REWIEW OF THE METHODS USED FOR LEAK RATE MEASUREMENTS FOR WWER-440/230 CONFINEMENTS AND WWER-440/213 CONTAINMENTS

Transcription

1 IAEAEBPWWER10 XA REWIEW OF THE METHODS USED FOR LEAK RATE MEASUREMENTS FOR WWER440/230 CONFINEMENTS AND WWER440/213 CONTAINMENTS A PUBLICATION OF THE EXTRABUDGETARY PROGRAMME ON THE SAFETY OF WWER AND RBMK NUCLEAR POWER PLANTS September M INTERNATIONAL ATOMIC ENERGY AGENCY

2 The originating Section of this publication in the IAEA was: Safety Assessment Section International Atomic Energy Agency Wagramer Strasse 5 P.O. Box 100 A1400 Vienna, Austria REVIEW OF THE METHODS USED FOR LEAK RATE MEASUREMENTS FOB WWER440/230 CONFINEMENTS AND WWER440/213 CONTAINMENTS IAEA, VIENNA, 1998 IAEAEBPWWER10 ISSN IAEA, 1998 Printed by the IAEA in Austria September 1998

3 FOREWORD The IAEA initiated in 1990 a programme to assist the countries of central and eastern Europe and the former Soviet Union in evaluating the safety of their first generation WWER440/230 nuclear power plants. The main objectives of the Programme were: to identify major design and operational safety issues; to establish international consensus on priorities for safety improvements; and to provide assistance in the review of the completeness and adequacy of safety improvement programmes. The scope of the Programme was extended in 1992 to include RBMK, WWER 440/213 and WWER1000 plants in operation and under construction. The Programme is complemented by national and regional technical cooperation projects. The Programme is pursued by means of plant specific safety review missions to assess the adequacy of design and operational practices; Assessment of Safety Significant Events Team (ASSET) reviews of operational performance; reviews of plant design, including seismic safety studies; and topical meetings on generic safety issues. Other components are: followup safety missions to nuclear plants to check the status of implementation of IAEA recommendations; assessments of safety improvements implemented or proposed; peer reviews of safety studies, and training workshops. The IAEA is also maintaining a database on the technical safety issues identified for each plant and the status of implementation of safety improvements. An additional important element is the provision of assistance by the IAEA to strengthen regulatory authorities. The Programme implementation depends on voluntary extrabudgetary contributions from IAEA Member States and on financial support from the IAEA Regular Budget and the Technical Cooperation Fund. For the extrabudgetary part, a Steering Committee provides coordination and guidance to the IAEA on technical matters and serves as forum for exchange of information with the European Commission and with other international and financial organizations. The general scope and results of the Programme are reviewed at relevant Technical Cooperation and Advisory Group meetings. The Programme, which takes into account the results of other relevant national, bilateral and multilateral activities, provides a forum to establish international consensus on the technical basis for upgrading the safety of WWER and RBMK nuclear power plants. The IAEA further provides technical advice in the coordination structure established by the Group of 24 OECD countries through the European Commission to provide technical assistance on nuclear safety matters to the countries of central and eastern Europe and the former Soviet Union. Results, recommendations and conclusions resulting from the IAEA Programme are intended only to assist national decision makers who have the sole responsibilities for the regulation and safe operation of their nuclear power plants. Moreover, they do not replace a comprehensive safety assessment which needs to be performed in the frame of the national licensing process.

4 EDITORIAL NOTE In preparing this publication for press, staff of the IAEA have made up the pages from the original manuscript(s). The views expressed do not necessarily reflect those of the IAEA, the governments of the nominating Member States or the nominating organizations. Throughout the text names of Member States are retained as they were when the text was compiled. The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries. The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement or recommendation on the part of the IAEA.

5 CONTENTS SUMMARY 7 1. INTRODUCTION Background Requirements of national nuclear regulatory authorities Process of preparation and approval of the test programmes Definition of hermetic boundary Leakage rate testing as the basis for safety analysis Extrapolation of leakage rates obtained at reduced pressure tests to the design pressure level Leak rate definition LOCAL LEAK RATE TESTS Components of the hermetic boundary Test methods Local leak test methods for the purpose of leak identification Method of local leak rate evaluation Conclusions INDIVIDUAL COMPARTMENT TESTS Compartments tested Methods of testing Evaluation methods of semiaccessible compartment tests Results of individual compartment tests INTEGRATED LEAK TEST FOR LEAK IDENTIFICATION Aims Methods of establishing pressure difference Overpressure Underpressure Methods of leak identification and leakage rate determination Results and experience PREOPERATIONAL INTEGRATED LEAK RATE TESTS Aims and procedures of preoperational ILRT Compartments tested and volumes in hermetic compartments Systems tested during preoperational ILRT Safety measures under ILRT conditions Preoperational ELRT procedure schedule Preoperational ILRT preparation Pressurization of hermetic compartments Stabilization of parameters in hermetic compartments Leakage rate measurements and evaluation Depressurization Periodical leaktightness tests Leak rate determination Pressure measurements 46

6 Temperature measurements Air humidity measurements and water vapour partial pressure determination in hermetic compartments Measuring and evaluation system Hermetic boundary examination and leak Preoperational ILRT results and experience acquired when implementing the tests VERIFICATION INTEGRATED TESTS Aims Methods Verification test evaluation Experience STRUCTURAL INTEGRITY TEST (SIT) Aims, principles and frequency of SIT Inspection methods and measurements Structural integrity test procedure SIT evaluation CONCLUSIONS Regulatory authorities' approach Extrapolation of leakage rate values to the design pressure Local leak rate tests Leak identification Structural integrity Integrated leakage rate tests (preoperational, periodical) 58 REFERENCES 59 ABBREVIATIONS 61 SYMBOLS 62 CONTRIBUTORS TO DRAFTING AND REVIEW 63

7 SUMMARY The design and leaktightness characteristics of the confinements in WWER440/230 nuclear power plant (NPP) units and those of the containments in WWER440/213 units operating in many countries of central and eastern Europe differ significantly from those of the containments built in western type pressurized water reactors (PWRs). One of the characteristic features of confinements in WWER440/230 reactors is their comparatively high leakage rate, ranging from several hundred per cent per day upwards. The leak rates in containments of WWER440/213 reactors are significantly lower than those of WWER 440/230 units, but in both types of containments the task of leakage rate determination presents complex problems which do not appear in western type units. The methods for leakage rate testing and leakage rate definitions are not the same in all countries operating WWER440 units. The differences are significant and influence strongly the leakage rates published for each reactor. The present publication presents in detail the actual methodology being used in different countries and compares it with the western practice. This report spells out the requirements for containment leak rate testing and for integral tests in the countries operating WWER reactors and compares them with the requirements in force in the USA, Germany and France. It describes the leak rate testing principles and leakage rate extrapolation methods, discusses in detail the differences in leakage rate definitions due to significant leakages in WWER containments and explains their consequences. The report addresses the methods of leak rate testing for local leakage and for individual tests of compartments. It reviews the integrated leak tests and describes in detail and analyses the main leak rate determination methods the time total method and momentary leakage rate determination method. Requirements for instrumentation and accuracy of tests are provided. The procedures for structural integrity tests of WWER440 containments are described and recommendations given. Finally, Section 8 draws conclusions on all the topics discussed, based on the operational experience gathered in the countries operating WWER440 reactors and on joint expert discussions. The recommendations are aimed at operational staff conducting actual leakage and structural integrity tests of WWER440 containments, regulatory bodies evaluating the results of these tests, and national and international organizations active in the field of WWER safety evaluation.

8 1. INTRODUCTION 1.1. BACKGROUND In the system of defence in depth being applied in nuclear power plants, which consists of the fuel itself, fuel cladding, the reactor coolant system boundary and the containment envelope [1] it is the final barrier i.e. the reactor containment which is required to protect the environment in cases of accidents leading to the release of radioactive substances from the primary cooling system. The effectiveness of this barrier prevented any significant radiological consequences in the case of Three Mile Island accident and, contrarely, the lack of containment made it impossible to confine the radioactivity released from the core of the RBMK reactor in the Chernobyl accident. The requirements concerning the leaktightness and integrity of containment have been always very high in western types of nuclear power plants. In the case of eastern reactor technology, the role of confinement was initially regarded as secondary to the importance of an easy access to all elements of the plant, which should have assured prevention of leaks from the primary system, thus enhancing the role of the third and reducing that of the fourth barrier in the defenseindepth approach to Nuclear Plant Safety. As the role of the containment became recognized, the design of containment in WWER reactors was improved, passing from the confinements of WWER440/230 units to bubbler condenser containment of WWER440/213 reactors. The design and leaktightness characteristics of the confinements in WWER440/230 nuclear power plant (NPP) units and those of the containments in WWER440/213 units operating in many countries of central and eastern Europe differ significantly from those of the containments built in western type pressurized water reactors (PWRs). One of the characteristic features of confinements in WWER440/230 reactors is their comparatively high leakage rate, ranging from several hundred per cent per day upwards. The leak rates in containments of WWER440/213 reactors are significantly lower than those of WWER 440/230 units, but in both types of containments the task of leakage rate determination presents complex problems which do not appear in western type units. The improvements of confinement leaktightness are very important for the upgrading of the safety of WWER440/230 nuclear power plants. For example, in Bohunice Vl, the leaktightness is checked after each reactor maintenance period and the regulatory body requires that the leakages be regularly reduced. This active approach brings results the leakages in Bohunice Vl have been reduced from initial values being far above 1000% per day to the present day values of less than 100% per day 1. The importance of containment has been always recognized by the IAEA. The Safety Guide on Safety Functions and Component Classification [2] places the containment in safety class 2, as necessary "to limit the release of radioactive material from the reactor containment during and after the accident", and adds that this may be achieved by a combination of containment envelope and the use of components that limit the leakage from the containment 'The value of leak rate is defined as the fraction of air contained in the reactor containment which can leak out to the surrounding atmosphere under maximum pressure difference conditions in a unit time, for example in a second or in a day. In some reports the degree of leaktightness is illustrated by determining the equivalent crosssection area of a leak, or the diameter of a pipe which would involve the same leakage from the containment to the atmosphere. In this report the leakage rate will be expressed as the fraction of the air mass which under maximum pressure difference conditions can leak out per day, and in accordance with the prevailing practice it will be denoted as ' % per day' without repeating that'%' means'% of the air mass in the containment'.

