Effect of NonCondensables on Natural Circulation Passive Safety Systems

1
2 RCM of the IAEAs CRP Oregon State
University, Corvallis, USA Aug-Sept 2005
Effect of Non-Condensables on Natural
Circulation Passive Safety Systems José Luis
Muñoz-Cobo, Luis Herranz, Juan Carlos de la
Rosa, Alberto Escrivà Presented by Juan Carlos
de la Rosa Instituto de Ingeniería Energética
Universidad Politécnica de Valencia Edificio I4
Camino de Vera s/n 46022 Valencia -
SPAIN delarosablul_at_yahoo.es
2
Main Contents

• General Presentation of the UPV
• Main aspects of the developed work
• Condensation inside vertical tubes in presence of
  NC gases
• Condensation on Finned Tubes Heat Exchangers in
  presence of NC gases

3
General Presentation of the UPV

• The UPV is one of the leading Technical
  Universities in Spain with 35000 Students and
  2500 Faculty lecturers, Professors and
  Researchers
• The Institute of Energy Engineering belongs to
  the UPV and has 60 researchers that works in the
  following areas
• Thermal Engineering
• Nuclear Engineering and Thermalhydraulics
• Electric Engineering
• Renewable Energy Sources

4
General Presentation of the UPV

• Thermalhydraulics and Nuclear Engineering Group
  Participation in Projects related with Natural
  Circulation Passive Safety Systems
• CEE-TEPSS Technology Enhancement of Passive
  Safety Systems 1996-1999, European Fourth
  Framework.
• CONGA (Subcontract with CIEMAT). Containment
  Behaviour in the Event of Core Melt with Gaseous
  and Aerosol Releases 1996-1999.
• CEE-NACUSP NaturalCirculation and Stability
  Performance of BWR 1999-2004. Fifth European
  Framework Project.
• EPP-1000, Small and Large Break LOCA Analysis in
  collaboration with ANSALDO and DTN. 1999-2000.

5
Main Aspects of the developed work
By means of the proposed PIRT -which has to be
finished and consolidated in this 2 RCM-, the
UPV and Ciemat have already made the following
issues

• Related with the 2 Phenomena of the proposed
  PIRT -Tracking of Non-Condensables-, we have
  studied the main different models which simulate
  the condensation inside vertical tubes in the
  presence of NC gases. Also, we have made a model
  for the condensation on finned tubes heat
  exchangers in presence of NC gases, which gives a
  good agreement. The main prototypes of passive
  safety reactors which can incorporate this models
  in its simulations are the ESBWR and SBWR for the
  Passive Containment Cooling Condensers (PCCC),
  and the SWR-1000 for the Finned Tubes Heat
  Exchanger.
• Related with the 3 Phenomena of the Proposed
  PIRT -Condensation on the containment
  structures-, we have developed an analytical
  model for the condensation on the containment
  structures, explicitly done for the AP600 passive
  safety reactor.

6
Main Aspects of the developed work

• Finally, related with the 9 Phenomena of the
  proposed PIRT -Liquid Temperature
  stratification-, we have studied the
  Stratification that occurs in the hot legs of a
  Pressurized Water Reactor (PWR) using commercial
  codes (like CFX) and developing new codes (like
  TUBO3D).

7
Condensation inside vertical tubes in presence of
NC gases

• The ESBWR and the SBWR extract the containment
  heat by means of the passive containment cooling
  condensers (PCCC). Each condenser is formed by a
  set of vertical tubes connected to common lower
  and upper headers or drums, and submerged into a
  water pool. A tube drains the steam plus
  non-condensable mixture from the containment and
  drives it to a distributor that transports the
  mixture by natural circulation to the upper
  headers of the PCCCs. Once the gas enters into
  the tubes, the steam condenses on the walls of
  the tubes, and the condensation heat is
  transferred to the PCCC pool.
• The condensate is drained by gravity to the
  lower header, where accumulates at the lower part
  of the lower drum where it is removed by gravity
  through the drain line which is directly
  connected to the reactor vessel. The
  non-condensed steam plus the non-condensable
  gases are discharged through the vent line into
  the wetwell pool.

