Failure Modes, Effects and Diagnostic Analysis

Transcription

1 Failure Modes, Effects and Diagnostic Analysis Project: Temperature Transmitters TT*300-*H with ma output Customer: ABB Automation Products GmbH Minden Germany Contract No.: ABB 06/05-29 Report No.: ABB 06/05-29 R012 Version V4, Revision R1; January 2016 Stephan Aschenbrenner The document was prepared using best effort. The authors make no warranty of any kind and shall not be liable in any event for incidental or consequential damages in connection with the application of the document. All rights reserved.

2 Management summary This report summarizes the results of the hardware assessment according to IEC carried out on the Temperature Transmitters TT*300-*H with ma output. The Temperature Transmitter TT*300-*H is a configurable single or dual sensor channel (1 or 2 x RTD 2/3/4 wire, 2 x TE, 2 x mv, 1 x RTD 2/3 and 1 x TE / mv) analog 4..20mA device output. Table 1 gives an overview of the different types that belong to the considered Temperature Transmitters TT*300-*H with ma output including hardware and software version The hardware assessment consists of a Failure Modes, Effects and Diagnostics Analysis (FMEDA). A FMEDA is one of the steps taken to achieve functional safety assessment of a device per IEC From the FMEDA, failure rates are determined and consequently the Safe Failure Fraction (SFF) is calculated for the device. For full assessment purposes all requirements of IEC must be considered. Table 1: Version overview Type Description HW Version SW Version TTH300-*H Head mounted temperature transmitter 1.06 / / TTF300-*H Field mounted temperature transmitter 1.06 / / For safety applications only the ma output was considered. All other possible output variants or electronics are not covered by this report. The failure modes used in this analysis are from the exida Electrical Component Reliability Handbook ([N2]). The failure rates are the basic failure rates from the Siemens standard SN ([N3]). The Temperature Transmitters TT*300-*H with ma output can be considered to be Type B 1 elements with a hardware fault tolerance of 0. The failure rates do not include failures resulting from incorrect use of the Temperature Transmitters TT*300-*H with ma output, in particular humidity entering through incompletely closed housings or inadequate cable feeding through the inlets. The listed failure rates are valid for operating stress conditions typical of an industrial field environment similar to IEC class C (sheltered location) with an average temperature over a long period of time of 40ºC. For a higher average temperature of 60 C, the failure rates should be multiplied with an experience based factor of 2.5. A similar multiplier should be used if frequent temperature fluctuation must be assumed. It is assumed that the connected logic solver is configured per the NAMUR NE43 signal ranges, i.e. the Temperature Transmitters TT*300-*H with ma output communicate detected faults by an alarm output current 3,6mA or 21mA. Assuming that the application program in the safety logic solver does not automatically trip on these failures, these failures have been classified as dangerous detected failures. The following tables show how the above stated requirements are fulfilled. 1 Type B subsystem: Complex subsystem (using micro controllers or programmable logic); for details see of IEC Stephan Aschenbrenner Page 2 of 29

3 Table 2: Summary - Failure rates according to IEC 61508: Failure category Failure rates (in FIT) Safe Detected (λsd) 0 Safe Undetected (λsu) 0 Dangerous Detected (λdd) 313 Dangerous Detected (λ dd); by internal diagnostics or indirectly High (λ H); detected by the logic solver 22 Low (λ L) ; detected by the logic solver 78 Annunciation Detected (λ AD) 0 Dangerous Undetected (λdu) 34 Annunciation Undetected (λ AU) 6 No effect (λ #) 118 No part (λ -) 145 Total failure rate of the safety function (λtotal) 347 Safe failure fraction (SFF) 4 90% DC 90% SIL AC 5 SIL 2 A complete temperature sensor assembly consisting of the Temperature Transmitters TT*300- *H and a thermocouple or RTD can be modeled by considering a series subsystem where a failure occurs if there is a failure in either component. For such a system, failure rates are added. The failure rates are valid for the useful life of the Temperature Transmitters TT*300-*H with ma output (see Appendix 2). Appendix 3 gives typical failure rates and failure distributions for thermocouples and RTDs which were the basis for the following tables. Assuming that the Temperature Transmitter TT*300-*H will go to the pre-defined alarm state on detected failures of the thermocouple or RTD, the failure rate contribution for the thermocouple or RTD in a low stress environment is as follows: 2 It is assumed that practical fault insertion tests can demonstrate the correctness of the failure effects assumed during the FMEDAs. 3 indirectly means that these failure are not necessarily detected by diagnostics but lead to either fail low or fail high failures depending on the transmitter setting and are therefore detectable. 4 The complete sensor element will need to be evaluated to determine the overall Safe Failure Fraction. The number listed is for reference only. 5 SIL AC (architectural constraints) means that the calculated values are within the range for hardware architectural constraints for the corresponding SIL but does not imply all related IEC requirements are fulfilled. In addition it must be shown that the device has a suitable systematic capability for the required SIL and that the entire safety function can fulfill the required PFDAVG / PFH value. Stephan Aschenbrenner Page 3 of 29

