Ohio, West Virginia, Pennsylvania Fracking Story Map

FracTracker’s aerial survey of unconventional oil & gas infrastructure and activities in northeast PA to southern OH and central WV

Compressor station within Loyalsock State Forest, PA.

Air Pollution from Pennsylvania Shale Gas Compressor Stations – REPORT

Air pollution from Pennsylvania shale gas compressor stations is a significant, worsening public health concern.

By Cynthia Walter, Ph.D.

Dr. Walter is a retired biology professor who has worked on shale gas industry pollution since 2009 through Westmoreland Marcellus Citizens Group, Protect PT and other groups. Contact: walter.atherton@gmail.com

Executive Summary

Compressor Stations (CS) in the gas industry are sources of serious air pollutants known to harm humans and the environment. CS are permanent facilities required to transport gases from wells to major pipelines and along pipelines. Additional operations and equipment located at CS also emit toxins. In the last 20 years, CS abundance and sizes have dramatically increased in shale gas extraction areas across the US. This report will focus on CS in and near Southwestern Pennsylvania. Numbers of CS there have risen more than tenfold in the last decade in response to well completions and pipelines after the local fracking boom began in 2005. For example, Westmoreland County, Pennsylvania, had two CS before 2005 and now has 50 CS corresponding with about 341 active shale gas wells. In Pennsylvania, state regulations allow CS to be as close as 750 feet from homes, schools, and businesses. Emission monitoring relevant to public health exposure is limited or absent.

Current Pennsylvania policies allow rapid CS expansion. Also, regulations do not address public health risks due to several major flaws. First, permits allow annual totals of emitted toxins using models that assume constant releases, but substantial emissions from CS occur in peaks that expose citizens to concentrations may impair health, ranging from asthma to cancer. Second, permits do not address the fact that CS simultaneously release many serious air toxins including benzene and formaldehyde, and particulates that carry toxins into lungs. This allowance of multiple toxin release does not reflect the well-established science that public health risks multiply when people are exposed to several toxins at once. Third, permit reviews rarely consider nearby known air pollution sources contributing to aggregate air toxin exposures that occur in bursts and continually. Fourth, permits do not require operators to provide public access to real-time reports of air pollutants released by CS and ambient air quality near CS.

Poor air quality causes harm directly, e.g. respiratory distress, and indirectly, e.g., through increased vulnerability to respiratory viruses. The annual cost of damages from air pollution from CS was estimated at $4 million-$24 million in Pennsylvania based on emissions from CS in 2011. These damages include harm to human and livestock health and losses of crops and timber. After 2011, CS and gas infrastructures continue to expand, with increasing air pollution and damages, especially in shale gas areas. These costs must be compared to the benefits of using alternative energy sources. For example, in a neighboring state, New York, shifting to renewable energy will save tens of billions of dollars annually in air pollution costs, prevent thousands of premature deaths each year, and trigger substantial job creation, based on peer-reviewed research using US government data.


  1. Constant air monitoring must occur at current compressor stations and nearby sites important to the public, such as schools. The peak concentrations and totals for substances relevant to public health must be recorded and made available to the public in real time.
  2. Air pollution from compressor stations must become an important part of measuring and modeling pollution exposures from all components of the shale gas industry.
  3. Permits for new compressor stations must be revised to better protect the public in ways including, but not limited to the following:
    • Location, e.g., increased general setback limits and expanded limits for sensitive sites such as schools, senior care facilities and hospitals
    • Emission limits for criteria air pollutants and hazardous air pollutants including Radon, especially limits for peak concentrations and annual totals
    • Monitoring air quality within the station, at the fence-line and in key sites nearby, such as schools, using information from air movement models to select locations and heights.
    • Limits for CS size based on aggregate pollution from other local air pollution sources.
  4. Costs of harm from CS and other shale gas activities must be compared to alternatives.

Table of Contents

Chemistry of Compressor Station Emissions

Health Effects of Compressor Station Emissions

Regional Air Toxins and Cancer Risk in Southwestern Pennsylvania

Measurements of Compressor Station Emissions

Compressor Station Locations

Costs of Compressor Stations and Air Pollution

Appendix – Compressor Station Locations in Westmoreland County, Pennsylvania

Chemistry of Compressor Station Emissions

CS emissions contribute major air pollutants to the total pollution from unconventional gas development (UCGD), but their role in regional air quality problems has not always been noted. In 2009, when UCGD operations were only a few years in this region and many CS had not yet been built, CS emissions were estimated to be a small component. Now, in 2020, gas transport requirements have increased, leading to many more and larger CS. The amounts of CS emissions have increased accordingly, based on estimates by Carnegie Mellon University atmospheric researcher, Robinson (Figure 1). Part of the reason that CS are such a major pollution source is that they run constantly, in contrast to machinery for well development and trucking that fluctuate with the market for new wells.

Relative contribution of compressor stations and other components of shale gas industry to Nitrous Oxides (NOx). Relative contribution of compressor stations and other components of shale gas industry to Volatile Organic Compounds (VOC).

Figure 1. Relative contribution of compressor stations and other components of shale gas industry to Nitrous Oxides (NOx) and Volatile Organic Compounds (VOC). Source: Clean Air Council- adapted from webinar by Alan Robinson.


Air pollutants in CS emissions vary substantially in chemistry and concentrations due to differences in equipment (Table 1). Emissions in CS can come from several types of sources described below.

  1. Engines: Compression engines powered with methane release nitrogen oxides (NOx), carbon monoxide (CO), volatile organic compounds (VOCs) and hazardous air pollutants (HAP). Diesel engines release those pollutants as well as sulfur dioxide (SO2) and substantial particulate matter. In addition, diesel storage on site is a hazard. Electric engines produce less pollutants, but they are much less common than fossil fuel engines in southwestern Pennsylvania. CS operators can vary the use of engines at a station, and therefore, emissions vary during partial or full shutdowns and start-up periods.
  2. Blowdowns: Toxic emissions dramatically increase during blowdowns, a procedure that is scheduled or used as needed to release the build-up of gases. Blowdown frequency and emissions vary with the rate of gas transport and the chemistry of transported gases. The full extent of emissions from any CS, therefore, is not known. Blowdowns can release a wide range of substances, and when flaring is used to burn off gases, the combustion creates new substances and additional particulates. Blowdowns are the most likely source of peaks in emissions at continuously operated CS. For example, Brown et al. (2015) used PA DEP measures of a CS in Washington County, Pennsylvania, alongside likely blowdown frequencies and weather models to predict peak emission frequency. They estimated nearby residents would experience over 118 peak emissions per year.
  3. Non-compression Procedures: CS facilities are often the location for equipment that separate gases, remove water and other fluids, and run pipeline testing operations called pigging. These activities can be constant or intermittent and release a wide range of substances which may or may not be included in estimates for a permit. In addition, some of the processing releases gases which are flared at the facility, thus releasing a range of combustion by-products and particulate pollution. For example, the Shamrock CS operated by Dominion Transfer Inc. includes equipment for dehydration, glycol processing and pigging. The Janus facility operated by EQT includes dehydration and flaring. Permitted emissions for those facilities are listed in Table 1.
  4. Storage Tank Emissions: CS often include storage tanks that hold substances known to release fumes. For example, the Shamrock CS was permitted to have an above ground storage tank of 3000 gallons for drip gas and a 1000-gallon tank for used oil, both of which release volatile organic compounds. The EQT Janus CS has two 8,820-gallon tanks. Gas releases from such tanks could be controlled and recorded by the operator or they could be unrecorded leaks.
  5. Fugitive emissions: Gas leaks, called fugitive emissions, occur readily from many components in CS facilities; such problems will increase as equipment ages. A study of CS stations in Texas is an example.

“In the Fort Worth, TX area, researchers evaluated compressor station emissions from eight sites, focusing in part on fugitive emissions. A total of 2,126 fugitive emission points were identified in the four month field study of 8 compressor stations: 192 of the emission points were valves; 644 were connectors (including flanges, threaded unions, tees, plugs, caps and open-ended lines where the plug or cap was missing); and 1,290 were classified as Other Equipment. The Other category consists of all remaining components such as tank thief hatches, pneumatic valve controllers, instrumentation, regulators, gauges, and vents. 1,330 emission points were detected with an IR camera (i.e. high-level emissions) and 796 emission points were detected by Method 21 screening (i.e. low-level emissions). Pneumatic Valve Controllers were the most frequent emission sources encountered at well pads and compressor stations.”

