In 2014 the oil and gas industry overtook cow farts to become the leading source of methane emissions. These leaks are unintended, cost the industry millions in lost production, and a very bad for our planet. Unburned, methane a potent green house gas, some 80 times the impact of CO2.
Today, the only method for detecting these leaks is to drive to each wellsite with an infrared camera and physically inspect each location. The FLIR IR cameras cost over $50,000 and sending a person 30 miles down a dirt road to look at a remote location that might be leaking is incredibly inefficient. To make matters worse, many of these leaks are intermittent.
There are over 800,000 producing oil and gas wells in the US alone. Clearly this is a problem that required an innovative sensing solution. The challenge is that methane is slightly lighter than air. Detecting a methane leak with a ground level sensor is next to impossible.
Not only do these 'fugitive gas emissions' impact our planet, there is real money at stake. In 2015 Noble Energy entered into a $74M settlement with the EPA acknowledging the unintended emissions from its wells in Colorado.
During the 2017 Spring Semester at the University of Denver, a group of innovative students teamed up with a successful entrepreneur, David Armitage to tackle this challenge; come up with a low cost, highly scalable system to continuously monitor remote locations for fugitive gas emissions.
It was a daunting challenge.
Dozens of large corporations have research and development teams with multi-million dollar budgets working on the 'methane problem'. And they have been working on it for years.
Our first thought was to build a drone, hoist a sensor aloft, and record GPS location and methane concentrations. These oil and gas fields are VAST, so we would need a long flight time. And the best sensor we could find was a core from FLIR, the folks that build the fancy (expensive) camera.
Costs started to get crazy (>$100K for the plane and the sensor) and we weren't too excited about the idea of having to manually conduct arial inspections of all over Texas, North Dakota, and Colorado (etc.) every few weeks. Furthermore, the drone approach didn't achieve an important one of our objectives...low cost continuous monitoring at scale.
During a site visit to a gas well late in the semester, one of the students asked a couple of questions that turned everyone around and stopped them in their tracks. "Why does this well stink". He was right. Most wells we had visited didn't smell, but this one stunk like a can of kerosene on a hot summer day. We all knew that methane was odorless and as I mentioned, lighter than air. We shouldn't be smelling a leak, but we were. And then we all realized that what we were detecting was not methane but the associated Volatile Organic Compounds (VOC's). That was it! Our 'eureka' moment. We had been trying to come up with a direct detection method when what we need to do instead was to focus on the associated gases. Natural gas is made of a 'soup' of hydrocarbon compounds. Methane may be the most significant constituent, but there are lots of other heavier than air VOC's in the goulash.
Within weeks we had found a low cost, highly stable VOC sensor capable of reading down to parts per billion concentration levels. It is worth noting that in our observations, the fugitive gas leaks tend to be at very low concentration levels. We laid out boards which enabled us to integrate this with an LTE-enabled microprocessor (thank you SEEED for the quick turn-around).
Many of these remote sites have no line power, and we wanted our installation to be fast and inexpensive. We needed an enclosure that could house a small solar panel. We decided to use a blue bird house which provided just the right roof pitch for the solar, gave us a reasonably well protected place to mount the sensor, and would provide habitat. Project Canary was born.
(BTW, the VOC concentration levels are exceeding low and do not pose a risk to local habitat. Any oil and gas wells that have serious, large scale, persistent leaks are quickly remediated.)
In late November 2017 we installed a number of these around a well in Northeast Colorado. Every 15 minutes they wake up, take a reading, report to our Cloud servers, and go back to sleep. We worked hard on power management. During the winter we went through many snowy days and long nights, and we never missed a single report. Each data payload includes VOC concentrations, temperature, humidity, signal quality, and battery level.
The results have been phenomenal. You can review live data here. We have detected numerous, intermittent leaks lasting from minutes to hours. While we don't have intimate knowledge of the equipment on this particular well site, it appears that periodically the vapor recovery systems which are designed to capture an unintended emissions become overwhelmed. These systems which are supposed to either recapture or flare the gas cannot keep up, and it is instead released into the atmosphere. In this particular case, these periodic emissions are not large or persistent, but they do warrant inspection and remediation. The operator of this particular facility has taken steps to mitigate the leaks and recently we have noticed a substantial reduction.
What is important to note is that these leaks were not persistent. Had an air quality inspector driven up with a FLIR camera they likely would have missed one of these events and reported this well to be in compliance. In this particular case, the neighbors had been complaining for years about this well location emitting odors, but each time it passed it's periodic evaluation. Having hundreds or thousands of field inspectors driving around randomly looking for leaks is NOT going to solve the problem of fugitive gas emissions from these site. Project Canary is a scalable, low cost, and practical approach to reduce unintended natural gas emissions, and reducing methane is critical for the health of our planet.
Comments