Failure database and tools for wind turbine availability
Contents
1. Laskeplader Fish plate Lejer Bearings Lejer pakninger Bearing seal Luftcirkulation Air circulation Lynbeskyttelse Lightning protection Lobskkersel Over speed L s Loose Magnetventil Magnet valve Maksimalafbryder Maximum circuit breaker Montage Mounting Monteret Mounted Motorv rn Protective motor relay Multistik Multipin connector Nacelle Nav Net Mains N dstop Emergency stop Olie Oil Olie fedt pumpe Oil grease pump Oliek ler Oil cooler Olievarmer Oil heater Overl bsventil Overflow valve Overtryksventil Pressure valve Pitchet Pitched Produktion Production Pumpe Pump P fyldt Filled P f rt Placed RC enhed RC unit Rel er kontaktorer Relays Power relays Rel kort Relay card Rengjort Cleaned Renoveret Renovated Renset Cleaned Repareret Repaired Revnet Fracture Rotor Hydraulik Rotor hydraulics RPM sensor beslag RPM sensor mounting Rusten Corroded Samling Connection Sidder fast Stuck Sikkerhedsventil pop out Safety valve pop out Sikringer Fuses Slanger Hoses Slidt Worn out Slip ring Slip ring Sleebering Slip ring Smerenipler Grease nipples Snoede kabler Twisted cables Spreengblik Sp ndt Tensioned St dd mper Shock absorber St jer Noisy 46 Ris R 1200 EN TAC computer TAC computer Temperaturmodul Temperature module Temperaturs2. Component failure Oil flow by control system Oil flow by centrifugal 1 tip brake deploys 2 tip brakes deploy 3 tip brakes deploy Sequence formula ABCD ABC ABCD ABC Figure 6 Event tree for initiating event Gearbox failure and condition 20 m sSws lt 25 m s A B C D E Component Oil flow by Oil flow by 1 tip brake 2 tip brakes 3 tip brakes Sequence failure control system centrifugal deploys deploy deploy formula ABCDE ABCD ABC ABCDE ABCD ABC ABCDE ABCD ABC AB Figure 7 Event tree for initiating event ws225 m s A B E ER ane onan ABE ABE ABE APPENDIX II Diagram of the Hydraulic System Risg R 1200 EN 39 Sroqayus 0098 0 S v ANINVAGAI NAY Br OL fy ty Gyt MOE 4 yO Gy Cy Gyt 2u0qd Aqpuing 0992 40 sabojog Gt 96 Z8 98 Spi X64 96 2B 98 Gy Buoys 6 09a4NS 0099 40 faasuuousaboj S V AMAVYGAH NAV NAV OFT VE AVYSVIC 1102315 een WGOW FOA ZOA IOA NAV LO f6 90 GE ML LINA ONAVEGAH oy bumo 1al qns G 0 JONVHO decd LOA NAV Ot ca ome OA COA NAV 69 ii 22 Pactra COA NAV O JouvHoRid ee LN Z ae Sega SOA NAV OL pai ae are OA NA
3. Failure state Figure 2 Event tree for initiating event Grid loss and condition 10 m sSws lt 20 m s A B C D F Component failure Oil flow by control system Oil flow by centrifugal 1 tip brake deploys 2 tip brakes deploy Mechanical brake Sequence formula ABC DF ABC ABC DF ABC ABCDF AB Figure 3 Event tree for initiating event Grid loss and condition 20 m sSws lt 25 m s A B D E F Component failure Oil flow by control system Oil flow by centrifugal 2 tip brakes deploy 3 tip brakes deploy Mechanical brake Sequence formula ABDE F ABD ABDEF ABD ABDEF ABD AB Figure 4 Event tree for initiating event Gearbox failure and condition 5 m ssws lt 10 m s A B C D E Component Oil flow by Oil flow by 1 tip brake 2 tip brakes 3 tip brakes Sequence failure control system centrifugal deploys deploy deploy formula ABC ABC ABC Figure 5 Event tree for initiating event Gearbox failure and condition 10 m sSws lt 20 m s A B C D E Component failure Oil flow by control system Oil flow by centrifugal 1 tip brake deploys 2 tip brake
4. state On the event trees the systems involved are denoted as follows A B C D E F Magnetic Pop out valve 1 tip brake 2 tip brakes 3 tip brakes Mechanical valve system deploys deploy deploy brake It should be noted that the event trees for the initiating events Grid loss Generator failure and Driveline failure will be identical because the occurrence of any of these events means that the WTB cannot be maintained in a safe mode by the grid generator connection In conclusion it is only necessary to develop seven event trees to model the un reliability of the SS Initiating event Component failure This initiating event implies that when there is a demand to stop a WTB due to a component failure in the WTB both the generator and the grid are operational Thus under all wind speeds except ws225 m s which will be considered separately the WTB will be kept safe by the generator connected to the electrical grid Hence under these circumstances an undesired outcome will not occur and all the three conditional probabilities P SS I C will be equal to zero Initiating event Grid loss Generator failure Driveline failure and wind speed 5 m s lt ws lt 10 m s To ensure WTB safety under these conditions it must be possible for either one tip brake to deploy or the mechanical brake to deploy The event tree depicting all possible failure scenarios for initi
5. Bolte St dd mper Split bushing Lejer pakninger Olie Suspension Temperatursensor Oliek ler Olievarmer Overtryksventil Slanger Sl bering Pumpe The probability of gearbox failure is thus calculated using an equation similar to equation 23 Ris R 1200 EN 19 Mechanical brake failure P F The mechanical brake consists of the following components the failure of each of which will cause generator failure Bolte Bremseskiver Bremseklodser Filter Hydraulikstation Hydraulikslanger Ikke returnventil Magnetventil Motor Microswitch Olie Overlabssventil Trykswitch Bremsescoop Akkumulator The probability of mechanical brake failure is thus calculated using an equation similar to equation 23 Tip brake failure P T A tip brake consists of the following components the failure of each of which will cause generator failure Bolte Vinge tip guides Vinge cylinder Vinge tip spring Vinge tip Vinge root Wire Surface Forkant bagkant Sensor Lynbeskyttelse Hydraulic system P A and P B The deployment of the tip brakes is activated by the hydraulic system To enable the tips brakes to de ploy oil must flow through either the magnetic valve Appendix II Position 13 or the pop out system Appendix II Positions 19 and 55 These are considered independent subsystems of the SS The magnetic valve is processed in the database as one component and valve failure is designated A The pop out system is p
6. Information services data analysis Screening convoluting and reporting information in accordance with user requirements at the Component level System level and Whole WTB level 3 Reliability analysis e Reliability assessments of all components and systems and the reporting of them e Reliability analysis of the SS subsystems and the SS as a whole e Reliability analysis of the safety related systems 6 Logical Data Model In order to meet the objectives and theoretical requirements three basic tables Tables 1 3 and several subsidiary tables were created Appendix III The basic tables comprise the actual database and con tain the basic information needed to make a decision on WTB performance especially availability and reliability The main reason for dividing the database into a series of tables is to avoid the redundancy in the data The links between tables are provided by Primary and Foreign keys The table fields contain the option codes in those cases where it has been possible to classify them but not full names and definitions Thus for the fields System Component Failure causes Detection methods Responsibility WTB modes see Table 1 all the options can be foreseen in advance and classified The codes for these fields are consequently short The decoding tables for these fields are the subsidiary tables Except for the decoding function these serve as menu options under data entry Each type of WTB has its own set
7. le sessments of D D and T a where N is the number of failures of the ith com on No ae i j l ponent during 0 T that the average availability of a system in 0 T is the expected proportion of time the system is operating during 0 T i e i j l T a 2 In the specific case of a WTB we are also interested in the average availability A of WTBs of the same type where A is defined as NYTB NYTB WTB T PR T we ie 1 rd 2 1 2 NTS 2 i NB T NIB T 3 where NY is the total number of WTBs of the same type and T is the total operation time of the ith WTB within the interval 0 T Equation 3 is only really valid if all NY are put into operation at the same time 0 and are in operation until the present time In reality this is not the case as the various starting dates are generally different and some of the WTBs can be put out of operation completely Equation 3 thus has to be rewritten 1 ee eg z NIB LT 4 where T is the total calendar time worked by the ith WTB Equation 4 is used to assess the average availability of specific types of WTB Knowledge of the availability of each type of WTB will enable identification of the most unreliable specific WTBs Availabilities A are ranked in order of increasing A Starting from the lowest A A specific WTB occupying the first place is the most unreliable as re gards availability In order to identify the most unreliable subsystems and