Eastern Research Group (2011).

Table 1. Examples of air pollutants allowed for release by compressor stations. Air pollutants (pounds/year) are estimates provided by the companies for permits in West Virginia and Pennsylvania in recent years. Total compressor engine horsepower (hp) is noted. Sources: Janus and Tonkin CS Permits at WV DEP website. Shamrock CS permit. Buffalo CS, Washington, Co PA – PENNSYLVANIA BULLETIN, VOL. 45, NO. 16 APRIL 18, 2015.

Pollutant  Term Janus (WV)

22,000 hp

Tonkin (WV)

4390 hp

Shamrock* (PA)

4140 bhp

Buffalo ** (PA) 20,000 hp + 5,000 bhp
Nitrogen Oxides NOx 254,400 248,000 170,000 155,800
Volatile Organic Compounds VOC 191,200   30,000  66,000  77,000
Carbon Monoxide CO 118,200   80,000 154,000 144,400
Sulfur Dioxide SO2   1,400       400  10,000   5,400
Hazardous Air Pollutants-Total HAP  48,200    3,280  19,400  30,000
   Formaldehyde   1,080  12,800  12,200
   Benzene      540
   Ethylbenzene        60
   Toluene      140
   Xylene      200
   Hexane      500
   Acetaldehyde      600
   Acrolein      160
Total Particulate Matter

(PM-2.5, PM-10-separate or combined)

PM 18,200  11,000  32,000 PM-10       32,000

PM-2.5      32,000

TOTAL TOXINS 631,600 372,680 417,400 444,600
Carbon Dioxide Equivalents CO2-e 29,298,000 27,200,000 367,000,000 214,514,000


Health Effects of Compressor Station Emissions

Several toxic chemicals are released by individual CS in amounts that range from a few thousand pounds to a quarter of a million pounds per year (Tables 1 & 2) as described below.

  • Nitrous Oxides (NOx) are often the largest total amount of emissions from fossil fuel machinery. In CS, these oxides are formed when a fossil fuel such as methane or diesel is combusted to produce the energy to compress and propel gases. NOx contribute to acid rain. Excess acids in rain lower the pH of waters, in some cases to levels that dissolve toxic metals in drinking water supplies. NOx also trigger the formation of ozone, a substance well known to impair lungs.
  • Ozone forms when oxygen reacts with nitrous oxides, carbon monoxide, and a wide range of volatile organic compounds. Ozone exposure can trigger asthma and heart attacks in sensitive individuals, and for healthy people, ozone causes breathing problems in the short term and eventual scarring of lungs and impaired function.
  • Volatile Organic Compounds (VOCs) are gaseous compounds containing carbon, such as benzene and formaldehyde. In air pollution regulation, the EPA lists many compounds as VOC, but excludes carbon dioxide, carbon monoxide, methane and butane. Many VOC’s are toxic in themselves (Tables 2, 3 and 4). Also, several VOC’s react to form ozone.  https://www.epa.gov/air-emissions-inventories/what-definition-voc
  • Carbon Monoxide (CO) is another product of fossil fuel combustion and another contributor to ozone formation. CO is directly toxic because it prevents oxygen from binding to the blood.
  • Sulfur Dioxide (SO2) adds to lung irritation. It also contributes to acid rain, lowering the pH of water and increasing the ability of toxic metals to dissolve in water supplies.
  • Hazardous Air Pollutants (HAP) include highly toxic substances such as formaldehyde and benzene, which are known carcinogens, as well as the other substances known to be emitted from CS (Tables 3 & 4). The EPA lists 187 substances as HAP, which include many VOC’s as well as some non-organic chemicals such as arsenic and radionuclides including Radon. (https://www.epa.gov/haps/initial-list-hazardous-air-pollutants-modifications)
  • Particulate Matter (PM) usually refers to particles in small size classes. Most state or federal regulations address measures of particles less than 10 microns (PM-10) and some monitoring systems separate out particles less than 2.5 microns (PM-2.5). Particles in either of those size ranges are not visible, but highly damaging because they travel deep into the lungs where they irritate tissues and impair breathing. Also, these tiny particles carry toxins from air into the blood passing through the lungs. This blood transports substances directly to the brain where toxins can quickly impair the nervous system and subsequently impact other organs. (https://www.epa.gov/pm-pollution/particulate-matter-pm-basics)

Health impacts from many of the substances released by CS are well-known in medical research. For example, many of the VOC and HAP compounds permitted for release by state agencies are known carcinogens (Table 3). Many of these substances also impact the nervous system as shown in the organic compounds measured in CS in PA and listed in Table 4. Also, a study of 18 CS in New York by Russo and Carpenter (2017) found that all 18 CS released substances with known impacts on the nervous system and total annual emissions were over five million pounds, among the highest of all types of emissions (Table 5). Russo and Carpenter also found high annual emissions of over five million pounds for substances known to be associated with each of the following other health problems: digestive problems, circulatory disorders, and congenital malformations.

Congenital defects were significantly more common for mothers living in a 10-mile radius of denser shale gas development in Colorado compared to reference populations (MacKenzie et al. 2014). Currie et al. (2017) examined over a million birth records in Pennsylvania and found statistically significant increased frequencies of low birth weight and negative health scores for infants born to mothers within 3 km of unconventional gas wells compared to matching populations more distant from shale gas developments. Such developments include a wide range of gas infrastructure including CS and also high truck traffic and fracking. One plausible mechanism for harm to developing babies is exposure to VOCs such as benzene, toluene and xylene associated with CS and well operations. These VOC’s are classified by the Agency for Toxic Substances and Disease Registry as known to cross the placental barrier and cause harm to the fetus including birth deformities.

In sum, CS are a significant source of air pollutants with direct and indirect impacts on health. One indirect impact especially important during the COVID-19 pandemic in 2020, is the increased incidence and severity of respiratory viral infections in populations living in areas with poor air quality. Ciencewicki, and Jaspers (2007) write, “a number of studies indicate associations between exposure to air pollutants and increased risk for respiratory virus infections.”

Table. 2. Health effects of air pollutants permitted for release by compressor stations.

Pollutant Health Effects
Particulate Matter Impairs lungs and transfers toxins into body when microscopic particles carry chemicals deep into lungs and release into bloodstream.
Nitrogen Oxides

Forms ozone that impairs lung function which can trigger asthma and heart attacks and scars lungs in the long term.

Forms acid rain that dissolves toxic metals into water supplies.

Volatile Organic Compounds Includes a wide variety of gaseous organic compounds, some of which cause cancer. Many VOC react to form ozone that impairs lungs as noted above.
Carbon Monoxide Blocks ability of blood to carry oxygen.

Also forms ozone that impairs lungs as noted above.

Sulfur Dioxide Irritates lungs, triggering respiratory and heart distress.

Forms acid rain that dissolves toxic metals into water supplies.

Hazardous Air Pollutants Category of various toxic compounds many of which impact the nervous system. Includes formaldehyde, benzene and several other carcinogens.
Total Toxins Sum of emissions of all toxins. Exposure to multiple toxins exacerbates harm directly through impairment of lungs and circulatory system and indirectly through injury to detoxification mechanisms, such as liver function.
Carbon Dioxide Equivalents A measure of the combined effects of greenhouse gases such as CO2 and Methane expressed in a standard unit equivalent to the heat trapping effect of CO2. Greenhouse gases trap heat and worsen climate change and related harm to health when increased air temperatures directly cause stress directly and indirectly accelerate ozone formation.


Table 3. Gas industry list of carcinogenicity rating for Hazardous Air Pollutants (HAPs) released by compressor stations in a factsheet prepared by EQT for Janus compressor, WV. 2015 Source: DEP.

Substance Type Known/Suspected Carcinogen Classification
Acetaldehyde VOC Yes B2-Probable Human Carcinogen
Acrolein VOC No Inadequate Data
Benzene VOC Yes Category A – Known Human Carcinogen
Ethyl-benzene VOC No Category D Not Classifiable
Biphenyl VOC Yes Suggested Evidence of Carcinogenic Potential
1,3 Butadiene VOC Yes B2-Probable Human Carcinogen
Formaldehyde VOC Yes B1- Probable Human Carcinogen
n-Hexane VOC No Inadequate Data
Naphthalene VOC Yes C- Possible human Carcinogen
Toluene VOC No Inadequate Data
2,3,4-Trimethlypentane VOC No Inadequate Data
Xylenes VOC No Inadequate Data


Table 4. Center for Disease Control list of health effects for volatile organic carbons measured by PA DEP near compressor station. Source: CDC.

