Energy Service Security for Public Health Resilience in Michigan’s Western Upper Peninsula

Publication Date: 2023

Abstract

The Western Upper Peninsula of Michigan is characterized by long winters, legacies of the extractive mining economy, and the infrastructure features of extreme rurality—including aging housing, low health service density, and very high electricity prices. This project examines the extent to which health facilities in this region are prepared to navigate the increasing intensity, severity, and occurrence of severe weather that is likely to disrupt energy services, as well as how they can continue to provide energy services to support public health in communities during and after disruptions. This exploratory study uses a mixed-methods approach combining qualitative interviews with public health stakeholders and survey data from health facilities. We found that energy services are considered essential; however, given the self-reliant culture of the community and the perception of a limited risk of disaster impacting energy services, disruptions are not considered an urgent threat. Health facilities consider oil and gas as the most reliable backup energy supply in case of electrical grid failure. The self-reliant nature of the community, perceived as figuring out solutions to problems as they arise, could be both a resource and a challenge in case of unanticipated disaster. This could be challenging because the rural region is at the tail end of infrastructure, and extreme weather conditions risk energy service disruption impacting public health. Considering the critical role of reliable energy services in community health, these findings indicate that public health policymaking should use an energy service approach to build resilient communities in the face of climate-induced disasters.

Introduction

Energy use is central to human health, with access to electricity serving as a fundamental social condition that promotes and maintains the health of people and their communities. Security of energy services shapes public health outcomes both directly and indirectly (Cook et al., 2008¹; Khan et al., 2020²). For instance, electricity plays a critical role in supplying essential healthcare services, such as providing lighting for medical facilities, power for life-saving electrical devices, and refrigeration for medication and essential foods such as breastmilk. Disruptions in electricity, such as during acute disasters that impact energy grid operations, place communities at risk for poor health outcomes (Rose, 2007³; Dargin et al., 2020⁴). This is most evident in the case of vulnerable communities, both within the United States and worldwide, where inadequate infrastructure and/or economic resources limit energy services access.

This study is an exploratory evaluation of the energy services needs of health facilities (hospitals) in six counties of the Western Upper Peninsula. The region is historically, geographically, socially, and culturally unique and is significantly at risk of adverse disaster impacts. Here, single-season snowfall averages over 300 inches in parts of the peninsula (Manzullo & Siacon, 2019; Keweenaw County, 2022⁵). Households in this region also pay the highest electricity rates in the contiguous United States, despite having a lower-than-average income level for the state of Michigan (Prehoda et al., 2019⁶; Michigan Public Service Commission, 2023⁷). Consequently, people in the peninsula are particularly vulnerable to climate change related hazards and disasters that impact access to energy services.

Through interviews and surveys with public health stakeholders in the region, this project identifies current energy services provision and vulnerabilities in healthcare facilities. More specifically, we focus on the preparedness of these facilities to navigate climate-induced disasters, such as snowstorms and flash floods, that have a high likelihood of energy service disruption. Disruptions to energy services provision have significant public health implications and we seek to identify opportunities to improve community energy services access (Cox, 2021⁸; Chobanov, 2021⁹).

Literature Review

Public Health

Historically, health in a community was examined by understanding the rates of diseases, such as diabetes, cancer, heart disease, or stroke, as well as looking at personal health behaviors like tobacco use, physical activity, and eating habits and their impact on overall health (Winslow, 1920¹⁰; Link & Phelan, 1995¹¹; Susser et al., 1985¹²). However, with more evidence-based research, it has become increasingly clear that while these factors are extremely important, focusing on specific disease prevention and management alone is not enough to create healthy and resilient communities—it also requires looking at social indicators, such as in income, age, and class (Hill-Briggs et al., 2021¹³; Braveman & Gottlieb, 2014[^Braveman & Gottlieb, 2014]).

