How big data will soon tackle pollution from industrial sites near you

Beth Trask

One day in the not-so-distant future, oil refineries, chemical plants, natural gas wells and other industrial sites will be fully wired.

A network of sensors will detect a harmful pollutant the moment a leak occurs, and within minutes, a maintenance crew will be deployed to complete the repair. Government inspectors will maintain records of such pollution events and of company response times – and people living near industrial sites who signed up for alerts on their smart phones will rest easy, knowing it’s safe for their kids to play outside.

We can see this utopia on the horizon. Big data paired with sensors and strong science are changing how we detect and prevent pollution, and fast. Projects are already on the ground, propelling us toward that future.

Oil and gas companies test emerging technology

After four years of research and development, several cutting-edge – and affordable – technology solutions have emerged to help oil and gas companies rein in methane leaks. They point the way to those fully wired industrial sites coming our way.

Norway’s Equinor (formerly Statoil) and global energy giant Shell became the first energy producers to install a new solar-powered device in 2017 that continuously detects methane leaks. It was created by Quanta3, a Colorado-based startup that developed the technology through the Methane Detectors Challenge.

PG&E, meanwhile, installed a similar laser-based technology, developed by San-Francisco startup Acutect – another stand-out technology from the challenge – at a natural gas storage facility in northern California.

The key to bringing these and other innovations to scale is big data.

Data from 80K wells helped expose reporting gaps

State officials in Pennsylvania learned recently that pollution from oil and gas wells within its borders is likely five times higher than industry had reported. By crunching data from more than 80,000 active wells and pairing it with peer-reviewed research, we were able to show what was really going on.

The new estimates were based on recent pollution measurements conducted by Carnegie Mellon University researchers. Suspecting that the industry-reported data was off, they had used sensors and state-of-the-art mobile monitoring methods to gather accurate emissions data from oil and gas wells in the Marcellus Shale basin.

Our scientists then used this data to develop a new algorithm for estimating total emissions. Along with finding higher-than-reported methane emissions, they discovered that air toxics and smog-forming chemicals were nine times higher than what operators said.

A user-friendly platform now gives policymakers and citizens access to this massive trove of data, along with tools for identifying ways to cut emissions.

At least one other state is already using these tools and we’re planning to expand their use elsewhere.

Operators can miss 50 percent or more of what they emit.

By contrast, most well operators today still use a decidedly low-tech approach to report their emissions: They multiply each piece of equipment on their site by an outdated emissions rate that doesn’t account for malfunctions or human error.

It means they can miss 50 percent or more of what they emit and that the full extent of the pollution from most of America’s roughly one million oil and gas wells is unaccounted for.

So entrepreneurs are also looking to the next frontiers of science to fill in the blanks.

Satellites, data science will pinpoint pollution hotspots

A new “audacious” project to launch a satellite in early 2021, dedicated solely to tackling methane pollution, is our next step. The MethaneSAT will help companies and governments pinpoint the location and magnitude of emissions virtually anywhere on Earth.

The transmission of such data from space will eventually help nations build out those real-time monitoring networks and make them ubiquitous.

At the same time, ongoing advances in data science – especially predictive analytics and machine learning – will soon allow computers to make data-driven predictions, identifying likely pollution events before they even happen.

BP is already piloting such technologies in Wyoming in partnership with Silicon Valley’s Kelvin Inc.

This way, companies and governments will soon have the information they need – and the responsibility – to take action.Importantly, it will empower citizens to hold them all accountable. That is what democracy looks like and how our kids will soon breathe cleaner air. This is not a utopia; it’s our future.

Get innovation updates

We’ll send regular updates about developments in technology, science and the environment.

Thanks for subscribing to Innovation and the Environment.