Model Predicted Distribution of PM2.5 Exposure-Related Health Effects from Marcellus Shale Gas Development in Pennsylvania, 2005-2017

Background Development of Marcellus shale gas in recent years has turned into a public concern due to air quality changes and associated health impacts caused by emissions from shale gas development. Controlling standards and regulations were devised to provide health protection from exposure to such emissions, but health impact incidences have been reported in the areas where shale gas activities were prevailing. This study estimates the changes in expected health impacts associated to ne particulate matter (PM 2.5 ) emissions from shale gas development in Pennsylvania between 2005 and 2017. Methods The change in incidence rates is estimated using a Gaussian plume model to simulate PM 2.5 concentrations over time and distance and a health risk model to calculate the relative risk of exposure associated with different concentration levels. Results Simulation results indicate that 2.5 emissions from gas could increase the incidence of at specic the of this increase could be as high as 20,411 over 13-years. spatial


Introduction
The technological innovations in drilling and hydraulic fracturing led to a growing rate of shale gas production in Pennsylvania in the beginning of 21st century. According to the US Energy Information Administration (EIA), production from shale gas was 1.1 trillion cubic feet in 2011 and it increased to more than 5 trillion cubic feet in 2016 (EIA -Shale gas production, 2019). While shale gas supports energy security, provides fewer negative environmental impacts than other fossil fuels such as coal, and boosts the local economy, it may yet result in environmental degradation and net economic losses if not done properly (Sovacool, 2014). Jaramillo and Muller (2016) listed Pennsylvania with Ohio and Indiana as the U.S. states with the highest annual monetary damages due to exposure to air pollutants from energy production. This study estimated the marginal health cost of oil and gas extraction to increase from 1. 23  Total emissions in areas highly concentrated with shale gas development may lead to concentrations higher than the EPA National Ambient Air Quality Standards (Banan and Gernand, 2018), and recorded data show that in some cases, concentrations increased up to 20 to 40 times higher than the permitted level (Sovacool, 2014). State and local governments designate the minimum distance between a shale gas wellsite and any residential building (known as setback distance) to control the associated health risks from emissions due to shale gas activities and to protect public safety (McKenzie et al., 2016).
However, studies showed that setback policies were not successful in achieving their purpose to the fullest (Haley et al., 2016;Banan and Gernand, 2018).
Research shows an association between health risks and air pollutants originating from shale gas development (Shonkoff et al., 2014). Fine particulate matter (PM 2.5 ) is one of these pollutants (Shonkoff et al., 2014) whose health burden is known as a serious health concern around the world (WHO, 2007). According to Pascal et al. (2016), anthropogenic PM 2.5 emissions are thought to be responsible for 9% of mortality incidences in France.
According to Werner et al. (2015), despite a lack of highly relevant evidence on association between shale gas development activities and any direct health impacts, the claim of shale gas development causing speci c health outcomes might be valid if the lack of causal link is due to recent emergence of this industry. Other epidemiologic studies have demonstrated associations between these activities and nasal and sinus issues, migraine headaches, and fatigue symptoms (Tustin et al., 2017), and mild to severe asthma exacerbations (data from over 35 counties in PA) (Rasmussen et al., 2016). Weinberger et al. which contained greater number of reported health symptoms by people who were living within 1 km from these sources compared to ones living farther away.
One common approach in the evaluation of health burdens from exposure to PM 2.5 emissions is to estimate the increase in relative risk (RR) due to changes in concentrations by means of risk functions. Different studies have applied relative risk models in form of linear models (Cohen et al. 2004), integrated exposure-response (IER) models (Burnett et al. 2014;Maji et al 2018), and log-linear models (Burnett et al., 2014; RTI International, 2015;Lu et al. 2016;Pascal et al. 2016) with the latter recommended by WHO in estimation of health burdens due to air pollution (Ostro, 2004).
Risk models estimate the health risk based on pollutant exposure as their main input. Exposure changes as a function of time and location (Brown et al., 2015) and allocation of shale gas wells in a more dispersed pattern could cause a decrease of 80% or an increase of 600% in the number of unhealthy exposures based on population density within the vicinity of these sources (Banan and Gernand, 2020). Also, Morelli et al. (2016) argued that evaluating exposure risk based on emission measurements at background monitoring stations resulted in an underestimated risk value by 8 to 20%. Therefore, it is important to account for spatial and temporal changes in air pollution concentrations in estimating the health risks attributed to shale gas development.
This study investigates the expected burden of disease associated with exposure to PM 2.5 emissions from shale gas development in Pennsylvania between 2005 and 2017. We used a risk model to estimate the rate of health incidents based on the ndings by epidemiologic studies on associated health effects to these emissions. This study contributes to the previous studies on environmental and health damages due to shale gas activities (Litovitz et al., 2013;Sovacool, 2014) by providing more spatial resolution in the overall burden of health impacts imposed by such development in the state.