8. R9 Overlobsventil R10 Tryk switch R11 Sikkerhedsventil pop out R12 Accumulator R13 Spreengblik R14 Kabler Riso R 1200 EN Repair Action Code Name Responsibility Genanvendes Code Name Konstruktion og udvikling Genstartet via remote Indk b Genstartet 1 Vindm llen Montage Hejsning Produktion Inspiceret Garanti service Installeret Nl Nn WI NO Service Justeret Pafort Pafyldt Renset Samling Sp ndt Udf rt Udluftet Udskiftet Repareret Detection Method Name Planned maintenance Operator Overspeed Vibration Inverter temperature Inverter control SVD OY BY WI NO Grid failures Risg R 1200 EN Failure Causes Code Name Brandt Defekt Defekt Lynnedslag Kontrolleret For stor Irret Knekket Justeret Los Lebskersel Monteret Rengjort Pitchet Renoveret Revnet Rusten Sidder fast Slidt St jer de af justering df rt dtaget dluftet NI lt XI lt CLAY A AO 9 O Z lt A SA o E m D a w gt U U U Udskiftet U U t t 43 Appendix IV Creating a new Microsoft Access workgroup informa
9. WTB the contents of which have been described above in Section 5 Logical Data Model These tables store all the information used in reliability analyses as well as some reliability related data 24 Risg R 1200 EN The option Analyses offers three suboptions 1 Data Analysis 2 Availability Analysis and 3 Reliability Analysis The option structure of Data Analysis is as follows Specific WTB gt WTBID Component amp Component Specific WTB WTB ID all components Component All WTBs all components Specific 2 Component component All WTBs System Specific WTB All WTBs WTB Specific WTB The result of any of the options is a report consisting of predefined data selected from the database For example the result of the option Data Analysis System All WTBs is a report such as shown on the following page The other options represent similar reports depending on the request The option structure of Availability Analysis is as follows Data Analysis 1 Dynamic 2 All WTBs Availability Analysis 3 Systems 4 Component 5 System Component Option 1 Dynamic presents the average availability of the whole sample of WTBs by year i e the option provides the possibility to see the interannual variation in availability Option 2 All WTBs provides a summarizing report of the average availability of the type of WTB in question and the unavailability of each WTB within
10. components it is useful to determine the unavailability U represented as Risg R 1200 EN 9 N WTB 1 1 NYTB T down NYT 1 N gih U e LY ee 21 gt Ae NUT l NU T Nye i l Je T w N NIB mm down yei N 1 DE u FT S l gt T 1 T down 5 NYTB 4 i NIB T T AI j T i l i i l T j l down where T is the total down time due to failures of the ith WTB within interval 0 T N is the number of systems in a WTB of a certain type T down is the total down time due to failures of the jth y ype L J N T down ik 1 system in the ith WTB The term Nm gt i l i was extracted from the average unavailability to calculate the contribution to this reliability characteristic made by the kth system Comparing two sys NYTB N tems k and m their contributions differ by the terms Z and gt T m When ranking the sys i l i N tems contribution to unavailability it is thus sufficient to rank the terms SEY Ranking the sys i l tems according to the summary down time will portray the weakest systems with respect to their con tribution to the unavailability Ranking by systems might be not informative enough to show what design improvements would have the greatest effect on WTB availability Identifying the weakest components in a particular type of WTB can thus be very useful when trying to identify the most effective improvements A measure of the contribution at a component level can be introduced using t
11. follows Users Member of group Admin default Users default Administrator Admins Users default Analyst NotAdmin Users default Typistl Typists Users default The permitted operations within each group account are as follows Permitted operations Group Not Admin Tables Read Design Read Data Queries all Read Design Read Data Forms all Open Run Reports all Open Run Macros all Open Run Modules all Open Run Read Design Typists Tables all Read Design Read Data Insert Data Forms Components Form0 Form1 Main Table WTB Open Run Macros all Open Run Read Design Riso R 1200 EN 23 Some tables have more permitted operations which are created and deleted when running the reliability calcu lations These tables are not the objects of the logical model Only one table WTB can be changed by a No tAdmin group user These extra permitted operations are necessary to enable insertion of the current data into the field Date of Finish and to enable calculation of the total working experience The Admin group is the default group The Administrator group is not listed in the table of per mitted operations since all operations are permitted especially the assignment of new group and user accounts and the granting of permitted operations The security system is organized in such a way that information about groups users regi
12. lt 20 m s To ensure WTB safety under these conditions it must be possible for two tip brakes to deploy The other subsystems of the SS are not available The event tree for this case is shown in Appendix I Fig 5 The final canonical expression for failure of the SS is SS LC AB OR COR D 18 i e the SS fails if the magnetic valve fails and the pop out system fails or one tip brake cannot de ploy or two tip brakes cannot deploy Initiating event Gearbox failure and wind speed 20 m s lt ws lt 25 m s To ensure WTB safety under these conditions it must be possible for three tip brakes to deploy The other subsystems of the SS are not available The event tree for this case is shown in Appendix I Fig 6 The final canonical expression for fail ure of the SS is 16 Risg R 1200 EN SS IsC ABOR COR DOR E 19 i e the SS fails if the magnetic valve fails and the pop out system fails or one tip brake cannot de ploy or two tip brakes cannot deploy or three tips cannot deploy Initiating event ws gt 25 m s To ensure WTB safety at wind speeds exceeding 25 m s it must be possible for three tip brakes to de ploy The event tree developed for this case is given in Appendix I Fig 7 The final canonical logical expression for the SS failure is SS l AB OR E 20 i e the SS fails if the magnetic valve fails and the pop out system fails or three tip brakes cannot deploy 4 4 A
13. m s C1 LC Component failure I 10 m s lt ws lt 20 m s C2 LC 20 m s lt ws lt 25 m s C3 LC 5 m s lt ws lt 10 m s C1 LC Grid loss L 10 m s lt ws lt 20 m s C LC 20 m s lt ws lt 25 m s C3 LC 5 m s lt ws lt 10 m s CG LC Generator failure I 10 m s lt ws lt 20 m s C3 LC 20 m s lt ws lt 25 m s C3 LG 5 m s lt ws lt 10 m s C LC Drive line failure L 10 m s lt ws lt 20 m s C3 LC 20 m s lt ws lt 25 m s Gs LC 5 m s lt ws lt 10 m s C I C Gearbox failure I 10 m s lt ws lt 20 m s C3 L C 20 m s lt ws lt 25 m s Cs L G ws225 m s I Ig 4 3 Event Tree Construction In working out the event trees we were concerned with identifying the sequences ending in SS failure states Sequences that end in success states were therefore generally omitted Since we are analysing 6 systems that might affect SS reliability each of which can have two states success and failure the total number of states is 2 64 As only a few are failure states there is no need to depict them all On the event trees shown in Appendix I sequences ending in failure states are shown by bold lines with the logical function representation written above them In this context AB denotes that the se quence takes place if system A is in a success state and system B is in a success state while AB 14 Risg R 1200 EN denotes that the sequence takes place if system A is in a success state and system B is in a failure
14. of three basic tables and separate entrance i e a separate data base Risg R 1200 EN 21 Tables 2 3 were mainly created to provide for quantitative reliability calculations The field Date of start thus contains numbers for ti 2 Date of finish for tt 2 and Test interval for assess ing the un availability of periodically inspected components Number for m 1 The number of wind turbines in question Nwrp is equal to the number of records in the WTB table Foreign Foreign Main Table Table 1 mal Ee ped ar La a a BEER BER HEER EE Repair Repair P Failure Detect Respon Repair Remark cause method sibility action a o eae j WTB Table Table 2 WTB ID Date of finish record fo Primary key recordk ff L Component Table Table 3 Primary key HEEL 3 4 5 6 Component Component System Test Number Price name interval record 1 record n 22 Risg R 1200 EN 7 Security Features The database is secured and access to it requires a user account and a password The security system encompasses both group accounts and user accounts Each user account must belong to a group ac count There are default user and group accounts and newly created user and group accounts The fol lowing group accounts are relevant for the security of the current database Groups Admins default Users default NotAdmin Typists The user accounts are as
