Asset Health CIM - UCAIug



Common Information Model (CIM) for Asset ManagementJune 2014Table of Contents TOC \o "1-3" \h \z \u 1Introduction PAGEREF _Toc263602096 \h 32Asset Management – ISO 55000/1/2 PAGEREF _Toc263602097 \h 43Use Cases PAGEREF _Toc263602098 \h 63.1Aggregate Asset Data PAGEREF _Toc263602099 \h 63.2Asset Registry PAGEREF _Toc263602100 \h 73.3Support Life Cycle Management PAGEREF _Toc263602101 \h 73.4Asset Planning and Procurement PAGEREF _Toc263602102 \h 73.5Setting Asset Operational Limits PAGEREF _Toc263602103 \h 83.6Rating Assets using Health Index PAGEREF _Toc263602104 \h 83.7Attributing Risk to Assets and Asset Groups PAGEREF _Toc263602105 \h 83.8Asset Health Event PAGEREF _Toc263602106 \h 84Interface Reference Model (IRM) Business Function PAGEREF _Toc263602107 \h 95Requirements and Discussion Topics PAGEREF _Toc263602108 \h 115.1Asset Modeling Considerations PAGEREF _Toc263602109 \h 115.2Missing Assets and Materials PAGEREF _Toc263602110 \h 125.3Asset Group Considerations: AssetInfo and AssetModel PAGEREF _Toc263602111 \h 135.4Relocation of Assets PAGEREF _Toc263602112 \h 135.5Asset and Network Model Relationship PAGEREF _Toc263602113 \h 135.6Procedure and Dataset PAGEREF _Toc263602114 \h 145.7Test Activity PAGEREF _Toc263602115 \h 145.8Lifecycle-dependent Testing Events PAGEREF _Toc263602116 \h 155.9Measurements Not Made at the Asset PAGEREF _Toc263602117 \h 165.10Retrieval of Measurements as a Time Series Data PAGEREF _Toc263602118 \h 165.11Measurement Metadata PAGEREF _Toc263602119 \h 185.12Waveform Measurements PAGEREF _Toc263602120 \h 185.13Health Index and Risk PAGEREF _Toc263602121 \h 19IntroductionThe purpose of this effort is to enable analytics-driven, condition based asset management concepts by developing the use cases and requirements for a CIM standards framework to support best-of-breed practices. While, at the present time, condition-based asset management (CBAM) practices are often unique to a given utility, they have a number of general characteristics in common. They typically:Aggregate data from multiple sources.The data sources pertain to different aspects of an asset – electrical and non-electrical.The data is indicative of the current and future asset condition and performance.Multiple data sources may exist for any particular aspect of the asset – e.g., dissolved gases may be measured by online field monitors as well as lab analysis of oil samples. The multiple data sources have different time-based and data resolution characteristics that arise from different sampling and data storage approaches.In many cases, the history of data (pattern or trend over time) is the needed input, not just the current value.Generate asset health and risk assessments based on the data. Asset health assessment is characterized by a score, commonly known as health score or health index.Many factors that could contribute to the health of the asset are assessed and scored independently. These scores are combined to produce an overall health score for the asset.Assets in the same “family” (collection of similar assets) may be compared and ranked. Risk assessment is done by considering both asset importance (cost, criticality, replacement availability, etc.) and asset health score.Asset health and risk assessment are approached using a multitude of different methodologies and there could be multiple metrics used for different purposes.Generate asset health events when assessment indicates significant change in health conditions has occurred.Recognition that results of assessment represent a situation that merits annunciation.Asset health events could include annunciations like:User notifications and alarms.Work order requests – e.g., for additional testing or preventive maintenance.Triggers and operational limits for a desired health behavior.Make data available to other systems/tools as needed/requested. The work of the Asset Health Focus Community aims to clarify data exchange requirements and explore Common Information Model (CIM) model usage to support the utility CBAM processes described above. There are two primary points in the above process where data is shared among applications, devices or systems:In Step 1 (Aggregate data from multiple sources), where “raw” or unprocessed data is collected from multiple field devices, systems, databases, spreadsheets or applications to be used as input to asset health and risk assessments In Steps 3 and 4 (Generate asset health events / enable data consumption), where the results or outputs of asset health and risk assessments are shared with and subsequently acted upon by other applicationsThe information models and interfaces to enable these two data exchanges are the focus of CIM for asset health/management.Asset Management – ISO 55000/1/2The capabilities enabled by the current CIM effort is a subset of strategic asset management discipline described in the ISO 55000/1/2 standards, which were published in January 2014. Of particular relevance are the asset management elements in the lower half of the following figure. In this section, we also quote specific clauses from ISO 55001/2 that apply to the CIM asset management effort.Figure SEQ Figure \* ARABIC 1—Illustration from ISO 55002, shows relationship between the key elements of an asset management system, together with related clauses in ISO 55001.Actions to address risks and opportunities for the asset management system – Clause 6.1:The organization should determine the actions that are necessary for addressing risks when planning for its asset management system.When addressing