Geologist of the Future

Nenad Stanojevski 12 June 2018 • 20 min

Over the past few years, STC has initiated modernisation of IT equipment and raised the business support in this area to a much higher level.  The virtualisation of professional graphics workstations has been completed, the High-Performance Computing (HPC) is currently in use and the testing of the VR headset intended for analysing infrastructure and 3D geological models has started.

Scientific and Technological Centre (STC) is in charge of the preparation of geological investigations and the calculations of reserves, processing and interpretation of seismic data, design and monitoring of the oil and gas reservoir rock characterisation, preparation of the well completion designs and infrastructure projects, including laboratory testing.

In 2013 and 2014, NIS started creating its digital databases and modernising its IT equipment, while also raising the level of business support provided in this area. According to Goran Jajic, Automation Manager in STC, the Centre’s first project was the virtualisation of professional graphics workstations which was launched in 2014.

– The most demanding computer users in the Company are those working at STC. Software used in STC is exceptionally demanding in technical terms and thus regular computers are unable to support them. The key demanding factors for these computers are special professional certified graphic cards which support the presentation of highly demanding models and the handling of huge files. As an example, they can provide a review of the Regional Geological Model of the Pannonian Basin whose size exceeds 180GB. Employees need to upload the files of that size on their own computers, which is highly demanding in terms of capacity. The tasks also often require large monitor displays as well as a portrayal on multiple monitor displays which additionally increases the complexity of the entire system – says Jajic.

According to him, the server room at STC stores everything. All users of virtual machines, whether they are physically in STC, Belgrade, Russia or at home, are actually working on a system which is stored here. All calculations, simulations and analysis are performed on the servers here while the users only get to see them at their current locations. – Right now, we have 12 virtualisation servers which provide support to approximately 140 users. However, the tremendous power of these machines is what we can offer in the case it is needed for project purposes and we can change the machine configuration on a daily basis according to users’ needs — says Jajic.

STC is now making preparations for the virtualisation of the Engineering Department where traditional workstations are still in use and where a complex 3D infrastructure design was implemented in 2017. These workstations are not connected to the STC’s electric switchboard and practically need links to be connected to the existing servers which enable the use of virtual machines. – We have been using virtualisation technology for three years now and I can say that the benefits obtained far surpass the challenges we had to face. The number of users with remote access is continuously increasing. Previously, some employees had to come back to the office on weekends to finish their work and now they can do it from the comfort of their own homes. This technology facilitates the work of our employees, reduces the company’s expenses and increases the productivity of employees who are willing to dedicate a portion of their own spare time to work or additional education after the work hours — says Jajic.

The project was launched in 2014 after the first tests were completed. It is interesting to note that the vendor of graphic cards for virtualisation servers announced the existence of the system in December 2013 and that the STC started its first tests only a few months after the system hit the global market. Since then, three generations of these cards have been created and our system uses specimens of all three generations of these cards which goes to show our commitment to the continuous improvement of the system and to enabling employees in STC to work on the state-of-the-art equipment only a few months after its global premiere.

— Now, our virtualisation system practically enables us – continues Jajic — to allow our Russian colleagues, who work on joint projects with our employees, to be physically absent here while doing the work in their own offices on our machines, using our documents, models, etc., enabling us to communicate and observe within a single virtual environment. It enhances the quality of our cooperation and mutual projects — explains Jajic.

Sanja has been working in the Section for Geological Works (GIR) since April 2017, after shed had completed an internship program in our Company. She holds a master diploma in geological engineering and, as she puts it, she has always wanted to be involved in research. She is currently engaged on projects for seismic and geological interpretation of 3D cubes where she carries out 3D modelling and calculation of hydrocarbon reserves.

Sanja Marković
Sanja Marković
Junior Engineer for Geological Interpretation at NIS

Digital Data

In our contemporary world, data have strategic importance and real monetisable value. Work with geological data and reservoir rock characterisation data is an important element of NIS STC’s business. Management project for data of this type started with the digitalisation process i.e. by scanning of the hard copies of documents and their conversion into digital format. More than 300 thousand documents were digitalised. The creation of a digital database enabled us to fully utilise contemporary, high tech IT solutions for analysing reservoir rock characterisation, selection of new drilling sites, etc. STC works with data utilised by modern software – Petrel (digital geological modelling), TechLog (interpretation of geophysical measurements in wells), NGT-Smart (reservoir engineering dynamic control).

