June 2016 Machinery PIP REEE002 Reliability Indicators for Rotating Machinery
PURPOSE AND USE OF PROCESS INDUSTRY PRACTICES In an effort to minimize the cost of process industry facilities, this Practice has been prepared from the technical requirements in the existing standards of major industrial users, contractors, or standards organizations. By harmonizing these technical requirements into a single set of Practices, administrative, application, and engineering costs to both the purchaser and the manufacturer should be reduced. While this Practice is expected to incorporate the majority of requirements of most users, individual applications may involve requirements that will be appended to and take precedence over this Practice. Determinations concerning fitness for purpose and particular matters or application of the Practice to particular project or engineering situations should not be made solely on information contained in these materials. The use of trade names from time to time should not be viewed as an expression of preference but rather recognized as normal usage in the trade. Other brands having the same specifications are equally correct and may be substituted for those named. All Practices or guidelines are intended to be consistent with applicable laws and regulations including OSHA requirements. To the extent these Practices or guidelines should conflict with OSHA or other applicable laws or regulations, such laws or regulations must be followed. Consult an appropriate professional before applying or acting on any material contained in or suggested by the Practice. This Practice is subject to revision at any time. Process Industry Practices (PIP), Construction Industry Institute, The University of Texas at Austin, 3925 West Braker Lane (R4500), Austin, Texas 78759. PIP Member Companies and Subscribers may copy this Practice for their internal use. Changes or modifications of any kind are not permitted within any PIP Practice without the express written authorization of PIP. Authorized Users may attach addenda or overlays to clearly indicate modifications or exceptions to specific sections of PIP Practices. Authorized Users may provide their clients, suppliers and contractors with copies of the Practice solely for Authorized Users purposes. These purposes include but are not limited to the procurement process (e.g., as attachments to requests for quotation/ purchase orders or requests for proposals/contracts) and preparation and issue of design engineering deliverables for use on a specific project by Authorized User s client. PIP s copyright notices must be clearly indicated and unequivocally incorporated in documents where an Authorized User desires to provide any third party with copies of the Practice. PRINTING HISTORY March 1998 Issued June 2016 Complete Revision December 2006 Withdrawn May 2010 Reissued as Complete Revision Not printed with State funds
June 2016 Machinery PIP REEE002 Reliability Indicators for Rotating Machinery Table of Contents 1. Introduction... 2 1.1 Purpose... 2 1.2 Scope... 2 2. References... 2 2.1 Process Industry Practices... 2 2.2 Industry Codes and Standards... 2 3. Definitions... 2 4. Reliability Indicators... 4 4.1 General... 4 4.2 Terminology... 4 4.3 Equations... 6 5. Application of Reliability Indicators... Error! Bookmark not defined. 5.1 Data Acquisition and Sorting... 8 5.2 Basic Maintenance Uses... 8 5.3 Advanced Maintenance Uses... 9 Data Forms REEE002-F - Reliability Benchmarking Data Form Process Industry Practices Page 1 of 10
1. Introduction 1.1 Purpose This Practice provides standardized reliability terminology and a basis for measurement and data gathering for benchmarking the reliability of rotating machinery. 1.2 Scope This Practice describes reliability terms and definitions that can be applied to economic performance data generation. This Practice describes reliability data gathering techniques. This Practice provides basic steps to implementing reliability indicators in process industry facilities. This Practice describes levels of data gathering and analysis that may be used to provide the appropriate amount of reliability information. 2. References Applicable parts of the following Practice and industry codes and standards shall be considered an integral part of this Practice. The edition in effect on the date of contract award shall be used, except as otherwise noted. Short titles are used herein where appropriate. 