EXPERIMENTAL RESULTS ON A TWO-PHASE CLOSED LOOP SYSTEM

Transcription

1 XXV Congresso Nazionale UIT sulla Trasmissione del Calore Trieste, Giugno 2007 EXPERIMENTAL RESULTS ON A TWO-PHASE CLOSED LOOP SYSTEM C.Lombardi, M.Ricotti, L.Santini, B.Zito Nuclear Engineering Department, Politecnico di Milano, via Ponzio 34/3, Milano marco.ricotti@polimi.it ABSTRACT In this work some experimental results obtained on a two-phase natural circulation loop are presented. The loop simulate the passive safety system of an innovative nuclear reactor (IRIS reactor) that is under study by an international consortium. The circuit is an electrically heated loop with a water cooled heat sink (a condenser submerged into a pool) built and operated at SIET labs in Piacenza (Italy). The main scope of the work is to investigate the effect on loop performances of several parameters variations. The impact of the following parameters has been studied: loop water content, test section electrical power, test section inlet throttling, noncondensables presence and heat losses compensation. Circuit operative pressure have shown to be mainly dependent on pool condenser rejected power and on the so called Filling ratio, i.e. the ratio between the total mass in the closed circuit and the total mass of cold water that could be stored in the loop. The obtained results underlined the complex physics involved in a natural circulation loop, showing that such a system is prone to several type of instabilities involving oscillations of the main parameters such as pressure, flow rate and temperature. 1.1 INTRODUCTION The need to evacuate the decay heat from the core of a nuclear power reactor has brought to the introduction of costly and critical components in nuclear industry. The accident of Three Miles Island-2 in 1979 has shown the importance of core heat removal both in shutdown and post accident operation mode. Today almost all new water cooled nuclear power reactors are designed to remove the decay heat by natural circulation in case of complete loss of pumping power. For example the Advanced Light Water Reactors AP600, AP1000, SBWR, ESBWR, SWR1000 and IRIS are studied to operate in a natural circulation mode for removing such heat. In AP600 (figure 1-a) and AP1000 the primary system is connected with a heat exchanger (the heat sink of the natural circulation loop) immersed in the in-containment refueling water storage tank (IRWST). In SWR-1000 [1] the emergency condensers (figure 1-b) are located in the core reflooding pool: the bundles are permanently connected to the reactor pressure vessel (RPV) and start their operation when water inside the vessel goes below a certain level. Each emergency condenser consists of a steam line leading from the pressure vessel nozzle to a heat exchanger tube bundle. Given the normal water level inside the RPV, there prevails a stratified condition inside the emergency condenser (fig. 1-b left side). The upper part of the steam line is filled with steam, while the lower part is filled with water. When, during accident conditions, water level in the emergency condenser drops by more then 0,5 m, steam enters the heat exchanger bundle. The steam then condenses inside the heat exchangers tubes and the resultant condensate flows via the refolding line back into the RPV [2]. b Figure 1-Schematic of AP600 (a) and SWR-1000 (b) passive emergency devices a

2 IRIS reactor is an innovative medium size PWR of 335 MW e, whose main safety feature is a natural circulation loop, EHRS (Emergency Heat Removal System), that will transfer core decay heat and sensible heat from the reactor coolant to the environment during transients, accidents or whenever the normal heat removal paths are lost. IRIS reactor is composed of eight steam generators each one of 125 MW th of nominal power, every two steam generators there is a dedicated EHRS subsystem. An EHRS subsystem is a closed circuit with two parallel steam generators, an hot leg (the riser of the circuit), the heat sink composed by an heat exchanger bundle made up of 100 tubes of 59 mm of inner diameter and 10 meters long submerged in a suppression pool and the cold leg that close the circuit by bringing cold condensed water to the steam generators (schematic principle in figure 2). During plant normal operation valves V1 and V2 are open and valve V3 is closed, preheated water coming from the regenerator line is pushed into the steam generators where is evaporated, slightly superheated and sent to the high pressure turbine. If a reactor trip occurs the core decay heat will normally be removed by the steam generators with feed water supplied by the start up feed water system and the steam is directed to the condenser via the steam dump valves. In case of malfunction of the start