9 envelope or reduce the pressure and temperature inside the containment envelope during and after accident conditions. A special safety guide has been published by the IAEA to discuss the design of the containment and the tests and inspection of containment and its safety systems [3]. It states that "To keep the release of radioactivity to the environment within acceptable limits in accident conditons a system of confinement shall be provided...such a system can have different engineering solutions, depending on its design requirements". In another paragraph the guide [3] states that "the design of the containment envelope shall be consistent with the leakage rate adopted to meet the dose limits stipulated by the regulatory body". As the requirements to those dose limits for WWER NPPs have been subject to significant changes over the years, the present day requirements are difficult to satisfy without significant safety upgradings of the original design. The guide [3] indicates that the leak rate is affected by factors such as leaktightness, elevation of internal pressure resulting from a postulated initiating event (PIE) and reduction of pressure resulting from the effects of energy management features. These energy management features in WWER440 units are conceptually different for 230 and 213 types, and the approach to evaluation of their effectiveness has been changing over the years. In the case of WWER440/230 units the definition of PIE has been also changed, from a small break in the RCS (of 32 mm equivalent diameter) to a medium or large size break of 200 mm equivalent diameter. This influenced the evaluation of the maximum possible pressures in the confinement, and the requirements to the confinement leaktightness and integral strength. The guide [3] states that prior to commissioning the containment with its extensions and penetrations shall be subjected to structural integrity tests at specified overpressures and to a leak rate test programme. According to the world practive, such preoperational test cover the whole range of possible pressures, including the calculated peak containment accident pressure with proper uncertainty margins. In the case of WWER containments however, the preoperational tests in many units did not go up to the upper pressure limit, following the statement of [3] that the testing methods and conditions should not themselves produce and unacceptable degradation of safety performance. As the approach to maximum possible overpressures could have involved permanent deformations of the containment envelope, the decision of most plant operators during commissioning was to limit the test pressures to values below the maximum accident conditions. The WWER440/230 reactors being operated in several countries of central and eastern Europe were designed in the 1960s assuming that the design basis accident (DBA) could be limited to a break of a pipe of a diameter of 100 mm, fitted with an orifice of a diameter of 32 mm. This assumption was used to determine the parameters of the emergency core cooling system and the confinement structure. For larger breaks, until full rupture of one of the two pressurizer surge lines (of 200 mm in diameter each), large pressure release valves in the confinement boundary were designed to open during the initial pressure peak after the accident and to release into the environment large amounts of steam escaping from the disabled reactor coolant system (RCS). It was assumed that the pressure in the confinement would be much decreased after that, since the bulk of the coolant from the RCS would have been released in this first phase of the accident. As the question of leaktightness of the confinement under such assumptions was not of primary importance, the allowable leakage rates in WWER440/230 confinements were high. The shortcomings of the WWER440/230 confinements were recognized and the newer WWER440/213 plants were provided with an improved building of much higher strength

10 and better leaktightness containing a system of water trays and air traps designed to condense steam and remove some of the air from the reactor compartments after a large loss of coolant accident (LOCA). In such an event, the system provides fast pressure decrease in the containment to subatmospheric levels and thus prevents any significant radioactive releases into the atmosphere. As the main process is that of steam bubble condensation in water, the system is called bubbler condenser containment. Leak rates in WWER440/213 containments are significantly lower than in WWER 440/230 confinements, but higher than in western PWR containments. In most plants the leakage rates at the maximum design overpressure range from 5% to 15% per day, but in some plants they are higher, reaching 100%/day or more. Taking into account the rapid decrease of pressure inside containment after design basis accident (DBA) foreseen in the plant design with bubbler condenser containment, it has been shown that the leakage rates of 15% per day at maximum overpressure correspond to an overall integrated leakage after LB LOCA comparable to the overall integrated leakage at full pressure of western PWR plants with a dry containment. Nevertheless, the question of containment leak rates is important for WWER440/213 plants and the IAEA has included this issue in the list of safety concerns requiring further action. In WWER440/230 plants the leakage rates are even higher, exceeding as a rule 400% per day at a design maximum overpressure. There is positive experience from the safety upgrading programme already carried out in the Bohunice NPP, where the initial leakage rates of 5000% and 7000% per day were reduced to values of 400% and 560% per day. Further improvements of confinement leaktightness are planned with the aim of reaching leak rates of 300% per day. In both types of plants the important factor in reducing leakages is the ability to detect existing leakages. This involves integral leak rate tests for the whole containment (confinement) at overpressures as high as possible, local leakage tests and appropriate methods for the interpretation of results. With the improvement of leaktightness the leak rate testing pressure can be increased, which makes it possible to reveal new locations of leaks and eventually repair them. The initial shortcomings are difficult to correct. The work in this area is being pursued and this report presents the actual status of regulatory framework, scientific and technical background for present practices and considerations of resulting situation. Further progress is needed, one of the issues being the unification of approaches to the definition of leakages in containments with high leakages rates, so that the data on leaktightness of reactor containments could be understood in the same way no matter which country they come from. The methods for leakage rate testing and leakage rate definitions are not the same in all countries operating WWER440 units. The differences are significant and influence strongly the leakage rates published for each reactor. The present publication, based on a consultants meeting organized in the week 812 May 1995 within the framework of the Extrabudgetary Programme on the Safety of NPPs with WWER reactors, presents in detail the actual methodology being used in different countries and compares it with the western practice. The methods used for containment/confinement leak rate testing in the countries operating WWER reactors are partly based on the old Soviet Union regulations, partly on western rules such as US ANSI or German KTA regulations, and partly on original work conducted individually by each country. As the containment/confinement leaktightness is of 10

11 high importance for the safety of the public and significantly influences the level of public acceptance of WWER plants, the exchange of experience concerning leakage testing methods should be encouraged and the best methods made available to all interested parties. The IAEA has supported the efforts of the countries operating WWER reactors, organizing meetings on containment and confinement performance of WWER440/213 and WWER440/230 NPPs [4] and on WWER440/230 confinement improvement options [5]. The problem of methods used for leak rate measurements in containments and confinements of WWER440 reactors was found to be important enough to be treated separately. At the request of the Nuclear Regulatory Authority of the Slovak Republic, the IAEA organized a meeting on the subject within the framework of the Extrabudgetary Programme on the Safety of WWER NPPs. The main objectives of the meeting were: to present the methods of leak rate measurements used in each country operating WWER NPPs; to compare the advantages and deficiencies of each method and to make proposals for their further improvement. The questionnaire for the meeting was developed by Slovak specialists on the basis of the experience gathered in leak rate testing in the Bohunice NPP. The participants from the countries operating WWER440 model 230 and model 213 NPPs prepared technical papers describing leak rate testing practices in their plants and providing direct answers to the questions formulated in the questionnaire. The first two days of the meeting were dedicated to the presentation of the papers, after which the meeting was divided into working groups which prepared the experts' positions on the following topics: requirements of the national regulatory authorities on confinement/containment leak rate testing methods; local leak rate test methods; integral leak rate test methods; structural integrity testing; methods of evaluating leak rate test results. The reports prepared by each working group were reviewed in plenary sessions and the final conclusions were discussed and agreed upon in the plenary meeting. This report describes the national requirements for containment leak rate testing and for integral tests in the countries operating WWER reactors and compares them with the requirements in force in the USA, Germany and France. It describes the essential elements of leak rate testing and leakage rate extrapolation methods, discusses in detail the differences in leakage rate definitions due to significant leakages in WWER containments and explains their consequences. The report addresses the methods of leak rate testing for local leakage and for individual tests of compartments. It reviews the integrated leak tests and describes in detail and analyses the main leak rate determination methods the time total method and momentary leakage rate determination method. Examples of national requirements for 11

12 instrumentation and accuracy of tests are provided. The procedures for structural integrity tests of WWER440 containments are described and examples of good practice given. Section 8 draws conclusions on all the topics discussed, based on the operational experience gathered in the countries operating WWER440 reactors and on joint expert discussions. The content of this report is aimed at operational staff conducting actual leakage and structural integrity tests of WWER440 containments, regulatory bodies evaluating the results of these tests, and national and international organizations active in the field of WWER safety evaluation REQUIREMENTS OF NATIONAL NUCLEAR REGULATORY AUTHORITIES Actual status The requirements for containment leakage testing in various countries are generally similar, aimed at periodical confirmation that leaktightness of the containment lies within the admissible limits. In general, the purposes of these inspections and tests can be summarized as follows: to verify the quality and integrity of hermetic boundary component joints; to detect leaks applying local leak tests (pressure drop in test volume, medium flow rate) and repair leaks; to detect leaks on the external and internal hermetic boundary via separate pressurization (underpressure creation in) of individual compartments and integral overpressure (underpressure) creation within the whole hermetic compartment system, and subsequently to repair leaks; to evaluate the approximate leak rate of hermetic compartments via control pressurization before the integrated leakage rate test (ILRT); to verify the structural integrity of walls and resistance of the liner against maximum design pressure, to detect and repair contingent deformations (or destructions); to evaluate the leakage rate from hermetic compartments in preoperational (or operational) conditions and to compare it with the allowable leakage rate value; to determine pressure (or temperature) in semiaccessible compartments (or air locks) in accident conditions; to compare the leakage rate of hermetic compartments evaluated during operational periodical ILRT with that evaluated during previous tests. To meet the above aims, the national nuclear regulatory authorities require that: (1) Leak tests in compliance with this Programme be conducted and evaluated solely by an organization authorized for such activities. (2) The authorized organization prove that sufficient readiness exists in the following fields: 12

13 methodology (prepared and verified test programmes); material and instrumentation (complete instrumentation for tests); personnel (technicians with certificates for test conduction). However, specific regulations are different for several areas. In particular, the test pressures required in periodical integral leakage testing are different, e.g. in the USA they are equal to full accident pressure P a [68], while in the Russian Federation they are equal to 0.05 MPa [9], which corresponds to 0.33 P a. The main national requirements in force in various countries for preoperational integral tests and leakage rate tests are shown in Table I. Worldwide, the general tendency is to increase test pressures, and not the contrary. As the standards currently in force in the Russian Federation require the periodical test overpressure to be 0.05 MPa (Section 8.3.8, [9]), and those in other countries are even more stringent (e.g. in Hungary: 0.07 MPa [10]), it is suggested that this value should not be decreased. In Table II the national requirements for periodical integral tests are presented, and in Table III the requirements for local tests. Local tests in which the measurement of outflow (pressure drop) in test volumes is done with a pressure gauge, or the inflow into the test volume is measured with a flowmeter, are called "type B tests", according to ANSI/ANS [8]. The isolation valves and hermetic pressure release valves are not locally leak tested. These valves are only factory tested, and some of them are tested during integral leakage tests for the purpose of leak identification. Such local tests are called "type C tests" [8] PROCESS OF PREPARATION AND APPROVAL OF THE TEST PROGRAMMES Current status In the Russian Federation the test programmes are of two types: general and plant specific. The process of their preparation and approval is shown in Table IV. hi the countries of central Europe, where the number of NPPs is much less, there are no general programmes, only plant specific programmes. The process of their preparation and acceptance is also shown in Table IV. In particular, in the Slovak Republic, the authorization for tests is acquired and certified in compliance with SUBP Ordinance 66/89 of the Code on National Technical Supervision over the Safety of Technical Equipment in the Field of Power Engineering ( 7 and 15) [16] and with CSKAE (former Czechoslovak Nuclear Regulatory Authority) Ordinance 436/90 of the Code on Quality Assurance for Selected Equipment with Regard to Nuclear Safety of Nuclear Facilities ( 11) [14]. The authorized organization, in accordance with the above regulations, should prove sufficient readiness in the following fields: methodology (prepared and verified test programmes); material and instrumentation (complete instrumentation for tests); personnel (technicians with certificates for test conduction). VUEZ (the Power Equipment Research Institute) of Tlmace meets the aforementioned conditions and holds licences valid in both the Slovak and Czech Republics for performing leak and structural integrity tests as well as for evaluating and documenting them. Text cont. on p

14 TABLE I. NATIONAL NUCLEAR REGULATORY AUTHORITY REQUIREMENTS FOR INTEGRAL TESTS USA GERMANY FRANCE (for single containment with steel liner) RUSSIA UKRAINE TYPES OF TESTS A integral B local C valve test INTEGRAL TESTS Document 10 CFR Pt50, App.J [6] ANSI/ANS 56.8/1987 [8] PREOPERATIONAL LEAKAGE RATE TESTS KTA /1979 [11] RCCG/88, Part 3 [12] PNAE G [9] PNAE G [9] Test method Absolute Absolute Absolute A.M. A.M. Reduced pressure Min. 0.5 P a 0.5 bar Intermediary step 0.1 MPa 0.5 bar 0.5 bar Peak pressure Pa Pa 1.15 Pa P a (1.5 bar) 3 P a (1.5 bar) Duration of tests Min. 8 h, 20 data sets 36 h, can be reduced to 18 h 24 h, can be reduced to 16 h A.M., 9 data sets A.M., 9 data sets Acceptance criteria L <0.75 U L AL <L a L AL <0.75 U and L e = 0.74 L, AL/L <0.5 AL/L <0.5 AL/L s0.25 AL/L <0.3 Lpenet <0.6 L, a = 0.95 a = 0.95 L AL <L a 4 L 4 AL il a 5 (this gives approximately L c = 0.74 L a ) Key: P., accident (design) pressure; A.M. any method approved by the regulatory body; N.S. not specified; L a permissible leak rate; Lpe ne i leak rate of penetrations; L c maximum acceptable leakage rate under test conditions; Te = 273 I 20 = 293 K; Me = 29 g/mole dry air; Ta = = 413 K; Ma = 21.5 g/mole, LOCA. Footnotes 1 to 5, see Table II.