4 Table 3: TT*300-*H with thermocouple (close coupled) 0 FIT 0 FIT 408 FIT 39 FIT 91% Table 4: TT*300-*H with two thermocouples (close coupled) 0 FIT 0 FIT 518 FIT 32 FIT 94% Table 5: TT*300-*H with 2/3-wire RTD (close coupled) 0 FIT 0 FIT 352 FIT 42 FIT 89% Table 6: TT*300-*H with two 2/3-wire RTDs (close coupled) 0 FIT 0 FIT 414 FIT 32 FIT 92% Table 7: TT*300-*H with thermocouple and 2/3-wire RTD (close coupled) 0 FIT 0 FIT 466 FIT 37 FIT 92% Table 8: TT*300-*H with 4-wire RTD (close coupled) 0 FIT 0 FIT 360 FIT 36 FIT 90% Stephan Aschenbrenner Page 4 of 29

5 Table 9: TT*300-*H with thermocouple (with extension wire) 0 FIT 0 FIT 1213 FIT 134 FIT 90% Table 10: TT*300-*H with two thermocouples (with extension wire) 0 FIT 0 FIT 2309 FIT 41 FIT 98% Table 11: TT*300-*H with 2/3-wire RTD (with extension wire) 0 FIT 0 FIT 693 FIT 129 FIT 84% Table 12: TT*300-*H with two 2/3-wire RTDs (with extension wire) 0 FIT 0 FIT 1259 FIT 41 FIT 96% Table 13: TT*300-*H with thermocouple and 2/3-wire RTD (with extension wire) 0 FIT 0 FIT 1784 FIT 41 FIT 97% Table 14: TT*300-*H with 4-wire RTD (with extension wire) 0 FIT 0 FIT 808 FIT 39 FIT 95% Stephan Aschenbrenner Page 5 of 29

6 Assuming that the Temperature Transmitters TT*300-*H will go to the pre-defined alarm state on detected failures of the thermocouple or RTD, the failure rate contribution for the thermocouple or RTD in a high stress environment is as follows: Table 15: TT*300-*H with thermocouple (close coupled) 0 FIT 0 FIT 2213 FIT 134 FIT 94% Table 16: TT*300-*H with two thermocouples (close coupled) 0 FIT 0 FIT 4309 FIT 41 FIT 99% Table 17: TT*300-*H with 2/3-wire RTD (close coupled) 0 FIT 0 FIT 1100 FIT 207 FIT 84% Table 18: TT*300-*H with two 2/3-wire RTDs (close coupled) 0 FIT 0 FIT 2221 FIT 48 FIT 97% Table 19: TT*300-*H with thermocouple and 2/3-wire RTD (close coupled) 0 FIT 0 FIT 3265 FIT 139 FIT 95% Table 20: TT*300-*H with 4-wire RTD (close coupled) 0 FIT 0 FIT 1263 FIT 84 FIT 93% Stephan Aschenbrenner Page 6 of 29

7 Table 21: TT*300-*H with thermocouple (with extension wire) 0 FIT 0 FIT FIT 2034 FIT 90% Table 22: TT*300-*H with two thermocouples (with extension wire) 0 FIT 0 FIT FIT 231 FIT 99% Table 23: TT*300-*H with 2/3-wire RTD (with extension wire) 0 FIT 0 FIT 7913 FIT 1934 FIT 80% Table 24: TT*300-*H with two 2/3-wire RTDs (with extension wire) 0 FIT 0 FIT FIT 221 FIT 98% Table 25: TT*300-*H with thermocouple and 2/3-wire RTD (with extension wire) 0 FIT 0 FIT FIT 226 FIT 99% Table 26: TT*300-*H with 4-wire RTD (with extension wire) 0 FIT 0 FIT FIT 134 FIT 98% Stephan Aschenbrenner Page 7 of 29

8 Table of Contents Management summary Purpose and Scope Project management exida Roles of the parties involved Standards / Literature used exida tools used Reference documents Documentation provided by the customer Documentation generated by exida Description of the analyzed module System description Failure Modes, Effects, and Diagnostics Analysis Description of the failure categories Methodology FMEDA, Failure rates FMEDA Failure rates Assumptions Results Temperature Transmitters TT*300-*H Using the FMEDA results Example PFD AVG calculation Terms and Definitions Status of the document Liability Releases Appendix 1: Possibilities to reveal dangerous undetected faults during the proof test.. 22 Appendix 1.1: Possible proof tests to detect dangerous undetected faults Appendix 2: Impact of lifetime of critical components on the failure rate Appendix 3: Using the FMEDA results Appendix 3.1: TT*300-*H with thermocouple Appendix 3.2: TT*300-*H with RTD Appendix 3.3: TT*300-*H in redundant mode with drift monitoring Stephan Aschenbrenner Page 8 of 29

9 1 Purpose and Scope This document shall describe the results of the FMEDA carried out on the Temperature Transmitters TT*300-*H with ma output. Table 1 gives an overview of the different types which have been assessed including the considered hardware and software versions. The FMEDA is one part of the complete functional safety assessment according to IEC The FMEDA builds the basis for an evaluation whether whether the Temperature Transmitters TT*300-*H with ma output meet the average Probability of Failure on Demand (PFD AVG) requirements and the architectural constraints / minimum hardware fault tolerance requirements per IEC It does not consider any calculations necessary for proving intrinsic safety. Stephan Aschenbrenner Page 9 of 29

