Impacts of an Ammonia Leak on the Cabin Atmosphere of the International Space Station Stephanie M. Duchesne1 Wyle Integrated Sciences and Engineering, Houston, TX, 77058 and Jeffrey J. Sweterlitsch, Ph.D.2 National Aeronautic and Space Administration (NASA) Johnson Space Center (JSC), Houston, TX, 77058 and Chang H. Son, Ph.D.3 The Boeing Company, Houston, TX, 77059 and Jay L. Perry4 National Aeronautic and Space Administration (NASA) Marshall Space Flight Center (MSFC), Huntsville, AL, 35811 Toxic chemical release into the cabin atmosphere is one of the three major emergency scenarios identified on the International Space Station (ISS). The release of anhydrous ammonia, the coolant used in the U.S. On-orbit Segment (USOS) External Active Thermal Control Subsystem (EATCS), into the ISS cabin atmosphere is one of the most serious toxic chemical release cases identified on board ISS. The USOS Thermal Control System (TCS) includes an Internal Thermal Control Subsystem (ITCS) water loop and an EATCS ammonia loop that transfer heat at the interface heat exchanger (IFHX). Failure modes exist that could cause a breach within the IFHX. This breach would result in high pressure ammonia from the EATCS flowing into the lower pressure ITCS water loop. As the pressure builds in the ITCS loop, it is likely that the gas trap, which has the lowest maximum design pressure within the ITCS, would burst and cause ammonia to enter the ISS atmosphere. It is crucial to first characterize the release of ammonia into the ISS atmosphere in order to develop methods to properly mitigate the environmental risk. This paper will document the methods used to characterize an ammonia leak into the ISS cabin atmosphere. A mathematical model of the leak was first developed in order to define the flow of ammonia into the ISS cabin atmosphere based on a series of IFHX rupture cases. Computational Fluid Dynamics (CFD) methods were then used to model the dispersion of the ammonia throughout the ISS cabin and determine localized effects and ventilation effects on the dispersion of ammonia. Lastly, the capabilities of the current on-orbit systems to remove ammonia were reviewed and scrubbing rates of the ISS systems were defined based on the ammonia release models. With this full characterization of the release of ammonia from the USOS TCS, an appropriate mitigation strategy that includes crew and system emergency response procedures, personal protection equipment use, and atmosphere monitoring and scrubbing hardware can be established. 1 ISS Environmental Control and Life Support Systems Subsystem Manager, EC6 Crew and Thermal Systems Division, 1290 Hercules Avenue Suite 120, Houston, TX, 77058. 2 Insert Job Title, Department Name, Address/Mail Stop, and AIAA Member Grade for second author. 3 Insert Job Title, Department Name, Address/Mail Stop, and AIAA Member Grade for fourth author (etc). 4 Insert Job Title, Department Name, Address/Mail Stop, and AIAA Member Grade for fourth author (etc). 1 American Institute of Aeronautics and Astronautics Nomenclature atm = Atmosphere °С = Celsius °F = Fahrenheit g = Grams in = Inch kg = Kilograms kPa = Kilopascal mm = Milimeters m3 = Cubic meters ppm = Parts per million psi = Pounds per square inch I. Introduction T HERE are three classes of emergencies identified on the International Space Station (ISS): Fire, Depressurization and Toxic Release. In the case of a toxic release, a leak of Ammonia into the cabin atmosphere from the External Active Thermal Control System (EATCS) is considered the most catastrophic for both crew health and vehicle. Though this type of a leak is considered low probability, the consequence justifies steps to mitigate the hazard magnitude. To this end, it is necessary to understand the mechanism for ammonia entry into the cabin atmosphere, the dispersion rate of ammonia through the ISS cabin, and the capability to remove ammonia from the atmosphere once the leak is isolated. A detailed description of the ammonia chemistry, and ISS ammonia leakage scenarios are discussed in Ref 1. II. Ammonia Release into the ISS Atmosphere: EATCS IFHX Rupture The largest source of ammonia aboard ISS resides in the EATCS which provides heat rejection for the U.S. On- orbit Segment (USOS) and International Partner (IP) modules. Ammonia is not used on the Russian On-orbit Segment (ROS). The EATCS is a pumped single-phase ammonia system that collects heat from cold plates and heat exchangers and rejects it to space. The