15. our case different wind speeds In this case we have to calculate the resultant probability conditional on all possible combi nations of the initiating events and the conditions C The final equation for calculating the probabil ity that the SS will fail is thus P SS gt gt P SS I C PU C gt gt gt POSS 1 C PU P C 13 4 2 Assumptions for the Safety System The mission of the SS is different from the rest of a WTB in that it is activated on demand in order to bring the WTB to a safe condition i e one in which the rotor is either completely halted or is rotating at a permissible rotation speed SS reliability can thus be defined as The reliability of a WTB safety system is the probability that the safety system is able to bring the WTB to a safe condition on demand under given conditions for a specified time interval As pointed out above a characteristic of WTBs is that some of the components and systems are per manently operational and at the same time affect the performance of the SS The components in ques tion are the generator the gearbox and the driveline connecting the gearbox and the generator When the demand arises to stop a WTB the generator connected to the electrical grid acts as an additional SS Riso R 1200 EN 13 subsystem and the reliability of the SS thus depends on the reliability of the SS A failure of the gear box affects the reliability of the SS such that rotation of the rotor cannot b
16. the type The WTBs are ordered by their unavailabilities starting with the most unreliable WTB Options 3 4 and 5 provide the user with reports showing the contributions to the unavailability of 3 all the systems 4 all the components reported in the database and 5 all the components constitut ing a specified system The option structure of Reliability Analysis is as follows Risg R 1200 EN 25 Generator Gearbox Driveline Steering System External Event Nacelle Reliability Analysis WTB Reliability Tower SS Reliability Print Board Yawing System Main Shaft Cover The option External Event does not perform any calculations and is an information option informing the user of the rates of the external events used in the SS reliability calculations These rates can be subject to periodic correction if more precise data become available The option WTB Reliability calculates the reliability of any WTB system excluding the SS Clicking any of the system name buttons will initialize the chain of the resulting reliability assess ments of all the components constituting the system and the system as a whole The option SS Reliability calculates the reliability on demand for all the subsystems of the SS and the SS as a whole 9 Notes for the System Developers The relationships between the objects in the database are generally uncomplicated and the developer familiar with the basics of MS Access will easi
17. Downloaded from orbit dtu dk on Dec 18 2015 Technical University of Denmark lI Failure database and tools for wind turbine availability and reliability analyses The application of reliability data for selected wind turbines Kozine Igor Christensen P Winther Jensen M Publication date 2000 Document Version Publisher final version usually the publisher pdf Link to publication Citation APA Kozine l Christensen P amp Winther Jensen M 2000 Failure database and tools for wind turbine availability and reliability analyses The application of reliability data for selected wind turbines Denmark Forskningscenter Risoe Risoe R No 1200 EN General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights e Users may download and print one copy of any publication from the public portal for the purpose of private study or research e You may not further distribute the material or use it for any profit making activity or commercial gain e You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details and we will remove access to the work immediately and investigate y
18. Mechanical brake Sequence for mula Success state Failure state Figure 2 Event tree for initiating event Grid loss and condition 10 m sSws lt 20 m s A B C D F Component failure Oil flow by control system Oil flow by centrifugal 1 tip brake deploys 2 tip brakes deploy Mechanical brake Sequence formula ABC DF ABC ABC DF ABC ABCDF AB Figure 3 Event tree for initiating event Grid loss and condition 20 m sSws lt 25 m s A B D E F Component failure Oil flow by control system Oil flow by centrifugal 2 tip brakes deploy 3 tip brakes deploy Mechanical brake Sequence formula ABDE F ABD ABDEF ABD ABDEF ABD AB Figure 4 Event tree for initiating event Gearbox failure and condition 5 m ssws lt 10 m s A B C D E Component Oil flow by Oil flow by 1 tip brake 2 tip brakes 3 tip brakes Sequence failure control system centrifugal deploys deploy deploy formula ABC ABC ABC Figure 5 Event tree for initiating event Gearbox failure and condition 10 m sSws lt 20 m s A B C D E
19. N Appendix V Glossary of WTB construction and maintenance terms Aftasteplade Akkumulator Accumulator Anemometer Anemometer Bolte Bolts Bremse Brake Bremseklodser Brake shoes Bremsescoop Brake scoop Bremseskive Brake disc Brokobling Bridge coupling Breendt Burnt Carbon besning Carbon bushing Kardanaksel Cardan shaft Defekt Defective Defekt Lynnedslag Defective Struck by lightning Fasebatteri Phase battery Filter Filter For stor Too large Forkant bagkant Front edge back edge Garanti service Warranty service Gearkasse Gear box Genanvendes To be reused Genstartet 1 vindmellen Restarted in the wind turbine Genstartet via remote Restart via remote Hejsning Hoist Hovedaksel Main shaft Hovedleje Main bearing Hydraulikslanger Hydraulic hoses Hydraulikstation Hydraulic station Ikke returventil Non return valve Inddekning Cover Indk b Purchase dept Inspiceret Inspected Installeret Installed Irret Corroded Justeret Adjusted Kabelaflastning Cable relief Kabler Cables Klemmeelement Clamp Kn kket Broken Kobberb rste Copper brush Kobling Coupling Kommunikation Communication Konstruktion og udvikling Construction and development Kontaktor Power relay Kontrolleret Controlled Kr jegear Yawing gear Kr jesystem Yawing system Ris R 1200 EN 45
20. P C P A P B P D P E P F P D P F P SS 1 P P C P A P B P C P C P A P B P D P C P A P B P E P SS I PU P A P B P E The final expression for calculating the probability that the SS will fail is P SS P PC PU ILP C P A P B P F P C P F P F P C P A P B PCC P D P F PCC P F P C P A P B P D P E P F P D P F PU LP C P A P B P C P C P A P B P D P C P A P B P E PU P 4 P B P E 22 Probabilities P C P D and P E have to be calculated differently from the rest of the probabili ties and we have to employ the binomial distribution of probabilities P C is the probability of an event where one tip brake cannot deploy If we denote one tip brake failure by T and take into account that there are three tip brakes then P C PTY D is an event where two tip brakes cannot deploy This event takes place if none of the tip brakes can deploy or any two out of the three tip brakes cannot deploy Assuming binomial distribution P D PITY 3P T 2 1 PT E is an event where all three tip brakes cannot deploy together This event takes place if none of the tip brakes can deploy or any two of the three tip brakes cannot deploy or all three tip brakes cannot deploy simultaneously Thus P E PITY 3P T 1 P T 3P T P T 2 4 5 System Reliability Modelli
21. V COA NAV OL LOA NAV Gb C an Risg R 1200 EN 40 Appendix III Subsidiary Tables Providing the System and Component Codes Risg R 1200 EN 41 Code Styring S1 Net S2 Rel kort S3 Transformerkort S4 TAC Computer S5 WP3000 S6 WP2060 S7 Kommunikation S8 Nodstop S9 Maksimalafbryder S10 Relzer kontaktorer S11 Multistik S12 Modem S13 Lynbeskyttelse S14 Brokopling S15 Fasebatteri S16 E prom S17 Motorveern S18 Kontaktor S19 Thyristor S20 Sikringer S21 Program Code Nacelle NI Bolte N2 Vibrationssensor N3 Topboks N4 Kabler N5 Kabeaflastnig N6 Multistik Code Mekanisk bremse B1 Bolte B2 Bremseskiver B3 Bremseklodser B4 Filter B5 Hydraulikstation B6 Hydraulikslanger B7 Ikke returnventil B8 Magnetventil B9 Motor B10 Micro switch B11 Olie B12 Overlobssventil B13 Trykswitch B14 Bremsescoop B15 Akkumulator Code Vinger VI Bolte V2 Vinge tip guides V3 Vinge cylinder V4 Vinge tip spring V5 Vinge tip V6 Vinge root V7 Wire V8 Surface v9 Forkant bagkant V10 Sensor Vil Lynbeskyttelse 42 Code System Tarn Styring Printkort Krojesystem Nacelle Hovedaksel Gearkasse Makanisk Bremse Driveline Generator Ind
22. ating event Grid loss and condition 5 m s lt ws lt 10 m s is shown in Appendix I Fig 1 It can be seen that there are 5 different ways to reach a failure state The summarizing logical formula for failure of the SS is thus SS LbC ABCF OR ABCF OR ABCF OR ABF OR ABCF which can be simplified as SS LC CF AB OR AB OR AB OR AB OR ABF Furthermore the expression in parentheses is a certain event that takes place with a probability of 1 and hence can be omitted The final canonical expression for failure of the SS is thus SS LC F AB OR O 14 ie the SS fails if the mechanical brake fails and the magnetic valve fails and the pop out system fails or one tip brake cannot deploy Initiating event Grid loss Generator failure Driveline failure and wind speed 10 m s lt ws lt 20 m s To ensure WTB safety under these conditions it must be possible for either one tip brake to deploy and the mechanical brake to deploy or for two tip brakes to deploy The event tree depicting all possible failure scenarios for the initiating event Grid loss and wind speed 10 m s lt ws lt 20 m s is shown in Appendix I Fig 2 It can be seen that there are 6 different ways to reach a failure state The summarizing logical formula for failure of the SS failure is thus Risg R 1200 EN 15 SS LC ABCDF OR ABC OR ABCDF OR ABC OR ABDF OR AB with the corresponding final canonical expression be