10 2 Project management 2.1 exida exida is one of the world s leading knowledge and certification companies specializing in automation system safety and availability with over 300 years of cumulative experience in functional safety. Founded by several of the world s top reliability and safety experts from assessment organizations and manufacturers, exida is a global company with offices around the world. exida offers training, coaching, project oriented consulting services, internet based safety engineering tools, detail product assurance and certification analysis and a collection of on-line safety and reliability resources. exida maintains a comprehensive failure rate and failure mode database on process equipment. 2.2 Roles of the parties involved ABB Automation Products exida Manufacturer of the Temperature Transmitters TT*300-*H with ma output. Performed the FMEDA. ABB Automation Products contracted exida in May 2014 with the update of the corresponding FMEDAs and this report according to IEC 61508: Standards / Literature used The services delivered by exida were performed based on the following standards / literature. [N1] IEC :2010 Functional Safety of Electrical/Electronic/Programmable Electronic Safety-Related Systems [N2] Electrical Component Reliability Handbook, 3rd Edition, 2012 [N3] SN : SN H1: SN : SN : SN : SN : SN : SN : SN : SN : SN : SN : SN : exida LLC, Electrical Component Reliability Handbook, Third Edition, 2012, ISBN Siemens standard with failure rates for components [N4] IEC : , second edition Industrial-process measurement and control equipment Operating conditions Part 1: Climatic condition [N5] Summary of failure rates and distributions for RTD and TC V1R4.xls of exida summary of failure rates and distributions for RTD and TC Stephan Aschenbrenner Page 10 of 29

11 2.4 exida tools used [T1] FMEDA Tool V5 FMEDA Tool [T2] exsilentia Ultimate V3 SIL Verification Tool 2.5 Reference documents Documentation provided by the customer [D1] DS_TTF300-EN-C-09_2011.pdf Data Sheet DS/TTF300-EN Rev. C [D2] DS_TTH300-EN-B-12_2010.pdf Data Sheet DS/TTH300-EN Rev. B [D3] TTX_ELECTRONIC_SCH_SAP218533_REV107.pdf Circuit diagram UHTE Rev 1.07 of [D4] [D5] Fault Insertion Test Sunrise_Final_ pdf Fehlerversuche_ _Emulator_M16C_MFW01_01_01.pdf Fehlerversuche_ _Emulator_M16C_SFW01_01_10.pdf Tuev_assessment_ _Emulator_M16C_Tests.pdf Fault Insertion Test Sunrise_TÜV_Audittermin_ _Test_Protocol.pdf EMI_Impacttest_FW pdf Fault Insertion Test Sunrise_FW _ pdf Fault Insertion Test Sunrise_FW _ _Test_Enviroment.pdf Fehlerversuche_ _Emulator_M16C_MFW01_01_03.pdf Fault Insertion Test AD Converter.pdf IIM-IA Impact Analysis_TTX_300- HART_FW_1 1 8_to_FW_1 3 0_Rev_1.1.pdf Fault insertion tests Impact analysis for changes from hardware version 1.06 to 1.07 and firmware changes from to [D6] EMC TestReport TTX200_TTX3x0-HART 1_07.pdf EMC Test Report TTH300-.H HW. Rev. 1.07/ TTF300-.H Rev [D7] [D8] II-A TT - Test Report TTX Stability Test HART 7 Mode.pdf II-A TT - Test Report TTX Stability Test HART 5 Mode.pdf Regression test TTH300 Software Rev HART 7 Mode Regression test TTH300 Software Rev HART 5 Mode Stephan Aschenbrenner Page 11 of 29

12 [D9] IIM-A TT - Test Protocol TTX Regression Test.pdf [D10] IIM-A TT - Test Summary Report TTX300 Firmware pdf [D11] IIM-A TT - Software Code Review TTX300 HART7 FW pdf [D12] IIM-A TT - Factory Calibration Certificate TTH300 with Software Revision pdf [D13] IIM-FG TT-Release Report TTX300 MFW pdf Test Protocol summary of all executed tests Test report regarding Change Impact Measures and Regression Tests Software Code Review regarding Code changes from Firmware to Calibration Certificate Release Report TTX300 with Main Firmware Version Documentation generated by exida [R1] FMEDA V5 UHTE V5R3 Temperature.xls of [R2] FMEDA V5 UHTE V5R3 Temperature dual mode with drift alarm.xls of [R3] Zusammenfassung Fehlerraten V5R3.xls of Stephan Aschenbrenner Page 12 of 29

13 3 Description of the analyzed module 3.1 System description The Temperature Transmitters TT*300-*H are isolated two-wire ma devices used in many different industries for both control and safety applications. Combined with a temperature sensing device, the temperature transmitters become a temperature sensor assembly. The Temperature Transmitters TT*300-*H with ma output and with a temperature sensing device can be considered to be Type B elements with a hardware fault tolerance of 0. The temperature sensing devices that can be connected to the Temperature Transmitters TT*300-*H are listed underneath: 2-, 3-, and 4-wire RTD Thermocouple The FMEDA has been performed considering the worst-case input sensor configuration. Figure 1 gives an overview of the Temperature Transmitters TT*300-*H with ma output. Figure 1: Temperature Transmitters TT*300-*H with ma output Stephan Aschenbrenner Page 13 of 29