Public health aims to protect and improve health of people and communities and is concerned with protecting the health of entire populations (Centers for Disease Control and Prevention [CDC], n.d.¹⁴; Link & Phelan, 1995). These populations can be as small as a local neighborhood or as big as an entire country. Much of public health focuses on how socioeconomic factors are linked to health and risk factors (Link & Phelan, 1995; Susser et al., 1985; Wells et al., 2017¹⁵). Link and Phelan (1995) asserted that it is essential to contextualize social conditions that may be the fundamental cause of diseases. These social conditions include socioeconomic status, gender, community dynamics, and, the focus of this project, energy security and energy access. Social conditions play a causal role in illness, and illness affects social conditions; thus, increasing medical care alone cannot adequately improve public health outcomes—the social conditions that impact illness must also be addressed (Johnson, 1991¹⁶; Braveman et al., 2011¹⁷).

The social determinants of health (SDOH) public health framework considers the conditions in which people are born, grow, live, work, and age, as fundamental drivers of the health of individuals and communities (Braveman & Gottlieb, 2014; Wells et al., 2017). The Centers for Disease Control and Prevention (CDC) outlines SDOH in five areas: economic stability; education; social and community context; health and health care; and neighborhood and built environments (CDC, n.d.). Broader community health assessments incorporate data on poverty, education, employment, etc.

The link between SDOH and health inequalities is exacerbated during and after a disaster (Marmot & Allen, 2014¹⁸). In the current Anthropocene (time period when human activities have impacted and changed the earth’s environment), the increasing effects of disasters on global public health are widespread (Leaning & Sapir, 2013[^Leaning & Sapir, 2013]). Climate change exacerbates the frequency and intensity of hazards, such as extreme heat/cold events, floods, droughts, and the distribution of vector-borne diseases (Haines et al., 2005¹⁹). Accordingly, public health strategies must focus on the critical role that such impacts have on population health.

Public health can help ensure that living conditions can support healthy lives, including playing a critical role in addressing disaster preparedness and post-disaster needs (Shoaf et al., 2000²⁰). Ensuring disaster resiliency involves protecting against environmental hazards, preventing injuries, responding to disaster, and assisting the community in post-disaster recovery (Tiwari et al., 2022²¹). Policies with consequences that impact human health are not created in a vacuum and require convergence across multiple fields of social and behavioral sciences, public health, and hazards and disaster research (Peek et al., 2020²²). Hence, there is an increasingly important emphasis on interdisciplinarity in hazard and disaster research, with implications for public health (Peek & Guikema, 2021²³). In this project, we consider the role of energy services access within healthcare facilities during acute hazards and disasters and bring a novel consideration and approach to studying SDOH.

Role of Energy Services in Public Health

Electricity is an instrumental good, meaning humans do not need the electrons that flow through the electric circuit but rather the multiple services derived from the electricity (Fell, 2017²⁴; Tiwari et al., 2021²⁵). Yet, few existing studies examine electricity as a means to different services, instead focusing on electricity as the end goal itself. The electrical service needs of any organization or facility increase as the number of functions or services increases and becomes more complex, and this is particularly true for healthcare facilities (Federal Emergency Management Agency [FEMA], 2019²⁶). Energy services are central to functional, modern healthcare facilities and for community resilience (Smith et al., 2013²⁷; Hunt & Ryan, 2015[^Hunt & Ryan, 2015]). People rely on electricity for a multitude of services, such as thermal comfort, refrigeration, communication, and lifesaving devices (Tiwari et al., 2022). Patients of healthcare facilities often require acute medical care that heavily relies on electricity (FEMA, 2019; Skarha, 2021²⁸). Individuals with functional access needs or chronic health conditions rely on power-dependent durable medical equipment, such as powered wheelchairs and ventilators. Some individuals also have implants, such as cardiac devices and blood pressure monitors, that must be plugged in to recharge or require internet connection (U.S. Department of Homeland Security, 2017²⁹). Individuals, particularly immunocompromised or older populations who cannot regulate their own body temperatures, are especially dependent on climate control, such as air conditioning (FEMA, 2019).