Substance Exposure Symptoms Target Organs
Ethylbenzene Irritation to eyes and nose; nausea, headache; neuropath; numb extremities, muscle weakness; dermatitis; dizziness Eyes, skin, respiratory system, central nervous system, peripheral nervous system
n-Butane Drowsiness Central nervous system
n-Hexane Irritation to eyes, skin & respiratory system; headache, dizziness; nausea Eyes, skin, respiratory system, central nervous system
2-Methyl Butane n/a n/a
Iso-butane Drowsiness, narcosis, asphyxia Central nervous system


Table 5. Amounts of pollutants known to be associated with health impacts in a review of 18 New York compressor stations. Emissions were grouped and tallied based on their impacts on disorders classified by ICD codes as defined by the International Statistical Classification of Diseases and Related Health Problems (ICD), a medical classification list by the World Health Organization. Source: Copy of Table 3.17b, Russo and Carpenter 2017.

ICD-10 Facilities Chemicals Pounds
# Description ‘08 ‘11 ‘14 Tot ‘08 ‘11 ‘14 Tot 2008 2011 2014 Total
1 Q00-Q89 Congenital malformations and deformations 18 18 17 18 57 54 54 57 4,393,806 6,607,676 5,900,691 16,902,175
1.1 Q00-Q07 Nervous system 18 18 17 18 16 16 16 16 4,068,877 5,882,704 5,258,344 15,209,926
1.2 Q10-Q18 Eye, ear, face and neck 15 15 12 15 4 4 4 4 5,825 19,569 11,475 36,869
1.3 Q20-Q28 Circulatory system 18 18 17 18 10 10 10 10 4,269,779 6,336,905 5,651,896 16,258,581
1.4 Q30-Q34 Respiratory system 14 8 7 14 4 4 4 4 150 107 113 372
1.5 Q35-Q45 Digestive system 18 18 17 18 17 17 17 17 4,386,043 6,586,345 5,884,324 16,856,713
1.6 Q50-Q56 Genital organs 6 7 8 8 2 2 2 2 1,399 4,373 2,612 8,385
1.7 Q60-Q64 Urinary system 18 17 16 18 9 9 9 9 119,382 254,922 237,359 611,663
1.8 Q65-Q79 Musculoskeletal system 18 18 16 18 19 19 19 19 122,314 262,300 243,932 628,547
1.9 Q80-Q89 Other 18 18 17 18 55 52 52 55 2,124,445 3,614,575 3,413,375 9,152,395
2 Q90-Q99 Chromosomal abnormalities, nec 18 18 16 18 30 31 31 32 120,669 256,739 239,709 617,118
Q00-Q99 Total 18 18 17 18 57 56 56 59 4,393,806 6,607,676 5,900,691 16,902,175

Regional Air Toxins and Cancer Risk in Southwestern Pennsylvania

Cancer risks from HAPs have been elevated for many years in several areas of Southwestern PA, as noted in a map from 2005 (Figure 2), when most air pollution was from urban traffic and single sources such as coke works and unconventional gas development (UCGD) had just begun in the region. The cancer risk pattern changed by 2014 (Figure 3). The specific numbers of excess cancer risk predicted for each location cannot be compared between the two maps because each map was produced using different sources of information and models. The pattern, however, can be compared and shows that elevated cancer risk is now more widespread across Southwestern PA and no longer primarily in Allegheny County.

Cancer risk maps are constructed by the EPA office of National Air Toxics Assessment (NATA) using models of reported air toxics and their relationship to cancer as a risk factor, as defined by NATA: “A risk level of “N”-in-1 million implies that up to “N” people out of one million equally exposed people would contract cancer if exposed continuously (24 hours per day) to the specific concentration over 70 years (an assumed lifetime). This would be in addition to cancer cases that would normally occur in one million unexposed people.” (https://www.epa.gov/national-air-toxics-assessment/nata-glossary-terms) In the current context, the NATA models are useful to compare the relative differences in air quality from a public health perspective, assuming the data on air pollutants is complete.

Another, very different statistic regarding cancer is the rate of cancer, also called the incidence. This number is based on actual reported cases and applies to cancers that occur due to all causes. The cancer rate, therefore, is a much higher number than a risk factor. For example, according to the US Center for Disease Control, the annual rate of new cases of cancer in PA in 2016, the most recent year reported, was 482.5 per 100,000 people. Compared to other states, PA is among the ten states with the highest cancer incidence. In the US, one in four people die from cancer, placing it second to heart disease as a leading cause of death. (https://gis.cdc.gov/Cancer/USCS/DataViz.html). Compared to other nations, the US has the fifth highest cancer rate, with 352 new cases each year per 100,000 people. (https://www.wcrf.org/dietandcancer/cancer-trends/data-cancer-frequency-country)

Compressor station emissions contribute to air pollutants known to be associated with cancer. For example, in a review of emissions for 18 CS in New York, Russo and Carpenter (2017) found that most or all CS released substances associated with a wide range of cancers (Table 6). Up to 56 such chemicals were emitted in amounts that totaled over 1 million pounds each year.

Maps of cancer risk are likely to be under-reporting risk levels in both the amount rates of risk and also the locations. Cancer risks from serious air pollutants cannot be properly mapped for several reasons. First, reports on concentrations of HAP in emissions are limited. HAP emissions are in accounts required only from large facilities, and thus, smaller operations, such as many CS, are likely be ignored. Second, general air quality monitoring stations are limited in location and do not measure HAP. For example, the PA DEP maintains 47 air quality stations dispersed among over 60 counties (http://www.dep.state.pa.us/dep/deputate/airwaste/aq/aqm/pollt.html). Most stations report hourly measures of Ozone and PM-2.5, and only a handful also monitor one or more other substances such as CO, NOx, SO ₂ or H2S. One county in Southwestern PA has additional air quality stations. Allegheny has a county health department that maintains 17 stations to report real-time air quality based on Ozone, SO2 or PM-2.5 (https://alleghenycounty.us/Health-Department/Programs/Air-Quality/Air-Quality.aspx).

In sum, cancer risk estimates from air pollution fall short in the following ways:

  • Estimates of air quality do not reflect the reality of air pollution from CS as well as many other new sources such as increased truck traffic associated with shale gas development.
  • Tallies of annual emissions do not represent the actual exposures of individuals to pulses of toxins.
  • Models of air pollution and cancer are not sufficiently based on real world studies of impacts from multiple toxins in short and long-term exposures.

Cancer risk map in Southwestern Pennsylvania in 2005

Figure 2. Cancer risk map in Southwestern Pennsylvania in 2005 from the National Air Toxics Assessment program in the EPA. Total Lifetime Cancer Risk from Hazardous Air Pollutants (HAP) per million. Colors indicate yellow for 28-78, gold for 79-95, light orange for 99-148, orange for 149-271, bright orange for 272-517, and red for 518-744 excess cancer risk per million. (https://www.epa.gov/national-air-toxics-assessment)

Cancer risk map in Southwestern Pennsylvania in 2014 from the National Air Toxics Assessment in the EPA.

Figure 3. Cancer risk map in Southwestern Pennsylvania in 2014 from the National Air Toxics Assessment in the EPA. Facilities are locations where air quality information was available for modeling. Total Risk of cancer as a baseline was assumed to be 1 per 1,000,000.  Estimates of risk predict known air pollution sources alone will cause 1-24 excess cancers per million in Light Pink areas, 25-49 excess cancers per million in Gray areas, and 50-74 excess cancers per million in Blue areas. Source: EPA.

Table 6. Amounts of pollutants known to be associated with cancer in a review of 18 New York compressor stations. Emissions were grouped and tallied based on their impacts on disorders classified by ICD codes as defined by the International Statistical Classification of Diseases and Related Health Problems (ICD), a medical classification list by the World Health Organization. Source: Copy of Table 3b, Russo and Carpenter 2017.


