Community Noise and Air Pollution Exposure During the Development

Subscriber access provided by ALBRIGHT COLLEGE

Energy and the Environment

Community Noise and Air Pollution Exposure During the Development of a Multi-Well Oil and Gas Pad William B. Allshouse, Lisa M. McKenzie, Kelsey Barton, Stephen Brindley, and John L. Adgate Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.9b00052 • Publication Date (Web): 28 May 2019 Downloaded from http://pubs.acs.org on June 1, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

Environmental Science & Technology

1

Community Noise and Air Pollution Exposure During the Development of a Multi-Well Oil

2

and Gas Pad

3

William B. Allshousea*, Lisa M. McKenziea, Kelsey Bartona, Stephen Brindleya, John L. Adgatea

4 5

a Department

6

University of Colorado Anschutz Medical Campus, Aurora, Colorado 80045 USA

of Environmental and Occupational Health, Colorado School of Public Health,

7 8 9 10 11 12 13 14

15

*

16

Occupational Health, Colorado School of Public Health, University of Colorado Anschutz

17

Medical Campus, 13001 E 17th Pl, Mail Stop B119, Aurora, Colorado 80045 USA. Telephone:

18

303.724.8835. E-mail: [email protected] ([email protected]

19

beginning July 1, 2019)

Corresponding Author: William B. Allshouse, Department of Environmental and

1 ACS Paragon Plus Environment


Page 2 of 26

20 21 22

Abstract

23 24

Unconventional oil and gas development (UOGD) in the United States is increasingly

25

being conducted on multi-well pads (MWPs) and in residential areas. We measured air pollution,

26

noise, and truck traffic during four distinct phases of UOGD: drilling, hydraulic fracturing,

27

flowback, and production. We monitored particulate matter (PM2.5), black carbon (BC), A-

28

weighted (dBA) and C-weighted (dBC) noise using real-time instruments on 1 and 5-minute

29

timescales, and truck traffic for 4-7 days per phase at a large 22-well pad sited in a residential

30

area of Weld County, Colorado. Hydraulic fracturing, which requires frequent truck trips to

31

move supplies and diesel engines to power the process, had the highest median air pollution

32

levels of PM2.5 and BC and experienced the greatest number of heavy trucks per hour compared

33

to other phases. Median air pollution was lowest during drilling at this MWP, possibly because

34

an electric drill rig was used. The equivalent continuous noise level (Leq) exceeded guidelines of

35

50 dBA and 65 dBC for A-weighted and C-weighted noise, respectively, during all development

36

phases. Our data show that these multiple stressors are present around the clock at these sites,

37

and this work provides baseline measurements on likely human exposure levels near similarly

38

sized MWPs. 2 ACS Paragon Plus Environment

Page 3 of 26

39 40


Introduction Extensive adoption of advanced petroleum extraction technologies, such as horizontal

41

drilling and hydraulic fracturing, provide an economically viable means of developing previously

42

inaccessible unconventional oil and natural gas resources (i.e., shale oil and gas).1 These

43

technologies have also allowed operators to concentrate unconventional oil and gas development

44

(UOGD) surface infrastructure (e.g., drill rigs, tanks, etc.) at increasingly large distances from

45

the target mineral resource by using below ground laterals up to 3 kilometers (km) (1.9 miles)

46

long at depths of up to 3 km (1.9 miles).2 As a consequence multiple wells are now being drilled

47

on one surface location, known as multi-well pads (MWPs).3 Larger MWPs can include 20-40

48

wells4-6 and each MWP can cover a surface area >40,000 square meters (m2).7

49

Development of MWPs involves four phases (drilling, hydraulic fracturing, flowback,

50

and production) some of which may occur concurrently.8 For example, new wells may be drilled

51

while existing wells are in production. Limited data on phase-specific emissions suggest that air

52

pollutant emissions vary by phase.9, 10 Diesel emissions from trucks and heavy equipment and