up feed water system, the EHRS is available to remove the decay heat following the closure of V1 and V2 valves and the opening of V3 valve. EHRS hot leg EHRS cold leg Reactor Heat sink V3 Steam Generator (SG) V1 V2 To turbine From pre-heater line Figure 2-Schematic of IRIS Emergency Heat Removal System By properly timing the closure of the V1 and V2 valves it is possible to select the water inventory of the natural circulation loop, to which is related the so-called filling ratio defined as the ratio between the total mass in the closed circuit and the total mass of cold water that could be stored in the loop FR = M loop / M ), thus changing its ( max_ cold thermohydraulic behavior in terms of operating pressure, flow rates, qualities, stability and extracted power. In fact, some preliminary calculations performed with RELAP-5 code [3] have shown the importance of selecting the proper filling ratio of the system to obtain the desired power level extracted from the reactor and the desired degree of stability of the system (in terms of system main parameters oscillations). In figures 3 some results of RELAP-5 calculations have been reported for IRIS EHRS indicating the effect of the filling ratio on extracted power of the system and operating steady state pressure. The primary heating fluid is schematized as a constant inlet temperature, constant flow rate pressurized water. The heat sink is a pool of water at atmospheric pressure and at a constant temperature of 293 K. Figure 3-RELAP-5 preliminary calculations on IRIS EHRS: effect of filling ratio on extracted power The physical role of the filling ratio on the behavior of the system is not yet deeply understood. It seems that an increase in the filling ratio, i.e. in the loop water content, has the main effect of rising the length of the single phase subcooled zone in the condenser tubes thus reducing its overall heat transfer coefficient. Another qualitative effect of an increasing filling ratio is the reduction of the available space for the expansion in the circuit of the generated steam. The combination of the two mentioned effects can be the cause of the experienced increase in circuit pressure following an increase in the filling ratio (see next). Up to the authors knowledge in the majority of the experimental investigations, the natural circulating system desired pressure have been imposed via a pressurizer [4] and only in some few cases by adjusting the flow rate of the heat sink [5]. The experimental campaign described in the following has been carried out with an electrically heated sliding pressure natural circulation loop and had the main goal of exploring the effect on the loop of variations of the following parameters: filling ratio, power given to the flow, heated test section inlet orificing, non-condensable gases presence and heat losses compensation. 1.2 ADVANTAGES AND DRAWBACKS OF NATURAL CIRCULATION Natural circulation is based on elements that always exists in a nuclear reactor: the heat source, the gravity and the tubes that compose the circuit. The apparent simplicity of the physical principle, in reality covers the complexity of the phenomena that mutually interact in a natural circulation loop: operating pressure, flow rate, qualities, heat transfer coefficients and loop oscillations are all linked together in a complex manner. The physics of such systems have been extensively studied both with theoretical [6], [7] and with experimental investigations [8]. The main advantages of natural circulation are related to the elimination of circulating pumps with consequences on capital, operating and maintenance costs and elimination of possible failures of rotating machinery or electric power supplying systems. This big advantage under the profile of the safety of the system is in part balanced by some challenges brought by natural circulation: -Low driving force: the driving force of a natural circulating loop is related to the height of the system which value is strongly affected by economic considerations. In

3 addition tall and slender structures could rise specific problems of seismic stability of the system. The low driving force bring together the necessity of large pipe diameters for reducing frictional pressure drops. In a natural circulating loop the typical flow rates are quite small resulting in small heat transfer coefficients (especially in the single phase region) and thus the need to have big exchanging surfaces in fluid heated systems. Moreover, the slow fluid speed affects circuit time constant, which in its turn might influence the dynamic behavior of the loop. -Problems in the start-up procedure: experiences with natural circulation loops have shown, in certain circumstances, the possibility of difficulties in priming