IAEA Coordinated Research Project

on

NATURAL CIRCULATION PHENOMENA, MODELING AND RELIABILITY OF PASSIVE SYSTEMS THAT UTILIZE NATURAL

CIRCULATION

Research Contract No. 12758/R0

Report

On

Integral System Tests in ITL simulating AHWR (Section 4.4 of Chapter 4)

By

R. K. Bagul, M. R. Gartia, Naveen Kumar, P.K. Vijayan and D. Saha

Reactor Engineering Division Bhabha Atomic Research Centre Trombay, Mumbai 400085, India

March 2008

Integral System Tests in ITL Simulating AHWR

4.4.1 INTRODUCTION The Advanced Heavy Water Reactor (AHWR) [1] is a light water cooled and heavy water moderated pressure tube type boiling water reactor. In AHWR, it is proposed to remove the core heat by natural circulation during start-up, power raising, normal operation, transients and accidental conditions. AHWR uses several passive concepts with a view to simplify the design and to enhance safety. An integral test facility (Fig. 1 shows the building housing this facility) simulating MHT and safety systems of AHWR has been set up at BARC to investigate the overall system behaviour under different operating conditions, transients and accidents like LOCA. In addition to main heat transport system (MHTS), the Integral Test Loop (ITL) simulates the emergency core cooling system (ECCS), Isolation Condenser (IC) system and Gravity Driven Cooling System (GDCS) of AHWR. The scaling philosophy adopted for ITL is based on a three level approach. At the global level, power-to-volume scaling philosophy is used. However, due considerations are given to simulation of local phenomena also. 4.4.2 AHWR MHTS AND DECAY HEAT REMOVAL SYSTEM In the Main Heat Transport System (MHTS) (Fig. 2), the sub-cooled water flows from Reactor Inlet Header (RIH) to core through 452 feeder pipes. The subcooled water gets heated up as it rises through the 452 no of fuel assemblies. Boiling takes place in the fuel assemblies and steam-water two-phase mixture comes out of the channel. Low quality steam-water mixture flows from the core outlet to 4 horizontal steam drums through 452 tail pipes. In the steam drum, steam gets separated from water by gravity action. Under normal operating conditions, steam flows to the turbine, which is returned to steam drum as feed water at 130 oC. The feed water, after mixing with the saturated water flows through downcomers to the inlet header and cycle repeats itself.

Besides core heat removal by natural circulation under normal operating conditions, a

no of passive systems are incorporated in AHWR in order to simplify the design and to enhance safety. The following passive systems are incorporated in AHWR

• Advanced accumulator using a fluidic device to inject emergency coolant during the initial stages of LOCA.

• Gravity driven cooling system (GDCS) for uninterrupted core cooling for at least 3 days after the accumulators are exhausted.

• Passive decay heat removal system with isolation condensers

4.4.3 OBJECTIVES OF ITL The main objectives of the test facility are:

• Generation of database for

o Plant transients o Accidental scenarios like LOCA.

• Evolution and validation of a start-up procedure. • Generation of database for the performance evaluation of the following in the plant

environment: o Natural circulation in the PHT loop, including parallel channel

behaviour, o Steam separation process in the steam drum, o Fluidic device in the advanced accumulator, o Gravity Driven Cooling System.

• Investigations of asymmetric behaviour like operation with one steam drum at a higher pressure or break in a steam line connected to one steam drum.

4.4.4 SCALING PHILOSOPHY The scaling philosophy adopted for ITL is based on a three level approach [2, 3 and 4]. At the global level (commonly known as integral scaling), power-to-volume scaling philosophy [5 and 6] is used. Boundary flows such as feed water flow rate, steam flow rate, safety injection etc. which can affect the integral behaviors are also required to be scaled. In addition, important local phenomena which can influence the integral system behavior such as Critical Heat Flux (CHF), steam-water separation in the steam drum, flow pattern transition etc. are also preserved [7]. The details of the scaling philosophy are presented in following sections. 4.4.4.1 Power to volume scaling philosophy For scaling, the primary heat transport system of nuclear reactors is assumed to consist of a large number of pipe sections, which are either vertical or horizontal. One-dimensional homogenous two-phase flow (EVET model) through a vertical pipe can be represented by the following governing equations:

0z

)v(t

=+∂ρ∂

∂ρ∂ Conservation of mass (1)

ρτζ

∂∂

∂∂ρ

∂∂ρ g

Azp

zvv

tv

c

w −−−=+ Conservation of momentum (2)

zpv

tp

Aq

zhv

th

c

h

∂∂

∂∂ζ

∂∂ρ

∂∂ρ ++=+ Conservation of energy (3)

