The RV inner diameter is identical to that of the original. However, the RV outer diameter, not related to the measurement, was designed so as to be thinner, making it a light RV mockup. Thus, it was designed to be different from the actual size. The RV upper flange was simply designed.

The RV upper flange-face was assembled with the CSB and was applied to actual conditions in the same manner. The RV cylinder was designed with a thin plate, including a RV core- stabilizing lug. The height of the RV mockup was minimized to 2,665 mm.

The assembled parts of the RV core-stabilizing lugs, the RV core stop lug, and the shim were designed to be identical to their real-life counterparts.

Six supports of 70 cm in height were attached to the bottom of the RV mockup to monitor the condition of the measurement sensors. The lifting lugs were welded onto the RV upper flange to facilitate assembly and separation from the CSB mockup.

Fig. 17 and Fig. 18 show the designed 3D model and an image of the manufactured RV mockup.

Fig. 16. Important parts of the design of the mockup

Other parts of the actual RV and CSB were simply designed for usability of the experiment and out of concerns for the manufacturing budget. Thus, the thickness of cylinders was designed to be thinner than that of an actual RV and an actual CSB.

Fig. 16 shows the important parts in the design of the RV mockup and the CSB mockup.

The purpose of the OSU-MASLWR-001 test (Modro et al. 2003; Reyes et al., 2007; Pottorf et al., 2009; Mascari et al., 2011e), a design basis accident for MASLWR concept design, was to determine the pressure behavior of the RPV and containment following an inadvertent actuation of one middle ADS valve. The test successfully demonstrated the blowdown behavior of the MASLWR test facility during one of its design basis accident.

Following the inadvertent middle ADS actuation the blowdown of the primary system takes place. A subcooled blowdown, characterized by a fast RPV depressurization, takes place after the Start Of the Transient (SOT). A two-phase blowdown occurs when the differential pressure, at the break location, results in fluid flashing. A choked two-phase flow condition prevails and a decrease in depressurization rate of the primary system is experimentally observed. When the PRZ pressure reaches saturation, single phase blowdown occurs and the depressurization rate increases. The RPV and HPC pressure and the primary saturation temperature are shown in figure 7. The Psat, saturation pressure, is based on the temperature at the core outlet.

At 539 s after the SOT the pressure difference between the RPV and the HPC reaches a value less than 0.517 MPa, one of the high ADS valve is opened and, with approximately 10 s of delay, the other high ADS valve is opened equalizing the pressure of the primary and HPC system.

At 561 s after the SOT the pressure difference between the RPV and the HPC reaches a value less than 0.034 MPa, one of the sump recirculation valve is opened and, with approximately 10 s of delay, the other sump recirculation valve is opened terminating the blowdown period and starting the refill period. The refill period takes place for the higher relative coolant height in the HPC compared to the RPV. Figure 8 shows the RPV level evolution experimentally detected during the test. The RPV water level never fell below the top of the core during the execution of the test 1.

During the saturated blowdown period, the inlet and the outlet temperature of the core are equal each other assuming the saturation temperature value. A core reverse flow and a core coolant boiling off at saturation is present in the facility during this period. When the refill takes place the core normal flow direction is restarted and a delta T core is observed depending on the refill rate and core power, figure 9.

When the refill of the reactor takes place the level of the coolant reaches the location of the flow rate HL measurement point, therefore an increase of the RPV flow rate is detected for this phenomenon, figure 10.

Cycle counting is the prerequisite for any fatigue or service durability assessment method dealing with arbitrary operational load sequences. Consequently, an appropriate cycle counting algorithm is required.

Cycle counting methods in general are characterized by the following features [10]:

• decomposition of a given course of load (stress) — time history into a sequence of reversal points

• definition of a relevant elementary event (e. g. hysteresis)

• formulation of an algorithm for the detection and processing of elementary events.

The superposition of transients according to the design code (ASME Code, NB 3222.4, see Figure 12) is based on the peaks and valleys method. The largest stress ranges are usually determined from "outer combinations" (e. g. load steps across different transients respectively events). The associated frequency of occurrence results from the actual number of cycles of the participating two events with the smaller number of cycles. This event provides the associated contribution to the partial usage factor Ui. The summing up of all partial usage factors according to Miner’s rule delivers the accumulated damage (usage factor U) or cumulated usage factor CUF.

2007 SECTION III, DIVISION 1 — NB NB-3222.4 Analysis for Cyclic Operation

(5) Cumulative Damage.

■Usual counting method according to ASME-Code

■Search for the largest stress range across events („external combinations11)

■e. g. implemented in ANSYS®- postprocessing

Additionally, a counting of sub cycles within the events should be carried out according to the rain-flow cycle-counting method [e. g. 10] although it is not explicitly addressed by the design code [1]. This is standard practice in the framework of the AFC. The Hysteresis Counting Method (HCM) according to Clormann and Seeger [10] is applied for this purpose. Additionally, the introduction of so called basic events allows a more realistic consideration of the load time sequence [13].

Fuzzy models can be divided into three classes: Linguistic Models (Mamdani Models), Fuzzy Relational Models, and Takagi-Sugeno (TS) Models. Both linguistic and fuzzy relational models are linguistically interpretable and can incorporate prior qualitative knowledge provided by experts (Zadeh, 2008). TS models are able to accurately represent a wide class of nonlinear systems using a relatively small number of parameters. TS models perform an interpolation of local models, usually linear, by means of a fuzzy inference mechanism. Their functional rule base structure is well-known to be intrinsically favorable for control applications.