the circulation. -Instabilities: it is well known that two-phase systems are often prone to termohydraulics instabilities in terms of oscillations of the main parameters (pressure, flow rate, temperatures and so on). Such instabilities, particularly in natural circulation loops, could rise problems of mechanical fatigue for long times of operation. -Limitations on Critical Heat Flux : it is well known from many years of experience on boilers [9] that a two-phase pulsating flow rate has the effect of reducing the CHF, particularly of the dryout type, of a boiling equipment. This aspect has to be taken into account if large amplitude oscillations are expected to rise in the natural circulation loop. 2 EXPERIMENTAL FACILITY The test section at SIET labs in Piacenza is a closed loop electrically heated built with the scope of simulating the physical behavior of the Passive Emergency Heat Removal System of IRIS reactor. The loop is briefly composed of a heat source, a riser, a heat sink and a downcomer (figure 4). The heat source is an electrically heated steam generator previously built for the study of two-phase pressure drops [10] and thermal crisis inside a coiled tube. The total length of the steam generator is 32 meters, corresponding to one full length and full height tube of IRIS steam generator, but only 24 meters were powered with DC current. The tube of the steam generator is entirely insulated with rock wool and the small thermal dispersions were accurately characterized as a function of the temperature difference between external tube wall and ambient air. The riser is a 21,3 meters long AISI 316 stainless steel tube with an inner diameter of mm and an outer diameter of mm. Riser and downcomer diameters have not been scaled with respect to IRIS EHRS riser and downcomer expected pressure drops. The pipes were accurately insulated with rock wool. Nevertheless in a part of the experimental runs (nearly one half), the small thermal losses along riser and downcomer were compensated with an electrical wire coiled along the tubes whose power can be regulated during operation. The pool condenser (figure 5) is submerged into a 250 liters pool. The tube of the condenser is slightly inclined (3 ) to avoid water draining during condensation. A metallic slab is placed few centimeters under the vapor escaping duct of the pool to reduce the presence of liquid droplets in the exit stream. The evaporating water in the pool is continuously replaced via a submerged drilled tube placed in the bottom of the pool. The tube in the bottom of the pool is connected to a tank whose water level is maintained to a constant value thanks to a floating device. Extracted steam condenser Loop Filling Ratio Controlling Valve V3 Riser throttling valve Vapour to atmosphere V4 Condenser tube. Din=59 mm Dout=73mm L=1m Inclination=3 Balance Riser Din=20,93mm Dout=26,27mm Total Coiled Steam Generator Din=12,53mm Dout=17,15mm Toltal Lenght of the tube=32m Heated Length=24m Total Height=7,9m V1 Test section inlet throttling valve V2 Flowmeter Orifice DC Electric Heating Loop Height =19,9m Downcomer Din=20,93mm Dout=26,27mm Total Lenght=31,75m Figure 4-Schematic of IRIS Emergency Heat Removal System Facility The facility in Piacenza operate with one only SG tube, 1m long with 59 and 73 mm of inner and outer diameters. The height of the experimental loop is the same of the real system (20 m). Figure 5- The pool condenser The measured quantities in the loop are flow rates, pressures (absolute and differential), temperatures and powers. The circuit flow rate has been measured by a calibrated orifice of 5 mm placed at steam generator inlet and instrumented with a differential pressure transducer calibrated at SIET labs (SIT certified) with an estimated maximum uncertainty of 2%. The circuit absolute pressure is measured at steam generator inlet via an absolute pressure transducer calibrated at SIET labs and with a maximum uncertainty on read value of 0,1%. Differential pressure transducer are placed across the throttling valves and along the downcomer with the scope of evaluating the possible presence of mixture at condenser tube outlet. Fluid temperature measurement is obtained with K-class thermocouples (calibrated in SIET labs and with a maximum error at 100 C of 0.4 C). Fluid temperatures are measured at