23. bility point of view No component can ever be completely reliable however If no failures have been reported for a component the lower reliability can be estimated on the basis of the conservative assumption that a failure might possibly take place within the next small time interval i e Nj 1 Since the upper limit is not known for this case one can hypothetically optimistically assume that Nj 0 The lower and upper reliabilities can consequently be determined using equation 12 by employ ing optimistic and pessimistic assumptions 3 6 Reliability Analysis of Components and Systems Working on Demand Subsystems and components constituting the SS are not operational all the time but need to be acti vated when there is a demand to halt the rotor or to maintain it in a safe mode The probability that they will work on demand is expressed as A i O z APem where A is the failure rate of the ith subsystem or component A is the average number of demands per year A is determined using equation 10 In the present case the number of demands per year is defined as 12 Risg R 1200 EN Eon aan ee ee Grid r Driveli box where A041 Generator qpriveline ang 4699 are determined using equation 10 and and A are the average rates of grid loss and wind speed exceeding 25 m s respectively determined from external sources 4 Reliability Analysis of the Safety System 4 1 Basic Concepts of Event Tree A
24. dzekning Nav Rotor hydraulik Vinger Code Tarn Tl Bolte T2 Laskeplader gt T3 Snoede kabler T4 Kabler Code Printkort P1 RC enhed P2 Temperaturmodul P3 Transformer Pa P4 Triggerkort Code Kr jesystem K1 Kr jegear K2 Bremse K3 Lejer pakninger K4 Motor K5 T nder K6 Sensor K7 Aftasterlade Code Hovedalsel K8 Smerenipler HI Bolte H2 RPM sensor beslag H3 Carbon besning H4 Lejer pakninger HS RPM sensor Code Gearkasse H6 Slip ring G1 Bolte H7 Olie fedt pumpe G2 St dd mper H8 Hovedleje G3 Split bushing H9 Klemmeelement G4 Lejer pakninger H10 Sm renipler PUG Olie H11 Kobberb rste G6 Suspension G7 Temperatursensor G8 Oliek ler G9 Olievarmer Code Generator G10 Overtryksventil El Bolte G11 Slanger E2 RPM sensor G12 Sl bering E3 St dd mper G13 Pumpe E4 Kobling Code Driveline E Lejer DI Bolte E6 Temperatursensor E7 Terminaler D2 Kardanaksel D3 Kobling ES Kabir E9 Slanger E10 Pumpe Code jnddekning Ell Vanddekning Il Luft cirkulation 2 Vindfane B Anemometer 14 Temperatursensor Code Rotor Hydraulik RI Bolte R2 Filter Code Nav R3 Hydraulikstation Al Bolte R4 Hydraulikslanger R5 Ikke returnventil OS TII R Magnetventil R7 Microswitch R8 Olie
25. e availability divided by the system contribution NIB T down NIB imk down gt STA T u il 4i Gel 8 mk NU dow NYTB T i l i i l 3 5 Reliability Analysis of Permanently Working Components and Systems Reliability assessments are measures of the frequency of failure but without taking into account repair time They characterize individual properties of components without relating to the performance of the system of which they are a part A usual way of representing reliability is through the failure rates The general formula for assessing the failure rate of a component is ON NT 9 where N is the number of failures of a component during interval 0 T and N is the total number of components under observation As regards WTBs 9 can be rewritten NUTB 2N j l pg T a where N is the number of failures of the ith component of the jth WTB M is the number of similar components in one WTB and T is the total operational time worked by NY WTBs of a given type The parameter T can be calculated in different ways depending on the nature of the available data If failure data are collected for all wind turbines in operation from the point they entered service then N E 27 j Where t is the total time of operation of the jth wind turbine of a given type In our case j l the failure data are collected from a fixed time in the past tpasn which supposes that some failures have not bee
26. e halted by the grid con nected to the generator and the mechanical brake cannot be engaged Failure of the driveline disrupts the grid generator connection thereby increasing the probability of SS failure Six groups of demands activating the SS can be identified 1 a component failure in the WTB 2 grid loss 3 generator failure 4 driveline failure 5 gearbox failure and 6 wind speed exceeding 25 m sec These events are assumed to be mutually exclusive i e any two of them cannot happen simul taneously This further implies that while the WTB is being brought to a stop state the other initiating events demands cannot take place This simplification is necessary to simplify the reliability models and is practically justifiable Different weather conditions wind speed can require different functionality of the SS and must therefore be taken into account when working out scenarios for possible SS failures The following five wind speed ranges are chosen for the reliability analysis 1 ws lt 5 m s 2 5 m s lt ws lt 10 m s 3 10 m s lt ws lt 20 m s 4 20 m sSws lt 25 m s and 5 ws225 m s The lowest range ws lt 5 m s is of no inter est in the present context since a WTB cannot experience over speed under such conditions The high est range ws225 m s is regarded as an initiating event The following six different scenarios will thus be analysed Initiating event Wind speed Designation 5 m s lt ws lt 10
27. ensor Temperature sensor Terminaler Terminals Topboks Top box Transformerkort Transformer card Triggerkort Trigger card Trykswitch Pressure switch T nder Teeth Ude af justering Out of adjustment Udf rt Carried out Udluftet Ventilated Udskiftet Exchanged Udtaget Removed Ut t Leaky Vandd kning Water cover Vibrationssensor Vibration sensor Vindfane Wind vane Vinge cylinder Wing cylinder Vinge root Wing root Vinge tip Wing tip Vinge tip guides Wing tip guides Vinge tip spring Wing tip spring Wire Wire WP2060 WP2060 WP3000 WP3000 Ris R 1200 EN 47 Bibliographic Data Sheet Rise R 1200 EN Title and authors Failure Database and Tools for Wind Availability and Reliability Analyses Igor Kozine Palle Christensen and Martin Winther Jensen ISBN ISSN 87 550 2732 6 0106 2840 87 550 273 1 8 internet Department or group Date Systems Analysis Department January 2000 Safety Reliability and Human Factors SPM Groups own reg number s Project contract No s 1210066 Pages Tables Illustrations References 47 6 11 3 Abstract max 2000 characters The objective of this project was to develop and establish a database for collecting reliability and reli ability related data for assessing the reliability of wind turbine components and subsystems and wind turbines as a whole as well as for assessing wind turbine availability while ra
28. f the connected actions Only the content of the Tables 1 2 3 and those shown in Appendix III and the table External Events can be changed The table External Events can be periodically up dated when more precise data become available The contents of all the other tables cannot be changed An exception is the table Components The components Magnetventil Sikkerhedsventil pop out and Sprzengblik are referred to via their IDs in the queries Query Magnet Comp Fail Query Magnet Failure Query Pop Out Fail and Query Pop Out Failure If due to some reason their ID numbers get changed proper corrections must be done in the queries Risg R 1200 EN 29 References 1 M Modarres What every engineer should know about reliability and risk analysis Marcel Dek ker 1993 2 L Rademakers et al Reliability Analysis Methods for Wind Turbines Task 1 of the project Probabilistic Safety Assessment for Wind Turbines 3 N J McCormick Reliability and Risk Analysis Academic Press 1981 30 Risg R 1200 EN APPENDIX I Event Trees for Wind Turbine Failure States Risg R 1200 EN 31 Figure I Event tree for initiating event Grid loss and condition 5 m ssSws lt 10 m s A B C D E F Component fail ure Oil flow by con trol system Oil flow by cen trifugal 1 tip brake de ploys 2 tip brakes de ploy 3 tip brakes de ploy
