2. Dispersion We’re going to move on to the second major step in doing dose assessment.

2009 National Radiological Emergency Planning Conference
Meteorological Fundamentals We’ll start this discussion by looking at some of the meteorological fundamentals that are important to dispersion analyses. 2009 National Radiological Emergency Planning Conference

Diffusion, Transport, Dispersion
Diffusion answers “How much of the pollutant reaches a location?” Transport answers “Where does the pollutant travel to following release?” “What was the footprint on the ground?” As the terrain gets more complex, transport takes on more significance. A large source of uncertainty. Dispersion is the result of diffusion and transport. Lets get some terminology down so that we can talk the same language. Diffusion answers “How much of the pollutant reaches a location?” Transport answers “Where does the pollutant travel to following release?” “What was the footprint on the ground?” As the terrain gets more complex, transport takes on more significance. A potential large source of uncertainty. Dispersion is the result of diffusion and transport. 2009 National Radiological Emergency Planning Conference

Snapshot, Spatial, and Temporal
Snapshot refers to a projection based on a initial set of conditions, persisted without change. e.g., “straight line” Gaussian. Spatial refers to the ability to represent impacts on the plume after release from site. e.g., plume bending to follow river valley, sea breeze circulation. Temporal refers to the ability of model to reflect input data changes over time. e.g., change in release rate; meteorology. Snapshot refers to a projection based on a initial set of conditions, persisted without change. e.g., “straight line” Gaussian. Spatial refers to the ability to represent impacts on the plume after release from site. e.g., plume bending to follow river valley, sea breeze circulation. Temporal refers to the ability of model to reflect input data changes over time. e.g., change in release rate; meteorology. 2009 National Radiological Emergency Planning Conference

Earth’s Heat Balance (typical)
With the definitions out of the way, lets have a look at the earth’s heat balance. For purposes of discussion, lets set the intake of short wave solar radiation at an arbitrary 100 units. Starting on the left: 12 units were are scattered by the atmosphere– 6 units to the Earth, 6 units back out of the atmosphere. 4 units are reflected off the Earth 17 units are absorbed by the atmosphere 20 units pass through the atmosphere and are absorbed by the Earth 47 units reflect off clouds in the atmosphere—24 units to the Earth and 20 back out of the atmosphere, with 3 units absorbed by the cloud. 50 units ultimately are absorbed by the Earth. These fractions can vary greatly from location to location. For example, increase the cloud cover and reflection is greater. If the surface is covered with snow, reflection is greater. Of course, Solar radiation is available only during daylight hours 2009 National Radiological Emergency Planning Conference

Earth’s Heat Balance (typical)
Now, lets look what happens to the 50 units absorbed by the Earth. The Earth dissipates its absorbed solar radiation in the form of infrared radiation. 20 units radiated as long waves from the Earth go to heating the atmosphere. 10 units are transferred by convection. 20 units are dissipated in the form of heat of vaporization used in evaporation.—some of this latent heat is released when the vapor condenses to form clouds. Again, local conditions can change these fractions The energy transferred as heat of vaporization will differ from arid areas to bodies of water Remember that the fractions shown on this figure and the last are illustrative—the exact values will vary due to local conditions. 2009 National Radiological Emergency Planning Conference

Energy Distribution Since the heat balance varies as a function of: Time of day Cloud cover Surface cover Vegetation / Desert / Bodies of water Urban heat islands --there is an unequal distribution of energy across the earth, causing differences in temperature and pressure. Weather as we know it is largely the result of these differences in energy distribution. Lets summarize to this point: Since the heat balance varies as a function of: Time of day Cloud cover Surface cover Vegetation / Desert / Bodies of water Urban heat islands there is an unequal distribution of energy across the earth, causing differences in temperature and pressure. Weather (and dispersion) as we know it is largely the result of these differences in energy distribution. 2009 National Radiological Emergency Planning Conference

Horizontal Winds Horizontal winds are created by differences in local pressures, both local and global. Wind flows from higher pressure air masses to lower pressure air masses. Since the distribution of air masses varies, so do winds. Lower surface winds typical at night. Because of surface friction, wind speeds generally increase with height. Gradient depends on surface texture and turbulence in atmosphere. Lets move on to horizontal winds. Horizontal winds are created by differences in local pressures, both local and global. Wind flows from higher pressure air masses to lower pressure air masses . Since the distribution of air masses varies, so do winds. Lower surface winds typical at night. Because of surface friction, wind speeds generally increase with height. Gradient depends on surface texture and turbulence in atmosphere. 2009 National Radiological Emergency Planning Conference