Methodology
This evaluation comprises of two analytical steps. The rst step is the simulation of PM 2.5 concentrations originated from each developed shale gas well over the period of 2005 to 2017 in the Marcellus shale region of Pennsylvania. We identi ed the locations where annual mean concentration of PM 2.5 changed due to shale gas development and estimated the number of people who could have experienced this exceedance. In the second step, we calculated the health risk associated to the simulated concentrations and determined the potential number of health problem cases, based on the results from epidemiological studies.

Model
The main model, implemented MATLAB (R2015a), is a health risk function which simulates the health risk due to exposure to PM 2.5 emissions. For the purpose of this study, concentration-response function is a better method than dose-response function due to uncertainty in actual internal dose which serves as the input for the latter function. Concentration-response function (CRF) is as formulated below (Evans et al., 2013): where is relative risk at exposure compared to the reference exposure , and represents the pollutant effect coe cient which indicates association between chronic emission exposure and cause of mortality or a disease. Coe cient is retrievable from epidemiological studies.
Results from CRF will be the input of the health impact function. In this study, we used a typical log-linear health impact function, introduced by where indicates resulting change in the number of adverse health outcomes, is the baseline incidence rate associated to reference exposure , is the population affected by the change in air quality.
In this evaluation, we estimate exposure and the population affected by the change in air quality at each speci c distance from every developed well between 2005-2017. For this purpose, we applied the

Data Sources
This study focuses on estimation of changes in the incidences of health impacts associated with residents' exposure to PM 2.5 emissions from shale gas developments between 2005 and 2017 in Pennsylvania. Thus, we simulate such impacts in association with any developed well during this period of time in Marcellus shale region of Pennsylvania.
Well data comprising of location, number of wells per wellpad, and "SPUD Date" (the date that drilling operation started) was retrieved from the reports by Pennsylvania Department of Environmental Protection (Department of Oil and Gas Reporting website, 2019). We used the wind data provided by Iowa Environmental Mesonet (IEM, 2018), measured at seventeen weather monitoring stations in Pennsylvania between 2005 and 2017, comprising of hourly wind direction, wind speed, relative humidity, and cloud coverage.
Solid information is not available on the rate of pollutant emissions from shale gas development stages.
We repeated the Monte Carlo simulation by Roy  We used the latest U.S. Census block data (Census Data, 2010) to investigate potential number of people who might have experienced air quality changes due to emission from shale gas development.
Corresponding data was population, block area, and block geographic location (latitude and longitude).
Pollutant coe cient ( , Eq. 1) and baseline incidence rate ( , Eq. 2) for different health outcomes are the two main inputs of our model. Epidemiologic studies have provided estimated pollutant coe cients for the health issues which evaluations show associations between exposure to PM 2.5 emissions and increase in their incidence. Table-1 provides the list of pollutant coe cients and baseline incidence rates used in this evaluation. The U.S. EPA developed the environmental Bene ts Mapping and Analysis Program (BenMAP) which has been used as an analytical tool by scientists, policy analysts, and decision makers for air quality management purposes and policy assessment and to estimate the human health impact associated with changes in air quality ( RTI International, 2015). Therefore, we used the same reference studies included by BenMAP in order to keep the same ground for our analytical estimates as the ones used in policy and regulatory purposes before. [

Analysis
This study provides an estimation of the health impacts imposed by shale gas development as a contribution to similar efforts by studies such as Litovitz  For each wellpad, our model uses the "SPUD DATE" as well as the total drilled and fractured length of wells to specify the activities time window. It then uses the hourly wind data measurements at the closest weather monitoring station to that wellsite during the corresponding time window to simulate the PM 2.5 concentrations at each hour of operation. The model assumes zero accumulation of emissions from hour to hour, but it accounts for accumulation of emissions dispersed from multiple sources at every hour to simulate PM 2.5 concentrations at each grid.
The reference exposure ( in Eq. 1) at any location is assumed to be equal to the level of PM 2.5 before the air quality was changed due to emissions from shale gas development. Therefore, the model estimates the increase in the incidence of any health impact from the expected number of incidences corresponding to no shale gas development in those areas. Incidence rate data was not available for all identi ed health impacts which were identi ed to be associated to chronic exposure to PM 2.5 emissions.
Thus, the model estimated the increase in the health impacts with available incidence rate.
For the purpose of this study, we used the unit of person-years in calculating the potential number of affected people by PM 2.5 emissions from shale gas development over years. Person-years is an epidemiologic jargon which allows to accommodate for persons coming into and leaving the study when quantifying incidence rate of a health outcome. Therefore, it accounts for the possibility that same person be counted in multiple years.
Our model only accounts for emissions from diesel engines at the wellsite and does not consider emissions from trucks, fugitive dust, mineral dust from proppant handling, or road dusts. We also assumed a at terrain (no buildings or similar structures) around the wellsite, as most of the shale gas wellsites in Pennsylvania are generally located in rural areas with few or no such a structure within their close vicinity. We assumed constant weather conditions for each hour of operation at each wellsite as well as the same hourly emission rates for drilling and hydraulic fracturing stages of all wells. The model assumes zero precipitation which is a typical assumption in Gaussian plume modeling. This study does not account for the role of delays due to weather, equipment failures, or other factors on the actual duration of development operations at the wellsites, as this information is not publicly available. Such delays would extend the period of emissions being released from the diesel engines in-use and could C 0 increase the annual average concentration within the vicinity of shale gas wellsite. This increase implies higher relative risk value (Eq. 1) and therefore, greater increase in the incidence of different health impacts (Eq. 2).

