U. Minnesota team develops Lagrangian technique to identify NOx hotspots; opportunity for connected vehicles and big data
Researchers at the University of Minnesota have introduced a new method for identifying NOx emissions hotspots using high-fidelity Lagrangian vehicle data to explore spatial interactions that may influence emissions production.
Their study, published in the ACS journal Environmental Science & Technology, finds that the two transit buses under study—a conventional powertrain transit bus and a series electric hybrid bus—emit higher than regulated emissions because on-route operation does not accurately represent the range of engine operation tested according to regulatory standards. Using Lagrangian hotspot detection, they demonstrated that NOx hotspots occurred at bus stops, during cold starts, on inclines, and for accelerations.
On the selected routes, bus stops resulted in 3.3 times the route-averaged emissions factor in grams/km without significant dependence on bus type or climate. The buses also emitted 2.3 times the route-averaged NOx emissions factor at the beginning of each route due to cold selective catalytic reduction aftertreatment temperature.
Despite stringent government regulations, studies have shown that NOx emissions observed for in-use vehicles are far higher than mandated. A study by Wu et al. showed that real world NOx emissions from Euro IV certified heavy-duty buses using SCR exhaust aftertreatment systems were substantially higher than the certified limit suggesting that the certification cycle does not accurately represent real-world conditions. Andreae et al. found that poor vehicle integration and misalignment of engine and vehicle test cycles can be significant contributors to the emissions discrepancy. Low-load driving at slow speeds can also cause elevated NOx emissions in heavy-duty diesel engines equipped with selective catalytic reduction (SCR) devices for reducing NOx emissions due to low temperature inactivity of SCR catalysts.
Where much research has explored engine-operating conditions that lead to high NOx emissions, aligning those emissions with spatial information can indicate at-risk geographic areas. Further, identifying spatiotemporal influences on production of these emissions can provide researchers and vehicle manufacturers with a better understanding on where to focus their emissions reduction efforts and provide regulators with data for improved standards.—Kotz et al.
Regulatory agencies and researchers commonly map spatial emissions using stationary or movable ambient air monitoring stations. Such spatiotemporal data sets are used to calibrate vehicle emissions and travel demand models.
|Eulerian and Lagrangian methods|
|There are two basic methods of describing a fluid flow: Eulerian and Lagrangian.|
|The Eulerian method observes fluid flow from a fixed position; the Lagrangian method observes the trajectories of specific fluid parcels.|
Looking at the problem as one of fluid flow analysis, the Minnesota team said, such analyses can be viewed as Eulerian—they measure emissions of passing vehicles from a stationary frame of reference to give a sense of the emissions from traffic flowing by.
However, they argue, Eulerian approaches do not have the resolution to determine individual contributions by vehicles in traffic flow and therefore cannot be used to accurately determine acute health effects from highly localized sources of emissions from vehicles.
Conversely, Lagrangian data taken from the vehicle frame of reference can more precisely indicate the spatiotemporal distribution of mobile emissions sources.… In this work, we show how Lagrangian spatial emissions mapping techniques using data recorded from manufacturer-installed tailpipe NOx sensors in two transit buses can identify systematic and physical causes for elevated in-use emissions from hybrid and conventional powertrain transit buses over a gradient of route and ambient temperature. Performed on a large data set collected over three seasons, we demonstrate spatial influences on vehicle emissions and develop a method for spatial vehicle emissions analysis that could be applied to larger datasets.—Kotz et.al
The researchers used two MY 2013, 2010 EPA-compliant transit buses. The first bus was a Gillig 40-foot, low-floor transit bus with a conventional Cummins ISL 8.9L diesel engine and an adaptive ZF Ecolife automatic transmission. The second was a New Flyer Xcelsior 40-foot, low-floor transit bus powered by a Cummins ISB 6.7L diesel engine with a BAE HybriDrive series hybrid drivetrain and selectively enabled engine start−stop technology.
They measured NOx emissions by accessing data from sensors associated with the SCR system through the SAE J1939 heavy-duty controller area network (CAN). These sensors are necessary for proper exhaust aftertreatment operation—NOx input and output measurements direct the system’s urea injection for the catalytic reduction process.
They manually identified Lagrangian NOxhotspots by plotting emitted NOx as a function of location using internally developed software and the Google Earth mapping program. For the study, they used an emissions factor in grams of NOx per kilometer; however, they noted, the Lagrangian hotspot detection technique can be applied to any desired vehicle attribute.
They found that bus stops, cold starts, inclines, and accelerations had the most noticeable impact on elevated NOx emissions for the tested routes.
The data techniques applied here have further implications for manufacturers and policy makers. With ever increasing vehicle connectivity, manufacturers could improve vehicle performance by providing real-time suggestions for augmenting operation based upon spatial hotspot information. While results shown…indicate engine start−stop had little impact on overall route NOx emissions, hypothetically engine start−stop technology could be used to limit human exposure when a series hybrid transit bus enters a known high NOx location in proximity to a bus stop or location of high population density. For instances of cold start with anticipation of transient driving based on traffic conditions…the bus could preheat its aftertreatment system to allow for effective NOx reduction. From a regulatory standpoint, spatial analysis could assist with emissions based taxing for certain areas through a practice known as Geo-Fencing, whereby different rules are set based upon geographic location.
Connected vehicles with enhanced sensor quality, low data storage costs and improved computing power will enable Lagrangian emissions analyses to be performed at a larger scale and with ever increasing accuracy. … The sheer size of data possible as more vehicles are connected presents a significant challenge to extending Lagrangian analyses. Spatial big data techniques such as those employed in routing applications where directions are calculated based on previous journeys and real time traffic information will be required. We have publically advocated for future vehicle research to focus on spatial big data techniques for better understanding connected vehicle emissions from the Lagrangian frame of reference. Using these new tools, we can effectively bridge the critical gap between Eulerian monitoring and vehicle emissions modeling to accurately quantify emissions from in-use vehicles at high spatial resolution.—Kotz et al.
Andrew J. Kotz, David B. Kittelson, and William F. Northrop (2016) “Lagrangian Hotspots of In-Use NOx Emissions from Transit Buses” Environmental Science & Technology doi: 10.1021/acs.est.6b00550