internal thermal control system (ITCS) uses pumped single-phase water to remove heat from USOS and IP systems and payloads and rejects heat to the EATCS via an interface heat exchanger (IFHX). The ITCS consists of a low temperature loop (LTL) and a moderate temperature loop (MTL). The lowest design pressure within the ITCS is the gas trap located in the pump package assembly. For this reason, it is theorized that the gas trap would be the point of entry of ammonia into the ISS cabin as the ITCS loop overpressurizes after the IFHX failure. There are six prime gas traps aboard the USOS: 2 in the US Lab, 2 in Node 2, and 2 in Node 3. Hazard assessments identified three scenarios that could lead to a breach in the water/ammonia barrier of the IFHX that in turn would cause high pressure ammonia to pass from the EATCS into the ITCS and then into the cabin: 1) IFHX core freeze/thaw, 2) Structural Failure of the IFHX, or 3) Over-pressurization of the IFHX. The IFHX has a low probability of failing due to proper selection of materials, design, and testing as well as the inclusion of relief valves, bleed lines, software controls and procedures. The existing NASA probabilistic risk assessment (PRA) model shows that the likelihood of an ammonia leak due to IFHX core freezing is 1 in 1,406,074.2 Assuming proper software control and procedures are followed, that number drops to 10-19.2 2 American Institute of Aeronautics and Astronautics Figure 1. IFHX Leak The IFHX failure scenario is broken down into 3 types of leaks: micro leak, moderate leak, and rupture. The micro leak is considered the result of micro cracks or corrosion of the IFHX and is detected through periodic sampling of the ITCS water loop. Remediation steps are taken as necessary upon detection. A moderate leak is detected through monitoring of the ITCS accumulator quantity. This type of leak could be detected through sampling of the ITCS, crew sense of smell, or a slow unexplained rise in the ITCS accumulator quantities. A rupture is detected by the rapid filling of the ITCS accumulator and/or pressurization of the affected ITCS loop. A rupture is considered anything larger than a pin hole sized breach (> 1.27 mm (0.05 inches)). In response to a rupture of the IFHX and filling of the ITCS accumulator, ISS software will automatically isolate the affected IFHX. However, as seen in Fig. 1, a bleed line exists within the IFHX that prevents full isolation of the IFHX. The bleed line is 0.81 mm (0.032 inches) 3 in diameter. The EATCS operates around 2000kPa (300psi) with a maximum design pressure of 3447 kPa (500psi). The ITCS operates around 190 kPa (28psi) and has a maximum design pressure of 690kPA (100psi). With this high pressure differential it is possible that an IFHX breach can propagate from a micro leak to a rupture scenario. In order to understand the impacts of an ammonia breach on the ISS vehicle and to develop a crew emergency ammonia response strategy, only the rupture case has been analyzed. III. Toxicity of Ammonia As discussed in Ref. 1, ammonia is highly irritating to the eyes, mucous membranes, and respiratory tract. At low levels ammonia acts as an irritant and quickly becomes lethal as levels rise. The National Institute for Occupational Health and Safety (NIOSH) standard for immediate danger to life and health (IDLH) is 300ppm. As will be shown in this paper, the levels of ammonia that could be present in the ISS cabin as a result of an IFHX ammonia leak far exceed this standard. Little data exists on the health effects of ammonia beyond low level exposure. Industry standards dictate immediate evacuation for moderate to high levels of ammonia release. Unfortunately immediate evacuation into a clean zone (area not requiring Personal Protection Equipment (PPE)) is not practical on ISS. NASA medical teams and toxicologists in partnership with Russian specialists at RKK Energia and the Institute of Medico-biological Problems (IMBP) have worked together to define the impacts of ammonia on crew health based on available research as shown in Table 1. mg/M3 Zone PPM Ammonia Limit or Effect Rounded Rounded E= Energia; I = IBMP; N = NASA 21 000 30 000 Skin Blistering within several minutes dermal only contact(E) 14 000 20 000 Discernable irritation of open skin dermal only contact (E) 7000 10 000 Faint irritation of open skin dermal only contact(E) 5000 Rapidly Fatal dose (Henderson, 1927 & Mulder,1967)(N) 2500 3520 Deadly concentration(I) 2500 30-min exposure causes