7.3 Population Categories.
Section 450.101 contains different safety criteria for people in different categories. One category is “public,” meaning, for a particular licensed or permitted launch or reentry, people that are not involved in supporting the launch or reentry. Public includes those people who may be located within the launch or reentry site, such as visitors, individuals providing goods or services not related to launch or reentry processing or flight, and any other operator and its personnel as defined in § 401.7. Public includes a second category “neighboring operations personnel (NOP)” meaning those members of the public located within a launch or reentry site, or an adjacent launch or reentry site, who are not associated with a specific hazardous licensed or permitted operation currently being conducted, but are required to perform safety, security, or critical tasks at the site and are notified of the operation as defined in § 401.7. People that are involved in supporting the launch or reentry do not meet the definition of “public” in § 401.7. The FAA does not have safety oversight of people that are involved in supporting the launch or reentry, and therefore does not have safety criteria that apply to these personnel. It is important to know the definitions of each of these categories so that population centers are categorized appropriately. The population model should identify the appropriate category for each population center. A region may have more than one population center if there are people in different categories at the same location. The acceptable risk to NOP in §§ 450.101(a)(1)(ii) and 450.101(b)(1)(ii) is twice that of the rest of the general public in §§ 450.101(a)(1)(i) and 450.101(b)(1)(i).
7.3.1 Personnel Involved in Supporting a Launch or Reentry.
Personnel involved in supporting a launch or reentry do not meet the definition of “public” in § 401.7. These personnel may include individuals providing goods or services related to launch or reentry processing or to flight, and those ensuring safety, such as safety inspectors or security guards who are preventing people from entering flight hazard areas. Employees of vehicle operator licensee or launch or reentry site licensee may be designated as being involved in supporting a launch or reentry if they provide goods or services related to the launch or reentry processing or flight. Likewise, employees of contractors, subcontractors, payload operators, or government personnel may be involved in supporting a launch or reentry. Visitors not involved in supporting the launch or reentry, and individuals providing goods or services not related to launch or reentry processing or flight are considered public, per the definition in § 401.7.
No one who is a guest without a necessary function in the operation, nor any visitors or individuals providing goods or services not related to launch or reentry processing or flight should be designated as operational support. An operator should ensure that the numbers used in any population model are accurate and that personnel involved in supporting a launch or reentry are not counted as public.
7.3.2 Neighboring Operations Personnel (NOP).
Per § 401.7, NOP means those members of the public located within a launch or reentry site, or an adjacent launch or reentry site, who are not associated with a specific hazardous licensed or permitted operation currently being conducted, but are required to perform safety, security, or critical tasks at the site and are notified of the operation.
While still considered members of the public, NOP are subjected to different individual and EC risk criteria. Per § 450.101(a)(1)(ii) and (b)(1)(ii), the collective exposure to risk for NOP must not exceed an expected number of 2 × 10-4 casualties, and per § 450.101(a)(2)(ii) and (b)(2)(ii) the risk to any individual categorized as NOP must not exceed a probability of casualty of 1 × 10-5 per launch. The NOP definition’s reference to safety, security, or critical tasks is intended to maintain flexibility to include various tasks as industry practices evolve over time. These tasks may include maintaining the security of a site or facility or performing critical launch processing tasks such as monitoring pressure vessels or testing safety-critical systems of a launch vehicle for an upcoming mission. These tasks may also include business operations that cannot be reasonably conducted off site, such as onsite hardware work as well as data processing that must be conducted in a secure facility.
Note: Personnel who are not required to perform safety, security, or critical tasks at the site are not considered NOP pursuant to § 401.7. Examples of such individuals may include individuals conducting normal business operations that need not be conducted in hazardous areas, individuals in training for any job, and individuals performing routine activities such as administrative, office building maintenance, human resource functions, or janitorial work.
7.3.3 Public.
Everyone who falls under the definition of “public” in § 401.7 and who does not qualify as NOP is subject to the same collective and individual risk requirements in § 450.101.