These equations are non-dimensionalised using the following substitutions:

t tt

+ =0

; z zz

+ =0

; ρ ρρ

+ =0

; ;vvv0

=+ p pp

+ =0

; (4a)

200vfρ

ττ =+ ; h hh

+ =0

; q qq

+ =0

; ζ ζζ

+ =0

; A AAc =

0

; (4b)

Where, the quantities having subscript ‘0’ are some known reference values. The non-dimensional governing equations are:

+

++

+

+

⎥⎦

⎤⎢⎣

⎡+

z)v(

ltv

t 0

00

∂ρ∂

∂∂ρ = 0 (5)

+

+++

+

++

⎥⎦

⎤⎢⎣

⎡+

zvv

ltv

tv

0

00

∂∂

ρ∂∂

ρ = ++

++

+

+

⎥⎦

⎤⎢⎣

⎡−⎥

⎦

⎤⎢⎣

⎡−⎥

⎦

⎤⎢⎣

⎡ρτζζ

∂∂

ρ 0

0

c

w00

0c000

0

vgt

Atvf

Azp

vltp 00

(6)

+

+++

+

++

⎥⎦

⎤⎢⎣

⎡+

zhv

ltv

th

0

00

∂∂

ρ∂∂

ρ =+

++

+

+

+

++

⎥⎦

⎤⎢⎣

⎡+⎥

⎦

⎤⎢⎣

⎡+⎟

⎟⎠

⎞⎜⎜⎝

⎛⎥⎦

⎤⎢⎣

⎡

zpv

hlptv

tp

hp

Aq

hAtq

000

000

00

0

c

h

000c

000

∂∂

ρ∂∂

ρζ

ρζ (7)

From equations (5) to (7), we can say that for similarity to be achieved between processes observed in a model (denoted by subscript ‘m’) and in the prototype (denoted by subscript ‘p’), the following equalities must be satisfied:

ml

vt⎥⎦⎤

⎢⎣⎡ =

plvt⎥⎦⎤

⎢⎣⎡ (8a)

mlv

tp⎥⎦

⎤⎢⎣

⎡ρ

= plv

tp⎥⎦

⎤⎢⎣

⎡ρ

(8b)

mc

w fvtA ⎥

⎦

⎤⎢⎣

⎡ζ=

pc

w fvtA ⎥

⎦

⎤⎢⎣

⎡ζ (8c)

mv

gt⎥⎦⎤

⎢⎣⎡ =

pvgt⎥⎦⎤

⎢⎣⎡ (8d)

mc

h

hqt

A ⎥⎦

⎤⎢⎣

⎡ρ

ζ=

pc

h

hqt

A ⎥⎦

⎤⎢⎣

⎡ρ

ζ (8e)

mh

p⎥⎦

⎤⎢⎣

⎡ρ

= ph

p⎥⎦

⎤⎢⎣

⎡ρ

(8f)

The subscript ‘0’ has been omitted from the above equations for convenience. Assuming that the same fluid is going to be used in the model and prototype and imposing the conditions of isochronicity and equality of pressures and temperature (and hence properties), then equation (8e) can be written as:

mc

h

Aq⎥⎦

⎤⎢⎣

⎡ζ=

pc

h

Aq⎥⎦

⎤⎢⎣

⎡ζ

Multiplying both the sides by l,

mc

h

lAql⎥⎦

⎤⎢⎣

⎡ ζ =

pc

h

lAql⎥⎦

⎤⎢⎣

⎡ ζ

Since l hζ is heat transfer area, l qhζ is the power (φ ). Similarly, lAc is the volume (V). In other words, equation (8e) can be expressed as:

φφ

p

m

= VV

p

m

(9)

where, V is the volume. Similarly, equation (8a) can be expressed as:

m

vl

⎛ ⎞⎜ ⎟⎝ ⎠

= p

vl

⎛ ⎞⎜ ⎟⎝ ⎠

or c

c m

A vA l

⎛ ⎞⎜ ⎟⎝ ⎠

= c

c p

A vA l

⎛ ⎞⎜ ⎟⎝ ⎠

This can be expressed as, QQ

p

m

= VV

p

m

Since the properties are same in the model and prototype, this can be expressed as

QQ

p

m

= WW

p

m

= VV

p

m

(10)

From equation (8d) we get,

p

m

vv

= 1 (11)

From equation (8b) we get,

( )( )p

m

vlvl

= 1 (12)