death (Helmers, 1971 & Millea 1989)(N) 1700 2390 Life threatening level (I) 1720 Coughing (E) 1200 1700 Laryngospasm/airway swelling life threat (Helmers 1971 & Grant)(N) 700 1000 ERPG#3 maximum 1-hr exposure without death(N); Hypostasis of lungs possible(I) 700 Immediate eye injury (Helmers 1971 & Grant 1974)(N) 350 500 Work without mask impossible (E);Life threatening level and Coughing (I); Heavy irritation of eyes, nose and throat (N) 280 390 Throat Irritation (I,N) 360 Work without mask can barely be withstood (E) 300 NIOSH Immediate Danger to Life and Health Red 200 ERPG#2 maximum 1-hr exposure without serious health effect & work without mask is difficult (E) 110 Strong eye irritation within one hour (NASA, SMAC vol 1 p 42)(N) 100 NRC Emergency 1-hour exposure limit 90 Begin nose and mouth irritation (E) 40 60 Begin eye and upper airway irritation; reflex inspiratory hesitation is 3 American Institute of Aeronautics and Astronautics possible (E,I,N) 50 OSHA max permissible exposure limit (8/hr/d time-weighted avg)(N) 35 NIOSH max 30 min exposure limit (N) Yellow 30 NASA one-hour SMAC(N) 20 28 IBMP one-hour MPC(I) 25 ERPG#1 maximum 1-hr exposure to cause only transient mild symptoms;NIOSH max recommended exposure limit (8/hr/d time- weighted avg)(N) 14 20 NASA 24-hr SMAC(N) 5 7 IBMP 24-hr and 7-day MPC(I) 2 3 NASA 7-day SMAC(N) Green 0,7 Odor threshold (E, I ,N) Table 1. Ammonia Effects and Limits in Emergent Acute Exposure IV. Current ISS Vehicle Response The current USOS system control software detects an IFHX leak condition and automatically activates a caution and warning (C&W) toxic alarm on ISS, shuts down intermodule ventilation (IMV) between USOS and IP modules and initiates isolation of the affected IFHX. As previously noted, the IFHX cannot be completely isolated due to the presence of the bleed line. It is possible that a crewmember will smell an ammonia leak before it is detected by the system control software. In that case, a crewmember would manually activate the C&W toxic alarm. The current ammonia response strategy calls for the crew to immediately don oxygen masks, ensure that the C&W toxic alarm has been activated, move as quickly as possible towards the ROS and close the hatch between segments to isolate themselves from the ammonia leakage source. The primary goals of the ISS ammonia leakage response are to evacuate the crew to the ROS as quickly as possible, limit ammonia dispersion into the ROS, and limit ammonia leakage into the ISS atmosphere. V. Characterization of the EATCS IFHX Rupture In order to properly mitigate an ammonia release on board the ISS, it is essential to first characterize the IFHX breach, the flow through the ITCS, and the entry into the cabin. Characterization of the ammonia leak scenario was achieved using three different modeling techniques which include mathematical modeling of the ammonia flow through the thermal control system and out the gas trap completed in 2007 with a follow-on two-dimensional analysis on dispersion through the ISS as well as removal using existing ROS systems in 2012. Computational fluid dynamics (CFD) analysis was then run on the ammonia dispersion through the cabin towards the ROS which was also completed in 2012. For consistency, both the two-dimensional dispersion analysis and the CFD model assumed a 0.05 inch hole is formed in the IFHX, and the hole does not propagate. The 2012 analysis also uses the calculations completed in 2007 for the flow rate of ammonia out of the gas trap. The analysis does not address detailed failure mechanisms, or the likelihood of system failures. A. Modeling Ammonia Release into ISS Atmosphere In 2007 a mathematical model was developed to predict how much ammonia would be present in the ISS ROS of the ISS by the time the crew is able to isolate the USOS, which at the time was assumed to be approximately 5 minutes, following a EATCS IFHX rupture. The model was also used to determine if an atmosphere could be generated that could present structural (overpressurization) or flammability concerns within the USOS. The scenario that was modeled assumed a mechanical/structural failure of an IFHX causing liquid anhydrous ammonia at high pressure to flow into the ITCS LTL, fully stroking the loops accumulator. The flow of ammonia would then pressurize the ITCS LTL until the weakest part of the loop, the gas trap, fails, after which the water in the loop, as well as the ammonia from the ETCS, would flow into the cabin in an uncontrolled manner as shown in Fig 2. 