Measuring the Emission, not the Leak
This may strike you as a strange title for an article on the monitoring of fugitive emissions but bear with me as I try to explain what I mean, and why its central to the fugitive emissions work we’re doing here at Sedolabs.
Traditionally, leak detection and repair programs, together with the regulatory frameworks that run alongside them, are concerned with capturing point-source concentrations readings around a leak and generalising a picture of the emissions situation at a facility. In order words, measurements are taken at the component, micro level and, using some constructs, a macro level picture regarding the state of an entire facility is established. These mechanisms have served the industry remarkably well; factoring algorithms developed many years ago have held up well when dealing with the estimation of overall emissions at oil and gas facilities.
Back in 2016, we decided to look at some alternative approaches. It struck us that, rather than looking at point source concentration readings as a way of estimating mass emission rates, it would be worth trying to model the emissions plume itself. Of course, an emissions plume is not a static object. It is a dynamic, three-dimensional mass of gas within a gas; a directional column of methane supported within the medium of the surrounding atmosphere. Taking a static snapshot of an emissions plume is akin to taking a photo of a horse race and trying to predict the outcome; like the horse race, the emission plume exists in the time domain. It is only when you take into account enough time bound information about the plume that you have enough data to make predictions and quantifications.
Once you have an accurate model of the emissions plume, you can ‘ask’ it questions. You can analyse it and take measurements from it. You can ‘look into’ the future and make predictions about its scale and about how it will evolve.
Where, before, you might have taken a single concentration reading from around the leak point, using a relatively inaccurate ‘sniffer’ probe, we can now model an entire emission plume based on what we ‘see’ in the field.
Of course, the model we create is only as good as the data on which we have based it. The more data, and the more accurate the data, we can capture about the emission, the more reliable as model we can create. The key is to pull in data from multiple sources, and data taken at multiple points in time, and to combine this data with accurate environmental data taken from the location of the emission.
One reason that, traditionally, point source concentration readings have been the source data for emissions monitoring campaigns is that they are quick and simple to take in, what can be, demanding circumstances. Surveys tend to be done at intervals, such as annually or biannually. The weather can be unpredictable. Locations can be remote. What was needed was a way of continually monitoring a facility without human involvement, reliably and at low cost. Sedolab’s response to this requirement is the SEMS Static Monitoring Solution, an in-situ system that can be used to monitor a facility, or the assets within a facility, constantly and from anywhere on earth. Aside from all the cost-saving benefits of a static monitoring solution, perhaps the greatest scientific benefit is that measurements can be taken continually, not just as snapshots. For the first time, operators can view a dynamic emissions profile for a facility, with the time domain considered. As a result, the emissions quantifications, predictions and alerting produced are more reliable, more accurate and are gathered in an automated manner.
Looking again at the title of this article, you can hopefully see what I was alluding to; we used to find a leak, ‘sniff’ its concentration and then use factoring tables to estimate the size of the resulting emission. Now, we’re modelling the emission plume itself by observing its characteristics over time, and from the model we create we can take measurements and make predictions about its quantification and future behaviours. In essence, we’re measuring the emission, rather than the leak.