risks in the asset management system, the organization should determine the risk assessment criteria (e.g. likelihood and consequence, and risk attitude) with the asset management decision making for its asset management system.Asset Management Objectives – Clause 6.2.1Asset management objectives should be specific, measurable, achievable, realistic and time-bound (i.e. “SMART” objectives).They can be both quantitative measurements (e.g. mean time between failure) and qualitative measurements (e.g. customer satisfaction).Typical issues that are addressed by objectives for asset portfolios:Return on investment (or return on capital employed, or return on asset).Typical issues that are addressed by objectives for asset systems:Asset system availability;Asset system performance (e.g. uptime, efficiency);Unit cost of product or service.Typical issues that are addressed by objectives for assets:Reliability (mean time/distance between failures);Asset condition, performance, or health score;Life cycle costs;Life expectancy;Asset energy performance.Planning to achieve asset management objectives – Clause 6.2.2:A risk ranking process can determine which assets have a significant potential to impact on the achievement of the asset management objectives, i.e. which are the critical assets.Standard risk management practices tend to overlook events that are very low probability/high consequence as not worth considering in detail. It is important that there be an additional dimension in the risk analysis to include the capability of the asset management system to monitor and continually assess the probability of these rare, but potentially catastrophic events.Life cycle costs, which may include capital expenditure, financing and operational costs, should be considered in the decision-making process.Performance evaluation: Monitoring, measurement, analysis and evaluation – Clause 9.1The organization should develop processes to provide for the systematic measurement, monitoring, analysis and evaluation of the organization’s assets, asset management system and asset management activity on a regular basis. In the development of these processes (and any associated procedures) the following should be taken into account:Setting of performance metrics and associated indicators, e.g. condition or capacity indicators;Confirmation of compliance with the requirements;Examination of historical evidence;The use of documented information to facilitate subsequent corrective actions and decision making.Improvement – Clause 10:Nonconformity and corrective action.Corrective actions are actions taken to address the root cause(s) of identified non-conformances, or incidents, in order to manage their consequences, and to prevent or reduce the likelihood of recurrence.Preventive action.May include predictive actions.Actions taken to address the root cause(s) of potential failures or incidents, as a proactive measures, before such incidents occur.Continuous improvement.Opportunities for improvement should be identified, assessed and implemented across the organization as appropriate, through a combination of monitoring and corrective actions for the assets, asset management, or asset management system.Use CasesIn this section, we capture the use case narratives that apply to the CIM asset management effort.Aggregate Asset DataThis use case covers obtaining data pertinent to asset health from various sources.Obtain nameplate and catalogue information for the asset.Obtain off-line data.Field test results.Lab test results.Inspection results.Obtain on-line data.On-line monitors.Operational data.Obtain activity history.Lifecycle stages and dates.Procedures performed and dates.Corrective actions taken and dates.Obtain data (metadata) indicative of the conditions of original data measurement – i.e., quality.Asset RegistryAn asset registry that provides organization for assets, asset systems, asset models, and asset families is necessary for asset management and analytics.Incorporate every asset that is subject of asset management program, including non-conducting assets such as poles and towers.Capture the components of the assets / material related to assets – such as bushings and oil – that are the subject of asset management programs. For instance, regulations require tracking PCB lifecycle.Able to group the assets in many different ways to enable aggregate analysis in the asset management system – e.g., based on operational asset systems they belong to, particular asset model, and particular asset functional class/family.Able to associate asset-pertinent data that isn’t measured at the asset – e.g., flow for a breaker.Support Life Cycle ManagementAsset management involves a structured approach that can be applied over different life cycle stages of the asset.There are many significant stages/events in an asset’s lifecycle.Planning and procurementInstallationOperationRepair Refurbishment ReinstallationEnd-of-LifeRelocationWhen an asset moves through its lifecycle, including events that involve moving the asset to a different location, the data and history pertaining to that asset should still be available for analytics. This includes sufficient operational history.Support application of asset management concepts pertinent to the current life cycle. For instance, the asset planning and procurement use case described below. Asset Planning and ProcurementDuring the Asset Planning stage, asset maintenance history is typically taken into account (e.g., for the particular model), and this is something that would be expected from an asset health analytics