Application of layers of protection analysis to prevent coronavirus infection

Abstract Layers of protection analysis (LOPA) methodology is applied to an encounter with the SARS‐COV‐2 infection as an initiating event, and subsequently, independent protection layers (IPLs) (namely health safeguarding protocols), such as social distancing, ventilation, hand hygiene, face masks, and vaccinations. LOPA is applied considering numerical quantification of the COVID fatality index in order to manage the transmission risk to a tolerable level, namely the fatality risk due to seasonal flu. This measurement tool quantifies the ratio of the annual death rate due to the SARS‐COV‐2 infection to the annual death rate of the common flu, and it is applied to a chemical plant. The lower this quantified value is, the more the COVID‐19 infection death rate approaches that of the common flu. Thus, any improvement in safeguarding protocols should reduce this index. The input data is based on public domain COVID‐19 infection statistical data and websites accessible in the United Kingdom. The COVID‐19 transmission rate is statistically analyzed with random number sampling to simulate the random pattern of the virus' person‐to‐person infection in the community. The success of the COVID‐19 protection protocols is probabilistic and depends on the public's compliance, which is modeled by observational surveys.


| INTRODUCTION
From a chemical engineering point of view, the transmission of the SARS-COV-2 virus is a process, and the disease COVID-19 can be managed like any other process hazard. Considering that layers of protection analysis (LOPA) methodology has successfully been applied to safeguarding process plants, it can also perform risk assessments to evaluate the relative probabilities of virus transmission, infection, and death. This is vital moving forward due to the high transmission rates of the SARS-COV-2 virus and its severe consequences. Lockdowns and heavy restrictions were imposed to help limit the SARS-COV-2 contagion spreading; however, in order to make life sustainable with these restrictions, new methods/standards need to be incorporated into day-to-day life to help manage this virus and guarantee business continuity. [1][2][3] LOPA is a semi-quantitative risk assessment methodology and well documented in literature sources. [4][5][6] LOPA requires three main inputs: "Risk Tolerability Criteria," "Initiating Event  Greenhalgh et al. 8 present evidence and guidelines for the use of face masks. The key message from the paper states that we should sometimes act without definitive evidence as a precaution. There have been widespread debates as to the effectiveness of the use of face masks; however, it can be agreed that even limited protection is beneficial and could prevent some transmission of this disease, which, in turn, would save lives due to COVID-19 being such a serious threat.
Therefore, wearing masks in public should be advised. The World Health Organization also provides guidelines on the use of face masks for children and adults. 9 A report written by the European Centre for Disease Prevention and Control defines targets and instructions for the use of face masks and hand hygiene methods. 10 Increases in the encouragement of hand hygiene is also recommended in the following reports and have been supported with scientific facts. 11,12 The latter reference states that when hand washing is carried out, it is essential to limit skin damage by the use of a moisturizer each time the hands are washed.

Research and investigations have been conducted on whether
there is an association between hand hygiene and COVID-19 transmission. 13 Their conclusion is as follows; in a population-based sample of Polish adolescents, individuals from regions of low COVID-19 morbidity presented more beneficial hand hygiene habits than those from regions of high COVID-19 morbidity.
The origin of the 2 m safe distancing rule has been investigated by Jones et al., 14 and the authors concluded that investigations referring to "safe distancing" started in late nineteenth century. Despite limitations in the accuracy of these early study designs, especially for longer ranges, the observation of large droplets falling close to a host reinforced and further entrenched the assumed scientific basis of the 1-2 m distancing rule. Computational fluid dynamics simulations have been used by Blocken et al. 15 in order to model the safe distancing for people standing still (1.5 m), walking (5 m), and running (10 m).

| COVID RISK MANAGEMENT WITH LOPA METHODOLOGY
The COVID-19 pandemic can be viewed in the same way as a typical process hazard. It is possible to apply LOPA to the issue of the virus transmission in particular settings, its likelihood of transmission considering layers of protection, or indeed multiplication, and its impact upon individuals, taking into account their demographics and state of health.
The initiating event is the frequency of encountering a person infected with the virus, and the IPLs are social distancing, free air movement (ventilation and open space), face masks, hand hygiene and the vaccine efficacy. These safeguards are collectively referred to as COVID-19 protection layer health protocols. For LOPA modeling, IPLs are used as barriers to the spread of the virus with their probability of failure on demand (PFD). A "PFD" is a probability between 0 and 1.0, with 1.0 indicating no IPL is present or 100% failure, decreasing as the probability of failure of an element decreases. The calculation does not consider common cause failure of human noncompliance with health protocols.
The input data for LOPA modeling and calculations are based on regional statistical COVID-19 infection rates and fatalities obtained from websites in the United Kingdom.

| Basis of COVID transmission rate
COVID-19 is atypical of process hazards as it is all-pervasive, often carried by asymptomatic individuals, without any obvious sign of infection. However, it is possible to evaluate the frequency of an "initiating event" defined as an "effective" contact with an infected person or the transmission rate, based on the following inputs: These factors are used to evaluate the number of effective infections per year, that is, the transmission rate, which is the initiating event in the LOPA calculation.