Wind Speed Gradients This figure illustrates what we just talked about. The figure shows three surface conditions: An urban area A suburb, and, A rural area. This graph shows two axes—the gradient is also dependent on the amount of turbulence in the atmosphere.—during stable conditions, the increase is steeper than it is for unstable conditions . As you can see, the wind speed increases with height, with the increase being steeper for areas with high surface friction. Some dose assessment codes will use curves such as these to determine a wind speed when the elevation at which the wind speed is measured is not the same as the plume centerline elevation. 2009 National Radiological Emergency Planning Conference

Atmospheric Stability
The temperature of the lowest layer of the atmosphere (troposphere) generally increases with decreasing height This gradient is known as the atmospheric lapse rate: On average, the atmospheric lapse rate is about +6.5ºC per kilometer of elevation Typically referred to as delta-T in nuclear power community By convention, lapse rate is always the higher elevation temperature less the lower elevation temperature We’re now going delve into Atmospheric Stability. We’re going to buildup to this topic slowly. The temperature of the lowest layer of the atmosphere (troposphere) generally increases with decreasing height. Remember the atmosphere is heated by the radiation emitted by the Earth—so its warmer near the surface This gradient is known as the atmospheric lapse rate: On average, the atmospheric lapse rate is about +6.5ºC per kilometer of elevation. Typically referred to as delta-T in nuclear power community. By convention, lapse rate is always the higher elevation temperature less the lower elevation temperature. This has always seemed backwards to me—its just something you get used to. 2009 National Radiological Emergency Planning Conference

Overall lapse rate through the troposphere is positive ( i.e., “positive” with decreasing height) In the lower kilometer, lapse rate varies with the daily cycle of heating and cooling This daily variation has significant impact on diffusion Thermal gradients are created, resulting in thermal currents and turbulence Atmospheric stability is an index of the amount of atmospheric turbulence present The greater the turbulence, the better the diffusion Overall lapse rate through the troposphere is positive. (“positive” with decreasing height) In the lower kilometer, lapse rate varies with the daily cycle of heating and cooling. This daily variation has significant impact on diffusion. Thermal gradients are created, resulting in thermal currents and turbulence. Atmospheric stability is an index of the amount of atmospheric turbulence present. The greater the turbulence, the better the diffusion. 2009 National Radiological Emergency Planning Conference

Unstable Atmosphere Positive lapse rate (increases with decreasing height) Warmer buoyant air near surface rises Strong convective air currents favor diffusion Neutral Atmosphere No change of temperature with height Stable Atmosphere Negative lapse rate (decreases with decreasing height) Little or no vertical currents disfavor diffusion Mostly at night; also known as inversion Unstable Atmosphere Positive lapse rate Warmer buoyant air near surface rises Strong convective air currents favor diffusion We’ve probably all seen this as heat rising from parking lots on a hot summer day Neutral Atmosphere No change of temperature with height Stable Atmosphere Negative lapse rate (decreases with decreasing height) Little or no vertical currents disfavor diffusion Mostly at night; also known as inversion 2009 National Radiological Emergency Planning Conference

Impact of Stability on Diffusion
Consider parcel of air injected into the atmosphere by a motive or buoyant force: Parcel does not mix with surrounding air Parcel will expand or contract as the pressure of air surrounding the parcel changes If parcel is less dense than the surrounding air, the parcel will rise As the parcel expands or contracts, its temperature and density will change We’re going to discuss stability on dispersion by considering a parcel of air injected into the atmosphere by a motive or buoyant force: No mass transfer between parcel and surrounding air. Parcel will expand or contract as the pressure of air surrounding the parcel changes. If parcel is less dense than the surrounding air sufficient to overcome gravitation pull, the parcel will rise. As the parcel expands or contracts, its temperature and density will change following the ideal gas laws. 2009 National Radiological Emergency Planning Conference

The parcel temperature decreases at a constant rate of 10ºC per kilometer Known as the dry adiabatic lapse rate The atmosphere temperature is also changing as a function of the measured atmospheric lapse rate As such, the parcel movement is a function of the measured atmospheric lapse rate The more positive the measured atmospheric lapse rate is, the greater the parcel rise, the better the diffusion For reasons well beyond the scope of this presentation, the parcel temperature will decrease at a constant rate of 10ºC per kilometer. Known as the dry adiabatic lapse rate. The figure here shows the dry adiabatic lapse rate. Remember, however, that the atmosphere temperature is also changing as a function of the measured atmospheric lapse rate. Since the dry adiabatic lapse rate is a constant, the parcel vertical movement is a function of the measured atmospheric lapse rate. The more positive the measured atmospheric lapse rate is, the greater the parcel rise, the better the diffusion. 2009 National Radiological Emergency Planning Conference