system – i.e., reliability study that is used for asset planning.When a new asset needs to be purchased, access historical performance data that helps decide on the asset to procure.Identify asset models for purchase.Obtain performance history of models being considered.For new asset models with no history, identify and obtain performance history of similar models.Obtain pricing information for asset models.Setting Asset Usage LimitsEnable usage of assets in a way that helps achieve asset management objectives, as measured by risk, aging factor, etc.Identify target values for risk, aging factor, etc. in due course of asset operation.Determine asset usage limits for the objectives.Provide the limits to appropriate systems.Rating Assets using Health IndexAsset health assessment is characterized by a score, commonly known as health score or health index.Many factors that could contribute to the health of the asset are assessed and scored independently. These scores are combined to produce an overall health score for the asset.Assets in the same “family” (collection of similar assets) may be compared and ranked. Attributing Risk to Assets and Asset GroupsRisk-based asset management.Risk assessment is done by considering both asset importance (cost, criticality, replacement availability, etc.) and asset health index.Calculate risk scores for individual assets and asset aggregates.Asset condition, geography, operating conditions, environmental factors, etc. used to calculate risk.The risk scores can be used to identify potential projects and facilitate decision making.Asset Health EventGenerate asset health events when assessment indicates significant change in asset conditions has occurred.Generate event notification upon detected and predicted health events.Asset health events could include annunciations like:User notifications and alarms.Work order requests – e.g., for additional testing or preventive maintenance.Triggers and usage limits for a desired health behavior.The event notification may include:Level indicator – e.g., informational, warning, action required.Metrics such as health index, risk, etc.Recommended action.Time to action. The system can provide the asset health event information to other systems – e.g., the WAMS to trigger a preventive or corrective work order.Interface Reference Model (IRM) Business FunctionThe IRM business sub-function of Records & Asset Management (AM) / Asset Investment Planning (AIP) seems to fit the capabilities that are being enabled by the current work. The following changes to AM-AIP seem appropriate: (1) Change of the business sub-function name to Asset Decision Support (ADS), (2) Change the “Decision support” abstract component to “Decision support systems.” The resultant IRM business sub-function and abstract components read as follows:Records & Asset Management (AM) / Asset investment planning (AIP) Asset Decision Support (ADS)Asset investment planning decision support involves strategy definition and prioritisation, maintenance strategy planning, risk management, programme management and decision-making. It drives the condition, configuration, performance, operating costs, and flexibility of the asset base, with the aim of maximising value.Maintenance strategyMaintenance strategy that balances risk, cost, and schedule while applying schedule-based maintenance, reliability-cantered maintenance, condition-based maintenance as appropriate.Life-cycle planningPlanning the use of an asset throughout its life-span, from the time a decision is made to purchase it until the time it is properly discarded.Reliability centred analysisIdentifying reliability trends and corrective actions. For example, looking at all the faults for a given year (planned and outage) and performing root cause analysis and suggesting corrective actions.Engineering and design standardsThe establishment of measurable conditions and performance of assets and how they are employed. Characteristics or attributes of a standard describe its required level of performance, which typically describe “how much”, “of what nature” and “how frequently”. Often engineering and design standards reference industry standards and manufacturing specifications. Performance measurementsAsset management decisions are based on lifecycle costs. Assets are measured so that full economic costs of activities are understood. Risk managementTo monitor, understand, and manage the risks involved in business activities and to ensure that policies, processes, and practices are committed accordingly.Environmental managementAssets must be installed and maintained with an awareness of both environmental impacts on the assets (salt corrosion, ice, water, sand, etc.) as well as impacts on the environment by the asset. The later is subject to many forms of government regulation. Decision support systemsData warehousing and business intelligence used to support asset investment planning decision asset planning and management decisions.Thermal ratings of network equipment and linesChanges in electrical asset performance limits based on temperature and wind speed. Maintain work triggersConverting schedules into work plans such as preventive maintenance for an aging pole.Asset maintenance groups (lists)A list of assets organized into a category, for example to add efficiency to certain types of maintenance and inspection work.Asset failure historyThe history of failures related to all assets in the distribution network including burnout and accidents.Asset financial performanceReturn On Investment (ROI) for assets. Understanding the