Predictive Maintenance Reliability

Predict The Future To Improve Operations

EP Editorial Staff | May 10, 2020

TGI linked Emerson’s OpenEnterprise SCADA system with PipelineManager simulation software to collect data in real time and create a digital replica of the company’s entire pipeline network.

Digital-twin technology allows natural-gas provider to react to problems before they begin.

Natural-gas providers rely on operators who can make fast, informed decisions to serve a variety of customers with constantly changing needs. However, as the size and complexity of the pipeline network increases and several variables influence their objectives of reliability, safety, quick response, accuracy, and regulation, these companies have begun looking to digital-transformation strategies to provide the fast, thorough information that their businesses need.

Transportadora de Gas Internacional (TGI) is the largest natural gas transporter in Colombia, with more than 25 years in the business. The organization has seen firsthand the value of implementing digital solutions to upskill and support decision making. For years, TGI engineers had been using complex spreadsheets to handle day-to-day changes and disaster planning. Responses to issues such as shut down of fields, shut down of compressor plants, pipeline maintenance, and pipeline leaks or ruptures all had their impacts calculated on complicated spreadsheets. This solution allowed the reliability team to gain visibility of its assets and manage the complications of running a pipeline, but it was difficult to teach and cumbersome to maintain for even the most experienced employees. Relying on these paper records made it particularly difficult to identify where a problem occurred whenever a delivery discrepancy was detected.

To better align analytics and operations, the reliability team developed a comprehensive prescriptive- and predictive-maintenance solution to digitize operations across the organization. At the heart of this solution is a powerful digital-twin simulation of TGI’s entire operation. With 24/7 access to advanced simulation technology, operations and maintenance now can not only closely monitor the day-to-day status of equipment and processes, but also peer into the future to improve the ways they prepare for and react to change.

An Emerson digital-twin system has replaced spreadsheets and allowed TGI reliability professionals to better understand current and future conditions.

Why change?

TGI safely maintains an extensive network of 4,000 km of gas pipelines in Colombia with a compression capacity of more than 193 hp. These pipelines deliver gas to many different sectors of the economy including residential, commercial, industrial, and power generation, so the company has a strict social and political responsibility to meet contracts successfully.

The pipeline operating environment is extremely dynamic. Maintenance activities, plant and pipeline shutdowns, and variation in client consumption all mean that company personnel must be able to handle daily planned and unplanned operational conditions—a difficult task for even the most experienced technicians, made even more difficult when all operational data was recorded on paper and spreadsheets. The complexities of daily operations, and the importance of maintaining consistent results, regardless of changes in personnel, made TGI a perfect candidate for improved decision support through digitization.

To provide better operational control and to design a better safety net for equipment in high-consequence areas, the reliability team developed a real-time digital simulation. Working with Emerson engineers (St. Louis,, the team linked Emerson’s OpenEnterprise supervisory control and data acquisition (SCADA) system to PipelineManager simulation software. The simulation collects data in real time from the SCADA system to create a digital replica of TGI’s entire pipeline network.

Digital future view

Not only does the digital-twin simulation help to maintain and improve operations in real time, but the operations and reliability teams also use it to “see the future.” Every half hour, a “look-ahead” model runs at 10 to 50 times normal speed using current-state operating conditions to tell the operator what conditions will be in six hours. Either team can also run this simulation on demand. Operators no longer need to wait for incidents to happen. The high-speed model tells them the future results.

The operations team has set the look-ahead model to monitor critical conditions. If the team wants to push the system for any reason, conditional alerts are in place to tell them how many hours it will take to break boundary conditions such as losing a compressor or plant, losing an injection plant, or creating an imbalance somewhere in the network.

Within a few months of initial startup, the reliability team used its look-ahead model to determine the consequences to downstream plants and stations if a single production site were shut down. Engineers learned that when the production site was offline, suction pressure at specific compressor stations would decrease until it reached its minimum, at which point they would shut down. Using the look-ahead model, the team was able to determine the actual time it would take to reach the minimal suction pressure at the compressor stations and the time it would take to reach the minimal arrival pressure at a site. This allowed the reliability team to identify an exact window of opportunity for any necessary shutdowns to the production site. Should the site need to be shut down, maintenance crews will go in knowing exactly how much time they have available to perform their tasks, and the planning crew can schedule any maintenance activities around that timeframe.

In another instance, the team used the look-ahead model to see what would happen at customer delivery points when a compressor station was brought down for maintenance. Again, using the model, the team was able to identify the time it would take to reach minimal arrival pressure at customer sites, providing them with a window of time during which they could work on the compressor without creating any risk to contract obligations with those customers.

The simulation is also extremely useful for testing hypothetical scenarios for process improvement and disaster preparedness. Engineers regularly make simulated changes to pressure and flow or install new virtual equipment to see, in real time, exactly how those changes will affect transportation of gas elsewhere in the pipeline. Working with emergency crews, the reliability team also tests response plans, such as simulating an emergency shutdown on a compressor plant to see if low suction creates a problem for gas delivery. Responses to emergency situations can all be designed and tested in the simulated environment to help keep the organization safe, efficient, and competitive without any risk to current operations.

TGI also uses its simulation to closely monitor energy regulations from customers and national agencies. Colombian Regulatory Entity (CREG) guidelines require that pipeline operators communicate any restrictions or shutdowns in advance and must also report estimated gas composition. Any data that TGI needs can be exported at any time to be used in reports to management, the government, or clients.

The digital-twin “look-ahead” model provides the reliability team with an exact window of opportunity for any necessary shutdowns. This allows maintenance crews to know exactly how much time they have available to perform their tasks.

The future of prediction

Digital-transformation solutions work best when they seamlessly integrate with the operating environments they are designed to impact. The digital twin not only provides TGI with clear analysis of data that the reliability team relies upon to keep the pipeline running safely and efficiently, but also provides the organization a platform for future improvements to run systems even more efficiently.

Reliability team members are working to have the digital twin aggregate and analyze data to predict how corrosion, buckling, and fatigue will impact the company’s real-world assets. Engineers will use the data to predict maintenance cycles, prevent spills, further reduce downtime, and increase throughput.

The reliability team is also examining options to connect the digital twin to machine-learning algorithms that predict future performance based on historical data, sensor data, and the digital twin’s output. Engineers will be able to perform even more prognostics, health management, and predictive maintenance. From there, machine-learning algorithms will be able to connect with artificial intelligence systems to directly control the asset’s operations, allowing the pipeline network to respond almost instantaneously to changing conditions. EP

For more information, visit


  • Show More Videos

    Featured Video Play Icon
    Clear Advantages for Window Machine Maker

    Manufacturing custom windows for architects is both an exacting business and an extremely cost-sensitive undertaking. By automating their window fenestration equipment, DeMichele Group meets their customers’ price and precision goals in a way that makes designing each new machine quicker and easier. See how strategic alliances with key suppliers yield a system with superior productivity […]


View Comments

Sign up for insights, trends, & developments in
  • Machinery Solutions
  • Maintenance & Reliability Solutions
  • Energy Efficiency
Return to top