This chapter deals with Takagi-Sugeno (T-S) fuzzy models because of their capability to approximate a large class of static and dynamic nonlinear systems. In T-S modeling methodology, a nonlinear system is divided into a number of linear or nearly linear subsystems. A quasi-linear empirical model is developed by means of fuzzy logic for each subsystem. The whole process behavior is characterized by a weighted sum of the outputs from all quasi-linear fuzzy implication. The methodology facilitates the development of a nonlinear model that is essentially a collection of a number of quasi-linear models regulated by fuzzy logic. It also provides an opportunity to simplify the design of model predictive control. In such a model, the cause-effect relationship between control u and output y at the sampling time n is established in a discrete time representation. Each fuzzy implication is generated based on a system step response (Andone&Hossu, 2004), (Hossu et al., 2010) , (Huang et al. 2000).

IF y(n) is A0, y(n -1) is A1,….,y(n — m +1) is A’m_a,

and u(n) is B0, u(n -1) is B,…, u(n -1 +1) is Bj-1 (1)

T

THEN yl (n +1) = y(n) + ^ hjAu(n +1 — j)

j=1

where:

Aij fuzzy set corresponding to output y(n-j) in the ith fuzzy implication

Bj fuzzy set corresponding to input u(n-j) in the ith fuzzy implication

hij impulse response coefficient in the ith fuzzy implication

T model horizon

Au(n) difference between u(n) and u(n-1)

A complete fuzzy model for the system consists of p fuzzy implications. The system output y(n +1) is inferred as a weighted average value of the outputs estimated by all fuzzy implications.

£аУ (n+1)

y(n+1) = ——- (2)

TMj

j=1

where

Mj =*Aj kB’ (3)

i k

considering

®j =~^~ (4)

ZMj

j=1

p

y{n +1) = ^a’y’ (n +1)

j=i

Added mass and damping are known to be dependent on fluid properties (in particular, fluid density and viscosity) as well as functions of component geometry and adjacent boundaries, whether rigid or elastic. Nuclear reactor components are typically immersed in a liquid coolant and are often closely spaced (Wambsganss, et al., 1974).

3.1 Added mass

(TEMA, 7th Edition, 1988) defines hydrodynamic mass as an effect which increases the apparent weight of the vibrating body due to the displacement of the shell-side flow resulting from motion of vibrating tubes, the proximity of other tubes within the bundles and relative location of shell-wall. The so-called "virtual mass" for a tube is composed of the mass of the tube, mass of the fluid contained in the tube and the inertia M’ imposed by the surrounding fluid. This hydrodynamic mass M’ is a function of the geometry, the density of the fluid, and the size of the tube. In an ideal fluid, it is proportional to the fluid density and to the volume of the tube (Moretti & Lowry, 1976), and hence may be expressed, per unit length as:

M’ C 2 (-1 8)

— = CmP^r (18)

where L is the tube length, r the radius of the tube and p is the mass per unit volume of the surrounding fluid, Cm is called the inertia coefficient which is a function of the geometry, and is discussed by Lamb (Lamb, 1932, 1945). If the moving tube is not infinitely long, the flow is three-dimensional and leads to smaller values of Cm (Moretti & Lowry, 1976). For a vibrating tube in a fluid region bounded by a circular cylinder, Stokes (Endres & Moller, 2009) has determined hydrodynamic mass per unit length as given by:

where Cm = — — , R is the outer radius of annulus and, ma = ржт2 , where p is fluid

m R2 _ r2

density, and r is the tube radius.

(Wambsganss, et al., 1974) have published a study on the effect of viscosity on Cm. Hydrodynamic mass M’ for a tube submerged in water was determined by measuring its natural frequency, fa, in air and, fw, in water. Neglecting the density of air compared to water, the following equation may be obtained from beam equation (Moretti & Lowry, 1976).

(20)

where p is the tube mass per unit length. The inertia coefficient, Cm, can be obtained from Equation 18. Figure 13 gives the results showing the variation of Cm with pitch-to-diameter ratios (Wambsganss, et al., 1974).

C„

P/D ratio

3.2 Damping

System damping has a strong influence on the amplitude of vibration. Damping depends upon the mechanical properties of the tube material, the geometry of intermediate supports and the physical properties of the shell-side fluid. Tight tube-to-baffle clearances and thick baffles increase damping, as does very viscous shell-side fluid (Chenoweth, 1993). (Coit et al., 1974) measured log decrements for copper-nickel finned tubes of 0.032 in still air. The range of most of the values probably lies between 0.01 and 0.17 for tubes in heat exchangers (Chenoweth, 1993). From (Wambsganss, et al., 1974), damping can be readily obtained from the transfer function or frequency response curve as

with / = /n (1) — /n (2),

where / is the resonant frequency and /N (1 and/N( 1 are the frequencies at which the response is a factor —^ of resonant response.

(Lowery & Moretti, 1975), have concluded that damping is almost entirely a function of the supports. More complex support conditions (non-ideal end supports or intermediate supports with a slight amount of clearance) lead to values around 0.04. From analytical point of view, (Jones, 1970) has remarked that in most cases, the addition of damping to the beam equation re-couples its modes. Only a beam, which has, as its damping function, a restricted class of functions can be uncoupled. (Chen et al., 1994) have found the fluid damping coefficients from measured motion-dependent fluid forces. (Pettigrew et al., 1986, 1991) outlines the energy dissipation mechanisms that contribute to tube damping as given in Table 6:

(Pettigrew et al., 1986) further outlines the parameters that influence damping as given below:

The Type of tube motion. There are two principal types of tube motion at the supports, rocking motion and lateral motion. Damping due to rocking is likely to be less. Rocking motion is pre-dominant in lower modes. Dynamic interaction between tube and supports may be categorized in three main types, namely: sliding, impacting, and scuffing, which is impacting at an angle followed by sliding:

Effect of number of supports. The trend available in damping data referenced in (Pettigrew et al., 1986), when normalized give

£n = £N / (N -1) (22)

where £n is the normalized damping ratio, N is the number of spans, and £ is the damping ratio.

