Winter Design Storm Factor Determination for Airports (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 This section describes a decision support tool in the form of a winter design storm identification process, shown in Fig- ure 3-1. On the left side of the figure are data reflecting design storm factors discussed in Section 2. The balance of the fig- ure illustrates the analyses and decisions required to identify an appropriate winter design event or multiday storm in the context of a specific project and its circumstances. 3.1 Nomenclature Certain terminology in this guidance should be clearly understood before the reader applies the process shown in Figure 3-1. 3.1.1 Design Events and Storms An important distinction is made between design condi- tions that are referenced to a single, isolated, peak event and those that reflect a multiday progression of weather and deic- ing activities. The terminology is as follows: • Winter event. An isolated period of significant wet and/ or freezing precipitation that requires applications of air- craft and pavement deicers/anti-icers. The duration of an event in this context is generally in the range of hours (for example, 6 hours, 24 hours). Isolated design events will be used most frequently where there is no significant storage requirement involved and the primary temporal criterion is time of concentration, operating day, or similar. • Multiday winter storm. A prolonged (multiday) period of continuous or intermittent significant wet and/or freezing precipitation that requires applications of aircraft and pave- ment deicers/anti-icers. The multiday winter storm is rel- evant where the design of a project, such as a storage facility, must take into consideration the cumulative effects of the precipitation and runoff occurring during a prolonged peri- od, often involving multiple back-to-back events. Multi day winter storms may encompass entire deicing seasons where, for example, processing of collected deicing runoff contin- ues through, and even beyond, the deicing season. 3.1.2 Project The project is the component, facility, or system that is driving the need to define a winter design condition. Exam- ples of relevant projects are a deicing pad, an apron diversion and collection system with storage and treatment compo- nents, and a pond that stores dilute airfield runoff for con- trolled discharge via a NPDES-permitted outfall. 3.1.3 Factors at Risk This term refers to what is at risk if facilities designed for the selected winter design storm fail. Examples are violation of NPDES permit requirements, flooding of areas of the air- field, or damage to airport infrastructure. Additional exam- ples are given in the case studies in Section 4. • Regulatory factors. Regulatory risk factors involve non- compliance with FAA and environmental (that is, EPA and/or state-level) regulatory requirements. For example, FAA Federal Aviation Regulation (FAR) Part 77 airfield contours might be at risk if deicer-affected snow piles were to be placed in certain locations or allowed to grow beyond a certain elevation. The most common environmental reg- ulatory risk factor is noncompliance with discharge limits in an airport’s NPDES permit. Other examples are viola- tion of an airport’s sanitary sewer discharge permit and failure to maintain downstream water quality standards as a result of airport deicing discharges. • Physical factors. Other potential factors at risk are infra- structure features and safety features. Examples of the for- mer would be risking the integrity of storm water pond containment structures if pond volumes reach overflow S e c t i o n 3 Decision Support Tool

10 levels, or the risk of flooding airfield pavements or air- field equipment infrastructure such as electrical vaults. An example of the latter would be increased wildlife hazard risk with flooded areas with standing water. • Operational factors. Deicing infrastructure and operations must accommodate peak-hour departures under weather conditions that pose the greatest demands on effectively deicing and anti-icing the aircraft fleet mix at the airport. The risk factors associated with maintaining these opera- tional levels include inadequate throughput capacity of the aircraft deicing pads and teams, inability to maintain snow- and ice-free airfield pavement, inadequate supplies of deicing products, and excessive amounts of ponded water in terminal apron areas. Undersizing deicing system components will result in increased risk of not maintaining these regulatory, infrastruc- ture, and/or operational factors. 3.1.4 Target Level of Service Once the factors at risk have been identified, the critical question then becomes, “How much risk are we willing to accept?” This question is addressed through defining the Figure 3-1. Winter design storm identification process.