| COVID testing regime and asymptomatic infection modeling
The COVID-19 testing regimes are as follows: • Polymerase chain reaction (PCR) tests are sent away to a lab to diagnose the disease.
• Lateral flow tests (LFTs) can diagnose COVID-19 within 30 min of taking a sample but are not as accurate as PCR tests.
• Antibody (or serology) tests cannot diagnose active infection but can help to indicate if a person has immunity to COVID-19.
In this study, LFTs are used to account for the worst-case scenario for person-to-person infection. According to Mahase et al.,18 studies have shown that, while false positives are rare with the commonly used lateral flow test, false negatives are much more common. Three results from Public Health England showed that the test's overall sensitivity was 76.8%, meaning that 23.2% were false negatives. Sensitivity dropped to just 57.5% when carried out by selftrained staff at a track and trace center.
The pooled estimate of the asymptomatic portion of COVID-19 is 28%, which was used to calculate the transmission frequency. 19 PCR testing is more accurate than LFT and can be an input if required. Pooled For comparison and benchmarking purposes, a COVID fatality Index is introduced in the calculation, which determines the improvement measures in testing regimes and IPL compliance to bring down the cases relating COVID-19 to seasonal flu fatality level.

| STATISTICAL ANALYSIS
The transmission rate is the initiating event for the LOPA modeling.
This rate is based on the person-to-person infection rate.
The infection rate is evaluated by statistical regression analysis that is used to look at the correlation between the dependent and independent variables. This method aims to explain the dependent variables in terms of the independent variables through a mathematical relationship, to obtain a prediction of one variable given the value of the other. The regression analysis was done using spreadsheet calculation modules.
The Regression model is being used here to help look at the mean infection rates of COVID-19 in different locations, to then mathematically simulate the way that the virus infection is spreading among the population. Thus, the input data was randomly selected and fed into a regression analysis model by an analyst; however, in the future, it is envisaged to use a computer software program for the random selection process.
The COVID-19 infection is unpredictable and could be arbitrary. It is therefore required to describe and predict how the virus transmits itself in the community. Data collection and sampling with statistical models can predict the virus propagation. The infection is also random, which means it is impossible to predict future human infections based on past or present ones. The modeling, therefore, requires probabilistic assessment to account for the randomness. The statistical modeling algorithm uses an "arbitrary random population sampling" approach that is meant to randomize the virus' person-to-person transmission in the community. The mathematical calculations are designed to simulate the real-life virus transmission randomness and develop predictive tools on virus behavior in a given population sample.
For calculation of the infection rate, statistical modeling was performed on a hypothetical Chemical Plant in the United Kingdom.
Once the data was collected and each case rate was randomly assigned to a population sample, a scatter graph was produced, and then for each case, a regression analysis was performed to obtain the straight-line regression equation.
The geometric mean of the sampling population was used as the numbers in these series are not independent of each other, and in were useful tools in obtaining information, such as the case rate per 100,000 people, the total number of cases in a given area, and also the rate of change in percentage from the previous week.

| BASIS OF INFECTION RATE CALCULATION IN A CHEMICAL PLANT
For the case of the Chemical Plant in the United Kingdom, infection data from the urban municipalities surrounding the site were used.
The population of the plant is assumed to be 637, and it is surrounded by 19 local boroughs that are of a reasonable commuting distance; therefore, 19 arbitrary samples were randomly taken. Tables 1 and 2 show the "local home addresses," "random population sample," and the corresponding "infection/case rate per 100,000 people." The sample and infection rate are inputted into the regression analysis for the last 2 weeks of June 2021. It is noted that while the "random population sample" is arbitrary selected, the infection/case rate per 100,000 The upper and lower fatality limits are calculated using the method presented by Coggon et al. 23 per 1000 people infected (see eq. 4), and their geometric mean of a one person fatality is therefore evaluated and used for LOPA modeling, Table 5. Similar to process safety LOPA, for COVID-19, the probability of failure of the IPLs also needs to be calculated. Table 6 (social distancing), There are many types of face masks that are commercially available. For this study, PM2.5 Surgical Masks are used, which are more widely used, as shown in Figure 3.

| Ventilation
The main types of air cleaning that are likely to be effective at reducing infection risks include high-efficiency particulate air (HEPA) filters and ultraviolet light.
According to Reference 25, the ventilation system factors that can minimize the virus spread are filtering, the number of air changes per hour (ACH) and recirculation. Based on ventilation design information, the following "rules" are proposed to estimate the ventilation system efficacy to combat virus spread.  Note: Please refer Table  3 for calculation tables color key.

| Vaccination
For vaccination efficacy, there are numerous sources of data depending on the type of vaccines, the real-life data, and various interpretations of the results. Most vaccine manufacturers note efficacy above 90% and some between 70% and 80%; others go as low as 65%. 26   variables that can be changed in order to get the final outcome.

| Conclusions
The advantage of this paper is that the main outcome is a calculated COVID fatality index value. The magnitude of this calculated value determines how much improvement in the transmission rate variables and the safeguarding protocols should be made in order to bring down the annual death rate in parity with the common flu. This Index can be used by stakeholders in the chemical industry to manage and control the spread of infection.

| Recommendations
By now, it is well established that the SARS-COV-2 virus will be in circulation with variants in the human population permanently, as demonstrated by recent outbreaks of Omicron, Delta, or the original Wuhan strain. 28,29 The SARS-COV-2 infection must therefore be considered as a major hazard and should be treated similarly to a process plant, nuclear accident, or transportation risks. Process safety engineering has long demonstrated that probabilistic risk assessment and various qualitative risk reviews can reduce the process plants' accidental injuries and fatality risks.
It is, therefore, recommended that other process safety risk assessment tools may also be applied to the analysis of the COVID-19 infection spread. Structured process safety reviews such as hazard identifications, with relevant modifications, may be applied to identify the COVID-19 infection risks and make appropriate recommendations.
There are many other examples of process safety assessment tools that can potentially be applied to the pandemic situation, such as Performance Standards and Safety Critical Elements. 30 In this case, it is concluded that efficient ventilation in closed spaces is the key to safeguarding against the virus spread in confined areas. 31 The ventilation system can be treated as a safety critical element with the rigorous safeguarding performance standards as applied to process engineering critical equipment.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.