29. f the possibility to bring the WTB to a safe halt The reliabil ity analysis of these three categories has specific features that are examined below The reliability and availability analyses are two different types of failure analysis each reflecting a different facet of WTB performance These two analyses are provided based on the data collected in the database The availability analysis takes into account all the failures that affect the WTB s ability to produce electricity due to the time spent on repairing the WTB It does not allow for the frequency of failures In general when the average repair cost is a fraction of the initial equipment cost and the latter is high and the duration of down time affects the volume of production losses one is interested in considering system repair In such a system time between failures repair time and percentage of operating time in an interval are of more interest when analysing system performance The availability function A t is defined as the probability that the system is operational at time t In contrast the reliability function R t is the probability that the system has operated over the interval 0 to t If repair is not permitted then A t reduces to system reliability R t The reliability analysis thus takes into account the fre quency of failure without considering the time spent for recovering the failed components In performing these analyses the following assumption is employed Failu
30. ggregated Model for Calculating the Probability of SS Fail ure The effect of any of three initiating events I Grid loss I Generator failure and L Coupling fail ure is similar in the sense that the WTB loses one of the possibilities to be maintained in a safe mode by the connection grid generator The probability of SS failure can thus be expressed as P SS P SS I OR I OR I P SS 1 P SS I 21 Each term in 21 is calculated on the basis of equations 14 20 as follows P SS 1 OR I OR I P 1 PU PU P C P ABF OR CF P C P AB OR C OR DF P C P AB OR D OR EF P SS I P I P C P AB OR C P C P AB OR COR D P C P AB OR COR DOR E P SS I PU P AB OR E In order to determine the final expression for calculating the probability that the SS will fail it is nec essary to take into account the fact that the intersections of some of the events in the above expressions are not empty The probabilities of those non empty intersections must appear with a minus sign Some of the intersecting events do not appear implicitly i e CM D C two tip brakes cannot de ploy if none can deploy and DM E D all three tips cannot deploy if two cannot deploy either The above conditional probabilities can be written as Risg R 1200 EN 17 P SS I OR I OR I PU PU PU I P C P A P B P F P C P F P F P C P A P B P C P D P F P C P F
31. he following unavailability represen tation which is similar to equation 5 should be done 1 NYTB NYTB 1 ver NIB T down NYT 1 N Comp Lon 1 own I im own NW gt T NIB a T 27 T p gt T ae n T 27 ie 6 i l i i l i i where NY is the number of components in a WTB and F R is the total down time due to failures 1 T down of the jth system in ith WTB Unlike 5 the term Nit gt T was extracted from the average i l i N WTB unavailability in order to be able to calculate the contribution made to the unavailability by the mth component As with the system contribution it can be inferred that in order to rank the components NYTB according to their contribution to the unavailability it is sufficient to rank the terms gt j i l Another useful ranking characteristic is that expressing the contribution to the unavailability made by the system components namely 1 NYTB NYTB 1 Neo 1 NYTB 1 Noor 1 D 1 own own Own NIB gt T Wu 2 Ti p Lr gt Te T i l i i l j l i l j l 1 NYTB NYT 7 Ne omp 1 T down a NM gt Got Tae Fe gt T 7 ml 1 T 10 Ris R 1200 EN where N is the number of components in the kth specific system 7 47 is the total down time caused by the jth component situated in the kth system in the ith WTB Thus the coefficient of relative component contribution to the unavailability of a system can be calculated as the component s contri bution to th
32. horized access to the data and the calculation results Risg R 1200 EN 5 3 Theoretical Foundation 3 1 Definitions The most widely accepted definition of reliability is the ability of an item product system etc to operate under designated operating conditions for a designated period of time or number of cycles 1 This ability can be designated in terms of probability with reliability being defined as follows 2 Re liability is the probability that a product or a system will perform its intended functions satisfactorily i e without failure and within specified performance limits at a certain time for a specified length of time operating under specified environmental and usage conditions Availability analysis is performed to verify that an item has a satisfactory probability of remaining operational so that it can achieve its intended objectives An item s availability can be considered as a combination of its reliability and maintainability When no maintenance or repair is performed reli ability can be considered as instantaneous availability 1 The following two definitions can be em ployed when defining availability 2 Availability is the probability that a product or system will op erate satisfactorily at any point in time where the total time considered includes operating time active repair time administrative time and logistic time An alternative definition 3 is that Availability is the probability that a syste
33. ility and the reliability of all the components and systems especially the safety system The report consists of a description of the theoretical foundation of the reliability and availability analyses and of sections devoted to the development of the WTB reliability models as well as a de scription of the features of the database and software developed Those who are not interested in the mathematical details and who focus on the analysis of the results can skip the theoretical part and pro ceed with the understanding of how to use the database The project was carried out by Ris National Laboratory in collaboration with NEG MICON A S Funding was provided by the Danish Energy Agency Project No 51171 97 0021 2 Objectives The objective of this project was to develop and establish a database for collecting reliability and reli ability related data for assessing the reliability of WTB components and subsystems and WTBs as a whole as well as for assessing WTB availability while ranking the contributions at both the compo nent and system levels The reliability analysis supposes development of the necessary reliability models i e event trees and fault trees All the calculations need to be embedded into a software package which together with the database is self sufficient in performing all the analyses laid down in the methodology of the software system Different levels of access and security features need to be provided to exclude non aut
34. ing SS LC AB OR C OR DF 15 i e the SS fails if the magnetic valve fails and the pop out system fails or one tip brake cannot de ploy or two tip brakes cannot deploy and the mechanical brake fails Initiating event Grid loss Generator failure Driveline failure and wind speed 20 m s lt ws lt 25 m s To ensure WTB safety under these conditions it must be possible for either two tip brakes to deploy and the mechanical brake to deploy or for three tip brakes to deploy The event tree for this case is shown in Appendix I Fig 3 Under these conditions there are 7 pos sible undesired outcomes The final canonical expression for failure of the SS is SS LC AB OR D OR EF 16 i e the SS fails if the magnetic valve fails and the pop out system fails or two tip brakes cannot de ploy or three tip brakes cannot deploy and the mechanical brake fails Initiating event Gearbox failure and wind speed 5 m s lt ws lt 10 m s To ensure WTB safety under these conditions it is sufficient that one tip brake can deploy The other subsystems of the SS are not available The event tree for this case is shown in Appendix I Fig 4 The final canonical expression for fail ure of the SS is SS I C ABOR C 17 i e the SS fails if the magnetic valve fails and the pop out system fails or one tip brake cannot de ploy Initiating event Gearbox failure and wind speed 10 m s lt ws
35. ing event tree and fault tree analyses 3 4 Availability Analysis of Permanently Working Components and Systems Let T represent the random length of the jth operating period having mean for the ith component and D the random length of the jth replacement having mean D for the ith component where j 1 2 1 1 2 n and n is the number of components in the system under consideration Figure 3 illustrates such an alternating sequence of operating and replacement periods State of component i Ti Di To Do Ti Time t Figure 3 Alternating failure and repair for component i 8 Ris R 1200 EN To assess WTB availability the model assumes the following a The system is in series and a system failure thus coincides with a component failure b During replacement of a failed component all other components remain in suspended animation When replacement of the failed component is completed the remaining components abstract op eration At that instant they are not as good as new but only as good as they were when the sys tem stopped operating As t becomes large the availability function reaches the following steady state value 2 1 1 1 1 4 where A and y are the mean failure rate and repair rate of the ith component respectively It is known see for example 2 or can be inferred from 1 independently by inserting the as A D A i T a i 1 a 1