4 American Institute of Aeronautics and Astronautics Figure 2. IFHX Rupture Model Several assumptions were made to generate the model, including physical, chemical, and operational assumptions. The physical assumptions of the 2007 model on Ammonia Release into the ISS Cabin were as follows: 1) Nominal cabin pressure of 99.2 kPa (14.4 psia) and temperature of 23.9°C (75 °F) 2) The quantity of ammonia was assumed to be 272.2 kg (600 lbs) at 2688.9 kPa (390 psia), distributed amongst the bellows accumulator and the plumbing between the bellow accumulator and the heat exchanger 136.1 kg (300 lbs), the pump accumulator and the plumbing between the pump accumulator and the heat exchanger 13.6 kg (30 lbs), and heat exchanger and its relevant plumbing 122.5kg (270 lbs) 3) The gas trap of the ITCS would rupture at 1461.7 kPa (212 psig) 4) The size of the rupture in the heat exchanger was expressed as an equivalent hole size, and different diameters of the hole were assumed between 1.27 mm (0.05 inches) and 12.7 mm (0.5 inches) 5) The LTL accumulator’s capacity was 10651.6 cm3 (650 in3), but was initially 75% full 6) The LTL operates in dual loop mode, initially at 124.1 kPa (18 psia), with 62.08 liters (16.4 gallons) of water 7) The internal free volume of the ISS is 636 m3 (2007 projection of assembly complete volume) These values and assumptions may not hold today at the time this paper is prepared, but were what was available late 2006 / early 2007 when this model was developed. The assumptions related to chemistry were: 1) Ammonia and water do not interact chemically 2) The entire system is isothermal 3) Once the liquid ammonia/water mixture enters the cabin environment, evaporation of the water and ammonia is instantaneous, but only the evaporation of ammonia contributes to pressure build-up in the cabin 4) Ammonia vapor instantly and evenly distributes throughout the ISS cabin. 5) The operational assumptions of the 2007 model were that there were no operational controls in place: 6) Nitrogen backing pressure from the bellows accumulator was not removed 7) Isolation valves were not closed 8) Caution and warning alarms were not enunciated 9) No automatic software controls were utilized 10) Once ammonia entered the cabin environment, the status of IMV and hatches were unchanged prior to the rupture. The physical system was modeled as four dependently-coupled subsystems, and it is the flow from one system to the next that was modeled at the subsystems’ interfaces. These four subsystems were: 5 American Institute of Aeronautics and Astronautics 1) The bellows accumulator, the pump accumulator, and the plumbing between these components up to the EATCS 2) The EATCS, including the heat exchanger and relevant plumbing 3) The ITCS, including the LTL and its accumulator and the gas trap 4) The cabin volume, including the Pressure Control Assembly (PCA) vent valve (in case the ISS over- pressurizes due to rapid expansion of liquid ammonia to vapor ammonia). The first interface was between the bellows accumulator and the EATCS, and the flow of liquid ammonia was modeled as liquid flow with friction through the plumbing. The second interface was between the EATCS and the LTL (the IFHX), and the flow of liquid ammonia was modeled as choked liquid flow through an orifice of 1.27 mm (0.05 inch) diameter, representing the rupture of the heat exchanger. The third interface was between the gas trap between the LTL and the cabin environment, and the flow of the ammonia-water mixture was modeled as choked liquid flow through an orifice. The size of the orifice was iteratively calculated such that the pressure in the LTL did not exceed 1461.7 kPa gauge (212 psig), i.e., the hole in the gas trap was sufficiently large enough to maintain 212 psig in the LTL. As liquid ammonia and water exit the gas trap into the cabin environment, the ammonia would undergo a phase-changing throttling process that could cause local pressure effects, but these effects were beyond the scope of this modeling effort. The fourth interface was PCA vent between the cabin environment and space vacuum, and was the flow of ammonia and water was modeled as gaseous sonic flow through an orifice. However, because the focus of this paper is the flow rate of ammonia exiting the gas trap into the cabin environment, cabin pressure effects will not be discussed further. The flow rate of the water/ammonia mixture flow out of the rupture gas trap was