value and costs of assets and the financial resources needed to appropriately sustain them (short and long term). Budget allocationAllocation of budget for procurement, maintenance, and replacementThe following is a first-cut interaction diagram for AM-ADS.Requirements and Discussion TopicsThe following are requirements and pertinent concepts obtained from the AHFC work.Asset Modeling ConsiderationsSpecialization of AssetUse of AssetContainer child classes (EndDevice, Seal, Cabinet, Duct Bank, Facility, Structure, Tower, Pole, Bushing) – philosophy of when there is specialization vs use of Asset.typeSome Asset child classes exist because they do not have a corresponding role in the network model. “Childing” Asset or (AssetContainer) is appropriate:If there is a need for a relationship to a network model class other than PowerSystemResource (like Terminal for instance)If there is no corresponding network model role and there is asset-specific (not asset type-specific) information that needs to be defined. See EndDevice and EndDeviceInfo example below. Note that the EndDeviceInfo class would be used for both catalog information and typical planned asset information.Relationship between Asset and AssetInfoSome assets may have both asset-specific and model-specific nameplate information. Be able to associate multiple AssetInfo instances to a single Asset instance to support this?(Tanja) The idea of distinguishing instance vs. model *Info is good, but it gives a problem of many on the side of *Info (i.e., non-unique relationship). I would prefer we either:add relationship AssetModel – Asset for this purpose and keep 0..1 on the AssetInfo for everybody. Asset would then go through its model to fetch the model-specific nameplate data, and would be able to “own” its own instance if required). This means that the existing Asset-AssetInfo relation would be for the instance case exclusively. Or,(probably preferred, because self-describing) add a second relationship [0..1]-[0..1] between Asset and AssetInfo, with different role names, for the instance case. The existing AssetInfo role would be the one shared by many (all with the same model) and thus “owned” by the model, while the new one would be for the asset instance, for that instance’s exclusive use – if given.Both of these ensure we still keep [0..1] on the AssetInfo side, to avoid ambiguities.Missing Assets and MaterialsThere are assets that are subjects of asset management programs missing from the model. For instance Bushing and Battery.We have an INF Bushing class and a couple more related to bushing tests, mainly for transformer bushings. This needs to be reworked and adapted (to make it useful for both breakers and transformers).There are materials that are assets management subjects – Oil and SF6 – missing from the model. Asset Group Considerations: AssetInfo and AssetModelThe informative class GenericAssetModelOrMaterial is means for planning purposes. GenericAssetModelOrMaterial is a subclass of AssetModel peer to AssetInfo. Also, look at ElectricalAssets diagram & DCIMAssetsInfo diagram (+ GenericAssetModelOrMaterial class).The connection between Asset and GenericAsset… would be indicative planning aspects of the asset.While there may be many unique designs for transmission assets, there is a lot of commonality and scale with distributions assets such as pole-top transformers.AssetModel & AssetInfo generally created for Asset Procurement. For assets that are found in relatively large numbers such as pole-top transformers, AssetModel may already have been created from prior procurements.AssetModel is somewhat general, where there is a generic item in the asset catalog, such as a 25 kVA pole-top transformer. ProductAssetModel is more specific, e.g., GE 25 kVA pole-top transformer of a particular vintage. AssetModel is keyed for CompatibleUnit and can be used together to great effect. A macro CompatibleUnit can be used to very quickly grab the information needed to put together the design for a new project. The CompatibleUnit should be considered in the asset health modeling.Relocation of AssetsThe health-related measurements should be associated with the assets and be able to move with the asset if the asset if moved.With the SCADA measurements being associated with the PSR and not with the Asset, this could be a challenge – moving operational history with the physical asset. Requires duplication. Also, possible that only a summary of the operational history has to be duplicated – e.g., averages, max, min, etc.Asset and Network Model RelationshipConnection between Asset and PSR. Currently this is many-to-many, which is a problem. There is a proposal to make this one-to-one, which means Assets would have to be aggregated by AssetContainer, which is then associated to the PSR. But components have correspondence as well, such as bushings correspond to terminal?How many connections make sense for different kinds of assets? Suggested usage (by type of asset, by type of network model (balance, unbalanced) is probably fundamental to interoperability, just like asset templates areHow the AssetInfo attributes relate to those of ConductingEquipment and the reference architecture assumptions that underlie their populationReference architecture assumptions related to Measurement relationships to network model objects and to asset-related objectsProcedure and Dataset Cleanup and standardization of the informative procedure-related classes for tests and inspections.Test procedures are applied to an asset or asset component.The test procedures are usually widely applied