36. lability analysis provides a characterization of system behaviour ena bling some features of maintainability to be modelled 3 In multicomponent systems one can rarely apply approaches 1 and 2 however primarily because the number of components is too great and the interrelationships between the different subsystems are less trivial Fault tree analysis is a method whereby a large number of events that interact to produce other events can be related using simple logical relationships AND OR etc thereby enabling methodical construction of a structure representing the system 4 Some systems can require a more comprehensive analysis with the involvement of different exter nal conditions such as wind speed lightning strikes etc in which case event tree analysis is the appropriate technique for assessing the probabilities of possible outcomes 6 Risg R 1200 EN 3 3 Assumptions for Reliability Modelling All the systems of a WTB can be classified as either permanently working or working on demand Permanently working systems encompass all systems except the safety system SS When undertak ing reliability analysis it is convenient to allocate a third category safety related systems Failure of safety related systems affects not only the ability of the WTB to produce electricity but also the per formance of the SS The systems in question are the generator the driveline and the gearbox The fail ure of any of these systems causes loss o
37. le to predict the reliability of a system one needs a reliability model and information about component fault frequencies Failure data are usually available in most companies in the form of repair reports Reliability modelling is a well known tool in other areas and was introduced in the wind turbine industry by the EFP project Safety Systems for Wind Turbines Method for Evaluation of Failure Modes and Reliability initiated in 1994 This project focussed on WTB Safety Systems i e the systems preventing the turbine from going into over speed under accidental circumstances A need remained though to extend these reliability considerations to the whole wind turbine with the additional aim of predicting its availability i e its ability to produce electricity when wind speeds are adequate The need to embrace the whole wind turbine including the safety system by a com prehensive reliability and availability analysis has necessitated revision and reconstruction of the data base and software developed within the framework of the EFP project The present project comprises analysis of WTBs NM 600 44 600 48 750 44 and 750 48 all of which have similar safety systems As these types of WTB differ from those analysed in the EFP pro ject it was necessary to revise the previously developed reliability models and establish new ones The project resulted in a software package combining a failure database with programs for predicting WTB availab
38. ly be able to trace most of the actions performed Nev ertheless there are some chains of related actions that can be difficult to fathom The most compli cated operations are those carried out for the reliability analyses All the objects involved in these cal culations are described below 26 Risg R 1200 EN Option Reliability Analysis gt WTB Reliability Generator activates the following sequence of actions MS Access Object Description Macro Macro Generator Reliability DeleteObject Table Tab Gen Failure RunCode DateOfFinishEqualNow OpenQuery Query Gen Failure RunCode Generator OpenReport Report Gen Failure RunCode DateOfFinishOriginal Table Tab Gen Failure A subsidiary intermediate table deleted and recreated prior to each analysis Code DateOfFinishEqualNow Inserts into the WTB table Date of Finish field the current date for those WTBs that are in operation at the time of the analysis Action is needed for the calculation of the net work ing experience years Make Table Query Query Gen Failure Creates the table Tab Gen Failure based on the table Components and the select query Query Gen Comp Failure Query Query Gen Comp Failure Selects the generator component failures from the Main Table Counts the number of fail ures for each of the components Code Generator Based on the table Tab Gen Failure and the query WTB T
39. lysis Approaches 6 3 Assumptions for Reliability Modelling 7 4 Availability Analysis of Permanently Working Components and Systems amp 5 Reliability Analysis of Permanently Working Components and Systems 11 6 Reliability Analysis of Components and Systems Working on Demand 2 4 Reliability Analysis of the Safety System 73 4 1 Basic Concepts of Event Tree Analysis 3 4 2 Assumptions for the Safety System 13 4 3 Event Tree Construction 14 4 4 Aggregated Model for Calculating the Probability of SS Failure 7 4 5 System Reliability Modelling 5 Conceptual Database Structure and Capabilities 27 6 Logical Data Model 2 Security Features 23 B User Manual 24 9 Notes for the System Developers 26 References 30 APPENDIX I 31 APPENDIX II 39 Appendix III 4 Appendix IV 44 Appendix V 45 Ris R 1200 EN Risg R 1200 EN 1 Introduction Although Danish wind turbines WTBs are manufactured to very high standards a permanent need exists for manufacturers to maintain performance records for the turbines they sell This is no different from the situation with other types of quality goods Such performance records facilitate the undertak ing of reliability analyses for the preparation of performance documentation for use by the manufac turers research and development departments and potential new customers Reliability analyses can also be used to predict the performance of new designs In order to be ab
40. m will perform a specified function or mission under given conditions at a prescribed time Maintainability is the probability that a product or system will conform to specified conditions within a given period of time when maintenance action is performed according to prescribed proce dures and resources 3 2 Failure Analysis Approaches Systems analysis approaches vary depending on the complexity of the system the diversity of possible failure scenarios and factors such as reparability nonreparability on demand or permanent operation etc The following four approaches can be used for failure analysis of WTBs 1 With systems regarded as simple with nonrepairable components the manner in which they func tion is portrayed by connecting the units in a reliability block diagram 3 All reliability block diagrams are classified as either series parallel k out of n or cross linked structures These sys tems can be in one of two states either operational or failed To some extent such reliability dia grams can also be employed with repairable systems to assess the probability of failures between two down states and mean time between failures as well as probabilities characterizing random time between two failures 2 As system repair is generally initiated after a system has failed a system is either operational or under repair Hence knowledge of system reliability is of less interest than knowledge of the sys tem availability 3 An avai
41. mand rates Code DateOfFinishOriginal Restores the WTB table Date of Finish field to its original state The current dates for those WTBs that are in operation at the time of the analysis are deleted and are easily recognisable The macros Macro Magnet Reliability Macro Pop Out Reliability and Macro Tips Reli ability are similar in structure to Macro Brake Reliability which is described above The differences lie in the names of the objects which are related to the names of the subsystems Macro Macro Demand Rates Since the failure of the generator the gearbox and the driveline are considered initiating events for activation of the SS their failure rates are re garded as demand rates additional to the grid loss and wind speed gt 25 m s This macro activates all the actions needed to calculate the total demand rate DeleteObject Table Tab Drive Failure DeleteObject Table Tab Gear Failure DeleteObject Table Tab Gen Failure RunCode DateOfFinishEqualNow OpenQuery Query Drive Failure OpenQuery Query Gear Failure OpenQuery Query Gen Failure RunCode DemandRate RunCode DateOfFinishOriginal 28 Risg R 1200 EN Table Tab Drive Failure A subsidiary intermediate table deleted and recreated prior to each analysis Tables Tab Gear Failure and Tab Gen Failure have the same design as Tab Drive Fail ure a
