ML24215A105
Text
Response to SDAA Audit Question Question Number: A-5.2.5-2 Receipt Date: 08/28/2023 Question:
In the FSER for the NuScale standard plant design application, the staff accepted the leakage approach based on demonstration of the CNV pressure and CES tank level for reactor coolant leakage detection satisfying the quantitative criteria specified in RG 1.45, Regulatory Positions C.2.1 and C.2.2. NRC was able to verify sensitivity of 0.05 gpm and leakage detection response time of 1 gpm leakage detection within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. NuScale FSAR response to eRAI 8841 (NuScale ref RAIO-0717-55060 - ML17205A650) provided clarification how design quantified leakage using methods described in FSAR. As the staff cannot find similar information within the SDAA, provide the following: a) Clarify how the instrument output of CNV pressure and CES tank level timing correlate to reactor coolant leakage rate,. b) Demonstrate how the leakage sensitivity of 0.05 gpm and the leakage detection, response time of one gpm within one hour are satisfied by using CNV pressure and the correlation, c) Demonstrate how the leakage sensitivity of 0.05 gpm and the leakage detection response time of one gpm within one hour are satisfied by using CES sample tank level and the correlation and d) TS BASES B 3.4.7 indicates, In addition to meeting the OPERABILITY requirements, the monitoring instrumentation is typically set to provide the most sensitive response without causing an excessive number of spurious alarms. Clarify how settings are defined for instrumentation sensitivity.
Response
Clarification of how design leakage is quantified is provided in response to parts (a), (b), and (c) of the question below. The equations use standard unit conversions, which are not specified for brevity.
(a) The reactor coolant leakage into the containment vessel (CNV) is calculated using pressure instruments using the ordinary differential equation for flow in a pumped vacuum system (Equation 1) and solving for the reactor coolant leak rate. The reactor coolant leak rate is given as a quantity of gas entering the CNV measured as pressure times volumetric flow. Therefore, NuScale Nonproprietary NuScale Nonproprietary
dp dt = Q Sp Equation 1 where p = pressure inside the CNV, t = time from reactor coolant leak initiation, Q = reactor coolant leak rate, and S = vacuum pump speed (volumetric flow).
An initial pressure is assumed for the initial condition at time t = 0 (Equation 2).
p 0 = p0 Equation 2 Then, Equation 1 is solved for containment pressure as a function of time (p(t)) given a constant reactor coolant leak rate (Equation 3). Then, p t = p0 Q S etS V + Q S
Equation 3 where V is the free volume inside containment.
The ideal gas law (Equation 4) is used to define the reactor coolant leak rate in terms of mass flow. To obtain a volumetric flow rate, Equation 3 is solved for the reactor coolant leak rate (Q) in Equation 5.
p = RT Equation 4 Therefore, Q = p = RT Equation 5 where = volumetric flow rate, = initial density of reactor coolant leak, R = specific gas constant of water vapor, and T = initial temperature of reactor coolant leak.
Equation 5 is then solved for the volumetric flow rate (Equation 6).
=
S p0e tS V p RT e 1
Equation 6 As the pressure approaches steady state over time (t ), Equation 6 reduces to Equation 7.
NuScale Nonproprietary NuScale Nonproprietary
= Sp RT Equation 7 Thus, given a pressure reading and assuming the leaking fluid is at reactor coolant system (RCS) conditions, a volumetric flow rate for the reactor coolant leak is calculated. The module control system performs this calculation automatically.
As reactor coolant leakage enters the CNV from the RCS, the fluid pressure descends below the vapor pressure and the fluid vaporizes. The CNV is held at a pressure below the vapor pressure associated with the lowest temperature surface within the CNV to preclude condensation. At least one containment evacuation system (CES) vacuum pump operates continuously so that vapor is constantly removed.
As the vapor passes through the vacuum pump, it is condensed back to liquid in the condenser.
The CES sample vessel collects the liquid, where its temperature, radioactivity, and the CES sample vessel level is measured over time.
Given that the dimensions of the CES sample vessel are known, the volumetric flow rate can be calculated using Equation 8. The internal geometry of the vessel is assumed to be a vertical cylinder.
= r2h ft Equation 8 where f = condenser bypass factor, r = internal radius of CES sample vessel, and h = liquid level in CES sample vessel.
The calibrated output of the level sensor is used to measure the liquid level in the CES sample vessel. The measurement begins at time t = 0 with the CES sample vessel drain valve closed.
As liquid accumulates, Equation 9 calculates an average reactor coolant leak rate for the time it takes to cycle the CES sample vessel. However, real time leak rate trending can be performed by using continuous CES sample vessel level indication over time. After the calculation is performed, the results can be correlated to a volumetric leak rate at RCS conditions. A condenser bypass factor is also included to account for the fraction of vapor that is not condensed in the condenser stage; this factor is dependent on the condenser design.