Members of the public can be anywhere, even on-facility during the launch (although the applicant may want to restrict public population movement within hazard areas; this is discussed in § 450.161, Control of Hazard Areas). In accordance with § 450.101(a)(1)(i) and 450.101(b)(1)(i), the collective exposure to risk must not exceed an expected number of 1 × 10-4 casualties and the risk to any individual must not exceed a probability of casualty of 1 × 10-6. However, only members of the public in uncontrolled areas, excluding NOP, are considered in the conditional expected casualties (CEC) calculation pursuant to § 450.101(c)(2) (see AC 450.101-1 for a discussion of which are controlled and uncontrolled areas).
8 POPULATION EXPOSURE ANALYSIS – CONSTRAINTS.
8.1 Characterize Distributions of People Geographically and Temporally.
In accordance with § 450.123(b)(1), the exposure analysis must characterize the distribution of people both geographically and temporally. The following paragraphs in this AC provide a means of compliance to characterize these distributions within the identified region(s) of interest. Resolution in space and time should be sufficient to demonstrate compliance with the safety criteria when the operation is initiated. In the context of population center fidelity, resolution in space should be defined as the geospatial area in which a population is defined, and resolution in time should be defined as the recency of the data.
8.1.1 Geographical Distributions of People.
A valid population exposure model may use different resolutions (i.e. population center sizes) for different locations. The key factor is that risk should be constant across the population center (as defined in this AC). Different resolutions of population data could occur either because the population density or the probability of casualty is sufficiently uniform across the population center. For example, the resolution of population data for an on-facility area generally should be much finer than the resolution of population data downrange, where debris impact probability distributions typically span a much larger geographical area. Figure 1 of this AC graphically illustrates the concept of matching population resolution to the size of debris impact probability dispersions (also known as footprints). A series of debris footprints within which the debris could impact are displayed along with modeled population centers of varying sizes.
8.1.1.1 The top left image in figure 1 of this AC shows population centers that are too large as compared to the debris impact probability dispersions. If the resolution of the population centers is this large relative to the debris dispersions, then it can lead to inaccurate risk results. However, if population centers can be adequately modeled with uniform distributions of buildings and population, then larger population centers may be used without loss of precision in casualty expectation calculations.
8.1.1.2 The top right image in figure 1 shows populations that are too small as compared to the debris dispersions. While this will allow for accurately calculated risk, using this resolution everywhere might not be computationally feasible.
8.1.1.3 The bottom two images in figure 1 show appropriately sized population centers.
opul ation cente rs too lar ge (unle ss p opul ation is uniforml y distr ibuted ) Popul ation cente rs comp arable to dispers ion size Popul ati on center s unne cessaril y small Popul ation centers vary in size , each w ith ro ughly uniform den sity 10/12/2022 AC 450.123-1
8.1.1.4 To demonstrate compliance with the safety criteria, an adequate risk assessment should resolve debris impact dispersions so that the impact probability contours are smooth and continuous, in accordance with § 450.119(b). Multiple overlapping fragment groups will result in smoother probability contours. An assessment of the dimensions of the probability contours can then be used to assess the degree of refinement that should be applied. In practical applications, population center sizes are usually aggregated over groups of these high-resolution cells to arrive at a compromise between the maximum population center resolution and total number of population centers that need to be processed in risk analysis tools.
Figure 1. Population Data Resolution as a Function of Dispersion3
8.1.1.5 On-Facility Populations Data Resolution.
On-facility populations usually experience the highest risk during a mission. In the launch area, debris footprints typically span several hundred to more than ten thousand feet. Due to the small impact probability dispersion size, the resolution of the population data at the launch area should be by individual building.
3 Figure 1 of this AC is derived from Figure 4.1.37 in the following reference: Allahdadi, Firooz A., Isabelle Rongier, Tommaso Sgobba, and Paul D. Wilde, Safety Design for Space Operations, Sponsored by The International Association for the Advancement of Space Safety, (Watham, MA: Elsevier, 2013).
8.1.1.6 Spectators.
Some missions may have spectator areas. Due to collecting many people in one location near the mission, there is the potential for a single accident causing many casualties. This group can be the largest contribution to risk, and particularly important for the FAA’s calculation of Maximum Probable Loss (MPL) determinations for financial responsibility purposes.
It may be necessary to manage location and access to spectator areas to meet the safety criteria and remain consistent with the MPL determination.