8
Condensation inside vertical tubes in presence of
NC gases
9
Condensation inside vertical tubes in presence of
NC gases
10
Condensation inside vertical tubes in presence of
NC gases
The main objective is to propose several models
which can be easily implemented in the existing
thermal-hydraulics codes, like RELAP, TRAC-BF,
and TRACE. There exists two main paths to arrive
to construct a model

• The first one is an a priori or analytical
  method, which sees the physical system as a white
  box. This means that the model will be
  constructed by means of the conservation
  equations along with closure relations, and/or
  using the heat and mass transfer analogy.
• The second one is an a posteriori or
  semi-empirical method, which sees the physical
  system as a black box, in a way that the model
  will be constructed based on an analytical
  previous model, making the correct adaptation to
  the real physical system by means of an empirical
  correlation.

11
Condensation inside vertical tubes in presence of
NC gases

• Following this aim, we have compared three
  models
• The University of California at Berkeley model,
  called the UCB model, developed by Vierow and
  Schrock, all together with the improved UCB
  model, called UCBA, and developed by Kuhn.
• The Second model is the one from the Middle East
  Technical University at Ankara, called the
  METU-TAEA model, and performed by Anglar and
  Tanrikut.
• Finally, the third analyzed model is the UPV-FIT
  model, performed by José Luis Muñoz Cobo et al.

12
Condensation inside vertical tubes in presence of
NC gases

• All these models belong to the second
  way to construct a model, that is, they use an
  empirical factor to account for the phenomena not
  implemented in the source analytical model.
• As we have previously said, these
  simple models are based on obtaining the real
  heat transfer coefficient (HTC) by means of one
  previously obtained analytically, in our case,
  through the Nusselt theory
• In this way, the positive and negative
  aspects of each model will depend on the
  phenomena not included analytically, that is, in
  the degree of dependence that each one has with
  the experiments that were used to get the
  degradation factor.

13
Condensation inside vertical tubes in presence of
NC gases

• The UCB model
• We can obtain a Nusselt number for the
  laminar and turbulent region which will be used
  to calculate the film thickness

14
Condensation inside vertical tubes in presence of
NC gases

• The UCB model
• The degradation factor must account for two kind
  of phenomena not included in the Nusselt theory
• The enhancement of the HTC produced by the
  shrinkage caused by the transfer of momentum from
  the gas mixture to the condensate film
• The degradation produced by the NC effect, which
  would almost counterbalances the forces created
  by the depression that occurs when the steam
  condensates in the interface

15
Condensation inside vertical tubes in presence of
NC gases

• The UCBA model
• This model was improved by the
  Khun-Schrock-Peterson correlation denoted as
  UCBA, that takes on account the waviness of the
  condensate layer and the interfacial shear
  stress, but the model is more complex because
  involves to solve the Nusselt liquid film
  boundary layer equation with specified shear
  stress at the interface and it is not as direct
  as the Vierow Schrock model, because it is
  necessary to iterate to solve this equation.
• In this model
• where

16
Condensation inside vertical tubes in presence of
NC gases

• The METU-TAEA model
• In this case, the degradation factor is given
  by
• Where f1 is the enhancement factor of
  the HTC produced by the shrinkage caused by the
  transfer of momentum from the gas mixture to the
  condensate film. Also, this enhancement factor
  takes into account the effect produced by the
  onset of disturbance waves at the interface at
  relatively low condensate Reynolds numbers

17
Condensation inside vertical tubes in presence of
NC gases

• The METU-TAEA model
• with
• where d1 is the film thickness without
  interfacial shear stress, and d2 is the film
  thickness with interfacial shear stress. We note
  that the interfacial shear stress is influenced
  by both the interface velocity and the mixture
  side velocity. For this reason fOTHER is
  correlated as
• Where C1, C2, Z1, and Z2, are constants
  of the model, Rel and Reg are the Reynolds
  numbers of the condensate and the gas plus NC
  mixture respectively.

