating heat transfer between lower head vessel
with internal or external ambient. In order to conduct simulation and calculation in MELCOR code, a
nuclear power plant was divided and nodalized as control volumes and heat structures.

Figure 2: Nodalization scheme for volumes Figure 3: Nodalization scheme for structures in
in reactor vessel reactor vessel

Figure 4: Nodalization scheme for Figure 5: Input model of cavity of

lower head vessel VVER1000/V320 in MELCOR
2
 The VVER1000/V320 is a kind of pressurized water reactor with 3000 MW thermal power and
1000 MW electric power. It has four primary loops, each loop consists of hot leg and cold leg, main
coolant pump and a horizontal steam generator. The reactor has 163 fuel assemblies and 121 control
rod assemblies.
The volume in reactor vessel was modeled as 5 control volumes (Fig.2), such as lower plenum,
downcomer, reactor core, upper plenum, upper head. The cold and hot legs, four accumulators were
also modeled as control volumes (Fig.2). All control volumes were connected by flow paths. The
structure in reactor vessel was divided into 6 concentric rings and 12 axial levels (Fig.3). Among 6
rings, the first five rings are for reactor core structures and the last ring is for downcomer area
(MELCOR 1.8.6 can model downcomer area as a core ring). The lower head vessel was divided into 6
layers and 9 segments (Fig.4).
In order to calculate the thermal heat flux on outside surface of lower head vessel and
implemented IVR method for VVER1000/V320 by MELCOR, we changed and modeled geometry of
VVER1000/V320 cavity as Figure 5.

3. Results and discussion

Assuming that the LBLOCA and SBO simultaneously happened at zero second, due to total loss
of power, all active core cooling system were available. By using MELCOR, the paper studied two
sub-cases, such as: with and without passive core cooling system (four hydro-accumulators) in order
to evaluate the effect of four accumulators for IVR strategy. In order to evaluate the heat flux on the
external surface lower head vessel, IVR strategy would be implemented as following: water from
water resource was injected to flood cavity and submerge the reactor vessel when temperature of vapor
in reactor core reached 923.25oK (650oC). The water resource to flood cavity was assumed to be
unlimitted, and water level in cavity was stably sustained at the height of cold leg as required to
implement IVR.
a) Case1: The four accumulators were unavailable
The severe accident progression is listed in the below table 1.
Table 1: The severe accident progression of Case 1

Events Time (second)

Start of the accident 0.0
Reactor core totally loss of water 18.0
The oxidation begins 170.0
Start to flood cavity 528.0
Fuel rods start to melt and degrade 1300.0
Appearance of core debris in lower plenum 1500.0
Whole fuel debris relocated to lower plenum 6000.0
Lower plenum totally loss of water 6800.0
Failure of lower head due to creep-rupture 14000.0

When accident happened, the reactor core quickly lost all water and became to uncover at 18.0s
(Fig.6).The oxidation started at 170s and total mass of generated hydrogen was 80 kg (Fig.7).
Temperature of steam in reactor core existed 650oC at 528s which initiated cavity flooding. Water was
injected to cavity and kept at required level (Fig.8).
At 1500.0s, core debris appeared in lower plenum area. Whole water in lower plenum completely
evaporated at 6800 seconds (Fig9). Whole core structure degraded and relocated to lower plenum at
8000.0 seconds (Fig 10). The results showed that the maximum heat flux of outer face of lower head
3
vessel reached 0.59 MW/m2 at 5th segment (Fig 11), the height was around 0.4185 to 0.5105 meter.
According to result, the lower head vessel was predicted to fail after 14000.0 second due to creep-
rupture according to the creep-rupture failure rule in MELCO code.

Figure 6: Water level in reactor core Figure 7: Total mass of generated hydrogen

Figure 8: Water level in cavity and annulus Figure 9: Water level in lower plenum
channel

Figure 10: Mass of debris in lower plenum Figure 11: Heat flux on external surface of
lower head vessel
b) Case 2: The four accumulators were available
In this case, the four accumulators were available. They were initiated when the pressure in
reactor vessel below 5.9 MPa. The sequence of accident of Case 2 and Case 1 were listed as in the
4
table 2 with the aim of comparison. In the reference [6], a scenario of LBLOCA accompanied with
SBO as the same as Case 2 but with a smaller diameter of break (200mm) was studied by MELCOR
1.8.6, the results in the Reference were listed in table 2 as well, there was big difference between the
time of events in [6] and Case 2, this can be explained that was mainly due to the smaller diameter of
break in [6].
When the accident happened, reactor core quickly lost of water as in Case 1. But with the
availability of four accumulators, the reactor core was recovered and sustained about 19s, before water
level started to decrease again (Fig.12). The oxidation happened slightly later than Case 1 but more
violently, evidentially the total mass of generated hydrogen in core (Fig. 13) was 230 kg much larger
than 80 kg. This led the degradation of fuel rods in Case 2 slightly earlier than Case 1. The cavity
flooding (Fig.14) was initiated at 727.0s which was 199s later than Case 1.
However, the water injection contributed to sustain significant amount of water in lower plenum
(Fig.15) much longer than in Case 1, therefore delaying the failure of core lower support plate and the
time of whole reactor core debris relocating to lower head in Case 2 is 9000.0s (Fig.16), was later than
Case 1.
Table 2: The major events of accident of Case 2

