Fuji Oil leverages Cognite Data Fusion® to realize the construction of a virtual refinery

Fuji Oil specializes in petroleum refining. It is headquartered in Shinagawa, Tokyo, and has a refinery in Sodegaura, Chiba Prefecture. Established in 1964, the company began operating the Sodegaura Refinery in 1968. Since then, it has consistently engaged in petroleum refining and the sale of petroleum products, playing a crucial role in supplying energy to support the lives and industries of the Tokyo metropolitan area, one of Japan's major consumption centers.

The left side: Hideaki Takahashi, Manager, Technical Planning Section, Technical Department at Fuji Oil
The right side: Takuya Nishimura, Supervisor, Technical Planning Section, Technical Department at Fuji Oil

The company has identified four key issues in its near-term business plan:

1. Maintaining and enhancing the operational reliability of refinery equipment
2. Strengthening cost competitiveness and establishing a competitive advantage
3. Thoroughly reducing the environmental impact of refineries
4. Pursuing the decarbonization of its business

The company aims to use digital technology to the fullest extent possible to enhance operations. The Technical Planning Section of the company's Technical Department is at the core of digital promotion and is responsible for considering the introduction of technology, implementing it, and promoting its subsequent use.

The decline in the working population in Japanese society is an urgent issue for the oil industry too. In order to make up for any future shortage of engineers, it is necessary to secure and create medium- to long-term resources by making work more efficient through the use of digital technology. At Fuji Oil, the issue of a younger generation of engineers in the operation, maintenance, and technical departments had become apparent, but it took a long time to consider and implement specific plans to solve the problem.

The Technical Planning Section first surveyed the work content of engineers in the operation, maintenance, and technical departments and conducted an inventory of their work. It was found that a great deal of time was being spent on collecting technical data in all departments. In particular, a lot of time was spent on researching technical information used by the operation, technical, and maintenance departments.

It has been about 50 years since operations began. The number of drawings for the equipment that has been continuously upgraded and repaired is huge. Although some of the equipment was integrated into the existing system, not all of the drawings were managed in a centralized location. When engineers were drawing up maintenance plans, they often had to visit the site where the specific drawings were stored to find the files, and with limited resources, the time spent on the process of carrying out their normal duties was a major issue for improving engineer productivity.

The time spent by engineers collecting materials and searching for data is estimated to be over 10,000 hours per year. If data can be centralized through digitization and direct access to the necessary data is made possible, the time spent collecting materials can be reduced, leading to significant improvements in operational efficiency.

The deciding factor was the demo that worked with real data

The Technical Planning Section held a number of meetings with potential vendors with the aim of introducing technical information infrastructure software. In order to accelerate the digitization process, it was necessary to make the most of the measures and systems that had been put in place up to that point.

In selecting a vendor, the key point was whether or not the system would be able to smoothly link with the PFD (Process Flow Diagram) and P&ID (Piping and Instrumentation Diagram) data for each device on the existing system. Although some PFDs and P&IDs had already been collected and consolidated to a level that could be used in practical work on each device, whether or not it would be possible to easily link various types of information together without taking a long time using existing resources and solutions would affect the speed at which the medium-term business plan could be realized.

The proposal from Cognite was detailed. They had heard that Cognite Data Fusion® could flexibly link with all kinds of data through Extractors and APIs, and that it could make great use of existing systems, but when they gave Cognite some of our actual technical materials, they were able to see a demo environment Cognite Data Fusion® that had completed data linking and contextualization in less than a week. Seeing how the data could be contextualized much more quickly and accurately than they had imagined when it would have taken several months to do manually, They became convinced that starting a project with a small initial investment would lead directly to improved business efficiency.

The official decision to introduce Cognite Data Fusion® within the company was made in March 2023, and the project was launched in May of the same year. In order to proceed with the implementation as quickly and smoothly as possible and achieve results, the project was divided into two phases: Phase 1, which involved creating model cases for specific devices, and Phase 2, which involved large-scale implementation.

Phase 1, the initial introduction, introduced Cognite Data Fusion to 11 devices, equivalent to about 1/4 of all the devices at the refinery. To maximize the effect and make it easier to obtain feedback from users who actually use the devices within the company, a small start was made targeting the main devices with particularly high operational utilization rates. The implementation period was only 6 months, which was about half the expected period.

Hideaki Takahashi said, “Cognite came up with a precise proposal that would realize Fuji Oil's vision in a short period of time. The short project period until implementation, as well as the after-sales support system, including maintenance and management after implementation, were also major factors in the decision-making process. We were also impressed that they were able to complete Phase 1 in just six months, which was half the time we had originally estimated, by making full use of the innovative technology of contextualization.”

To date, the system has been well received within the company. As many departments have requested that the system be introduced to all refinery equipment as soon as possible, we are moving forward with the introduction of Cognite Data Fusion® for all refinery equipment as Phase 2, starting in May 2024, without a break from Phase 1.

Takuya Nishimura said, “It is now possible to obtain the materials that were previously managed in a complicated way and depending on each department, through the P&ID linked to the Cognite Data Fusion's UI. The Technical Department often holds meetings to identify energy-saving and profit-improving projects for the refinery. During discussions, it is often necessary to check the design information, drawings, and data sheets for many pieces of equipment, but using Cognite Data Fusion® has greatly reduced the time taken to search for documents, and this has actually started to have a positive effect on the efficiency of meetings and the speed of decision-making.”

In addition, in terms of operational management, the Charts function, which visualizes data in real-time in Cognite Data Fusion® and allows users to easily analyze and monitor time-series data, is used for the operation management of refinery equipment.

The Results

As a result of Phase 1 of the project, Fuji Oil reduced the amount of time spent searching for information by approximately 2,300 hours per year.

In Phase 2, which involves introducing Cognite Data Fusion® to all equipment, we are expecting to reduce information search time by at least 7,900 hours per year.

In addition, the introduction of Cognite Data Fusion® has enabled the Technical Department to spend the time it has saved on other important considerations, such as the operation and management of refinery equipment, energy conservation, and profit improvement studies, and new business studies, leading to the advancement of operations.

In the future, they aim to achieve improvements by further reducing working hours through the use of VR and by enhancing operational management in areas not covered by these figures. In particular, they also expect to see improvements in terms of work style reform and health and safety through the use of VR to reduce on-site work.

Diagram of data integration using Cognite Data Fusion®

Aiming to build a Virtual Refinery

In the previous project, as Phase 1, the introduction of the system to approximately one-quarter of the main equipment used in oil refinery operations was completed, and its use in actual operations has already begun. By the end of FY2024, the introduction of the system to the remaining equipment is scheduled to be completed as Phase 2, and this is expected to complete the information infrastructure linkage of the equipment used in oil refinery operations.

With the completion of Phase 2, the infrastructure solutions for the refinery operations that the company aims to achieve will be in place, and the company is expecting to see full-scale use of the system in actual operations for all devices. Furthermore, Fuji Oil has started to use Cognite to create a VR space for its equipment group, and is currently conducting a PoC for one piece of equipment. In the future, we aim to further improve productivity across the organization by integrating the use of VR space and robots to further reduce the labor required for on-site work.

“The combined use of Cognite Data Fusion® and VR has great potential to solve various operational issues that have remained unresolved in the manufacturing industry for many years. We want to build a virtual refinery through further collaboration with Cognite.” said the project team member. The project is being promoted with enthusiasm to realize the concept."