ICD-10 Facilities Chemicals Pounds
# Code Description ‘08 ‘11 ‘14 Tot ‘08 ‘11 ‘14 Tot 2008 2011 2014 Total
1 C00-C97 Malignant neoplasms 18 18 17 18 53 54 54 56 744,394 1,679,621 1,583,745 4,007,761
2 C00-C14 Lip, oral cavity and pharynx 18 18 16 18 12 14 14 14 118,992 254,897 238,943 612,833
3 C15-C26 Digestive organs 18 18 16 18 37 38 38 38 121,690 258,670 241,866 622,227
4 C30-C39 Respiratory system and intrathoracic organs 18 18 17 18 36 37 37 38 740,798 1,673,574 1,579,882 3,994,254
5 C40-C41 Bone and articular cartilage 18 18 17 18 33 34 34 35 694,106 1,551,399 1,492,704 3,738,210
6 C43-C44 Skin 16 15 13 16 12 12 12 14 2,362 5,008 4,029 11,400
7 C45-C49 Connective and soft tissue 17 17 15 17 17 17 17 17 1,929 5,074 4,639 11,643
8 C50-C58 Breast and female genital organs 18 18 16 18 23 25 25 25 361,015 823,303 663,237 1,847,556
9 C60-C63 Male genital organs 18 17 16 18 12 13 13 13 111,217 233,176 224,147 568,541
10 C64-C68 Urinary organs 18 18 16 18 24 24 24 25 119,062 255,474 238,596 613,133
11 C69-C72 Eye, brain and central nervous system 18 18 16 18 20 20 20 20 121,282 258,655 241,954 621,892
12 C73-C75 Endocrine glands and related structures 18 17 16 18 10 10 10 10 112,911 235,120 225,269 573,300
13 C76-C80 Secondary and ill-defined 17 16 14 17 6 6 6 6 2,054 5,690 5,771 13,516
14 C81-C96 Malignant neoplasms, stated or presumed to be primary, of lymphoid, haematopoietic and related tissue 18 18 16 18 31 31 31 31 364,338 833,140 671,245 1,868,724
15 C97 Malignant neoplasms of independent (primary) multiple sites 0 0 0 0 0 0 0 0 0 0 0 0
16 D00-D09 In situ neoplasms 16 15 13 16 3 3 3 3 3,313 7,557 6,606 17,477
17 D10-D36 Benign neoplasms 17 17 14 17 27 27 27 27 12,499 35,013 23,068 70,580
18 D37-D48 Neoplasms of uncertain or unknown behavior 18 18 16 18 39 40 40 41 121,277 257,142 240,115 618,535

Measurements of Compressor Station Emissions

Studies of real-world concentrations of air pollutants from CS emissions are lacking, but some reports exist. Of these, a few records are in peer-reviewed studies, and cited in reviews such as Saunders et al. 2018.  A few published reports are described below. They all show the high variation over time for CS emissions and the occurrence of peak concentrations.

Macey et al. (2014) observed ambient air near CS contained toxins at concentrations that impair health. They collected grab samples of air from industrial sites including CS in Arkansas and Pennsylvania and analyzed them for toxins using EPA approved methods. Most of the CS studied in Arkansas (Table 6) and Pennsylvania (Table 7) released formaldehyde at amounts associated with a cancer risk from exposure to this substance of 1/10,000 which is equivalent to 100 times higher risk than the widely accepted baseline risk of 1 per million. This means the amounts of formaldehyde found near CS substantially increased the risk of cancer using well-established federal analyses (https://www.atsdr.cdc.gov/hac/phamanual/appf.html).  Some toxins Macey et al. recorded are less well studied than formaldehyde and benzene. For example, 1,3-butadiene is classified by the EPA as a known human carcinogen, but a calculation of cancer risk for this substance is lacking. Air samples in the Macey study were collected close to the CS (e.g., 30-42m) and at greater distances (e.g., 254-460m). Those distant samples were well beyond the 750-foot set-back rule for Pennsylvania. At all these distances, air movement modeling predicts that toxins released from a source such as a CS are likely to travel downwind within the air mass under most weather conditions, thus exposing residents near and further from CS. Many people, therefore, in homes, schools and businesses that are downwind of CS are likely to experience serious air toxins at concentrations that harm their health.

Air toxins were also measured by the Pennsylvania Department of Environmental Protection in 2010 in a variety of unconventional gas extraction facilities including one CS in Washington County, PA. Brown et al. (2015) reported these data, showing the concentrations that citizens could experience near a compressor station varied greater than tenfold within a day and among consecutive days (Table 8). The length of time for peak concentrations was unknown, but Brown et al. used a model of weather including wind patterns to estimate citizens are likely to experience 118 peak concentrations per year.

Goetz et al. (2015) sampled air in Marcellus shale regions of Pennsylvania for short periods (1-2.5 hrs.) at distances 480-1100 meters from eight CS, four with relatively small capacity (5,000-9,000 hp) and four with moderate capacity (14,000-17,000 hp). They found that each CS had a different pattern of relatively higher concentrations of some pollutants, such as NOX versus other pollutants, e.g., CO. Also, totals of all pollutants did not correlate with compressor engine capacity, probably because the CS they sampled include a mix of engines using fossil fuels and electric power. Goetz et al. concluded with recommendations for more comprehensive and longer-term monitoring to better understand air pollution from CS and all components in shale gas development.

Radionuclides in CS emissions are almost never measured, even though Marcellus shales are well known for containing elevated amounts of radiologic substances such as uranium, radium and radon. The only published report of testing for radionucleotides in CS emissions in PA was a test of a single CS emission for one period of time. In a review of radiation in shale gas industry components, the Pennsylvania Department of Environmental Protection (PA DEP) measured radon (Rn) in ambient air at one CS by deploying sample bags in four cardinal directions at the fence line at a height of 5 feet for 62 days. They reported Rn concentrations of 0.1-0.8 pCi/L, values they stated were within the range of outdoor air in the US.  (https://www.dep.pa.gov/Business/Energy/OilandGasPrograms/OilandGasMgmt/Oil-and-Gas-Related-Topics/Pages/Radiation-Protection.aspx)  Given the high variation of amounts of emissions from CS and variable chemistry in sources of gases released from combustion, blowdowns and leaks, frequent testing for radionucleotides should be standard in monitoring CS emissions.

Methane is the substance tracked most often in emissions from CS and other gas industry facilities because of its central role in operations, requirements to avoid explosive concentrations, and readily available measurement technology, in comparison to other substances emitted from CS. Although methane emissions from CS are not always correlated with amounts of other, more toxic emissions, patterns observed in plumes of methane from CS are likely to reflect elevated concentrations of other harmful substances from CS.

Nathan et al (2015) sampled methane emissions from one CS in the Barnett shale region using a sensor carried on a model aircraft. The open-path, laser sensor produced measures with a precision of 0.1 ppmv over short intervals, allowing researchers to see emission variation in time and space as the aircraft changed position. Based on 22 flights within a week period, they observed a substantial range in methane released from 0.3 – 73 g CH4 per second. These values calculate to 0.02 – 6.3 metric tons of methane per day, a range that matches that estimated by Goetz of 0.5 – 9 metric tons per day. In addition, Nathan et al. found high variability in concentrations at different heights, as the emission plumes shifted in response to wind velocity, direction and topography. They recommend caution in interpretations of ground-based emission monitors and called for more monitoring of air movements and emissions at different elevations.

Payne et al. 2017 confirmed these ideas when they mapped plumes of methane in CS in New York and Pennsylvania using a sensor capable of recording methane in parts per million (ppm) every 0.25 – 5 seconds. The sensor was located on a mobile unit that marked GPS location. They found high variability in the shape and extent of plumes. For example, one of most extensive plumes was recorded near Dimock, Pennsylvania in a locale with CS as the only major source of methane. Researchers recorded the highest concentrations of methane in the study, 22 ppm, at 500 m from the CS, with a second peak of 0.6 ppm noted over 1 km from the CS and elevated methane as far as 3 km from the site (Figure 4). Wind direction did not always predict the shape of the plume, but data collection was restricted by the path of the sensor and the transport vehicle (Figure 8). Most importantly, they found that …“during atmospheric temperature inversions, when near-ground mixing of the atmosphere is limited or does not occur, residents and properties located within 1 mile of a compressor station can be exposed to rogue methane from these point sources.” These residents are likely to also experience excess toxins from CS as well, especially under such weather conditions.