11 target level of service or the level of risk the airport is willing to assume in its deicing facility design approach. There are two distinct elements in defining the risk associated with each risk factor: severity and likelihood. If the risk factor is operational—for example, the risk that deicing facilities may not be adequate to serve all of the flights—then the severity of the risk might be associated with the number of flights delayed, and the likelihood might be associated with the probability of critical storm factors exceeding the selected design storm. The actual risk is a combination of the severity (one or 100 flights affected) and the likelihood (10 times a year or once in 10 years). Often, regulatory agencies initially assume a very low risk tol- erance. Commonly, this tolerance level is later balanced against the costs to achieve that initial low level of risk. For example, the environmental permitting agency may expect NPDES effluent limits to be met 100% of the time under the maximum his- torical deicing storm conditions. After estimating the cost and practicality of achieving that high standard, the airport may propose a compromise level of service that meets regulatory agency requirements at a marginally higher level of risk. Thus, the process of defining level of service often becomes an itera- tive one where the cost of achieving lower risk levels is balanced against the value of incremental reductions in risk. 3.1.5 Existing System Components This term refers to the relevant components that make up the current system within which the project will be installed. Some examples of relevant system components are drainage basins, storm drainage networks, deicing runoff collection elements, storage facilities, permitted storm water outfalls, permitted sanitary sewer discharge points, and on-site treat- ment facilities. Definition of some or all of these components is needed to support modeling conducted as part of the fre- quency analysis, as discussed in Section 3.4. 3.2 Decision Points Four decisions, represented by the diamonds in Figure 3-1, determine which of three possible approaches is used to identify the winter design event or multiday storm. The discussions that follow describe the nature of each decision point and the thinking that goes into each. It must be rec- ognized that each airport and deicing project represents a unique set of circumstances, and it is not feasible to antici- pate the details of every possible situation. As a result, the users of this guidance should be prepared to interpret the principles represented by each of these decision points and apply them to their specific project. • What is the duration of the design condition? The nature of the project and the design criteria will determine if the design condition is a single, isolated, peak event or a mul- tiday progression of weather and deicing activities. Typi- cally, if conveyance capacity is the key design issue, then intense, short-duration events will define the critical con- dition because they represent the greatest challenge to run- off conveyance. An example is the inlets and pipes at a cen- tral deicing facility that must be sized to move runoff away from the facility quickly enough to avoid flooding under peak runoff conditions. On the other hand, if the project involves designing a storage tank for a deicing apron or central deicing facility, then typically it will be necessary to consider the dynamics of runoff flowing into and out of the tank from a series of events between which the tank cannot be completely emptied. • Is the design issue volume or load? This question applies only to projects where the critical design condition is an isolated design storm. If the risk is associated with exceed- ing a parameter involving volumes of runoff resulting from wet precipitation, such as flooding an apron area, then short-term precipitation and runoff phenomena are the drivers, and conventional storm water design event definitions are applicable. On the other hand, if the risk is associated with exceed- ing a deicer load or concentration design criterion, such as a numerical effluent limitation in storm water discharges under a NPDES permit, then the sources and magnitudes of deicing loads must be incorporated into defining the design event or multiday storm. • Are long-term data records available? At this junction in the process, the need for data to support modeling anal- yses must be addressed. Long-term historical records of weather parameters that affect deicing operations are gen- erally available through the National Weather Service. The resolution of these data may be hourly or shorter periods, or daily totals and averages, depending on the reporting station and length of record. Where weather records are available for several decades or more, the representative- ness of older data should be critically evaluated in light of local long-term trends in climate change (see Section 3.5). In addition, non-weather data—such as aircraft opera- tions and fleet mix, deicer types and usage (see Box 3-1), deicing runoff collection performance, treatment and recy- cling volumes and concentrations, and outfall discharge records—are often needed to support the analyses. The exact data requirements will vary with the type of project, factors at risk, and target level of service. Typically, recent data on these aspects of airport operations are going to be the most representative of conditions of interest to the design objectives. Table 3-1 provides general guidance regarding the data that may typically be needed for different types of projects and the relative importance of the data in supporting analy- ses. It is rare for all of the data requirements to be met with available records, which is why skilled analysts and mod- elers with experience in conducting analyses with limited