15 TABLE I. (cont.) BULGARIA CZECH REPUBLIC HUNGARY SLOVAK REPUBLIC V213 V230 TYPES OF TESTS A integral B local C valve test : : : I INTEGRAL TESTS Regulations CSKAE/78 [13] Regulations 436CSKAE/90 [14] Regulations 66SUBP/89 [16] Document Based on KTA /79 [11] Regulations CSKAE/78 [13] Regulations 436CSKAE/90 [14] Regulations 76CUBP/89 [15] Programme No. P75 [10] PREOPERATIONAL LEAKAGE RATE TESTS Test method Absolute Absolute Absolute Absolute Absolute Reduced pressure 0.6 bar lbar 0.3 bar 0.5 bar(l bar) 0.5 bar Peak pressure Pa P. (1.5 bar) P a (1.5 bar) P. (1.5 bar) 1 bar Duration of tests 24 h 24 h 24 h 3h Acceptance criteria N.S. L AL <L,,(13%) L AL <L a (14.7%) L <L a (16.3%) N.A. Key: P a accident (design) pressure; A.M. any method approved by the regulatory body; N.S. not specified; L kl permissible leak rate; L^, leak rate of penetrations; L c maximum acceptable leakage rate under test conditions; Te = = 293 K; Me = 29 g/mole dry air; Ta = = 413 K; Ma = 21.5 g/mole, LOCA.

16 O\ TABLE II. NATIONAL NUCLEAR REGULATORY AUTHORITY REQUIREMENTS FOR PERIODICAL INTEGRAL TESTS PERIODICAL TESTING USA GERMANY FRANCE RUSSIA UKRAINE Pressure Pa 0.5 bar possibility of Pa Pa 0.5 bar 0.5 bar Frequency of tests 3/10 y 6 7 1/10 years I/year I/year Duration of tests Min. 8 h, 20 data sets 24 h, can be reduced to 10 h 24 h, can be reduced to 16 h Min. 9 data sets Min. 9 data sets Acceptance criteria L <0.75 L a L AL <L a L AL <0.75 L c L ii <L Lji <L; BULGARIA CZECH REPUBLIC HUNGARY SLOVAK REPUBLIC V213 V230 PERIODICAL TESTING Pressure 0.3 bar 0.5 bar Pa1/10 year 0.2 bar 0.7 bar 0.5 bar 0.2 bar Frequency of tests I/year I/year 0.2 bar I/year 0.7 bar 1/10 years I/year I/year Duration of tests 3h 6h 0.2 bar 6 h 0.7 bar 12 h 6h 3h Acceptance criteria <2500m 3 /h L <L a (13%) L <L a (14.7%) L <L a (16.3%) L it < L; Key: P a accident (design) pressure; A.M. any method approved by the regulatory body; N.S. not specified; L., permissible leak rate; L^.^i leak rate of penetrations; L o maximum acceptable leakage rate under test conditions; L r leak rate in year i, L i leak rate in year i 1, L,, u leakage measured in test i. Te = = 293 K; Me = 29 g/mole dry air; Ta = = 413 K; Ma = 21.5 g/mole, LOCA. 'In Russia and Ukraine, tests B and C are treated similarly. 2 L e = maximum allowable containment leakage rate under the leaktightness test conditions, expressed in % per 24 hour period. 3 In Kola NPP, Unit 3, P a = 1 bar owing to a twofold increase in the number of check valves between air traps and the volume above water trays. "For Kola 3, L a = 160 %/d, for Kola 4, L, = 25%/d. 5 For Rovno 1 and 2, L, = 98.7%/d. 6 In case of negative results and corrective work, the frequency of tests is increased to 1/18 months. 7 In Germany, recurrent tests of the integral leak rate shall be carried out at the end of each outage after completion of all maintenance and repair work which might change the tightness of the containment vessel.

17 TABLE III. NATIONAL NUCLEAR REGULATORY AUTHORITY REQUIREMENTS FOR LOCAL TESTS REQUIREMENTS FRANCE USA BULGARIA 1. Test pressure Not less than P a Not less than P a Not less than P a 2. Acceptance criteria (leak rate %/24 hours) Less than 0.6 L a for all type B and C penetrations Less 0.6 La for all type B and C penetrations Defined in design documentation TypeB Electrical penetrations Less than 0.01 L a for all penetrations Air locks Less than 0.01 L a for each Air lock Hatches Less than 0.01 L a for all hatches Blind flanges Less than 0.01 L a TypeC Less than L Defined in design documentation 3. Periodicity TypeB After refuelling or every 2 years After repair and periodically according to operational instructions Electrical penetrations Every month Air locks Every 6 months Every 6 months Hatches After closing or every 2 years Blind flanges After closing or every 2 years TypeC Every 2 years After refuelling or every 2 years After repair and periodically according to operational instructions

18 TABLE III. (cont.) REQUIREMENTS RUSSIA UKRAINE HUNGARY SLOVAK REPUBLIC 1. Test pressure Not less than P a Not less than P a Not less than P a Not less than P a 2. Acceptance criteria (leak rate %/24 hours) TypeB Defined in documentation design Defined in documentation design Defined in documentation design Defined in design documentation Electrical penetration Defined in documentation design Defined in documentation design Defined in documentation design Defined in design documentation Air locks Defined in documentation design Defined in documentation design Defined in documentation design Defined in design documentation Hatches Defined in documentation design Defined in documentation design Defined in NB5746 Defined in design documentation Blind flanges Defined in documentation design Defined in documentation design Defined in NB5746 Defined in design documentation TypeC Defined in documentation, design Defined in documentation design Defined in documentation design Defined in design documentation 3. Periodicity Types B and C After repair and periodically according to operational instructions After repair and periodically according to operational instructions After repair and periodically according to operational instructions Electrical penetration 3040%/year Air locks Every 2 years Hatches Blind flanges

19 TABLE IV. PREPARATION AND APPROVAL OF THE TEST PROGRAMMES TYPICAL PROGRAMME RUSSIA SLOVAK REPUBLIC HUNGARY UKRAINE BULGARIA Prepared by: Atomtechenergo 1 Not applicable 2 NPP staff As in Russia Empresarios Agrupados Energoproekt, Sofia Accepted by: Atomenergoproekt 3 Gidropress 4 Kurchatov Institute 5 Chief engineer NPP Kozloduy Confirmed by: Rosenergoatom 6 Approved by: GAN 7 Nuclear Safety Inspectorate BNSA 8 PLANT SPECIFIC PROCEDURE Prepared by: NPP staff NPP staff NPP staff Atomenergoexport (Russia) Eroterv Design Company VEIKI Research Institute for Electric Power NPP staff Energoproekt, Sofia NPP staff Confirmed by: Chief engineer of NPP Deputy director of operations Chief engineer Chief engineer of NPP Head of engineering, NPP support division Approved by: Local inspector of GAN NRA SR (central office) Nuclear Safety Inspectorate GAN BNSA 1 Support organization for NPP operation. 2 No typical programme exists. Such programmes were carried out only during the initial period of NPP commissioning. 3 NPP designer. 4 Reactor designer. 5 Leading scientific organization. 6 Operating organization of NPPs. 7 Regulatory authority. 8 Bulgarian Nuclear Safety Authority.

20 TABLE IV. (cont.) TEST Performed by: NPP staff NPP staff NPP staff NPP staff NPP staff Confirmed by: Chief engineer of NPP Head of the technical dept. Deputy director of operations Plant shift supervisor Chief engineer of NPP Head of engineering, NPP support division Approved by: Local inspector of GAN NRA SR (central office) 9 Nuclear Safety Inspectorate GAN BNSA DECISIONS Development of technical decisions concerning containment acceptance: Atomtechenergo Rosenergoatom & According to test According to test results 10 results Accepted by: Atomenergoproekt VNIIAES" Gidropress Kurchatov Institute Confirmed by: Rosenergoatom Approved by: GAN BNSA 9 Minor changes of equipment can be accepted by the local inspector. The local inspector performs field inspections during the test. 10 Fulfillment of acceptance criteria is stated in the test results. If they are fulfilled, further operation is allowed without additional documents. If they are not, operation is not allowed, corrections must be made and the test repeated. 11 Scientific institute for NPPs.

21 1.4. DEFINITION OF HERMETIC BOUNDARY The hermetic boundary for the WWER440/213 units is determined in the design and is the same for all units. The main hermetic zone of the WWER440/213 NPPs is designed for an overpressure of 150 kpa and consists of three main components: steam generator (SG) compartment; bubble condenser tower; vent centre. These three parts are mutually interconnected and separated from the environment by means of reinforced concrete walls lined with an hermetic liner. Furthermore, the hermetic zone also comprises associated hermetic compartments which are situated on the outer or inner sides of the main hermetic compartments. These hermetic 'subcompartments' are partially isolated from the main hermetic compartments during operation. They consist of: seven semiaccessible compartments; four air locks; four bubble condenser air traps; the reactor platform. A representative list of compartments is provided in Table V, based on the information given in [17]. Hermetic compartments of the WWER440/230 NPPs are even simpler than those of the WWER440/213 models and are designed for an overpressure of 100 kpa. The hermetic zone of the WWER440/230 consists of two main parts: the SG compartment; the recirculation vent system. These are complemented by associated hermetic compartments consisting of: two semiaccessible compartments; one air lock; the reactor compartment. We can distinguish: an external hermetic boundary, separating the main and associated hermetic compartments from the outside atmosphere; and an internal hermetic boundary, separating the main hermetic compartments from those associated with them. The equipment and components which form the external hermetic boundary are subject to leakage rate tests. Those forming the internal boundary are subject to less stringent requirements. Individual hermetic boundary components are verified applying local leak test methods and the hermetic boundary as a whole is tested using integrated leak test methods. 21

22 TABLE V. COMPARTMENTS INCLUDED IN THE CONTAINMENTS OF WWER 440/213 UNITS The leaktight volume is a rectangular compartment of a crosssection of 54 x 75 m, limited at the bottom by the lowest floor at a level of6.5 m and at the top by the ceiling at a level of m ELEVATION (in mm) COMPARTMENTS VOLUME (m 3 ) FLOOR CEILING Compartment of SGs. 9100* Drive compartments associated with SG compartments Venting room of SG compartment As above Corridor for release of gassteam mixture from the compartments of pumps and SGs into the tower of the bubbler condenser (up to but not including the water trays) Compartments of drives for the venting system of the SG compartment As above Compartments of heat exchangers of RCS blowdowns Compartments of filters of SWO Hydroaccumulator compartments Compartment of the pressurizer and its bubbler condenser Venting room of the pressurizer Compartment of drives for the venting system of the pressurizer a Compartments of the I&C for SGs and RCPs Air locks on all levels, except for compartments of RCPs and SGs Compartments of the recirculation system. 4x Reactor shaft The volume under the reactor cap Other air locks of the containment.** Bubbler condenser tower: (a) air volume above the water trays (b) volume of air traps * The volume of RCPs and SGs is given taking into account the air locks, cable penetrations and corridors adjoining them. Coefficient of filling up the volume with equipment K =

23 With regard to WWER440/230 units, the differences among the various NPPs are significant and no common definition of confinement boundary can be given. An example of the situation in the Kozloduy NPP is described in [18] LEAKAGE RATE TESTING AS THE BASIS FOR SAFETY ANALYSIS The leakage rate test conditions are different from those existing during accident conditions. The temperatures during accidents are higher and the fraction of steam in the air is significant. This has given rise to doubts over which leakage values those obtained in the experiment or some other values recalculated for accident conditions should be used in a safety analysis. The consultants confirmed that in all countries which operate WWER NPPs, the values found in the experiments are assumed to be valid for accident conditions. This approach is conservative, as illustrated by the formula given in the French regulations which expressly address this issue. The French regulation document, RCCG, Part 3 [12], gives an example for containments: where: L e is the maximum acceptable leakage rate under test conditions, dry air, 20 C La is the maximum acceptable leakage rate of airsteam mixture under LOCA conditions T e is the = 293 K, dry air Me is the 29 g/mole, dry air T a is the = 413 K, LOCA is the 21.5 g/mole, LOCA M a which gives Le = 0.74 La. This illustrates that the conditions of tests (room temperature and dry air) impose stricter requirements. In other words, leakage rates determined under test conditions are conservative EXTRAPOLATION OF LEAKAGE RATES OBTAINED AT REDUCED PRESSURE TESTS TO THE DESIGN PRESSURE LEVEL It is a common understanding among experts that the testing of the containment should be performed at the pressure levels up to the peak accident pressure (or design pressure), as required by various international standards. During subsequent tests at a reduced pressure level, the values of leakage rates, measured at reduced pressure, should be recalculated (extrapolated) to the peak accident (design) pressure. 23