42. n reported and that these unreported periods of time have to be excluded Moreover different WTB life histories have to be taken into account when calculating the total operational time see Fig 4 expressed as N JWTB NS WTB T tres past JN gt tres gt past 1 1 i l i l where tpast is the date of the first failure reported in the database NTE is the number of wind turbines removed from operation within the interval tpas tps N WTB is the number of wind turbines entered into service within the interval tpast teres The remaining variables are explained in Fig 4 Risg R 1200 EN 11 5 4 3 2 1 TE al i o p L e ea t tpres time Figure 4 Possible wind turbine life history t time of initiation of the ith wind turbine operation tf the cessation of the ith operation tpas the time data collection was initiated tp es present time The reliability P t of the ith component at time t can be defined as follows P t exp A t 1 At When all the components are connected in series the reliability of the WTB is calculated by the ex pression PO JJRO 0 49 12 There might be some practical difficulty associated with the use of equation 10 due to the fact that some of the component failures might not occur within the analysed time interval and because equa tion 10 thus gives the failure rate as zero i e the component is ideal from the relia
43. nalysis Event trees are inductive logic methods for identifying the various possible outcomes of a given initi ating event The initiating event of an event tree is either a system failure or an external event that can end in an undesired outcome The effect of an initiating event on a system depends on what might happen next and the sequence of occurrences As a result several possible scenarios can be developed that could possibly have severe impact on the system and the environment From a theoretical point of view the probability that a specific system will fail is conditional on the initiating events To analyse a SS which is activated on demand we have first to identify all possible initiating events The probability of its failure can then be calculated as P SS gt P SS 1 PU where SS designates the failure of the SS and is the ith initiating event All the initiating events together constitute the partitioning of the set of all possible initiating events This means that S PU 1 ie all possible initiating events are included in the analysis Sometimes it is unrealistic L to consider all the events and instead only those making the greatest contribution to the final probabil ity are taken into account This is usual practice when analysing the reliability and risk of large scale technical systems such as chemical plants or nuclear power plants Initiating events can be considered under different external conditions in
44. nd finally will keep the results of the calculations of their failure rates Code DateOfFinishEqualNow See above Make Table Query Query Drive Failure Creates the table Tab Drive Failure based on the table Components and the select query Query Drive Comp Failure Query Query Drive Comp Failure Selects the driveline component failures from the Main Table Counts the number of fail ures for each of the components Queries Query Gear Failure and Query Gen Failure have the same design as Query Drive Failure Code DemandRate Based on the tables Tab Gen Failure Tab Gear Failure Tab Drive Failure and Ex ternal Events and the query WTB Total Time Calculates the total demand rate and transfers the results to the table Demand Rates Code DateOfFinishOriginal See above Code SSRates Based on the tables Tab Brake Failure Tab Magnet Failure Tab Pop Out Failure Tab Tips Failure External Events SS Subsys tem Reliab and Demand Rates Calculates the reliabilities of the SS subsystems and the SS as a whole on demand and transfers the results to the table Tab SS Reliability Report Report SS Subsystems Reports the results of the SS reliability calcu lations based on the table SS Subsystem Re liab All the objects in the database are interlinked and any changes in the design will lead to failure of some o
45. ng Some of the events in equation 22 are trivial and their probabilities can be defined directly from the data collected in the database or in the case of wind conditions from external sources of informa tion These events or conditions do not need to be broken down and no failure models are required for them The probability of the grid loss P L and probabilities P C P C2 and P C3 thus cannot be determined from the data stored in the database but only from other data and hence have to be stored in the database in the form of constants that can be periodically updated when more precise data be come available 18 Risg R 1200 EN Generator failure P I3 The generator consists of the following components the failure of each of which will cause generator failure Bolte RPM sensor St dd mper Kobling Lejer Temperatursensor Terminaler Kabler Slanger Pumpe Vandd kning From the reliability standpoint all of the components are connected in series i e the generator fails if the bolte fails or RPM sensor fails or the st dd mper fails etc The probability of generator failure is thus P Generator failure 1 P bolte x P RPM sensor x P steddeemper x 23 where all the terms are the probabilities that the individual components are in non failure state Gearbox failure P Is The gearbox consists of the following components the failure of each of which will cause generator failure
46. nking the contributions at both the component and system levels The project resulted in a software package combining a fail ure database with programs for predicting WTB availability and the reliability of all the components and systems especially the safety system The report consists of a description of the theoretical foun dation of the reliability and availability analyses and of sections devoted to the development of the WTB reliability models as well as a description of the features of the database and software developed The project comprises analysis of WTBs NM 600 44 600 48 750 44 and 750 48 all of which have similar safety systems The database was established with Microsoft Access Database Management System the software for reliability and availability assessments was created with Visual Basic Available on request from Information Service Department Riso National Laboratory Afdelingen for Informationsservice Forskningscenter Riso P O Box 49 DK 4000 Roskilde Denmark Telephone 45 4677 4004 Telefax 45 4677 4013 Figure I Event tree for initiating event Grid loss and condition 5 m ssSws lt 10 m s A B C D E F Component fail ure Oil flow by con trol system Oil flow by cen trifugal 1 tip brake de ploys 2 tip brakes de ploy 3 tip brakes de ploy Mechanical brake Sequence for mula Success state
47. otal Time this code calculates the reliabilities of all the generator compo nents and the generator as a whole and trans fers the results to the table Tab Gen Failure Query WTB Total Time Counts the number of WTBs under observa tion and listed in the table WTB and the net working experience accumulated by all the WTBs Report Report Gen Failure Reports all the calculated reliabilities and shows them on the screen Is based on the ta ble Tab Gen Failure Code DateOfFinishOriginal Restores the WTB table Date of Finish field to its original state The current dates for those WTBs that are in operation at the time of the analysis The listed sequence of the actions is repeated each time the user makes a request and is similar for each of the following systems Generator Gearbox Driveline Steering system Nacelle Tower Print board Yawing system Main shaft and Cover The differences lie in the names of some of the objects which clearly indicate what system they are related to In the case of the gearbox for example the names are Macro Gearbox Reliability Tab Gear failure Query Gear Failure Gearbox and Report Gear Failure The reliability calculations for the Safety System and its subsystems are carried out differently Risg R 1200 EN 27 MS Access Object Description Macro Macro SS Rates RunMacro Macro Brake Reliabili
48. our claim Riso R 1200 EN Failure Database and Tools for Wind Turbine Availability and Reliability Analyses The Application of Reliability Data for Selected Wind Turbines Igor Kozine Palle Christensen and Martin Winther Jensen Rise National Laboratory Roskilde January 2000 Abstract The objective of this project was to develop and establish a database for collecting reliabil ity and reliability related data for assessing the reliability of wind turbine components and subsystems and wind turbines as a whole as well as for assessing wind turbine availability while ranking the con tributions at both the component and system levels The project resulted in a software package com bining a failure database with programs for predicting WTB availability and the reliability of all the components and systems especially the safety system The report consists of a description of the theo retical foundation of the reliability and availability analyses and of sections devoted to the develop ment of the WTB reliability models as well as a description of the features of the database and soft ware developed The project comprises analysis of WTBs NM 600 44 600 48 750 44 and 750 48 all of which have similar safety systems ISBN 87 550 2732 6 ISBN 87 550 273 1 8 internet ISSN 0106 2840 Information Service Department Ris 2000 Contents I Introduction 5 2 Objectives 5 3 Theoretical Foundation 6 1 Definitions 6 2 Failure Ana
