DAQ-2024-010346

23 September 2013 Page 1

Summary of Findings from the Uintah Basin
Ozone Study:
Preliminary Update from 2013 Field Study 1

BACKGROUND
The Uintah Basin is a rural area of northeastern Utah where the majority of the state’s oil and
gas production occurs. Ozone concentrations in excess of the current national air quality
standard have been measured in the Basin during the winter. These high ozone levels are only
observed in the Basin during winter inversion periods when the ground is covered by snow;
ozone levels outside of these periods have remained below the air quality standard and
conditions resulting in exceedances of the standard do not occur every year.
In the first quarter of 2012, a multi-phased study (the Uintah Basin Ozone Study, UBOS) was
begun to identify the emissions sources and the unique photochemical processes that cause
elevated winter ozone concentrations, and to identify the most effective strategies to reduce
winter ozone. UBOS 2012 included measurements of ozone and ozone precursor
concentrations and meteorological conditions throughout the Basin. Meteorological conditions
during UBOS 2012 were not conducive to ozone formation due to a lack of snow cover; no
exceedances of the 8-hour average 75 parts per billion (ppb) National Ambient Air Quality
Standard (NAAQS) were observed during UBOS 2012. Key findings from UBOS 2012 are
described in a summary report
(http://www.deq.utah.gov/locations/uintahbasin/2012study.htm).
UBOS 2013 took place between January and March 2013. In contrast to UBOS 2012, conditions
during UBOS 2013 were favorable to ozone formation and numerous exceedances of the
NAAQS were observed.
Presented here is an update to the interim findings from UBOS 2012 to reflect the considerable
additional information obtained during UBOS 2013. These findings are drawn from preliminary
analyses of data and results and will be updated in the coming months. A final report for UBOS
2013 is currently under preparation and planned for completion by December 2013. Additional
analysis and results will also be included in peer-reviewed papers to be published by
participating researchers.

1 Prepared by researchers and air quality managers at Utah State University, University of Utah, National Oceanic
and Atmospheric Administration, ENVIRON, University of Colorado, Utah Department of Environmental Quality
and EPA.
23 September 2013 Page 2

This important work is made possible by funding and in-kind support from the following:
Uintah Impact Mitigation Special Service District (UIMSSD), Western Energy Alliance, Bureau of
Land Management (BLM), National Oceanic and Atmospheric Administration (NOAA),
Environmental Protection Agency (EPA), Utah Department of Environmental Quality (UDEQ),
Utah Science Technology and Research Initiative (USTAR), and Utah School and Institutional
Trust Lands Administration (SITLA). Work on UBOS 2013 was conducted by researchers from
Utah State University (USU); National Oceanic and Atmospheric Administration (NOAA); Utah
Department of Environmental Quality (UDEQ); University of Colorado, Boulder (CU); University
of Wyoming (U of WY); and University of Washington (UW).
PRELIMINARY FINDINGS
Key findings from UBOS 2013 are summarized below:
I. Air Quality
a. Maximum 8-hour average ozone concentrations measured at Ouray, which
typically has the highest readings in the Basin, reached 142 ppb during the
December 2012 – March 2013 winter study, exceeding the EPA 8-hour standard
(75 ppb) by 89%. Exceedances of the standard (i.e., a daily maximum 8-hour
average in excess of 75 ppb) occurred at 17 of the 20 monitoring sites operating
during the study. Monitors in the major Basin population centers exceeded the
standard on a total of 22 days at Vernal and 29 days at Roosevelt during 2013
with all of the exceedances occurring between January 9th and March 6th. Note
that the occurrence of an exceedance does not by itself constitute a violation of
the EPA standard; a violation only occurs when the annual fourth highest daily
maximum 8-hour average, when averaged over three consecutive years, is
greater than 75 ppb. Exceedances occurred in the Basin during seven separate
multi-day ozone episodes which were separated by periods of lower ozone levels
coinciding with the passage of storm systems. These observations are in sharp
contrast to conditions during the 2011-2012 winter study, when daytime 8-hour
average ozone levels did not exceed 63 ppb.
b. Individual ozone episodes ranged from 3 to nearly 15 days in length, with ozone
concentrations generally increasing from one day to the next during each
episode, indicating a lack of ventilation.
c. Daytime ozone concentrations at locations close to but outside of the Uintah
Basin, at the time of high surface ozone events inside the basin, ranged from 40
to 60 ppb. Ozone concentrations measured inside the Basin above the
temperature inversion during these periods were also in the 40 to 60 ppb range.
These facts, combined with observations of winds around the Basin, confirms
that high surface ozone concentrations within the basin are not influenced to
23 September 2013 Page 3