Exposure to peak concentrations of air pollutants have dramatic effects on health for several reasons. First, lungs carry toxins into the blood within seconds, and the blood quickly transfers compounds to the brain and other vital organs. Many of the substances released by compressor stations impact the central nervous system as seen in Table 3, and these toxins are released simultaneously. Citizens, therefore inhaling a plume of emissions will have impacts from the total of these compounds. The health impacts for these combined toxins are unknown, and especially of concern during pregnancy and child development. Exposure studies in animals and humans test individual substances and the Center for Disease Control and NIOSH use these to develop exposure guidelines for a healthy adult in a work-place. In contrast, residents near compressor stations will include citizens of all ages with various health conditions. For example, the American Lung Association determined that over 50% of the 360,000 residents of Westmoreland County are at greater risk for health impairment due to air pollution because they have one or more of these conditions: asthma, diabetes, heart disease, respiratory illness, advanced age (https://www.lung.org/our-initiatives/healthy-air/sota/key-findings/people-at-risk.html).

In sum, the research on CS emissions of methane, air pollutants such as NOx, and hazardous air pollutants such as formaldehyde and benzene, all indicate exposures to CS emissions pose a threat to public health, but the emissions have not yet been fully quantified and modeled. Documenting CS contributions to harmful ambient air quality is feasible, however. The published studies from as far back as 2011 indicate that instrumentation to record substances and weather are readily available. Activities within a station such as compressor function, blowdowns, venting and flaring are all recorded by operators, but such reports are not released to researchers or the public. The science of models that predict public health risks in response to air pollution exposure are highly developed. In sum, operators of CS have the technology to measure emissions and ambient air quality and scientists have the models, but lack of industry data prevents the public from knowing impacts from CS.


Table 6. Air toxins found in grab samples near Arkansas compressor stations including concentrations, the Agency for Toxic Substances and Disease Registry (ASTDR), Minimum Risk Level (MRL) exceedance, and the Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) cancer risk. Source: Copy of Table 4 from Macey et al. 2014.

State/ID County Nearest infrastructure Chemical Concentration (μg/m3) ATSDR MRLs


EPA IRIS cancer risk exceeded
AR-3136-003 Faulkner 355 m from compressor Formaldehyde 36 C 1/10,000
AR-3136-001 Cleburne 42 m from compressor Formaldehyde 34 C 1/10,000
AR-3561 Cleburne 30 m from compressor Formaldehyde 27 C 1/10,000
AR-3562 Faulkner 355 m from compressor Formaldehyde 28 C 1/10,000
AR-4331 Faulkner 42 m from compressor Formaldehyde 23 C 1/10,000
AR-4333 Faulkner 237 m from compressor Formaldehyde 44 C, I 1/10,000
AR-4724 Van Buren 42 m from compressor 1,3-butadiene 8.5 n/a 1/10,000
AR-4924 Faulkner 254 m from compressor Formaldehyde 48 C, I 1/10,000

C = chronic; I = intermediate.


Table 7. Air toxins found in grab samples near Pennsylvania compressor stations including concentrations, the Agency for Toxic Substances and Disease Registry (ASTDR), Minimum Risk Level (MRL) exceedance, and the Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) cancer risk. Source: Copy of Table 5 from Macey et al. 2014



County Nearest infrastructure Chemical Concentration (μg/m3) ATSDR MRLs


EPA IRIS cancer risk exceeded
PA-4083-003 Susquehanna 420 m from compressor Formaldehyde 8.3 1/10,000
PA-4083-004 Susquehanna 370 m from compressor Formaldehyde 7.6 1/100,000
PA-4136 Washington 270 m from PIG launcha Benzene 5.7 1/100,000
PA-4259-002 Susquehanna 790 m from compressor Formaldehyde 61 C, I, A 1/10,000
PA-4259-003 Susquehanna 420 m from compressor Formaldehyde 59 C, I, A 1/10,000
PA-4259-004 Susquehanna 230 m from compressor Formaldehyde 32 C 1/10,000
PA-4259-005 Susquehanna 460 m from compressor Formaldehyde 34 C 1/10,000

C = chronic; A = acute; I = intermediate.

aLaunching station for pipeline cleaning or inspection tool.


Table 8. Variation in air pollutants measured in ug/cubic meter by PA DEP during two sampling times per day for three consecutive days near a compressor station in Southwest PA. Source: Copied from Table 1. Brown et al. 2015 based on data from Southwestern Pennsylvania Short Term Marcellus Ambient Air Sampling Report, Pennsylvania Department of Environmental Protection, Nov. 2010.

May 18 May 19                                 May 20
Chemical Morning Evening Morning Evening Morning Evening 3-day Average
Ethylbenzene No detect No detect 964 2015 10,553 27,088 13,540
n-Butane 385 490 326 696 12,925 915 5,246
n-Hexane No detect 536 832 11,502 33,607 No detect 15,492
2-Methyl Butane No detect 230 251 5137 14,271 No detect 6,630
Iso-butane 397 90 No detect 1481 3,817 425 2070



Methane emission plumes from compressor stations near Dimock, Pennsylvania Methane emission plumes from compressor stations near Springvale, Pennsylvania 

Figure 4. Methane emission plumes from compressor stations near Dimock, Pennsylvania (left) and Springvale, Pennsylvania (right). Source: Copied from Payne et al. 2017.


Compressor Station Locations

Prior to 2008, compressor stations were infrequent with one or a few per county broadly distributed across PA as part of gas transport from locations outside of PA (Figure 5). These pipelines were mainly an issue for public health in the case of explosions. Major transmission pipelines use pressures up to 1500 psi. Leaks, therefore, release large amounts of gas much of which is not noticed because it lacks the mercaptan odorant added to household methane. For example, the 30-inch Spectra gas pipeline that exploded in 2016 in Westmoreland County caused a hole 12 feet deep and1500 square feet in area and burned 40 acres. The PA DEP claimed to have measured air quality, but they did not arrive until long after the plume from the fire traveled downwind. This pipeline was transporting gas from one of the largest gas storage facilities in the country, the Sunoco Gas Depot in Delmont, Pennsylvania to New Jersey as part of over 9,000 miles of pipelines in the Texas Eastern system from the Gulf Coast to the Northeast. That section of pipeline was built in 1981 and had recently been increased in pressure, probably using older or newer compressors in nearby locations. Faulty joints between pipeline sections were blamed for the catastrophic release of gas. (Phillips, S. 2016. State Impact, NPR). Immediately after the explosion, while gas continued to pour out of the pipeline, emergency workers needed at least one hour to locate shut-off locations. In general, pipeline shut-offs are sited at compressors stations or at intervals along a pipeline.

CS abundance in counties with shale gas extraction increased over tenfold in the decade after 2005 when the gas industry obtained exemptions to the Clean Water Act and began unconventional gas extraction in Pennsylvania (Figure 6). Permit applications for new wells, pipelines and CS continue throughout southwest Pennsylvania. In PA, the Oil and Gas law states the following: “ In order to allow  for the reasonable development of oil and gas resources, a local ordinance … Shall authorize natural gas compressor stations as a permitted use in agricultural and industrial zoning districts and as a conditional use in all other zoning districts, if the natural gas compressor building meets the following standards:….(i) is located 750 feet or more from the nearest existing building or 200 feet from the nearest lot line, whichever is greater, unless waived by the owner of the building or adjoining lot;”  (Pennsylvania Statutes Title 58 Pa.C.S.A. Oil and Gas §3304). CS and many aspects of the shale gas industry are controlled by this state law.

Each stage of gas extraction involves emissions that can be close or far from the well pad. Most emissions involve diesel engines. Diesel engines are well-known to produce substantial amounts of VOC’s, NOx and particulate pollution (PM-2.5, PM-10). Well pad construction requires intense activity by diesel trucks and earth moving equipment. Well drilling uses diesel engines. From 3 – 5 million gallons of water are used for each fracking event and up to 300 truck visits are needed to transport water for the many wells that are not close to water supplies from piped sources. Trucks are used to transport the 1 – 2 million gallons of produced water that emerges from the well for disposal in injection wells likely to be distant from most wells. Additional waste is carried long distances as well, including drill cuttings and sludge. For example, shale gas industry waste was handled for years in Max Environmental, one of the largest industrial waste sites in the eastern US located in Yukon, Westmoreland County since the 1960’s. Within one mile of Yukon is Reserved Environmental, a waste facility with operations focused since 2008 on processing sludge from fracking into solid cakes to be trucked to other landfills. In sum, all stages of shale gas industry contribute to many poorly documented sources of air pollution likely to be near CS.

The density of CS in some areas such as southwest Pennsylvania impacts the local and regional air quality. For example, Westmoreland County has 50 CS and 341 shale gas wells (https://www.fractracker.org) and some neighboring counties have even more shale gas emission sources. People in Westmoreland County receive pollutants from shale gas activities in their immediate vicinity and additional air pollutants from CS and other industries in neighboring counties. Wind patterns shown in Figure 7 indicate Westmoreland County is frequently downwind from Washington County, a county with a very high density of shale gas operations, and Eastern Allegheny County where large industries such as coke works release substantial amounts of air pollutants.