12 data records are normally required. If it is determined that the data needed to support a robust analysis do not exist, then a simplified default approach is recommended, as dis- cussed in Section 3.3.1. • Is a robust risk analysis needed? Developing a statistical frequency analysis of long-term weather and deicer usage can be a significant undertaking in terms of time and resources. As such, it is important to consider whether the investment is appropriate to the value of the information that the analysis will provide, even if the data to support an analysis are available. On projects where the factors at risk have a high value, and the consequences of over- or under-designing the system are significant and potentially costly, a robust frequency analysis will typically be valu- able in supporting confident decision making and justify- ing costs. An example is the sizing of an on-site treatment facility where under-design risks noncompliance with regulatory permits and overdesign is unnecessarily costly. On the other hand, where the risk associated with under- designing a project, or the incremental cost of overdesign- ing it, is small, a sophisticated frequency analy sis may not be needed to adequately define the winter design event or multiday storm. The Baltimore/Washington Thurgood Marshall International Airport (BWI/Marshall) case study (see Section 4.1) is an example of such a situation, where new storage volume was estimated on the basis of expe- rience with system performance to estimate additional storage requirements. 3.3 Winter Design Event/Storm Outcomes Applying the decision process outlined in Figure 3-1 will lead to three possible characterizations of the winter design event or storm. These outcomes are described in the following subsections, presented in order of increasing complexity. It should be emphasized here that in all cases the characterization of the winter design event or storm should be in the context of the deicing season months and conditions. 3.3.1 Design Storm or Multiday Event of Memory In the absence of long-term, facility-specific records that support a frequency analysis, the default is to examine weather records as well as any supporting information (newspaper accounts, recollections of long-term airport staff, etc.) to identify an extreme winter season storm or multiday event in the past (i.e., a storm “of memory”) that arguably represents appropriate design conditions. Commonly, the identification of such an event or storm is done as part of negotiations with a regulatory agency. 3.3.2 Design Event Based on AC 150/5320-5C If it is determined that an isolated winter precipitation event is appropriate as the design event, then the design stan- dards for airfield drainage systems in FAA’s AC 150/5320-5C, “Surface Drainage Design,” (2006) are used to identify the appropriate design event. Published precipitation records can be used to identify an event with the target frequency of occurrence (for example, 5-year/24-hour). The National Oceanic and Atmospheric Administration (NOAA) publishes Box 3-1. Obtaining deicer usage data. Obtaining accurate data on deicer usage is often challenging but is essential where the project design requires consideration of deicer loads or concentrations. The data will typically be obtained from the entities that apply the deicers. It is advisable to obtain Material Safety Data Sheets or similar product specification documents for all deicers used. Aircraft deicer usage. Aircraft operators and deicing service providers are the best sources for data on aircraft deicer (Type I ADF) and anti-icer (Type II/IV AAF) usage. Critical usage information includes whether the products are propylene- or ethylene glycol-based, the mixture strength or dilution (for Type I ADF), and the volumes applied at mixture strength. Mixture can be reported in different ways, so providing the reporting organizations with specific guid- ance (e.g., always report mixture strength as the ratio of neat ADF concentrate to water) will promote data consistency. Other useful informa- tion is time and location of application, and type of aircraft deiced. Daily or finer timescale usage data are typically best for modeling analyses. Pavement deicer usage. Airfield pavement deicer usage is obtained from the airport department responsible for airfield pavement deicing. Critical data include the brand names of the products, the volumes or weights of each product applied, and the location of those applications. Normally, daily totals are adequate for pavement deicer usage data.

13 precipitation frequency analyses that provide the frequency of precipitation depths associated with different durations (depth-duration-frequency analyses). The latest versions of NOAA’s precipitation frequency analyses are available at http://hdsc.nws.noaa.gov/hdsc/pfds/index.html (accessed November 4, 2011). The design event can be identified from the NOAA data with a definition of the critical duration and the target fre- quency of occurrence. For example, if a deicing apron is being designed for a 5-year/24-hour storm for an airport near Louisville, KY, the NOAA reference would indicate that the design event would have a rainfall depth of 3.9 in. The frequency data published by NOAA seldom dis- tinguish between seasons. Some site-specific studies have identified distinct differences between precipitation depth frequency statistics calculated for winter months and the annual statistics readily available from NOAA. This may be important where separate deicing runoff conveyance is Type of Project Weather Data Facility and Opera�ons Data Deicing Prac�ces Deicing Controls no it ati pi ce rP re ta W( )t ne la vi uq E no it ati pi ce rP ep yT ri A er ut ar ep me T )e ga re vA /x a M/ ni M( re ht O tf ar cri A sn oit ar ep O dn a te el F xi M tf ar cri A gn ic ie D dn a it nA - gn ic i di ul F eg as U gn id ul cn I( )s no it ac oL dl eif ri A tn e me va P re ci eD eg as U gn id ul cn I( )s no it ac oL gn ic ie D ff on uR no it ce ll oC se mu lo V dn a sn oit ar tn ec no C mr ot S re ta W ff on uR eg ra hc si D se mu lo V dn a re ta W yti la u Q tn e mt ae rT ro /d na gn il cy ce R se mu lo V dn a re ci eD sn oit ar tn e c no C Apron collec�on system Glycol collec�on vehicles — — Block-and-pump system — Airfield drainage planning/design/retrofit — — Centralized deicing facility — Deicer-laden snow management — — — — Storage for deicing runoff — — — Manual and automated diversion valves — — — — — — Real-�me monitoring technology — — — — — — Catch basin inserts/valves — — — — — — POTW discharge — — — — — — On-site treatment facility — — — Recycling program — — — — — Legend: High priority Poten�ally useful but not essen�al — Not typically needed Table 3-1. Types of data records commonly needed for different types of projects.