2.1 Process Industry Practices (PIP) PIP REIE686/API RP686 - Recommended Practices for Machinery Installation and Installation Design 2.2 Industry Codes and Standards 3. Definitions American Petroleum Institute (API) API689 - Collection and Exchange of Reliability and Maintenance Data for Equipment active machinery: A permanently installed machinery item (e.g., a pump that is in active service and available for operation), whether continuous, standby, or intermittent. Active machinery includes machinery in all types of services (e.g., process, lubricating oil, transfer, utility, and effluent services). Active machinery does not include machinery that is associated with maintenance or office facilities (e.g., air conditioning or portable maintenance pumps). Active machinery does not include machinery located in facilities or units that have been indefinitely shutdown, preserved or abandoned. critical service: Categorization of process machinery based on the operating service and includes unspared or partially spared machinery. Downtime of this machinery causes production losses and/or violates regulatory agency requirements. economic performance: Various measures of the costs associated with the operation of process equipment, typically reported in monetary based measures Process Industry Practices Page 2 of 10
equipment size: Size division for machinery based on the driver power. Typical power division are as follows: a. 4 kw (5 HP) and less b. Greater than 4 kw 40 kw (5 HP 50HP) c. Greater than 40 kw 185 kw (50 HP 250 HP) d. Greater than 185 kw 750 kw (250 HP - 1000 HP) e. Greater than 750 kw (1000 HP). installed power (Kw, HP): Nameplate power rating of the machinery, typically the driver of the machinery train. Typically, summaries of kw (HP) for a unit, plant or criticality are totals of the pertinent machinery driver ratings. life cycle cost: Total cost of a rotating machinery installation through its lifetime, including total cost of purchase, installation, operation, maintenance, and disposal of the equipment machinery train: Driven equipment, driver, and auxiliaries (e.g., sealing systems, integral lubrication components, speed changing gears and couplings). A machinery train does not include process control system components or switch gear. machinery class: Grouping of equipment with similar function and general characteristics (e.g., pumps, turbines, compressors, etc.) machinery unit: Individual machinery item (e.g., pump, motor, turbine, compressor, gear box, etc.). Component machines of a machinery train. Machinery boundaries are defined by API689. material costs: Purchase price for equipment, accessories, and dedicated spare parts. Does not include labor costs. non-critical service: Categorization of process machinery based on the operating service and may include either spared or unspared equipment. Downtime of this machinery does not cause direct production losses, or the losses are minor and acceptable and do not violate regulatory agency requirements. operating performance: Various measures of the reliability of operation for process equipment, typically reported in time based measures predictive maintenance techniques: Methods used to help determine the condition of in-service equipment in order to predict when maintenance should be performed preventive maintenance: Maintenance activities performed at predetermined intervals or according to prescribed criteria that are intended to reduce the probability of failure or unacceptable degradation of machinery function. Preventive maintenance activities are considered either normal operating tasks (i.e., activities accomplished during normal machinery operation) or preemptive repairs (i.e., activities requiring machinery shutdown for safe completion). Reliability Centered Maintenance (RCM): A facility maintenance process that emphasizes the use of predictive maintenance techniques in addition to traditional preventive maintenance measures. Process Industry Practices Page 3 of 10
