From waste to watts: The commercial opportunity from methane emissions

If you're in oil and gas, there's a good chance you're sitting on revenue you don't know you're losing. We sat down with Bill Barna, Principal Architect, Google Cloud, and Raiford Smith, Global Market Lead for Power & Energy, Google Cloud, to find out how remote sensing, AI, and operations research are turning one of the industry's oldest inefficiencies into a compelling commercial opportunity.

Bill has spent 15 years in oil and gas, in the field in West Texas, at U.S. Diplomatic missions in Saudi Arabia, Nigeria, and Colombia, and at Google Cloud, where he advises energy companies on methane emissions solutions. Raiford leads Google Cloud's Power and Energy vertical and brings over 30 years of industry experience, including years overseeing the energy procurement strategy behind Google's global data center infrastructure.

Together, they make the case that addressing one of the industry's biggest waste problems goes beyond sustainability footprints and regulatory compliance. It can deliver meaningful financial returns, too.

Q: Let's start with the basics. Why are methane emissions such a problem? And why are oil and gas companies interested in addressing it now?

Bill: There are approximately a million producing wells across the U.S., and a lot of this infrastructure is ancient. I’ve worked on wells in West Texas that were drilled in the 1920s. Older equipment was designed to produce oil, not to prevent methane emissions. Equipment like pneumatic controllers release methane by design to regulate pressure. Nobody engineered these process emissions out of the system because nobody had a reason to. Methane also enters the atmosphere through fugitive emissions, or leaks. 

Things have changed over time. First, natural gas is more important than ever to our economy. Also, public awareness of the harm from methane emissions has shifted and regulations have tightened. In fact, European LNG buyers are starting to build emissions standards directly into their supply contracts. And finally, the technology to find and quantify methane emissions has made the economics of addressing it very different than even five years ago.

Raiford: The demand side has also shifted over the last few years. No data center powers itself. They all rely on significant energy assets, and Google and the other hyperscalers buy responsibly-sourced energy. We want energy to be as reliable, affordable, and sustainable as it can be, not just for us but for everyone. The methane currently vented or flared from wells represents wasted energy and an environmental hazard. I think the estimates of how much methane could be recovered are probably understated, because the measurement tools are only now catching up to the actual scale of the problem. Addressing this is a real opportunity to improve operational efficiency, lower costs, and improve the environment. 

Q: Methane emissions are invisible. How do you find them?

Bill: You use a stack of technologies to locate, identify, and visualize methane emissions: IoT sensors at the asset level, optical gas imaging (OGI) cameras, drones, fixed-wing aerial surveys, and satellites. Each has a role, depending on the scale and resolution you need. Most operators are already familiar with at least some of these. 

New technologies are making it easier and less expensive to address process and fugitive emissions. Remote sensing capabilities are improving rapidly while prices are coming down. New AI models are extracting methane plumes from legacy data sources. 

For example, Google Earth Engine comes pre-loaded with Sentinel-2 satellite imagery captured globally every three days. With AI, Sentinel-2 shortwave infrared imagery can identify methane plumes. Add wind data, and you can calculate flux rates. This approach cost-effectively augments the other remote sensing technologies. 

Q: Once you've detected a plume, what happens next?

Raiford: The next step is attribution: figuring out exactly which asset is causing the plume. We use computer vision to map the plume to a specific piece of equipment on the ground, with GPS coordinates. Then we match it to data in BigQuery, including asset operational data, maintenance history, equipment age, and service records. From there, Gemini Enterprise turns that data into a specific result: a 12-year-old compressor with a likely seal failure, step-by-step repair instructions for the field technician, and an automatic parts order through the ERP system. The time from detection to remediation drops dramatically, which saves operators money and supports continuous operations.

The scale of the waste is significant. Some estimates show that the methane emitted in the Permian Basin annually exceeds 500 BcF. That’s enough to produce more than 5 gigawatts of continuous electricity for a year, enough to power several AI data centers.

Q: How do you make the economics work — especially for smaller, independent operators?

Bill: If you’re a small independent operator, you may be leaving money on the table when process or fugitive emissions could be converted into products and brought to market. Most analysis is one problem at a time, and many operators still rely on Excel workbooks without a clear view of the highest-return interventions. Larger operators have access to more data and data science capabilities, though in some cases methane gets treated as a local issue rather than a portfolio-wide opportunity for capital investment. 

Google Cloud’s Gemini Enterprise can help any of these companies turn methane from a waste stream into a revenue stream. Vizier on Gemini Enterprise Agent Platform uses advanced optimization models to rank remediation opportunities by ROI across your entire portfolio of leases and your full asset base. The output is a prioritized, fundable plan you can take to your CFO. Google Cloud also has solutions to integrate insights into Copperleaf, SAP, or Maximo. The result is three concrete advantages: lower costs to find and fix fugitive emissions, lower costs to upgrade equipment that produces process emissions, and better financial models for deciding when to capture methane for pipelines instead of flaring or venting it.

Raiford: Lowering energy costs and sustainable energy sources are top of mind for everyone right now, which makes a strong case for improving domestic production. Recovering enough methane to power a gigawatt-scale data center is also enough to power over 500,000 homes. That’s new revenue at very low incremental cost, with environmental benefits on top. The same remote sensing, AI, and subsurface modeling capabilities extend beyond methane recovery. For example, sequestering emissions with these technologies also improves carbon dioxide capture rates, which helps make energy cleaner for everyone.

Bill: Google’s geospatial tools, AI capabilities, and public data sources can help companies identify potential revenue that’s currently hauled off as waste. Most operators pay to dispose of produced water without knowing what's in it, and that water may contain lithium or other valuable minerals. 

Q: Detection and optimization are one thing. But what about the actual work of capturing and redirecting those emissions? What role do cloud technology and AI play there?

Bill: The equipment for capturing and redirecting methane is readily available. Using AI, we can model redesigns of your operations to evaluate different configurations, run the economics, and identify the path to positive ROI. This is similar to our work helping Westinghouse optimize the construction of nuclear power plants with technologies like Vizier. Vizier improves the hyperparameter performance of operations research and machine learning models used for scheduling and financial analysis, replacing trial-and-error with rigorous optimization.

If you're evaluating whether to retrofit a compressor station or redesign a separator, we can help you build that business case at a speed and cost that wasn't possible before.

Q: Where does this go from here — for the industry and for the operators who move first?

Raiford: There’s a lot of untapped potential here. What I’m most excited about is the chance to help energy companies make energy safer, more reliable, affordable, and more sustainable. Google brings strong capabilities to this work through our foundational AI models, geospatial tools, and ongoing research. We also have a direct stake in the outcome as a major energy buyer, a technology provider, and a company committed to a clean energy transition. Done right, this approach delivers new revenue, better margins, safer operations, and a stronger environmental record at the same time.

Interested in learning more about Google Cloud’s solutions for the oil and gas sector? Check out:

https://cloud.google.com/solutions/energy/oil-gas