53

chemicals in drilling muds used during the drilling of the wells can increase air pollutant levels

54

in the drilling phase.11, 12 Similarly, diesel emissions from the numerous truck trips to transport

55

water, sand, and chemicals and engines used to power the process can increase air pollutant

56

levels during hydraulic fracturing.13 Following hydraulic fracturing, flowback involves the return

57

of injected liquids, produced water, and hydrocarbons to the surface. Flowback has been

58

associated with some of the largest air emissions of volatile organic compounds (VOCs).14

59

Shortly after the well is completed, it goes into production, where fugitive emissions from tanks

60

and other equipment are likely the dominant source of air pollutant emissions.15 Air pollutants

61

associated with UOGD have been suggested as the basis for health effects such as fatigue, 3 ACS Paragon Plus Environment


62

migraines, chronic rhinosinusitis, asthma, adverse birth outcomes, congenital heart defects, and

63

acute lymphocytic leukemia.16-27

64

In addition to air pollutants, drilling and the use of heavy equipment during all stages of

65

well development produces noise at levels that can affect health.11, 28 This includes both audible

66

A-weighted (measured in dBA) noise and low frequency sound pressure characterized as C-

67

weighted (measured in dBC) noise.29, 30 Approximately 37% of all complaints filed with the

68

Colorado Oil and Gas Conservation Commission (COGCC) in 2015 were related to noise.31

69

Health effects associated with A-weighted noise greater than 50 dBA include increased stress,

70

sleep disturbance, nausea, headaches, and cardiovascular disease.32-35 To date, there has been

71

little research on possible health effects from C-weighted noise, though guidance suggests that

72

sustained exposure be limited to under 60 dBC.36

73

Page 4 of 26

While MWPs offer efficiency of scale and fewer overall well pads, MWPs also increase

74

the intensity and duration of exposures to UOGD-related air pollutants and noise for nearby

75

residents due to the development of many wells in succession. As MWPs have increased in size

76

and intensity, many of these sites resemble industrial facilities. In Colorado and other states,

77

MWPs are typically not subject to the same zoning codes as other industrial facilities and it is not

78

uncommon for MWPs to be in the middle of residential areas and near schools.4, 5, 37 In Colorado

79

the current setback, i.e., the distance from which a new well can be sited from an existing home,

80

is 152.4 m (500 ft) in most areas and 304.8 m (1000 ft) in urban mitigation areas and established

81

well pads can be expanded without meeting the current setback requirements.38 In the state of

82

Colorado, there are at least 378,000 people within 1609 m (1 mi) of a well and this population is

83

growing at a faster rate than the overall population.39 As the Denver-Julesburg Basin in


Page 5 of 26


84

northeastern Colorado continues to experience rapid growth in both oil and gas development and

85

population, it is expected that more people will be living near MWPs.39-41

86

Relatively few studies have concurrently measured air pollution and noise levels near

87

MWPs, and even fewer characterize levels during specific phases of UOGD. To fill this

88

knowledge gap, we collected continuous data documenting particulate matter less than 2.5

89

microns in diameter (PM2.5), black carbon (BC), and A and C-weighted noise levels as well as

90

truck counts during each phase of the development of a MWP in a residential area of Weld

91

County, Colorado.

92 93

Materials and Methods

94

Site Information

95

We monitored PM2.5 and BC, as well as A-weighted and C-weighted noise at four

96

residences located between 217.9 m (715 ft) and 392.6 m (1288 ft) from the sound wall of a large

97

MWP during drilling operations from January 27 - April 24, 2017 (Northeast, East, South, and

98

Northwest monitoring sites). We continued to monitor these pollutants at the Northeast and

99

South (224.6 m (737 ft) and 264.6 m (868 ft) from the sound wall, respectively) residences from

100

July 24 - August 21, 2017, October 2 - November 20, 2017, and January 8 - March 2, 2018,

101

which corresponded to periods of hydraulic fracturing, flowback, and production activities on the

102

well pad, respectively. The development phases of drilling, hydraulic fracturing, and production

103

were later verified using information from nearby residents and data from the COGCC database.