Equations (11) and (12) can be satisfied only if elevations (in case of horizontal pipes, length) are preserved between model and prototype, i.e.,

llm

p

= 1

This also implies that,

( )( )A

Ac p

c m

= VV

p

m

(13)

i.e., the flow areas are scaled in the volume scale ratio and hence the diameters are scaled in the square root of the volume scale ratio. From equation (8c) we get, ζ w

c m

fA

⎛⎝⎜

⎞⎠⎟ =

ζ w

c p

fA

⎛⎝⎜

⎞⎠⎟ ⇒

ζ w

m

fV

⎛⎝⎜

⎞⎠⎟

= ζ w

p

fV

⎛⎝⎜

⎞⎠⎟

⇒ ( )( )ζ

ζw p

w m

f

f =

VV

p

m

(14)

Thus, the power to volume scaling relations can be expressed as (1) equal pressures and properties i.e.,

pp

p

m

= ρρ

p

m

= hh

p

m

= 1 (15)

(2) power to volume scaling relation

φφ

p

m

p

m

p

m

p

m

WW

QQ

VV

S= = = = (16)

Which also implies isochronicity i.e., tt

p

m

= 1

(3) Equal elevations

ll

p

m

= 1 which also implies ( )( )A

AVV

Sc p

c m

p

m

= = (17)

(4) Equal frictional effects

ζ ζw

c m

w

C p

fA

fA

⎡

⎣⎢

⎤

⎦⎥ =

⎡

⎣⎢

⎤

⎦⎥ Or

( )( )ζ

ζw p

w m

p

m

f

fVV

S= = (18)

Thus, once the power to volume scaling ratio ‘S’ is fixed, all other quantities are automatically fixed. Hence, this method of scaling is known as ‘power to volume scaling philosophy. Since the full-size, full-length components are proposed to be used for most components in ITL, all the ratios expressed by equations (15) to (18) can be met in ITL. However, some components like the header, downcomer and steam drum have to be necessarily downsized for which it is difficult to preserve the relationship expressed by equation (18).

4.4.4.2 Local phenomena scaling Important local phenomena that can affect the integral system performance need to be scaled appropriately. The local phenomenon to be simulated depends on the component of the loop. The following local phenomena are considered.

• Critical Heat Flux (CHF): CHF is the most important local phenomena to be simulated in the core simulator. Use of full scale, full power and full pressure fuel channel simulator ensures that CHF can be simulated in ITL.

• Flashing: Flashing can occur in the riser due to decrease in the static head. It can

cause instability during start-up. Since full size, full height tail pipes (risers) are used in ITL, flashing can be simulated.

• Geysering: Geysering can occur due to the condensation of large slug bubbles that are generated in the core when they mix with subcooled water in the riser pipes. This phenomenon depends on size of slug bubble formed in the core, flow velocity, pressure and temperature in the riser. It also depends on subcooled boiling. Geysering is a highly thermal non-equilibrium phenomenon by nature. Since full scale core and riser pipe are simulated along with the operating conditions, Geysering can be simulated in ITL.

• Flow Stratification: Flow stratification occurs in horizontal pipes/large pipe sections

at low steaming rate and hence expected to occur at low power. To simulate flow stratification in ITL, it requires Froude number simulation.

• Flow pattern transition: Flow pattern transition can occur in the two-phase regions of

the loop. It can cause instability generally known as flow pattern transition instability. Again since full size, full height and operating conditions are simulated in ITL, flow pattern transition instability can be simulated in ITL.

• Steam-Water Separation: Steam-water separation (including carryover and carry

under) is the most important local phenomena to be simulated in the steam drum. Steam-water separation at the interface depends on the interface area. Carryover depends on the steam velocity above the interface and carry under depends on the water velocity in the downcomer section of the steam drum. Since the steam drum is downsized the interface area is larger than required and since the steam velocity and downcomer velocity in the steam drum are not scaled, steam-water separation phenomenon is not exactly scaled.

4.4.4.3 Boundary flow scaling For simulating the effect of certain components or systems on the integral system performance, it is sufficient to simulate only the boundary flow of mass and energy.

• Feed water flow: The feed water temperature needs to be maintained in the ITL same as that of AHWR. The feed water mass flow rate at normal operating condition has to be scaled in the same ratio as the volume scale.

• Accumulator Injection: For accumulator, the following are to be simulated:

o Transient non-dimensional injection flow rate. o Loss coefficient of the fluidic device and pressure loss in the injection pipe. o Water temperature and nitrogen pressure in the accumulator.