Results
[ Figure-1: Simulated changes in the incidence of health impacts due to shale gas activities (person). For the cases of asthma, chronic bronchitis and lung cancer, the lower and upper bounds are the estimated incidence cases based on available data for pollutant coe cients and baseline incidence rates. For other health impacts only one set of data was retrieved.]  Also, Figure-  Butler, Fayette, Westmoreland and Allegheny are the next four counties (in decreasing order) with high shares of the estimated total number of people affected by low air quality due to shale gas development (see Additional Excel File, Table-S2 to Table-S27). Table-2 provides more details on the wellsite-speci c factors that could explain the contribution of listed counties in Figure-4 to the estimated changes in incidence of associated morbidity outcomes. Corresponding population and activity statistics in these counties ( Table-3) indicate that the contribution of Washington County to the simulated changes is attributable to level of activity (number of developed wells and wellpad density) as well as proximity to more populated areas. Results presented in Figure-3 and Figure-4 are based on high level of PM 2.5 emission rate during drilling and hydraulic fracturing stages. Refer to the Additional Excel File, Table-S28 for corresponding estimations based on mean and low levels of PM 2.5 emission rates.
The density of shale gas development in terms of the number of wells per area is found to be a signi cant factor in case of Fayette, Westmoreland and Allegheny counties. For instance, data indicate that 5.2% of the estimated total people affected by low air quality due to shale gas development in Fayette is attributable to the cumulative effect of emissions from these activities. In case of Butler County, the level of activity (number of developed wells and wellpad density) found to be a less signi cant factor, but proximity to more populated areas explains the high contribution of this county to the estimated total incidence of morbidity cases.

Discussion
Results from this study are simulation of the contribution of PM 2.5 emissions from shale gas development to the increases in incidence of speci c diseases. These results are consistent with the ndings by Jaramillo and Muller (2016) and Litovitz et al. (2013) on the increases in the incidence of health issues due to PM 2.5 emissions from shale gas development, while they add more resolutions in spatial distribution of such increase. The current analysis also estimates the incidence of speci c health outcomes by county.
Our results show the ine cacy of current policies to achieve public health protection at every location, especially when accounting for accumulation of PM 2.5 emissions from multiple wellsites. In the case of asthma, one of the predominate health impacts of exposure to these emissions, we estimated emissions from shale gas development wellsites to result in an increase of 20,157 (95% CI: 20012, 20411) incidents in Pennsylvania. The mean value of this range represents a 2.3% increase in the total cases in the state Washington County ranked rst. From our results, part of the potential change in the incidence of associated health impacts are attributable to the counties with either higher level of shale gas development (i.e. more developed wells per wellsite or per area) (such as Washington, Bradford and Susquehanna), or higher population density within the vicinity of developments (Westmoreland).
However, there is still a considerable portion of estimated increase in health problems such as asthma in the less-populated counties (Greene) or counties with a relatively lower level of shale gas development (Allegheny), or both (Fayette). Our results indicate that estimated total change in the incidence of associated health impacts should be investigated with respect to a combination of factors (i.e. number of wells, wellpad-density, well density per area, and population density). For example, despite a lower number of developed wells in Allegheny County, a higher number of wells per wellpad and proximity of development activities to more populated areas help explain the estimated change in the incidences of associated health impacts in this county (Table-3). Table- Banan and Gernand (2018), have previously discussed the insu ciency of current setback policies for shale gas wellsites to achieve compliance with EPA's standards on PM 2.5 exposure. However, results presented here indicate that even though increasing setback distance would reduce the exposure of people to high PM 2.5 concentrations, it does not seem to have much protective effect with respect to population health. The area experiencing low increases in concentrations is much larger and much more populated than the close-in area adjacent to the wellsites where concentrations tend to exceed NAAQS. This presents a unique environmental management challenge that could be addressed more e ciently by reducing emissions at the wellsites rather than pushing these wellsites further from occupied buildings.