14 4 Failure Modes, Effects, and Diagnostics Analysis The Failure Modes, Effects, and Diagnostic Analysis was done by exida together with ABB Automation Products. The results are documented in [R1] and [R2]. When the effect of a certain failure mode could not be analyzed theoretically, the failure modes were introduced on component level and the effects of these failure modes were examined on system level (see fault insertion test report [D4]). This resulted in failures that can be classified according to the following failure categories. 4.1 Description of the failure categories In order to judge the failure behavior of the Temperature Transmitters TT*300-*H with ma output, the following definitions for the failure of the product were considered. Fail-Safe State The fail-safe state is defined as output reaching the user defined threshold value. Fail Safe A safe failure (S) is defined as a failure that plays a part in implementing the safety function that: a) results in the spurious operation of the safety function to put the EUC (or part thereof) into a safe state or maintain a safe state; or, b) increases the probability of the spurious operation of the safety function to put the EUC (or part thereof) into a safe state or maintain a safe state. Fail Dangerous A dangerous failure (D) is defined as a failure that plays a part in implementing the safety function that: a) deviates the output current by more than 2% of full span and prevents a safety function from operating when required (demand mode) or causes a safety function to fail (continuous mode) such that the EUC is put into a hazardous or potentially hazardous state; or, b) decreases the probability that the safety function operates correctly when required. Dangerous Undetected Failure that is dangerous and that is not being diagnosed. Dangerous Detected Failure that is dangerous but is detected by internal or external testing. Fail high A fail high failure (H) is defined as a failure that causes the output signal to go to the maximum output current (> 21mA). Fail low A fail low failure (L) is defined as a failure that causes the output signal to go to the minimum output current (< 3.6mA). Annunciation Failure that does not directly impact safety but does impact the ability to detect a future fault (such as a fault in a diagnostic circuit). Annunciation failures are divided into annunciation detected (AD) and annunciation undetected (AU) failures. No effect Failure mode of a component that plays a part in implementing the safety function but is neither a safe failure nor a dangerous failure. No part Component that plays no part in implementing the safety function but is part of the circuit diagram and is listed for completeness. Stephan Aschenbrenner Page 14 of 29

15 4.2 Methodology FMEDA, Failure rates FMEDA A Failure Modes and Effects Analysis (FMEA) is a systematic way to identify and evaluate the effects of different component failure modes, to determine what could eliminate or reduce the chance of failure, and to document the system in consideration. A FMEDA (Failure Modes, Effects, and Diagnostic Analysis) is a FMEA extension. It combines standard FMEA techniques with extension to identify online diagnostics techniques and the failure modes relevant to safety instrumented system design. It is a technique recommended to generate failure rates for each important category (safe detected, safe undetected, dangerous detected, dangerous undetected) in the safety models. The format for the FMEDA is an extension of the standard FMEA format from MIL STD 1629A, Failure Modes and Effects Analysis Failure rates The failure modes used in this analysis are from the exida Electrical Component Reliability Handbook ([N2]). The failure rates are the basic failure rates from the Siemens standard SN ([N3]). The rates were chosen in a way that is appropriate for safety integrity level verification calculations. The rates were chosen to match operating stress conditions typical of an industrial field environment similar to IEC , class C. It is expected that the actual number of field failures will be less than the number predicted by these failure rates. For hardware assessment according to IEC only random equipment failures are of interest. It is assumed that the equipment has been properly selected for the application and is adequately commissioned such that early life failures (infant mortality) may be excluded from the analysis. Failures caused by external events however should be considered as random failures. Examples of such failures are loss of power, physical abuse, or problems due to intermittent instrument air quality. The assumption is also made that the equipment is maintained per the requirements of IEC or IEC and therefore a preventative maintenance program is in place to replace equipment before the end of its useful life. The user of these numbers is responsible for determining their applicability to any particular environment. Accurate plant specific data may be used for this purpose. If a user has data collected from a good proof test reporting system such as exida SILStat TM that indicates higher failure rates, the higher numbers shall be used. Some industrial plant sites have high levels of stress. Under those conditions the failure rate data is adjusted to a higher value to account for the specific conditions of the plant. Stephan Aschenbrenner Page 15 of 29