calculated to be 104.3 kg/hr (230 lb/hr) for an equivalent heat exchanger hole size of 1.27 mm (0.05 inch) diameter with the LTL at 1461.7 kPa gauge (212 psig) and the backing pressure inside the heat exchanger was 2688.9 kPa (390 psia). Fig. 3 depicts the timeline of the ammonia release. As the pressure decreases with time, the flow rate also decreases, but that analysis was beyond the scope of the 2007 model. Initially, the flow is primarily water, but as more ammonia flows into the LTL from the heat exchanger, the concentration of ammonia in the mixture increases. For the case of a 0.05 inch hole in the IFHX, the ITCS gas trap breach occurs 10.75 minutes after the IFHX leak initiates. Figure 3. Timeline of 0.05” IFHX Rupture 6 American Institute of Aeronautics and Astronautics B. Two-Dimensional Ammonia Dispersion Assessment Assessing ammonia dispersion in the ISS cabin at a two-dimensional level assumes that the cabin behaves as two interconnected well-mixed volumes. It is expected that time predictions using this technique are conservative because they neglect the mixing time within each successive module that a more detailed three-dimensional analysis addresses. The two-dimensional analysis uses an ammonia generation rate derived from the analysis of relative timing from interface heat exchanger failure through gas trap failure (as described in the 2007 mathematical model) and the cabin concentration reaching 10000 ppm (1%). A simultaneous mass balance on each individual segment is conducted for the two-dimensional analysis. The mass balance equations for the USOS and ROS are provided by Eq. (1) and Eq. (2). These equations define the change in contaminant mass as a function of time. dM v v v U  R M  U M  M g dt V R V U V U U R U U (1) dM v v v R  U M  R M  M g dt V U V R V R R U R R (2) In Eq. (1) and Eq. (2), M is the total mass of contaminant in the USOS, M is the total mass of the contaminant in U R v the ROS, V is the USOS free volume, V is the ROS free volume, U is the intermodule ventilation flow from the U R v USOS to ROS, R is the intermodule ventilation flow from the ROS to USOS, Σηv is the removal capacity in the respective segment, g is the generation rate in the USOS, and g is the generation rate in the ROS. U R Simultaneous solution of Eq. (1) and Eq. (2) provide an equation for each segment in the form of Eq. (3). In Eq. (3), M is the total mass of contaminant in the reference cabin volume; α, β, and γ are constants calculated from the segment cabin free volume, ventilation flow, removal capacity, and contaminant generation rate; and x and x are 2 3 integration constants. The integration constants are calculated from the segment free volume, ventilation flow, and removal capacity parameters. Concentration is calculated by simply dividing the contaminant mass by the segment free volume. M ex2t ex3t (3) If the entire cabin volume is assumed to be well mixed, then the total cabin mass balance equation can be defined more simply as Eq. (4). Derivation of Eq. (4) can be found in Ref. 4. In Eq. (4), M is the contaminant mass at time, t; Mo is the contaminant mass at time equal to zero; V is cabin volume, Σηv is the contaminant removal capacity, g is the contaminant generation rate, and t is time. However, because ammonia exhibits a gradient, Eq. (4) cannot be readily used. As such, the USOS and ROS are assumed to be separate well-mixed volumes. A composite concentration is then obtained by summing the contaminant mass in each segment and then dividing by the total station free volume. vVt gV  vV t M  M e     1e   o  v       (4) C. Ammonia Removal Routes The equipment on board the ISS that possess ammonia removal capacity include the USOS trace contaminant control (TCC) equipment (268 grams), ROS harmful contaminants filter (HCF or Russian acronym FVP) equipment (2 grams), and the humidity control equipment. Under the circumstances of bulk ammonia leakage the TCC and HCF capacity must be assumed to be fully saturated. This leaves ammonia absorption in humidity condensate as the primary removal route for ammonia under bulk ammonia leakage circumstances. D. Ammonia Removal via Humidity Control Equipment A mass balance on a typical condensing heat exchanger, assuming co-current condensate and process air flow, provides a general equation relating bulk liquid (condensate) and gas (atmospheric) phase mole fraction.5 Eq. (5) 7 American Institute of Aeronautics and Astronautics shows the basic form solved for liquid