24 However, it is known that a number of WWER440 units have not been tested at design pressure levels due to various reasons, hi the case of such units, the leakage rates measured at reduced pressure levels should be extrapolated to the full design pressure using the results of the previous tests under maximum achieved level of overpressure. The consultants believe that there is no universal theoretical relationship for the extrapolation of the results, since each unit has its own individual hydraulic characteristics of leakages. Therefore, in extrapolating the values of leakage rates measured at test pressures, the data experimentally obtained for the given unit should be used LEAK RATE DEFINITION The experts are of the opinion that the definitions of leakage rate stated in various western standards which use similar wording are in general applicable to the WWER440 NPPs. However, in view of the significant leakage rates of WWER440 bubbler condenser containments and WWER440/230 confinements, it is not possible to consider the test pressure to be constant during the 24 hour test. Therefore, the notion of leakage rate "at peak accident pressure Pa" given for instance in the American standards can be interpreted either: (1) as the leakage determined during the test in which the initial pressure was Pa, and then the pressure was dropping due to the containment leakage; or (2) as the momentary leakage rate in the time interval when the average pressure was equal to Pa. In practice both approaches are used in the countries operating WWER440 NPPs. The first approach is applied in Czech and Slovak Republics and in Hungary [1923,10], while the second is used in Bulgaria [18], in the Russian Federation [24] and in Ukraine [17]. The leakage rates quoted in the official documents are related to these two different approaches. In order to avoid misunderstandings, the experts recommend that the definition of the leak rate (for initial test pressure or for average test pressure) be given for each test result. The leak rate is usually defined as the loss of air in the containment during a 24 hour test, measured in per cent (%/day). hi reality the tests can be conducted for shorter time periods, especially when periodical testing is carried out. Then the leak rate is determined within a shorter time span and recalculated for the 24 hour test interval. The method of recalculation is especially important for tests in which the initial test pressure is taken as the peak accident condition. For small leakage rate values such as those observed in full pressure PWR containments, the decrease of containment pressure during testing will have little influence on the containment leak rate. For WWER440/213 containments with leakage rates of 1 20%/day, 24 hour tests can still reasonably be conducted and the results can be correlated to the initial test pressure. With respect to WWER440 confinements with higher leakage rates, e.g. reaching 100%/day or more, the approach of relating the measured leakage rate to the average pressure in the test interval is applied in Bulgaria, Russia and Ukraine. This is in conformity with Russian regulations which require that the measured leakage rate be assigned to the average test pressure within the measurement interval. 24

25 According to the information presented by Russian experts, the measurements are performed in such a way that the experiment starts with a pressure slightly higher than the calculated peak accident pressure. The leak rate is measured during the time interval when the pressure falls to a value slightly below the peak accident pressure. The average pressure value is determined during this interval, and the leakage values are used to determine the equivalent diameter of the opening in the containment. This procedure can be repeated for several points as the pressure drops. The value of the equivalent diameter of the opening is calculated for each of the average pressure values in each time interval. Therefore, the value of the leakage rate finally given for the plant, if it is measured in %/day, is the value measured for the time interval in which the average pressure value was the same as the peak accident pressure value. In the Slovak Republic the determination of the leakage rates for Bohunice VI was performed using another approach [1923]. It consists in predetermining the final pressure of the test, so that the leakages are measured for a time interval in which the pressure falls from the initial value PI to the final test value P2. The leakage rate at the accident pressure PI is defined as: L = 2400 *«I" [%/day] where: x H P Pr T L is the measurement interval [h], is the absolute pressure in hermetic compartments [Pa], is the water vapour partial pressure in hermetic compartments [Pa], is the temperature in hermetic compartments [K], is the leakage rate from hermetic compartments [ %/24 h]. Subscripts 1 and 2 designate the beginning and the end of measurement interval, respectively. The leakage rate calculated in such a way is always related to the initial pressure value, Pi, in hermetic compartments. More detailed information on the subject is provided in Section 5.6 below. The experts from the Slovak Republic explained that the leakage rates defined in this manner are used only for checking the progress of leaktightness improvements of the WWER440/230 confinements [18]. It is clear that, in view of the differences in the leak rate definition, they cannot be directly compared with the leakage rates quoted for Bulgarian, Russian or Ukrainian units. The French approach in the presence of significant leak rates consists in stabilizing the containment pressure at P a by entering a controlled flow of air from outside. The leakage rate is then equal to the injected flow, corrected for temperature and humidity. 25

26 2. LOCAL LEAK RATE TESTS Local leak rate tests are effective steps for the improvement of the leaktightness of confinement COMPONENTS OF THE HERMETIC BOUNDARY All components and equipment located at the boundary of the hermetic zone should be subject to local tests. It is clear that global leaktightness rests on the leaktightness of these components. With respect to testing methods and possibilities for testing, hermetic boundary components can be divided into several groups. The component groups for both WWER 440/230 and WWER440/213 type reactors are presented in Table VI. TABLE VI. COMPONENTS OF THE HERMETIC BOUNDARY EQUIPMENT 440/ / Liner / / 2. Pipe penetrations / / 3. Electrical penetrations / V 4. Hermetic doors, manholes V" / 5. Valve stem extensions / / 6. Hatches / 7. Isolation valves 1 V" 8. Sump drain valves / 9. Venting flaps / 10. Heating, ventilation, air conditioning elements / 11. Reactor dam seal 12. Refuelling pool gate / 13. Ionization chamber covers / 1 With respect to 440/213 model WWER plants, all isolation valves of heating, ventilation, air conditioning and all sump drain valves have been included in this item 7. Acceptance criteria for leaktightness Establishing the acceptance criteria is important in local leak tests. The acceptance criteria are defined depending on the construction of the components and test methods involved. The quantitative acceptance criteria were defined for the components provided with a control volume for pressure difference determination. In the original design of the WWER 440/213 the following criteria are defined: Overpressure test: pressure drop in the test volume shall not exceed 10 kpa/20 minutes and when soap bubble solution is applied to the joints, no bubbles shall appear on the tested surface at a test overpressure of 150 kpa. At the present time, for WWER 26

27 440/213 units no pressure decrease should be detected for electrical and pipe penetration (new criteria). Quantitative overpressure test: flow rate (equal to the flow rate through the leak) shall not exceed 0.03 mvh per 1 m of sealing (i.e., 30 L/h at a test overpressure of 150 kpa.). Underpressure test: no bubbles shall appear on the tested surface. Ultrasonic or acoustic tests: no air leakage shall be audible through the tested joints. Adhesive dye test: the imprint shall be clearly visible on at least on 60% of the sealing width. The acceptance criteria for WWER440/230 units are not fully defined in the original documentation. According to the consultants, the criteria for WWER440/213 plants should equally be applied for WWER440/230 plants. Test frequency The local test frequency for WWER440/230 and WWER440/213 plants is usually defined in the operational procedures. The local test frequency practiced at different units is illustrated in Table VII TEST METHODS With regard to the design of individual technological nodes, different methods of leak testing are applied to hermetic boundary components for the purpose of leak identification and local leak rate evaluation Local leak test methods for the purpose of leak identification Sensory check Applying the visual method, cracks from the width of 0.1 mm can be identified with good lighting (500 Lx). The acoustic method can be used to identify major leaks. This method is very effective but difficult to use as it causes loud and disturbing background noise. The tactile method also requires a pressure difference and direct access to the examined joint as well. Soap bubble method The most frequent and effective method of leak identification. Leakages of about 1.10" 2 (l/s)/pa can be detected. The soap solution used for detection shall have the proper characteristics. Underpressure (vacuum) method This method is used to test the liner field welds on flat compartment wall, floor or ceiling surfaces, welds of door frames and equipment hatches, i.e. all welds not equipped with their own test volumes. Leakages of about 1.10' 3 (l/s)/pa can be detected. 27

28 N 00 TABLE VII. LOCAL TEST FREQUENCY BULGARIA WWER440/230 SLOVAK REPUBLIC HUNGARY WWER440/213 RUSSIA UKRAINE WWER440/213 WWER 440/230 WWER 440/213 WWER 440/230 WWER 440/213 l. Liner After repair After repair After repair After repair After repair After repair After repair 2. Pipe penetrations /4 years or after repair 1/4 years or After repair 3. Electrical penetrations After repair After repair 30%/yearly 30 40% /yearly 0 1/4 years or after repair 1/4 years or After repair 4. Hermetic doors, 0 manholes 5. Valve stem extensions Hatches 7. Isolation valves of No No 0 0 technological systems 8. Sump drain valves No Venting flaps 10. HVAC elements 0 No Reactor dam seal No 12. Refuelling pool gate 13. Ionization chamber 0 No covers Key: Equipment is not present in this unit; 0 Periodicity is not clearly defined; No Tests are not performed. 0 0

29 Ultrasonic method This way of leak localization can be successfully applied during local leak testing of hatches and doors with single sealing as well as for the reactor cap test performed in the framework of individual compartment tests. Likewise, it is suitable for identification of major leaks in unaccessible hermetic compartment places (e.g. ceilings) from a considerable distance (1012 m). Imprint method This method is used only for the identification of leaks in hermetic joints which can be dismantled, e.g. joints with rubber sealings. Such sealings are on hermetic doors, manholes, equipment hatches or the paddle and reactor cap. Helium method This method is not used for local leak testing in WWER440 NPPs as it presents several disadvantages: the instrumentation is too complicated; the detector sensitivity is reduced rapidly; the sensitivity is too high for the purpose of local leak testing of the hermetic compartments of the WWER440/213 or WWER440/230 reactor types. Leakages of about 1.10" 5 (l/s)/pa can be detected. Gas mixture method This method is used for individual compartment tests. The goal of the test is to define defects in the civil structures and equipment. During the overpressurization of the individual compartments SF6 gas is injected into the air supplied. During the pressure drop measurement of tested volume, the SFs gas concentration is measured by taking samples from the surrounding compartments. The results allow to estimate the location of leakages and the value of leakage rates. Parallelism measurement This method is intended to measure the parallelism of the hermetic doors and their sealing surfaces as well as the flatness of hatches to assure equalization of pressure around all their edges Method of local leak rate evaluation The tests are conducted with the aim of detecting local leaks and measuring the leakage rate of hermetic components such as: electric and piping penetrations; doors and hatches with resilient seals or gaskets; liner welds with test volume; HVAC system valves. Measurement of outflow (pressure drop) in test volumes with a pressure gauge Pressure drop (from an overpressure of 150 kpa) measurement is a simple, reliable and sufficiently precise qualitative method allowing to evaluate joint quality and also to find leaks. 29

30 Measurement of inflow into the test volume with aflowmeter Measurement of inflow into the test volume is a quantitative method of local testing which, in addition to leak localization, can also determine leakage rate of the pressurization medium per unit of time. Supplied pressurization air compensates for the air lost through leaks, i.e. the measured flow rate corresponds to the leakage rate. A summary of the local test methods applied at different WWER440 units is provided in Table VIII CONCLUSIONS (1) Local leak rate tests are effective steps towards improving the WWER440 confinement leaktightness. (2) Local leak rate tests have to be performed periodically at all units according to national requirements and international practice. (3) Tests and inspections are performed after any job which could affect the confinement leaktightness. (4) For improvement of local test quality, unsuitable penetration elements should be modernized. 3. INDIVIDUAL COMPARTMENT TESTS The individual compartment tests performed at the Paks NPP are preoperational. In Russian and Slovak NPP units, individual compartment tests are performed periodically COMPARTMENTS TESTED Associate hermetic compartments (subcompartments), which are separated from the main hermetic compartments by an internal hermetic boundary, can be divided into: semiaccessible compartments and air locks; bubble condenser air traps in the case of WWER440/213 plants. Acceptance criteria exist currently only for Slovak NPPs and are plant specific as follows: semiaccessible compartments extrapolated pressure increase: Skoda type doors: dp/dt = 10 kpa after the first minute BLR type doors: dp/dt = 10 kpa after the second minute Air lock extrapolated pressure increase: dp/dt = 3 kpa after 1 hour 30