Healthcare facility (hospital) buildings rely on electricity for almost all their operations, including but not limited to lighting, security systems, fire alarms, environmental controls, electronic health records, refrigeration, and an array of electricity-dependent durable medical equipment and devices to provide care (FEMA, 2019; Bawaneh et al., 2019³⁰). In addition, they provide energy-dependent support, including food, water, and transportation (Western Upper Peninsula Health Department , n.d.,³¹ 2021³²). These facilities may also be required to meet the energy service needs of local communities during and after a disaster, when people do not have energy access in their homes but require lifesaving energy services. These requirements, and the fact that healthcare facilities generally cannot temporarily suspend operations, make these facilities uniquely susceptible to infrastructural failure due to power outages or inadequate supply or distribution of oil and gas, which can be the difference between life and death in the event of disaster (FEMA, 2019; Casey et al., 2020³³). Without power, facilities may require evacuation, which carries significant dangers in rural areas that are difficult to access during winter or other harsh weather conditions, such as the Western Upper Peninsula.

Different types of end-use energy entail some level of health risk, and public health experts have focused on the public health impacts of different energy sources and their local and global disease implications (Smith et al., 2013; Rabl et al., 2000³⁴). To date, research on the connection between public health and energy has primarily focused on the effect of fuel use type on the health of the community. For instance, there is substantial research on the effect of indoor air pollution on women and adolescent girls due to biomass used for cooking (Khandelwal et al., 2017³⁵; Gall et al., 2013³⁶; Pratiti, 2021³⁷; Rabl et al., 2000; Wilkinson et al., 2009³⁸). Within the context of disasters, people may also rely on biomass when there is limited access to relatively cleaner end-use energy fuel, such as gas cylinders or electricity for cooking (Ali, 2007³⁹; Lask et al., 2017⁴⁰). Aside from this limited research related to public health and energy, few studies examine the role of energy services in determining the resilience of health facilities and healthy communities more broadly (Modi et al., 2005⁴¹; World Health Organization, ⁴²; Khandelwal et al., 2017).

Secure and reliable energy service access is paramount to community preparedness, resilience, and disaster response, and more research is needed on energy in the context of public health in rural areas (Casey et al., 2021). Solutions like temporary power through generators only address the problem in the short term and fail to account for the potential for long-term outages in remote, isolated communities like those in the Western Upper Peninsula. This project addresses this gap in the literature by examining how professionals working in public health facilities in this region perceive the energy services risks they face and the potential pathways they could use to provide energy services security for rural communities in the face of increasing hazards and disasters.

Public Health and Energy Service Resilience in the Upper Peninsula of Michigan

The Western Upper Peninsula is prone to extreme weather events, experiencing over seven months of long winters with snowfall averaging over 300 inches (Keweenaw County, 2022). Winter temperatures frequently go below 10°F during winter months (National Weather Service, 2018⁴³) and the area is increasingly impacted by extreme weather events such as polar vortexes; in 2019, the temperature dropped below -6°F with wind chills of -25°F to -30°F (Manzullo & Siacon, 2019⁴⁴). More recently, a major disaster struck Houghton County—named the “Father's Day flood” as it occurred during Father's Day weekend of 2018 (Blackmon, 2018⁴⁵; National Weather Service, 2018; Keweenaw Report, 2018⁴⁶). Over seven inches of rain fell in a three-hour period. The estimated damages were $30 million to the public infrastructure, creating extreme financial and service access challenges for the community (Keweenaw Report, 2018).

The geographical location, demographics, and the historical legacies of extractive copper and iron mining industries impact the health of the people living in the peninsula. The region is one of the most rural, northern places in the Great Lakes region of the United States and home to the oldest recognized Tribal Nation in Michigan, the Keweenaw Bay Indian Community. It is also an ageing community, with nearly 20% of the non-incarcerated population aged 65 or older, compared with 15% statewide. In two of the six counties in the study, more than 30% of the population is 65 years or older (U.S. Census Bureau, 2010⁴⁷). Since the chronic disease burden is higher in older adults and they have greater needs for home health services, assisted living, and nursing home care, a shift in a community's age distribution to older cohorts has profound implications for their health care needs. This becomes even more true during and after a disaster. Furthermore, the population of these counties is declining, which may exacerbate the lack of resources and regional stability. Table 1 summarizes the demographic changes in these six counties between 2000 and 2015.