Effect of tube Frequency. Frequency does not appear to be significant parameter (Pettigrew et al., 1986).

Effect of vibration amplitude. There is no conclusive trend of damping as a function of amplitude. Very often, amplitude is not given is damping measurements (Pettigrew et al., 1986).

Effect of diameter or mass. Large and massive tubes should experience large friction forces and the energy dissipated should be large. However, the potential energy in the tube would also be proportionally large in more massive tubes. Thus, the damping ratio, which is related to the ratio of energy dissipated per cycle to the potential energy in the tube should be independent of tube size or mass (Pettigrew et al., 1986).

Effect of side loads. In real exchangers, side loads are possible due to misalignment of tube — supports or due to fluid drag forces. Side loads may increase or reduce damping. Small side loads may prevent impacting, and thus reduce damping, whereas large side loads may increase damping by increasing friction (Pettigrew et al., 1986).

Effect of higher modes. Damping appear to decrease with mode order, for mode order higher than the number of spans, since these higher modes involve relatively less interaction between tube and tube-support (Pettigrew et al., 1986).

Effect of tube support thickness. Referenced data in (Pettigrew et al., 1986) clearly indicates that support thickness is a dominant parameter. Damping is roughly proportional to support thickness. (Pettigrew et al., 1986) corrected the damping data line for support width less than 12.7mm such that

£nc =£n [ ^ J (23)

where L is support thickness in mm and £nc is the corrected normalized damping ratio.

Effect of clearance. For the normal range of tube-to-support diametral clearances (0.40mm — 0.80mm), there is no conclusive trend in the damping data reviewed (Pettigrew et al., 1986).

Design Recommendations (Pettigrew et al., 1986, 1991, Taylor et al., 1998)

Friction damping ratio in a multi-span tube (percentage)

where N is the number of tube spans, L is the support thickness, lm is the characteristic span length usually taken as average of three longest spans.

Viscous damping ratio

Rogers simplified version of Chens’ cylinder viscous damping ratio (percentage) of a tube in liquid.

where p is the fluid density, m is the mass per unit length of tube (interior fluid and hydrodynamic mass), De is the equivalent diameter to model confinement due to surrounding tubes, D is the tube diameter, / is the frequency of tube vibration and и is

Squeeze film damping ratio

(For multi-span tube) £

Support damping

(Pettigrew, et al., 1991) has developed a semi-empirical expression to formulate support damping, using Mulcahys’ theory (Mulcahy, 1980).

(28) (TEMA, 6th Edition, 1978, TEMA, 7th Edition, 1988)

According to TEMA standards, £ is equal to greater of £1, and £2 (For shell-side liquids)

C = l^L

Wafn

where v is the shell fluid velocity, do is the outside tube diameter, po is the density of shell — side fluid, /n is the fundamental frequency of tube span, and Wo is the effective tube weight.

(For shell-side vapors)

where N is the number of spans, tb is the baffle or support plate thickness, and l is the tube unsupported span. A review of two-phase flow damping is presented by (Khushnood et al., 2003).

The production and consumption of electricity lead to environmental impacts which must be considered in making decisions on the way in which to develop energy systems and energy policy. The key to moving towards rational energy development lies in finding the ‘balance’ between the environmental, economic and social goals of society and integrating them at the earliest stages of project planning, programme development and policy making. The environmental consequences of energy production and use must be known in order to manage and choose energy products and services. The requirements for information in support of corporate and/or government planning and decision making are changing, there being a clear emergence of concerns for environmental accountability. Thus, there is a need to integrate the environment more effectively into all aspects of energy planning and decision making, in order to make current decisions environmentally prudent, economically efficient and socially equitable, both now and for the future. Assessing environmental impacts associated with different energy systems through the use of a framework which facilitates comparison will permit consistent and transparent evaluation of these energy alternatives.

Tiering of environmental evaluation

Appraising sustainability

Sustainability appraisal (SA) has recently emerged as a policy tool whose fundamental purpose is to direct planning and decision-making towards sustainability. Its foundations
lie in well-established practices such as strategic environmental assessment (SEA), applied to policies, plans and programmes, and in project environmental impact assessment (EIA). The distinguishing feature of sustainability appraisal, when compared with others, e. g. SEA, is that the concept of sustainability, not just the environment, lies at its core. However, as explained below, comprehensive SEAs also deal with all three components — environment, economy and society — in a balanced way. No matter which type of assessment is applied at the highest planning level, either SA or SEA, its aim is to provide answers in a comparative manner and to assist in the process of identifying the most suitable alternative, e. g. energy option.

Alulsio Sousa Reis, Jdnior, Eliane S. C. Temba, Geraldo F. Kastner and Roberto P. G. Monteiro

Centro de Desenvolvimento da Tecnologia Nuclear — (CDTN)

Brazil

1. Introduction

For legal and regulatory purposes, the International Atomic Energy Agency (IAEA, 1994) defines radioactive waste as "waste that contains or is contaminated with radionuclides at concentrations or radioactivity levels greater than clearance levels as established by the regulatory body". The radioactive wastes are residues that have been produced by human nuclear activity and for which no future use is foreseen. Besides the nuclear power plants, the nuclear weapons testing, medical uses and various research studies involve a large number of radionuclides. In particular the nuclear accidents such as Three Mile Island Nuclear Power Station, where some gas and water were vented to the environment around the reactor, Chernobyl Nuclear Power Plant, the effects of the disaster were very widespread and Fukushima II Nuclear Power Plant have also released a large amount of radionuclides to environment.