Contribution by other counties listed in
Results from this study imply the need for a health protection policy which accounts for spatial characteristics within the vicinity of developed wellsites as well as the level of activities. Such a policy should consider the impact of cumulative emissions from multiple sources, even in low-populated areas (similar to the case of Beaver County). Said otherwise, a policy which sets a cap on number of wells to be developed per site and/or the minimum distance from development sites may not be su cient to achieve health protection from PM 2.5 emissions due to shale gas development in Pennsylvania.
A positive association between exposure to PM 2.5 emissions and cardiovascular mortality has even been observed at PM 2.5 concentrations of 8.7 in Canada (Crouse et al., 2012). Nevertheless, there is no evidence of a lack of adverse health impacts below a speci c threshold (WHO, 2013). Therefore, we conducted this evaluation for any level of increase in PM 2.5 exposure due to shale gas development activities and did not suggest that the health risk is zero at any speci c concentration level. Our goal was to evaluate the effectiveness of current policy in providing public health protection and our results indicate the extra incidences of any speci c health impact due to emissions from shale gas development.
There are some limitations pertaining to this evaluation. The even population distribution may re ect a rough assumption to incorporate the impact of population spatial distribution. Results from this study are an estimation of changes in the incidence of health impacts due to shale gas development, based on the described associations between exposure to PM 2.5 emissions and such impacts by peer-reviewed epidemiologic studies. While such estimates are made with the best methodology available, predicting effects from exposures to speci c pollutants when the studied populations were exposed to complex mixtures leaves some irresolvable uncertainty. The best way to reduce such uncertainty for future research involves monitoring of exposures as they occur rather than reconstructing them from historical data and recording health effects of the affected population during and after such exposures. Such studies would be resource intensive, but this analysis presented here should offer information to enhance the feasibility of these future studies. Our results indicate that the overall impact of this industry is non-trivial, and greater scrutiny of the air quality impacts of the industry as a whole is warranted.

Conclusion
This study indicated the inadequacy of current policy to provide public health protection at all locations by estimating the increases in incidence of adverse health effects associated with PM 2.5 emissions from shale gas development. Such increases are not correlated with the total number of developed wells in all μ g / m 3 locations. Rather these effects are attributable to the case of wellsites with high wellpad density or multiple wellsites with fewer wells per wellpad within a close proximity to each other being developed in more populated areas.
Results from this study show that potential changes in the incidence of health impacts are not uniform across the Marcellus shale region in Pennsylvania. There is no one speci c factor that governs the pattern of these changes, and rather it is the complex combination of wellpad density, well area-density (number of wells per area), the number of developed wells, and population density that explain the differences in the changes of health impacts incidences at different locations. Therefore, an environmental policy that seeks to reduce the health burden of hazardous emissions such as PM 2.5 resulted by shale gas development should consider spatial characteristics of the neighborhood alongside the level of activities at the future development wellsites.

Declarations
Declarations Ethical Approval and Consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its supplementary information les. More detailed explanations and descriptions are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This research did not receive any speci c grant from funding agencies in the public, commercial, or notfor-pro t sectors.
Authors' contributions ZB contributed to the conceptualization and design of the work, devising the methodology, data gathering and investigation, data curation, development of the MATLAB model and conducting the scienti c analysis, interpretation and evaluation of the results, and preparation and writing the original draft.
JG contributed to the conceptualization and design of the work, development of methodology, interpretation and evaluation of the results, and conducting the review and revising of the draft. JG provided the resources.
Both authors approved the submitted version of this scienti c work and agreed to be personally accountable for their own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which they were not personally involved, are appropriately investigated, resolved, and the resolution documented in the literature.

Figure 1
Simulated changes in the incidence of health impacts due to shale gas activities (person). For the cases of asthma, chronic bronchitis and lung cancer, the lower and upper bounds are the estimated incidence cases based on available data for pollutant coe cients and baseline incidence rates. For other health impacts only one set of data was retrieved.

Figure 2
Estimated reduction in the number of asthma outcomes due to PM2.5 emissions from shale gas development at different setback distances compared to current setback distance in Pennsylvania (152.4 m or 500 ft.) for the case of Asthma.

Figure 3
Simulated contribution of emissions from shale gas activities to changes in incidence of mortality outcomes by county: a) Coronary heart disease (CHD), b) Cardiovascular disease (CVD), and c) Lung cancer.

Figure 4
Simulated contribution of emissions from shale gas activities to changes in the number of morbidity outcomes by county: a) Acute upper respiratory disease, b) Asthma, c) Chronic bronchitis, d) Cardiovascular disease (CVD), e) Heart attack, and f) Ischemic heart disease (IHD).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. AdditionalExcelFile.xlsx AdditionalTextFile.docx