14 being designed. In such cases, using the annual statistics may result in oversizing the project components. However, the procedures documented in NOAA’s Atlas 14 (2004– 2011) can be applied to derive deicing season-specific depth frequency statistics from available long-term precipitation data. The deicing season would be defined by reviewing deicing records and identifying those months during which significant deicing is typically conducted. The procedure can be simplified by using only the local airport data and using a site-specific analysis not adjusted for regional pat- terns. The simplification might also use a less rigorous testing of the most appropriate extreme value frequency distribution, defaulting to the lognormal, log Pierson, or Gumble distributions. One final note on this approach concerns the implications of snow removal. If the operation of the planned facility includes removing snow from within the project boundar- ies before it can melt, then some adjustment to the design event precipitation depth may be appropriate to account for removing the water equivalent of that snow. 3.3.3 Design Storm or Multiday Event Based on Frequency Analysis The frequency analysis approach identifies a winter design storm or multiday event specific to the airport and project. This is accomplished by conducting a continuous simulation modeling of runoff and deicing operations and of deicer usage and loading over a relatively long period of historical winter weather conditions. The output of the simulation is in parameters that are relevant to the factor(s) at risk and that represent the performance of a current or planned deicing project or system under the full range of deicing season weather conditions in the historical record. Statistical determination of the frequency of conditions that correspond to failure of the factor(s) at risk is used to identify the characteristics of the winter design storm or multiday event. This approach involves relatively sophisticated modeling and statistical techniques applied to the circumstances spe- cific to the airport and project being evaluated. A significant amount of professional skill and judgment is involved in con- structing the analysis to adapt to these factors, and the reality of available data and resources is also a concern. As a result, the guidance presented here cannot provide a cookbook approach for conducting these analyses. Rather, the guidance is intended to equip readers with a working familiarity with the concept and steps in the process, to serve as a basis for making informed decisions about its potential applicability to their situation, and to provide a foundation for under- standing the results of its application. 3.4 Frequency Analysis The steps in applying the frequency analysis approach are described in the following subsections and correspond to the process steps illustrated in Figure 3-2. It should be noted that the example tables and graphics that are shown in Figure 3-2 and accompany the following discussions are for illustrative purposes only and do not reflect actual data or analysis. As noted in Section 3.2, the need to conduct a frequency analysis at the level of detail described will depend on the specific design issues and needs. If either the risk associated with under-designing the system or the incremental cost of overdesign is small, then the costs of conducting an analysis to precisely define the winter design event or storm may not be justified. Instead, a less rigorous analysis combined with professional judgment may be adequate in defining appro- priate design conditions. 3.4.1 Develop Time Series Water Budget from Historical Weather Records The first step is to construct a time series runoff model of the system that describes the relevant components of the water budget (see example water budget in Box 3-2) within the boundaries of the project. These may include the following: • Precipitation inputs in the form of rain, freezing rain, and snow; • Storage in the form of accumulated snow and ice pack, depression storage, storage of collected deicing runoff, and storm water pond, basin, or tank storage; and • Outputs in the form of conveyance from a project area to the airport’s storm water system, plowing of accumulated snow outside of a project area, permitted surface water dis- charges, discharges to sanitary sewers, and volumes of col- lected deicing runoff transported off site for treatment or recycling. The complexity of the runoff model used for developing the water budget can range from a relatively simple spread- sheet model to one of the more sophisticated continuous hydrology and hydraulics models, such as the EPA’s Storm Water Management Model (SWMM). Often, an existing storm water runoff model developed for general drainage at the airport may be adapted to this purpose. ACRP Report 14 (CH2M HILL et al., 2009) provides an overview of various approaches that can be applied to modeling airport runoff quantity and quality.