18
Condensation inside vertical tubes in presence of
NC gases

• The METU-TAEA model
• As the NC gases are being accumulated in the
  interface creating a gradient concentration, the
  Ficks law assures that a mass transfer of NC
  gases will produce from the interface to the
  bulk. According to Aglar and Tanrikut, this can
  be expressed with the help of the Sherwood
  number. Then, the degradation factor can be
  written in the following expression
• where C3, and Z3 are model constants,
  ync is the NC molar fraction and Sh is the
  Sherwood number.

19
Condensation inside vertical tubes in presence of
NC gases

• The UPV-FIT model
• This model only depends empirically on the NC
  mass fraction, because the other phenomena not
  included in the Nusselt theory -waviness and
  shear stress effects- have been implemented
  through d, in a way that its influence has not
  been accounted for experimentally. This goal has
  been achieved introducing a different d from the
  one of Nusselt. In this way, instead of
• We have

20
Condensation inside vertical tubes in presence of
NC gases

• The UPV-FIT model
• Where
• d is obtained solving
• where the interfacial shear stresses
  and the wavy effects of the condensate layer
• have been introduced, and dN is the one
  obtained from the Nusselt theory.

21
Condensation inside vertical tubes in presence of
NC gases

• The UPV-FIT model
• In this way, we can use a degradation factor that
  only takes account for the NC mass fraction
• Where
• and

22
Condensation inside vertical tubes in presence of
NC gases

• Comparisons
• First, we compare the results of the
  UCBA and METU-TAEA models, defining an average of
  the absolute values of the relative differences
  in between the calculated hcal and experimental
  hexp heat transfer coefficients as
• The second comparison is made with all
  the models for the NC effect, using the Vierow
  and Schrock experiments. Also, we evaluated a
  mixed model defined as

23
Condensation inside vertical tubes in presence of
NC gases
24
Condensation inside vertical tubes in presence of
NC gases
Average relative error
25
Condensation inside vertical tubes in presence of
NC gases

• Conclusions
• The actual models which can be
  implemented in the Thermal-Hydraulic Codes to
  simulate the Condensation inside tubes in
  presence of NC gases, are all semi-empirical
  models.
• The problem of this kind of model
  construction is that we almost can not enhance
  it, that is, as the models use an empirical
  factor to account for the phenomena not included
  in the Nusselt theory, it is almost impracticable
  to get a better approach.
• The next phase will be to implement
  them in a Thermal-Hydraulic Code like TRAC,
  RELAP, or TRACE, and to compare its results
  between them, and with the models now implemented
  in the Thermal-Hydraulic Codes.

26
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The finned tube HX consists in bundles
  of finned tubes cooled internally by natural
  circulation as displayed in the figure below .

27
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The working principle of FTCCC can be
  explained as follows if the temperature in the
  drywell atmosphere increases over the one in the
  dryer-separator storage pool, then the steam
  condenses on the tubes and heat up the water
  inside the HX finned tubes. This water of less
  density moves upward by natural circulation due
  to the slope of the exchangers tubes, and flows
  through the outlet line, discharging the higher
  temperature cooling water in the pool. The outlet
  line ends at a higher elevation level than the
  inlet line, so the lifting forces are increased
  for the whole system.
• If the medium on the heat exchanger
  after an accident is a nitrogen-steam mixture,
  then the natural convection flow over the finned
  tubes is downward toward the core flooding pool,
  because the density of the nitrogen-steam mixture
  increases with decreasing temperature. However,
  the opposite is true for a hydrogensteam
  mixture.
• These passive HX condenses the steam
  inside the containment and transport the heat by
  natural circulation to a large pool with
  capability to act as a heat sink at least during
  72 hours.

28
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases
29
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The total thermal resistance from the bulk gas to
  the coolant is formulated as a parallel
  combination of the convective and condensation
  gas resistances coupled in series to those of
  condensate layer, the wall, and the coolant.
• The condensate layer thermal resistance is
  calculated by means of an Adamek based
  condensation model.
• The Murata model is implemented to compute the
  fins efficiency, because it is calculated
  through a direct expression, and not using an
  iteration procedure as the Adamek model requires.
  