Table 1. Population Changes in the Six Western Upper Peninsula Counties in Michigan

Before the implementation of the Affordable Care Act in 2014, 19% of peninsula residents aged 18 to 64 did not have health insurance. By 2017, that rate had declined to an estimated 7% due to Michigan's Medicaid expansion and the newly created health insurance marketplace. Each of the six counties also has areas with Health Shortage Population Area designation. This designation is given when a place, population, or facility meets standard federal criteria for shortages that may be geographic, population, or facility based. It is also a Health Professional Shortage Area, indicating that they experience provider shortages in primary care, dental health, and mental health. Table 2 provides additional demographics demonstrating the region's population vulnerability.

Table 2. Select Social Statistics of Six Western Upper Peninsula Counties

The Western Upper Peninsula is at an elevated risk of power system failures due to its cold weather, aging infrastructure, and location at the far end of the electrical power system (Chen et al., 2005⁴⁸; Casey et al., 2020). This is a matter of particular concern for the region’s healthcare facilities, which experience the challenges of serving under-resourced individuals and communities within the context of the ever-present threat of disasters. Severe and changing weather is a real threat to energy services resilience and overall grid reliability. Given the potential consequences of infrastructure failure, especially for anyone who relies on electricity-dependent equipment, ensuring the resilience of energy service systems is paramount to protecting public health (Casey et al., 2020).

In addition to its limited and underfunded health facilities, the peninsula falls victim to the urban-rural health disparities that plague much of the rural United States. (Leider et al., 2020⁴⁹). As a rural environment, its residents have worse health outcomes and higher age adjusted death rates than their urban counterparts (Harris et al., 2016⁵⁰; Moy et al., 2017⁵¹). The peninsula also experiences high patient-physician ratios, making what care exists within the region more difficult to access. These factors, combined with the limited energy services security seen in the region, are a matter of concern for residents and planners of the region alike, warranting research attention.

Research Questions

This project seeks to understand current energy services provision, needs, and vulnerabilities of health facilities in the six counties of the Western Upper Peninsula. Understanding these dynamics is key to improving community energy services access, which directly impacts public health. Through this transdisciplinary study, we aim to provide health facilities with a framework to identify indicators of vulnerability as well as opportunities and translational tools to inform decision-making around future energy development and energy services prioritization. The aim of the project is to answer the following two research questions:

1. To what extent are health facilities in the Western Upper Peninsula prepared to navigate the increasing occurrence and intensity of storms with a high likelihood of power system disruption?
2. How can an evaluation of preparedness help inform decision-making that can enhance energy service access and promote health resilience in rural and Tribal communities in this region?

Research Design

This project utilized a mixed methods approach (Creswell, 2014⁵²) that included interviewing relevant stakeholders and implementing a short survey to understand the energy service needs of health facilities in the six-county study region: Baraga, Gogebic, Iron, Houghton, Keweenaw, and Ontonagon. The interdisciplinary nature of this study aligns with calls for more convergent research involving the social and behavioral sciences, public health, and hazards and disaster research, which warrants the use of mixed methods for studying issues of social vulnerability in diverse settings (Adams et al., 2022⁵³; Peek et al., 2020; Peek & Guikema, 2021).

Accordingly, interview data were first collected to gather insights from individuals in the public health and healthcare sector related to their facility's energy service resiliency and priorities in disaster situations. Given the significant impact of disasters on public health and the unique physical and social conditions of the Western Upper Peninsula, interviews were chosen as the primary way of data collection, given their flexibility and the ability to gather detailed information from each respondent on potentially sensitive topics. In total, 14 people were interviewed. Following the interviews, an online short survey was utilized to collect data related to the health facility operations in each of the counties. The survey was utilized to ensure the same data was collected about each facility and to protect facility anonymity. The survey does not identify the facility and asks only basic demographics of the responding party. In addition, the survey does not ask questions related to disasters or traumatic events, which have been shown to potentially cause harm or distress to respondents than use of more personal data collection methods on these topics (Adams et al., 2022).

Interviews

Sampling Strategy

The interview participants belonged to two different stakeholder groups. The first group consisted of professionals working in hospitals in the Western Upper Peninsula and the second group of professionals working more broadly in public health, including at a health foundation, an emergency management agency, a planning department, and a state service commission. Two separate interview protocols were used (see Appendices A and B) and individuals were selected based on their role within the participating organizations.

