Search past winners/finalists

Tüpras - IT Team of the Year

Gold Stevie Award Winner 2016, Click to Enter The 2017 Stevie Awards for Sales and Customer service

Company: Tüpras, Kocaeli, Turkey
Company Description: Tüpras is Turkey’s sole oil refiner, largest industrial company by revenue and added-value generated. It operates 4 refineries in Izmit, Izmir, Kirikkale and Batman, with a total annual crude oil processing capacity of 28.1 million tons. As Europe’s 7th largest refining company, Tüpra? ranks among the most complex refiners in the Mediterranean region, with an average Nelson Complexity Index of 9.5
Nomination Category: Information Technology Categories
Nomination Sub Category: IT Team of the Year

Nomination Title: Tüpras Big Data & Analytics Team

Tell the story about what this nominated team has achieved since 1 January 2016 (up to 650 words). Focus on specific accomplishments, and relate these accomplishments to past performance or industry norms.

In an oil refinery everything happens in big proportions: liquids flow in tons/hour rate, temperatures are measured in hundreds to thousands °C, and electricity is produced in Megawatts. Millions of dollars are at stake every moment as one little system malfunction or mistake can cause serious damage to the entire plant and the workers or at least cause losses in revenue. Thus, it is necessary to continuously monitor and verify the accuracy of measurements streaming in from numerous and various types of sensors placed around the refineries. Proactive monitoring is the term that describes monitoring paradigm, that enables operators to take actions before unwanted events occur. Through monitoring proactively, human/system would notice problems while they are visible as small events, and intervene before there are more drastic consequences such as larger equipment failures, plant destabilization, human injury or loss of production efficiency.

Because of the reasons above, a team has been initiated in order to implement “Big Data platform”, “Proactive monitoring platform” and “Analytics (Predictive & Prescriptive) platform” can work together helping to achieve better level of decision making for;

what is the system doing now
why is it doing that it is doing
what will it do in near future

This project is started with the identification of the primary needs of refinery and determination of the key performance indicators and become functional with connections to different data sources, integrated data management, development of managerial reporting tools and customizable online dashboards.

For the first time in literature, we describe the real-time big data validation problems witnessed in oil refineries and simultaneously demonstrate application of a carefully-selected and integrated set of analytical and system tools from different domains (statistics, signal processing, and distributed systems) as a solution.

The mentioned data is stored and processed on the developed “Big Data platform (Hadoop, MapR)”. “Spark”, “R”, “TSDB” is used for large-scale data processing and statistical computing. https://en.wikipedia.org/wiki/Data_analytics">Analyzing (processing) streams of data takes place on “Esper CEP” engine and visualization is improved with “.Net” and “HTML 5” based graphical user interfaces. “iOS” based applications are developed for accessibility from mobile devices. Furthermore, special reporting tools are developed using “SQLServer OLAP” services for designed multi-dimensional data models.

We are working on a cyber-system that develops predictive models by using machine learning and deep learning algorithms fed by historical and online sensor data stored in our big data platform. We currently work on several cases to enhance the efficiency of processes, quality of products and prevent unexpected failures.

The two main cases are briefly described as:

Corrosion Control: The system predicts the corrosion rate in critical static equipments in crude-oil unit to prevent unexpected failures.

On-Line Spalling Control: On-Line Spalling (OLS) is an operation implemented in a Delayed Coker Unit (DCU) to remove coke deposits from the inside of the coker furnace tubes. The main input data for the decision of spalling time are coker furnace process tube temperatures. The system predicts and model the trend of the furnace process tube temperatures and minimize the number of OLS to maximize the profit.

On-Line Spalling Control: On-Line Spalling (OLS) is an operation implemented in a Delayed Coker Unit (DCU) to remove coke deposits from the inside of the coker furnace tubes. The main input data for the decision of spalling time are coker furnace process tube temperatures. The system predicts and model the trend of the furnace process tube temperatures and minimize the number of OLS to maximize the profit.

With our “Big Data platform” and “Proactive monitoring platform”, by processing millions of raw data that comes from thousands of sensors in our refineries, we continuously monitor and analyze every stage of oil production.

By combining the financial data, environmental/safety metrics and process data, we obtained the power of turning data into KNOWLEDGE.

We are proud of the project’s contribution to our company’s decision support mechanism which has been developed completely with our own resources.

In bullet-list form, briefly summarize up to ten (10) accomplishments of the nominated team since the beginning of 2016 (up to 150 words).

-Establishing a new “Big Data platform based on Hadoop, MapR”
-Integrating daily 300 billion raw process data depended on 200K sensors of 4 refineries into Big Data
-Historical export of 10 years data intagrated into Big Data
-Reducing the data frequencies of 30s and 60s with the new platform to 1s frequency
-Developing “Tüpras Historian Database-THD” for access and analysis of all refinery data with a single web-based application
- “Management Information System-MIS” platform developed for visual analysis of approximately 50K metric/KPI calculated from process data. The platform supports self service reporting and Decision Making Support tools for “proactive monitoring & analysis”.
-Developing “Engineering Platform” to run fast What-IF scenarios
-Developing “Alarm Management” system to centralize and analyze DCS (distributed control system) alarms
-Implementing of “Predictive Maintenance” scenarios based on Machine Learning
-Corrosion Control
-On-Line Spalling Control
-Coke Control in Furnace Tubes
-Developing IOS based mobile applications of MIS and THD