16 4.2.3 Assumptions The following assumptions have been made during the Failure Modes, Effects, and Diagnostic Analysis of the Temperature Transmitters TT*300-*H with ma output. Failure rates are constant, wear out mechanisms are not included. Propagation of failures is not relevant. Failures during parameterization are not considered. The device is locked against unintended operation/modification. The HART protocol is only used for setup, calibration, and diagnostics purposes, not during normal operation. Practical fault insertion tests can demonstrate the correctness of the failure effects assumed during the FMEDAs. Sufficient tests are performed prior to shipment to verify the absence of vendor and/or manufacturing defects that prevent proper operation of specified functionality to product specifications or cause operation different from the design analyzed. Materials are compatible with process conditions. The mean time to restoration (MTTR) after a safe failure is 24 hours. The worst-case internal fault detection time is 2 minutes. Depending on the application, this interval needs to be considered directly in the SIL verification. The common cause factor for the temperature sensing devices in redundant mode is considered to be 5%. The end-user is responsible for determining the appropriate common cause factor for a particular application. The listed failure rates are valid for operating stress conditions typical of an industrial field environment similar to IEC class C with an average temperature over a long period of time of 40ºC. For a higher average temperature of 60 C, the failure rates should be multiplied with an experience based factor of 2.5. A similar multiplier should be used if frequent temperature fluctuation must be assumed. For safety applications only the considered ma output is used. The ma output signal is fed to a SIL 2 compliant analog input board of a safety PLC. The device is operated in the low demand mode of operation. External power supply failure rates are not included. When using the ma output signal the application program in the safety logic solver is configured according to NAMUR NE43 to detect under-range and over-range failures and does not automatically trip on these failures; therefore these failures have been classified as dangerous detected failures. Stephan Aschenbrenner Page 16 of 29

17 5 Results DC = λ DD / (λ DD + λ DU) λ total = λ SD + λ SU + λ DD + λ DU According to IEC the architectural constraints of an element must be determined. This can be done by following the 1 H approach according to of IEC or the 2 H approach according to of IEC The 1 H approach involves calculating the Safe Failure Fraction for the entire element. The 2 H approach involves assessment of the reliability data for the entire element according to of IEC This assessment supports the 1 H approach. According to of IEC , the Safe Failure Fraction is the property of a safety related element that is defined by the ratio of the average failure rates of safe plus dangerous detected failures and safe plus dangerous failures. This ratio is represented by the following equation: SFF = (Σλ S avg + Σλ DD avg) / (Σλ S avg + Σλ DD avg+ Σλ DU avg) When the failure rates are based on constant failure rates, as in this analysis, the equation can be simplified to: SFF = (Σλ S + Σλ DD) / (Σλ S + Σλ DD + Σλ DU) Where: λ S = Fail Safe λ DD = Fail Dangerous Detected λ DU = Fail Dangerous Undetected As the Temperature Transmitters TT*300-*H with ma are only one part of a sensor element, the architectural constraints should be determined for the entire sensor element. Stephan Aschenbrenner Page 17 of 29

18 5.1 Temperature Transmitters TT*300-*H The FMEDA carried out on the Temperature Transmitters TT*300-*H with ma output leads under the assumptions described in section to the following failure rates: Failure category Failure rates (in FIT) Safe Detected (λsd) 0 Safe Undetected (λsu) 0 Dangerous Detected (λdd) 313 Dangerous Detected (λ dd); by internal diagnostics or indirectly High (λ H); detected by the logic solver 22 Low (λ L) ; detected by the logic solver 78 Annunciation Detected (λ AD) 0 Dangerous Undetected (λdu) 34 Annunciation Undetected (λ AU) 6 No effect (λ #) 118 No part (λ -) 145 Total failure rate of the safety function (λtotal) 347 Safe failure fraction (SFF) 7 90% DC 90% SIL AC 8 SIL 2 6 indirectly means that these failure are not necessarily detected by diagnostics but lead to either fail low or fail high failures depending on the transmitter setting and are therefore detectable. 7 The complete sensor element will need to be evaluated to determine the overall Safe Failure Fraction. The number listed is for reference only. 8 SIL AC (architectural constraints) means that the calculated values are within the range for hardware architectural constraints for the corresponding SIL but does not imply all related IEC requirements are fulfilled. In addition it must be shown that the device has a suitable systematic capability for the required SIL and that the entire safety function can fulfill the required PFDAVG / PFH value. Stephan Aschenbrenner Page 18 of 29

19 6 Using the FMEDA results The following section describes how to apply the results of the FMEDA. It is the responsibility of the Safety Instrumented Function designer to do calculations for the entire SIF. exida recommends the accurate Markov based exsilentia tool for this purpose. The following results must be considered in combination with PFD AVG values of other devices of a Safety Instrumented Function (SIF) in order to determine suitability for a specific Safety Integrity Level (SIL). 6.1 Example PFDAVG calculation An average Probability of Failure on Demand (PFD AVG) calculation is performed for a single (1oo1) Temperature Transmitter TT*300-*H with ma output considering a proof test coverage of 74% as indicated in Appendix 1.1. A mission time of 10 years has been assumed, an average Mean Time To Restoration of 24 hours and a maintenance capability of 100%. The failure rate data used in this calculation are displayed in section 5.1. The resulting PFD AVG (for a variety of proof test intervals) is displayed in Table 27. For SIL2 applications, the PFD AVG value needs to be < 1.00E-02. Table 27: PFD AVG values PFD AVG for T[Proof] 1 year 2 years 5 years 10 years 5.01E E E E-03 As the Temperature Transmitter TT*300-*H with ma output is part of an entire safety function it should only consume a certain percentage of the allowed range. Assuming 25% of this range as a reasonable budget it should be better than or equal to 2.50E-03. The calculated PFD AVG values are within the allowed range for SIL2 according to table 2 of IEC and do fulfill the assumption to not claim more than 25% of the allowed range for a proof test interval of up to 5 years. The resulting PFD AVG graphs generated from the exsilentia tool for a proof test of 1 year are displayed in Figure 2. Figure 2: PFD AVG (t) values Stephan Aschenbrenner Page 19 of 29