31 TABLE VIII. LOCAL TEST METHODS NO. METHOD TYPE BULGARIA 440/230 SLOVAK REPUBLIC HUNGARY 440/213 RUSSIA UKRAINE 440/ / / / / Visual 2. Acoustic 3. Tactile 4. Soap bubble 5. Vacuum 6. Ultrasonic 7. Imprint 8. Helium freon used 9. Gas mixture 10. Pressure drop measurement 11. Inflow rate measurement 12. Parallelism measurement

32 3.2. METHODS OF TESTING Three basic test methods are used for individual compartment tests: (1) Individual compartments tests with overpressure: overpressure created in the compartments tested, and overpressure created around the compartment tested. (2) Individual compartment tests with underpressure. (3) Individual tests of compartments with the gas mixture injection method. The purpose of the overpressure test is to: identify leaks on the external boundary of the given compartment and to repair them; determine extrapolated pressure increase in semiaccessible compartments and air locks with regard to personnel safety during operation (used in Czech and Slovak Republics); determine maximum duration of personnel stay in an air lock in accident conditions and leaks greater than those stated by criteria for semiaccessible compartment tests (used only in Czech and Slovak Republics). Underpressure in compartments is used primarily to detect leaks on walls of those compartments with an internal hermetic liner. Individual tests with additional gas injection point up the defects which cannot be detected by other methods (used only in Russia) EVALUATION METHODS OF SEMIACCESSIBLE COMPARTMENT TESTS The evaluation method can be used only if results of inflow (outflow) measurements of the given compartment are available. The extrapolated overpressure increase, dp/dt, is calculated directly from the measured (or corrected) values of overpressure increase (or drop) and compared to the respective criterion (used in the Czech and Slovak Republics). The results of individual compartment tests consist of: tables of measured and corrected values of overpressure drop (outflow) in semiaccessible hermetic compartments (air lock); plots of measured and extrapolated values; protocols on outflow evaluation within the time limit; a time limitation for the personnel stay in an air lock in case of nonconformance with the basic condition (3 kpa/lst hour); a list of leaks and defects identified during semiaccessible compartment tests. As to time limitation for the stay of personnel in an air lock, this requirement shall be included in the operating instructions covering the next operational period for the given unit. 32

33 3.4. RESULTS OF INDIVIDUAL COMPARTMENT TESTS There has been a demand to improve the leaktightness of the semiaccessible compartments at the Bohunice2 NPP. The leaktightness improvement applies mainly to the resealing of cable gland penetrations by special foam sealing compounds. The tests of this NPP semiaccessible compartment have been conducted by VUJE Trnava. The measurement values have been virtually stable for several years, therefore the frequency of tests in the Slovak Republic has been decreased to one every two years [20, 21] AIMS 4. INTEGRATED LEAK TEST FOR LEAK IDENTIFICATION This type of test represents, in practice, the first stage of integrated leakage rate tests implemented during preoperational and sometimes operational periods at an NPP unit. The aim of the test is: (a) (b) (c) (d) to identify leaks on the external hermetic boundary of hermetic compartments, from outside and inside, before the preoperational (or periodical) ILRT; to evaluate approximately leakage rates from hermetic compartments; to verify the function of the measuring and assessing system, installed and planned to be used for ILRT; to identify quantitatively the distribution of the integral leakages over particular segments around the hermetic boundary. Comments The procedures outlined in (a) and (b) are performed in all NPPs; (c) is performed only if the system is already installed (there are also other options for doing verification). Although currently (d) is not being performed, further efforts towards the effective implementation of this practice are desirable. The following test procedures might be helpful: mixing SF6 into pressurized air in the hermetic zone (HZ) and measuring its contents in surrounding rooms (areas) by e.g. mass spectrometry of samplings; mixing coloured gas into pressurized air in the HZ and its visual registration from outside of the HZ; interim hermetization of rooms surrounding the HZ and measuring inflow into them METHODS OF ESTABLISHING PRESSURE DIFFERENCE The aim of the integrated leak test (ILT) can be achieved by performing it either at overpressure or at underpressure in the HZ. 33

34 Overpressure The method of overpressure ILT is based on global (integral) pressurization of all hermetic compartments, i.e. on the creation of an initial overpressure within the hermetic compartments compared to the outside atmosphere. The gradual decrease of this overpressure is then used for leak identification on the hermetic boundary and for the first approximation of leakage rate from hermetic compartments. To create overpressure in hermetic compartments, low pressure compressor plant and/or inflow ventilation system(s) are used. According to good practice, already implemented in some WWER440/230 plants, the leaktightness of their confinements should be improved in such a way that the ILT for leak identification at overpressure can be performed by creating this overpressure with the aid of the compressor station only Underpressure Vacuum conditions in the hermetic compartments as a whole enable the localization of leaks on the hermetic boundary from the inside. Underpressure is created by using the suction ventilation system(s). According to good practice, an additional vacuum system (e.g. serial connected blowers) can be installed to create appropriate underpressure in the tested hermetic compartments METHODS OF LEAK IDENTIFICATION AND LEAKAGE RATE DETERMINATION As follows from the above, leaks are identified: at overpressure on the hermetic boundary from outside; at underpressure on the hermetic boundary from inside. It is useful to conduct both tests immediately in succession, regardless of their sequence. The benefits of proceeding in this manner are as follows: leaks are identified by the same leak identification inspection teams; preparation work and the total duration of ILT can be reduced; results of leak detection are available with regard to both sides of the hermetic boundary, which reduces the time period necessary for leak repair. Leak identification is assured by an appropriate number of inspection teams. All members of inspection teams shall be completely familiar with the ILT programme and especially the activities of their own team. The operating organization therefore organizes special theoretical and practical training for them. Inspection teams are provided with the necessary equipment (sprayers with soap solution, brushes, ladders, flashlights, notepads, etc.). 34

35 Activities of inspection teams The aim of these teams, applying suitable methods, is to identify, mark and register existing leaks on the hermetic boundary. Soap bubble and ultrasonic methods are applied to all accessible hermetic boundary components, namely: suspicious (damaged, improperly welded) liner joints; all plugs on nozzles of liner and hermetic penetration test volumes; glands of electrical penetrations (spraying in short intervals); active drain valves; all sealings on the hermetic boundary. Leakage rate determination Simultaneously to or after leak identification, an approximate determination of leakage rate is performed. If the temperature and humidity changes are not considered, the formula for leakage rate approximation is as follows: 2400.,, [%/day] where: L P i TH [h] Pi [kpaabs] P2 [kpaabs] is the leakage rate at pressure Pi (initial pressure during the test), is the measurement period, is the initial pressure in hermetic compartments, is the final pressure in hermetic compartments. For all other aspects, see Section 5 on ILRT RESULTS AND EXPERIENCE Integrated leak testing for the purpose of leak identification is evaluated based on: results of leak identification as a basis for repair work; comparison of measured leakage rates to the values acquired during the previous (if any) overpressure tests of this type. ILTs for the purpose of leak identification have been performed before or during preoperational ILRT on all NPPs. At some NPPs ILTs are being performed at the very beginning of operational outages, and ILRTs in the final stage of such outages. The heightened practical importance of ILT for leak identification is illustrated by the effects of hermetization work on NPPs with WWER440/230 units. At all NPPs where this type of test was conducted by applying over and underpressure, defects were repaired either fully or partially using available repair techniques, so that the leakage rate was improved or at least maintained at the previous level. The approximate 35

36 leakage rate evaluation method was also used and the results obtained compared against the results of previous tests or expected (allowable) values. The most difficult problems of leak identification arise when leaks and flaws which are not physically accessible (e.g. covered by concrete on both sides or situated in narrow ventilation shafts) must be located. Further efforts of all involved towards solving these problems are needed, especially if the methods applied up to now have shown negligible results. 5. PREOPERATIONAL INTEGRATED LEAK RATE TESTS Preoperational integrated leak rate testing represents the final stage of preoperational tests of hermetic compartments, and can be carried out following completion of local leak tests, individual tests, ILT for the purpose of leak identification and structural integrity tests AIMS AND PROCEDURE OF PREOPERATIONAL ILRT The aim of preoperational ILRTs is to determine leak rates from hermetic compartments at design initial overpressure and at newly determined initial overpressures. Leakage rate determination is based on pressure, temperature and humidity measurements within hermetic compartments, as described below. The testing procedure is as follows: All partial hermetic compartments are connected to one another and all openings on the external hermetic boundary, i.e. doors, hatches, valves, flaps, etc. are set to the "operation at nominal pressure" mode according to the Operating Instructions. After pressurization by means of a low pressure compressor plant, the atmosphere in the hermetic compartments is stabilized for a certain period of time. Then the pressure source is disconnected and the measurements start by recording barometric pressure and absolute pressure, mean temperature and mean humidity values in the hermetic compartment atmosphere. The values of temperature and humidity, from several sensors, are weighted by volume. Readings are repeated at regular intervals and the leakage rate from hermetic compartments is evaluated after each measuring cycle in compliance with the calculational algorithm. The leakage rate is determined by applying the socalled absolute method, which is an application of equation of state. This method requires precise temperature, pressure and humidity measurements within the hermetic compartments at the beginning and at the end of the reading interval. It is assumed that the hermetic compartment volume is not changed during the test, i.e., the impact of temperature and pressure variations on volume can be neglected. Acceptance criterion The national nuclear regulatory authorities or other responsible bodies determine for each NPP the maximum allowable leakage rate from hermetic compartments into the environment, La, which corresponds to the peak accident pressure Pa. 36

37 There is a certain uncertainty, AL, in establishing the actual leakage rate. This factor is added to the leakage rate Lm determined through measurements and calculations, and the sum of both is then compared to the permissible leak rate La. Lm AL <La 5.2. COMPARTMENTS TESTED AND VOLUMES IN HERMETIC COMPARTMENTS A representative list of compartments tested is provided in Table V. Free volumes should be determined individually for all units by measuring the real geometry SYSTEMS TESTED DURING PREOPERATIONAL ILRT The following systems or auxiliary systems are subject to ILRTs: supporting structures of hermetic compartments and components assuring their leaktightness; all valves specified in ILRT programme; low pressure compressor plant; low pressure air distribution (pressurization); depressurization system; intake system for refuelling; exhaust system for refuelling; I&C in hermetic compartments; high pressure air distribution to fast acting valves; active drains; system of aerosol sampling; complete measurement and evaluation system according to specification in preoperational ILRT programme SAFETY MEASURES OBSERVED UNDER ILRT CONDITIONS The scope of the safety measures is dependent on the overpressure level. Generally: access to and work by personnel in the hermetic compartments is prohibited; activities not related to the ILRT on the outer side of the hermetic boundary are prohibited; during a test overpressure, the persons involved (inspection teams) are allowed to enter the external hermetic boundary only after a stabilization period of one hour; the restricted zone shall be designated with warning signs; the persons involved shall be familiar with safety instructions and shall participate in a special training; to prevent damage, some equipment and instrumentation shall be protected against overpressure impact. Assurance of fire fighting, health protection and safety measures as well as protection of instrumentation shall be documented with protocols. 37

38 5.5. PREOPERATIONAL ILRT PROCEDURE SCHEDULE A preoperational ILRT is conducted as a final stage of hermetic compartment tests before putting the NPP unit into operation. It can be conducted either separately or in combination with the structural integrity test Preoperational ILRT preparation The low pressure compressor plant and/or mobile compressors are used to pressurize hermetic compartments during the ILRT. Preparation for pressurization consists in setting the valve in pressurization line, checking the compressor plant and preventing unnecessary air extraction. Setting of hermetic boundary components With regard to leaktightness, the most important stage of preparatory works before ILRT commencement is assurance of the prescribed initial state of components related to the hermetic boundary. Establishment of ILRT checking points During the preparatory phase, the NPP operating organization assures the establishment of the following ILRT checking points: ILRT supervision; pressure measurement; ILRT evaluation point; pressurization point; depressurization point. Preparation of instrumentation The following instrumentation for measurements and evaluation should have valid certificates and should be installed or activated prior to the tests by the organization authorized for ILRT implementation: installation of pressure measurement and pressure gauges to the pressure measuring line; installation of temperature and humidity sensors in hermetic compartments; installation of measuring, control and evaluation instrumentation; verification of all sensors using a verification computer programme or other calibration procedures; activation of the main calculational programme and its verification without overpressure in hermetic compartments Pressurization of hermetic compartments The pressurization of hermetic compartments begins when the compressor plant is started and the valves in the pressurization line are opened. Other valves in the pressurization line are open or closed as preset in the preparatory stage. 38