Events Time (second)

Case 2 Case 1 Reference [6]
Start of the accident 0.0 0.0 0.0
Start water injection from ACCs 15.0 - 153.41
Reactor core totally loss of water 18.0 18.0 -
Start to recover reactor core 33.0 - -
Fully recover reactor core 38.0 - -
All ACCs run out of water 54.0 - 291.98
Reactor core starts to unrecover again 57.0 - -
Reactor core totally loss of water again 65.0 - -
The oxidation begins 180.0 170.0 -
Start to flood cavity 727.0 528.0 2168.19
Fuel rods start to melt and degrade 1230.0 1300.0 -
Appearance of core debris in lower plenum 2250.0 1500.0 -
Whole fuel debris relocated to lower plenum 8000.0 6000.0 -
Lower plenum totally loss of water 8500.0 6800.0 -
Failure of lower head due to creep-rupture 15192.0 14000.0 19265.30

5
Figure 12: Water level in reactor core Figure 13: Mass of H2 generated in reactor core
 Figure 14: Water level in cavity and annulus Figure 15: Water level in lower plenum
channel

Figure 16: Mass of debris in lower plenum Figure 17: Heat flux on outer face of lower head
vessel
The maximum heat flux on external surface of lower head was predicted to be 0.62 MW/m2 (Fig.
17) at the segment 6th which was higher than that of Case 1. The result in Case 2 showed that the
failure of lower head vessel due to creep-rupture happened at 15192.0s which was later than in Case 1
about 1192.0s (20 minutes).
The results obtained from MELCOR 1.8.6 in two sub-cases showed that:
 In the LBLOCA accident simultaneously happening with SBO accompanied with
implanting IVR strategy, the lower head vessel was still failed due to creep-rupture
according to failure standard in MELCOR code.
 In Case 2, with the availability of four accumulators could not prevent the failure of
lower head vessel. Four accumulators only delayed the relocation of core debris to lower
plenum and the failure of lower head compared to Case 1.
 An important problem is due to the water injection from four accumulators, the oxidation
in Case 2 happened more violently than in Case 1. This not only introduced additional
heat to heat up core structures, but also changed the ingredient of components (table 3)
and configuration of molten pool in lower plenum which would impact to thermal load
from molten pool to lower head vessel.

6
  The peak of heat flux on external surface of lower head vessel was predicted in Case 2
slightly higher than that of Case 1, respectively were 0.62 MW/m2 and 0.59 MW/m2.
Compared to low margin of heat flux on external surface of lower head, 1 MW/m2, which
was averaged from various different calculations for bench mark calculation with the
same scenario [7] by stand-alone codes (MELCOR, MAAP, SOCRAT…) and combined
codes (ASTEC combined MELCOR, ALATS combined MELCOR…), both heat fluxes
predicted in the study are much lower than. This might be due to the mass of UO2 debris
predicted in the study (80600 kg) was smaller than in the bench mark calculation.
 By MELCOR stand-alone calculation as in Reference [6] and in Case 2 with the same
scenario, LBLOCA plus SBO, the results showed that there was a failure of lower head
vessel due to creep-rupture. The results were preliminary, and need to be improved by
integral calculations in the future.

4. Conclusion
The paper used MELCOR code to study VVER1000/V320 in-vessel severe accident and estimate
the external heat flux of lower head vessel in case of accident, LBLOCA and SBO accompanied with
implementing IVR strategy. With the aim of study the impact of four hydro-accumulators in severe
accident progression, the scenario was divided into sub-cases with and without the availability of four
accumulators. Compared to Case 1, with the presence of four hydro-accumulators in Case 2, the
results showed that the accident progression was delayed, but water injection from four hydro-
accumulators boosted oxidation in reactor core, it led to change in mass of ingredients of molten pool
which would change the configuration of molten pool and the thermal load mode from molten pool to
lower vessel. That explained the difference in external heat flux of lower head vessel in both sub-
cases.
Compared to result from benchmark calculation with the same scenario, LBLOCA plus SBO, the
heat flux on external surface of lower head vessel predicted in the study was much lower than, the
main reason might be the predicted mass of UO2 debris relocating lower plenum in the study was
smaller than in benchmark calculation [7]. And by MELCOR stand-alone calculation with the scenario
LBLOCA plus SBO, with the availability of four accumulators and implementation IVR strategy,
there was a failure of lower head reactor vessel due to creep-rupture in Case 2 and Reference [6].
The work done in the paper is the first step of long term IVR study. The result is the preliminary
result which contains a lot of uncertainties relating to initial mass of fuel elements, fuel cladding and
steel. The results will be improved and updated following up analyses of more precise simulation by
MELCOR and integral calculation using CFD tool (FLUENT) combined with PECM for more precise
estimation of thermal load to lower head vessel in the future.