rework: Repairs performed on equipment that failed shortly after a repair because of defective parts or inadequate repair root cause: Primary initial causal action, condition, or circumstance for an event. Typically, the term root cause is used in failure investigation. total installed cost: The cost of engineering, procurement, construction, installation, inspection, start-up and documentation for a machinery train 4. Reliability Indicators 4.1 General 4.1.1 Reliability, also called performance, is the ability of a machinery unit or train to perform a required function under given conditions for a given time interval. 4.1.2 Measurement of machinery operating and economic performance is effectively defined by the cost and time of operation. Reliability indicators are measures of operating and economic performance that quantify the reliability of rotating machinery. Typically, reliability indicators provide time and cost of machinery operation in various formats. 4.1.3 Effective comparability of reliability indicators depends on consistent, logical, and easily understood groupings of reliability data. 4.2 Terminology 4.2.1 Mean Time Between Repair (MTBR) 4.2.1.1 MTBR is the average operating calendar time between required repairs for a particular piece of machinery, type of machinery, class of machinery, or operating facility. 4.2.1.2 MTBR is not mean time between shutdowns caused by failures, planned maintenance, or any other categorization of shutdowns. 4.2.1.3 MTBR calculations include repairs caused by any of the following events: a. Failures b. Planned maintenance c. Any other categorization of repair events 4.2.1.4 MTBR is the most common measure of operating reliability because it is the most comparable and easily measured reliability statistic. 4.2.1.5 The longer the average time of operation between repairs, the greater the reliability of the machinery. 4.2.1.6 Data should be collected and submitted on broad categories of machinery types. The data should be arranged into specific groupings of machinery types divisions and then in association with the mated equipment (e.g., pumps, centrifugal pumps, single stage pumps, Process Industry Practices Page 4 of 10
4.2.2 Repairs API/ANSI/other, multiple seals / single seals / sealless, electric motor driven / turbine driven / other drivers, etc.). 4.2.2.1 Repairs may include maintenance activities (i.e., routine, preventive, predictive, failure correction, or rebuild) that typically require machinery shutdown and isolation from energy sources for safe completion. These maintenance activities include the following work: a. Restore machinery to its normal function or performance b. Replace failed, failing, worn or questionable parts c. Replace seals, including those that exceed fugitive emission regulations d. Disassemble machinery to remove deposits, debris, or plugging e. Discovery work performed during disassembly or partial disassembly of machinery for inspection f. Maintenance activities to correct defects identified by condition monitoring or surveillance programs g. Equipment modifications and engineering design improvement project appended to a repair h. Replacement of all or a portion of a coupling that was damaged by incorrect assembly, lubrication problems, misalignment, or failed flex elements 4.2.2.2 In order to assure that repairs are properly identified, the following checks should be made for each occurrence: a. Root cause of the failure was determined and recorded b. The specific part of the machinery unit that failed was accurately recorded 4.2.2.3 Flawed repairs discovered and corrected before the machinery is returned to operation s control should not be counted as rework, but rather as a continuation of the first repair. 4.2.2.4 Data gathering for repairs to new machinery installations should start after initial mechanical sign off/completion (i.e., equipment is turned over to operations/maintenance). 4.2.2.5 Maintenance and other activities should not be considered repairs if the activities are in accordance with one of the following criteria: a. Activity does not require machinery shutdown to safely complete b. Activity is normal, recurring, and necessary to ensure nominal run life of the machinery (e.g., on-line performance of oil and filter changes, topping up oil levels, checking oil mist systems, minor adjustments and cleanings) Process Industry Practices Page 5 of 10
