Why People-Intensive Gas Measurement Analysis Screams for Automation and AI

If you’re in the business of moving and measuring natural gas, you know just how challenging it is to monitor your pipeline system for measurement errors. Custody transfer metering stations are your cash register, only instead of a few point of sale registers like a typical grocery store has, you have hundreds or even thousands of meters to monitor. And with an industry average lost and unaccounted for rate of 0.4%, most teams are losing 4 out of every thousand dollars in their cash register.
Why is catching natural gas custody transfer measurement errors so hard? Simply put, information sprawl and SCADA data overload from the vast network of Industrial Internet of Things (IIoT) in the field, including myriad types of smart connected meters (multipath ultrasonic, Coriolis, etc.), flow computers, gas chromatographs, temperature and pressure sensors.
Savvy teams build dashboards to organize the chaos of monitoring so many data channels streaming data across pipeline systems. But even with the best visualizations, it’s still a hunt for a needle in a haystack with catastrophic errors often persisting for up to 60 days before being detected. And then the clock starts ticking on analyzing and resolving errors. The financial impact is substantial and the reputational damage from sending a prior period adjustment to a customer months later can be irreversible. At a time of fierce competition and rate battles, can you really afford to admit to customers that you undercharged, or worse overcharged them for moving their methane because your cash register was wrong?
Measurement teams don’t have to be told to improve throughput accuracy but they’ll also be the first to admit that it’s a people problem. Under strict capital discipline at a time of low fees and tariffs (not to mention gas prices that discourage transportation), measurement teams are overworked and underbudgeted. If finding and fixing errors is a function of how many eyes are looking at dashboards, then the natural method to improve measurement certainty and reduce LAUF is to start throwing people at the problem. But we don’t because of hiring freezes, arbitrary staff budgets, and an evaporating pool of qualified measurement professionals. Misinformed or not, Millennials and Gen Z aren’t looking for a career in hydrocarbons and so the next generation of measurement analysts looks for work elsewhere.
So if measurement accuracy is pegged to a fixed number of measurement staff, what happens when employees retire or hired away to another job? Your measurement certainty and LAUF increase as your measurement resources decrease. Not sustainable.
Some innovators seek to automate measurement analysis by building their own software to aggregate, analyze, and detect errors, perhaps with an eye towards AI for some breakthrough capability. DIY measurement analysis software takes years to build when the capability to improve accuracy is needed urgently NOW. And even if you could attract and train programmers and the data scientists needed to develop AI, developers can cost two or three times more than a measurement analyst, so DIY teams would be better to simply budget the same amount and hire more measurement staff. At least the performance improvement would be realized immediately rather than years into the future.
All of this is “most illogical” as Star Trek’s Spock might say. So, what’s the logical way? C-SMART has spent the past decade building the solution to today’s measurement analysis problem. Combining digital twins that compare your physical assets with digital specifications of their capability, NIST-traceable data, and a first of its kind AI analytics engine, C-SMART reaches into the digital haystack of SCADA diagnostics data and pulls out the needle where errors reside, in real time. Suddenly, your overloaded measurement staff can force multiply their workday to stop wasting time manually chasing down errors and put their skills to work on high value workflows. It’s a win-win for performance and the bottom line: improve LAUF by 80% and cut operations and maintenance costs by 30% by switching to exception based error monitoring.
If the measurement department were viewed as a profit center, management would throw every dollar their way to hire and increase output. By empowering measurement teams to improve accuracy and resolve errors in record time, C-SMART helps drop more revenue to your bottom line and make measurement more profitable.