Compressor Stations prior to 2008 and in around 2013

Figure 5. Compressor Stations prior to 2008 and in around 2013. Source: Copied from article by James Hilton in Pittsburgh Post-Gazette.

 Compressor Stations in Pennsylvania mapped in 2019

Figure 6. Compressor Stations in Pennsylvania mapped in 2019. Source: FracTracker Alliance. 2000.

Wind patterns at small airports around Pennsylvania

Figure 7. Wind patterns at small airports around Pennsylvania 1991-2005 showing predominant direction of wind and velocity in knots (Orange 0 – 4, Yellow 4 – 7, Turquoise 7 – 11, Medium Blue 11 – 17, Dark Blue 17 – 21). Source: The Pennsylvania State Climatologist.

Costs of Compressor Stations and Air Pollution

As permanent, constant sources of air and noise pollution and safety risks, CS add significant costs to communities. Poor air quality alone is well-established as an economic drain for a region due to many factors including increased health care, lower property values, a declining tax base, and difficulty in attracting new businesses or housing development. Litovitz et al. (2013) estimated that, compared to other activities of shale gas extraction, CS made up the majority of the annual emissions of important air toxins in 2011, and therefore a majority of the damages from air pollution, totaling 4 – 24 million dollars of the 7 – 32 million dollars of the aggregate air pollution damages from gas operations (Table 9).

Litovitz and others recognize that the costs of damages from the gas industry air pollution in 2011 may appear smaller than the state-wide impacts from other industries, such as coal burning power plants and coke production, but that appearance deserves a second look. First, shale gas extraction activities are concentrated in a few regions of Pennsylvania, and local air quality is most relevant to public health and local economics such as property values. Second, emissions from gas extraction in 2011 was only in its early stages in Pennsylvania and shale gas operations will expand greatly unless regulations change, while coal-fired power plants are declining due to the advanced age of most facilities. For example, in Westmoreland County, PA alone there are over 50 CS in 2020, the number currently in the entire state of New York, where unconventional gas development was suspended due, in large part, to concerns for public health. Costs from one aspect of an energy sector can be viewed in the context of economic and other benefits of alternative energy efforts. For example, Jacobson et al. (2013) estimated that shifting to clean, renewable energy in NY state would prevent 4000 premature deaths each year and save $33 billion/year through air pollution reductions that impact health care, crop production and other costs. Jacobson et al. used government data in their models regarding health benefits and also identified substantial job growth during and after the transition away from fossil fuels toward renewable energy. Pennsylvania has the potential to attain similar benefits in air quality, public health, savings and job growth gained from a shift to clean, renewable energy in place of fossil fuels.

Table 9.  a) Emissions from shale gas industry in 2011 throughout Pennsylvania in metric tons per year. b) Costs of damages due to air pollution from shale gas extraction in 2011 throughout Pennsylvania. Copied from Tables 5 and 6 in Litovitz et al. 2013.


Activities VOC NOx PM2.5 PM10 SOx
(1) Transport 31–54 550–1000 16–30 17–30 0.82–1.4
(2) Well drilling and hydraulic fracturing 260–290 6600–8100 150–220 150–220 6.6–190
(3) Production 71–1800 810–1000 15–78 15–78 4.8–6.2
(4) Compressor stations 2200–8900 9300–18 000 280–1100 280–1100 0–340
Totalᵃ 2500–11 000 17 000–28 000 460–1400 460–1400 12–540

ᵃ These totals are reported to two significant figures, as are all intermediate emissions values in this document. The activity emissions may not exactly sum to the totals.


Activities Timeframe Total regional damage for 2011 ($2011) Average per well or per MMCF damage ($2011)
(1) Transport Development $320 000–$810 000 $180–$460 per well
(2) Well drilling, fracturing Development $2 200 000–$4 700 0 $1 200-$2 700 per well
(3) Production Ongoing $290 000–$2 700 0 $0.27-$2.60 per MMCF
(4) Compressor stations Ongoing  $4 400 000–$24 000 000 $4.20-$23.00 per MMCF
(1)-(4) Aggregated Both $7 200 000–$32 000 000 NA

Major Studies Cited in Text:

Brown, David, Celia Lewis, Beth I. Weinberger and Heather Bonaparte. 2014. Understanding air exposure from natural gas drilling put air standards to the test. Reviews in Environmental Health. https://doi.org/10.1515/reveh-2014-0002

Brown, David, Celia Lewis and Beth I. Weinberger. 2015. Human exposure to unconventional natural gas development; a public health demonstration of high exposure to chemical mixtures in ambient air. Journal of Environmental Science and Health (Part A) 50: 460-472.

Ciencewicki, J. and I. Jaspers 2007. Air Pollution and Respiratory Viral Infection. Inhalation Toxicology 19:1135–1146, DOI: https://doi.org/10.1080/08958370701665434

Currie, J, M Greenstone and K Meckel. 2017. Hydraulic fracturing and infant health: New evidence from Pennsylvania.   Science Advances 2017;3:e1603021

Eastern Research Group, Inc. and Sage Environmental Consulting, LP. City of Fort Worth natural gas air quality study: final report. July 13, 2011. http://fortworthtexas.gov/gaswells/air-quality-study/final/

Goetz, J.D. E. Floerchinger, E., C. Fortner, J. Wormhoudt, P. Massoli, W. Berk Knighton, S.C. Herndon, C.E. Kolb, E. Knipping, S. L. Shaw, and P. F. DeCarlo. 2015. Atmospheric Emission Characterization of Marcellus Shale Natural Gas Development Sites. Environ. Sci. Technol. 49, 7012−7020. DOI: https://doi.org/10.1021/acs.est.5b00452

Jacobson, MZ, RW Howarth, MA Delucchi, ST Scobie, JH Barth, M Dvorak, M Klevze, H. Hatkhuda, B. Mirand, NA Chowdhury, R Jones, L Plano, AR Ingraffea. 2013. Examining the feasibility of converting New York State’s all-purpose energy infrastructure to one using wind, water, and sunlight. Energy Policy 57: 585-601.

Litovitz, A., A. Curtright, S. Abramzon, N. Burger and C. Samaras. 2013. Estimation of regional air-quality damages from Marcellus Shale natural gas extraction in Pennsylvania. Environ. Res. Lett. 8; 014017 (8pp) doi:10.1088/1748-9326/8/1/014017. https://iopscience.iop.org/article/10.1088/1748-9326/8/1/014017/meta

Macey, G.P., Breech, R., Chernaik, M. (2014) Air concentrations of volatile compounds near oil and gas production: a community-based exploratory study. Environ Health 13, 82 (2014). https://doi.org/10.1186/1476-069X-13-82

McKenzie, LM, G Ruisin, RZ Witter, DA Savitz, LS Newman, JL Adgate. 2014. Birth Outcomes and Maternal Residential Proximity to Natural Gas Development in Rural Colorado.  Environmental Health Perspectives Vol 22.  http://dx.doi.org/10.1289/ehp.1306722.

Nathan BJ, LM Golston, AS O’Brien , K Ross, WA Harrison, L Tao, DJ Larry, DR Johnson, AN Covington, NN Clark, MA Zondlo. 2015. Environ Sci Technol. 2015   Near-Field Characterization of Methane Emission Variability from a Compressor Station Using a Model Aircraft. Environ Sci Technol. 2015 Jul 7;49(13):7896-903 doi: 10.1021/acs.est.5b00705.

Payne, RA, P Wicker, ZL Hildenbrand, DD Carlton, and KA Schug. 2017. Characterization of methane plumes downwind of natural gas compressor stations in Pennsylvania and New York. Science of The Total Environment  580:1214-1221

Russo, PN and DO Carpenter 2017. Health Effects Associated with Stack Chemical Emissions from NYS Natural Gas Compressor Stations: 2008-2014 Institute for Health and the Environment, A Pan American Health Organization / World Health Organization Collaborating Centre in Environmental Health, University at Albany, 5 University Place, Rensselaer New York. Https://www.albany.edu/about/assets/Complete_report.pdf

Saunders, P.J., D. McCoy. R. Goldstein. A. T. Saunders and A. Munroe. 2018.   A review of the public health impacts of unconventional natural gas development Environ Geochem Health 40:1–57. https://doi.org/10.1007/s10653-016-9898-x



Compressor Stations in Westmoreland Co. PA in Dec 2019, based on information from FracTracker Alliance, Pennsylvania Department of Environmental Protection Air Quality Report, and the Department of Homeland Security.