In the case of radioactive wastes each country has its own classification, in general we can identify three types of wastes, that are, Low Level Waste (LLW), the LLW wastes contain primarily short lived radionuclides which refer to half-lives shorter than or equal to 30-year half-life, Intermediate Level Waste (ILW), radioactive non-fuel waste, containing sufficient quantities of long-lived radionuclides which refer to half-lives greater than 30 years. And a third one that is High Level Waste (HLW), arise from the reprocessing of spent fuel from nuclear power reactors to recover uranium and plutonium, containing fission products that are high radioactive, heat generating and long-lived. We would like to call attention to the fact that the waste classification LLW, ILW, HLW used here is only one of several alternative schemes; we adopted the simplest one.

Identification and characterization of radioactive wastes is a technical challenge because of their importance in choosing the appropriate permanent storage mode or further processing. Characterization definition of nuclear waste by IAEA (IAEA, 2003) is "the determination of the physical, chemical and radiological properties of the waste to establish the need for further adjustment, treatment, conditioning, or its suitability for further handling, processing, storage or disposal. Thus, it involves a collection of data that pertains to specific waste properties as well as processing parameters and quality assurance, some of

which include the following: thermal, mechanical, physical, biological, chemical and radioactivity properties (IAEA, 2007).

Testing and analyzes to demonstrate the radioactive content and the quality of final waste forms and waste packages are key components of this knowledge and control and are essential to accurate characterization of the waste. Physical characterization involves inspection of the waste to determine its physical state (solid, liquid or gaseous), size and weight, compactability, volatility and solubility, including closed waste packages which can be done using a variety of techniques, such as radiography (X-ray). Chemical waste characterization involves the determination of the chemical components and properties of the waste that is, potential chemical hazard, corrosion resistance, organic content, reactivity. This is most often done by chemical analysis of a waste sample. The radioactive inventory of various materials needs to be assessed for the classification of the nuclear waste. Radiological waste characterization involves detecting the presence of individual radionuclides and its properties such as half-life, intensity of penetrating radiation, activity and concentration and quantifying their inventories in the waste. This can be done by a variety of techniques, such as radiometric methods, mass spectrometric methods depending on the waste form, radionuclides involved and level of detail/accuracy required.

Furthermore, for developing a scaling factor (IAEA, 2009) to be applicable to the assessment of the radioactive inventory of the wastes with various matrices, it is indispensable to prepare a database compiled with a large numbers of information related to the radioactive inventory of long lived alpha and beta emitting nuclides which are difficult to measure (DTM) and gamma emitting nuclides which are easy to measure (ETM). It is necessary to develop analytical techniques for the DTM nuclides.

The aim of this work was to develop a sensitive analytical procedure for simultaneous determination of radionuclides difficult to measure. Between them is the 59Ni and 63Ni determination in low and intermediate level wastes from Brazilian Nuclear Power Plants — Eletrobras Termonuclear according to an analytical protocol developed based on sequential separation of different radionuclides presents in the waste matrices (Reis et al, 2011). Sources for 59Ni are austenitic steel in the reactor and activation of nickel dissolved in the coolant and in corrosion particles deposited on the core. The content of nickel in stainless steel is around to 10% and in Inconel in the range of 50-75%. Furthermore, nickel is found as an impurity in Zircaloy, ~ 40 ppm, and in reactor fuel, ~ 20 ppm (Lingren et al, 2007).

Six samples for Mossbauer effect experiments collected from different parts of NPP Bohunice unit were prepared by crushing to powder pieces (Table 9). These samples consisted of corrosion products taken from small coolant circuit of pumps (sample No. 3.1), deposits scraped from filters after filtration of SG — feed water during operation (sample No. 3.2), corrosion products taken from SG42 pipelines — low level (sample No. 3.3), mixture of corrosion products, ionex, sand taken from filter of condenser to TG 42 (sample No. 3.4), deposit from filters after refiltering 340 l of feed water of SG S3-09 during passivation 27. and 28. 5. 08 (sample No. 3.5) and finally deposit from filters after 367 l of feed water of SG S4-09 during passivation 27. and 28. 5. 08 (sample No. 3.6). All samples were measured at room temperature in transmission geometry using a 57Co(Rh) source. Calibration was performed with a-Fe. Hyperfine parameters of the spectra including spectral area (Arei), isomer shift (IS), quadrupole splitting (QS), as well as hyperfine magnetic field (Bhf), were refined using the CONFIT fitting software [27], the accuracy in their determination are of +0.5 % for relative area Are], +0.04 mm/ s for Isomer Shift and Quadrupole splitting and ±0.5 T for hyperfine field correspondingly. Hyperfine parameters of identified components (hematite, magnetite, goethite, lepidocrocite, feroxyhyte) were taken from [28].

All measured spectra contained iron in magnetic and many times also in paramagnetic phases. Magnetic phases contained iron in nonstoichiometric magnetite Fe3-xMxC>4 where Mx are impurities and vacancies which substitute iron in octahedral (B) sites. Another magnetic fraction is hematite, a-Fe2C3. In one sample also the magnetic hydroxide (goethite a — FeCCH) was identified.

Paramagnetic fractions are presented in the spectra by quadrupole doublets (QS). Their parameters are close to those of hydroxides e. g. lepidocrocite у — FeOOH or to small, so called superparamagnetic particles of iron oxides or hydroxides with the mean diameter of about 10 nm. It should be noted that there is no problem to distinguish among different magnetically ordered phases when they are present in a well crystalline form with low degree (or without) substitution. Both the substitutions and the presence of small superparamagnetic particles make the situation more complicated [29]. In such cases, it is necessary to perform other supplementary measurements at different temperatures down to liquid nitrogen or liquid helium temperatures without and with external magnetic field [30].