4.2.3 Availability c. Activity is required because of a change in the machine design, also called design repairs Availability is a measure of the percentage of the potential operating time that machinery is capable of performing its required function, typically expressed as a percentage of an operating calendar year. Comment: For further information about availability, see API689. 4.2.3.1 Criteria for measuring availability can vary depending on the type of assets being measured and how a specific company internally measures its performance. The criteria provided below are a function set of availability measurements but do not represent any sort of industry standard. 1. Machinery is counted as available for operation when: a. It is operating and meeting process requirements. b. It is considered to be capable of operation 2. Machinery is counted as unavailable when it cannot meet current process requirements. Note: Spare machinery needing repairs that are deferred due to business decisions are considered unavailable. 3. Time counted as unavailable for machinery does not include time for planned work performed during shutdowns or turnarounds if the shutdown is for reasons unrelated to the machinery. 4.2.4 Repair Cost per Installed Horsepower per Year 4.3 Equations 4.2.4.1 Repair cost per installed horsepower per year is the annual average repair cost for a particular machine, type of machine, size of machine, or similar machines in an operating facility, divided by the total nameplate horsepower for the machinery. 4.2.4.2 Repair cost data is typically grouped into equipment classes based on driver power (refer to equipment class and size definition) 4.2.4.3 Repair cost per installed power should be sorted by machinery type as the data and capability become available. Comment: Reliability data and the calculated reliability indicators should be tabulated and reported by machinery type to maximize comparability. 4.3.1 MTBR 4.3.1.1 MTBR should be calculated using the following equation: Number of Active Machinery Units MTBR= Number of Machinery Repairs for the Period Process Industry Practices Page 6 of 10
4.3.1.2 MTBR should be expressed as a unit of time (months, years, etc.). 4.3.1.3 MTBR is typically calculated on a fixed interval (appropriate for the equipment being analyzed, monthly is a common interval) for individual machinery units (drivers separate from driven), by class and in groupings by type, size, service, process unit, etc. 4.3.1.4 MTBR is typically calculated with a period of data that provides a statistically significant sample size resulting in the selected unit of time (months/years, etc., See Section 4.3.1.2) between repairs. 4.3.1.5 Reported data should be marked to properly distinguish the equipment arrangement. 4.3.2 Availability Availability should be calculated using the following equation: Number of Days Machinery Available x 100% Availability ( ) = Number of Unit Operating Days for the Year 4.3.3 Average Cost per Repair 4.3.3.1 Average cost per repair, also called average repair cost, is the total cost of all repairs divided by the total number of repairs for any of the following machinery groups: a. Particular machine b. Class and type of machine c. Size of machines d. Similar machines in an operating facility 4.3.3.2 Average cost per repair should be calculated using the following equation: Total Repair Costs Average Cost per Repair= Total Number of Repairs 4.3.4 Annual Cost per Installed Asset Type Note: The period needs to be long enough to contain a statistically significant sample size 4.3.4.1 Annual maintenance cost per installed asset type should be calculated using the following equation: Annual Maintenance Cost per Installed Asset Type= Annual Maintenance Costs for All Assets of a Subtype Number of Assets of the Subtype Process Industry Practices Page 7 of 10
4.3.4.2 As an example of the use of annual maintenance cost per installed asset type, if the maintenance costs for just centrifugal pumps in a facility can be separately determined, the annual maintenance cost per installed centrifugal pump can provide a cost to maintain a centrifugal pump for trending, future budgeting, or cost analysis of capital projects. 4.3.5 Annual Repair Cost per Installed HP Annual repair cost per installed horsepower should be calculated using the following equation: Annual Repair Cost Annual Repair Cost per Installed HP = Nameplate Horsepower 5. Application of Reliability Indicators 5.1 Data Acquisition and Sorting 5.1.1 The first step in developing a reliability monitoring program for an operating facility is to determine what items need to be monitored. 5.1.2 Monitoring can range from all rotating equipment to specific equipment types (e.g., canned motor pumps). 