ID # Facility # Name/Operator Municipality Latitude Longitude Status
627743 645570 CNX GAS CO/HICKMAN COMP STA Bell Twp 40.5174 -79.5498 Active
693305 696606 PEOPLES TWP/RUBRIGHT COMP STA Bell Twp 40.5278 -79.5561 Active
626482 644726 CNX GAS CO/BELL POINT COMP STA Bell Twp 40.5413 -79.5338 Active
na na NORTH OAKFORD Delmont 40.4018 -79.5597 Active
714057 713241 RW GATHERING LLC/ECKER BERGMAN RD COMP STA Derry Twp 40.3533 -79.3028 Active
760724 752063 RE GAS DEV/ORGOVAN COMP STA Derry Twp 40.3857 -79.4019 Active
736807 732436 RW GATHERING LLC/SALEM COMP STA Derry Twp 40.3908 -79.3361 Active
714057 713241 RW GATHERING LLC/ECKER BERGMAN RD COMP STA Derry Twp 40.3533 -79.3028 Active
774714 766854 EQT GATHERING LLC/DERRY COMP STA Derry Twp 40.4511 -79.3161 Active
na na Layman Compressor, Range Resources Appalachia, LLC East Huntingdon 40.1113 -79.6345 Unknown
na na Key Rock Energy/LLC East Huntingdon 40.1228 -79.6489 Unknown
662759 673466 Kriebel Minerals Inc./Sony Compressor Station (Inactive) East Huntingdon 40.181 -79.5882 Unknown
662781 673477 Lynn Compressor, Kriebel Minerals Inc. East Huntingdon 40.1798 -79.5557 Unknown
636316 660570 Range Resources Appalachia/ Layman Compressor Station East Huntingdon 40.1086 -79.6359 Unknown
na na Keyrock Energy LLC/ Hribal Compresor Station, East Huntingdon, Pa. (active) East Huntingdon 40.1353 -7905653 Unknown
761545 752755 KeyRock Energy LLC/ Hribal Compressor Station (Active) East Huntingdon 40.1333 -79.55 Unknown
649767 663499 Range Resources Appalachia/Schwartz Comp. Station East Huntingdon 40.0879 -79.601 Unknown
652968 665874 TEXAS KEYSTONE/FAIRFIELD TWP COMP STA Fairfield Twp 40.3363 -79.1786 Active
557780 572987 EQUITRANS LP/W FAIRFIELD COMP STA Fairfield Twp 40.3333 -79.1167 Active
675937 683303 DIVERSIFIED OIL & GAS LLC/MURPHY COMP SITE Fairfield Twp 40.3362 -79.1122 Active
812881 806928 TEXAS KEYSTONE INC/ MURPHY COMP STA Fairfield Twp 40.3543 -79.1123 Active
na na SOUTH OAKFORD/Dominion Greensburg 40.365 -79.5585 Unknown
na na OAKFORD Greensburg 40.3848 -79.5489 Active
na na DELMONT Geensburg 40.382 -79.5554 Active
496667 626720 Silvis Compressor Station, Exco Resources Pa. Inc Hempfield 40.2022 -79.5526 Unknown
na na  Dominion Trans Inc., Lincoln Heights Hempfield Township 40.3004 -79.6193 Active
812660 806731 CNX Gas Co. LLC Hempfield Township 40.2957 -79.6277 Active
812661 806732 CNX Gas Co. LLC/ Jackson Compressor Station, Status: Active Hempfield Township 40.2931 -79.6119 Unknown
601521 626775 PEOPLES NATURAL GAS CO/ARNOLD COMP STA Lower Burrell City 40.3623 -79.4316 Active
812883 806930 TEXAS KEYSTONE INC/LOYALHANNA Loyalhanna Twp 40.4514 -79.4727 Inactive
na na J.B. TONKIN Murrysville Boro 40.4629 -79.6402 Active
815083 809310 HUNTLEY & HUNTLEY INC/BOARST COMP STA Murrysville Boro 40.4686 -79.6417 Inactive
735725 731655 MTN GATHERING LLC/10078 MAINLINE COMP STA Murrysville Boro 40.4708 -79.65 Active
241708 276314 Dominion Trans Inc/Jeannette Penn Township 40.3317 -79.5935 inactive
na 701239 DOMINION ENERGY TRANS INC/ROCK SPRINGS COMP STA Salem Twp 40.4052 -79.5546 Unknown
na na OAKFORD Salem Twp 40.4052 -79.5546 Unknown
465965 495182 EQT GATHERING/SLEEPY HOLLOW COMP STA Salem Twp 40.3634 -79.5426 Inactive
465965 495182 EQT GATHERING/SLEEPY HOLLOW COMP STA Salem Twp 40.3634 -79.5426 Inactive
483173 512126 COLUMBIA GAS TRANS CORP/DELMONT COMP STA Salem Twp 40.3871 -79.5638 Active
707759 708010 LAUREL MTN MIDSTREAM OPR LLC/SALEM COMP STA Salem Twp 40.3782 -79.4929 Active
459024 488214 CNX Gas Co./ Jacobs Creek Compressor Station, South Huntingdon Twp 40.1172 -79.6681 Unknown
634559 650802 Rex Energy I LLC/Launtz Unity Twp 40.3325 -79.4295 Unknown
na 668776 Keyrock Energy LLC/ Unity Compressor Station Unity Twp 40.2251 -79.5109 Unknown
na na Nelson/RE Gas Dev LLC UnityTwp 40.3378 -79.4348 Unknown
657366 66932 People’s Natural Gas/ Latrobe Compressor Station Unity Twp 40.3075 -79.4369 Inactive
812662 806733 CNX Gas Co. LLC, Troy Compressor Station Unity Twp na na Unknown
657366 564168 Dominion Peoples (Inactive) Unity Twp 40.3073 -79.4371 Inactive
815196 809457 HUNTLEY & HUNTLEY INC/WASHINGTON STATION Washington Twp 40.4967 -79.6206 Active
605562 629821 PEOPLES NATURAL GAS/MERWIN COMP STA Washington Twp 40.5083 -79.6203 Active
815203 809466 HUNTLEY & HUNTLEY INC/TARPAY STA Washington Twp 40.5222 -79.6186 Active
na na Mamont (CNX GAS CO/MAMONT COMP STA) Washington Twp 40.5046 -79.5862 Unkown
741197 735870 CONE MIDSTREAM PARTNERS LP/MAMONT COMP STA Washington Twp 40.5067 -79.5644 Active


Feature image of a compressor station within Loyalsock State Forest, PA. Photo by Brook Lenker, FracTracker Alliance, June 2016.

Support this work

Stay in the know

National Energy and Petrochemical Map

FracTracker Alliance has released a new national map, filled with energy and petrochemical data. Explore the map, continue reading to learn more, and see how your state measures up!

View Full Size Map | Updated 4/13/21 | Data Tutorial

4/13/2021 Update

This map has been updated since this blog post was originally published, and therefore statistics and figures below may no longer correspond with the map

The items on the map (followed by facility count in parenthesis) include:

         For oil and gas wells, view FracTracker’s state maps. 

This map is by no means exhaustive, but is exhausting. It takes a lot of infrastructure to meet the energy demands from industries, transportation, residents, and businesses – and the vast majority of these facilities are powered by fossil fuels. What can we learn about the state of our national energy ecosystem from visualizing this infrastructure? And with increasing urgency to decarbonize within the next one to three decades, how close are we to completely reengineering the way we make energy?

Key Takeaways

  • Natural gas accounts for 44% of electricity generation in the United States – more than any other source. Despite that, the cost per megawatt hour of electricity for renewable energy power plants is now cheaper than that of natural gas power plants.
  • The state generating the largest amount of solar energy is California, while wind energy is Texas. The state with the greatest relative solar energy is not technically a state – it’s D.C., where 18% of electricity generation is from solar, closely followed by Nevada at 17%. Iowa leads the country in relative wind energy production, at 45%.
  • The state generating the most amount of energy from both natural gas and coal is Texas. Relatively, West Virginia has the greatest reliance on coal for electricity (85%), and Rhode Island has the greatest percentage of natural gas (92%).
  • With 28% of total U.S. energy consumption for transportation, many of the refineries, crude oil and petroleum product pipelines, and terminals on this map are dedicated towards gasoline, diesel, and other fuel production.
  • Petrochemical production, which is expected to account for over a third of global oil demand growth by 2030, takes the form of chemical plants, ethylene crackers, and natural gas liquid pipelines on this map, largely concentrated in the Gulf Coast.