8.1.1.7 Near-Facility Populations Data Resolution.
Near-facility is not precisely defined. The region where high resolution should be used depends on the size of the dispersions and the potential consequences of an accident. Generally, this region extends from the launch or reentry site boundary up to about 50 km from the flight path, where the vehicle IIP is moving slower than several km per second, but it may be smaller for smaller vehicles. As noted above, smaller dispersions could necessitate higher resolution, but the accuracy that is needed for a vehicle with small consequence can be less than for a large vehicle. The typical resolution for near-facility population sheltering data should be by small groups of buildings. This could be a city block, a housing tract, or a few square km for rural regions. Detailed population models typically include “point receptors” to account for high density population areas or other areas of special concern such as schools, hospitals, spectator areas, stadiums, cruise terminals, recreational areas (such as crowded beaches), chemical processing plants or nuclear plants, etc.
8.1.1.7.1 Grouping by Structure Class.
Although it is suggested that buildings are grouped when accounting for population, it is important to keep in mind that groups of buildings can have multiple construction types. This generalization issue can be solved by allocating people by approximate percentage to different building classes. More information about building classes can be found in paragraph 8.2.2 of this AC.
8.1.1.7.2 Accuracy.
While these population centers are larger than the on-facility centers, the accuracy of the near-facility population still should be good, as the risk to near-population centers has a significant effect on public safety. When considering near-facility populations, the applicant should ensure both “permanent” (people who live in this area and are there year-round) and “temporary” (people who are only visiting the area for vacation, business, etc.) populations are accounted for. The applicant should also consider significant areas where populations might be. This can include launch viewing areas, vacation locations (e.g. parks, beaches, and hotels), seasonal events (e.g. harvest, fishing), schools, stadiums, and airports.
Applicants should also address isolated or transient structures such as oil 13
a 1kmCellP opula tion Centers : ....... G3 • High Resolution Po . pulanon C.enters platforms, boats, and significant roads. Time variations also should be accounted for. Daytime, nighttime, weekdays, and weekends can determine if people are at work or home, outdoors or inside. Seasons or events can determine whether to expect more or less people in the area.
Given a known launch is occurring, an operator should account for people congregating outside (homes and office buildings), observing from rooftops, and pulling to the side of roads. More information about temporal variation can be found in paragraph 8.1.2 of this AC.
8.1.1.7.3 Example of Population Center near a Launch Site.
Figure 2 shows an example of the resolution of regional population centers near a launch site. The purple circle shows where the launch pads are.
Within about 5 km of the launch site, individual buildings are specified.
A higher resolution is also used in an area north of the launch pads because part of the base exists in that area. Out to about a 15 km radius from the launch site, 1 km cells are used.
Figure 2. Near Facility Population Data Example 14
8.1.1.8 Downrange Populations Data Resolution.
Population centers that are far downrange from the launch site are typically defined in terms of population density and structure types within latitude-longitude grid cells. Densely populated urban centers located far downrange should be treated as separate population centers and their population count deducted from the latitude – longitude cell(s) that contain the urban center. In a lower fidelity population exposure model, counties, states, or even countries with average population densities and demographically linked structure types might be used. However, when applying lower fidelity modeling assumptions, the effect of uncertainty in the population density and structure definitions on risk and hazard area predictions should be evaluated in accordance with § 450.115(b).
8.1.1.8.1 Extent and Resolution.
The population model should encompass the area that includes all foreseeable impact locations resulting from all credible failure modes. In situations where the land area subject to potential debris impacts is large and the population at risk generally dispersed, such that population distribution is uniform, a simple population model may be based on a database of cells defined by a grid covering the hazarded land area.
A latitude-longitude grid system is commonly used where each grid cell covers an area defined by ranges of latitude and longitude. The location of a cell should be defined in terms of the coordinates (usually latitude and longitude) of the centroid of the populated area. The area of the population center is the land area of the cell and this value is used in the probability of impact calculations for debris dispersions that fall on the cell. For each cell, the number of people and sheltering distribution should be specified over the time period subject to debris impacts. Population is then assumed to be uniformly distributed within each cell, and cell sizes should be selected to provide adequate resolution as discussed below. The area under the flight path should be higher resolution than areas further away from the flight path.