• Gravity Driven Cooling System (GDCS) Injection: Simulation of GDCS injection in

ITL requires the same transient non-dimensional injection flow rate as that of AHWR. ITL maintains an initial nitrogen pressure in the GDCS tank to compensate for the difference of elevation. The temperature of the GDCS liquid in the model and prototype is maintained same.

• Critical Flow: Critical flow simulation is important for LOCA following a pipe break.

It depends on break size, upstream conditions (P, h), etc. 4.4.5 DESCRIPTION OF ITL 4.4.5.1 Systems included in ITL The Integral Test Loop consists of the following systems:

• PHT system with all its components. • ECCS system including advanced accumulator and Gravity Driven Water Pool

(GDWP) • Secondary system up to turbine stop valve (CIES valve) • Feed water system (flow rate and inlet temperature) • Isolation Condenser • Control Systems for

o PHT Pressure o Steam Drum level o Power

• Normal power raising • Step back / Set back • Trip • Decay power simulation

o Feed water temperature and flow rate o Steam pressure and flow along with associated control and safety valves.

Brief descriptions of the various systems of ITL are given below. 4.4.5.2 Primary heat transport system Overall scaling parameters

The prototype loop is scaled down by a factor of 452. Fig. 3 shows a 3-D view of the test facility. It consists of a full size, 54-rod cluster with full size feeder and tail pipe. The components like steam drum; downcomer and header are scaled down in volume by a factor of 452. All primary loop piping is SS 316L and is designed for 3040C and 100 bar. Overall scaling parameters are as below

• Pressure Scaling: 1:1 (70 bar) • Temperature Scaling: 1:1 (285 0 C) • Power Scaling: 1:452 • Elevation Scaling: 1: 1 (only for PHT Loop) • Volume Scaling: 1:452 • Maximum power to the test section: 3 MW

Steam drum ITL consists of a horizontal steam drum (Fig. 4) wherein steam and water are separated by gravity action. While scaling the steam drum, the important local phenomenon to be scaled is the steam-water separation by gravity, simulation of which requires the scaling of the carryover and carry under phenomenon. The following parameters are considered to be important in scaling carryover and carry under.

• Steam and water velocity above the interface. • Interface area per unit of power. • Level in the steam drum (height of interface) and height of steam space above the

interface • Froude number simulation.

The above said parameters are inter-related and generally accepted non-dimensional groups for the steam separation process are not available as of now. Hence, a dimensional analysis was carried out to obtain the relevant non-dimensional groups. Downcomer In AHWR, each steam drum is connected to the header by means of four downcomers. In ITL, a single downcomer pipe of full height and reduced diameter replaces all these 16 pipes. The downcomer diameter is arrived at by scaling down the flow area in the volume scale ratio, which will also simulate the velocities and preserve the time scale. Header & feeder pipe The header in the AHWR is connected to all the sixteen downcomers and 452 parallel feeder lines emerge from the header with 113 tail pipe lines connected to each steam drum. The dimensions of the header are obtained by volume scaling. Feeder in the ITL is a pipe with reduced length due to space constraints in the existing building. To simulate the volume, the pipe diameter has been increased. However, the desired flow resistance can be simulated by placing an orifice plate in the flow path. Core simulator

A directly heated 54-rod full-scale fuel cluster simulator with central water rod for ECC injection is installed in ITL. The fuel cluster simulator, as fabricated, is shown in Fig. 5. The power of the test section during integral experiments will be 2.03 MW, which corresponds to 1/452 of the reactor nominal power (920 MWth). A cross sectional view of Fuel Cluster Simulator and assembly drawing is depicted in Fig. 6.