phase mole fraction. In Eq. (5), x is the volatile compound liquid phase mole fraction, y is the volatile compound gas phase mole fraction, C is the condensate flow rate in the heat exchanger core in moles/hour, A is the process air flow rate through the heat exchanger core in moles/hour, H is the Henry’s Law constant in atm/mole fraction, and P is total pressure in atmospheres. An adjustment, , to the Henry’s Law constant is necessary to account for the 4.4 C (39.92 F) heat exchanger operating temperature and liquid phase interactions.6 y x  C H  A P (5) The correlation of Eq. (5) is used to approximate published vapor-liquid equilibrium data at 4.4 °C (39.92 F).7 These data are plotted in Fig. 4 and presented in tabular form by Table 2. 6000000 y = 416.32x2.0891 R² = 0.9742 5000000 mv p p n ‐ o4000000 uti ol S r  e v o H33000000 N of  e  r u s es2000000 r P al  ti ar P 1000000 0 0 10 20 30 40 50 60 70 80 90 100 Percent NH molal concentration in Solution 3 Figure 4. Aqueous Ammonia Vapor-Liquid Equilibrium at 4.4 °C (39.92 F) 8 American Institute of Aeronautics and Astronautics NH3 Partial Pressure Mole % NH3 psia ppmv 5 0.33 24118.8 10 0.66 48237.6 15 1.14 83319.49 20 1.92 140327.6 25 3.16 230955.8 30 5.13 374937.7 35 7.98 583236.4 40 11.98 875585.5 45 17.14 1252716 50 23.33 1705126 55 30.15 2203581 60 37.15 2715192 65 43.69 3193183 70 49.56 3622205 75 54.4 3975948 80 58.31 4261719 85 61.62 4503638 90 64.77 4733863 95 68.31 4992592 Table 2. Tabular Vapor-Liquid Equilibrium Data at 4.4 °C (39.92 F) The co-current absorption operating curve using Henry’s Law as the equilibrium driving force is used with the Henry’s Law constant equal to 1.60 × 10-5 atm-m3/mole at 25 °C.8,9 This value is adjusted for temperature to the 4.4 °C (39.92 F) condition by using the vapor pressure ratio temperature dependence estimating technique yielding a value of 8.03 × 10-6 atm-m3/mole (0.444 atm/mole fraction).10 The calculation technique was shown to reasonably predict ammonia loading into cold water up to 15 mole percent and ammonia gas phase concentration up to 83000 ppm using this temperature-adjusted Henry’s Law constant. Once good agreement with the literature vapor-liquid equilibrium data was demonstrated, the Henry’s Law constant was adjusted further to account for reported reactivity with dissolved carbon dioxide thus yielding a final Henry’s Law constant value of 0.0231 atm/mole fraction (4.17 x 10-7 atm-m3/mole).11 This value is used for calculation purposes. The resulting single pass removal efficiency and condensate loading for the ROS SKV and USOS common cabin air assembly (CCAA) heat exchangers were determined using the co-current adsorption mass balance equation and adjusted Henry’s Law constant. The performance of humidity control devices to remove ammonia is summarized in Figs. 5, 6, and 7. 400 350 e) at ns de300 n o g c k H/3250 N g  g ( din200 a o e L nsat150 e d n o C e 100 g a er v A 50 0 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 Cabin Concentration (ppm) Figure 5. Predicted condensate loading in the ROS SKV as a function of cabin concentration. 9 American Institute of Aeronautics and Astronautics SKV CCAA @ 50% Bypass 0.14 0.12 0.1 nt) e c er al p0.08 m ci e d y (0.06 c n e ci Effi0.04 0.02 0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Condensate Collection Rate (liters/hour) Figure 6. Predicted single pass ammonia removal efficiency provided by the ROS SKV and USOS CCAA. Assessment shows the single pass removal efficiency can be assumed to be constant across the range of cabin concentration between 30 ppm and 10,000 ppm. The concentration reduction calculation takes this into account. SKV CCAA @ 50% Bypass 390 Cabinconcentration = 10000 ppm Vapor‐liquid equilibrium adjusted for temperature and dissolved carbon dioxide effects 380 g) k / g g (370 n di a o a L oni360 m m A e  at ns350 e d n o C 340 330 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Condensate Collection Rate (liters/hour) Figure 7. Condensate loading with ammonia for varying condensate collection rates with ammonia at 10,000 ppm in the cabin. The ROS SKV is predicted to result in condensate loading averaging 361.4 g/kg at a cabin concentration of 10,000 ppm. Note that at 30 ppm, the condensate loading with ammonia is reduced substantially. For the SKV the loading averages 1.08 g/kg condensate over the condensate removal rate range. E. Two-Dimensional Ammonia Dispersion Prediction 10 American Institute of Aeronautics and Astronautics