5.1.3 Monitoring priorities can be determined by asset criticality, class or type. The level of individual asset detail in the Computerized Maintenance Management System (CMMS) will facilitate filtering. For example, it may be possible to capture all pump work orders if the assets all have a P in the number but it may not be possible to isolate just mag-drive pumps unless there is a second field in the CMMS that identifies such pumps. If an equipment listing that identifies more detail (e.g., seal or pump type) is available from a seal survey or maintenance listing, the filtering can be performed using spreadsheet or database tools. 5.1.4 Data for basic reliability calculations can usually be obtained from the facility s CMMS system. Many CMMS systems can export work order data from the database into programs where the data can be manually manipulated. 5.1.5 Work orders can be evaluated against the criteria for a repair based on one of several possible work order entries. The work order closing comments should contain sufficient data to make an initial determination and/or, in facilities where codes are used as part of the work order package, the repair or action taken codes can be used. 5.1.6 After the work orders are evaluated and a determination is made on what meets the repair qualification, the number of repairs during the time period can be determined. This combined with an asset count from the CMMS or plant accounting system can be applied to the reliability formulas. 5.2 Basic Maintenance Uses 5.2.1 MTBR 5.2.1.1 Basic data acquisition and filtering can generate sufficient information to calculate MTBR. It is important to recognize what the MTBR is based on, especially when it is compared to MTBRs from other Process Industry Practices Page 8 of 10
operating facilities or industry data. MTBRs can significantly differ because of non-uniform definitions of a failure and by the quality of the data and analysis. 5.2.1.2 Work orders with incomplete or inaccurate close out comments, failure codes, or action taken codes degrade the ability to obtain accurate data. Also, work orders written for a machinery asset that are not for machinery failures (e.g., the discharge pressure gauge) can skew the failure rates. 5.2.1.3 If the data acquisition and analysis is consistent, it provides trend information for a specific facility even if the data quality is not sufficient to be used for comparison to an outside facility or standard. 5.2.2 Average Cost per Repair 5.2.2.1 Average cost per repair can be generated from the same data as the MTBR. If the work order costs are exported from the CMMS with the failure data, the costs can be summed and used in the average cost per repair formula. 5.2.2.2 Average cost per repair can be used as follows: a. Provide information on the trend of the costs for maintenance. b. If combined with the MTBR, to estimate the impact of adding equipment to the facilities maintenance budget. c. If the machinery is categorized into specific types or services, to identify areas for improvements in operations or maintenance. 5.3 Advanced Maintenance Uses 5.3.1 Measurements Subsets 5.3.1.1 Reliability indicators can be developed for and applied to progressively refined subsets of equipment based on the information available and the information s format. An example of a subset progression is as follows: a. Rotating Equipment a.1 Pumps a.1.1 Centrifugal Pumps a.1.1.1 Sealed pumps (1) Dual Seals (2) Gas Seals 5.3.1.2 Analysis of reliability indicators can permit comparisons of technology and help with determination of the most effective systems for new installations. It will also identify where additional resources are justified to reduce costs and/or downtime. Examples of this kind of use of reliability indicators are as follows: a. The MTBR of dual seals in a facility may be shorter than sealless pumps. However, the cost per repair of a sealless pump may be Process Industry Practices Page 9 of 10