39 The ILRT pressurization point is attended by technicians who are permanently in contact with pressure measurement point (telephone, transmitter) and pressure increase in hermetic compartments is maintained as a maximum at 20 kpa/h. After the test overpressure has been reached, pressurization is shut down with a valve. The period of stabilization follows, as described in the following section Stabilization of parameters in hermetic compartments During stabilization: voids in heat insulation, concrete and technological equipment are filled with air; the air temperature is equalized with the temperature of various structures in the hermetic compartments. Overpressure variations are eliminated during a 34 hour stabilization period by: additional pressurization (through the pressurization line), or depressurization. Air parameters in hermetic compartments are considered to be stabilized if overpressure drop per 1 h is stabilized at values which do not differ from one another by more than 20%. The stabilization period generally lasts from 1 h to 6 h Leakage rate measurements and evaluation After stabilization has been finished, the first stage of air parameter measurements starts for the purpose of leakage rate evaluation. All air parameters (pressures, temperature, humidity) in hermetic compartments are recorded automatically. After completion of the measurements in a respective measurement interval, the measured values are processed, the leakage rate from hermetic compartments is calculated and all the measured (averaged) values and output data are recorded on a printed protocol Depressurization This procedure is conducted at the end of each test but also between individual test steps. The atmosphere in hermetic compartments is depressurized through the isolation valves of the ventilation systems PERIODIC LEAKTIGHTNESS TESTS Leak rate determination Total time method The total leak rate of the containment is determined as follows: 39

40 L_ 100 \ l T,T, [, 2 2 j where: L is the total leak rate, %/d, P is the absolute pressure, (Pa), T is the temperature, K, R is the gas constant, J/(kg K), x is the time, d; indices: 1 is the start of test, 2 is the end of test. Intermediate measurements are used to estimate the leakage rate and to determine errors (systematic and statistical) of measurements. The hermeticity tests should be executed upon confirmation by the commission of full preparedness of confinement and receipt of permit from the supervising body of the operating utility [14, 15, 17]. In the Slovak Republic, the absolute method of measurement and leakage rate evaluation is used. It consists in precise measurements and calculation of state variables in hermetic compartments, which are substituted during evaluation into the equation of state in the form shown in Section 1.6, namely: ?H 1 ' ' r^/^ i [%/day] where: XH x XH P Pr T L is the T2 TI is the real time [h, min, s], is the measurement interval [h], is the absolute pressure in hermetic compartments [Pa], is the water vapour partial pressure in hermetic compartments [Pa], is the temperature in hermetic compartments [K], is the leakage rate from hermetic compartments [ %/24 h]. Subscripts 1 and 2 designate the beginning and the end of measurement interval, respectively. The leakage rate calculated in such a way is always related to the initial value of pressure, Pi, in hermetic compartments. 40 The measurement of quantities x, P, Pr and T is described in the following sections.

41 Momentary leakage rate determination method According to the requirements of Russian rules, the leakage rate is determined as follows [24]: G where: M is the mass of gas in the containment, in kg, G is the dm/dt. Therefore, the leakage rate is that of leakage which occurs during a specified period of time. The values of leak rates in WWER440/213 containments are calculated using the following procedure: For each moment of time the values of an intermediate parameter L are calculated: /, = In P, P' Rt Rj Ti Tj where: Pi, Pi are absolute values of pressure index " 1" is the value at the beginning of the test; i = 2 n, where n is the number of measurements of parameters inside containment. The overall number of measurements of parameters n after their stabilization should not be less than 9. The set of values L are processed according to the least squares method to find a linear dependence: /, = a b%i where: Ti is the current time, hour. The values of factors a and b can be derived from the following set of equations: a ~ 41

42 n n 2r, (L?iJ where: n is the number of measurements of parameters. The value of leak rate is calculated with the formula: L = 100 (I M) where: I is the a 24b; AI is the total error of determination of the value of a complex parameter I The statistical error of the determination of the value of the complex I is calculated with the formula: /. = tsj Mn n(t 0 r cp ) 2 where: jl(a br i hf s = i ~ 2 5,, =sj± where: TO Tav t is the time for which is a confidence interval determined is the average time is the Studentt distribution coefficient, determined on the basis of the number of measurements, and of the value of confidence level a >0.95. The systematic error ALys can be evaluated as: AI, VV (A? AT 42

43 where: AP, AR, AT are systematic errors. The influence of the total error AI on the leakage rate can then be evaluated as follows: if ALys/Ssi <0.8, then the systematic error can be neglected and the total error is equal to AL, ALi = t SM; if ALys/Sst >8, then the statistical error can be neglected and the total error is equal to Alsys; if 0.8 < ALys/Sst <8, then the total error is given with the formula: A/ = k, S s = J, 7 (Sst t V 3k~> with: k2 coefficient dependent upon the confidence interval (a = 0.95 fo = 1.1). ki coefficient resulting from the formula: A 7 ' AI, A/,. v. v 3kl The values of a complex I which do not satisfy the inequality (a b TI) < 2.3 S should be excluded from consideration as rough misses. Extrapolation of leakage rate in compliance with US rules The extrapolation method by means of extrapolation factors was developed at the Oak Ridge National Laboratory [25]. According to this method, the total leakage value from hermetic compartments is a combination of various types of 'flow through' leaks. Formulas for 'flow through' leaks have been developed based on thermodynamic laws and research of flowing gases and vapours. Since leakage from hermetic compartments is expressed in terms of a weight (or a percentage) proportion which leaks out per unit of time, formulas for various types of flow define the mass flow of the leaking medium. 43

44 The type of flow through a certain leak, and thus the character of leakage, depend on several factors, the most important being: pressure (overpressure) and its critical values; the shape of the leak path; the type of leaking medium (composition of gaseous mixture). In turn, these factors determine the type of flow: laminar flow; molecular flow; turbulent flow; outflow dirough an orifice. Formulas for all basic types of flow are given in Section of Ref. [25]. During hermetic compartment leak tests at reduced overpressure, it is necessary to know the relationship between pressure and leakage rate for various types of leaks which can occur, i.e. to know the formulas defining leakage rate as a function of pressure. Equations of leakage for all main flow types are provided in Section of Ref. [25], which expresses leakage with laminar flow as: L L = K L (P, Pi, / P,) where: Pat Pi KL is the atmospheric pressure = 0.1 MPa, is the test pressure [MPa] (initial pressure of measurements), is the proportionality coefficient for laminar flow. Extrapolation coefficients for individual flow types When leakage formulas for all basic flow types are known based on the definition of extrapolation factor, extrapolation coefficients for individual flow types can be derived. The extrapolation factor can be defined as follows: e f = L e p i The extrapolation factor is a dimensionless ratio of leakage rate at extrapolation pressure Pe [MPa] to the leakage rate at reduced test pressure Pi [MPa]. Extrapolation coefficients for individual flow types are obtained by substituting the leakage rate ratio as a function of pressure in the relationship for pressure extrapolation. Formulas of extrapolation coefficients for all basic flow types are given in Section of Ref. [25], which expresses the extrapolation coefficient for laminar flow as: 44

45 el p, Pi / p, The extrapolation coefficient for critical outflow through an orifice is defined by: (a)?z, Pe <0.19MPa n (P / P f m j 0 5 [1 ( fgt / re ) J r p. jo.7!s II I r / r^ i I (b) Pz < 0.19 MPa, Pe > 0.19 MPa (P u, / P z ) 0J!5 _ [1 (P M (c) Pz > 0.19 MPa, Pe > 0.19 MPA e K = 1 Determination of the extrapolation factor Extrapolation coefficients enable the definition of factors influencing a measured leakage rate when it is extrapolated to a higher pressure than the test pressure (e.g. to accident pressure), i.e. to determine the extrapolation factor. The extrapolation factor can be determined by means of two methods: (1) The first method is based on the assumption that it is impossible to conduct additional tests at various pressure levels. In this case, the only option is to assume a flow type through leaks which offers conservative extrapolated values of leakage rate (e.g. the laminar flow, which is the most conservative). In this case, the extrapolation factor would equal the extrapolation coefficient of the laminar flow (extrapolation in compliance with DIN). (2) The second method assumes implementation of several tests at various pressure levels. When analysing measured leakage rates by means of extrapolation coefficients of individual flow types, it is possible to select extrapolation coefficients of those flow types, which are closest to real extrapolation factors for tests implemented at various pressures. To determine the extrapolation factor in this study, the latter method has been used. The extrapolation factor has been calculated based on analyses of leakage rates measured in WWER440/213 units at the Bohunice and Dukovany NPPs at pressure levels of 50 and 45

46 100 kpa. From leakage rates measured at the above overpressures, real extrapolation factors have been determined. Based on analyses, it has been found that the flow character practically does not change within the whole range of pressures (30 to 150 kpa) and the decisive combination is laminar flow and outflow through an orifice. A comparison of the extrapolation curves for various flow types with the real extrapolation curve shows that selection of extrapolation according to the laminar flow would be overly conservative, and therefore to determine the extrapolation factor, a proportion of both flow types was calculated for the resultant leakage. The calculation was based on the solution of two equations with two unknowns: K(o) K (L) = L 50 E(o) x K(o) E(L) x K(L) = Lioo where: K(o) is the proportion of outflow through an orifice [%/day], K(L) is the proportion of the laminar flow [%/day], E(o) is the extrapolation coefficient for outflow through an orifice for extrapolation from 50 to 100 kpa, E(L) is the extrapolation coefficient for the laminar flow for extrapolation from 50 to 100 kpa, L50, Lioo are the real leakage rates measured. Solving these equations for leakage rate values measured at the Dukovany and Bohunice2 NPPs, the proportion of the laminar flow has been determined to be Kl = and that of outflow through an orifice Kv = when extrapolating from an overpressure of50kpatol50kpa[17, 18]. The resultant formula of the extrapolation factor for extrapolation to a higher pressure (e.g. design pressure) of leakage rates measured at a reduced pressure is: er = x E(L) x E(o) During preoperational ILRTs at the Bohunice and Dukovany NPPs, the suitability of individual extrapolation coefficients and their proportion in the total extrapolation factor were verified. In Hungary, extrapolation is used to recalculate the leak rate measured at reduced pressure during periodical testing to leak rate at peak pressure. The basis for recalculation is the experimental leakage rate plot of the unit measured at different pressure levels, including the peak pressure during preoperational tests Pressure measurements Absolute (or relative) pressure in hermetic compartments is measured by way of an impulse pipe leading from the SG compartment to the measurement room. All the instrumentation for pressure measurements in hermetic compartments is connected to the impulse pipe. 46

47 In the Bohunice NPP, the main measuring instrument is the MENSOR quartz digital pressure gauge with a precise absolute pressure sensor and digital output to the computer interface. The pressure measurement gauge parameters are as follows: Measurements of absolute pressure in hermetic compartments and atmospheric pressure instrument type measurement range precision quartz digital pressure gauge MENSOR, DPII 0 to 700 kpa ±0.025% of the scale Measurements of overpressure in hermetic compartments instrument tensometric digital pressure gauge type DIPTRON 3 measurement range 0 to 300 kpa precision ±0.05% of the scale In the Russian Federation, the measurement of overpressure during tests is performed with type MO250 Mod pressure gauges of range of 0100 kpa which satisfy the following requirements: in the range of pressures 0 to 1.15 Pop, (Pop overpressure for operating test) the class of accuracy is lower than The gauges for atmospheric pressure measurement satisfy the following requirements: in the range of measurement 0.09 to MPa the class of accuracy is Temperature measurements In the Bohunice NPP, the remote atmospheric temperature measurements in hermetic compartments are assured by automatic measuring and evaluation systems (AMVS, see below) provided with a PERIS calculational programme for PC/AT. The system for air temperature measurements in hermetic compartments of WWER 440/213 units consists of 39 dry bulb thermometers with the following parameters: instrument resistance thermometer, Pt type Pt 100 measurement range 200 to 350 C precision ±0.5 C In an automatic measuring cycle, AMVS enables the measurement of temperature by means of sensors distributed within hermetic compartments. 47