Mossbauer spectrum (Fig. 10) of sample no. 3.1 (corrosion products taken from small coolant circuit of pumps) consist of three magnetically split components, where the component with hyperfine field Bhf = 35.8 T was identified as goethite (a-FeOOH). Hyperfine parameters of remaining two magnetically split components are assigned to A — sites and B — sites of magnetite (Fe3C4). Cne paramagnetic spectral component has appeared. According to water environment and pH [31], this component should be assigned to hydrooxide (feroxyhyte 8-FeOOH).

The sample No. 3.2 (deposits scraped from filters after filtration of SG — feed water during operation) also consists of three magnetically split components, where two of them were assigned to magnetite (Fe3C4) as in previous spectra, and the remaining magnetically split component was identified as hematite (a-Fe2O3). Paramagnetic part of the spectra was formed by one doublet, whose hyperfine parameters were assigned to hydroxide (lepidocrocite, y-FeOOH). The spectrum is shown in Fig. 11.

The spectrum (Fig. 12) of the sample No. 3.3 (corrosion products taken from SG42 pipelines — low level) consists only of two magnetically split components with hyperfine parameters assigned to A — sites and B — sites of nearly stoichiometric magnetite (Fe3O4) with a relative area ratio в = 1.85.

The sample No. 3.4 (mixture of corrosion products, ionex, sand taken from filter of condenser to TG 42) also consists of a magnetically split component which corresponds to hematite (a-Fe2O3) and two magnetically split components were assigned to magnetite (Fe3O4) as in previous spectra, and the remaining paramagnetic component was identified as hydroxide. The spectrum of the sample No. 3.4 is shown in Fig. 13.

velocity (mm/s)

Fig. 13. Mossbauer spectrum of sample no. 3.4. Haematite (blue), A-site (red) , B-site (dark red) magnetite and hydroxide (green) was identified

Both the sample No. 3.5 (deposit from filters after 340 l of feed water of SG S3-09 during passivation 27. and 28. 5. 08) and the sample No. 3.6 (deposit from filters after 367 l of feed water of SG S4-09 during passivation 27. and 28. 5. 08) consist of three magnetically split components, identified as hematite (a-Fe2O3) and magnetite (Fe3O4) and the remaining paramagnetic component in both spectra was assigned to hydrooxide (lepidocrocite y — FeOOH). The spectra of the samples No. 3.5 and 3.6 are shown in Figs. 14 and 15. Based on comparison of results from samples 3.5 and 3.6 it can be concluded that the longer passivation leads more to magnetite fraction (from 88% to 91%) in the corrosion products composition.

As it was mentioned, above all hydroxides could be also small superparamagnetic particles.

The refined spectral parameters of individual components including spectral area (Arei), isomer shift (IS), quadrupole splitting (QS), as well as hyperfine magnetic field (Bf) are listed in Table 9 for room (300 K) temperature Mossbauer effect experiments. The hyperfine parameters for identified components (hematite, magnetite, goethite, lepidocrocite, feroxyhyte) are listed in [28].

Major fraction in all samples consists of magnetically ordered iron oxides, mainly magnetite (apart from the sample No. 3.1 and 3.2, where also goethite and hematite has appeared, respectively). Magnetite crystallizes in the cubic inverse spinel structure. The oxygen ions form

velocity (mm/s)

s

velocity (mm/s)

a closed packed cubic structure with Fe ions localized in two different sites, octahedral and tetrahedral. The tetrahedral sites (A) are occupied by trivalent Fe ions. Tri- and divalent Fe ions occupying the octahedral sites (B) are randomly arranged at room temperature because of electron hopping. At room temperature, when the electron hopping process is fast, the Mossbauer spectrum is characterized by two sextets. The one with the hyperfine magnetic field Bhf = 48.8 T and the isomer shift IS = 0.27 mm/ s relative to a-Fe corresponds to the Fe3+

ions at the tetrahedral A — sites. The second one with Бы = 45.7 T and IS = 0.65 mm/ s is the pe2.5+ — like average signal from the cations at octahedral B sites. Fe2+ and Fe3+ are indistinguishable due to fast electron transfer (electron hopping), which is faster (~1 ns) than the 57Fe excited state lifetime (98 ns). The magnetite unit cell contains eight Fe3+ ions and eight Fe2+ and Fe3+ ions, 16 in total at the B sites, therefore, the intensity ratio в = I(B)/I(A) of the two spectral components is a sensitive measure of the stoichiometry. Assuming that the room temperature ratio of the recoil-free fractions fB/ fA for the B and A sites is 0.97 [32], the intensity ratio в for a perfect stoichiometry should be 1.94. In non-stoichiometric magnetite, under an excess of oxygen, cation vacancies and substitutions at the B sites are created. The vacancies screen the charge transfer and isolate the hopping process. For each vacancy, five Fe3+ ions in octahedral sites become trapped. In the Mossbauer spectrum these trapped Fe3+ ions at the octahedral sites and Fe3+ ions at tetrahedral sites are indistinguishable without applying an external magnetic field. Therefore, in the spectrum of non-stoichiometric magnetite, intensity transfer from the Fe2.5+ to Fe3+-like components is observed. Therefore, the intensity ratio в decreases markedly with the oxidation process, until the stoichiometry reaches the y-Fe2O3 phase. It should be noted that in our samples the intensity ratio в is far from 1.94 (for perfect stoichiometry), varies from 0.97 up to 1.85.

2. Conclusions

Material degradation and corrosion are serious risks for long-term and reliable operation of NPP. The paper summarises results of long-term measurements (1984-2008) of corrosion products phase composition using Mossbauer spectroscopy.