any significant extent by transport of ozone or precursors from outside of the
Basin.
d. Days with high ozone in the Basin coincide with elevated levels of methane,
volatile organic compounds (VOC) and nitrogen oxides (NOx). Alkane
hydrocarbons constitute the main fraction of identified VOC. VOC and NOx are
the primary chemical precursors of ozone. A review of all emissions sources
within the Basin (based on WRAP Phase III and UDAQ triennial inventories)
indicates that activities associated with oil and gas exploration and production
are the predominant sources of ozone precursors.
e. VOC concentrations in the Basin are extremely high during ozone episodes due
to large amounts of VOC emissions and limited ventilation under a strong
temperature inversion. Average VOC levels at Horsepool in 2013 were
1,684 ppbC (based on C2 – C7 non-methane hydrocarbons) as compared to
232 ppbC in 2012 when inversions were not present and there was more mixing.
f. Vertical profiles show that the polluted air mass associated with ozone episodes
is confined to a shallow boundary layer that varies in height from 70 – 400 m
(230 – 1,300 ft) above ground level.
g. The Bonanza power plant plume does not appear to contribute any significant
amount of nitrogen oxides or other contaminants to the polluted boundary layer
during ozone episodes; the thermally buoyant Bonanza plume rises upwards
from the 183 m (600 ft) stack and penetrates through the temperature inversion
layer. As a result, emissions from the Bonanza plant are effectively isolated from
the boundary layer in which the high ozone concentrations occur.
II. Meteorology
a. Observations made during the 2013 winter study confirmed that high winter
ozone in the Uintah basin only occurs when the ground is covered with snow and
weather conditions promote the formation of a strong temperature inversion
which traps a layer of cold, stable air (a “cold pool”) within the basin. In the
absence of any snow cover, warming of the earth’s surface by the sun causes too
much convective mixing for a cold pool to form.
b. Chemical reactions resulting in ozone formation are driven by the illumination of
the atmosphere from direct, reflected and scattered ultraviolet solar radiation.
Reflection of light from the snow surface significantly increases the total flux of
ultraviolet radiation and thus the rate of ozone formation.
c. Ozone episodes in the Basin are characterized by complex, diurnally varying
patterns of light winds which have the potential to produce gradual but
significant intra-basin transport of ozone and precursors. Factors driving winds
within the polluted boundary layer include differential daytime heating
23 September 2013 Page 4

producing upslope flows (which can be modified by variations in snow cover),
nighttime drainage flows resulting in convergence along the river valleys at the
lowest Basin elevations, and spatial perturbations of the depth of the cold pool
air mass by winds blowing over the mountains surrounding the Basin. These
forcing factors appear to produce oscillations or an east-west “sloshing” of air
within the basin that further contributes to intra-basin mixing. These processes
result in transport of ozone and ozone precursors from one part of the Basin to
another but are not strong enough to mechanically breakdown the temperature
inversion.
d. Initial attempts at modeling meteorological conditions during episodes indicate
that obtaining accurate results with current modeling techniques will be very
challenging. Additional data collection and model development and testing will
be required to produce a good working model suitable for evaluating alternative
regulatory strategies.
III. Chemistry
a. Unique features of the chemical reactions involved in ozone production were
observed during the 2013 winter ozone episodes which make these episodes
very different from summer ozone episodes in urban areas. Nitrous acid (HONO)
and formaldehyde rather than ozone photolysis were found to be the biggest
contributor to the pool of chemical radicals responsible for ozone formation.
However, the contribution of HONO to the radical pool is uncertain at this time
due to potential chemical interferences with the HONO measurements
performed during the 2013 study. This in turn results in a significant amount of
uncertainty in modeling the chemistry of ozone formation.
b. Uncertainties remain regarding the likely impact of nitrogen oxides (NOx) and, to
a lesser extent, volatile organic compound (VOC) emission controls on ozone
levels:
i. Analysis of data from 2012 (when elevated ozone did not occur) suggests
that ozone formation was VOC limited, i.e., VOC reductions would
produce ozone reductions whereas marginal NOx reductions may result in
increases in ozone due to a lack of sufficient radicals to process the
available NOx. Further analysis of conditions during 2013 is needed to
determine if this was also true during the 2013 episodes or if instead the
radical pool was sufficiently increased to the point where NOx controls
start to become an effective ozone reduction strategy.
ii. The effectiveness of NOx controls will also be impacted by the extent to
which unreactive NOz species (including HNO3 and organic nitrates) are
being recycled back into reactive NOx by heterogeneous chemistry in
23 September 2013 Page 5