4 steam generator inlet and outlet headers, at condenser tube inlet and outlet and inside pool condenser. The electrical power is measured via a volt-amperometric digital instrument with a relative uncertainty guaranteed of 2.5%. 3 RESULTS The test matrix is composed of 45 runs including variations of the following parameters: filling ratio (0.18, 0.31, 0.49, 0.61, 0.79), power given to the flow, heated test section inlet orificing, non-condensable gases content and heat losses compensation. In the following the effect of the previously mentioned parameters variation on loop behavior is presented. 3.1 EFFECT OF TEST SECTION POWER AND FILLING RATIO ON LOOP PARAMETERS The most important effect of rising power on system behaviour at constant filling ratio is the increase of working pressure. A semi-quantitative physical explanation is related to the working principle of the pool condenser. The general formula describing the behaviour of the condenser, i.e. the heat sink of the circuit, is: Pressure [bar] 90,0 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 System Pressure as a function of pool condenser rejected power and filling ratio 0,61FR 0,49FR 0,79FR 0,31FR 0,18FR without compensation compensated Filling ratio 0,0 9,0 14,0 19,0 24,0 29,0 34,0 39,0 Pool Rejected Power [kw] Figure 6- Effect of filling ratio and pool condenser rejected power on system pressure In order to summarize the experienced loop behaviour (fig. 6 data) a simple correlation between system pressure and pool condenser rejected power has been proposed valid in the explored range of filling ratios (between 0,18 and 0.79): p T 0.7 sat W h gl 0,86 2 = 136FR 78FR + 49 (2) W = F U S (1) T ml With F=1 being the cooling fluid of the pool at constant temperature (100 C). In the condenser of our circulation loop, when the electrical power is increased, neglecting the thermal dispersion, the only way for transferring the thermal power to pool water is by increasing U or T (or both). The possibility of increasing ml U is very small due to the strong thermal resistance of tube wall and to the weak dependence of two phase flow in-tube condensing heat transfer coefficient with flow rate. So the only possibility for the system is by increasing the temperature difference between the two streams, but, as previously said, pool water temperature is fixed by atmospheric pressure and thus, loop pressure must increase. Filling ratio effect is physically more difficult to explain, but, as previously said in the introduction, it has two main effects: rising single phase zone in pool condenser tube, and reducing steam volume. The rising of single phase zone, i.e. the increased subcooling at condenser outlet, at fixed circuit pressure, reduces the mean temperature drop available for exchanging thermal power. But the thermal power to be exchanged is fixed by the electrical heating, and thus pressure must increase. In Figure 6 we collected all the runs done on the loop for various powers and filling ratio with and without riser and downcomer heat losses compensation. The controlling parameter on system operative pressure is resulted pool condenser rejected power whichever are electrical power and thermal dispersions (that mainly depend on system pressure). This conclusion is also included in equation (1). In particular the strong effect of filling ratio on operating pressure is evident especially at high filling ratios (when the circuit is going full of water). Where p is the pressure expressed in kpa, h gl is the heat of vaporization and is expressed in kj/kg, T sat is the saturation pressure expressed in K and W is pool condenser rejected power, expressed in kw. Formula (2) allows to predict system pressure with a mean error of 1% and a RMS of 7,3%. Nevertheless its predictive capabilities have to be considered strictly limited to the specific characteristics of the loop investigated in this work. The effect of filling ratio and electrical power on pool condenser inlet quality is shown in the next figure (figure 7) Pool condenser inlet quality 0,35 0,30 0,25 0,20 0,15 0,10 0,05 Effect of filling ratio and electrical power on pool condenser inlet quality (without heat losses compensation) 0,00 0,4 0,5 0,6 0,7 0,8 0,9 Filling Ratio 33kW 23kW Figure 7- Effect of filling ratio and electrical power on pool condenser inlet quality (without heat losses compensation) As it is possible to observe filling ratio always decreases pool condenser inlet quality (and thus steam generator outlet quality) and power always increases it. The effect of filling ratio and electrical power on pool condenser outlet subcooling is shown in the next figure (figure 8)

