Guidance for Treatment of Airport Stormwater Containing Deicers (2013)

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.

62 General considerations for deicer treatment system design and implementation are discussed in this chapter. 5.1 Sizing the Treatment System The basis of design for a treatment system provides information that is needed to complete the design and cost estimates for the selected deicer management system. A basis of design typically includes the following: • Treatment capacity needed. • Number and type of treatment units. • Sizes of supporting equipment. • Quantities of materials needed and generated (e.g., chemical, solids, biogas). • Expected performance (e.g., BOD removal efficiency). • Design condition (e.g., design event, design season). Establishing the required treatment system size or capacity is a key aspect of developing the treat- ment system basis of design because of its effect on cost, footprint, and performance. Considerations for determining the design treatment capacity are provided in the remainder of this section. 5.1.1 Understanding Treatment Capacity Parameters The capacity of a treatment system can be described using a number of different parameters, including: • Design flow rate, • Concentration range, and • Mass loading rate. Ensuring that the project team understands these parameters, how they relate to each other, and that the appropriate parameters are applied to each treatment technology is important to the treatment implementation process. 5.1.1.1 Definition of Treatment Capacity Terms and their Relationships Concentration. Concentration is the mass per unit volume, most often expressed in mg/L, of a single chemical constituent (e.g., PG, EG) or aggregate measurement parameter (e.g., BOD, COD, TOC) in stormwater. Flow Rate. The flow rate of water is the volume conveyed or processed per unit time, typically expressed as gallons per minute (gpm), million gallons per day (mgd), liters per minute, or cubic meters per day. C H A P T E R 5 Designing and Implementing Deicer Treatment Systems

Designing and Implementing Deicer Treatment Systems 63 Mass Loading Rate. The mass loading rate refers to the mass of stormwater constituents conveyed, processed, or treated per unit of time, expressed typically in pounds per day (lbs/day) or kilograms per day (kg/day). Mass loading rates (mass per unit of time) are calculated by multiplying the flow rate (volume per time) by the concentration (mass per unit of volume) of a given parameter: Mass Loading Rate (lbs/day) = Flow Rate (gpm) × Concentration (mg/L) × 0.0120 As seen from this equation, if the flow rates entering a treatment system are kept at a constant rate, changes in pollutant concentration result in an equivalent change in mass loading rate. Keeping the mass loading rate constant, a recommended operating condition for many treat- ment systems, requires changing the flow rate to adapt to the changing concentrations typical of deicing. If both flow rate and concentration fluctuate independently, such as has been the case with uncontrolled deicer management operations, the mass loading rate may be highly variable. Treatment performance in an uncontrolled system is likely to suffer. 5.1.1.2 Treatment Capacity Parameters Used with Various Technologies The capacities of recycling technologies, like reverse osmosis and mechanical vapor recom- pression, are typically described in terms of flow rate and the number of treatment units. The number of units required depends on the volume that must be processed, the PG or EG concen- tration, and the frequency of maintenance. Lower concentrations of PG or EG mean more water to separate from the glycol and a higher number of treatment units. Mass loading rate is not a term that is often directly used when considering recycling technologies, although a higher mass load to process will produce a larger volume of concentrated glycol product to sell. For biological treatment technologies, mass loading rate is most often the proper means of characterizing treatment system capacity. This is because the treatment system capacity is directly related to the mass of bacteria maintained in the treatment reactors. The mass of bacteria, in turn, is directly related to the mass of food (BOD) supplied to the system. Consideration of how to manage mass loading rate in a biological treatment system is important because a fluctuating BOD mass loading rate results in an unstable bacterial population. Unstable populations make the treatment system prone to poor treatment efficiency (e.g., when a spike load of deicer is applied to a low bacterial population, higher effluent BOD concentrations may result). Unstable bacterial populations also result in less than the full capacity of the system being used. Flow rates for biologi- cal treatment systems primarily affect the detention time, or the time that the BOD in the water is exposed to the bacteria. Unless the flow rate entering a biological treatment system is relatively constant, defining treatment system capacity strictly by concentration is not appropriate. Many of the most recently implemented biological deicer treatment systems, such as AFBRs and aerated gravel beds, are designed to maintain a near-constant mass loading rate by monitoring Treatment Tips Relationship of Treatment Capacity Parameters Understanding the relationships among mass loading rate, flow rate, and concentration in treatment system design and operation is essential. Determining which parameter will be used to control plant operations should be done early in the treatment design process.