20 7 Terms and Definitions DC FIT FMEDA HART HFT Low demand mode PFD AVG SFF SIF SIL Type B element T[Proof] Diagnostic Coverage of dangerous failures (DC = λ dd / (λ dd + λ du)) Failure In Time (1x10-9 failures per hour) Failure Mode Effect and Diagnostic Analysis Highway Addressable Remote Transducer Hardware Fault Tolerance Mode where the frequency of demands for operation made on a safetyrelated system is no greater than one per year and no greater than twice the proof test frequency.pfd AVG Average Probability of Failure on Demand Average Probability of Failure on Demand Safe Failure Fraction summarizes the fraction of failures, which lead to a safe state and the fraction of failures which will be detected by diagnostic measures and lead to a defined safety action. Safety Instrumented Function Safety Integrity Level Complex element (using micro controllers or programmable logic); for details see of IEC Proof Test Interval Stephan Aschenbrenner Page 20 of 29

21 8 Status of the document 8.1 Liability exida prepares reports based on methods advocated in International standards. Failure rates are obtained from a collection of industrial databases. exida accepts no liability whatsoever for the use of these numbers or for the correctness of the standards on which the general calculation methods are based. Due to future potential changes in the standards, best available information and best practices, the current FMEDA results presented in this report may not be fully consistent with results that would be presented for the identical product at some future time. As a leader in the functional safety market place, exida is actively involved in evolving best practices prior to official release of updated standards so that our reports effectively anticipate any known changes. In addition, most changes are anticipated to be incremental in nature and results reported within the previous three year period should be sufficient for current usage without significant question. Most products also tend to undergo incremental changes over time. If an exida FMEDA has not been updated within the last three years and the exact results are critical to the SIL verification you may wish to contact the product vendor to verify the current validity of the results. 8.2 Releases Version History: V4R1: Second HW and SW version added; January 20, 2016 V4R0: Updated to IEC 61508:2010; January 19, 2016 V3R0: TT*200-*H and TSP with TT*200-*H removed; November 6, 2012 V2R0: Product names changed, software and hardware versions updated; November13, 2009 V1, R0: Internal review comments incorporated, software version updated; April 2, 2007 V0, R2: External review comments incorporated; March 19, 2007 V0, R1: Initial version; January 11, 2007 Authors: Stephan Aschenbrenner Review: V4R0: Andreas Stelter (ABB); January 19, 2016 V0, R1: Harald Müller (ABB); March 18, 2007 V0, R2: Rachel Amkreutz (exida); April 2, 2007 Release status: Released to ABB Automation Products GmbH as part of a complete functional safety assessment by TÜV Nord according to IEC Stephan Aschenbrenner Page 21 of 29

22 Appendix 1: Possibilities to reveal dangerous undetected faults during the proof test According to section f) of IEC proof tests shall be undertaken to reveal dangerous faults which are undetected by diagnostic tests. This means that it is necessary to specify how dangerous undetected faults which have been noted during the FMEDA can be detected during proof testing. Appendix 1 shall be considered when writing the safety manual as it contains important safety related information. Appendix 1.1: Possible proof tests to detect dangerous undetected faults A suggested proof test consists of the following steps, as described in Table 28. Table 28 Steps for Proof Test Step Action 1 Bypass the safety PLC or take other appropriate action to avoid a false trip 2 Send a HART command to the transmitter to go to the high alarm current output and verify that the analog current reaches that value. This tests for compliance voltage problems such as a low loop power supply voltage or increased wiring resistance. This also tests for other possible failures. 3 Send a HART command to the transmitter to go to the low alarm current output and verify that the analog current reaches that value. This tests for possible quiescent current related failures 4 Perform a two-point calibration of the transmitter 5 Restore the loop to full operation 6 Remove the bypass from the safety PLC or otherwise restore normal operation This test will detect 74% of possible du failures in the transmitter. Stephan Aschenbrenner Page 22 of 29

23 Appendix 2: Impact of lifetime of critical components on the failure rate According to section of IEC , a useful lifetime, based on experience, should be assumed. Although a constant failure rate is assumed by the probabilistic estimation method (see section 4.2.3) this only applies provided that the useful lifetime 9 of components is not exceeded. Beyond their useful lifetime the result of the probabilistic calculation method is therefore meaningless, as the probability of failure significantly increases with time. The useful lifetime is highly dependent on the component itself and its operating conditions temperature in particular (for example, electrolyte capacitors can be very sensitive). This assumption of a constant failure rate is based on the bathtub curve, which shows the typical behavior for electronic components. Therefore it is obvious that the PFD AVG calculation is only valid for components which have this constant domain and that the validity of the calculation is limited to the useful lifetime of each component. It is assumed that early failures are detected to a huge percentage during the installation period and therefore the assumption of a constant failure rate during the useful lifetime is valid. Table 29 shows which components with reduced useful lifetime are contributing to the dangerous undetected failure rate and therefore to the PFD AVG calculation and what their estimated useful lifetime is. Table 29: Useful lifetime of components contributing to λ du Type Name Useful life at 40 C Capacitor (electrolytic) - Tantalum C210, C206 Appr hours electrolytic, solid electrolyte When plant experience indicates a shorter useful lifetime than indicated in this appendix, the number based on plant experience should be used. 9 Useful lifetime is a reliability engineering term that describes the operational time interval where the failure rate of a device is relatively constant. It is not a term which covers product obsolescence, warranty, or other commercial issues. Stephan Aschenbrenner Page 23 of 29

