Skip to main content

Doing more with data

Published by , Senior Editor
Hydrocarbon Engineering,

These days, there is no escaping the phrase ‘digital transformation’. There is no denying that the rise of Industry 4.0 is impacting day-to-day environments. Digital transformation is challenging traditional approaches and and can open up new opportunities, if the willingness to adopt is there.

The majority of the organisations (approximately 90%) in Fortune’s Global 500 are oil and gas companies.1 In general, these companies are aware of the importance of the data being accumulated and of the need for data analytics and integrated technology throughout their production lifecycle. How can these oil and gas processors capture this data in the most useful way?

Empowering the experts

Emerging analytics platforms that combine machine learning technologies (part of the artificial intelligence umbrella) are being more readily adopted with the goal of optimising all areas of operations. In order to achieve the highest levels of optimisation, however, the right information needs to be made available across all horizontal and vertical levels of the organisation, and the type of information needed can vary depending on who in the processing plant is accessing the data. When process and field engineers can easily gain insight into their processes, they can solve more day-to-day questions independently and enhance their own effectiveness. In turn, they will provide their organisations with new insights based on their specific expertise in engineering. This delivers value to the owner-operator at all levels of the organisation and leverages (human) resources more efficiently.

Enabling the subject matter experts with analytics does not mean asking them to become data scientists. They can remain vital within their section of the plant floor. New plug-and-play software solutions are available that are user-friendly and can begin creating value immediately ‘out of the box.’ These new, so-called ‘self-service’ industrial analytics tools allow engineers to be able to use the data coming from a variety of sources. They help to identify trends in their production processes, visualise similar historical events, monitor behaviour, and notify control room and engineering teams of potential deviations in live processes.

Enhance day-to-day operations

Self-service industrial analytics solutions provide refineries with the ability to maximise the value being stored in the wealth of data they already own. One example is that engineers can now easily identify their optimal process behaviours or operating zones, otherwise known as their ‘golden batch fingerprint’.

A case in point is seen with Ashland, a US$5 billion specialty chemical provider that had struggled with transitioning one of their plants to a good manufacturing processes (GMP) production environment. When the temperature of its distillation tower kept rising and causing a temperature spike, also known as a trip, Ashland was able to diagnose and fix its root cause: a measurement failure that triggered manual control. This was done by using self-service industrial analytics and by creating a ‘golden batch fingerprint’.

Key issues were identified and isolated as a pattern within the analytics solutions and used to find previous similar occurrences. Historical events were compiled to create a graphical fingerprint that could be used for monitoring. With monitoring in place, Ashland engineers could receive early notifications of any deviations in process behaviour, which gave the company enough time to prevent a trip from occurring. This increased the plant’s on-target production and delivered substantial cost savings. The on-target production of Ashland’s GMP products increased from 70% to 95%, assisted by the self-service industrial analytics software.

Time and time again, such analytics solutions have helped to guide the operators on the manufacturing floor to identify the best batches historically. This is done by doing a comparison using analytics and then building a fingerprint based on the best batches. As seen with Ashland and other petrochemical companies, corrective action can be taken sooner for batches to be produced according to the best proven efficiencies.

The power of meaningful data

When oil and gas companies look to leverage meaningful data collected by their myriad of systems, they realise they have amassed tremendous amounts of data but lack affordable and effective tools to search for actionable information to make better decisions. Self-service industrial analytics using machine learning, pattern recognition and artificial intelligence can now help.

Analytics can point experts in the right direction of meaningful data that can make a difference in their operations. For instance, in one scenario where heat exchangers were used in the refining process in fluid catalytic cracking units as well as in the liquefaction of natural gas, the rapid-fire pattern recognition capabilities of the self-service analytics tool pinpointed the relationships, uncovering potential quality issues with heat exchanger seals. It was determined that due to fouling, a batch cycle took longer to complete. Analytics uncovered that a switch made to a spare heat exchanger caused a potential fouling problem.

Thanks to self-service predictive analytics, several monitoring alerts identified in advance potential issues that were then able to be confirmed by operators. By having detection in advance, corrective action was immediate and lost production was avoided. Due to the clever contextualised cooling time patterns tracked, maintenance alerts ran, preventing the heat exchanger fouling from causing longer cooling times.