15 Figure 3-2. Winter design storm frequency analysis process.

16 situation is sizing a facility to store deicing runoff prior to its discharge to a wastewater treatment plant where the dis- charge rate to the plant is limited by concentration in the runoff. 3.4.3 Develop Time Series Deicer Usage for Historical Weather Conditions The next step is to develop a time series of deicer usage associated with the water budget time series over the simula- tion period (see example in Box 3-3). Deicer usage at each time step in the simulation period is estimated using defined relationships between weather parameters, airfield opera- tions, and deicer usage. Typically, these relationships are developed empirically using historical weather, operations, and deicer usage records from the airport. The expression of deicer usage relationships can take vari- ous forms, typically depending on the nature of available data. For example, if daily airport-wide deicer usage records are available, aggregate usage may be estimated as a func- tion of weather. Figure 3-3 presents an example of a predic- tive relationship based on regression analysis of daily aircraft deicing fluid (ADF) usage and snowfall. The decrease in usage predicted after a critical snowfall amount reflects the point at which aircraft departures begin to decline with the increasing severity of the snow event. Greater resolution and detail in estimated deicer usage may be possible where historical data include deicer appli- cation records by individual aircraft. Where adequate data are available, analysis of hourly weather data, aircraft type, and deicer usage can yield average application rates for dif- ferent group size aircraft under different defined ranges of weather conditions. Table 3-2 illustrates how rates of Type I ADF usage might be expressed on the basis of these Two important considerations at this point in the process are selecting the timeframe and selecting the time step of the model simulation. The implications of climate change should be considered in selecting a simulation period that reflects current and anticipated future weather conditions. This topic is discussed further in Section 3.5. With respect to the time step for the analysis, although weather data may be available for intervals as short as 5 min., data on aircraft and deicing operations are typically available only for longer intervals; deicer usage is rarely reported per aircraft and is more commonly reported on a daily or less frequent basis. Also, short time steps may not be necessary to adequately represent the mechanisms and phenomena affecting the factor(s) at risk. The product of this step is a description of the water budget at each time step over a simulation period that cap- tures the full range of weather conditions experienced at the facility. 3.4.2 Determine if Deicer Load or Concentration Is a Factor The complexity of the frequency analysis depends on whether deicer concentrations or loads in runoff affect the sizing of the project. If sizing is purely a function of runoff flows and volumes, then frequency analysis is conducted on the time series water budget record. An example is designing storage for high-strength runoff from a centralized deicing facility prior to transport off site for recycling. In this case, the constraining factor on emptying the storage facility is the rate at which stored runoff can be transported off site, regardless of the deicer concentration. Conversely, where deicer concentration or load is a deter- minant in sizing the project, then the time series water bud- get must be expanded to include deicer concentration, as described in the subsections that follow. An example of this Box 3-2. Example of a time series water budget. Da te Pr ec ip (i nc he s) Ru no ff (in ch es ) Sto ra ge (1,000 cf) Di sc ha rg e (1,000 cf) 1/ 2/ 195 1 0 0 0 0 1/ 3/ 195 1 0 .4 4 0 .4 0. 1 0 .0 4 1/ 4/ 195 1 6 .6 5. 6 1 .3 1 0 .7 5 1/ 5/ 195 1 3 .9 3. 4 1 .7 5 0 .7 5 1/ 6/ 195 1 1 .3 1. 1 1 . 385 0. 75 Box 3-3. Example of time series estimated deicer usage. Dat e Pr ec ip (in ch es ) App lied Gl ycol (g al s) 1/ 2/ 1 951 0 0 1/ 3/ 1 951 0. 44 1, 000 1/ 4/ 1 951 6. 6 1 5, 000 1/ 5/ 1 951 3. 9 1 2, 000 1/ 6/ 1 951 1. 3 5 , 600

17 types of records. Similar relationships can be developed to describe Type II/IV deicer usage and airfield pavement deicer usage. The relationships describing deicer usage as a function of weather and operations are applied to each time step in the simulation period to generate a time series record of esti- mated deicer usage, or load generation. 3.4.4 Distribute Time Series Deicer Loads among Fate Compartments The next step in the process is to develop a quantitative characterization of the distributions of applied deicers among the compartments shown in the material balance of deicers applied to aircraft and/or airfield pavement in Figure 3-4. (An example of the possible distribution of applied deicers is contained in Box 3-4.) Each compartment on the right side of the figure represents the fate of some portion of the applied deicers. It is difficult to generalize the quantitative distribu- tions or extrapolate them from other airports because of the influence of facility-specific factors. The best approach is to conduct mass balance evaluations of monitoring data to esti- mate facility-specific distributions for aircraft and pavement deicers. Guidance on approaches for these evaluations is pro- vided in ACRP Report 14 (CH2M HILL et al., 2009). Not all of the compartments shown in Figure 3-4 will apply under all weather conditions, and the description of distribu- tions must reflect that. For example, with freezing rain, there will be no “pink snow” compartment unless there is snow- pack remaining from previous precipitation. It is important to note that there should be a direct corre- spondence between compartments in the water budget and in the distribution of deicers, with the possible exception of fugitive losses. Da ily A D F U sa ge (g al s) Daily Snowfall (inches) Most Least Most Figure 3-3. Example of how the relationship between daily snowfall and Type I ADF usage can be expressed. Weather Condi�on Group I Group II Group III Group IV Group V Frost [Gallons of Type I ADF concentrate for each combination of weather condition and aircraft group size] Freezing rain Light snow Medium snow Medium-heavy snow Heavy snow Table 3-2. Example of how the relationship between weather, aircraft size, and Type I ADF usage can be expressed.