64 Guidance for Treatment of Airport Stormwater Containing Deicers concentrations of BOD, TOC, or COD and adjusting flow rates in response. Higher measured concentrations result in the need to reduce flow rates to maintain a constant loading rate and vice versa. The allowable mass loading may be linked to temperature or the onset of excessive bacterial growth that could cause clogging. These types of constant-load operations often use a greater percentage of their available treat- ment capacity in comparison to uncontrolled situations. They are also less prone to upsets from variable deicing conditions. If a biological treatment system’s loading rate is controlled, the variation in the pollutant concentration is not a particularly important operational parameter if the system is operated within design conditions. 5.1.2 Calculating Required Mass Loading Treatment Capacity A mass balance provides the basic means of approximating the required mass loading rate that must be removed to meet governing discharge criteria, as shown in simplified form in the following equation: Ltreatment = Lstormwater − Ldischarge Where: Ltreatment = Approximate mass loading rate of treatment system (lbs/day), Lstormwater = Mass loading rate of untreated stormwater water (lbs/day), and Ldischarge = Allowable mass loading rate of discharged stormwater (lbs/day). To obtain an order-of-magnitude approximation of the mass load that needs to be removed by treatment for use in establishing costs during the alternative analysis process, Ltreatment and Lstormwater can be calculated for average conditions. Care must be taken, however, to understand the source of the calculated values. Many deicer treatment systems in the early history of deicer management were sized based on limited stormwater sampling data and flow data. They sub- sequently proved to be undersized because the variety of flows and concentrations were not well understood. In practice today, dynamic models that simulate deicer application conditions, stormwater runoff and routing, and deicer management measures are used to assess a broader array of potential conditions. Models can capture variation in deicing under much shorter time steps than is the case with calculations based on average calculations. The treatment capacities needed under more extreme conditions can be better understood in this way. In many instances, extreme deicing and runoff conditions dictate the required system capacity in terms of treated mass loading rate, flow rate, and storage capacity. Dynamic models are typically devised to jointly simulate the characteristics of the storm- water runoff, the required treatment mass loading rate, and the required storage volume. The optimal balance between storage volume and the treatment mass loading rate required to meet the governing discharge conditions for a variety of deicing circumstances is assessed in an iterative fashion. The assumptions associated with the model development and simulations should be well documented. In many cases, sensitivity analyses can be used to understand the potential impacts of model simulations. The advantage to including the storage volume in a dynamic model is that the effects of equalization and attenuation of concentrations can more readily be understood and the required volumetric limit can be determined in one step. The nature of deicing, with unpredictable weather conditions and uncertain future flight activity, makes it impossible to quantify a true worst-case condition. Attempts to project true worst-case deicing have shown that the worst of conditions tend to be true outliers that occur

Designing and Implementing Deicer Treatment Systems 65 extremely rarely. The design capacities needed to manage worst-case conditions can be several times larger, for example, than the design treatment capacities needed to fully manage 90% of the deicing events. As a result, worst-case design basis is defined by the airport in coordination with the regulatory authorities. The required treatment system design capacity is not typically a distinguishing criterion for the treatment technology selection phase, except as it relates to costs and space requirements. The size and number of units for virtually any treatment technology can be adapted to meet the required treatment capacity. In other words, it is rare that a treatment technology could be eliminated from consideration based on the inability to size the technology to the required capacity. The required treatment capacity, however, does have a direct bearing on capital and operating costs. The required design capacity is frequently affected by a number of site-specific factors, which are discussed in the following. Design capacity may need to be increased to account for the following: • Cold temperatures. Cold-weather treatment inefficiencies that cannot be managed by other means. • Influent variation. Insufficient storage or control of influent flow and loadings results in the need to add capacity to treatment. • Effluent quality. The system should have enough capacity to ensure that capacity is not the limiting factor in achieving the desired effluent quality. • Expansion. The future year to which the treatment facility is designed to operate can have a significant effect on the design capacity. Design capacity could be decreased based on design decisions on other components of the deicer management system, including: • More storage. Increased storage capacity coupled with using more of the non-deicing season to operate the treatment plant will decrease the needed treatment capacity. • Multiple treatment technologies. Use of multiple treatment technologies could reduce the total required design capacity and the cost of the primary deicer treatment technology. For example, discharge of loads to a POTW during rare extreme deicing conditions could allow the maximum capacity needed for an activated sludge plant to be decreased. • Monitoring and control. Monitoring and control of the flow rates and BOD concentrations entering treatment to attenuate peaks could decrease the maximum treatment capacity needed. These conditions affecting design capacity are further discussed in the following. Treatment Tips Design Capacity A simple mass balance can be used to get an order-of-magnitude approximation of the BOD mass loading rate that needs to be removed by treatment. Model simulations are used to more accurately account for the treatment mass loading and flow rates needed for the more extreme conditions. Site-specific factors like adjustments for low temperature, storage capacity, and future design conditions can affect the design treatment capacity.