REFERENCES
[1] T.G.Theofanous et al. , “In-Vessel Coolability and Retention of Core Melt,” DOE/ID-10406
Volume 1, October 1996.
[2] O. Kyma ̈la ̈inen, H. Tuomisto, T.G.Theofanous, “In-vessel retention of corium at the Loviisa
plant,” Nuclear Engineering and Design 167(1997) 109-130.
[3] T.G.Theofanous et al. “Invessel coolability and retention of a core melt,” Nuclear Engineering and
Design 169(1997) 1-48.
[4] T.G.Theofanous and S. Angelini, “Natural convection for in-vessel retention at prototypic
Rayleigh numbers”, Nuclear Engineering and Design 200 (2000)1-9.
[5] T.N.Dinh, J.P.Tu, T.Salmassi and T.G.Theofanous, “Limits of coolability in the AP1000-ralated
ULPU-2400 configuration V facility,” the 10th International Topical Meeting on Nuclear Thermal
Hydraulics (NURETH-10), Seul, Korea, 2003.
[6] J.Duspiva “Analytical simulation of in-vessel retention Strategy for VVER1000/V320 reactor
using MELCOR code”, NURETH-16, Chicago, IL, August 30-September 4, 2015.
[7] JRC Technical Report “ In-Vessel Melt Retention Analysis of a VVER1000 NPP”, European
Comission, 2016.
[8] Plilipp Dietrich, “Coupling the PECM with MELCOR,” NURETH-16, Chicago, IL, August 30-
September 4, 2015.
7
[9] Sandia National Laboratories, MELCOR computer code manuals, Ver.1.8.6, Rev. 3 vol.
NUREG/CG 6119, 2005.

MÔ PHỎNG TÍNH TOÁN TÁC ĐỘNG NHIỆT VÀO VỎ ĐÁY THÙNG LÒ

VVER1000 TRONG ĐIỀU KIỆN XẢY RA TAI NẠN NGHIÊM TRỌNG
Đoàn Mạnh Long
Trung tâm Đào tạo hạt nhân
Email: longdoanmanh28@gmail.com

Giới thiệu: Trong trường hợp xảy ra tai nạn nghiệm trọng, các cấu trúc vùng hoạt
của lò phản ứng nóng chảy và di chuyển xuống khu vực đáy thùng lò, dẫn đến hình thành
một bể nhiên liệu nóng chảy. Quá trình hình thành và cấu trúc của bể nhiên liệu nóng
chảy quyết định cách thức tác động nhiệt vào vỏ đáy lò phản ứng. Nghiên cứu và tính
toán diễn biến tai nạn có ý nghĩa quan trọng trong nghiên cứu biện pháp giữ vật liệu nóng
chảy bên trong lò phản ứng.
Bài báo sử dụng chương trình tính toán và mô phỏng sự cố MELCOR 1.8.6 để
nghiên cứu diễn biến tai nạn và đánh giá tác động nhiệt của nhiên liệu nóng chảy vào vỏ
đáy lò phản ứng VVER1000/V320 trong trường hợp xảy ra sự cố vỡ lớn (LBLOCA) kết
hợp với sự cố mất hoàn toàn nguồn điện (SBO). Kịch bản được chia thành hai kịch bản
nhỏ lần lượt ứng với khả năng hệ thống cấp nước an toàn thụ động không hoạt động
(kịch bản 1) và có hoạt động (kịch bản 2), với mục đích đánh giá vai trò của hệ thống cấp
nước làm mát vùng hoạt thụ động trong diễn biến tai nạn.
Các kết quả cho thấy hệ thống cấp nước làm mát vùng hoạt thụ động có ảnh hưởng
lớn tới diễn biến tai nạn như tăng cường các phản ứng ôxi hóa, làm chậm quá trình di
chuyển của vật liệu vùng hoạt xuống khu vựa đáy thùng lò và làm thay đổi thành phần
khối lượng thành phần mảnh vụn rơi xuống khu vực đáy lò dẫn tới giá trị thông lượng tác
động vào vỏ đáy thùng lò thay đổi. Giá trị thông lượng nhiệt ở bề mặt ngoài vỏ đáy thùng
lò ứng với trường hợp không có và có hệ thống cấp nước làm mát vùng hoạt thụ động lần
lượt là 0.59 MW/m2 và 0.62 MW/m2 theo thứ tự. Cả hai giá trị này đều nhỏ nhiều so với
tính toán thu được từ bài toán chuẩn cùng kịch bản được tính cho lò VVER1000/V320
(1MW/m2). Kết quả cũng cho thấy trong cả hai kịch bản đều xảy ra hiện tượng hỏng vỏ
đáy thùng lò phản ứng do hiện tượng dão nhiệt lần lượt tại các thời điểm 15192 giây và
14000 giây tương ứng với kịch bản 1 và kịch bản 2.

Từ khóa: VVER1000, IVR, MELCOR1.8.6, sự cố nghiêm trọng.