Predictive analytics and artificial intelligence

It is becoming increasingly standard for downstream oil and gas operators to continually monitor equipment performance in real time. But often only 1 – 2% of most critical assets are monitored. With self-service industrial analytics, the other 98% of assets can be monitored too. Advanced analytics methods such as diagnostic, visual, and predictive analytics, when combined with artificial intelligence (AI) algorithms, can be useful; for example in providing time-series Industrial Internet of Things (IIoT) data to even further leverage IIoT applications.

Next generation analytics/AI combination solutions have demonstrated improvements in uptime, efficiency, and throughput. For instance, the International Energy Agency (IEA) cites that digital technologies will expand oil and gas reserves by about 5% and decrease production costs up to 20%.2 Managers must be able to make educated decisions about procurement, production scheduling, and delivery without having to spend a lot of time and money on modeling by data scientists. This is why advanced analytics solutions have emerged, which leverage the contextual information from other business applications while analysing operation performance. Not only is it needed at the production level but also at the enterprise level, where oil and gas executives require real time, accurate data to relate production to the larger business context and understand the impact of fluctuating costs, changing market conditions, and asset performance.

IIoT initiatives

Much of this digitalisation in the oil and gas industry is powered by the emergence of IIoT, which according to Research N Reports, is expected to grow from US$64 billion in 2018 to US$91.4 billion by 2023, at a compound annual growth rate of 7.39% during that forecast period.3 All one must do is look at the companies rapidly adopting (and expanding) the implementation of new technology to see that those numbers may well prove true.

Total, one of the world’s largest integrated petroleum refining companies, piloted self-service industrial analytics software at its refining and chemicals (R&C) division. The company quickly realised the great potential of easy data exploration and process analytics to increase productivity and plant availability and, after the pilot, rolled it out to their operations worldwide. Analytics improved equipment effectiveness and advanced the company’s data-driven transformation.

Total R&C remarked how beneficial it was to utilise such a user-friendly self-service analytics tool, noted the time savings and felt it delivered better quality results than data models. The company was able to better define its normal operating windows using historical data and identified the most reliable and effective patterns to create a golden batch fingerprint. These fingerprinted patterns can then be used for real time monitoring, easy comparison of batch production for different grades, and to trigger pre-maintenance notifications. Trend search and filtering capabilities saved time, increasing both employee and production process efficiency.

Benefits of self-service industrial analytics

On a granular level, self-service industrial analytics have become key because they help analyse performance, identify anomalies, and provide advisory services, as well as actionable recommendations to optimise operations. Ultimately it assists in reducing failures and increases machine availability. According to McKinsey’s benchmark, oil and gas operators have not maximised the production potential of their assets.4 It has been noted that bottlenecks caused by high gas pressure variations could be resolved with two algorithms. The first would predict the risk of unplanned pressure spikes and the second would minimise the size of the pressure spike. By predicting the probability of pressure build up and reducing the size of each pressure spike, a 1 – 2% gain in production could be made. Combining sensors and advanced analytics, oil and gas companies can leverage the data acquired through condition monitoring, improve uptime and reliability, and ultimately reduce costs while increasing production to drive profitable growth.

Continuously under huge pressures to ease cost burdens and augment production, one of the ways refineries are successfully tackling digital transformation is by implementing self-service industrial analytics. McKinsey benchmarks cited how the use of analytics will fill in the performance gaps that hold back production.

Advanced data analytics can improve a company’s bottom line by prolonging equipment life, increasing asset availability, and extending maintenance windows.

Substantial productivity gains are on the upswing as refinery and asset tasks become more autonomous and AI-assisted. While the oil and gas industry continues to manage big volume, variety and velocity, the industry is also starting to think beyond its self-made boundaries to truly capture the benefits of self-service industrial analytics, machine learning, AI, and IIoT.


  1. FARRIS, A., ‘How big data is changing the oil & gas industry’, Analytics,
  2. ‘Digitalization & Energy’, International Energy Agency,
  3. ‘Digital Transformation In Oil & Gas Market Sector With analytics Will Touch A New Level In 2025’,
  4. BRUN, A., TRENCH, M, and VERMAAT, T., ‘Why oil and gas companies must act on analytics’, McKinsey & Company, (October 2017),

Written by Nick Petrosyan, TrendMiner, USA.

Read the article online at:

You might also like

Enerflex and BASF to collaborate in CCUS applications

Enerflex Ltd. (Enerflex) and BASF have entered into a Memorandum of Understanding (MoU) to jointly facilitate the commercial scale deployment of carbon capture, utilisation and storage (CCUS) applications.


Embed article link: (copy the HTML code below):


This article has been tagged under the following:

Downstream news