Shown here is a situation involving an unstable atmosphere Atmospheric lapse rate is positive, temperature and pressure increases with decreasing height Parcel of air rises as long as its density is less than that of the surrounding atmosphere (How’s this for production value??) This illustration attempts to depict what going on. Note the temperatures on the left hand scale The temperatures increase with decreasing height, so this is an unstable atmosphere The parcel is injected with a Temperature of 15 degrees, which is greater than the temperature of the surrounding air. The parcel being less dense, rises. Since the surrounding air pressure is less, the parcel expands, and in doing so cools somewhat. The parcel temperature decreases and a slower rate than the surrounding air, so it keeps rising Once the temperature of the parcel balances with the surrounding air, the parcel stops rising. Note that the thermal rise completes with horizontal wind, so an actual plume will curve. 2009 National Radiological Emergency Planning Conference

Within the troposphere, it is possible to have multiple layers, each with its own atmospheric lapse rate Figure shows a neutral stability layer topped by a stable layer (e.g., Sunrise as earth heating burns away inversion layer) The stable layer aloft “caps” the vertical rise The release mixes in the neutral layer The height of the lower layer is known as the mixing depth Within the troposphere, it is possible to have multiple layers, each with its own atmospheric lapse rate. Figure shows a neutral stability layer topped by a stable layer. (e.g., Sunrise as earth heating burns away inversion layer.) The stable layer aloft “caps” the vertical rise. The release mixes in the neutral layer. The height of the lower layer is known as the mixing depth. Although turbulence favors dispersion, reducing concentration, the mixing brings the plume to the ground sooner than it might have otherwise—there could be a net increase in ground level concencentration. There are other combinations illustrated in the text. 2009 National Radiological Emergency Planning Conference

Diffusion Models We’re going to delve into diffusion models 2009 National Radiological Emergency Planning Conference

Basic Concept Release of a material to environment Certain volume Certain radionuclide concentration Release mixes with air due to turbulence Release increases in volume Release decreases in concentration Resulting concentration is not uniform Most mixing occurs at the surface of the release volume Concentration greatest at center of release volume, decreases asymptotically BASIC CONCEPT Release of a material to environment. Certain volume Certain radionuclide concentration Release mixes with air due to turbulence. Release increases in volume / is diluted Release decreases in concentration. Resulting concentration is not uniform. Most mixing occurs at the surface of the release volume Concentration greatest at center of release volume, decreases asymptotically from the center 2009 National Radiological Emergency Planning Conference

Atmospheric Dispersion Models
Gaussian plume model most widely used for estimating dispersion in dose projections Generally non-spatial and non-temporal Stylized, straight-line “snapshot” Simple, can be implemented in hand calculations May not be representative for a given site Advanced models are available Segmented-plume Gaussian models Modified potential flow numerical models Particle tracking models Gaussian plume model most widely used for estimating dispersion in dose projections. Generally non-spatial and non-temporal. Stylized, straight-line “snapshot.” Simple, can be implemented in hand calculations. May not be representative for a given site. The NRC categorizes these models as “Class A” models Advanced models are available. Segmented-plume Gaussian models Modified potential flow numerical models Particle tracking models Some of these models can be categorized as “Class B” Models 2009 National Radiological Emergency Planning Conference

Gaussian Model Plume concentration is assumed to diffuse in the vertical by a normal distribution, represent-ed by the standard deviation, y This is the structure of a Gaussian Model. The coordinate scheme is The Y-axis is cross-wind The X-axis is along wind The Z-axis is elevation. Plume concentration is assumed to diffuse in the vertical by a normal distribution, represented by the standard deviation, y Plume concentration is assumed to diffuse in the horizontal (cross-wind) by a normal distribution, represented by the standard deviation, z There is also diffusion downwind (x) but this is small compared to the distribution by the wind, represented by the wind speed, U Note that convention places the axes origin at the base of the release point on grade level. There is also diffusion downwind (x) but this is small compared to the distribution by the wind, represented by the wind speed, U Plume concentration is assumed to diffuse in the horizontal (cross-wind) by a normal distribution, represented by the standard deviation, z 0,0,0 2009 National Radiological Emergency Planning Conference

