You can measure everything.
What is next?

Having been on both sides of the table, firstly as an operator and now as a technology provider in our team, Peter Evans has some extra insights into why the technological evolution is crucial to deliver mitigation. In 11 minutes, the conversation covers:

1. No room for continuous emissions monitoring systems. Space at offshore oil platforms was at a premium, so retrofitting continuous emissions monitoring systems would be a massive hurdle.
2. “Can we just test it in the cloud?” We used the data being collected, fed it into a predictive model and calculated the emissions.
3. Patterns at scale. When you scale predictive emissions monitoring systems over hundreds of systems, you start to see a bigger picture.
4. Where artificial intelligence enters the chat. Trained large language models to analyze combustion and emissions data and turn it into direct fuel savings and mitigation proposals.
5. Secure and traceable. We work within the operator’s data environment, using their rules and systems to keep their data protected.
6. People decide. These are advisory tools, not control systems.

When many of the oil majors began to call for direct action on methane, Peter’s focus was on offshore process emissions from flares and engines. Space on offshore platforms is scarce, so retrofitting continuous emissions monitoring systems would have been a major hurdle. Instead, a colleague suggested testing predictive emissions monitoring for methane in the cloud, using the data the facility was already streaming. According to Peter, it worked better: less could go wrong, and the data was more reliable.

Since then, we have scaled predictive emissions monitoring to hundreds of systems. At that scale, patterns emerge that show how to run equipment efficiently and within its optimal design window, and the cloud-based reports can be customized for everyone from executives to technicians.

The next step is artificial intelligence. We are developing two features that use trained large language models to analyze combustion and emissions data and turn it into fuel savings and mitigation proposals. Peter also addresses the energy use of data centers and digital security: the features work within the operator’s own data environment, and the results go back to the operator’s historians for traceable verification.

The features are advisory tools, not control systems. An executive sees which facility is struggling, a technician sees their own equipment in detail, and in both cases the people make the decisions. For Peter, the bigger idea is to see emissions data not only as a burden, but also as a resource.