The interview team reached out to health facilities, including hospitals and county or tribal health departments, as well as broader planning stakeholders, such as emergency managers and health related non-profits in the region. Certain facilities, such as long-term care facilities and outpatient facilities, were excluded from initial data collection. Given the nature of this study, it was necessary to select initial organizations and individuals directly rather than take a random sampling approach to ensure participants were qualified to be interviewed on the topic. Utilizing a snowball sampling method, initial participants were asked to share the names and contact information of any other potential participants. This aided in allowing the research team access to additional interview participants, which may have been otherwise difficult to access (Babbie, 2015⁵⁴; Creswell, 2014; Naderifar et al., 2017⁵⁵).

Participant Recruitment and Consent

Based on identified stakeholders, 23 individuals were initially contacted for interviews. Of these individuals, 11 qualitative, semi-structured interviews were conducted with 14 individuals. Three interviews were joint interviews at the request of the participants. A listing of interview specifics can be found in Appendix C.

Potential participants were contacted primarily over email in three major waves. Participants were contacted with an initial email introducing the project and the possibility of being contacted for an interview, and they were provided an opportunity to redirect project communication if they felt someone else in their organization would be better suited to the interview. A subsequent email provided information about scheduling through an online Calendly link that was accessible by one member of the project team. A follow up email was sent to individuals who had not scheduled an online interview within one week of the second email.

Some interview respondents were unable or unwilling to be contacted via email and were instead reached via phone, text, or in person following the initial wave of emails. The project team attempted to contact all possible participants via phone call if they had not responded to email communication, using phone numbers that were publicly available or provided by another participant. The same approach was followed with participants who responded late, although they were asked to schedule via email or other forms of communication due to shorter turnaround times. One individual declined to participate in the study via phone call.

Participants were provided copies of the intended questions and the consent script via email. The consent script was additionally read aloud at the start of each interview to ensure each participant was able to provide informed consent. Participants were asked to verbally agree or disagree with participating in the interview. Participants were additionally asked to consent to the recording of the interview. All participants consented to recording.

Setting and Other Details

Participants were allowed to select their interview time and location when possible. All interviews were conducted in private locations via Zoom, telephone call, or in person. Interviews were slated for a one-hour time period and ran for 30 to 60 minutes in nearly all cases. All participants were offered a $50 gift card to the business of their choosing for their participation, which was mailed to the participant following the interview. Some participants declined this payment on the grounds of their employment terms.

Interviews began with introductions to the project team and an overview of the intended interview structure. Interviews were primarily conducted by graduate fellow Ketola of the project team, however, Tiwari and Cole attended and conducted interviews when possible or necessary. Interview questions were then asked from pre-determined sets depending on the participants' backgrounds. Follow-up questions were asked as appropriate and informed by participant responses to initial questions.

Data Analysis Procedure

Recordings and transcripts were cleaned of identifying or superfluous information and stored in a secure cloud server only accessible by the project team. Recordings and transcripts were numbered and stored without identifying information. All participants were provided an opportunity to review and revise their interview transcripts following transcript cleaning and in advance of data analysis. A project team member analyzed interview transcripts using the qualitative analysis software Nvivo, which other team members reviewed for consistency. Codes were developed iteratively and contained both parent and child codes to explain the data more efficiently. Codes were developed from the literature review and the interview questions, and some were developed based on the frequency of appearance in transcripts. Matching the literature were codes such as “weather event” and “outages,” which were concerns mentioned in the literature when exploring energy services security. There were also codes about how the facilities were handling outages and electrical disturbances, compiled under the parent code of “emergency systems.” Under that parent code, codes such as “backup generation” and “emergency planning” gave a clearer idea of how these health facility experts are thinking about handling the potential for disruption of services. Parent codes were primarily selected for ease of organization, grouping child codes that related to similar concepts within the literature and the interviews. The research team developed all codes, with no automatically generated codes used for analysis.