The first period (mostly results achieved in 80-ties) was important for improving proper Mossbauer technique [5]. The benefit from this period came via experience collection, optimization of measurement condition and evaluation programs improvement. Unfortunately, the samples were not well defined and having in mind also different level of technique and evaluation procedures, it would be not serious to compare results from this period to results obtained from measurement after 1998.

The replacement of STN 12022 steel (in Russian NPP marked as GOST 20K) used in the steam generator feed water systems is necessary and very important from the operational as well as nuclear safety point of view. Steel STN 17 247 proved 5 years in operation at SG35 seems to be optimal solution of this problem. Nevertheless, periodical inspection of the feed water tubes corrosion (after 10, 15 and 20 years) was recommended.

Based on results of visual inspection performed at April 19, 2002 at SG16 (NPP V1) it was confirmed, that the steam generator was in good condition also after 23 years of operation. Samples taken from the internal body surface of PG16 confirmed that the hematite concentration increases in the vertical direction (from bottom part to the top).

The newest results from 2008 confirm good operational experiences and suitable chemical regimes (reduction environment) which results mostly in creation of magnetite (on the level 70% or higher) and small portions of hematite, goethite or hydrooxides.

Regular observation of corrosion/ erosion processes is essential for keeping NPP operation on high safety level. The output from performed material analyses influences the optimisation of operating chemical regimes and it can be used in optimisation of regimes at

HEMATITE

MAGNETITE

Period 1998-99 1-13

Period 2002-03 2.11-2.15

Period 2006-08 3.1-3.6

Fig. 16. Summarized figure of corrosion products phase composition at NPP V-2 Bohunice (Slovakia) performed according to results from period 1998-2008

decontamination and passivation of pipelines or secondary circuit components. It can be concluded that a longer passivation time leads more to magnetite fraction in the corrosion products composition.

Differences in hematite and magnetite content in corrosion layers taken from hot and cold collectors at SG11 in 2004 show, that there is a significantly lower presence of magnetite in case of hot collector. This fact can be derived from 2 parallel factors: (i) difference in temperature (about 298°C — HC) and (about 223°C — CC) and mostly due to (ii) higher dynamic of secondary water flowing in the vicinity of hot collector, which high probably removes the corrosion layer away from the collector surface.

With the aim to summarize our results in the form suitable for daily use in the operational conditions a summarized figure was created (see Fig. 16). Corrosion products phase composition (limited on magnetite and hematite only) is presented in form of circular diagrams.

Basically, the corrosion of new feed water pipelines system (from austenitic steel) in combination with operation regimes (as it was at SG35 since 1998) goes to magnetite. In samples taken from positions 5 to 14 (see Fig. 16 — right corner). The hematite presence is mostly on the internal surface of SG body (constructed from "carbon steel" according to GOST20K). Its concentration increases towards the top of the body and is much significant in the seam part of SG where flowing water removes the corrosion layer via erosion better than from the dry part of the internal surface or upper part of pipeline.

The long-term study of phase composition of corrosion products at VVER reactors is one of precondition to the safe operation over the projected NPP lifetime. The long-term observation of corrosion situation by Mossbauer spectroscopy is in favour of utility and is not costly. Based on the achieved results, the following points could be established as an outlook for the next period:

1. In collaboration with NPP-Bohunice experts for operation as well as for chemical regimes, several new additional samples from not studied places should be extracted and measured by Mossbauer spectroscopy with the aim to complete the existing results database.

2. Optimisation of chemical regimes (having in mind the measured phase composition of measured corrosion specimens from past) could be discussed and perhaps improved.

3. Optimisation and re-evaluation of chemical solutions used in cleaning and/or decommissioning processes during NPP operation can be considered.

In connection to the planned NPP Mochovce 3, 4 commissioning (announced officially at 3.10.2008) it is recommended that all feed water pipelines and water distribution systems in steam generators should be replaced immediately before putting in operation by new ones constructed from austenitic steels. The Bohunice design with feed water distribution boxes is highly recommended and it seems to be accepted from the utility side.

3. Acknowledgement

This work was supported by company ENEL Produzione, Pisa and by VEGA 1/0129/09.

[1] Some countries (e. g. France) does not consider the fuel pellet as a safety barrier

[2] The analysis of GDH guillotine break at complete ECCS failure and at operation of one, two or three ECCS pumps demonstrated that for reliable cooling of the core at a long stage two ECCS pumps are enough. With such number of ECCS equipment, after one hour from the beginning of the accident, the ECCS water flow rate starts to exceed water discharge through the break. However, short-term increase in temperatures of fuel rod cladding and FC walls at the initial stage of accident in FCs, connected to the broken GDH, is inevitable in any case. The excess of acceptance criterion for fuel rod cladding (700 °С) is probable in 12 FCs. A more detailed analysis (using RELAP/SCDAPSIM model presented in Fig. 7) shows that this short temperature peak does not lead to the failure of any fuel rods.

• In the case of coolant flow rate stagnation in channels, multiple FCs breaks are probable after approximately 20 s from the beginning of the accident if the reactor is not shutdown on time. On the contrary, if a reactor is shutdown quickly (until the FCs heat up), the acceptance criterion for channels walls will not be violated and the channels

[3] The results of the analysis showed that at LOCA in Zone 4 the operation of ECCS short­term subsystem is not necessary, i. e. the temperature of fuel rod cladding and FC walls is much lower than acceptance criteria without operation of this subsystem.

• For reliable cooling of the reactor core in long-term post-accidental period, it is necessary to have not less than two ECCS pumps in operation in the case of two steamlines break, and not less than one pump in the case of one steamline break.