4.4.5.3 Steam and feed water system The steam supply system to the turbine up to the governor valve will be simulated. All the controls on the steam pressure including steam discharge, safety relief valve, dump valves and CIES valve will be simulated as per the plant. The steam produced in the facility will be condensed using the Jet condenser and pool boiling cooler. The condensed steam will be supplied as feed water at the required temperature (130 0C) to the steam drum. Jet Condenser The saturated steam coming from the steam drum is at 285 0C temperature and 70 bar pressure. This steam is to be condensed and recirculated to the steam drum at 130 0C under nominal power operation. For this purpose the jet condenser is used, in which the steam is condensed by a jet of cold water. The cold water for this jet is supplied by the pool boiling cooler. This spray of cold-water jet is maintained with the help of secondary feed water pump. Pool boiling cooler The pool boiling cooler (PBC) is included in the circuit as the ultimate heat sink and to provide cold water at the required temperature to the jet condenser. Part of the feed pump discharge at temperature of 130 0C is sent through the pool boiling cooler, which in turn cools the water to be sent to the jet condenser. The cooling water for the pool boiling cooler is supplied from service water system, which is normally at room temperature. The service water boils and the steam is let out to the atmosphere. The level of water in the PBC is maintained manually. However, a level alarm is provided to alert the operator to commence the make-up. 4.4.5.4 Emergency Core Cooling System (ECCS) The ECCS essentially consists of the advanced accumulator and the GDCS. Advanced accumulator The advanced accumulator is designed for core cooling in a passive manner in the event of LOCA in AHWR. The accumulator is provided with a passive fluidic flow control device known as “vortex chamber”. It allows a large flow rate at the initial stage of the LOCA. As the time elapses the requirement for the large flow rate comes down and the fluidic device allows only a smaller flow rate. As the primary system is scaled down in volume by a factor of 452, the accumulator also is volume scaled in the same ratio (a photograph of advanced accumulator is shown in Fig.7).

The initial pressure of nitrogen gas is maintained same as in the prototype plant. The fluidic flow control device is geometrically simulated. The proportion of water volume above and below the standpipe is kept same as in the plant. Pressure drop in the line from accumulator to the PHT system injection point is also simulated. Gravity Driven Water Pool (GDWP) The gravity driven cooling system (GDCS) in the AHWR is designed to provide long-term core cooling for 72 hours after a reactor trip following the occurrence of LOCA. The operation of the GDCS depends solely on the system pressure. In the event of any depressurisation in the PHT system, water enters the core under gravity. The system mainly consists of a water pool known as GDWP and associated piping and components. The total inventory of GDWP is 6000 m3 and level of water in the pool is 5.1 m. The GDWP in the ITL is scaled down by a factor 452 and that gives its volume to be 14.705 m3. Due to elevation constraints, in the existing building it is not possible to locate the GDWP at the same elevation as in the prototype. Hence, the GDWP is pressurized by nitrogen gas to compensate for the lower relative elevation in ITL. Also due to space constraints the GDWP volume is sufficient to cater to less than 24 hours instead of 72 hours. 4.4.5.5 Isolation Condenser System

In the prototype AHWR, during normal shutdown the main steam line valve to turbine is closed and decay heat is removed in four isolation condensers. Each isolation condenser is coupled to one steam drum and each steam drum is connected to 113 channels. For simulating IC water pool a separate water tank is installed at an elevation of 44.05m that is on the terrace of the building, which houses the facility. 4.4.5.6 Break simulation system To simulate LOCA in the event of a break in the PHT system, a break simulation system is employed. The system consists of a quick opening valve (QOV), break simulating orifice and a tank to collect the break flow (known as break flow storage tank (BF-ST). To avoid the effect of backpressure on the critical flow rate at the break location, the pipeline after the break simulating orifice is made sufficiently large. Also, the distance between the QOV and the break-simulating orifice is kept minimum possible. Break is simulated for header and feeder in ITL. 4.4.6 EXTENT OF SIMULATION OF AHWR BY ITL If Eqs. 15 to 18 are satisfied then simulated facility will have the same temporal variation of pressure, temperature and velocity in dimensional form as in the prototype plant, whereas mass flow rate and volumetric flow rates will be simulated in non-dimensional form. Looking at those equations, it is evident that no scaled facility can preserve all the above parameters, as some distortions are inherent in the scaling philosophy itself. The frictional term i.e. (ζ f) cannot be simulated for a down sized pipe. If pipe length is maintained same in the model as in the prototype then pipe area in the model has to be reduced by a factor of S as compared to

prototype to achieve volume scaling (Vp/Vm = S). Due to this frictional effects ((ζ f)p/(ζ f)m = S) cannot be simulated. The scale distortions due to frictional effects cannot be compensated in the model except in the cases where the local pressure losses due to orifice, valves etc. are dominant over pipe frictional losses. In cases, where pipe friction is dominant, (ζ f)p/(ζ f)m is given by

( )( )

( )( )

( )( )

( )( )

( )( )

( )( )

( )( )

( )( )

bb b 1 bw w w c cp p p p p p pm

w w w c cm m m m m p m m

f f A A DDf f A D A D

ζ ζ ζ

ζ ζ ζ

−⎡ ⎤⎡ ⎤ ⎡ ⎤ ⎡ ⎤= = =⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥

⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦ ⎣ ⎦⎣ ⎦ (19)

