All Climate Events

SUPER EMITTERS IN NATURAL GAS EMISSIONS AND THE FUTURE OF EMISSIONS MONITORING

Download as iCal file

Friday, 20 April 2018, 12:00

Friday, April 20, 2018.  12:00 pm. SUPER EMITTERS IN NATURAL GAS EMISSIONS AND THE FUTURE OF EMISSIONS MONITORING.  Dana Caulton, Princeton University.  Sponsored by Department of Environmental Sciences. Candidate for Atmospheric Chemistry position. Refreshments at 11:45 am.

ABSTRACT:

The Marcellus Shale Play is the highest producing shale basin in the U.S. Recent work suggests the distribution of methane emissions from natural gas ‘fracking’ operations is highly skewed with a small percentage of well pads contributing a large proportion of the total emissions. However, existing field observations of well pad emissions are limited to small numbers of samples, thereby complicating the representativeness of the observations. Further, experimental methods that are more suitable for observing larger sample sizes are also prone to significant uncertainties in their derived emissions. In order to determine a statistically robust distribution of emission rates, errors in current techniques need to be constrained and mitigated. The Marcellus shale has received less focus for well pad emissions than other basins and presents unique challenges relating to topography and collocation of other sources that interfere with the petrochemical emission signal.

 

Over the past two years I designed and conducted a field experiment to measure emission rates from well pads in the Marcellus.  Over 700 unique well pads were sampled accounting for ~ 20% of the total well pad population. Multiple field measurements (mobile transects, stationary towers, replicate sampling) and modeling studies (inverse Gaussian, large eddy simulation) were integrated into the experiment to quantify emission rate uncertainties. The findings indicate that the top 10% of sites are responsible for ~80% of emissions. Because of the large sample size, data pertaining to the timescale of emissions and the predictability of emissions can also be examined. Multiple regression and machine learning techniques were used to identify commonalities of super emitters among traits such as production rate, operator size, violation history and location. Overall, natural gas ‘super emitters’ are not predictable in space or time, suggesting that monitoring strategies need to be frequent and widespread but not with particularly sensitive approaches. The episodic nature of super emitters are likely to be present in other emission monitoring areas of interest, such as urban and agricultural sources. This work allows the development of improved emission monitoring techniques and protocols that will be critical to understand how fossil fuel, urban and agricultural emissions are changing and can be mitigated.

Location  Room 205, Institute for Food, Nutrition and Health, Room, 61 Dudley Road, Cook Campus