As COVID-19 continues to spread around the globe, people are struggling to establish best practices for themselves and their communities. Amid a cacophony of hearsay and contradictory directives, there are timely and verified real data from South Korea (a country that, as of early April, seemingly had passed its peak of infections) that can provide guidance to policymakers and building owners as well as the general public.

This article compares two buildings at the center of viral outbreaks in Korea and provides information on where COVID-19 clusters have occurred. By presenting this research, based on my review of recent data from the Korea Centers for Disease Control and Prevention (KCDC) and local news reports, I hope to contribute some hard facts to the discussion concerning building systems and COVID-19.

The first building is Shincheonji Church of Jesus in Daegu, South Korea, which was determined to be the epicenter of the COVID-19 epidemic in the country. Built in 1991, the church is a nine-story concrete structure with one basement floor. Each above-ground floor covers an area of 934 to 985 square meters (10,000 to 10,600 sq ft) and the basement covers 1,168 square meters (12,572 sq ft). According to a press release issued by the first confirmed infection case from this group, identified on February 18, was a 61-year-old woman who attended church services on the mornings of February 9 and 16, temporarily leaving the hospital where she had been staying since February 7 for cold-related symptoms.

The near-zero temperature (32 F.) at the time of the services, coupled with conditions inside the church, likely created an environment conducive to contagion. Based on news reports, the church was crowded with hundreds of people sitting close to each other during services. Moreover, based on publicly available photos, the number of air terminals appears to be inadequate relative to the density of the people and the size of the space; this would provide an inadequate amount of fresh air, says Wang Yan-Ping, a building management system expert currently finishing his PhD in civil engineering at Stanford University. And research shows that low humidity makes the human body more susceptible to getting infected by viruses.

It is important to note that while these building conditions likely permitted the virus to spread, social interaction among affected church members and the church’s verbally expressive form of worship also were contributing factors. In addition, it was reported that the 61-year-old woman did not wear a mask during services because the wearing of masks was not allowed.

As a direct result of the outbreak at this Shincheonji Church, 4,400 cases of coronavirus have been corfirmed as of March 28. As of March 28, Daegu—the country’s third-largest city with a population of about 2.5 million—had 6,610 confirmed cases, representing 69 percent of the 9,583 confirmed cases in the country. There were 5,089 cases of coronavirus among Shincheonji members (from this and other church locations) who tested positive, representing 53 percent of the country’s total.

The second building I want to discuss is a 19-story structure located near two major subway stations in Seoul’s Guro District. The building’s bottom six floors are used for a wedding hall and café, the seventh to 11th floors are used for call centers and offices, and the 13th to 19th floors are used for small office/home office (SOHO) units. The building also has six basement floors. The outbreak in the building was primarily concentrated in an insurance call center that took up the entire 11th floor, with roughly 1,103 square meters (12,000 sq ft) of floor area. The call center had about 200 employees, a combination of call operators, document processors, and managers.

The first confirmed case occurred on February 25, and by March 8 about 60 additional workers in the call center had confirmed cases. Concerned about a potential building-wide contagion, the government authority shut down the building on March 9 and over several days tested all 1,143 workers, residents, and visitors. When news of the outbreak was first reported, there was a great amount of concern that other people—especially workers who shared the elevator to odd-numbered floors and subway passengers at the two nearby stations—might be infected.

After a two-week investigation, on March 25, the KCDC announced the results of its investigation. The KCDC’s report said that as of March 25, the building had 97 confirmed cases of COVID-19, consisting of 94 cases from the call center on the 11th floor, two cases on the 10th floor, and one case on the ninth floor. In addition, 61 contacts of the 97 infected people were confirmed, including 34 family members, five acquaintances, and 22 members of a church. According to the March 25 investigation report from the KCDC, the data showed a high infection rate of 43.5 percent for the workers in the call center, while the infection rate was 7.5 percent for the offices on the 10th floor and 0.5 percent for another call center on the ninth floor. The KCDC’s report states that the virus spread by repeated exposure to droplets.

Dr. Jeong Eun-kyung, head of the KCDC, said on March 25, “Based on the fact that there were nearly no other infected people on other floors, we have confirmed that the probability of infection through sharing an elevator or an office lobby for a brief period as part of daily life is low.” She also identified the high density and inadequate ventilation in the call center as likely contributing factors for the contagion. “In an enclosed, crowded, and inadequately ventilated workplace [i.e., the 11^th-floor call center], there was a very high infection rate.”

Within the call center, proximity may have had an effect—there was a significantly higher rate of infection among call operators occupying one half of the floor versus managers and document processors occupying the other half. It was reported that the operators did not wear masks, given the nature of their work—speaking into headsets.

Based on information provided by the KCDC on March 28, here is a breakdown of the types of buildings identified with confirmed cases:

(Note: This table does not include contacts of confirmed cases, imported cases, and cases still under investigation. The figure for Shincheonji Church represents multiple church buildings, though the Daegu church represents the majority.)

By no means is the COVID-19 crisis in Korea over. The country is still experiencing about 100 new confirmed cases each day, many from Koreans returning from overseas looking for a safe haven. Also, despite the government’s plea, some churches continue to hold services, and most recent cluster outbreaks have been in churches. News reports of such outbreaks often cite crowding and expressive forms of worship as potential causes of transmission. So far, there have not been any known outbreaks in Korean movie theaters or subways. (When infections took place in mid-February to early March, cinemas and subways were still open, but there was a decrease in the number of users.)

It is important to note that since COVID-19 was first detected in Korea on January 18, the country has been on high alert, and the government has strongly encouraged that all people, even young and healthy people, practice proper personal hygiene—washing hands, disinfecting surfaces, not touching their face,and so on—and wear masks when outside their homes.

My review of the data from the KCDC and news reports clearly supports the need for social distancing. Many of the reported cluster outbreaks in South Korea involved people engaged in social activities, especially talking close to each other for prolonged periods of time. Affected people have often been in crowded indoor venues with inadequate ventilation and possibly low humidity, as it is often the case in the middle of Korea’s cold winter.

Obviously, this article does not constitute a sufficient study of all available data provided by the KCDC. Yet it is clear that the rich and granular data gathered over the past three months and made publicly available by the KCDC can be valuable resources (www.cdc.go.kr/). The data can help us understand how COVID-19 has spread in building environments and can inform government leaders, building maintenance experts, and building owners about how they can respond.

Note: The author wishes to thank Yan-Ping Wang, a PhD student in civil and environmental engineering at Stanford University and CEO of iDRC, a Shanghai-based building management system consultancy, for his help in reviewing the article and conducting the analysis of the conditions inside the Shincheonji Church.