5 Cond. outlet subcooling [ C] Effect of filling ratio and electrical power on pool condenser outlet subcooling (without heat losses compensation) 33kW 23kW 0,4 0,5 0,6 0,7 0,8 Filling Ratio Figure 8- Effect of filling ratio and electrical power on pool condenser outlet subcooling (without heat losses compensation) As it is possible to observe filling ratio always increases pool condenser outlet subcooling, as power does. The data at 0,18 and 0,31 of FR have been omitted in the last figure due to the presence of mixture in downcomer tube and to the impossibility of measuring system flow rate. The effect of filling ratio and electrical power on system flow rate is reported in the next figure (figure 9) Flowrate [kg/s] 0,10 0,09 0,08 0,07 0,06 0,05 0,04 Effect of filling ratio and electrical power on loop flow rate (without heat losses compensation) 33kW 23kW 0,03 0,4 0,5 0,6 0,7 0,8 0,9 FR Figure 9- Effect of filling ratio and electrical power on system flow rate (without heat losses compensation) As figure 9 shows power and filling ratio effects on flow rate are not monotonic. The effect is surely due to the complex non linear relation between flowrate and loop parameters in two phase natural circulation, but a detailed and satisfactory explanation is still missing. 3.2 EFFECT OF NONCONDENSABLE GASES ON LOOP OPERATING PRESSURE In order to investigate the effect of air presence on system overall behaviour some runs were done after the introduction of a certain amount of air into the system. Heat transfer coefficient in condensation are reduced if an incondensable gas is added into the condensing fluid [11]. In fact, non condensables collect at the interface between liquid film and condensing vapour, thus reducing vapour partial pressure and its temperature (that drives the thermal exchange). The degradation in convective coefficients is more pronounced if system pressure is reduced and is more pronounced in stagnant mixtures then in forced convective condensation. A typical parameter introduced in order to quantify the amount of noncondensable gases is the ratio between the mass of air and the mass of steam ( M / M ). The total quantity of steam, that is a function of filling ratio and system pressure, is obtained with the simplifying hypothesis that air content doesn t influence steam presence and that the circuit is schematised as a container with a defined portion occupied by liquid and the remaining portion by steam in saturated conditions. Thus: M V v FR M t l 0 v = vv vl The ratio between air content and steam content is thus a function of system pressure and reduces by increasing the pressure. That ratio is calculated assuming that all the steam inside the circuit is mixed with the total amount of air. In the following figure (figure 10) the effect of incondensables presence on system pressure is reported. System Pressure [bar] air (3) Incondensables concentration effect on system pressure (data normalized on pool condenser rejected power) Air mass/steam mass % [kg_air/kg_steam*100] v 23 kw 33 kw Figure 10- Incondensables concentrations effect on system pressure Air masses with respect to steam masses covered a range from 0% to 20%. The highest values must be considered non realistic in real system operation and were achieved only with the scope of emphasizing noncondensables content effect on system operation. The degradation of heat transfer coefficients at pool condenser tube is evidenced by system pressure increase, nevertheless the effect is to be considered small, being a 3% of pressure increase every 1% of air content increase the rate observed. 3.3 EFFECT OF STEAM GENERATOR INLET THROTTLING ON LOOP PARAMETERS IRIS real steam generator tubes will operate with a certain degree of inlet throttling. This is due to the necessity of avoiding two phase instabilities between the hundreds of parallel tubes of each steam generator and between the eight steam generators pods. The degree of throttling of each tube will derive from the hydraulic impedance of each tube row. The necessity of obtaining uniform steam conditions at SG outlet headers will impose the mentioned adjustment of tubes hydraulics impedance by ad hoc inlet throttling. The final ferrules dimensions, as AGRs experience has shown, will be decided after extensive experiences on a facility with several parallel tubes connected. In order to quantify tubes inlet throttling effect on natural circulating loop performances, twelve runs of the loop were performed with four different steam generator inlet throttling

6 (K o =30, K min =785, K med =1286, K max =1286), three steam generator power levels (23kW, 33kW, ) and one single filling ratio (0,49). The first value of throttling (30) corresponds to the throttling effect of measuring orifice, which is located upstream the throttling valve that gives most of the K value. The four degrees of throttling have been chosen with the role that the 1286 value represents the predicted IRIS steam generator ideal inlet orifice for a stable behaviour in nominal conditions. Mainly throttling affects three parameters: operating pressure, flow rate mean value and flow rate oscillations. Operating pressure have shown to be only slightly throttling affected (roughly 10 % percent in the explored K range). This behaviour is probably associated to the fact that the