49. re of any component of a WTB except the components constituting the safety system leads to the WTB being shut down for repair of the component during which time the WTB remains idle This assumption defines the structure of the reliability block diagram and the fault tree All the com ponents can thus be considered to be connected in series i e the block diagram is a series structure and all the events in the fault tree are connected through OR gates Figures 1 and 2 Thus the WTB fails if any of n components fail or if component 1 fails or component 2 fails and so on Figure 1 Reliability block diagram of a WTB n is the number of components in the WTB Risg R 1200 EN 7 WTB Fails Figure 2 Fault tree for WTB failure The SS failure analysis is twofold Firstly some SS failures affect WTB performance through extend ing the repair time Such failures can be revealed 1 when the SS is activated on demand or 2 dur ing a periodic test control If there is a failure in the SS and the WTB can be brought to a safe halt the repair time spent on recovering the SS simply decreases the availability of the WTB thus contributing to the total unavailability Such failures are taken into account in the availability analysis Secondly some SS failures might affect the WTB s ability to be maintained in a safe mode Such failures are of particular interest and their likelihood must be analysed by means of reliability analysis conducted us
50. rocessed in the database as a two component system that is considered to be in a state of failure if the Sikkerhedsventil pop out fails and the Spraengblik fails i e the com ponents are connected in parallel from the reliability standpoint This system failure is designated by B 20 Ris R 1200 EN 5 Conceptual Database Structure and Capabili ties Now that all the calculations necessary for the availability and reliability analyses have been defined and the event and fault trees developed the database structure and capabilities can be defined The following options are provided by the database 1 Availability analysis Assessment of the average availability of a certain type of WTB and ranking of all specific WTBs within a certain type sample according to availability Assessment of the contribution made by system unavailability to the average unavailability of a certain type of WTB and ranking the systems according to their contribution to unavailability Assessment of the contribution made by component unavailability to the average unavailability of a certain type of WTB and ranking the components according to their contribution to unavailability Assessment of the contribution made by component unavailability to the unavailability of the sys tem to which the component belongs and ranking of the components according to their contribution to system unavailability Assessment of interannual variation in availability 2
51. s deploy 3 tip brakes deploy Sequence formula ABCD ABC ABCD ABC Figure 6 Event tree for initiating event Gearbox failure and condition 20 m sSws lt 25 m s A B C D E Component Oil flow by Oil flow by 1 tip brake 2 tip brakes 3 tip brakes Sequence failure control system centrifugal deploys deploy deploy formula ABCDE ABCD ABC ABCDE ABCD ABC ABCDE ABCD ABC AB Figure 7 Event tree for initiating event ws225 m s A B E ER ane onan ABE ABE ABE
52. ss 3 In the Workgroup Administrator dialogue box click Create and then type your name and or ganization 4 In the Workgroup Owner Information dialogue box type any combination of up to 20 numbers and letters and then click OK Caution Be sure to write down your exact name organization and workgroup ID carefully distin guishing between upper case and lower case letters for all three entries and store the information in a safe place If you have to re create the workgroup information file you must supply exactly the same name organization and workgroup ID If you forget or lose these entries you cannot recover them and might lose access to your databases 5 Type a new name for the new workgroup information file and then click OK By default the workgroup information file is saved in the folder where you installed Microsoft Access To save in a different location type a new path or click Browse to specify the new path The new workgroup information file is used the next time you start Microsoft Access Any user and group accounts or passwords you create are saved in the new workgroup information file To have others join the workgroup defined by your new workgroup information file copy it to a shared folder if you did not already save it in a shared folder in step 4 and then have each user run the Work group Administrator to join the new workgroup information file 44 Risg R 1200 E
53. stration IDs and passwords is saved in the workgroup information file This file must be used with the database in order to be able to gain access to it The current valid workgroup information file is RELIAB MDW When it was created the following information was incorporated Name Wind Turbines Organization NEG MICON Workgroup ID Reliab For more about the workgroup information file and the security system see Appendix IV or Help in MS Access The current security settings are listed below It is necessary to know them in order to be able to gain access and change them in the future Newly created groups and users Group NotAdmin Registration ID 999999 Group Typists Registration ID 111111 Administrator Reg ID Igor Password Kozine Analyst Reg ID Palle Password 123456 Typist Reg ID Lise Password 234567 It should be remembered that the passwords and security data are case sensitive 8 User Manual The database interface is self contained in the sense that it is not necessary to use the Access toolbars but it is sufficient to use the command buttons on the forms created The structure of the interface is uncomplicated and hence does not need to be described in detail Some general notes on the options provided are given below The database is divided into two parts Data Tables and Analyses Data Tables allows viewing editing and data insertion in the tables Main Table Compo nents and
54. tion file When you install Microsoft Access the Setup program automatically creates a Microsoft Access workgroup information file that is identified by the name and organization information you specify Because this information is often easy to determine it is possible for unauthorized users to create an other version of this workgroup information file and consequently irrevocably assume the permitted operations of an administrator account i e an Admins group user in the workgroup defined by that workgroup information file To prevent this you should therefore create a new workgroup information file and specify a workgroup ID WID Thereafter only persons knowing the new WID will be able to create a copy of the workgroup information file 1 Exit Microsoft Access 2 To start the Workgroup Administrator do one of the following depending on which operating system you are using e If you are using Windows 95 use My Computer or Windows Explorer to open the System subfolder in the Windows folder and then double click Wrkgadm exe e If you are using Windows NT Workstation 4 0 use My Computer or Windows Explorer to open the System32 subfolder in the WinNT folder and then double click Wrkgadm exe If you are using Windows NT Workstation 3 51 open Program Manager and then double click the Workgroup Administrator icon in the program group where you installed Microsoft Acce
55. ty RunMacro Macro Magnet Reliability RunMacro Macro Pop Out Reliability RunMacro Macro Tips Reliability RunMacro Macro Demand Rates RunCode SSRates OpenReport Report SS Subsystems Macro Macro Brake Reliability DeleteObject Table Tab Brake Failure RunCode DateOfFinishEqualNow OpenQuery Query Brake Failure RunCode MechanicalBrake RunCode DateOfFinishOriginal Table Tab Brake Failure A subsidiary intermediate table deleted and recreated prior to each analysis Code DateOfFinishEqualNow Inserts into the WTB table Date of Finish field the current date for those WTBs that are in operation at the time of the analysis Action is needed for the calculation of the net work ing experience years Make Table Query Query Brake Failure Creates the table Tab Brake Failure based on the table Components and the select query Query Brake Comp Failure Query Query Brake Comp Failure Selects the mechanical brake component fail ures from the Main Table Counts the num ber of failures for each of the components Code MechanicalBrake Based on the table Tab Brake Failure and the query WTB Total Time the code calcu lates the failure rates of all the components of the mechanical brake and the system as a whole and transfers the results to the table Tab Gen Failure The results are intermedi ate and do not take into account the de
Download Pdf Manuals
Related Search