8.1.1.8.2 Example of Downrange Population Center Area.
A variable resolution should be used for downrange populations, as demonstrated in figure 3 of this AC. The red dots represent the nominal trajectory. The green and red squares underneath represent the population density. The size of the square represents the population area. The brighter the green, the smaller the population density is. Darker green and red squares indicate high population densities, as described in population density key in figure 3. The squares underneath the trajectory are smaller due to a higher resolution. Gaps can be seen between population areas.
These are areas where either no people or an extremely limited number of people are located.
.000000
100.000000 1000. 000000 10000. 000000 100000 . 000000 1000000 . 000000
•••
• • · ■ 10/12/2022 AC 450.123-1 Figure 3. Example of Downrange Population Center Areas
8.1.2 Temporal Distributions of People.
Section 450.123(b)(1) requires that the analysis must characterize the distribution of people temporally. People move between locations and structure classes for many different reasons, on timescales from daily (e.g. work vs. home) to annual (e.g.
vacationers). In addition, specific events, such as cruise ship sailings, can have a significant effect, and weather also affects the choice of activity. Sheltering and geographic location should be accounted for via multiple population exposure data sets as appropriate to support launch and reentry missions.
8.1.2.1 Movement Based on Time of Day, Day of Week, and Seasons.
The day vs. night movement is typically the most significant effect, as this changes both location and sheltering of people. People often move around significantly throughout the day due to events. For on-facility populations, the activities happening during time of launch will determine where people may be located. Off-base, people may be commuting based on time of day. The location to which they are commuting can also depend on time of day, or day of week. Customers may be eating indoors or outdoors at restaurants depending on time of day, day of week, and weather. Hotels and resorts can have different populations based on time of day and year.
Beaches and campgrounds can have different populations based on time of 16
day, day of week, and weather. At ports, ships can be loading and unloading, which is another influx of people. Likewise, airports are also generating a large rush of people coming and going. Mission times in locations around launch areas can also affect population movement as usually large open areas with high visibility can attract spectators.
High-interest launches can generate an extremely large spectator population, as opposed to an “ordinary” launch which may only attract a few hundred spectators. Seasonal factors can affect the number of students at schools, the number of agricultural workers in fields or gatherings at tourist sites.
8.1.2.2 Near-Facility Populations.
For near-facility populations where the resolution is relatively high, applicants should account for areas where people may be gathered, such as beaches, campsites, and parks. The populations in these areas will fluctuate with time, so applicants should obtain data on the number of people in these areas per month. Table 1 of this AC shows an example of data collected for the number of people that are typically at beaches near a launch site. The data collected considers seasons, day of week, time of day, and weather.
Table 1. Data Collected for Beaches near Launch Site Location Description Maximum Weekend In- Season Weekend Off- Season Weekday In- Season Weekday Off- Season Night / Rain Beach 1 Outside 2500 1400 600 400 200 10 Concessions Outside 40 20 20 10 5 2 Concessions Inside 40 40 40 20 10 0 Beach 2 Outside 400 280 120 80 40 2 Beach 3 Outside 600 420 180 120 60 4 Note: The units in the above table refers to the total number of people at these locations at these times.
8.1.2.3 Foreign Countries.
For missions that affect foreign countries, applicants should consider how temporal distributions can differ from the United States. Different countries can have different work hours and work weeks. For example, some countries in the Middle East have work weeks that last from Saturday to Thursday, or Sunday to Thursday. In the case of Saturday to Thursday, there is only a one-day weekend, which is Friday. In some countries, like India or Japan, there may be many companies where employees have workdays that last more than eight hours.
8.2 Sheltering.
According to § 450.123(b)(2), the exposure analysis must account for the distribution of people among structures and vehicle types. This is commonly referred to as “sheltering.” The key purpose of sheltering modeling is assigning each population center parameters that quantitatively link to the vulnerability modeling in the debris, far-field blast overpressure (FFBO), and/or toxic risk analyses. These parameters should characterize the structure(s) such that the accuracy of the risk metrics is sufficient to demonstrate compliance with the safety criteria when the operation is initiated.
8.2.1 Important Factors.
For most debris, a structure provides protection to people inside. However, for some debris, people in structures are more vulnerable than people outdoors.
