"Managing data is a bit like playing whispers. You start out with one truth. Then something changes, perhaps the data is transferred to another system, some decimal places are deleted and the number is rounded off. Then there are errors and consequential errors, but when you go back to the source, it is not certain that the source is correct either. Perhaps the figures have been updated with a change that has not been returned to the original source", begins Åge Løklingholm, Principal Data & Digitalization Consultant at Cegal.
Being sure that the data is correct, up-to-date and accurate is a challenge for most businesses. The issue has become even more relevant in the AI age, when artificial intelligence is taking over and automating more and more tasks. This makes it even more important to be sure that the data is correct. This is described by Cegal's Anaïs Kermagoret, Principal Solution Architect at the CTO Office.
Petoro's task is to manage the state's interests in the oil and gas sector on the Norwegian continental shelf. Petoro shall create the greatest possible value and achieve the highest possible income for the state from the State's Direct Financial Interest (SDFI). Through the SDFI, the Norwegian state has interests in a number of oil and gas fields, pipelines and onshore facilities. These interests are managed by Petoro, which prioritizes creating more investment opportunities, finding solutions across the portfolio and taking care of our surroundings. The cash flow from the SDFI is transferred in its entirety to the Pension Fund of Norway.
"A major challenge is that we have enormous amounts of data collected over a long period of time. In addition, the original data is often processed to adapt to tools that provide insight and as a basis for analysis. Over time, this creates uncertainty about data quality and what is the right information. This is because the interpretations we use to estimate the value of decisions and future revenues are updated and reassessed over time, or unintentional changes occur in the import of data," explains Ragnar Sandvik, Strategic Advisor at Petoro.
For Petoro, which manages highly valuable information from oil and gas production on the Norwegian continental shelf, incorrect data can have major financial consequences. Therefore, Petoro initiated a Data Management project to ensure data quality and be confident that the data sources are correct and up to date. Cegal assists with program developers, analysts and data expertise in the "Source of Truth" (SoT) project.
"We want the data to give the same answer, no matter what data source we use. We want there to be one truth for us, the authorities and the operators on the Norwegian continental shelf."
Ragnar Sandvik, strategic advisor in Petoro
The first part of the project involves the geographical location of about 100 wells. This was a PoC (Proof of Concept) to show the benefit of the SoT project and assess whether it was profitable to expand the project to cover larger parts of Petoro's operations.
"DISKOS is Norway's official data source and the authorities' archive. If the official data is not consistent with Equinor's, for example, we need to know why. Has there been an error at Equinor, has Equinor corrected incorrect data that has not been updated or is the error with us in our systems. There are many sources of error. For example, when data is moved between systems, decimals can be rounded off and removed or geolocation can be described in slightly different ways. Then there are small changes in the data that can have major consequences in the results of analyses," Sandvik continues.
Petoro started by examining the data associated with the wells on the shelf, so-called metadata; location geographically and in depth, whether they are in production, what the status is and so on. It is very important that the wells are correctly located in the data models Petoro uses. If the information is wrong, this can lead to incorrect decisions. Petoro's goal is to maximize the value of the fields on the Norwegian shelf, so we must avoid decisions being made on the wrong basis. "We first wanted to investigate whether the official data retrieved from wells and reported to the authorities is consistent with the latest data used by the operator and the data we use ourselves in our simulations," says Sandvik.
Ragnar Sandvik - Petoro
The challenge in the project is that parts of the data reported to the authorities are not structured in a database or as machine-readable files, but are stored in pdf reports and as images.
The information to be structured thus comes from four sources:"Our first test proved the usefulness of the system. The system read and evaluated different sources of well data from approximately 100 wells in a field over the course of a day. Unfortunately, the results were disheartening for the industry. The sources of the data were not consistent and there were a lot of discrepancies, and we didn't know where the errors lay," says Sandvik.
With the Source of Truth project, Cegal has provided Petoro with a tool for data evaluation and data comparisons. By utilizing advanced analytics and automated processes, SoT makes it possible to assess and compare data from different data sources efficiently. Microsoft Fabric and Microsoft Power BI are key components of the solution. Integration with Fabric and Power BI streamlines data ingestion and visualization.
"If we were to sort data manually, it would take weeks. It would also be easy to make mistakes, and for humans this is a demotivating job. By using Source of Truth, Petoro will be able to spend its time creating added value for the Norwegian people rather than checking data."
Ragnar Sandvik, strategic advisor in Petoro
Sandvik says that Petoro chose Cegal as a partner in the project because of Cegal's expertise in exploiting the possibilities of Microsoft Azure.