18 are identified by statistically evaluating the time series data set. Assuming for simplicity’s sake that the data set is based on a 1-day time step, the analysis can be described in the follow- ing steps: 1. Rank all days in the data set according to the parameter that reflects protection of the factor(s) at risk. An example is ranking the data set by daily outfall discharge concen- tration from lowest to highest because high concentra- tions represent risk to NPDES permit compliance. 2. Calculate the frequency of exceedance of each value in the data set (see Box 3-5). 3. Identify the day in the data set where the exceedance frequency corresponds to the target level of service. The conditions on this day reflect the threshold of the design condition, or the design event. 4. Evaluate the days immediately above and below the threshold to assess the variability in weather factors associated with the design condition. This is important because similar volumes of deicers may be used, or similar outfall concentrations may be observed under significant- ly different weather conditions. Insights gained through understanding the range of weather conditions that are associated with the target level of service will typically be valuable in the design process. The same basic steps apply to identifying the multiday winter design storm, but the analysis becomes more compli- cated. The key complexity is that the duration of the storm must be defined. Storm duration will be specific to the airport and context of the project. Determining an appropriate dura- tion involves examining the time series data set and possibly testing the implications of using different storm durations to identify the duration that best suits the needs of the design. Applying statistical techniques for time series analyses may be required, the details of which are beyond the scope of this document. Deicer Usage Aircra� deicers and an�-icing fluids + Pavement deicing materials = Collected and sent to treatment/ recycling + Entrainmentin pink snow + Storm waterdischarges + Fugi�ve Losses On aircra� drippage tracking wind dri� degrada�on other Figure 3-4. Material balance. Adapted from ACRP Report 14 (CH2M HILL et al., 2009). Box 3-4. Example of distribution of applied deicer. 3.4.5 Develop Time Series Water and Deicer Load Budget for Historical Weather Records Once the distributions of applied deicers are defined, they are applied to the time series of deicer usage to describe deicer loads in each compartment at each time step. Where an exist- ing storm water model has already been developed for an air- port, the deicing functions can often be added as pollutant load sources. The product of this step is a time series data set of weather parameters, distributed water volumes and deicer loads, and any other calculated parameters that directly relate to the factor(s) at risk. Some examples of the latter are deicer concentrations in collected runoff or discharges and storage facility volumes or surface elevations. Table 3-3 presents a conceptual example of what this data set might look like. 3.4.6 Analyze Exceedance Frequency The event or multiday storm conditions associated with the target level of service relative to the factor(s) at risk

19 3.4.7 Identify Design Event or Multiday Storm Based on Frequency of Delivery of Target Level of Service The process described in the preceding subsections will result in the identification of a subset of individual events or multiday storms that reflect conditions that recur at the defined target level of service. These events or storms may be similar or dissimilar in terms of meteorological, operational, and environmental factors, depending on a myriad of site- specific characteristics and mechanisms that drive runoff vol- umes and rates, deicer loads and concentrations, discharge rates, and so forth. As a result, it is difficult to generalize the steps in precisely characterizing design conditions based on the events or storms identified through the frequency analy- sis. It is at this point in the process where the collaboration of designers, planners, compliance specialists, airport manage- ment, and other airport stakeholders will result in a suitable definition of the design event. 