48 The average temperature is determined by the formula: where: m Vj[m 3 ] is the partial volume of hermetic compartments allocated to the j" 1 temperature sensor, V is the gas volume of containment, free from the equipment and building structures (V = IVJ), m is the number of temperature sensors, Tj[K] is the temperature at the j* sensor measured 10 times consecutively within the measuring cycle. Sixteen thermometers are distributed in the WWER440/230 hermetic compartments. Temperature sensors are distributed in hermetic compartments in such a way that a certain theoretical partial volume is allocated to each sensor. In the Russian Federation the measurements are made with Pt 100 TSP0879 resistance thermometers. Their measurement range is from 0 to 100 C, accuracy 0.2 C, resolution 0.01 C. There are ten such thermometers installed in the containment. The average temperature is determined by the formula: T = m V and the measurements should satisfy the following requirements: no temperature sensors are installed in compartments of a volume of less than 200 m 3 ; one temperature sensor is installed in compartments of a volume of between 200 and 700 m 3 ; two temperature sensors are installed in compartments of a height exceeding 5 m, one at every 5 m height level; temperature sensors are installed in compartments of a volume exceeding 700 m 3, at a rate of one per 700 m 3 of volume at compartment height intervals of 5 m [24]. 48

49 Air humidity measurements and water vapour partial pressure determination in hermetic compartments In the Bohunice NPP, in addition to air temperature, the relative air humidity is measured to determine water vapour partial pressure in the hermetic compartment atmosphere. The automatic measurement and evaluation system senses relative humidity from humidity sensors with the following parameters: instrument: capacity humidity detector and thermometer; type: HTC781; measurement range: 0 to 100% (20 to 60 C); precision: ±1.5% relative humidity. For the mean relative humidity: 1 ^ V i 10 1 where: n V = I V k 1 VK [m 3 ] n [] (j)k [%] is the partial volume allocated to k* humidity, is the number of humidity sensors, is the relative humidity of air on k* sensor measured 10 times consecutively within the measuring cycle. The system for measuring air humidity in the WWER440/213 hermetic compartments consists of 7 capacitance humidity sensors. In the WWER440/230 models, five humidity sensors are used. Water vapour partial pressure is determined from the relative humidity applying the Vesper polynomial function. In the Russian Federation the value of gas constant in volume Vj is calculated with the formula: Rj = (R a dj R x ) where: 1 1 Ra and Rs are the gas constants of air and steam, respectively, J/kg x K; dj is the humidity of air in Vj, kg/kg. One humidity indicator per 1000 m 3 of free volume should be installed. It is supposed that the average value of gas constant will be defined with the formula: 49

50 V where: V is the free volume of the containment, (V = SVj). At measurement of relative humidity inside containment the indicators should satisfy the following requirements: in the range of measurement of relative humidity 0% to 100%, the absolute error should not exceed 3 % Measuring and evaluation system In the Bohunice NPP, leakage rate measurement and evaluation is assured by the measuring and evaluation system. This system enables the conduct of all measurements and calculations without operator intervention, which eliminates human factor impacts on measurement results and leakage rate evaluation during the measurement itself. The measuring and evaluation system consists of: 46 temperature and humidity sensors; stable cabling for temperature and humidity sensors in hermetic compartments; 8 distribution boxes in hermetic compartments; stable switchboard in the measurement room; measuring centre; cabling to the switchboard; stable impulse piping for pressure measurements in hermetic compartments; absolute pressure gauge; overpressure gauge; computer with a monitor; printer. The accuracy of measurements in an automatic measuring and evaluation cycle is assured through precision, reliability and replacement of contingent defective temperature or humidity sensors. The programme for automatic data acquisition, data processing and leakage rate calculation also includes operations enabling verification of data from individual sensors within a predetermined tolerance. The total measuring error, Ler, is calculated according to the formula: r i l/2 K = ^ 2[A) 2 (^) 2 A) 2 ] [%/day] 50

51 where: P T P P x is the absolute pressure at measurement point, is the absolute temperature at measurement point, is the vapour partial pressure at measurement point, is the time, hours. The error in sensing of individual parameters is given by a general relationship: //2 s = * *~~ n where: 8sn is the error in pressure (temperature, humidity) sensing, 6 P r is the error in data transfer ( P, PP, T), n is the number of sensors to measure the given quantity Hermetic boundary examination and leak During the pressure stabilization in hermetic compartments, an inspection of the hermetic boundary is conducted from the outside. The inspection for the purpose of leak identification is carried out in the same way as the inspection performed during an ILT for the purpose of leak identification (see Section 4.3). In the case of tests at the full overpressure of 150 kpa (80 kpa at the Bohunice1 NPP) in addition to leak identification, the so called supervision of the state of the hermetic boundary is assured. This activity is described in more detail in Section 7.3 below PREOPERATIONAL ILRT RESULTS AND EXPERIENCE ACQUIRED WHEN IMPLEMENTING THE TESTS To evaluate the leakage rate, the leakage rate value is determined in the total time method at the end of measurements in the 24th hour of measurements. Usually a measurement interval of x = 30 minutes is applied and 49 measurements are taken. Continuously measured leakage rate values in short intervals (i.e. point to point) and in total time (from the beginning of measurements) provide a general picture of leakage rate measurements, mainly on air parameter stabilization in hermetic compartments. When evaluating the leaktightness of hermetic compartments as a whole, the main criterion of validity is that the leakage rate value measured at accident overpressure, including positive measuring error, shall not exceed the acceptance criterion expressed as: L C r < i'dov 51

52 for WWER440/213 NPPs, or L _i_ T ^ X MAX SO "" Lt C r < JUDOV for WWER440/230 NPPs. When measuring at a reduced overpressure level with a corresponding pressure extrapolation (as described in Section above) the above criterion shall be met by the extrapolated leakage rate: T, T ^ rmax Conditions for preoperational ILRT programme application The differences in WWER ILRT programmes vis a vis the requirements of App. J to 10 CFR50 result from the principal differences between the hermetic compartments of WWER440/213 and 230 NPPs and the classical containments of PWR reactors in the USA and western Europe. The differences are: Unlike the single cylindrical or spherical shape of western containments, WWER confinements consist of a system of rectangular compartments mutually interconnected by means of shafts and corridors; The WWER440/213 confinement consists of three main hermetic compartments: the SG compartment, the vent centre and the bubble condenser; A hermetic liner is placed on the internal walls of the WWER440/213 confinement; it also passes through walls and confines the SG compartment from outside. In floors and ceilings it is covered with concrete; In addition to air locks and penetrations, the WWER confinement is fitted with equipment hatches in the reactor hall where other hermetizing components are situated such as the reactor cap and refuelling pool paddle (and, in the case of the 440/230, flaps), all of which are potential sources of leaks; The WWER440/230 confinement air locks are not designed as compact hermetizing units but they represent separate rooms with internal hermetic liners; hermetic doors are located not only on inlets and outlets but also between adjacent semiaccessible compartments; At WWER440/230 plants, the socalled gland penetrations are a major source of leakages. All these differences represent potential sources of leakages, and the leak rates of WWER440 hermetic compartments are some 100 times higher than the required values in PWR containments. The absolute (or reduced) pressure drop history in hermetic compartments, with such leakages, cannot be considered as linear and when evaluating leakage rates, western procedures of linear regression (mass point method in compliance with ANSI/ANS or reduced pressure method in compliance with DIN) should be applied. 52

53 6. VERIFICATION INTEGRATED TESTS 6.1. AIMS Precise determination of hermetic compartment leaktightness requires a high degree of accuracy in the parameters of leaking air as well as a high precision of the measuring and evaluation systems, calculational methods and computer codes used. In spite of regular calibration of individual sensors and measuring instrumentation, there might be doubts as to the correctness of measured and calculated leakage rates from hermetic compartments during the main preoperational ILRT. Therefore, the calculational procedure and results of measurements are verified in some countries (Slovak and Czech Republics) by means of the socalled verification integrated tests METHODS In practice, three main types of verification tests can be applied: verification test with zero calibration; verification test with flow rate measurements during depressurization; verification test with flow rate measurements during additional pressurization. Zero calibration Zero calibration is always implemented during preparations for the ILRT. This test is conducted with isolated hermetic compartments at an atmospheric pressure where a leakage rate of zero should be measured. After preparatory work for the ILRT has been completed, the hermetic compartments are isolated from the outside atmosphere. Before pressurization, a few cycles of measurements are carried out as for leakage rate evaluation. Flow rate measurements during depressurization This calibration procedure is applied after the measurements have been completed during the main ILRT. The procedure is as follows: completion of leakage rate measurements during the main ILRT; conversion of leakage rate to flow rate through leaks; gradual opening of the depressurization line bypass with installed flow meter; setting by means of a control valve the calculated flow rate to the same magnitude as was the leakage rate value measured on the flow meter; stabilization of pressure and flow rates; measurement and leakage rate evaluation by means of measuring instrumentation, as for ILRT. Flow rate measurements during additional pressurization In this instance the flow meter is installed in the pressurization line and the inflowing air has to compensate the leakage in such a way that the overpressure in hermetic compartments remains stable. The remaining aspects are as described for flow rate measurements during depressurization. 53

54 6.3. VERIFICATION TEST EVALUATION Zero calibration is evaluated based on leakage rate calculation as for ILRT at the test overpressure. At zero overpressure in hermetic compartments, the measured leakage rate should also be zero with a certain tolerance: Lo = 0 allowable uncertainty [%/d] Verification test with flow rate measurements during depressurization is evaluated based on comparison of the leakage rate set on the flow meter with the quantity by which the original leakage rate measured during ILRT is increased. The air flow rate which leaks from hermetic compartments through leaks during ILRT is calculated from the determined leakage rate (%/day) based on the known free volume of hermetic compartments stated in the design specifications. If the same flow rate is continually released from the hermetic compartments, the original leakage rate measured during ILRT should be increased to a value two times higher: LVT = 2LILRT ± allowable uncertainties The accuracy of this verification method depends to a considerable extent on the precision of the flow meter as well as on the correct determination of the free volume of the hermetic compartments. The procedure which consists in measuring the flow rate of inflowing pressurized air that balances the integral leakages is similar, except for a different LVT: LVT = 0 ± allowable uncertainties [%/d] Note: this procedure is not to be considered a possible substitute for ILRT EXPERIENCE 'Zero' calibration tests are commonly used and their execution involves no problems. Tests with flow meters are rather complicated and were performed only once but with results that were more than satisfactory: the criteria were met with negligible uncertainties. Moreover, there is a common confidence in the method presently used for ILRT, and therefore there are no requirements for further application of these kinds of verification tests. 7. STRUCTURAL INTEGRITY TEST (SIT) 7.1. AIMS, PRINCIPLES AND FREQUENCY OF SIT The aims of the SIT at WWER440 NPPs are: to verify the structural integrity of the hermetic compartments of WWER440/213 plants at a maximum test overpressure of 150 kpa or 1.15 x 150 kpa and of WWER 54