Electricity generation

The “power plant” legend item on this map contains facilities with an electric generating capacity of at least one megawatt, and includes independent power producers, electric utilities, commercial plants, and industrial plants. What does this data reveal?

National Map of Power plants

Power plants by energy source. Data from EIA.

In terms of the raw number of power plants – solar plants tops the list, with 2,916 facilities, followed by natural gas at 1,747.

In terms of megawatts of electricity generated, the picture is much different – with natural gas supplying the highest percentage of electricity (44%), much more than the second place source, which is coal at 21%, and far more than solar, which generates only 3% (Figure 1).

National Energy Sources Pie Chart

Figure 1. Electricity generation by source in the United States, 2019. Data from EIA.

This difference speaks to the decentralized nature of the solar industry, with more facilities producing less energy. At a glance, this may seem less efficient and more costly than the natural gas alternative, which has fewer plants producing more energy. But in reality, each of these natural gas plants depend on thousands of fracked wells – and they’re anything but efficient.Fracking's astronomical decline rates - after one year, a well may be producing less than one-fifth of the oil and gas it produced its first year. To keep up with production, operators must pump exponentially more water, chemicals, and sand, or just drill a new well.

The cost per megawatt hour of electricity for a renewable energy power plants is now cheaper than that of fracked gas power plants. A report by the Rocky Mountain Institute, found “even as clean energy costs continue to fall, utilities and other investors have announced plans for over $70 billion in new gas-fired power plant construction through 2025. RMI research finds that 90% of this proposed capacity is more costly than equivalent [clean energy portfolios, which consist of wind, solar, and energy storage technologies] and, if those plants are built anyway, they would be uneconomic to continue operating in 2035.”

The economics side with renewables – but with solar, wind, geothermal comprising only 12% of the energy pie, and hydropower at 7%, do renewables have the capacity to meet the nation’s energy needs? Yes! Even the Energy Information Administration, a notorious skeptic of renewable energy’s potential, forecasted renewables would beat out natural gas in terms of electricity generation by 2050 in their 2020 Annual Energy Outlook.

This prediction doesn’t take into account any future legislation limiting fossil fuel infrastructure. A ban on fracking or policies under a Green New Deal could push renewables into the lead much sooner than 2050.

In a void of national leadership on the transition to cleaner energy, a few states have bolstered their renewable portfolio.

How does your state generate electricity?

Figure 2. Electricity generation state-wide by source, 2019. Data from EIA.

One final factor to consider – the pie pieces on these state charts aren’t weighted equally, with some states’ capacity to generate electricity far greater than others.  The top five electricity producers are Texas, California, Florida, Pennsylvania, and Illinois.


In 2018, approximately 28% of total U.S. energy consumption was for transportation. To understand the scale of infrastructure that serves this sector, it’s helpful to click on the petroleum refineries, crude oil rail terminals, and crude oil pipelines on the map.

Map of transportation infrastructure

Transportation Fuel Infrastructure. Data from EIA.

The majority of gasoline we use in our cars in the US is produced domestically. Crude oil from wells goes to refineries to be processed into products like diesel fuel and gasoline. Gasoline is taken by pipelines, tanker, rail, or barge to storage terminals (add the “petroleum product terminal” and “petroleum product pipelines” legend items), and then by truck to be further processed and delivered to gas stations.

The International Energy Agency predicts that demand for crude oil will reach a peak in 2030 due to a rise in electric vehicles, including busses.  Over 75% of the gasoline and diesel displacement by electric vehicles globally has come from electric buses.

China leads the world in this movement. In 2018, just over half of the world’s electric vehicles sales occurred in China. Analysts predict that the country’s oil demand will peak in the next five years thanks to battery-powered vehicles and high-speed rail.

In the United States, the percentage of electric vehicles on the road is small but growing quickly. Tax credits and incentives will be important for encouraging this transition. Almost half of the country’s electric vehicle sales are in California, where incentives are added to the federal tax credit. California also has a  “Zero Emission Vehicle” program, requiring electric vehicles to comprise a certain percentage of sales.

We can’t ignore where electric vehicles are sourcing their power – and for that we must go back up to the electricity generation section. If you’re charging your car in a state powered mainly by fossil fuels (as many are), then the electricity is still tied to fossil fuels.


Many of the oil and gas infrastructure on the map doesn’t go towards energy at all, but rather aids in manufacturing petrochemicals – the basis of products like plastic, fertilizer, solvents, detergents, and resins.

This industry is largely concentrated in Texas and Louisiana but rapidly expanding in Pennsylvania, Ohio, and West Virginia.

On this map, key petrochemical facilities include natural gas plants, chemical plants, ethane crackers, and natural gas liquid pipelines.

Map of Petrochemical Infrastructure

Petrochemical infrastructure. Data from EIA.

Natural gas processing plants separate components of the natural gas stream to extract natural gas liquids like ethane and propane – which are transported through the natural gas liquid pipelines. These natural gas liquids are key building blocks of the petrochemical industry.

Ethane crackers process natural gas liquids into polyethylene – the most common type of plastic.

The chemical plants on this map include petrochemical production plants and ammonia manufacturing. Ammonia, which is used in fertilizer production, is one of the top synthetic chemicals produced in the world, and most of it comes from steam reforming natural gas.

As we discuss ways to decarbonize the country, petrochemicals must be a major focus of our efforts. That’s because petrochemicals are expected to account for over a third of global oil demand growth by 2030 and nearly half of demand growth by 2050 – thanks largely to an increase in plastic production. The International Energy Agency calls petrochemicals a “blind spot” in the global energy debate.

Petrochemical infrastructure

Petrochemical development off the coast of Texas, November 2019. Photo by Ted Auch, aerial support provided by LightHawk.

Investing in plastic manufacturing is the fossil fuel industry’s strategy to remain relevant in a renewable energy world. As such, we can’t break up with fossil fuels without also giving up our reliance on plastic. Legislation like the Break Free From Plastic Pollution Act get to the heart of this issue, by pausing construction of new ethane crackers, ensuring the power of local governments to enact plastic bans, and phasing out certain single-use products.

“The greatest industrial challenge the world has ever faced”

Mapped out, this web of fossil fuel infrastructure seems like a permanent grid locking us into a carbon-intensive future. But even more overwhelming than the ubiquity of fossil fuels in the US is how quickly this infrastructure has all been built. Everything on this map was constructed since Industrial Revolution, and the vast majority in the last century (Figure 3) – an inch on the mile-long timeline of human civilization.

Figure 3. Global Fossil Fuel Consumption. Data from Vaclav Smil (2017)

In fact, over half of the carbon from burning fossil fuels has been released in the last 30 years. As David Wallace Wells writes in The Uninhabitable Earth, “we have done as much damage to the fate of the planet and its ability to sustain human life and civilization since Al Gore published his first book on climate than in all the centuries—all the millennia—that came before.”

What will this map look like in the next 30 years?

A recent report on the global economics of the oil industry states, “To phase out petroleum products (and fossil fuels in general), the entire global industrial ecosystem will need to be reengineered, retooled and fundamentally rebuilt…This will be perhaps the greatest industrial challenge the world has ever faced historically.”

Is it possible to build a decentralized energy grid, generated by a diverse array of renewable, local, natural resources and backed up by battery power? Could all communities have the opportunity to control their energy through member-owned cooperatives instead of profit-thirsty corporations? Could microgrids improve the resiliency of our system in the face of increasingly intense natural disasters and ensure power in remote regions? Could hydrogen provide power for energy-intensive industries like steel and iron production? Could high speed rail, electric vehicles, a robust public transportation network and bike-able cities negate the need for gasoline and diesel? Could traditional methods of farming reduce our dependency on oil and gas-based fertilizers? Could  zero waste cities stop our reliance on single-use plastic?

Of course! Technology evolves at lightning speed. Thirty years ago we didn’t know what fracking was and we didn’t have smart phones. The greater challenge lies in breaking the fossil fuel industry’s hold on our political system and convincing our leaders that human health and the environment shouldn’t be externalized costs of economic growth.

Support this work

Stay in the know