66 Guidance for Treatment of Airport Stormwater Containing Deicers 5.1.2.1 Effect of Temperature on Design Capacity The cold temperatures characteristic of deicer-affected stormwater may affect treatment in multiple ways: • Decreased levels of microbial activity reduce the mass of BOD that can be treated in a biological system per unit mass of bacteria. • Increased water viscosity reduces flow rates through membrane systems. • Chemical nutrients do not dissolve as readily. • More energy input is required for evaporation-based systems. For biological treatment, it is well established in theory and in practice that without imple- menting measures to adapt to cold temperatures, treatment effectiveness will suffer. As part of the research for this guidebook, the effect of water temperature on degradation rates for both aerobic and anaerobic systems was assessed using a lab-scale biological respirometer. Biological seeds from actively functioning full-scale deicer treatment systems were used for the analyses. Tests were conducted at a variety of PG concentrations and temperatures. The results of the testing under aerobic conditions are shown in Figure 17. The y-axis represents the rate of off-gas production in the respirometer, which corresponds to the degree of biological activity (higher values = higher degradation rate and better treatment). The x-axis is PG concen- tration. As shown, the degradation rates at 36°F were approximately half of the rates at 41°F and over five times less than the degradation rates above 48°F. The relationships were similar over a PG range of between 100 mg/L and 7,000 mg/L, indicating no relationship between concentration versus temperature, and degradation rate within that range. Not surprisingly, the temperature effects for anaerobic bacteria are even more pronounced (Figure 18). For anaerobic treatment, little or no degradation was observed below 60°F. There were marked differences in degradation rates between 75°F and 85°F. The PG concentration within the range tested (100 mg/L to 8,000 mg/L) did not change the basic temperature effects. The temperature data from the lab testing aligned with results observed in the field. Anaerobic systems like the AFBR are specifically designed to function between 85°F and 90°F because that is where the anaerobic bacteria function best. Many aerobic systems, especially those that do not take specific measures to help counter the temperature effects, have seen significant decreases in 0 5 10 15 20 25 30 35 40 45 50 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000PG D eg ra da ti on R at e (b as ed o n m ea su re d ga s) PG Concentration (mg/L) Aerobic Respirometer Test Data 60 Degrees 48 Degrees 41 Degrees 36 Degrees Figure 17. Relationship of water temperature and aerobic degradation rates.

Designing and Implementing Deicer Treatment Systems 67 performance during the coldest months. During the coldest period for many northern airports, the water temperatures are frequently less than 41°F when the stormwater enters the treatment system. Within a biological treatment system, the activity of the microorganisms actually releases energy. The extent to which water temperatures increase or decrease within a treatment system depends on the energy losses and gains. Factors in the heat balance include: • Local climate conditions (ambient air temperature, wind speed, solar heating), • Exposure of the treatment unit water surface to the air, • Insulation from the ground, • Insulation of tank walls, • Reactor surface area, and • Heat loss (evaporative cooling) from surface aeration or blowers. Without taking steps in design and operation to counteract the cold temperatures and subsequent effects on treatment, the treatment efficiency of a biological treatment will suffer. A number of measures can be taken to address temperature effects on treatment capacity, including: • Storage and flow control. – Holding water collected during winter in storage until temperatures increase (has net impact of increasing storage requirements). – Reducing mass loading rates during cold periods (has net impact of increasing storage requirements and running the treatment system for a longer period in a season). – Some systems have evidence that reducing mass loading rates prior to cold periods helps the bacterial population recover faster when the temperatures warm. – Reducing the volume of water that is collected from the runoff (e.g., with deicing pads) to reduce the volume of water that has to be heated. • Heat management. – Insulation of potential heat loss surfaces to better take advantage of heat generated by the biological activity. 0 10 20 30 40 50 60 70 80 90 100 0 2,000 4,000 6,000 8,000 PG D eg ra da ti on R at e (b as ed o n m ea su re d ga s) PG Concentration (mg/L) Anaerobic Respirometer Test Data 85F 75F 60F Figure 18. Relationship of water temperature and anaerobic degradation rates.

68 Guidance for Treatment of Airport Stormwater Containing Deicers – Providing external source of heat. � Heat exchange with hot water source fueled by: • Methane captured from anaerobic systems. • Natural gas. � Heat exchange between effluent and influent flows if the effluent has higher temperatures from the treatment process. � Geothermal heat. � Mixing with warmer wastewater such as sanitary sewage. � Treatment units contained in a heated building. • Bacterial population management. – Increasing the biomass concentrations in the treatment reactors prior to cold weather. An increase in the number of bacteria can at least partially offset reduced treatment rates per unit mass of bacteria. This is primarily a strategy for short-term management. • Increased reactor capacity. – Building in larger treatment capacity provides more opportunity to maintain a larger biomass population prior to treatment. However, a larger reactor can also make it more difficult to sustain the biomass during periods when the amount of deicer available to treat is low. The specific measures that can be employed depend on the specific treatment technology characteristics as well as local infrastructure and resources. 5.1.2.2 Effect of Effluent Quality Targets on Design Capacity A number of factors can affect the effluent pollutant concentrations for BOD, COD, PG, and EG that can be achieved by treatment: • It is generally thought that anaerobic treatment results in higher effluent concentrations for BOD and COD than aerobic treatment, although the difference may not necessarily be as significant for deicer-affected stormwater as it is for other wastewaters. • For a given flow rate, higher influent concentrations may result in higher effluent concentrations if the treatment system does not have enough treatment load capacity. • A lack of nutrients or other essential conditions necessary during certain times for biological treatment will result in higher concentrations. • When there are an insufficient number of membrane units in a reverse osmosis system, it will be unable to reduce the concentration of the dilute stream. When considering treatment capacity, the capacity should be large enough that it is not a limiting factor in achieving effluent limits. Understanding that variable conditions will occur, it is good practice to design with a safety factor in treatment capacity that can provide a buffer against those variations and allow target effluent concentrations to be met more frequently. 5.1.2.3 Effect of Storage Volume on Design Capacity In the process flow schematic of many deicer management systems, storage is located between the collection system and the treatment system. Storage is most typically viewed as a means of capturing large runoff events. Storage structures also provide an equalization function, where some of the variation in deicer concentrations in the collected runoff can be attenuated. Storage capacity is related to treatment capacity. Both are related to how long an on-site treatment system runs during a deicing season. As shown in the Treatment Examples: Balance Between Storage and Treatment at Portland International Airport text box, higher storage volumes generally allow lower treatment capacity. The trade-off does require longer running time for the treatment system into the warm months. The maximum degree that storage can be increased in this scenario is typically bounded by the practical value in being able to drain the storage structure by the start of the next deicing