Where

b

afRe

=

Assuming Blasius equation to be valid for friction factor f and using equation (17), equation (19) is written as ( )( )

w p 0.5(1 b) 0.625

w m

fS S

f

ζ

ζ+= = (20)

This is not equal to S as required by Eq. 18. The surface heat losses cannot be simulated for down sized pipe as surface area per unit volume changes with the pipe size. Besides this, the distortions also come in the model because of space constraints and appropriate scaled down pipe sizes are not commercially available. However, power to volume scaling philosophy is recommended for pressure tube type of reactor as diameters of most of the pipes are retained in the model as in the prototype. Therefore it is important to investigate the adequacy of power to volume scaling by comparing the prototype and model behaviour using the same analytical tool. A comparison of AHWR and ITL steady state behaviour is shown in Fig. 8. A similar comparison for system stability is shown in Fig. 9. 4.4.7 EXPERIMENTS CONDUCTED IN ITL 4.4.7.1 Steady State Performance

Steady state experiments were conducted in ITL to study the natural circulation behaviour. The Loop was operated at different pressures and channel inlet subcooling over a wide range of powers. Figure 10 and 11 show the comparison of steady state loop flow rates predicted and experimentally measured. A comparison of test data against some of the analytical tools is also shown. TINFLO-S [8] is a linear stability analysis code, developed in-house. The code is based on homogeneous (EVET) approach. A comparison with generalized flow correlation developed by Gartia et al. [9] has also been made. Generalized correlations for steady state flow prediction in two-phase natural circulation loops consider the steady state governing equations for homogeneous equilibrium model. These equations are then solved to obtain the dimensionless flow rate as a function of a modified Grashof number and a geometric number. Figure 12 shows a comparison of the measured and predicted flows in dimensionless form.

4.4.7.2 Start-up Experiments Reactor start-up is a procedure in which the pressure and temperature of main heat transport system are raised from initial ambient conditions to nominal operating pressure and temperature. In the proposed start-up procedure, reactor is made critical at cold condition and system is kept in single phase by external pressurization with steam till the system temperature and pressure reaches to 285 oC and 70 bar respectively. During this period the system pressure is varied in such a way that cold pressurization is avoided. Also, the reactor power is maintained low enough to prevent fast and uneven structural heating, but at the same time high enough to have a reasonably fast start-up. The external pressurization with steam keeps the system in single phase thereby suppressing the low-pressure low power two-phase instabilities. This phase of start-up is termed as cold start-up. Experiments have been conducted in ITL to study the start-up and stability of two-phase natural circulation at low pressures. Figure 13 (a) shows the variation of thermal hydraulic parameters during a typical cold start-up experiment. During the test, the test section power was maintained at 2% of full power. Steam from external boiler was supplied to maintain the system pressure at 10 bar until PHT fluid reaches the saturation temperature. Initially at the start of transient, single-phase natural circulation oscillations are observed; these can be attributed to the inertia effect of flow. As system temperature reaches the saturation, boiling is initiated. The boiling inception is accompanied by the flow oscillations due to flashing. Effect of initial steam drum pressure is also studied during start-up and it is found that external pressurization reduces the two-phase oscillations at the transition to two-phase natural circulation [Fig. 13 (b)]. The flow oscillations continue until the system pressure reaches sufficiently high (about 30 bar). At high pressures the oscillations are suppressed. The transient was simulated using numerical code RELAP5/MOD3.2 [Fig. 14 (a)]. It was observed that the boiling inception takes place close to the steam drum, indicating the flashing effects that are characteristics of the system having tall risers. The results predicted by the model were fairly acceptable in qualitative manner. Analysis was also carried out considering the heat losses from piping [Fig. 14 (b)]. Fig. 15 shows the start-up data plotted on stability map. The stable and unstable points are successfully characterised by the theoretical stability map. 4.4.8.0 CONCLUSION A brief description of the integral test loop, its scaling philosophy and the various proposed experiments in the facility have been presented. The steady state and stability performance has been predicted for the facility as well as the prototype reactor. The steady state data generated in the loop are in good agreement with the predictions of the in-house developed generalized correlation as well as computer codes like RELAP5/MOD3.2. The stability data obtained also showed good agreement with the predictions of the in-house developed stability code. In addition, test data on the proposed start-up procedure has been presented along with simulations using the RELAP5/MOD3.2 code. The code is found to reproduce all the essential trends of the observed transients.