8.2.1.1 Buildings.
The roof of a building provides protection from the most common hazard:
small inert debris. However, people in a weak building or a building with windows are typically more vulnerable to serious injury from a nearby explosion or FFBO event than those outdoors. Also, inert impacts of larger fragments can cause partial or full roof collapse, or other damage to a building, resulting in a larger casualty area than if the occupants were not sheltered. The ventilation system can also affect the likelihood of casualty due to toxics.
8.2.1.2 Vehicles.
Vehicles are of particular interest for two reasons. First, the motion can add to the velocity at impact, leading to a higher kinetic energy than just from the debris. Second, a driver could be affected by the hazard, and the driver’s response could lead to a casualty.
8.2.1.3 Waterborne Vessels.
Waterborne vessels pose a special challenge as there are many scenarios where everyone on board is affected. This is especially important for vessels out of reach of emergency response or where the survival time of people in the water is short (due to temperature or sea state). Debris can also puncture the deck and either ignite fuel (especially for tankers) or penetrate the hull. Large inert debris might miss the vessel but cause a 18
surface wave due to displacement, which can cause injuries. Likewise, nearby explosions can cause effects due to the shockwave travelling underwater, which transmits energy far more efficiently to the vessel than does the air shock. The properties of the vessel are very important to these vulnerabilities, especially as vessels range from very small (and vulnerable) private boats to large cruise ships (very strong, but extremely high density of people) to super-tankers (which can be transporting flammable material).
8.2.2 Characterizing Buildings.
Depending on the type of hazard, sheltering categories may be different. For explosive debris, the wall type and window parameters should be characterized. For inert debris, the roof type and number of floors should be characterized. Typically, inert debris will not penetrate more than two floors beneath the roof. For toxic hazards, the potential for ingress of toxic materials is dependent on the air exchange rate provided by construction type, e.g., whether a building is leaky or tight. All building types may be conservatively assumed to be leaky. If the risk thresholds cannot be met with that conservative assumption, then the permeability of the buildings should be accounted for. For overpressure blast effects, window type, size, and amount are important.
Therefore, a comprehensive sheltering model should characterize a building type by roofing material (e.g., wood/tile, steel), overall construction type (e.g., reinforced concrete, wood frame), and window types (e.g., annealed, dual pane, tempered) and sizes.
8.2.2.1 Example categories.
The number of sheltering categories depends on the level of fidelity of consequence modeling employed. A simple population model may allocate people to a few sheltering categories, which may be acceptable if the casualty models for people in structures lead to conservative risks results. Table 2 of this AC gives examples of different building classes when considering inert debris. In this case, roof material is important.
Table 3 gives examples of different building classes when considering explosive debris. For explosive debris, window size, frame type, and wall type are important factors to consider. Paragraph 8.2.6 gives guidance on how to assign these structure types to population data.
Table 2. Building Classes for Inert Debris Name Building Description Open Exposed people without benefit of an overhead roof Steel Steel frame construction with metal roof panels Concrete Reinforced concrete frame with concrete on steel roof and floor Wood Wood frame with wood roof and floors Light-Metal Roof of pre-engineered metal building (PEMB) or vehicle Composite Layered roof made up of lightweight, non-metallic materials Weak Concrete Poor quality concrete frame, roof, and floor Usually found in less-developed areas of the world Tile-Roof Wood frame, tile roof, wood floors Weak Frame Thin/weak wood frame, very thin roof surface (corrugated metal, thin thatch). Found in poorly-developed areas of the world Table 3. Building Classes for Explosive Debris Building Name Wall Type Frame Type Window Size Open None None None Small Metal Pre-Engineered Metal Light Steel Columns & Beams Small Small Wood Small Wood Wood Stud Small Med Wood Med Wood Wood Stud Medium Small Unreinforced Block Small Unreinforced Block Brick Shear wall Medium Medium/Large Unreinforced Block Unreinforced Block Shear Wall or Reinforced Concrete Frame Medium Reinforced Block Reinforced Block Shear Wall or Reinforced Concrete Frame Medium Small Reinforced Concrete Reinforced Concrete Reinforced Concrete Frame Medium Vehicle Pre-Engineered Metal None Small Trailers/Recreation Vehicles Pre-Engineered Metal Wood Stud Small 20