Designing and Implementing Deicer Treatment Systems 69 season. The inability to treat all water before the start of the deicing season could lead to several issues, including: • Degradation of the stormwater stored in the tanks, in particular over warm weather months. For a recycling system, extended storage means loss of glycol product because of the degradation. For biological treatment systems, extended storage may cause treatability issues because the biology may not be sufficiently acclimated to the breakdown products of the primary deicer constituents. • Breakdown of the deicer into volatile fatty acids during warm weather decreases the pH to as low as 3 to 5. • Starting the season with water in storage could lead to a cascading effect where the treatment facility cannot catch up, and alternative means of disposal of the tank contents becomes necessary. • Operations staff members, who often have other responsibilities, are needed to tend to the treatment facility for a longer period during the year. The relative costs of storage capacity and treatment capacity need to be weighed. Storage can also require a greater footprint than some treatment technologies, which could be an issue in situations Treatment Examples Achievable Effluent Quality at Akron–Canton Airport (CAK) It is a generally accepted view, derived primarily from experience with municipal and industrial wastewater, that achievable effluent COD concentrations from anaerobic treatment systems are substantially higher than achievable effluent concentrations from aerobic treatment systems. Several years of operating data from the CAK’s AFBR treatment system were reviewed in this research to further define the low range of anaerobic deicer treatment effluent quality. The conclusions indicate that anaerobic deicer treatment can achieve COD effluent concentrations significantly lower than previously thought, although not as low as some aerobic systems. The CAK AFBR system includes two anaerobic reactors running in parallel, where the vast majority of the COD is removed, and a dissolved air flotation system for removal of biological solids. For the 2008–2009 and 2009–2010 deicing seasons, the CAK AFBR ran for a total of 296 days (excluding the start-up periods) at average COD mass loadings 15% higher than the system design capacity (3,400-lbs COD/day). Over those two seasons, the treated effluent averaged 151-mg/L soluble COD (i.e., COD without solids) directly from the anaerobic reactor and averaged 39-mg/L soluble COD in the effluent from the dissolved air flotation unit that is integral to the anaerobic treatment system. (PG concentrations were non-detectable in lab analyses, and no BOD analyses were conducted.) During this period, the system removed 99.77% of the influent COD load. See the airport summary for CAK in Appendix D for graphs showing the range of effluent concentrations in the 2009–2010 season. It is theorized that the ability of the CAK system to achieve lower concentrations than thought with anaerobic treatment may be related to the vast majority of COD being derived from soluble glycols rather than more difficult to degrade soluble and insoluble COD found in municipal and industrial wastewater.

70 Guidance for Treatment of Airport Stormwater Containing Deicers where space is at a premium. The primary advantage to storage is that there is significantly less operational burden and cost to operate a storage structure compared to treatment. It is also often easier to add storage than to expand treatment capacity in response to airport growth. 5.1.2.4 Effects on System Influent Control on Design Mass Loading Rate Even with the partial equalizing effects of storage prior to treatment, most facilities experience significant periods of variability of deicer concentrations routed from storage to treatment. Biological treatment systems function best when the mass loading rate in the treatment influent is relatively constant because the bacterial populations tune themselves to the quantities of BOD to be treated. An increasing number of airports are finding that controlling the mass loading rate entering the treatment systems helps to stabilize operations and provide for more predictable effluent quality. Control of mass loading rates can be achieved by monitoring concentrations, typically of COD or TOC, at least once a day and making adjustments to the flow rate. The control can mostly be manual in nature (grab samples, on-site lab tests, manual flow adjustments) or more computer controlled (online monitoring, changing of flow rates through automatic valves or variable frequency drives on pumps). The level of control is related to the needed treatment capacity because the system can be designed for a lower mass loading capacity if the peak concentrations are dampened (Figure 19). Treatment Examples Balance Between Storage and Treatment at Portland International Airport (PDX) In the process of determining the required design mass loading capacity and design storage capacity for the PDX deicing system enhancement, a model was used to simulate the various combinations of storage and treatment capacity that would meet compliance objectives for discharges to surface waters. As shown in the graph, various combinations of storage and treatment would be able to meet the objectives. Once the storage–treatment capacity curve was determined, combined costs for storage and treatment were calculated to find the minimum cost point.