The main system for storage of geological and technological data is GeoBank which stores the most important information about wells, reservoir rocks and their productivity, core analyses, geophysical measurements and a lot of other info that well development engineers and geologists use in their day-to-day work. The system has special tools for data access and import in accordance with the adopted policies. In 2018, the automation of the part of the business process was initiated together with the implementation of tools for automatic data migration from other storage systems to GeoBank and for uploading data to NGT-Smart which should ensure that data in all information systems are complete and harmonised.

Another significant stage of data management began in 2017 – the Project for Verification of Geological and Technological Data. The goal of this project is to check the data sets contained in the main systems used for storage of geological and technological data in terms of their completeness, harmonisation and geological justification. The project envisages the organisation of automatic data checks by creating special program modules. The logic of the module will be prepared jointly by IT professionals and experts in geological and reservoir rock characterisation.

Likewise, STC is in charge of the support and development of an electronic database of knowledge – Thinker. Thinker contains over 3.500 books, 1,600 articles, 700 standards and 230 educational courses. Thinker is a system intended to share the knowledge among all employees of the Exploration and Production Block and the operation of this system is one of the priorities of the Unit for STC’s Databases.

The development of new technologies gives rise to new questions, goals and challenges in the area of data management. In the near future, STC NIS intends to start applying technologies such as Big Data and Machine Learning. Thanks to deep data analysis, these technologies will provide new answers to old questions as well as optimisation of routine processes.

Smart Computer

HPC system (High-Performance Computing) is a server cluster comprising 64 smaller servers interconnected into a single unit which processes geophysical data using Paradigm software suite. HPC consists of 64 servers i.e. computing nodes. Each node has a 2×8 core CPU and 128BG RAM. In total: 1024 core and 8 TB RAM.

After the data acquisition in the field is complete, the recorded data are brought to STC and entered into this supercomputer. These data are then used by geophysicists to create 3D cubes. So, the processing of these crude data on this supercomputer results in seismic cubes which are then sent to seismic interpreters for further processing. – The processing of the crude signals to obtain a three-dimensional cube is a highly demanding mathematical task requiring an extremely powerful machine to process it. Our computer is among the most powerful in Serbia. In the last four years, HPC processed more than 10,000 km2 of seismic data obtained from NIS’s exploratory areas in Serbia, Romania and B&H. By the end of 2017, the processing of 3D seismic data obtained from the shelf of the Pechora Sea (Russia) was completed as part of the program for the development of cooperation with Gazpromneft-Sahalin Company.

He adds that STC has big plans for future HPC. – Currently, the concept envisages the use of this system only for geophysical purposes and the current method does not envisage other purposes. Our plan is to test the possibilities for some kind of virtualisation of this system which would enable it to be used by other organisational units of STC when the needs arise. For example, new trends in petrophysics are such that software can use supercomputers for reservoir rock characterisation and analysis. If we were to install a software which enables HPC to share its resources to multiple parties, and not to just one as it currently does, this would literally mean that available computing resources would be shared and distributed as per current needs – on any given day, these resources would be required either by geophysicists or petrophysicists for reservoir rock characterisation, etc.

As a part of development of the business unit for seismic data processing, STC currently analyzes potential procurement of an “heir” which will be up to 100 times more powerful (speed above one petaFLOPS), while keeping the old one in use – says Jajic, adding the third part of the story which is an attempt to implement the state-of-the-art technologies in our business environment.

While analyzing justification behind the purchase of the new HPC, we are also reviewing the option to create a single large HPC system. With intelligent analysis, it could provide computing for various areas and, upon purchase of additional graphic cards, it could be used by the entire company i.e. other organisational units outside of STC (e.g. procurement or marketing) for BIG data analyses, special machine learning and use of programs with AI (artificial intelligence) elements. All of the abovementioned areas of application are very attractive and topical at this moment but a relatively small number of IT professionals deals with them due to enormous technical requirements for computing capacities needed to perform the tasks. Also, such independent systems are very rare outside of the Cloud – says Jajic.

Virtual Reality

A new technology currently implemented in STC is the use of VR headsets. They are used for two purposes. In engineering, it shows a 3D image of the facilities which offers, at the design phase, a virtual tour through the facility. In practice, the project investor has the option to see the facility “live” and thus avoid modifications and errors in the construction and exploitation stages. Alternatively, VR headsets can be used in Petrel software for the presentation of models created in Petrel. Also, it would provide to geological interpreters the option to “dive into” a three-dimensional geological facility i.e. model which enables them to see layers, faults, geological structures and channels of a potential hydrocarbon reservoir and thus significantly assisting the geologists and making them a lot more efficient.