greater, thus having a higher maintenance cost. If a unit cannot be shut down for extended periods, the reduction of loss of production from using a sealless pump may justify the higher initial and maintenance cost. Some industries have experienced lower maintenance costs with sealless pumps. b. The MTBR of single seals at a facility may be greater than dual seals indicating that additional training or maintenance tasks may be needed and justifiable for dual seal systems. 5.3.1.3 The best source of reliability measurement information is a field in the CMMS that indicates type of pump, seal, or whatever subset is desired. If this source of information is not available, a look-up function can be used between the CMMS machinery failure data and an equipment list containing the desired subsets. 5.3.1.4 Equipment surveys can provide a good start for obtaining reliability measurement information. For example, a seal survey in spreadsheet or database format can be compared to the CMMS machinery failure history, and any field in the seal survey can be paired with the associated asset. This provides failure by seal type which can be used with seal type counts to generate MTBR and cost per repair by seal subtype. 5.3.2 Availability 5.3.2.1 Availability data may be obtained from an asset utilization system, corporate management system, or operating logs. The data may also be obtained manually from shift logs or turnover logs. 5.3.2.2 Availability data can be used to identify assets affecting product shipment and to justify expenditures for modifications or upgrades. Analysis of availability is most effective if combined with an estimated cost of downtime for the assets involved. 6. Reliability Benchmarking Data Form A Reliability Benchmarking Data form, PIP REEE002-F, is provided with this Practice. The form may be used to record reliability metrics for an operating facility. Examples of completed benchmarking forms are also provided as follows: a. Example 1 has been prepared based on reliability data accumulated by machinery unit. b. Example 2 has been prepared based on reliability data accumulated by machinery train. Process Industry Practices Page 10 of 10
ASSOC. PIP REEE002 DATA SHEET REEE002-F LOCATION: / PERIOD MEASURED: / -- / COMPANY /' FACILITY RELIABILITY BENCHMARKING DATA PAGE 1 OF X JUNE 2016 1 2 3 4 5 6 7 8 9 10 11 12 Service Mach. Driver HP # of Availability # of Total Total MTBR $/Repair $/HP/YR Unit Type Class Type Type Class Items (where measured) Repairs HP (M) Repair $ (M) Calculated Values Plant DATA ACCUMULATED Machinery Unit Item Machinery Train SPECIAL EXPLANATION: Submitted by:
ASSOC. PIP REEE002 DATA SHEET REEE002-F PAGE 1 OF 1 RELIABILITY BENCHMARKING DATA JUNE 2016 LOCATION: XYZ Chemical / Houston PERIOD MEASURED: 1 / 96 -- 12 / 96 COMPANY /' FACILITY 1 2 3 4 5 6 Service Mach. Driver HP # of Availability # of Class Type Type Class Items (where measured) Repairs Non-critical Centrifugal Pump (ASME) Elect. Motor/ 620 800 <50 HP Centrifugal Pump (ASME) Elect. Motor/ 48 67 >50 HP <100 HP Centrifugal Pump (API) Stm. Turb. 32 58 >100 <1000 HP 7 8 9 10 11 12 Total Total MTBR $/Repair $/HP/YR Unit Type HP (M) Repair $ (M) Calculated Values Plant 15.25 312 2/Houston 37.50 52 13.10 102 Steam Turbine ----- 32 70 >200 <1000 HP 13.10 89 Critical Centrifugal Compressor Elect. Motor/ 12 92% 29 >1000 HP Centrifugal Pump (API) Elect. Motor/ 14 18 >200 <1000 HP Recip Compressor Gas Engine/ 3 94% 17 >1000 HP 50.00 123 4.00 46 4.90 38 EXAMPLE #1 Formulas: Column 9 = Col. 4 Col. 6 Column 10 = Col. 8 Col. 6 Column 11 = Col. 8 Col. 7 DATA ACCUMULATED Machinery Unit Item Machinery Train SPECIAL EXPLANATION Submitted by: John Doe, XYZ Chemical
ASSOC. PIP REEE002 DATA SHEET REEE002-F PAGE 1 OF 1 RELIABILITY BENCHMARKING DATA JUNE 2016 LOCATION: JKL Refinery / Channel PERIOD MEASURED: 1 / 96 -- 12 / 96 COMPANY /' FACILITY 1 2 3 4 5 6 Service Mach. Driver HP # of Availability # of Class Type Type Class Items (where measured) Repairs Critical Cont. Comp > 1000 HP Elect. Motor/ 14 97% 18 Steam Turbine 12 98% 11 Centrifugal Pump (ASME) Gas Turbine 1 96% 8 7 8 9 10 11 12 Total Total MTBR $/Repair $/HP/YR Unit Type HP (M) Repair $ (M) Calculated Values Plant 27.0 150.0 1/Channel 30.0 200.0 15.0 3.0 Cent. Pumps (ASI) Elect. Motor 57 98% 53 >200 HP < 1000 HP Steam Turbine 15 95% 22 12.0 30.0 4.2 31.0 Recip. Comps. > 1000 HP Elect. Motor 10 93% 23 Gas Engine 3 91% 14 15.0 45.0 8.0 21.0 Non-Critical Cen. Pumps (ASME) Elect. Motor 2000 3000 >50 < 200 HP Steam Turbine 20 39 36.0 100.0 3.7 27.0 P. D. Pumps < 50 HP Elect. Motor 17 23 0.3 19.0 Fans Other EXAMPLE #2 Formulas: Column 9 = Col. 4 Col. 6 Column 10 = Col. 8 Col. 6 Column 11 = Col. 8 Col. 7 DATA ACCUMULATED Machinery Unit Item Machinery Train SPECIAL EXPLANATION: Submitted by: John Dokes, JKL Refining