• In case of breaks in Zone 4 without the reactor shutdown (break of steamlines), the temperature rises much more slowly (especially the temperature of FCs walls). This specifies that such breaks in Zone 4 are less dangerous, than breaks in Zone 1 (Fig. 19). In the case without the reactor shutdown, the melting of the core at low pressure in RCS is probable, but does not result in the immediate damage of the reactor cavity.

[4] loss of intermediate cooling circuit;

• loss of service water;

• station blackout.

The station blackout (the most likely initiating event) is the loss of normal electrical power supply for local needs with an additional failure on start-up of all diesel generators. In the case of loss of electrical power supply MCPs, the circulating pumps of the service water system and feedwater supply pumps are switched-off. The failure of diesel generators leads to the non-operability of the emergency long-term core cooling subsystem. It means the

[5] For the first group of accidents (accidents with no severe damage of the core) it was showed: (1) In the case of erroneously withdrawn of a group of control rods, the local power increase appears in the adjacent fuel channels, but this do not lead to overheating of the fuel in these channels. The operators have possibility to compensate this local power increase by inserting remaining control rods. In the case the local power exceeds limits — the reactor will be shutdown automatically by activation of emergency shutdown system. (2) In the case of loss of long-term cooling,

[6] Turbulent buffeting

• Vorticity excitation

• Fluid-elastic excitation

• Acoustic resonance

Turbulent buffeting cannot be avoided in heat exchangers, as significant turbulence levels are always present. Vibration at or near shedding frequency has a strong organizing effect on the wake. Vorticity or vortex shedding or periodic wake shedding is a discrete, periodic, and a constant Strouhal number phenomenon. Strouhal number is the proportionality constant between the frequency of vortex shedding and free stream velocity divided by

[7] First, information about the present status of the waste was gathered. Attention was paid to the variability and accuracy of data on quantities, the radionuclide inventory and the activity of different types of waste.

• Then, based on this information, what may most likely be expected (with regard with these waste characteristics) by the end of the anticipated operational period of Krsko NPP, i. e. 2023, was estimated. The ORIGEN2 computer code was used for calculating isotope generation, activity build-up and depletion, and the decay heat of spent fuel (Croff, 1983), while a specific code was developed for calculations associated with LILW.

• The total activity, its time dependent change and the identification of radionuclides that mainly contribute to the activity in long timeframes, were applied as key information for discussing waste-disposal options for spent fuel (HLW). Changes (variations, uncertainty) in these characteristics were evaluated based on the technical specifications that are in place after the replacement of steam generators at the plant in the 17th fuel cycle in 2002. The variations considered were 3-5 % of U-235 in the fuel, and an operational period of the plant of five years more or five years less than that envisaged. The basic estimate was that Krsko NPP uses fuel with 4% U-235 in all future cycles and that it operates for 35 years.

The key input data for calculating burn-up and fuel characteristics in future cycles is not available at the moment. Consequently, certain assumptions had to be made. These were:

• 35 fuel cycles are assumed for the operational period of Krsko NPP.

• The average cycle burn-up is 12,000 MWd/tU. This value was adopted based on the following: The average number of effective days of full power operation per cycle is

[8] Annular flow: Heat flux continues increasing, and the dryness fraction in the channel increases too. When the vapor content is higher than that of churn flow, the liquid block is smashed and the vapor unites to be a continuous axle center in the core of the tube-bundle channel. The liquid film goes upward along the wall of PMMA pipe. Thus, annular flow occurs. The liquid film might be broken due to the effect of the vapor wave, as shown in Fig. 4 and Fig. 13 (d).

[9] Kliuchnikov Olexander, Seniuk Olga, Gorovyy Leontiy, Zhidkov Alexander, Ribalka Valeriy, Berezhna Valentyna, Bilko Nadiya, Sakada Volodimir, Bilko Denis, Borbuliak Irina, Kovalev Vasiliy, Krul Mikola, Petelin Grigoriy

Institute Cell Biology & Genetic Engineering of NAS of Ukraine, Ukraine Institute for Safety Problems of Nuclear Power Plants NAS of Ukraine, Ukraine National University "Kyievo-Mogiljanskaja Academy", Ukraine

[10] Samples m006, m008, m010 were taken from outside surface, samples M007, M009, M012 from inside surface of the feed water pipeline according to the same positions 1, 2 and 3, respectively. Sample M15 — see Fig. 7, position 7).

[11] Samples l754-l757 were taken from the feed water pipelines in situ during the reactor shut down. Samples l758-l790 were taken from the same steam generator from selected parts of feed water dispersion box (see Table 3 and Fig. 6, positions 1-14)

1.2 Analysis of accidents with no severe damage of the core

As it was mentioned, the first category of BDBA (damage of the core or its components with the reactor maintaining its overall structural integrity see Figure 5) is divided into two groups: (1.1) no severe damage of the core; (1.2) severe core damage accompanied by the containment of core fragments in the reactor core, accident localization system or other reactor buildings.

The examples of accidents for RBMK-1500 in the first group (1.1, see Figure 5) are the failure of the final heat sink systems with the loss of their functions or staff error, failure of group of reactor Control & Protection System (CPS) rods, failure of one or more channels in the reactor Emergency Core Cooling System (ECCS) short-term or long-term subsystems, failure of the system which supplies feedwater during a prolonged period of time, and other failures. This group of accidents involves additional failures of the equipment in the safety systems (additional to failures considered by the single failure principle). Usually the accidents of first group (no severe damage of the core) are analyzed in order to justify the effectiveness of the functional backup, as well as to assess the conditions and the time available for the backup systems to be actuated. As a rule, this belongs to the field of the accident management on the basis of symptom-oriented emergency operating procedures.