across the industry.Many are governed by standards.Others are de facto industry practices.The test procedures can be applied in a laboratory or in the field.Examples of laboratory testing are testing of oil samples for DGA and oil quality; forensic analysis of failed assets/components.Examples of field testing are power factor testing; trip coil current on circuit breaker; Sweep Frequency Response Analysis (SFRA) and Partial Discharge (PD); Infra-Red (IR) survey etc.The tests produce measurements of different types.DGA testing produces analog values.Oil quality testing produces enum / string.Inspections and Surveys: Visual inspections are a key part of asset health maintenance regimes. [13 June 2014] ProcedureDataSet, when produced through a work process, always captured with the asset reference – i.e., Procedure instructions are sent to mobile workforce, who then perform the procedure and return the ProcedureDataSet produced along with a reference to the Asset.Test ActivityAHFC discussions have determined that there is value in explicitly modeling testing activity. For instance, a single trip to the field could result in multiple procedures (Assets::Procedure) being performed and multiple, disparate data set being collected (Assets::ProcedureDataSet). We need a way of tracking these multiple procedures and data sets under the single activity.[13 June 2014] ActivityRecord going to be used to track lifecycle stages? If we are going to use ActivityRecord to document test activity as well, shouldn’t it be associated with Procedure?Lifecycle-dependent Testing EventsFactory Acceptance Test (FAT) vs Site Acceptance Test (SAT) vs Reconditioning Test vs Routine Tests.Routine tests are less comprehensive than SAT & Reconditioning Test, which in turn are less comprehensive than FAT.When a transformer is first energized (or re-energized), it is monitored to detect infant mortality. For instance, oil samples @ 24 hr intervals at first, then 1 week interval, and then 1 month interval.As for FAT vs SAT, results from the latter are more important since they represent in-field condition. But if there are symptoms like unusual gassing, may refer back to factory test data.If oil is cleaned, a new baseline is set.The following figures show typical DGA and Oil Quality FAT for transformer. These tests could be done multiple times over a period of several days.To model lifecycle-dependent testing events (comment from Tanja): AssetLifecycleEvent is a compound and at any one time, the Asset instance has exactly one “set” of values in attributes of AssetLifecycleEvent. But, if we need to track what’s happening in time, we have the class ActivityRecord (generic) and should provide a more specific subtype and link to Asset (thus replacing overly generic relation Asset-ActivityRecord currently in the model; I think this relationship was used in one of metering profiles and people were reluctant to remove it, but we can review the real need and maybe negotiate a subclass with part 9 team J).Measurements Not Made at the AssetSome measurements not made at an asset is important from asset health perspective – e.g., flow for breakers. How do we consider such data?Retrieval of Measurements as a Time Series DataBeing able to retrieve time series data for trending etc is important for asset health.On this subject, the following is from Tanja’s notes on WG14 modelling call on 2013-12-17: The pic below shows the existing time series model for meter readings; this seems to be specific usage for meter readings. Note that here, every reading will get a new mRID (which is different from how MeasurementValue is intented to work).We made an exercise of adding the means to specify multiple triples {value, timestamp, quality} (will call them {vtq}) to the existing CIM Measurement model. See pic below.It would require to:- define additionally 4 new compounds (one per MeasurementValue subtype) and add an attibute [0..n] values. Note: name AnalogueReading is not recommended (don't want to mix with meter readings).- agree to allow multi-valued attributes in CIM (open combined issue 1126).Note that this potential solution does allow for smooth transition from the existing model to something that would semantically support time series of measurements. However, we are polluting the model with unnecessary redundancy.(added after the call) We should maybe give another name for the new 'AnalogValue.values' and mark it as deprecated, for in the future simply changing the type of an existing 'AnalogValue.value' from Float to AnalogueReading?Measurement MetadataIn addition to units/multiplier, need other meta-data about the measurement data, such as whether the data is max, or min, or peak, or average, etc. Such qualifiers are relevant to analytics and aren’t captured in CIM.Waveform MeasurementsThere is need to represent the test data that is in the form of a time or frequency waveform.Trip coil current test produce time waveform (figure below).Sweep Frequency Response Analysis (SFRA) and Partial Discharge (PD) produce frequency response.These test results make sense only as the complete waveform, not as individual measurement points within the waveform.Health Index and RiskWe’ve had preliminary discussions on Asset Health Index and Risk. Numerous issues to resolve:Attributes for AHI and Risk.Risk for individual assets as well as asset groups.Data that goes into computing AHI and Risk. The latter in particular involves data such as financial data – insurance claims, project costs, O&M, capital etc. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download