Designing and Implementing Deicer Treatment Systems 71 5.1.2.5 Effects of Using Multiple Treatment Technologies on Design Mass Loading Rate Many airports use multiple treatment technologies within a single treatment system. Frequently, the collected runoff is segregated into fractions based on PG, BOD, COD, or TOC concentration. Treatment technologies that best align with the individual fractions are selected. Using multiple treatment technologies may allow the capacities of the individual technologies to be reduced because of the additional flexibility gained in managing flows during variable conditions. At times, the features of multiple technologies can be used in a complementary way to help reduce design capacity. In that scenario, the treatment technologies would be applied to the fraction of the collected stormwater stream that is optimal for their conditions. For example, if an AFBR and aerated lagoon were used in combination, the excess biogas captured from the AFBR could be used to heat water entering the aerated lagoon, thus improving the lagoon’s cold-weather efficiency. 5.1.2.6 Allowances for Variation in Weather Conditions An important design point for the treatment plant capacity (as well as storage) is the extent to which the systems are sized for the most extreme conditions. Most often this is represented as a “design event,” “design storm,” or “design season.” The basis-of-design decision for sizing based on weather conditions is typically something that needs to be discussed with regulators. No specific regulatory standard exists in the United States for setting a design event or design season for deicer management. Treatment situations in some other industries, such as the design storms associated with combined sewer overflows, do have design standards. The basis- of-design condition can have significant consequences. For example, past modeling studies have shown that the treatment capacity needed for a 50-year-recurrence design season can be 1.5 to 2 times larger than the capacity needed for a 10-year-recurrence season. ACRP Report 81: Winter Design Storm Factor Determination for Airports discusses design storm methodologies. 5.1.2.7 Allowances for Future Growth and Expansion Airports will typically design their storage and treatment systems to have the capacity needed at some future date. In most cases, this means having allowances in capacity for future growth in airport operations. Deicer application volumes at a future date driven by changes in future operations can be subject to many variables, so it is important to have a thorough and agreed- upon understanding of the assumptions for future conditions. Most airports will design for a point in time between 10 and 20 years in the future. If the treatment capacity is designed for a point in the future, the system may have excess capacity in the short-term. Operating at a lower capacity than designed may have an impact on the treatment system’s short-term performance (positive or negative). M as s Lo ad in g Ra te Time in Deicing Season Required Treatment Design Capacity with No Influent Flow Control Required Treatment Design Capacity with Influent Flow Control Figure 19. Accounting for effects of flow control on maximum required capacity.

72 Guidance for Treatment of Airport Stormwater Containing Deicers 5.1.2.8 Use of Modular Systems to Split Required Design Capacity Many times, treatment systems are constructed as multiple smaller units rather than one large unit. Using multiple smaller units offers a number of advantages, including: • Ability to run the system in series or parallel mode, • Ability to have one or more treatment units offline for maintenance while continuing operations, • Operational costs managed by only operating enough units to meet the current treatment demand, • Better process control (e.g., more efficient mixing), • Easier expansion to meet future demand, and • Better effluent quality often achieved than with single, larger treatment unit. As a result, the basis of design for many treatment systems provides the capacity of individual units and the number of units to reach the required treatment capacity. 5.1.3 Design Concentrations and Flow Rates Conventional thinking for municipal and industrial wastewater treatment is that design concentration and design flow rate are defining design parameters for biological treatment system sizing. However, the huge variability in BOD concentrations and flow rates in deicer- affected stormwater result in the need for a different perspective on the significance of the terms “design concentration” and “design flow rate.” In most situations with biological treatment of deicers, the need to control the mass loading rates resulting from the variability in deicer-affected stormwater results in there being relatively little significance to the term “design concentration.” In a facility where mass loading rates are controlled, the fluctuation in concentrations does not really affect the biomass population in the treatment system or the treatment efficiency. Design flow rates are somewhat more important because of the impact on pump and pipe sizes, although the design flow rates are less a selected design point than a consequence of the selected mass loading rate and the BOD concentrations in the influent flow. For recycling-based systems, design concentrations are important because they relate to the volume of water that must be evaporated. Therefore, for them to be economical, it is important that minimum thresholds for PG concentration be met (typically 1% or greater for reverse osmosis and MVR recycling technologies). Design flow rates for recycling systems typically affect the number of treatment units (MVR or reverse osmosis) that are required. 5.1.4 Relationship of Design Capacity, Cost, and Risk When final treatment capacity decisions are made, it is important to understand the risks associated with the selected capacity. The decision may be made to construct a smaller system (usually because of insufficient funds) or oversize the system (usually to account for future growth or to add a greater degree of certainty of compliance). There are identifiable risks associated with undersizing or oversizing treatment, including: • Undersizing a treatment system can potentially result in: – Performance risk from added stress on the treatment process if the system has to be over- loaded to avoid overtopping storage during heavy deicing periods. – Compliance risk if the treatment system cannot process all stored volume within a calendar season. – Planning and development risk if there is insufficient capacity to accommodate future growth at the airport.