overall heat transfer coefficient is scarcely affected by flow rate variations. This is caused mainly by the fact that pool condenser overall heat transfer coefficient is strongly controlled by tube conductive thermal resistance. Furthermore heat transfer coefficients during condensation are scarcely affected by flow rate variations. Loop flow rate ( Γ ) have shown to be slightly sensitive to inlet throttling variations according to the following empirical correlation: Γ Γ K = med K med (4) In which Γ med is circuit flow rate in correspondence to K med in natural circulation. Next figure (Fig. 12) shows throttling effect on pressure drops oscillations amplitude. DP_orifice [kpa] 4,5 4 3,5 3 2,5 2 1,5 Throttling effect on flow rate measuring orifice pressure drops (23kW@0,49FR) No throttling (Ko=30) [18] Medium throttling (Kin=1256) [2.6] Time[s] Minimum throttling (Kin=785) [3.3] Maximum throttling (Kin=1831) [1.8] Figure 12- Inlet throttling effect on pressure drops oscillations; the number in square brackets is the ratio between hot leg over throttling device pressure drops. As it is possible to observe throttling damp down strongly the oscillations reducing the average deviation with respect to the mean value of the signal, in spite of the fact that flow rate is appreciably reduced. 3.4 OSCILLATIONS During the natural circulation flow loop operation several oscillations of the main parameters were observed. Both high and low frequencies oscillations were detected. High frequencies pressure drops excursions at flow meter orifice were always observed during the loop operation and their frequencies were of the order of ten seconds. This type of oscillation (see figure 12), detected in orifice flow meter recordings, are probably caused by loop vapour compressibility. Small perturbations in heater inlet power could induce sustained oscillations in such a manner that vapour behaves as a spring and liquid content as a mass. This oscillations have shown a frequency proportional to system pressure, in fact an increase in pressure have the main effect of making vapour spring more stiff, causing an increased frequency of the oscillator. A possible remedy for reducing Temperature [ C] kw 0,79FR with heat losses compensation Orifice pressure drops Pool condenser outlet temperature Time [s] high frequency oscillations amplitude is by throttling steam generator inlet flow (as fig. 12 shows). Low frequency oscillations, with periods of the order of 300 seconds, were detected only in the runs with the highest filling ratio, i.e. 0,79, with the lowest powers, i.e. 23 and 33 kw (figure 13). The similarity between oscillations period and particle residence time in the loop suggests that this oscillatory mode is somehow related to very slow enthalpy waves that travels with the same mean speed of the mixture in the circuit. This waves creates a strong coupling between heated test section flow rate, riser pressure drops and condenser outlet subcooling. This type of instability, already studied by other authors [13], is probably related to the peculiar shape of hot leg (steam generator+riser) static characteristic (flowrate VS pressure drops) at low steam generator exit qualities. In this condition, small variations in flow rate cause sensible variations in riser void fraction that, in turns, change loop driving force. In particular a small increase in flow rate, being the power imposed, has the main effect of reducing steam generator outlet quality that, for small values of quality, cause a strong reduction in void fraction. This reduction in void fraction increases riser gravitational pressure drops that reduce the flow rate reversing the phenomenon and thus causing oscillations. This simple explanation doesn t take into account pool condenser exchanged power oscillations that, as previously said, create enthalpy waves that influence steam generator quality variations. A complete explanation of this long period oscillations must take into account all the mentioned effects and, up to now, is still missing. 3.5 LOOP OPERATIVE BOUNDARIES Steam generator inlet temperature Figure 13- Low frequency oscillations. Loop operation have shown that not all the values of the filling ratio, always referred to cold water overall mass, are actually achievable in a natural circulation loop. In particular our experimental campaign showed that for the two smallest filling ratios experienced, i.e and 0.31, very high amplitude oscillations develop in the loop. We explained this fact observing that at 0.18 and 0.31 of filling ratios the circuit has not enough stored mass in order to completely fill the downcomer with liquid water. For this reason a two phase mixture flows at heater tube inlet, thus causing the observed Orifice Pressure Drops [KPa]

