Question: What must I do for "data integration"?
Answer: Data integration is an important concept in the IM rule. In principle, this is an action that will help assure that operators learn about their pipelines the things that data from disparate activities can tell them. An example is provided in ASME/ANSI B31.8S:
An operator suspects that a possible corrosion problem exists on a large diameter pipeline located in a populated area. However, a CIS indicates good cathodic protection coverage in the area. A Direct Current Voltage Gradient (DCVG) coating condition inspection is performed and reveals that the welds were tape-coated and are in poor condition. The CIS results did not indicate a potential integrity issue but data integration prevented incorrect conclusions.
The analytical process considering the synergistic effect of multiple and/or independent facts or data constitutes data integration.
Data aggregation is a first step. Often, data that is generated about the pipeline from routine activities has not been seen by other groups within the company. Data aggregation should bring together all relevant information so that it can be better evaluated in context with available data. However, data aggregation, by itself, is not sufficient. Operators must also evaluate the aggregated data to look for problems that might not have been identified absent such an evaluation. There is no single means of performing this evaluation. GIS systems can be significant aids in performing data integration, but use of these systems is not required. The models used for risk analyses required by the rule can also be a valuable tool for performing data integration. In some cases, use of subject matter experts (SME) may be sufficient. An operator needs to consider the types of data available and the relative complexity of their pipeline system and its environment and then develop and implement processes for data integration that are appropriate for its particular circumstances.