Nomenclature: A -area (m2) Subscripts D -diameter (m) c -cross section f -friction factor h -heated g -acceleration due to gravity (m/s2) m -model h -enthalpy (kJ/kg) 0 -reference value l -length (m) p -prototype p -pressure (bar or Pa) s -surface q -heat flux (W/m2) w -wetted Q -volume flow rate (m3/s) S -scale ratio t -time (s) Superscripts v -velocity (m/s) + -non-dimensional V -volume (m3) W -mass flow rate (kg/s) z -elevation (m) Greek symbols ρ -density (kg/m3) ζ -perimeter (m) φ -power factor τ -shear stress (Pa)

References: (1) R. K. Sinha and Anil Kakodkar Advanced Heavy Water Reactor, INS NEWS Vol. 15

Nos. 2-4 (2002), Vol.16, No.1 (2003). (2) Ishii, M., et al., “Scientific design of Purdue University Multi-Dimensional Integral Test

Assembly (PUMA) for GE SBWR”, NUREG/CR-6309, PU-NE 94/1. (3) Loomis, G.G., Soda, K., “Results of the semiscale Mod-2A natural circulation experiments”,

NUREG/CR-2335, (1982). (4) Multi-Loop Integral Test Facility (MIST) Specifications, NRC-04-83-168, (1983). (5) D’Auria, F., et al., “Scaling of natural circulation in PWR systems”, Nuclear Engineering and

Design, 132, 187-205 (1991). (6) Zuber, N., “Problems in modeling small break LOCA”, NUREG-0724, (1980). (7) Rao, G.S.S.P., Vijayan, P.K., Shinde, U.A., Saha,D., Sinha, A. and Sarkar, P.S., Experimental and

theoretical investigations in a two phase natural circulation loop, ISME-2003, XIII National Conference of Indian Society of Mechanical Engineers, December 30-31, 2003, IIT Roorkee.

(8) Nayak, A.K., Vijayan, P.K., Saha, D., Venkat Raj, V., Aritomi, M., 1998, “Linear Analysis of thermodynamic instabilities of the advanced heavy water reactor (AHWR), J. Nucl. Sci. Tech. 35, 768-778.

(9) Gartia, M. R., Vijayan, P.K., Pilkhwal, D.S., “A generalized flow correlation for two-phase natural circulation loops”, Nuclear Engineering and Design, 236, 1800-1809 (2006).

Fig. 1. Photograph of ITL building

IC INLET HEADER

IC2 IC1 IC TUBES

IC OUTLET HEADER

GDWP

STEAMDRUM

VALVEGOVERNOR

TURBINESTEAM TO

RETURN VALVEIC CONDENSATE

FEED WATER

RIH

DOWN COMERS

INLET FEEDERS CORE

COOLANTCHANNELS

TAIL PIPES

Fig. 2: MHT and Isolation Condenser systems of AHWR

ICSD

GD

WP

FCS

BFST

QOVHEADER

PBC

SFP

JC SB

AA

ICSD

GD

WP

FCS

BFST

QOVHEADER

PBC

SFP

JC SB

AA

Fig. 3: 3-D view of ITL

Fig. 4: Steam Drum as installed in ITL

FCS : Fuel channel simulator BFST : Break flow storage tank QOV : Quick opening valve GDWP : Gravity driven water poolSFP : Secondary feed pump JC : Jet condenser SB : Star-up boiler AA : Advanced accumulator SD : Steam drum IC : Isolation condenser

Fig. 5: Photograph of Fuel cluster simulator as installed in the facility

SHELL ASSEMBLY

HEATED SECTION

ELECTRICAL BUSBAR

OUTLET

ELECTRICAL BUSBAR

INLET

SECTION A-A

A A

SHELL ASSEMBLY

HEATED SECTION

ELECTRICAL BUSBAR

OUTLET

ELECTRICAL BUSBAR

INLET

SECTION A-A

A A

Fig. 6: Fuel Rod Cluster simulator set-up

Fig. 7: Advanced Accumulator of ITL

0.0 0.5 1.0 1.5 2.0 2.52.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

AHWR ITL

Sin

gle

chan

nel f

low

rate

- kg

/s

Average channel Power - MW Fig. 8: Comparison of predicted steady state flow rates using TINFLO

0 5 10 15 20 25 300

1000

2000

3000

4000

5000

0

30

60

90

120

150

180

AHWR ITL

Pressure = 7 MPa

Unstable

Unstable

Stable

Cha

nnel

pow

er (k

W)

Subcooling (0C)

Type-I instability threshold

Cha

nnel

Pow

er (%

)

Fig. 9: Stability Map predicted by TINFLO-S for AHWR and ITL

0 100 200 300 400 5000

1

2

3

4

5

6

Pressure=50 bar

Tsub

=4-5oC

Mas

s Fl

ow R

ate

(kg/

s)

Power (kW)

Experimental Data RELAP5/Mod 3.2 Generalised Flow Correlation

(with Blassius model for friction factor) Generalised Flow Correlation

(With Colebrook model for friction factor)

Fig. 10: Comparison of experimental steady state loop flow rate at 50 bar.