104

This database does not provide dates for flowback, but this phase could be deduced given the

105

timing and activity on the pad and was verified by nearby residents. The Northeast monitoring 5 ACS Paragon Plus Environment


Page 6 of 26

106

site was located along the well pad access road and the three other monitoring sites were away

107

from roadways and thus less influenced by traffic. The MWP is permitted for 22 wells, 22 oil

108

tanks, 22 separators, four vapor recovery units, two water tanks, three modular large volume

109

tanks, and two lease automatic custody transfer units. All 22 wells were producing oil and

110

natural gas when we finished monitoring in March 2018. For this MWP, the operator used

111

electrically powered drilling rigs, constructed a 9.8 m (32 ft) high sound wall around the site, and

112

used “green completion” techniques as well as other noise, traffic, and pollution mitigation

113

measures.42 A sketch of the well pad site including the sound wall and pad access road and the

114

four monitoring locations is shown in Fig. 1.

115 116 117 118

Figure 1. Map of the 22-well pad under construction with a sound wall. The Northwest (NW), Northeast (NE), East (E), and South (S) sampling sites collected black carbon, PM2.5, A-weighted, and C-weighted noise during drilling operations. The NE and S locations collected these variables during hydraulic fracturing, flowback, and production. In addition, a weather station and camera were installed at the NE site to record wind speeds and trucks driving to the pad on the access road.

119 120 121

Monitoring 6 ACS Paragon Plus Environment

Page 7 of 26

122


DustTrak II (TSI Incorporated, Shoreview, MN, USA) instruments were used to measure

123

1-minute average PM2.5 concentration. The equipment was cleaned and calibrated according to

124

the manufacturer’s protocol prior to each of the four field monitoring campaigns that occurred

125

over the course of the MWP development. The DustTrak was zeroed and re-calibrated once a

126

week, during which time the impaction plate was also cleaned and re-oiled. Data was

127

downloaded approximately once every three weeks. Average PM2.5 concentrations were recorded

128

every minute in mg/m3 and converted to µg/m3. The limit of detection was 1 µg/m3. All

129

measurement data was visually inspected to ensure reliability. All PM2.5 measurements were

130

adjusted using a humidity correction factor43 based on the relative humidity levels measured at

131

our on-site weather station.

132

MicroAeth AE51 (AethLabs, San Francisco, CA, USA) aethalometers were used to

133

measure 5-minute average BC concentration during the drilling phase and changed to 1-minute

134

average BC concentration during flowback and production processes to match the time scale of

135

other pollution measurements. The microAeth was set to draw an air sample at a flow rate of 100

136

ml/min through a 3 mm diameter portion of filter media. The absorbance or attenuation of the

137

sensing spot is measured relative to an adjacent reference portion of the filter. Measurements

138

were converted to a mass concentration of BC and expressed in ng/m³ using the known optical

139

absorbance per unit mass of BC material. In accordance with manufacturer recommended

140

standard operating procedures, the device had its filter replaced twice-weekly and data

141

downloaded once every three weeks. After evaluating the raw data, the instrument output was

142

adjusted using an equation (Eq 1) from Kirchstetter and Novakov44 to adjust for accumulation of

143

BC on the filter substrate over time:

144

BCAdj=0.73*e-ΔATN/100+0.298

(1) 7

ACS Paragon Plus Environment


Page 8 of 26

145

We did not have access to a working aethalometer during the hydraulic fracturing phase, so for

146

this period, we estimated BC from the PM2.5 data. This was done by calculating the mass fraction

147

(MF) of BC to PM2.5 for each space/time co-located sample of the respective pollutants during

148

the other three well development phases.45 The distribution of the log-transformed mass fraction

