Prognostic Asset Management Using Artificial Intelligence

#AssetManagement #AI #IIoT #Algo8

Nandan Mishra

Any unplanned shutdown of the supercritical equipment even for a small duration can reduce plant availability for more than a day as plant start-up, stabilization and on-spec product routing takes considerable time and man-hour extra efforts.

The Need for Predictive Maintenance

Unplanned downtime from rotating equipment, such as pumps and compressors, has a high impact on a plant’s Overall Equipment Effectiveness (OEE) metrics, plant availability and profitability.

For example, in oil and gas downstream industry, the cost of unplanned downtime of a Recycle Gas Compressor in Hydrocracker unit or Wet Gas compressor in Delayed Cocker or RFCCU even for a day can be as high as $5.0 million (USD) depending on plant capacity and refinery configuration. In fact any unplanned shutdown of these supercritical equipment even for a small duration can reduce plant availability for more than a day as plant start-up, stabilization and on-spec product routing takes considerable time and man-hour extra efforts.

The unplanned shutdown and subsequent start-up also raises safety risk in-addition to extra maintenance cost. A major part of most organizations’ maintenance budget is spent on annual or bi-annual inspections for failure analysis of rotating equipment. However, because these inspections are not conducted in real-time, the operator lacks control over unplanned downtime.

Live Solution Deployment

Industry: Oil and Gas, Power, Metals & Mining

Customer: A Fortune 500 company with a revenue of USD 55 Billion

Business Issue: A single unplanned showdown due to failure of a rotary equipment can cost a refinery USD 500K

Impact Created: 8’s PlantBrain was able to predict the failure of MAB with lead time of upto 70 days

Premise: Fluidised Catalytic cracking unit( FCCU) in a refinery plays a very important role in conversion of a low value hydrocarbon to high value Petrochemical feedstock like Propylene, Liquified Petroleum Gas (LPG), High Octane Gasoline and Diesel components. FCCU is called the heart of the refinery and its importance has increased as with state of Art technology, it is capable of maximizing Petrochemical feedstock and thereby the profitability of the refinery.

Critical Equipment: Reactor-Regenerator section coupled with Main Air Blower (MAB) to supply Air for continuous Catalyst Regeneration is the most important section in FCCU. The health of MAB is important for steady state operations and maximization of unit through-put. As most of the places, MAB does not have provision of standby, this rotary equipment is put under supercritical equipment category and thus the highest priority is accorded for regular monitoring of health. The key parameters to monitor may be the axial and radial vibration trend, bearing temperatures, process conditions including flow, temperatures and humidity etc.

Current Technology: The current process runs on the periodic visibility reports provided by some OEMs with respect to the performance of the MAB. The asset management system relies on data from 20 such equipment with no single source of truth, leading to a lack of a comprehensive view of the overall system.

Approach & Way Forward: The advancement of IIOT platform and computation techniques has made it possible to monitor MAB health on a real time basis and even do advanced predictions of equipment failure to the tune of 70-90 days.

Algo8, one of the leading Industrial AI company in India, leverages vast stores of floor-level OT data collected in process-driven industries (in addition to existing IT data) and other internal & external data, to derive unique insights using machine an deep learning that will be key not only to predict asset-failure events in advance, but also help operators take proactive decisions and corrective measures around asset health and maintenance. This has been successfully deployed for a MAB - Main Air Blower in a Fortune 500 company and has reduced maintenance costs by 25% and increased equipment availability by 10%.

The successful deployment of Artificial Intelligence (AI) and machine learning (ML) techniques by companies like Algo8 has made it possible to avoid unplanned shutdown of critical equipment in asset heavy industries, thereby improving the plant availability and profitability through proactively predicting failures to analyze and optimize maintenance. This will eventually lead to improved plant economics through::

● Reduced Operation and Maintenance (O&M) spend from unscheduled maintenance downtime
● Improved operational efficiency
● Minimized production losses.

Algo8 is an industrial AI company that learns from manufacturing big data and empowers all plant stakeholders to run their operations in a profitable and sustainable way.

About Nandan Mishra: Nandan is the Founder of Algo8. The company has 30+ clients including 9 Fortune 500 Companies. He is an alumnus of Indian Institute of Technology, Kanpur- Btech/Mtech in Chemical Engineering & Advanced Analytics. He is a key member of the Working Group for the National Mission of Interdisciplinary Cyber Physical system, set up to function like the think-tank for this mission under the Department of Science & Technology.