24 Appendix 3: Using the FMEDA results The Temperature Transmitters TT*300-*H with ma output together with a temperature sensing device become a temperature sensor assembly. Therefore when using the results of this FMEDA in a SIL verification assessment, the failure rates and failure modes of the temperature sensing device must be considered. Appendix 3.1: TT*300-*H with thermocouple The failure mode distributions for thermocouples vary in published literature but there is strong agreement that open circuit or burn-out failure is the dominant failure mode. While some estimates put this failure mode at 99%+, a more conservative failure rate distribution suitable for SIS applications is shown in Table 30 and Table 31 when thermocouples are supplied with the Temperature Transmitters TT*300-*H with ma output. The drift failure mode is primarily due to T/C aging. The Temperature Transmitters TT*300-*H with ma output will detect a thermocouple burn-out failure and drive their output to the specified failure state. Table 30 Typical failure rates for thermocouples (with extension wire) Thermocouple Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 900 FIT FIT Short Circuit (Temperature measurement in error) 50 FIT 1000 FIT Drift (Temperature measurement in error) 50 FIT 1000 FIT Table 31 Typical failure rates for thermocouples (close coupled) Thermocouple Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 95 FIT 1900 FIT Short Circuit (Temperature measurement in error) 4 FIT 80 FIT Drift (Temperature measurement in error) 1 FIT 20 FIT A complete temperature sensor assembly consisting of the Temperature Transmitters TT*300- *H with ma output and a temperature sensing device can be modeled by considering a series subsystem where a failure occurs if there is a failure in either component. For such a system, failure rates are added. Assuming that the Temperature Transmitters TT*300-*H with ma output will go to the pre-defined alarm state on detected failures of the thermocouple, the failure rate contribution for the thermocouple is: Low stress environment (extension wire) λ dd = 900 FIT λ du = 50 FIT + 50 FIT = 100 FIT High stress environment (extension wire) λ dd = FIT λ du = 1000 FIT FIT = 2000 FIT Low stress environment (close coupled) λ dd = 95 FIT λ du = 4 FIT + 1 FIT = 5 FIT High stress environment (close coupled) λ dd = 1900 FIT λ du = 80 FIT + 20 FIT = 100 FIT This results in a failure rate distribution and SFF to: Stephan Aschenbrenner Page 24 of 29

25 Table 32: TT*300-*H with thermocouple (low stress with extension wire) 0 FIT 0 FIT 1213 FIT 134 FIT 90% Table 33: TT*300-*H with thermocouple (low stress close coupled) 0 FIT 0 FIT 408 FIT 39 FIT 91% Table 34: TT*300-*H with thermocouple (high stress with extension wire) 0 FIT 0 FIT FIT 2034 FIT 90% Table 35: TT*300-*H with thermocouple (high stress close coupled) 0 FIT 0 FIT 2213 FIT 134 FIT 94% Appendix 3.2: TT*300-*H with RTD The failure mode distribution for an RTD also depends on the application with the key variables being stress level, RTD wire length and RTD type (2/3 wire or 4 wire). The key stress variables are high vibration and frequent temperature cycling as these are known to cause cracks in the substrate leading to broken lead connection welds. Failure rate distributions are shown in Table 36 to Table 39. The Temperature Transmitters TT*300-*H with ma output will detect open circuit, short circuit and a certain percentage of drift RTD failures and drive their output to the specified failure state. Table 36 Typical failure rates for 4-Wire RTDs (with extension wire) RTD Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 410 FIT 8200 FIT Short Circuit (Temperature measurement in error) 20 FIT 400 FIT Drift (Temperature Measurement in error) 70 FIT FIT 11 Table 37 Typical failure rates for 4-Wire RTDs (close coupled) RTD Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 41,5 FIT 830 FIT Short Circuit (Temperature measurement in error) 2,5 FIT 50 FIT Drift (Temperature Measurement in error) 6 FIT FIT It is assumed that 65 FIT are detectable if the 4-wire RTD is correctly used. 11 It is assumed that 1300 FIT are detectable if the 4-wire RTD is correctly used. 12 It is assumed that 3.5 FIT are detectable if the 4-wire RTD is correctly used. 13 It is assumed that 70 FIT are detectable if the 4-wire RTD is correctly used. Stephan Aschenbrenner Page 25 of 29