3.5 Consideration of Climate Change A detailed discussion of climate change and its implica- tions to the design of storm water infrastructure is beyond the scope of this investigation. Nonetheless, a general over- view of the issue is provided here. There is clear evidence in the historical weather records from many airports that weather conditions have changed significantly over the period of record, and the implications of these changes should be considered in the process of defin- ing a winter design storm. Depending on region, airports may experience changes in temperature regimes that affect the duration of the deicing season, necessitating more or less deicing product storage or more frequent deep freezing days that shift airport operations from defrosting aircraft to fully deicing aircraft and airfields. Temperature regime changes may also change the character and volume of deicing fluid- laden storm water that an airport needs to accommodate during storm events. Climate change may also result in an increase in the frequency or magnitude of intense precipita- tion events. Such an increase could result in more frequent exceedance of design conditions than those expected from use of the historic weather statistics. A simplified approach to addressing climate change impacts may be taken by looking at the historic winter precipitation records in isolated periods, distinguishing recent decades Date Precipita�on (in.) Runoff (in.) Storage (000s �3) Discharge (000s �3) Glycol (Gallons) Applied Fugi�ve Collected StormWater 1/2/1951 0 0 0 0 0 0 0 0 1/3/1951 0.44 0.4 0.1 0.04 1,000 360 300 340 1/4/1951 6.6 5.6 1.31 0.75 15,000 5,400 9,000 600 1/5/1951 3.9 3.4 1.75 0.75 12,000 4,320 7,200 480 1/6/1951 1.3 1.1 1.385 0.75 5,600 2,016 3,360 224 1/7/1951 0.5 0.4 0.775 0.75 275 99 83 94 1/8/1951 0 0 0.025 0.25 0 0 0 0 1/9/1951 0 0 0.025 0 0 0 0 0 Table 3-3. Example of time series data set of water and deicer budgets. Box 3-5. Estimating frequency of exceedance. The frequency of exceeding for a given level of a parameter (i.e., risk) can be estimated from historical data using the following Weibull formula: T NYRS M= +( )1 Where: T = return period in years, NYRS = number of years available in the data, and M = rank of the given event (events are ranked in descending order).

20 from earlier periods. An example of how this might be accomplished in identifying the winter design storm is pro- vided in the Portland International Airport (PDX) case study presented in Section 4.5. In this case, the full historical record showed significant climate change effects, with much more severe winter weather in Portland during the early part of the 20th century as compared to the latter decades of the century. With that recognition, the design storm was developed based upon just the most recent 22 years of weather data in order to more accurately represent current conditions. The reader is directed to the growing body of work on the topic of climate change adaptation at airports for detailed information on the implications of climate change to air- port infrastructure and design. In particular, the products of ACRP research on this topic (e.g., ACRP Synthesis 33: Airport Climate Adaptation and Resilience and research to come from ACRP Project 02-40) should be consulted. 3.6 Documenting the Basis for the Winter Design Event/Storm Figure 3-5 presents a template that can be used to docu- ment the application of the decision process to identify a win- ter design storm/event. Descrip�on of the project: _______________________________________________________________________ Factors at risk: ________________________________________________________________________________ Target level of service: __________________________________________________________________________ Dura�on of the design condi�on: One-day/short period Mul�day period Ra�onale: ________________________________________________________________________________ Key design driver: Volume of runoff Deicer load in runoff Ra�onale: ________________________________________________________________________________ Nature of available long-term records (parameter, period of record, frequency, other): Weather data: Aircra� deicer usage: Pavement deicer usage: Aircra� opera�ons/fleet mix: Deicing runoff controls: Is a robust risk analysis needed? Yes No Ra�onale: ________________________________________________________________________________ Frequency Analysis (if applicable) Period of meteorological record analyzed, and ra�onale: ___________________________________________ _________________________________________________________________________________________ Basis and ra�onale for water budget calcula�ons: ________________________________________________ _________________________________________________________________________________________ Basis and ra�onale for deicer usage es�mates: ___________________________________________________ _________________________________________________________________________________________ Basis and ra�onale for distribu�on of deicer loads among fate compartments: _________________________ _________________________________________________________________________________________ Design condi�on/storm corresponding to target level of service: ____________________________________ _________________________________________________________________________________________ Figure 3-5. Template for documenting the basis for the winter design event/storm.