Thus, in case of a postulated failure of CPS rods movement during reactor operation at power, the actions taken by the operators to shutdown the reactor and to hold it in a subcritical state were determined. The reactor can be shutdown and maintained subcritical by inserting one CPS rod into the core, decreasing the water temperature in the CPS cooling circuit or decreasing the temperature of the graphite stack [8]. At Ignalina NPP the Additional Hold-down System is implemented in case of CPS malfunction to prevent reactor re-criticality by injecting the liquid poison (Gadolinium Nitrate) into the CPS cooling circuit [19].

During reactivity initiated accidents the situation can occur when the group of control rods is withdrawn erroneously. Under adverse conditions it is possible that the signal for local Automatic control will not be generated and the local power increase can occur in a group of fuel channels. Validated calculations using the experimental data showed that this increase could reach 2-2.5 times of the nominal channel power, but it does not cause significant coolant flow decrease in the fuel channels (due to the increased channel resistance in the steam zone) and overheating of the fuel channels [9]. Since this chapter mainly deals with the thermal-hydraulics, more detailed examples with reactivity initiated accidents are not presented there.

Another example could be the loss of long-term cooling. The performed deterministic calculations showed that the reactor core cannot be damaged without the make-up by feedwater during any transient after full reactor shutdown in approximately 1.5 hours [8, 20]. One high pressure pump is sufficient to cool down the reactor for one hour after the reactor shutdown. If the water supply from one pump is re-established, all the parameters of RCS and reactor remain within safe operation limits. However, if the high pressure pumps are not available, the manual operators’ actions are required. In [18] the optimal reactor core cooldown scenario for RBMK-1500 in case of station blackout was developed (see Figure 8 — Figure 10). The modeling of such scenario was performed using RELAP5 model presented in Figure 6.

In the analysis presented below, it is considered that the operator takes early actions: 15 minutes after the beginning of the accident the operator begins to supply cold water from ECCS hydro-accumulators into GDH of both RCS loops. After approximately 1.5 hours from the beginning of the accident, the peak fuel cladding temperature exceeds 400 oC (Figure 8). According this signal, the operator opens one steam relief valve to decrease pressure in RCS (Figure 9). At the same time the operator takes actions to maintain the water supply by gravity from deaerators and prepares the connection for water supply from the artesian water source. The activation of ECCS hydro-accumulators after 15 minutes from the beginning of the accident provides only a small amount of water due to equalization of pressures in hydro-accumulators and GDH. Additional amount of water from hydro­accumulators is injected when the pressure in RCS decreases (~2 h after the beginning of the accident). The water supply from ECCS hydro-accumulators and opening of one steam relief valve would result in the increase of water level in RCS (Figure 10). At the time moment t = 2.5 h, the water supply from ECCS hydro-accumulators stops. Approximately 170 m3 of water is injected from ECCS hydro-accumulators.

When the pressure in RCS decreases down to 1.2 MPa (pressure in deaerators), the water flow from the deaerators begins (t = 3.8 h). There are four deaerators, which contain 480 m3 of water with temperature of ~ 190 oC. After the connection of deaerators to RCS, the pressure decrease leads to boiling of water in deaerators. The initial flow rate of water from deaerators does not warrant adequate cooling of the core — the temperature of core components is increasing at the time interval t = 3 — 5 h (Figure 8). To increase the flow rate of water from deaerators, at the time moment t = 5.3 h the operator opens one additional Steam Relief Valve (SRV). As it is seen from Figure 8, this action improves the core cooling conditions and the temperature of core components starts to decrease. The pressure in RCS decreases down to the pressure in artesian water system (~ 0.6 MPa) only more than 13 hours after the beginning of the accident. A complete connection of artesian water to RCS is permitted only after the decrease of pressure in RCS down to the level of pressure in the

artesian water source. These measures should prohibit the injection of coolant from RCS into the pipeline of artesian water. After the connection of artesian water source to supply water into reactor, the water level in reactor core starts to increase, which means the success of core cooling. The mentioned operators’ actions lead to a slow decrease of pressure in RCS, the fuel rods claddings and channels walls temperatures become not higher than 600 oC.

Fig. 10. RCS de-pressurization and water supply into reactor from ECCS hydro­accumulators, deaerators and artesian water source in case of station blackout. Calculated water level behavior in RCS

The results of the mentioned neutron-physical and thermal-hydraulic investigations have served as the basis for expanding the region where the accidents of the first group with multiple failures can be controlled using the symptom-oriented emergency operating procedures, and they have made it possible to determine the actions to be taken by personnel in order to prevent severe core damage.

Significant tube-to-restraint interaction can lead to fretting wear. Large amplitude out-of­plane motion will result in large impact forces and in-plane motion will contribute to rubbing action. Impact force and tube-to-restraint relative motion can be combined to determine work-rate. Work-rate is calculated using the magnitude of the impact force and the effective sliding distance during line contact between the tube and restraint (Chen et al., 1995). The work-rate is given below in Equations 54 and 55.

W = 1 J FA (51)

T /=0

1 n 1 n F + F

W = — Z FAS. = — Z F-2J±L^Si (52)

where F. is the instantaneous normal force, AS. is the sliding distance during line contact and n is the number of points discretized over the sample duration Ts. As the work-rate increases, the effective wear rate increases and the operational life of the U-bend tube decreases. Implementation of the technology is described in detail by (Fisher et al., 1991). Measured values of wear work-rate for pitch velocity and mass flux (Chen et al., 1995) are presented in Figures 22a and 22b respectively. The effect of fluid-elastic forces is very evident in the measured work-rates.

It is interesting to note that at higher pitch velocities and/or mass fluxes, the wear work-rate does not increase. Further study is required to understand why the flow-rates do not affect the work-rates. This may be related to the fact that at high void fractions and high flow rates the random excitation forces are constant with increasing flow rate (Taylor, 1992).