149

was approximately normally distributed and did not change much across development phases, so

150

we used the overall log(MF) mean (Eq 2) to derive estimated BC from measured PM2.5 during

151

hydraulic fracturing, BCHF (Eq 3). 𝐵𝐶

𝑛

∑𝑖 = 1log (𝑃𝑀 ) 2.5

152

log(MF)=

153

BCHF,i=exp(log(MF)+logPM2.5,HF,i)

154

𝑛

(2) (3)

Noise measurements were collected with Larson Davis (Depew, NY) Spark 703+ and

155

Spark 706RC dosimeters each with a detachable 10.6 mm microphone/preamp with integrated 1

156

m cable (MPR001) and a windscreen. Noise measurements were recorded as the 1-minute

157

equivalent continuous noise level (Leq) as described in a previous manuscript that collected noise

158

during drilling at this location.30 Measurements made when wind speeds were greater than 16.1

159

kilometers per hour (10 mph) were omitted from our data analysis. A Davis Vantage Pro2

160

(Vernon Hills, IL, USA) weather station collected wind speed and temperature every 15 minutes

161

at the Northeast monitoring site for the hydraulic fracturing, flowback, and production phases.

162

During the drilling phase, a nearby weather station that recorded hourly information and reported

163

to Weather Underground was used in its place.

164 165

In addition to the pollution monitoring equipment, a GoPro Hero 4 Silver (San Mateo, CA, USA) was set up to take and record an image every 30 seconds at the Northeast monitoring


Page 9 of 26


166

site. This location was adjacent to the access road used by any vehicle entering or leaving the

167

well pad. Since this access road for the pad was the only entry point to the pad from the main

168

road, any large vehicle traveling on it was part of the MWP development. We reviewed four days

169

of images taken during drilling and seven days of images for each of the other well development

170

phases to obtain vehicle counts. Each time a vehicle was identified exiting the access road, the

171

time-stamped image was saved for analysis. The images were categorized as heavy trucks (i.e.,

172

semi-trailer), non-heavy trucks (i.e., light duty pickup), or another type of vehicle. An hourly

173

count of the heavy trucks and non-heavy trucks was produced for the four or seven-day

174

surveillance windows. A summary of dates and durations for all monitoring data can be found in

175

Table 1.

176

Table 1. Description of dates and duration of air and noise measurements collected at a new Multi-Well Pad in Weld County, Colorado.