Designing and Implementing Deicer Treatment Systems 73 • Oversizing a treatment system can potentially result in: – Unnecessary capital and operating costs. – Compliance risk because of the difficultly in maintaining a continuous and stable treatment operation. – Compliance risk because of greater challenges with seasonal start-ups. Finding the sweet spot for selecting the right treatment system size is one of the greater challenges in treatment system design. Over much of the range of potential treatment capacities, there is a direct relationship between the treatment system capacity (cost) and the probability of an effluent exceedance. Generally, increasing the treatment system capacity will decrease the risk of noncompliance, at least until the point where continuous treatment facility operation cannot be achieved. Eventually, a point of diminishing returns is reached where additional investment in treatment capacity yields smaller reductions in the risk of noncompliance. There- fore, the decision-making team should consider both the benefits and costs of potential design capacity points. 5.2 Treatment Support System Design In this guidebook, treatment support system is defined as a unique process that is typically required to allow the primary treatment system process to function appropriately. Five categories of support systems are discussed in the following in general terms. Specific discussion of the support systems needed for each technology can be found in the treatment technology fact sheets. 5.2.1 Pretreatment Some deicer treatment systems may require pretreatment of the deicer-affected stormwater prior to the stormwater entering the primary treatment process. The pretreatment is typically needed to modify the characteristics of the stormwater to either protect the primary treatment process or make it more effective. Typical pretreatment processes that may be applied include: • Removal of large debris (through screening), • Removal of grit (through sedimentation), • Removal of TSS (through settling or flotation), • Increase in water temperature (typically using heat exchangers), • Adjustment of pH (through chemical addition), • Addition of chemicals to reduce likelihood of biofouling, and • Removal of oils, grease, and other petroleum products (through oil–water separator). The physical treatment processes for recycling tend to be more sensitive to stormwater contami- nation from non-deicer constituents because of the potential fouling of the treatment structures, which hurts processing efficiency and increases maintenance costs. Therefore, these technologies have evolved to include multistep pretreatment operations that protect the primary treatment units and allow them to function optimally. In some cases, certain pretreatment processes can be combined, such as pretreatment units that remove large debris, TSS, and oil and grease. The presence of dissolved solids (measured as TDS) in stormwater can be a particularly problematic issue. At times, the presence of pavement deicers can cause very high spikes in TDS concentrations. High TDS concentrations can negatively affect both biological and physical treatment systems. It can be difficult to treat deicer-affected stormwater to remove TDS. Biological systems will not remove dissolved solids. The effect of TDS and the need for pretreatment are usually mitigated through storage and equalization prior to treatment. While the peak TDS concentration during events of heavy pavement deicer use and low runoff volumes can be high,

74 Guidance for Treatment of Airport Stormwater Containing Deicers the dilution provided during the remainder of the time usually reduces average TDS concentrations in storage to avoid treatment impacts. However, the effect of TDS on deicer treatment systems has not been well-studied. If more significant TDS removal is needed, pretreatment processes such as chemical softening, ion exchange, and reverse osmosis can be used to remove certain constituents contributing to TDS. 5.2.2 Nutrient Management Nutrient addition is an essential component to biological treatment. The microorganisms performing the treatment need nutrients for new cell synthesis. There is significant evidence that lack of nutrients in biological deicer treatment systems will severely restrict, if not inhibit, treatment of deicers because not enough new cells can grow and use the deicer constituents. Lab testing performed for this research, for example, indicated that withholding nutrients can decrease the treatment rate by as much as 40% in a short period (1 week). This effect will be continued as long as an insufficient nutrient concentration exists in the system. While some wastewaters, such as sanitary wastewater, contain sufficient nutrients, deicer- affected stormwater does not. As a result, nutrients need to be added to most biological deicer treatment systems. The need for nutrients and their absence from deicer-affected stormwater mean that biological deicer treatment technologies need a support system for storing, mixing, and metering nutrients into stormwater prior to or within the primary biological treatment process. Deicer-treatment plant operators coming from a sanitary wastewater treatment background must adapt to the idea that nutrients are essential and that regular additions must be made. One of the more complex job functions for deicer system operators is the balancing of nutrient additions such that enough nutrients are added to adequately support the bacteria without over- loading the stormwater to the point that effluent limits in permits (if present) for ammonia-nitrogen or phosphorus are exceeded. This can be especially difficult if the COD load to be treated fluctuates. Often, nutrients in biological deicer treatment systems that need to be added continuously are classified as macronutrients. Nutrients that only need to be added on an occasional basis are classified as micronutrients. Nitrogen and phosphorus are almost always macronutrients. Other nutrients that may need to be added include: • Sulfur, • Iron, • Magnesium, • Potassium, • Calcium, • Sodium, and • Small amounts of additional minerals for anaerobic bacteria. Most often, nitrogen and phosphorus are the most critical nutrients to add. In anaerobic systems, sulfur is also critical and therefore considered a macronutrient requiring continuous feeding. Deicer-affected stormwater also often has insufficient amounts of other nutrients. While these nutrients do not have to be added in the same quantities as nitrogen and phosphorus, their absence can negatively affect deicer treatment in a biologi- cal system. Nutrient additions are typically paced to the organic (COD) load to reduce the likelihood of overfeeding or underfeeding of the biomass. Typically, the nutrients come in solid form and are mixed into solutions for larger systems. For smaller systems, prepurchased chemical solutions may be economical. A typical Treatment Tips Nutrients in Biological Treatment Understanding the role of nutrients in biological treatment is essential. In most situations, nutrients must be added regularly for a biological deicer treatment system to function well.