• The gas mixture (Steam plus NC) thermal
  resistance is formulated based on a diffusion
  layer theory formulated by Peterson, which uses
  the heat and mass transfer analogy.
• The model results are compared with available
  experimental data of the thermal-hydraulic phase
  (steamnon-condensable gases) of the experiments
  of the CONGA project of the European Community.

30
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The overall heat flux from the
  containment bulk gas mixture at temperature Tb,
  to the coolant circulating inside the finned tube
  at temperature Tc, is given by Newtons law of
  cooling
• To compute the total HTC, we use the
  electrical analogy

31
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• And the total HTC is
• where
• and

32
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• For the Condensate Layer Thermal
  Resistance we use the Adameks model, which is
  based on calculating the total mass condensation
  rate. For this goal, the finned tube is divided
  into 3 circumferential zones, according to the
  amount of condensate that we have in the interfin
  space

33
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The heat transfer coefficient for the
  condensate layer is
• M1/2 denotes the total mass condensation rate per
  one half fin at one side of the tube.
• L is the condensing length of the tube.
• Nf is the number of fins per unit length.
• ?T is the difference of temperatures between the
  temperature at the liquid-gas interface and the
  temperature at the surface of the tube.

34
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The heat transfer coefficient for the
  gas diffusion layer is
• hg,cond is the condensation heat transfer
  coefficient for the gas.
• Tcond is the suction factor for condensation.
• hg,conv is the convection heat transfer
  coefficient for the gas.
• Tconv is the suction factor for convection.
• These suction factors are caused by the
  boundary-layer shrinkage produced by the transfer
  of momentum from the gas mixture to the
  condensate film.

35
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• The convection HTC of the gas mixture
  is related to the Nusselt number, where the
  Nusselt number depends on the convection regime
• The Sherwood number is obtained from
  the same correlations quoted earlier and using
  the heat and mass transfer analogy, which allows
  Sh to be correlated as the Nu, provided that the
  Prandtl number of the gas mixture Prg, is
  substituted by the Schmidt number

36
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• Finally, the Condensation Conductivity, taken
  from Peterson, is given by the following
  expression
• D denotes the air-steam diffusion coefficient
  taken at Tavg temperature.
• hpfg is the specific enthalpy of phase change
  plus the subcooling energy to the average
  condensate temperature.
• Cg is the gas mixture molar concentration.
• Mst is the molecular weight of the steam.
• Tavg is the average temperature in the boundary
  layer.
• Xst,avg and Xnc,avg are the molar fraction
  logarithmic averages of the steam and
  non-condensable respectively.

37
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• Comparisons
• The results of the experiments
  performed at JRC by De Santi and at PSI by
  Suckow, are compared with the results of the
  program FINSTUBOAE that contains the model for
  finned tube condensation explained before.
• De Santi
  experiments Suckow
  experiments

38
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• Conclusions
• The model to simulate the behaviour of
  the Finned Tubes Heat Exchanger in presence of
  NC gases is constructed analytically.
• As we can observe, the results given by
  this model are better than the ones obtained with
  the models for condensation inside tubes, all of
  which were constructed semi-empirically.
• Our model keeps a non-dependency of any
  specific group of experiments, which permits a
  broad range of use of our model to different
  containment conditions.
• Although this model has been
  obtained analytically, it rests partially in some
  empirical correlations for the non-dimensional
  numbers as Nusselt or Sherwood number.

39
Condensation on Finned Tubes Heat Exchangers in
presence of NC gases

• Conclusions
• As the model is constructed
  analytically, it can be enhance by suppressing or
  getting better the empirical correlations or the
  simplificative hypothesis, as the followings
• The use of the Clausis-Clapeyron equation, which
  implies a saturated condition of the vapor at the
  entrance.
• To include the convection energy due to the NC
  gases, which actually is not considered.
• To include the conduction therm in the Heat
  Transfer of the boundary layer to the interface.
• To solve Ficks law considering the variation of
  the total gas density.
• The next step will be study the
  influence of aerosols deposition in the Heat
  Transfer and implement it in the FORTRAN
  FINSTUBOAE program.