SAFPWR home page
Unsymmetrical MSLB accident. Reference case.
rtvunsym_2.dat (download) is the counterpart of rtvsym_2.dat.
General trend
Although much less severe, the transient develops itself along the same successive transient phases observed on symmetric case.
Chart 01: During the (10,20) phase, when all the sg discharge to the break, power
qcn1 and
qcn2 extracted from
sg n1 and
sg n2 are similar but, from
sec=15
qcn2 < qcn1.
This because
wyan2 is slightly <
wyan1, or due to the fact that pressu is branched on loop1 ?
Chart 02: After completion of sg1 rapid voiding, at
sec= 170, core reactivity
rc (
chart 01) becomes negative and power (both
f2c and
q2c) drop rapidly until a new, semi stationary regime establishes.
wyan1 was chosen just high enough for allowing sg1 alone to extract the power of the 3 pumps, but still preventing the core to return partially to power due to excessive cooling.
q2c peak of 750e6 is observed (
chart 03) some 20 s after
rc peak;
f2c is barely in advance and higher than
q2c.
From that moment, at which the lowest margins are observed (
chart 04),
q2c starts decreasing steadily as the result of
tavc increase.
As anticipated, sg2 does not affect much the course of the transient.
From its isolation at 20 s, its power qcn2 (chart 01) drops rapidly and oscillates next around 0: the un-faulted sg do not succeed preventing or even mitigating core water cooling caused by sg1 rupture.
Assuming that, from sec= 10, sg2 remains totally thermally isolated, would thus not be an unreasonable simplifying assumption.
The small spurious transient that qcn2 exhibits when crossing zero power downwards are not relevant and are due to switching to "bottle" regime.
Core power tilting and safety margins
As for the sym case, a rapid axial-offset downward shift (chart 05), caused by return to power, is observed, but of lesser amplitude (.1 to -1.0 for sym case, vs .1 to -.25 here.
max u2c7 (chart 04) is 730 K, observed at q2c peak time.
As the result of the dramatic power shifting to the core bottom, the loss of dnb margin (chart 21) seems to be of less concern than excessive pellet temperature. This observation must of course be confirmed with a more realistic set of neutronic data and hot channel factors.
Comparing
fci profiles (
chart 06) with those of sym case shows that the squeezing of power at bottom is much less marked.
However the remarks about the necessity of refining the neutronic core description at its bottom remains pertinent.
The lesser severity of axial deformation is explained by absence of core boiling (
chart 07). All the primary system, including the dome remains sub cooled.
SG1 transient
Vapor quality
x2nj(j) (
chart 08) of the successive nodes of sg1 rise suddenly to 1 as the sg dries out at 170 s.
Until 170 s, all the nodes
j=1,4 of
sgn1 extract sensibly the same amount of heat
qnj (
chart 09). Thereafter, the first node takes over the extraction alone (
qcn1 = qn1 j1).Actually, the wet part only of it contributes to this removal [].
SG2 behavior
While, in sg n1, p3n1 (chart 10) drops steadily towards about 4e5, the pressure p3n2 in the other sg decreases from 70e5 to about 45e5 , and remains there after.
As already mentioned above, sg2 (chart 11, chart 12) switches alternatively from recirculation mode (qc2 and w2an > 0) to "hot bottle" mode when heat transfer flux reverses. Actually this mode transition should be smoother.
Anyway, the contribution of isolated sg is weak and the huge mass of hot "floating water" it contains does not succeed mitigating core cooling.
Thus, the common assumption of neglecting totally the contribution of isolated sg seems to be justified and not overly conservative.
Pressu response
With the initial steam region volume assumed , pressu (
chart 13) empties completely for a short time only (
j2x: transition element index).
After refill, it partially recovers the initial value of pressure and water level.
Steam quality
x2p (
chart 14) of vapor region is slightly <1 during outsurge and moderately > 1 after refilling.
Conversely, the nodes of the expansion column contain some vapor during voiding and become sub cooled with refill.
Effects of initial pressu region volume
v2p and heat exchange with wall will be looked at later.
Pressu heater is reactivated from
sec= 200 (
chart 15) but is unable to allow recovering the initial pressure value because
tavc (
chart 17) remains some 25 K lower than its initial value.
Remind that the pressu level regulation is not simulated here because the influence of feed-and-bleed on the transient should be minor.
Safety injection
Boration rate (chart 16) appears to be too low for modifying the course of transient significantly.
At long range, boration will just reinforce subcriticality.