7 non stable behavior. The minimum value for the filling ratio is thus a function of circuit operative pressure, due to liquid density dependence on temperature. Moreover, an upper limit for the filling ratio must exists too. This upper limit is imposed by the constrain of not reaching circuit solidity, i.e. the water completely fill the circuit volume, during loop operation. The latter condition depends on operative pressure and condenser outlet subcooling. By knowing circuit and downcomer volumes it is possible to plot loop operative boundaries in terms of filling ratio and pressure (fig. 14). Filling Ratio 1,2 1,0 0,8 0,6 0,4 0,2 Filling Ratio boundaries for Piacenza's loop FR_max (Dtsub=80 C) Subcoolin CIRCUIT WORKING REGION FR_min 0, System Pressure [bar] FR_max (Dtsub=40 C) FR_max (Dtsub=0 C) Figure 14-Loop allowable FR boundaries as a function of pressure and condenser outlet subcooling The minimum value of the filling ratio has been calculated considering downcomer full of saturated water (at the pressure under consideration), steam generator with mixture evaporating from x=0 to x=1, riser full of saturated steam and condenser with mixture condensing from x=1 to x=0. Figure 14 shows that the minimum value of the filling ratio in order to guarantee a downcomer full of liquid water is nearly 0.37, thus confirming the explanation of the observed oscillatory behaviour at the lowest FRs. 4 CONCLUSIONS An experimental parametric study on a natural circulation sliding pressure loop has been performed. The strong effect of filling ratio on loop behaviour have been underlined. The controlling parameter on loop operating pressure, at constant filling ratio, is resulted pool condenser rejected power (whichever are the thermal dispersion along the circuit). Non condensables presence has been studied and it resulted as a second order effect on loop performance even at high concentrations of gas. Mainly two types of oscillations were detected and an hypothesis on their physical origin has been proposed. NOMENCLATURE AGR Advanced Gas Reactor Mean log. Temperature drop K T ml EHRS Emergency Heat Removal System F Heat exchanger Correction factor FR Filling ratio h gl Heat of vaporization KJ/kgK M Mass kg M loop Mass of water stored in the loop kg M max-cold Maximum circuit storable cold water kg p Pressure Pa RMS Root Mean Square Error S Exchanging area m 2 U Overall heat transfer coefficient W/m 2 K V Volume m 3 v Specific volume m 3 /kg W Thermal power kw Γ Flow rate kg/s K Orifice constant x Thermodynamic quality REFERENCES [1] A.Schaffrath, E.F.Hicken, H.Jaegers, H.M.Prasser, Experimental and analytical investigationof the operation mode of the emergency condenser of the SWR1000, Nuclear Technology, May [2] N.Aksan, Application of natural circulation systema: advantages and challenges II, in Natural Circulation in Water cooled power plants, IAEA November 2005, IAEA- TECDOC [3] A.Cammi, A.Cioncolini, IRIS Passive Emergency Heat Removal ystem-pehrs: State of the art design and RELAP model, IRIS internal report, 24 June [4] Y-J.Chung, S-H.Yang, H-C.Kim, S-Q. Zee, thermal hydraulic calculation in a passive residual removal systaem of the SMART-P plant for forced and natural convection conditions, Nuclear Engineering and Design, 232 (2004), [5] C.Y.Wu, S.B.Wang, Chin Pan, Chaotic oscillations in a low pressure two-phase natural circulation loop under low power and high inlet subcooling conditions, Nuclear Engineering and Design, 162, 1996, [6] N.E.Todreas, M.S.Kazimi, Nuclear Systems I, Hemisphere Publishing Corporation [7] S.Guanghui, J.Dounan, K.Fukuda, G.Yujun, Theoretical and experimental study on density wave oscillation of two-phase natural circulation of low equilibrium quality, Nuclear Engineering and Design, 215, 2002, [8] J.T.Hsu, M.Ishii, T.Hibiki, Experimental study on twophase natural circulation and flow termination in a loop, Nuclear Engineering and Design, 186, 1998, [9] M.Ozawa, Critical heat flux induced by flow instability in boiling channels, 6 th International Conference on Boiling Heat Transfer, 7-12 May 2006, Spoleto (Italy). [10] L.Santini, C.Lombardi, M.Ricotti, Experimental investigation on two phase flow pressure drops in an helical coil steam generator tube, 6 th International Conference on Boiling Heat Transfer, 7-12 May 2006, Spoleto (Italy). [11] J.C.Collier, Convective Boiling and Condensation, McGraw-Hill second edition [12] A.J.Mathews, The early operation of the helical oncethrough boilers at Heysham 1 and Hartlepool, IAEA, Procedings of a specialists meeting on Technology of steam generators for gas-cooled reactors, Vienna 9-12 march [13] S.Y.Jiang, Y.J.Zhang, X.X.Wu, J.H.Bo, H.J. Jia, Flow excursion and its mechanism in natural circulation, Nuclear Engineering and Design, 202, 2000,