Phase

Drilling

Dates

January 27, 2017 to April 24, 2017

April 21-24, 2017

Hydraulic Fracturing

July 24, 2017 to August 21, 2017

August 714, 2017

Flowback

October 2, 2017 to November 20, 2017

5-minute

Number of Measurements Included from Northeast7 Site 14,752

Number of Measurements Included from South8 Site 7,939

Number of Measurements Included from Northwest9 Site 16,712

Number of Measurements Included from East9 Site 7,935

1-minute

84,846

40,265

80,995

39,663

dBA3

1-minute

57,594

25,851

57,433

22,323

dBC4

1-minute

57,586

25,854

57,426

25,791

Heavy Trucks5 Non-Heavy Trucks6

1-hour

75

0

0

0

1-hour

75

0

0

0

BC1

1-minute

0

0

0

0

1-minute

40,058

39,893

0

0

dBA3

1-minute

39,959

39,983

0

0

dBC4

1-minute

39,972

39,995

0

0


1-hour

167

0

0

0

1-hour

167

0

0

0

BC1

1-minute

65,546

69,731

0

0

1-minute

69,507

67,393

0

0

dBA3

1-minute

64,721

70,254

0

0

dBC4

1-minute

70,283

64,765

0

0

Pollutant

Time Resolution

BC1 PM2.52

PM2.52

PM2.52



October 613, 2017

Production

January 8, 2018 to March 2, 2018

Page 10 of 26


1-hour

169

0

0

0

1-hour

169

0

0

0

BC1

1-minute

73,264

73,329

0

0

1-minute

74,206

74,285

0

0

dBA3

1-minute

63,669

65,466

0

0

dBC4

1-minute

59,255

62,887

0

0

PM2.52

Heavy 1-hour 167 0 0 0 Trucks5 Non-Heavy 1-hour 167 0 0 0 Trucks6 1 Black Carbon (BC) was measured every 5-minutes during drilling and was changed to every 1-minute for the other phases to match the time resolution of other pollutant measurements 2 Particulate Matter less than 2.5 microns in diameter (PM2.5) 3 A-weighted noise measured in decibels (dBA) 4 C-weighted noise measured in decibels (dBC) 5 Heavy trucks (semi-trailer) and 6 Non-Heavy (light duty/pick-up) were counted and categorized by each hour 7 The Northeast site was a home located along the well pad access road where monitoring was conducted during drilling, hydraulic fracturing, flowback, and production 8 The South site refers to a home where monitoring was conducted during drilling, hydraulic fracturing, flowback, and production 9 The Northwest, and East sites refer to one of two homes where monitoring was conducted during drilling only January 1522, 2018

177 178 179 180 181 182 183 184 185 186

Statistical Analysis Descriptive statistics were computed for each air pollutant (PM2.5 and BC) over three

187

averaging times: short-term (1-minute or 5-minute sample), hourly, and daily by well

188

development phase and sampling site. Different averaging times are used to characterize the

189

temporal variability of these air pollutants during each phase of well development. Two-way

190

ANOVA of the short-term measurements was used to compare concentrations across the

191

monitoring sites and development phases.

192

Noise pollution statistics were computed for A-weighted and C-weighted noise across the

193

factors of well development phase, weekday/weekend, and daytime (7:00am – 7:00pm) or

194

nighttime (7:00pm – 7:00am) as defined by the COGCC. The one-minute Leq was compared to

195

the World Health Organization’s guideline of 50 dBA for outdoor living areas46 to calculate the

196

percentage of total monitoring time that noise exceeded this level. As there has been little

197

research to date on the health effects from C-weighted noise, we used the COGCC limit of 65


Page 11 of 26


198

dBC (Leq over 15 minutes) that necessitates a low frequency noise impact analysis47 to calculate

199

a similar exceedance percentage.

200

Finally, descriptive statistics for hourly truck counts were computed for both heavy and

201

non-heavy trucks for each well development phase for daytime and nighttime hours. We did not

202

compute statistics for weekend and weekday because we only had photo surveillance and truck

203

counts for up to one week during each phase.

204 205

Results and Discussion

206

Particulate Matter

207

Table 2 summarizes descriptive statistics for the air pollutants PM2.5 and BC, stratified by

208

averaging time, well development phase, and monitoring location. Hydraulic fracturing had the

209

highest median PM2.5 levels across sites (short-term medians of 4.6 and 4.7 µg/m3 at the sites),

210

which reflects the intensity of activity at the MWP during this phase. The mean short-term PM2.5

211

concentration was highest during production at the South monitoring site (5.5 µg/m3). Flaring

212

during production was reported by residents and may have contributed to spikes in PM2.5 levels

213

that elevated the mean at that site, but we were not able to confirm the exact time of flaring to

214

match it with our measurements. Average PM2.5 was lowest during the drilling phase, with

215

phase-specific means of 2.3-3.4 µg/m3 at the monitoring sites. Mean PM2.5 was typically higher

216

at the Northeast site compared to the other sites except during production. This could be due to

217

the combination of truck traffic and equipment on the pad during drilling, hydraulic fracturing,

218

and flowback affecting measurements at this site. On the contrary, truck traffic during production

219

should be minimal. Two-way ANOVA showed that PM2.5 concentrations differed by the 11 ACS Paragon Plus Environment


Page 12 of 26

220

monitoring site (p

Community Noise and Air Pollution Exposure During the Development

Recommend Documents