Survey

An online survey through SurveyMonkey was sent to interview participants who worked as administrators at local health facilities. This survey included questions about the size of the healthcare facility, information related to its electricity consumption and outages, and backup power and heating sources. There are six health facilities in the study context (one for each county) and survey responses were received from four of the six facilities. The relatively short survey was conducted to compliment interview responses from the health facilities. The survey instrument is included in Appendix D.

The survey was distributed via email to specific personnel within health facilities. Given the questions' nature, there was no reason to distribute the survey via other means or to other populations (including the second stakeholder group, representing professionals who work in public health broadly rather than at specific health facilities). Additional future work may include expanding survey distribution to other types of health facilities in the general study region that were not included in initial interviews, such as those outside of the Western Upper Peninsula, long term care facilities, or outpatient care facilities.

Participant Consent and Other Details

Health facility administrators or managers were asked to consent to receiving the online survey following their interview. Participants who agreed were provided with a survey link via email. The survey did not ask questions with identifying information about the individual nor about the health facility they were affiliated with.

Ethical Considerations, Researcher Positionality, and Reciprocity

Lead Investigator, Tiwari, and two project fellows, Ketola and Cole, have completed ten CONVERGE Training Modules. Co-lead Schelly has many years of experience with social science data collection and community engagement. Project fellow Ketola is from the Upper Peninsula. The team ensured that this research was co-produced, non-extractive, and conducted using best practices in the ethical conduct of research. The project was reviewed and approved by the Michigan Technological University Institutional Review Board (IRB) and the data was collected only after the approval date of November 11th, 2023; co-lead Schelly chairs the Michigan Technological University IRB Board and thus has experience to help ensure that materials are prepared appropriately.

As it is essential for researchers to go back to the community and share the findings of their research, we are working with our local collaborative partners to ensure that dissemination reaches beyond project participants to the most relevant stakeholders in the region and the state. We organized a resiliency workshop on May 10 and 11, 2023, and invited stakeholders to participate and provide feedback on the findings on the work. The workshop was open to a broader range of stakeholders than the initial study participants. The workshop was organized in person but also included an electronic (Zoom) component to accommodate the participants who could not attend the event in person. The workshop provided an opportunity to propose pathways to improve the preparedness of the community toward enhanced energy service resilience. It also provided a platform to develop a protocol from this study for other scholars to inform applied research on energy services’ role in public health resilience.

Findings

The interview analysis revealed several themes related to the 14 initial codes. Codes were divided into two categories: a) energy service needs, risks, and concerns, and b) energy sources and the provision of services. Some interview questions addressed these categories directly; however, some participants made mention of the topics independently.

Energy Service Needs, Risks, and Concerns

When discussing energy service needs, respondents outlined that they are dependent on three major types of end-use energy. Electricity use was primarily outlined for appliances, particularly for lighting, refrigeration, and emergency operations. Gas was described as being utilized for heating or to provide thermal comfort. Oil, the third type of fuel, was said to be largely stored and utilized for backup electricity generation. Survey results showed that most of the backup energy supplies are diesel generators or natural gas generators, with no large-scale battery storage in any of the facilities. General responses related to electricity carried a certain air of hesitancy about the electrical system, citing the limited number of electricity providers in the region, aging infrastructure, and uncertainty about the ability of utility companies to deal with outages. One respondent stated:

We have three to four different power companies that run through the county. They're supposed to be redundant. I mean, the grid is allegedly set up so that if one section goes down, it can reroute the other way. It has happened in the [Upper Peninsula] where that system is not very robust.

Another respondent said:

The utility companies were saying, well, it is a double circuit line, but they all go through the same poles. It was one truck driver away from plowing into one and the lines, and it could take us a month to replace that line.

Respondents from health facilities further outlined that they could meet energy service needs of the premises in case of electrical grid failure using the backup electricity generator. While each of the respondents provided estimates for the range of days for which they can meet all electricity service needs of the facility, they also highlighted that there are no known guidelines for how many days they could meet their peak energy service needs from fuel stored for backup electricity. The standard for storing oil and gas in health facilities to provide backup electricity service needs to be further explored to ascertain the energy service needs during a disaster.