snow and on particulates. Details of these potential heterogeneous
reactions are not yet understood well enough to be included in
photochemical models needed for evaluation of control strategies.
iii. Measurements made during 2013 suggest that VOC reactions in the snow
may contribute to ozone chemistry within the layer of air just above the
snow surface (the mixed layer). The potential implications of this finding
on the efficacy of VOC and NOx controls are not yet known and further
study is needed.
iv. Oxidation of aromatic VOCs (including toluene and xylene) is an
important source of radicals. Thus, VOC control measures focused on
sources of these species will be particularly effective. There do not
appear to be large sources of other highly reactive VOCs (alkenes) in the
Basin as alkene concentrations were very low relative to alkanes and
aromatics.
IV. Emissions
a. Some important progress has been made on developing emission inventories for
the Basin but the available data remain incomplete. Several groups, including
the Utah Division of Air Quality, EPA Region VIII, and Federal Land Managers
(through the Utah Air Resource Management Strategy project and the “3-States”
and “West Jump” air quality modeling studies) are currently working on
inventories for the Basin. Progress would be enhanced by more formal
coordination of effort between these groups and the energy producers.
b. Field measurements of emissions from produced water ponds were conducted
during the winter of 2012-2013 which showed that, while wintertime pond
emissions comprise a small part of total VOC emissions in the Uintah Basin, VOC
emitted from produced water ponds tend to be enriched in reactive VOC.
c. Field measurements made by the NOAA mobile laboratory during UBOS 2012 are
being further analyzed to extract additional information about emission
characteristics of sources sampled.
d. Obtaining sufficient emissions data for the Basin is made difficult in part due to a
mixture of private, federal, state and Indian surface and mineral rights
ownership which results in complex regulatory jurisdictions and inconsistent
reporting requirements.
e. Inventory data specific to the winter ozone season are needed as emissions from
some types of sources (such as evaporation ponds and methanol use) have large
seasonal variations.
23 September 2013 Page 6

f. Additional information is needed on methanol use and composition, especially
the degree of contamination with formaldehyde, which is a highly reactive
volatile organic compound (VOC) that accelerates the formation of ozone.
g. Important components of the inventory that need to be more carefully
quantified include emissions of VOCs from oil, condensate and produced water
storage tanks, as well as fugitive emissions from leaking components and other
sources and from venting, blowdowns and other intermittent events.
V. Recommendations for Future Analyses
a. Additional field data collection and analysis are needed to improve our
understanding of several key features of the winter episodes to the point where
meteorological and photochemical models suitable for analyzing alternative
regulatory scenarios can be developed. Two areas of particular importance were
identified:
i. The highest priority studies would focus on improved, more robust
measurements of HONO concentrations and related species, their vertical
distribution and temporal variations. The response of ozone to changes
in VOC and NOx emissions is highly sensitive to HONO, so a clear
understanding of HONO formation and removal mechanisms is critical.
ii. Development of a comprehensive and accurate winter emission
inventory for the Basin suitable for modeling applications (i.e., including
sufficient speciation of chemical compounds and spatial and temporal
detail) remains a critical requirement. There is a need to prioritize and
coordinate efforts among the various groups that are currently
developing inventories. In addition, work on “top-down” evaluation of
inventories using available air quality observations is needed.
b. Several other data collection and analysis needs were identified by the UBOS
research team; these will be detailed in the UBOS 2013 synthesis report.
VI. Implications for Control Strategies
a. Ozone levels in excess of the EPA 8-hour standard are only observed during
winter inversion periods; ozone levels on the majority of days during the year are
below the EPA standard and exceedances of the standard do not occur every
year. This suggests that episodic or seasonal controls may be a useful
component of an overall air quality management strategy for the Basin.
b. Emission reductions at the Bonanza power plant are unlikely to have any effect
on winter ozone episodes.
c. Reductions in emissions of VOC will be beneficial, especially reductions in highly
reactive VOC species, such as formaldehyde and aromatic VOC, that are sources
23 September 2013 Page 7

of radicals. Glycol dehydrators and produced water are two important sources of
aromatic VOC.
d. Ozone response to NOx reductions is less certain. NOx control strategies will
tend to be less effective if unreactive NOz is converted to reactive NOx by
heterogeneous chemistry on snow or particulates, e.g., the conversion of HNO3
to HONO in snow.
e. Reducing formaldehyde would be an effective way to reduce ozone but it is not
clear at this time which sources of formaldehyde (direct emissions from fuel
combustion and use of methanol contaminated with formaldehyde or secondary
formation of formaldehyde from VOC precursors) are most important.
f. Uncertainty in HONO sources and source strength (direct emissions or secondary
formation from nitrogen precursors via several potential reaction pathways)
makes it difficult to predict how responsive ozone will be to reductions in VOC
and NOx emissions.