26 Table 38 Typical failure rates for 2/3-Wire RTDs (with extension wire) RTD Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 370,5 FIT 7410 FIT Short Circuit (Temperature measurement in error) 9,5 FIT 190 FIT Drift (Temperature Measurement in error) 95 FIT 1900 FIT Table 39 Typical failure rates for 2/3-Wire RTDs (close coupled) RTD Failure Mode Distribution Low Stress High Stress Open Circuit (Burn-out) 37,92 FIT 758,4 FIT Short Circuit (Temperature measurement in error) 1,44 FIT 28,8 FIT Drift (Temperature Measurement in error) 8,64 FIT 172,8 FIT A complete temperature sensor assembly consisting of the Temperature Transmitters TT*300- *H with ma output and a temperature sensing device can be modeled by considering a series subsystem where a failure occurs if there is a failure in either component. For such a system, failure rates are added. Assuming that the Temperature Transmitters TT*300-*H with ma output will go to the pre-defined alarm state on a detected failure of the RTD, the failure rate contribution for the RTD is: 4-wire RTD with extension wire: Low stress environment λ dd = 410 FIT + 20 FIT + 65 FIT = 495 FIT λ du = 5 FIT High stress environment λ dd = 8200 FIT FIT FIT = 9900 FIT λ du = 100 FIT 4-wire RTD close coupled: Low stress environment λ dd = 41.5 FIT FIT FIT = 47.5 FIT λ du = 2.5 FIT High stress environment λ dd = 830 FIT + 50 FIT + 70 FIT = 950 FIT λ du = 50 FIT 2/3-wire RTD with extension wire: Low stress environment λ dd = FIT FIT = 380 FIT λ du = 95 FIT High stress environment λ dd = 7410 FIT FIT = 7600 FIT λ du = 1900 FIT 2/3-wire RTD close coupled: Low stress environment λ dd = FIT FIT = FIT λ du = 8.64 FIT High stress environment λ dd = FIT FIT = FIT λ du = FIT This results in a failure rate distribution and SFF to: Stephan Aschenbrenner Page 26 of 29

27 Table 40: TT*300-*H with 4-wire RTD (low stress with extension wire) 0 FIT 0 FIT 808 FIT 39 FIT 95% Table 41: TT*300-*H with 4-wire RTD (low stress close coupled) 0 FIT 0 FIT 360 FIT 36 FIT 90% Table 42: TT*300-*H with 4-wire RTD (high stress with extension wire) 0 FIT 0 FIT FIT 134 FIT 98% Table 43: TT*300-*H with 4-wire RTD (high stress close coupled) 0 FIT 0 FIT 1263 FIT 84 FIT 93% Table 44: TT*300-*H with 2/3-wire RTD (low stress with extension wire) 0 FIT 0 FIT 693 FIT 129 FIT 84% Table 45: TT*300-*H with 2/3-wire RTD (low stress close coupled) 0 FIT 0 FIT 352 FIT 42 FIT 89% Table 46: TT*300-*H with 2/3-wire RTD (high stress with extension wire) 0 FIT 0 FIT 7913 FIT 1934 FIT 80% Table 47: TT*300-*H with 2/3-wire RTD (high stress close coupled) 0 FIT 0 FIT 1100 FIT 207 FIT 84% Stephan Aschenbrenner Page 27 of 29

28 Appendix 3.3: TT*300-*H in redundant mode with drift monitoring This appendix shows the failure rates when the Temperature Transmitters TT*300-*H with ma output are used in redundant mode with two temperature sensing devices connected to it. To obtain the overall failure rates of the sensor assembly, use the failure rates of the Temperature Transmitters TT*300-*H with ma output for redundant mode and add failure rates of both temperature sensing devices. The temperature sensing device failure rates should be adjusted to reflect the additional coverage (95%) on the normally undetected failures provided by the drift alarm. Table 48: TT*300-*H with two thermocouples (low stress with extension wire) 0 FIT 0 FIT 2309 FIT 41 FIT 98% Table 49: TT*300-*H with two thermocouples (low stress close coupled) 0 FIT 0 FIT 518 FIT 32 FIT 94% Table 50: TT*300-*H with two thermocouples (high stress with extension wire) 0 FIT 0 FIT FIT 231 FIT 99% Table 51: TT*300-*H with two thermocouples (high stress close coupled) 0 FIT 0 FIT 4309 FIT 41 FIT 99% Table 52: TT*300-*H with two 2/3-wire RTDs (low stress with extension wire) 0 FIT 0 FIT 1259 FIT 41 FIT 96% Table 53: TT*300-*H with two 2/3-wire RTDs (low stress close coupled) 0 FIT 0 FIT 414 FIT 32 FIT 92% Table 54: TT*300-*H with two 2/3-wire RTDs (high stress with extension wire) 0 FIT 0 FIT FIT 221 FIT 98% Stephan Aschenbrenner Page 28 of 29

29 Table 55: TT*300-*H with two 2/3-wire RTDs (high stress close coupled) 0 FIT 0 FIT 2221 FIT 48 FIT 97% Table 56: TT*300-*H with thermocouple and 2/3-wire RTD (low stress with extension wire) 0 FIT 0 FIT 1784 FIT 41 FIT 97% Table 57: TT*300-*H with thermocouple and 2/3-wire RTD (low stress close coupled) 0 FIT 0 FIT 466 FIT 37 FIT 92% Table 58: TT*300-*H with thermocouple and 2/3-wire RTD (high stress with extension wire) 0 FIT 0 FIT FIT 226 FIT 99% Table 59: TT*300-*H with thermocouple and 2/3-wire RTD (high stress close coupled) 0 FIT 0 FIT 3265 FIT 139 FIT 95% Stephan Aschenbrenner Page 29 of 29