Respondents cited electrical outages as their most common energy disturbance, and many stated that the outages were typically weather related. They considered these outages to be low impact and infrequent, referring to them as “flickers” and “minimal.” This was also validated through survey response where all the outages were less than 30 minutes and only one of them over 15 minutes. Multiple respondents pointed out that it had been years since the last long-term power outage they remembered, so they were relatively unconcerned about them. Some respondents expressed more concern about the secondary effects of weather-related outages, such as increased use of indoor heating elements triggering house fires and other public health issues. When it came to weather, most were concerned about the ability to travel to health facilities or work, with one interviewee stating that following flooding, “our biggest challenge was getting people to work because so many roads were washed out.”

Respondents expressed confidence in the ability of health facilities to cope with outages, referring to facilities as “very well self-contained” and stating that “the hospitals at least have a backup power supply and probably back up heat supply.” Health facility personnel shared that backup power at their facility could last days to weeks and that testing procedures ensure the operation of these backup systems. Testing of backup power was something that health facilities did independently to comply with a health facility accreditation company and not because they were required by the state or federal government.

Respondents expressed little faith in their ability to access resources from the state of Michigan, such as disaster relief equipment or medical equipment. They pointed out the distance between the Western Upper Peninsula and equipment located “below the bridge,” a colloquialism that many residents use to refer to Michigan's lower peninsula as it is past the Mackinac Bridge. To cross the bridge by car from Houghton, Michigan takes upwards of 5 hours. During a recent major disaster event in the Western Upper Peninsula and other smaller events that have happened since, emergency managers mentioned using what they had access to in the region, be it police cruisers, county vehicles, or even personal vehicles, citing that it could be hours or days until they received state support. The counties are highly connected and appear to rely on each other for support before anyone else. Respondents talked about their collective self-reliance, seemingly out of necessity. Interviewees described how following a flood event, many residents cleaned up their own property before aid groups could even make it to the scene, and others refused help, seemingly believing there were bigger issues than theirs. One interview respondent stated:

That's just kind of stubbornness in our community. There is kind of a mindset, maybe, that in a disaster, well, my injury isn't that bad, which makes me wonder how many other injuries [there were] that were never reported because [people] did not think it was a big deal.

Others expressed concern about the ability to contact community members if power was lost, expressing the difficulty of contacting emergency services in many parts of the peninsula or sending out emergency bulletins in communities with limited cell phone coverage and aging, self-reliant populations.

Energy Sources and the Provision of Energy Services

Few respondents made mention of renewable energy resources without prompting. When asked if their healthcare facility had considered renewable energy for backup power, several facilities had, but only one was considering renewable technology installation. Others cited that the cost was too high for them to pursue installation. Respondents also highlighted that they would be interested in pursuing renewable energy if there was grant support from the state. Some facilities were planning to rely on trucking in new equipment for backup power should their primary backup generators fail, which can take the better part of a day if the weather is good. Other interviewees, however, pointed out that there are limited locations with backup power beyond the hospitals.

Broader responses on energy included a marked focus on energy conservation and fuel reliance (gas) in the region for standard and backup heating and power. Two participants made mention of energy conservation as a means of resilience and alternative power support, specifically referencing LED lighting conversions in one instance and LED lighting conversions plus temperature control in another. Facilities that possessed backup power did so through generators that ran primarily on diesel or propane, introducing additional complications should supplies expire, prices increase, or if more fuel cannot be brought in for lengthy outages. Other energy sources, like natural gas, are widely used and regarded as stable, given that they are piped into the community. Some respondents indicated that they believed increasing natural gas infrastructure was a solution to the region’s energy resilience. However, one respondent brought up an incident in the recent past that led to a natural gas system failure in a Western Upper Peninsula community. Another individual pointed out the possibility of natural gas pipeline failures due to the system's proximity to major roadways in the peninsula, stating, “Think of an ice storm scenario. It is icy, knocks the power out, and somebody goes off the road and hits your natural gas station. That is a real, real possibility.”

