atw 2018-02


atw Vol. 63 (2018) | Issue 2 ı February


| | Fig. 5.

MIDAS Nodalization Scheme with 6 Channels.

| | Fig. 6.

MIDAS Nodalization Scheme with 12 Channels.

Test No.

Steam Flow Rate


| | Tab. 2.

RMSE Calculated Results of Bypass Fraction with Measured Data.

DVI-2 and DVI-4, the steam flow

rate seems to govern the prediction

accuracy. In case of high steam flow

rate tests (≥ 1.1 kg/s), the SPACE

code predicted the bypass fraction

well regardless of the number of

channels chosen. For the low steam

flow rate tests (≤ 1.1 kg/s), the RMSE

is ≥ 0.16 as shown in Table 2. More

detailed examination is presented


4.1 Results of High Steam Flow

Rate Tests

The results for the high steam flow

rate tests (KM100 ~ KM103, and

Number of Channels

4 6 12

KM109 1.8086 0.056 0.078 0.005

KM100 ~ 103 ≥ 1.1 0.017 0.019 0.017

KM104 ~ 108

0.161 0.211 0.287

≤ 1.1

KM110 ~ 114 0.252 0.334 0.462

| | Fig. 7.

Comparison of the Measured and Calculated ECC Bypass Fraction for the

High Steam Flow Cases.

KM109) are presented in Figure 7. For

the high steam flow rate tests, the

SPACE code predicts the bypass

fraction relatively well for all

nodalization cases.

The liquid flow pattern for the

KM100 test (highest steam flow rate

test) of each nodalization case are

presented in Figure 8 to Figure 10.

The liquid flow pattern for the all

nodalization cases are quite similar.

The direct bypass phenomena occurs

in the upper region of the downcomer

as the ECCS flow joins the high

velocity steam from the intact cold leg

and is swept away through the broken

cold leg. In the case of tests with a

high steam flow rate, the result of

the 4 channels nodalization is similar

to that of 6 and 12 channels. Hence,

the 4 channels representation is considered

a reasonable approximation.

4.2 Results of Low Steam Flow

Rate Tests

The results for the low steam flow rate

tests (KM104 ~108 and KM110 ~114)

are presented in Figure 11. Contrary

to the high steam flow rate cases, for

the low steam flow rate tests, the

SPACE code over-predicts the bypass

fraction for the all nodalization cases.

The liquid and vapor flow patterns

of the 6 channels case for the lowest

steam flow rate test (KM114) are

presented in Figure 12 and Figure 13,

respectively. Most of the liquid

injected from the DVI nozzle is swept

with the steam flow through the

break. The test indicated some downward

liquid flow at this steam flow


In the SPACE code, the interfacial

friction model is dependent on the

flow regime of the control volume.

Thus, for quantitative agreement with

the MIDAS experimental measurements,

the estimation of the flow

regime has to be properly predicted to

accurately estimate the bypass flow in

the upper downcomer. The SPACE

code selects the annular mist flow

regime based on the volume average

conditions, which explains the deviation

between the code prediction and

MIDAS tests in the case of low steam

flow rate.

4.3 Results of Condensation


It is worthy to note that for all the

studied cases, the code under-predicts

the condensation fraction as shown in

the Figure 14. The RMSE based on

calculated condensation fraction with

the measured condensation fraction

data are presented in Table 3. The

under-prediction tendency is more

distinct for finer nodalization (e.g. 12

channels) as depicted in Table 3. This

may clearly be tied to the heat transfer

correlation which in turn depends on

the flow regime. Due to mass conservation,

the lower condensation rate

leads to over-estimation of the bypass

Environment and Safety

Sensitivity Analysis of MIDAS Tests Using SPACE Code: Effect of Nodalization ı Shin Eom, Seung-Jong Oh and Aya Diab