Designing and Implementing Deicer Treatment Systems 75 nutrient feed system includes mixing/storage tanks, metering pumps, and tubing to the injec- tion point. Many of the existing biological deicer management systems have experienced instances of unintentional or accidental impacts from lack of nutrients. Lessons learned from nutrient man- agement in biological deicer treatment systems include: • At one airport, lack of phosphorus addition for an extended period resulting from inaccurate laboratory analysis of the treated effluent resulted in treatment efficiency dropping by over 50%. When the phosphorus began to be added again, treatment efficiencies returned to nor- mal within days. • Nutrient addition needs, especially for nitrogen, are especially great during the system start- up at the beginning of the deicing season. This appears to be true for many different types of biological technologies. Nitrogen loadings several times higher than the normal loading may be necessary for several weeks at start-up. • One airport found that adding nutrients alone prior to the addition of BOD at the system start-up helped to speed up the overall start-up process. • The dying bacteria in a biological treatment system, especially after a summer shutdown, will typically release significant quantities of nutrients back into the water. It may be difficult to achieve the correct nutrient balance during this period. 5.2.3 Biogas Management The biological treatment systems using anaerobic processes produce methane, which can be captured and used as fuel. This includes the anaerobic fluidized bed reactor and the anaerobic digesters at POTWs. For AFBRs with influent COD concentrations of greater than 2,100 mg/L, the captured gas is enough to heat the incoming water for most of the deicing season once the start-up period is over. During the start-up period, natural gas is needed because the quantity of biomass in the treatment system is insufficient to generate the required methane. Methane is the primary component of biogas from an anaerobic treatment system and is the primary component in natural gas. As such, any system with the potential to use natural gas may be able to use the methane captured from an anaerobic reactor if the gas handling and burning equipment is adequately configured for both. If enough methane is produced to meet the treated water heating demands, excess methane can be used to heat buildings or sand, melt snow, or produce electricity. Treatment systems that produce biogas containing methane may have gas management support systems that contain the following components: sealed piping for gas collection, boilers for burning methane, temperature monitoring, heat exchangers, and flares for burning excess methane. Flares without visible flames can be used. 5.2.4 Monitoring and Control Systems One of the emerging trends in wastewater treatment in general, but more specifically in deicer treatment, is using online control of various elements of the treatment process to improve efficiency and predictability and reduce capital costs through reductions in the size of storage and treatment. Control system components include: 1. Instruments for monitoring temperatures, pressures, flow rates, pollutant concentrations, pH, and water level; 2. Programmable logic controllers (PLCs) for receiving the instrument inputs, performing calculations, and outputting signals to start, stop, and adjust operating conditions for equip- ment such as pumps, blowers, and valves;

76 Guidance for Treatment of Airport Stormwater Containing Deicers 3. Programming loaded onto the PLC to provide the process logic and control mechanisms; and 4. Means for recording process data. Together these components are typically called the supervisory control and data acquisition (SCADA) system. SCADA systems take some of the operating responsibilities from the operating staff for direct monitoring and control, but may increase operator requirements to manage and interpret data. The finer degree of control provided by a SCADA system helps to reduce the variation in the process and helps the treatment system respond more quickly to changes in conditions. It also provides additional information on process performance to help assess ways to optimize the system and help troubleshooting. In some situations, such as when a control system is applied to manage the influent mass load- ing to a treatment system, the finer degree of control has the net effect of reducing the required treatment or storage system capacity by reducing the need to size to peak conditions. 5.2.5 Post-Treatment Biological Solids Management In any biological treatment system, new microorganisms are continually formed and old organisms die. If a steady population of microorganisms is to be maintained, the excess micro- organisms must be removed. Microorganisms contained in the treated effluent must also be removed to meet effluent limits for TSS. The removed microorganisms are often called biological solids or sludge. The required biological solids management system varies depending on several factors, including: • The type of process used. Anaerobic processes produce quantities of biological solids that are approximately 10 times less than the biological solids produced by many aerobic processes because the anaerobic bacteria grow more slowly. • How well the biological solids settle. Anaerobic processes may have biological solids that are more difficult to settle by gravity. As a result, processes like dissolved air flotation systems may be required. • The degree of dewatering that is required. This depends to some extent on how the biological solids will be disposed of. If they are transported off-site, it is expensive to transport solids with a significant water content, and some dewatering on the site may be required. • The disposal method for the solids. The design and operational impacts of solids management should not be underestimated. In the design phase, the likely characteristics of the biological solids, as well as the quantities to be produced, should be carefully evaluated. If there is insufficient thought given or insufficient funds allocated to this phase of the system design, there may be significant operational costs and hassle. 5.3 Guidance on Deicer Treatment System Implementation 5.3.1 Construction and Commissioning The time required for construction of a deicer treatment system can vary from several months to over a year. The construction time depends on: • The complexity of the project, including the number of different systems that must be constructed, • Where the construction is occurring (e.g., inside or outside secure areas),