When asked about perceived energy risks and concerns, respondents expressed concern for vulnerable populations in the community, such as older adults, more so than access to power. Some expressed frustration about the inability to access local and state resources during outages and the lack of resources to deal with a large-scale disaster requiring triage and mass medical care. Staffing shortages, limited access to medical equipment, and lack of health facility capacity came up in multiple interviews. A potential issue cited by multiple respondents was uncertainty about the response if the local lift bridge connecting Houghton and Keweenaw counties was compromised. Both major hospitals in the area are located north of the bridge, and access would be severely compromised without bridge access. The next hospital is located over 30 miles away, and the next large hospital is nearly 100 miles away. The concerns for public health related to other service provisions in the region, which can be lacking.

Respondents questioned the ability of local systems to provide services or withstand disasters and outages. Their proposed solutions often relied on their local connections to deal with challenges or problems rather than institutions of public health or safety in their communities. Existing plans for backup power during more extended outages sometimes relied on agreements between local facilities, like health facilities or local schools, to ensure there was a place with power and heat for community members to go. Larger health facilities expressed that they could serve as a place for community members to shelter during disaster events; however, they were not sure of the number of days they could provide shelter and the kind of services that they would have to offer. Other interview respondents expressed disbelief that this would work unless there was a personal relationship between the facility administrator and a community official or manager. Respondents repeatedly cited the importance of their community relationships and the small community size when discussing disaster or hazard situations.

Conclusions

Public Health Implications

The provision of energy services, particularly electricity as an end-use energy service, is of prime importance for public health. The geography and rurality of the Upper Peninsula leads to a unique dependency on imported fuel, such as oil and natural gas, and its long-term storage to provide energy services in the case of service disruption. This study finds that while the public health facilities serving the region do not express significant concern about meeting energy services needs in case of an outage, they are not proactively planning to serve as an energy service provider for the community if needed during a disruption. The cultural context seems to create a perception of localized self-reliance, leading to a relatively informal approach that does not emphasize proactive planning.

Quantifying the energy services that health facilities can provide using backup fuel is of critical importance for the availability of services during and after a disaster. Currently, there is limited understanding of how many days and what kind of services health facilities can provide in case of a disaster that impacts the region for a longer duration. The self-reliant nature of the community, in which the preference is to figure out solutions when problems arise, could be challenging in the case of unanticipated disaster, particularly because of the region’s poor infrastructure and risk of extreme weather conditions that can disrupt energy service. Furthermore, this perception of self-reliance has the potential to exacerbate existing inequities and dynamics of social exclusion.

As rural communities prepare for the increasing intensity and severity of extreme weather events associated with the ongoing climate catastrophe, centering the essential role of energy services in promoting and maintaining public health is key to building resilient communities. In the context of the Western Upper Peninsula, how to address needs associated with heating energy services during a longer-term disruption was identified as particularly important and not well understood. Thermal comfort is of particular concern for those who are already experiencing health vulnerabilities, such as older adult populations. There are innovative technological options for both heating and cooling as well as for necessary electrical energy services. However, these are not being considered by the rural health facilities engaged in this work, largely because of concerns about cost, performance, and a sense that proactive planning for localized energy security is not necessary given the community’s ability to work together at a very local level to address needs as they arise. These cultural, infrastructural, and institutional dynamics intersect to exacerbate vulnerabilities for public health in this rural community.

Limitations and Future Research Directions

Due to the nature of the study and the compressed time period, one of the key limitations was that we did not quantify electricity demand for health facilities. In particular, future research is needed to assess the energy service requirements of local health facilities, the backup services that these facilities in the region can provide, and the number of days these services can be provided. This includes quantifying electrical and gas utilities and estimating the regional energy backup for different energy service scenarios in case of disaster—a recommendation that was proposed during the resilience workshop in May 2023. There is also a need to look at state regulation and incentives for developing distributive renewable energy systems in rural health facilities. This can support self-sufficiency during emergency situations. Furthermore, our regional partner, Western Upper Peninsula Planning & Development Region office, has developed a rural hazard resilience tool, including GIS visualization tools that simulate floods and natural hazards using data from past events (Western Upper Peninsula Planning and Development Region, 2023⁵⁶). We aim to work with them in the future and map health facilities and energy service indicators using this tool to support health system equity and resilience.

References