0 50 100 150 200 250 3000

1

2

3

Pressure = 70 bar

Tsub

= 10-20 oC

Mas

s Fl

ow R

ate

(kg/

s)

Power (kW)

Experimental data Generalised Flow Correlation TINFLO-S RELAP5 / Mod 3.2

Fig.11: Comparison of experimental steady state loop flow rate at 70 bar

107 108 109 1010 1011 1012 1013 1014 1015102

103

104

105

106

Re ss

Grm / NG

Theoretical 1.4*C(Grm/NG)r

0.6*C(Grm/NG)r

ITL experimental data

Fig.12: Comparison of experimental and predicted flow in dimensionless form

0 20000 40000 60000 800000

25

50

75

100

Flow

Power Pressu

reFCS inlet temperature

Feed

er F

low

- lp

mIn

let T

empe

ratu

re -

OC

Pow

er -

kWP

ress

ure

- bar

ITL Cold Start-up at 2% FP and 10 bar

Time - s

0

75

150

225

300

0 15000 30000 45000 60000 75000 900000

25

50

75

100

Pressure

Flow

Inlet temperature

Power

Cold start-up at 2% FP with stage-wise presurizationwith boiling inception at 15 bar

Inle

t Tem

pera

ture

- OC

Flow

rate

- lp

m

Pow

er -

kW

SD P

ress

ure

- bar

Time - s

0

75

150

225

300

Fig 13 (a): ITL Start-up at 10 Fig. 13 (b): ITL Start-up at 15 bar – Experimental Data bar – Experimental Data

0 5000 10000 15000 20000 25000 300000

50

100

150

200

250

300

Feed

er M

ass

flow

rate

(kg/

s)

Tem

pera

ture

(o C)/

Pres

sure

(bar

)

Time (s)

0

1

2

3

4

ITL Cold start-up at 2% FP and 10 bar

0 20000 40000 60000 800000

25

50

75

100In

let t

empe

ratu

re (o C

)Fe

eder

flow

(lpm

)

Pre

sssu

re (b

ar)

time (s)

0

50

100

150

200

250

300

Flow

Pressur

eInlet temperature

Post-test simulation accounting for heat losses

ITL Cold start-up at 2% FP and 10 bar

0 5000 10000 15000 20000 25000 300000

50

100

150

200

250

300

Feed

er M

ass

flow

rate

(kg/

s)

Tem

pera

ture

(o C)/

Pres

sure

(bar

)

Time (s)

0

1

2

3

4

ITL Cold start-up at 2% FP and 10 bar

0 5000 10000 15000 20000 25000 300000

50

100

150

200

250

300

Feed

er M

ass

flow

rate

(kg/

s)

Tem

pera

ture

(o C)/

Pres

sure

(bar

)

Time (s)

0

1

2

3

4

ITL Cold start-up at 2% FP and 10 bar

0 20000 40000 60000 800000

25

50

75

100In

let t

empe

ratu

re (o C

)Fe

eder

flow

(lpm

)

Pre

sssu

re (b

ar)

time (s)

0

50

100

150

200

250

300

Flow

Pressur

eInlet temperature

Post-test simulation accounting for heat losses

ITL Cold start-up at 2% FP and 10 bar

0 20000 40000 60000 800000

25

50

75

100In

let t

empe

ratu

re (o C

)Fe

eder

flow

(lpm

)

Pre

sssu

re (b

ar)

time (s)

0

50

100

150

200

250

300

Flow

Pressur

eInlet temperature

Post-test simulation accounting for heat losses

ITL Cold start-up at 2% FP and 10 bar

Fig. 14 (a): Simulation of ITL start-up at 10 Fig. 14 (b): Simulation of ITL start-up at 10 bar, using RELAP5/M3.2 without bar, using RELAP5/M3.2 with pipe heat loss pipe heat loss

020

4060

80

0

300

600

900

1200

510

1520

25

Experimental data

Unstable

Stable

Stable

Unstable

Subcooling (K)

Pow

er (k

W)

Pressure (bar)

Fig. 15: Experimental data plotted on theoretically obtained stability map