By Emily Newton, Editor-in-Chief at Revolutionized

Professionals who specialise in natural gas processing increasingly rely on modelling to shape their decisions. A model is not always certain in its predictions, but it can substantially reduce mistakes that may otherwise occur due to a lack of data about future insights. Here are some compelling examples of applying technology to achieve greater confidence about what’s on the horizon.

Showing Future Demand Trends

One way that modelling can assist in the natural gas sector is by showing how demands may change over the long term. Knowing about those shifts can help company leaders ensure they have the processing capabilities to keep an organization resilient and ready to meet society’s demands.

For example, a natural gas model from McKinsey indicates that natural gas demand will keep growing in North America until 2025 before leveling off. One identified source of the prolonged need for natural gas centered on North American exports to Mexico. However, the researchers concluded that demand varies across North America by region and that no universal trends exist.

That variability shows why it’s so useful to create different models for certain places. It’s then easier to determine where natural gas processing will occur, helping decision-makers determine whether it’s time to invest in more infrastructure.

Improving Process Control

Natural gas processing involves a series of steps that all play vital roles in production. In 2020, more than 77 million customers relied on natural gas pipelines in the United States. Keeping processes consistently trouble-free is essential for meeting their needs.

One of the key processes for natural gas preparation involves removing contaminants by pumping gas into a scrubber column. This is a fast and efficient way to get rid of pollutants. It’s also necessary to separate hydrocarbon liquid gases, water, and non-hydrocarbon gases before the natural gas begins transit via a pipeline. Models created by digital twins can help plant workers optimize processes, target abnormalities, and overcome challenges.

A Norwegian natural gas plant operated by Shell utilises digital twin modeling at various stages of production. More specifically, the models can show the numerous variables associated with stable operations, as well as various scenarios that could negatively impact production if not planned for before they happen. Shell representatives also report that the digital twins facilitate more equipment uptime and minimise energy consumption.

Having a strong handle on process control typically causes numerous advantages. For example, a company leader could pinpoint bottlenecks, increase safety, and make a facility more resilient. Models can also motivate people in charge to address certain operational vulnerabilities before they cause significant issues.

Reducing Pipeline Leaks

Statistics show that pollution causes more than 100 million human fatalities every year. It comes from numerous sources, but pipeline leaks tend to attract national attention. Outdoor natural gas leaks can contribute to smog and worsening respiratory ailments, including asthma. They can also contaminate natural habitats, often causing long-term ripple effects in affected areas.

It’s in the best interests of parties operating natural gas pipelines to monitor them closely and respond promptly to issues. Doing that can reduce the effects on society while protecting companies’ bottom lines.

Using machine learning models can help operators identify leaks faster, which should mitigate the overall effects of those issues. In one case, researchers built a model of normal behavior that triggered alarms for any deviations. In some instances, it only needed five minutes to identify a leak after training.

The researchers concluded that using this approach to manage leaks could be an effective way to get reliable data and reduce false positives. Their system did not generate any during tests.

Enhancing the Drilling Phase of Natural Gas Processing

Applying models for process control can help company representatives make their operations as efficient and stable as possible. That’s important, but it’s also necessary to enhance natural gas extraction efforts. Models can facilitate that aim.

A team at Texas A&M University developed models to predict the amount of natural gas extracted from new wells at a particular site. More specifically, this approach examines the flow characteristics from older gas wells in the same location. People then get a clearer idea of what to expect from newer drilling activity.

The models typically used for such forecasting require complex computations and specialized tools to create. However, this new option allows such modeling to take place in spreadsheets.

Ruud Weijermars, a petroleum engineering professor at the university, explained, “In the oil and gas industry, professionals use sophisticated reservoir simulators to get a sense of how much hydrocarbons can be recovered from the layers below the Earth’s surface.”

Weijermars continued, “These simulations are very useful but extremely time-consuming and computationally intense. We can now do the same kind of predictions as these simulations in a spreadsheet environment, which is much faster, saving a lot of time and cost for shale operators, without loss of accuracy.”

Models Remove Much Uncertainty From Natural Gas Processing

These examples show how modeling in the natural gas sector can help professionals get prepared for what’s on the horizon. Advanced models can also minimize the guesswork sometimes associated with where to drill, whether to upgrade facility equipment, and other necessities.

It takes time and effort to optimize a model, though. As people think about using them, they must make sure they’re ready to devote sufficient resources to their actions. Doing that greatly increases the likelihood of worthwhile results.