= r2dh fdt Equation 9 NuScale Nonproprietary NuScale Nonproprietary
(b) The reactor coolant leak detection sensitivity can be demonstrated by rearranging Equation 3 with Equation 5 substituted for Q to solve for time and by substituting the pressure inside the CNV for the initial pressure plus the instrument resolution. Once rearranged (Equation 10), this will be used to calculate the instrument response time for the containment atmosphere to change from its initial pressure to the initial pressure plus the minimum instrument resolution. Therefore, t = V S ln pi+pres RT S
piRT S
Equation 10 where pi = pressure inside the CNV and pres = pressure instrument resolution.
Assuming nominal RCS conditions of 540 degrees F (999.7 degrees Rankine) and 2000 psia with an initial containment pressure of 1 psia, a 1 gpm (0.134 ft3/min) reactor coolant leak with a vacuum pump operating at ((2(a),(c),ECI (effective pump speed for vapor for a 300 acfm vacuum pump considering system losses) will be detected by a pressure instrument with a resolution of 0.1 psi, which is much higher than a typical commercial vacuum pressure instrument, in well under one hour (Equation 11). Therefore, t = Vcontainment SCEpump ln PCNTinit+PCNTincrease densorigRvaporTRCSleakrate SCEpump PCNTinit densorigRvaporTRCSleakrate SCEpump = (( }}2(a),(c),ECI Equation 11 where Vcontainment is containment volume (6200 ft2), SCEpump is containment evacuation pump speed (( }}2(a),(c),ECI, PCNTinit is containment initial pressure (1.0 psi), PCNTincrease is containment pressure increase (0.1 psi), densorig is original density (45.95 lbm/ft3), Rvapor is the specific gas constant (85.78 ft lbf/lbm R), TRCS is RCS temperature (540 degrees F), and leakrate is the leak fate (1 gpm = 0.134 ft3/min). The leak flow rate is calculated using Equation 12: Qleak = densorigRvaporTRCSleakrate = (( }}2(a),(c),ECI Equation 12 Where Qleak is the gas leakage. While instrument loop uncertainty or signal processing time is not included in Equation 11, its inclusion will not cause the reactor coolant leak detection time to exceed one hour. NuScale Nonproprietary NuScale Nonproprietary
To demonstrate how CES pressure instrument supports monitoring a 0.05 gpm leak, Equation 13 is used to compare off-the-shelf pressure instrument resolution to the steady state containment pressure resulting from a 0.05 gpm leak. PCNT = Qleak
Equation 13 Then, Equation 14 calculates the gas leakage. Qleak = VleakleakRH2OT = (( }}2(a),(c),ECI Equation 14 Equation 15 calculates the containment pressure. PCNT = (( }}2(a),(c),ECI Equation 15 Because CES pressure instruments are specified to have better resolution than (( }}2(a),(c),ECI, they can monitor leaks as small as 0.05 gpm. (c) By rearranging Equation 8 to solve for time, a 1 gpm reactor coolant leak detection time is calculated (Equation 16). Therefore, t = r2hres f Equation 16 where hres = level instrument accuracy. Because the level instrument is calibrated to the differential pressure sensor at the bottom of the sample vessel, an accuracy of 0.004 psi is assumed based on typical industry differential pressure level instruments., Conservatively, the assumed temperature and pressure of water in the sample vessel is 212 degrees F and 1 atm. Thus, the minimum detectable level change in the vessel is calculated using Equation 17. hres = pres psampleg = (( }}2(a),(c),ECI Equation 17 NuScale Nonproprietary NuScale Nonproprietary
Using the level instrument accuracy and an assumed sample vessel diameter of 11 inches with a condenser bypass factor of (( }}2(a),(c),ECI, meaning (( }}2(a),(c),ECI of vapor is not condensed, the reactor coolant leak detection time for a 1 gpm leak is shown in Equation 18. t = (( }}2(a),(c),ECI Equation 18 Similarly, rearranging Equation 8, a 0.05 gpm leak results in the steady state rate of CES sample vessel level increase shown in Equation 19. leak = r2 = (( }}2(a),(c),ECI Equation 19 A change in water level of ((
}}2(a),(c),ECI is approximately (( }}2(a),(c),ECI. This change in level will be detected by the CES level instruments over time.
(d) For the reactor coolant leakage detection in the US460 design, the CES pressure and sample level instruments provide the monitoring and detection capabilities specified in Regulatory Guide 1.45, Regulatory Positions C.2.1 and C.2.2. For these RCS leakage monitoring instruments, setting to the most sensitive response without causing an excess number of spurious alarms means that the instrument sample rate, range, and alarm settings need to be selected to provide the greatest resolution practical to measure the smallest flow rate practical while still covering an appropriate range. The accuracy of an indicated value for most instruments is a function of the calibrated instrument span; therefore, there is a balance to selecting a sufficiently wide instrument span to cover expected process conditions while still achieving the best resolution possible for small process indications associated with low to no leakage. The CES pressure and sample level instruments are non-safety related and the calibration settings for instrument setpoints and uncertainty will be developed in accordance with plant procedures based on generally accepted industry standards. The SDAA technical specification 3.4.7 surveillance requirements show that there will be a channel check and a channel operational test in accordance with the Surveillance Frequency Control Program. No changes to the SDAA are necessary. NuScale Nonproprietary NuScale Nonproprietary}}