55 440/230 plants at the opening pressure of the small safety valve (thus analytically confirming their structural integrity at the design overpressure of 100 kpa); to verify the strength of technological components of the hermetic boundary using the above maximum test overpressure; to verify the leaktightness of hermetic compartments at a maximum test overpressure and to determine the approximate value of leakage rates; to localize all defects in hermetic boundary structural and technological components for the purpose of repair. An SIT is to be performed once in the lifetime of a unit (it is recommended and desirable to do it during commissioning). After the SIT has been conducted, no further SITs are recommended during the operational life of the NPP. Experts are of the opinion that the degradation of pertinent structural components (walls, technological elements) which could impact their integrity during DBA will not occur during the lifetime of the NPP. The duration of an SIT is determined by the scope of the work to be carried out (inspections, measurements). In principle the time required does not exceed 3 hours INSPECTION METHODS AND MEASUREMENTS Overpressure measurements in hermetic compartments The instrumentation used for overpressure measurements in hermetic compartments is the same as for ILRTs. Measurements of wall deflection Wall deformations are measured at the points specified by the designer. For deflection measurements optical methods have mainly been used and they are fully satisfactory for this purpose. Visual inspection and surveillance of hermetic boundary conditions With regard to the scope of the preparatory work and especially owing to safety reasons, the structural integrity test is the most demanding part of the hermetic compartment integrity tests. The test at a maximum overpressure requires extra safety measures, protection of instrumentation and equipment and the presence of trained staff for hermetic boundary inspection while the test is under way. Special attention should be devoted to training staff for hermetic boundary inspection. Members of inspection teams are equipped with protective aids and positioned on individual floors around the whole hermetic boundary at the respective unit. 55

56 Inspection teams fulfill two functions: searching for leaks up to 100 kpa; observation of deformations above 100 kpa. All inspection teams shall be provided with communication means and personal protective aids. When the inspection teams are searching for leaks, they are provided with equipment needed for leak identification STRUCTURAL INTEGRITY TEST PROCEDURE The SIT should be conducted shortly before the preoperational ILRT or in combination with it; it always precedes the ILRT. This precedence also applies if the SIT is conducted as a complement to a periodical ILRT. Pressurization This procedure is the same as for the ILRT. During the pressurization the following activities are implemented: measurements of deflection at particular overpressure steps; measurements of overpressure; search for leaks; inspection of the hermetic boundary. After the test pressure has been reached, pressurization is stopped. Deflection measurements With regard to SIT, deflection measurements are the most important stage of the test and therefore special attention should be devoted to them. Measurements are implemented by trained staff of the authorized organization. Contingent unallowable deflections shall be reported to the test supervisor. It is good practice to carry out deflection measurements during depressurization at the same overpressures as those at which they are measured during pressurization. Hermetic boundary surveillance and inspection This part of the test is crucial for the safety and success of the SIT procedure. The activities of inspection teams, in which regulators are also expected to take part, start after every particular pressurization step has been reached and pressurization stopped. Inspection teams observe components and structure on pertinent segments of the hermetic boundary (in compliance with the SIT programme). Leakage rate measurements The approximate leakage rate is measured and assessed by applying the methods of ILRT or ILT for leak identification. 56

57 7.4. SIT EVALUATION After processing the test results, the SIT is evaluated based on two criteria: the structural integrity of walls, ceilings and floors, hermetic liner and hermetic boundary components is satisfactory if during the test none of these components have been impaired (no residual deformations or destructions have been found); the structural integrity of the building part is satisfactory if no deflections higher than those specified in the design have been found. Results of leakage rate measurements are compared with the maximum allowable value, thus providing an indication of the ILRT outcome. 8. CONCLUSIONS 8.1. REGULATORY AUTHORITIES' APPROACH Licensing procedures, methods and limits for integral leakage rate testing for NPPs with WWER440/213 reactors are well defined and implemented in all countries. For some NPPs with WWER440/230 reactors, the methods used are less precise, mostly because of the high leakage rates found there. The work on the improvement of the leaktightness of WWER440/230 reactors will also make it possible to introduce essential improvements in this area EXTRAPOLATION OF LEAKAGE RATE VALUES TO THE DESIGN PRESSURE Regular leak rate tests are also performed at pressures lower than the design pressure. In such case, sufficient information should be available to justify extrapolation of measured values to the full containment pressure. This information can be obtained during plant leak rate tests and should include, inter alia, the contribution of various types of 'flow through' leakages to the integral value as well as the change of leakage size and flow characteristics. The extrapolation methodology should be verified by measurements at three different pressures at least, including the containment design pressure. Analytical methods are available for extrapolating a leakage rate measured at some lower pressure up to the design pressure value. Their application is not questionable, provided that they are successfully verified by the actually measured leakage rate values at the extrapolated overpressures LOCAL LEAK RATE TESTS There are many well developed and experimentally verified methods for local leak rate tests, used for type 'B' tests 1. With regard to 'C tests, some ventilation system designs do not allow for the implementation of methods used in western countries. This problem can be solved by means of small backfitting measures. It is believed that this approach is fully feasible and can be successfully implemented LEAK IDENTIFICATION In general, the experience accumulated in this area is good. The higher the integral test overpressure used for leak rate testing, the better the conditions for the identification of 'Definition of "B" and "C" type tests is given in Section

58 leakages. The major problem in this area seems to be the identification of 'inaccessible' leaks, i.e. located in inaccessible places or covered on both sides with concrete. Further efforts should be made to solve this problem STRUCTURAL INTEGRITY For a number of reasons, it has not been possible to perform structural integrity tests at design or 1.15 design overpressures in some NPPs. This lack of test results can presently be compensated by combining calculational analyses with the results of measurements at several overpressure levels. However, the implementation of such an approach is subject to the prior authorization of the regulatory bodies INTEGRATED LEAKAGE RATE TESTS (PREOPERATIONAL, PERIODICAL) Differences in the structure and elements of confinements/containments in WWER 440 units and western PWR containments caused much higher leakage rates in the WWER units than in western plants. Therefore the assumption commonly taken in western practice that the containment pressure will remain constant during the integral leakage rate testing cannot be fully applied to WWER440 reactors. The practice of interpreting the results of leak rate measurements actually differ in various countries. With respect to the WWER440/230 and WWER440/213 units in the Czech Republic, Hungary and Slovak Republic, the initial pressure PI is taken as the reference point, and the leakage rate measured over a period of time (during which the pressure is in effect decreasing from PI to a lower P2 value) is said to be "the leakage rate at the initial pressure PI". In Bulgaria, the Russian Federation and the Ukraine, for both WWER440 reactor types, the instantaneous value of the leakage rate at pressure PI is measured and is called the "the leakage rate at the average pressure PI." The results are different. The value of the discrepancy depends on the leakage rate, i.e. on the pressure decrease during the test. Both approaches are technically sound and can be logically explained. However, the differences outlined above make it impossible to compare the results directly among the plants in the various countries. The containment leak rate is defined as the time derivative of the containment air mass at the constant design pressure divided by the total air mass in the containment. It is usually expressed in terms of relative change of the air mass in 24 hours time. The containment leak rate for WWER440 reactors should also be defined in accordance with this definition. Since it is not possible to maintain the containment pressure constant during 24 hours, the measurement of the leakage rate made in a shorter period of time, during which the pressure can be treated as constant, can be used for determining the leakage during 24 hours. 58

59 REFERENCES [I] INTERNATIONAL ATOMIC ENERGY AGENCY, General Design Safety Principles for Nuclear Power Plants, A Safety Guide, Safety Series No. 50SG Dll, IAEA, Vienna (1986). [2] INTERNATIONAL ATOMIC ENERGY AGENCY, Safety Functions and Component Classification for BWR, PWR and PTR, A Safety Guide, Safety Series No 50SGD1, IAEA, Vienna (1979). [3] INTERNATIONAL ATOMIC ENERGY AGENCY, Design of the Reactor Containment Systems in Nuclear Power Plants, A Safety Guide, Safety Series No 50SGD12, IAEA, Vienna (1985). [4] Containment and Confinement Performance in NPPs with WWER 440/213 and WWER 440/230 Reactors, Report of a Consultants Meeting, WWERSC085, Vienna, 29 November to 3 December 1993 (Report to the Steering Committee). [5] Confinement Improvement Options for NPPs with WWER 440/230 Reactors, WWERSC101, Vienna, 1014 July 1995 (Report to the Steering Committee). [6] UNITED STATES NUCLEAR REGULATORY COMMISSION, Primary Reactor Containment Leakage Testing for WaterCooled Power Reactors, App. J., Title 10, Code of Federal Regulations, Part 50, Washington, DC (1991). [7] AMERICAN NATIONAL STANDARDS INSTITUTE, Leakage Rate Testing of Containment Structure for Nuclear Reactors, ANSIN45.4, La Grange Park (1972). [8] AMERICAN NATIONAL STANDARDS INSTITUTE/AMERICAN NUCLEAR SOCIETY, Containment System Leakage Testing Requirements, ANSI/ ANS56.8, La Grange Park (1987). [9] GOSATOMNADZOR, Regulations on Construction and Operation of Containment Safety Systems of NPPs PNAEG , Moscow (1991). [10] TECHY, Z., TAUBNER, R., BACSKO, G., "Leak rate measurements for WWER440/213 confinements in Hungary", paper presented at the Mtg on Methods Used for WWER440 Confinement/Containment Structural Integrity and [II] Leak Rate Tests, IAEA, Vienna (1995). GESELLSCHAFT FUR ANLAGEN UND REAKTORSICHERHEIT (GRS) mbh, Integral Leakage Rate Testing of the Containment with the Absolute Pressure Method, Kerntechnischer Ausschuss (KTA) 3405, Issue 2/79 Germany (1979). [12] ASSOCIATION FRANCAISE POUR LES REGLES DE CONCEPTION ET DE CONSTRUCTION DES MATERIELS DES CHAUDIERES ELECTRO NUCLEAIRES, Leaktightness and Resistance Tests of Containment, Part 3 of RCCG, "Design and Construction Rules for Civil Works of PWR Nuclear Islands", AFCEN, France, July 1988 Edition. [13] CZECHOSLOVAK ATOMIC ENERGY COMMISSION, Decree No 2 of October 27, 1978, on the Nuclear Safety Assurance during Design, Licensing and the Construction of a Building with a Nuclear Power Facility. [14] CZECHOSLOVAK ATOMIC ENERGY COMMISSION, Regulation No. 436/90 dated October 10th, 1990, on the Quality Assurance (QA) of Selected Equipment with Regard to Nuclear Safety of Nuclear Facilities. [15] MINISTRY OF LABOUR AND SOCIAL AFFAIRS, LABOUR SAFETY AUTHORITY OF CZECH REPUBLIC, Ordinance of the Code on National Technical Supervision over the Safety of Technical Equipment in the Field of Power Engineering, 76 UBP/89 (1989). 59

60 [16] MINISTRY OF LABOUR AND SOCIAL AFFAIRS, LABOUR SAFETY AUTHORITY OF SLOVAK REPUBLIC, Ordinance of the Code on National Technical Supervision over the Safety of Technical Equipment in the Field of Power Engineering, SUBP 66/89 (1989). [17] VASSIAGUINE, R.V., "Experience in conducting tests of accident localization system (containment) in unit No. 1 of Rovno, NPP, Ukraine", paper presented at the Mtg on Methods used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna, [18] DEMIREVA, E., BALABANOV, E., "Test methods, results and needed modernization for the improvement of WWER440/230 confinement leaktightness", paper presented at the Mtg on Methods Used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna, [19] BALAZ, P., "Integrated tests of WWER440 hermetic zones", paper presented at the Mtg on Methods Used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna, [20] PRANDORFY, M., "Local and individual tests of hermetic compartments of WWER440 NPPs conducted by VUEZ, Tlmace, Slovak Republic", paper presented at the Mtg on Methods Used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna, [21] RICANY, J., "Comments of the Nuclear Regulatory Authority of the Slovak Republic on the methods used for determining leak rates from confinement", paper presented at the Mtg on "Methods Used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests", IAEA, Vienna, [22] LISKA, P., KADLEC, L., HATAS, J., KLEPAC, J., "Measurement of leak rate from the WWER440 confinement, using a gas flow equation", paper presented at the Mtg on "Methods used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests", IAEA, Vienna, [23] HLINKA, M., KUCAK, L., URBAN, F., "Method of leak rate tests results evaluation", paper presented at the Mtg on Methods used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna [24] ANTOCHINE, V., VLASENKO, A., GRIGORIANTS, E., KOCHMANOV, E., "Integral leak rate tests performance for WWER440/230 confinements and WWER440/213 containments", paper presented at the Mtg on Methods used for WWER440 Confinement/Containment Structural Integrity and Leak Rate Tests, IAEA, Vienna, [25] US Reactor Containment Technology, Oak Ridge National Laboratory Report, ORNLNSIC5 (1965). 60