Designing and Implementing Deicer Treatment Systems 77 • When the construction is started (construction of some elements may not be feasible in winter), • Weather delays during construction, • The degree to which components or unit processes come prepackaged, • The extent to which existing infrastructure is used, and • When funding is available. For many airports, one of the primary construction considerations is minimizing interference with airport operations. A safety and phasing plan is often required to identify where and when the construction is occurring, in addition to haul routes and staging areas. Significant coordination between the contractor and owner is required to manage operational impacts. The last step of construction is the process for checking out and testing the individual pieces of equipment and the instruments, as well as the checking out of the deicer treatment system as a whole. Inadequate testing of the constructed treatment system can have significant effects on consistency of operations, compliance risk, the ability to reach design capacities, and cost. The process for checking and testing the constructed treatment system can take the form of a formal commissioning process conducted by a third party or the design engineer and contractor working together to check out operation of the system as a whole. For complex deicer treatment systems, the commissioning process can take 6 to 12 weeks, but in most cases it is well worth the time and investment. 5.3.2 System Start-Up and the First Year of Operation The system start-up and first year of operation make up a critical period. Because deicer treatment systems involve many pieces of equipment and instruments working together in a dynamic environment with changing stormwater conditions, some issues with design and con- struction only come to light when the system operates as a fully functional unit treating collected stormwater. In addition, the start-up and first year of operation are the period when operations and maintenance personnel first get hands-on experience in running the system. In the case of a biological treatment system, the start-up is also the period when the bacterial population is first established. Often, biological treatment systems require a source of seed to start the system. Aerobic systems require seed from another aerobic treatment system, and anaerobic systems require anaerobic seed. Obtaining a quality seed will speed the start-up process. A biological treatment system starting up for the first time can take 2 to 4 months to reach the full treatment capacity, with anaerobic systems expected to take longer than aerobic systems because anaerobic bacteria are slower growing. As a result of these conditions, the start-up and first year are usually the most difficult period of operation for a deicer treatment system. In some cases, early performance of new systems is not representative of the system’s long-term ability to treat. Effective design, efficient construction, operator training, and well-planned testing can reduce the likelihood of issues in the start-up period, but airport management should be aware that a breaking-in period for the treatment system is to be expected. 5.3.3 Long-Term Operations and Maintenance Most deicer treatment systems require some attention from operators in order to perform adequately. Typical operator functions include: • Making decisions on feeding of deicer-affected stormwater into treatment; • Adding the correct amounts of the right nutrients to biological systems at the right times; • Making process adjustments based on influent stormwater characteristics, effluent quality, and process monitoring data;

78 Guidance for Treatment of Airport Stormwater Containing Deicers • Troubleshooting; • Sampling, monitoring, and lab analysis; • Data entry, review, and analysis; • Performing maintenance; • Aligning treatment system operation with the rest of the deicer management system; and • Reporting to regulatory authorities. It is not always required or essential to have licensed wastewater treatment operators for deicer treatment facilities, but training of the operators in the specifics of the treatment system operation is essential. It is generally recommended that operators begin work on the treatment system no later than the testing and commissioning phase at the end of construction. If other airports use similar technologies, training with operators of those facilities is recommended. The number of operators required depends on the complexity of the deicer treatment system and other duties assigned to operators. Many deicer treatment systems can be run with one to two full-time operators. Choosing the right operators is critical. The airport’s management team also plays a critical role in successful deicer treatment. Managers should understand the basic system operations. An understanding of the capacities and limitations of the deicer treatment system based on system capabilities is also critical. This includes understanding what constituents can be treated, stormwater constituents than can affect system performance (e.g., presence of spilled fuel), the maximum treatment capacity, expected effluent quality, and expected treatment efficiency. Since most deicer treatment systems include a variety of electromechanical equipment, a preventative maintenance program is a necessity. It is also critical to work out procedures with those performing maintenance to have fast maintenance response for system components in need of repair. Owner/Operator Management Tips for Successful Treatment Systems • Have design engineers demonstrate the relationship between cost and compliance risk. • Document and understand the system’s capacities and operational limitations. • Hire and train qualified and engaged operators. • Implement a monitoring system for appropriate parameters. • Track and regularly assess system operational parameters. • Implement a preventative maintenance system. • Ensure that short-term maintenance support is available in a timely manner.