Effective Plume Height
For elevated plumes, the plume will raise above the stack height due to thermal buoyancy and other forces Increases in terrain height change the position of the receptor relevant to the plume and, if high enough, can obstruct the plume For elevated plumes, the plume will raise above the stack height due to thermal buoyancy and other forces. There are models available to calculate these rises. As horizontal wind speeds increase, plume rise decreases. As plume temperature increases, plume rise increases Increases in terrain height change the position of the receptor relevant to the plume and, if high enough, can obstruct the plume. Remember that the axes origin is at the base of the stack when assigning heights. NRC practice has been to consider an elevated plume to be a ground level plume once the plume centerline impacts terrain 2009 National Radiological Emergency Planning Conference

Gaussian Model Essential conditions: Non-zero wind speed Wind direction constant over time and downwind area Release rate constant over time for the duration of the release Atmospheric stability constant over time and downwind area Because of these conditions: Gaussian assessment is a straight-line, “snapshot” Gaussian model is not temporal nor spatial Essential conditions: Non-zero wind speed. Wind direction constant over time and downwind area. Release rate constant over time for the duration of the release. Atmospheric stability constant over time and downwind area. Because of these conditions: Gaussian assessment is a straight-line, “snapshot.” Basic Gaussian model is not temporal Note all the “constant with time” conditions Basic Gaussian model is not spatial. Conditions (wind direction) at the release point are persisted 2009 National Radiological Emergency Planning Conference

Gaussian Model Xxyz = Downwind concentration at coordinates X,Y,Z (e.g., Ci/M3) Q’ = Release rate (e.g., CI/sec) x = Receptor downwind distance (along wind) y = Receptor horizontal (crosswind) offset from plume centerline z = Receptor height (ground level =0) he = Plume centerline effective height above terrain u = Wind speed y,z = Dispersion Coefficient on Y-axis and Z-axis Here is the basic Gaussian model expression Xxyz = Downwind concentration at coordinates X,Y,Z (e.g., Ci/M3) Q’ = Release rate (e.g., CI/sec) x = Receptor Downwind distance (along wind) y = Receptor Horizontal (crosswind) offset from plume centerline z = Receptor Height he = Plume centerline effective height above terrain u = Wind speed y,z = Dispersion Coefficient on Y-axis and Z-axis 2009 National Radiological Emergency Planning Conference

Dispersion Coefficients
Input parameters can be measured or projected y and z vary as functions of the downwind distance and atmospheric stability class Values were determined by correlations to experimental field measurements. Correlations by Pasquill and Gifford typically used There are other datasets Input parameters can be measured or projected. y and z vary as functions of the downwind distance and atmospheric stability class. Values were determined by correlations to experimental field measurements. Correlations by Pasquill and Gifford typically used. There are other datasets. 2009 National Radiological Emergency Planning Conference

P-G Dispersion Coefficients
These two graphs plot the diffusion coefficients, sigma-Y on the left and sigma-Z on the right, as a function of downwind distance and stability class The sigma-Y and sigma-Z values are expressed in units of meters. There are various polynomial expressions of these graphs in the literature. σ y z 2009 National Radiological Emergency Planning Conference

Gaussian Model If the receptor is at ground level, z = 0, and: The Gaussian model expression discussed above addresses most receptor locations and plume heights. The expression can be simplified as shown on this slide. If the receptor is located at ground zero, Z = 0 the equation reduces to the first equation on this slide. If the receptor is located at ground zero immediately below the plume centerline, Z and Y = 0 and the second equation applies If the effective height of the plume is 0, the case for a ground level release, the third equation applies. You will see other variations in the literature. For example, the NRC has variants to address plume meander and building wake. These are generally not used for emergency dose assessment If the receptor is at ground level under plume centerline, z = 0, y=0, and: If the receptor and plume is at ground level, he= 0, and: 2009 National Radiological Emergency Planning Conference

Normalized Air Concentration
The preceding equations are often restructured to calculate normalized airborne concentration, or  /Q Normalized concentration =  divided by Q where  = airborne radionuclide concentration, e.g., Ci/m3 Q = source term, Ci/s and  /Q = s/m3 In the expressions above, the downwind concentration (chi) was the result and the release rate was part of the expression. However, in practice, the equation is restructured to place the release rate on the left side of the equation, yielding a parameter known as the Normalized air concentration or Chi over Q. The X/Q can be calculated without regard to the magnitude of the release rate, and then applied to various release rate, saving calculation. 2009 National Radiological Emergency Planning Conference

Gaussian Model Enhancements
Most limiting aspect of basic Gaussian model is inability to evaluate spatial and temporal differences in model inputs Enhanced Gaussian models can address these inabilities Puff model Segmented plume models Enhanced Gaussian models generally address both diffusion and transport Most limiting aspect of basic Gaussian model is inability to evaluate spatial and temporal differences in model inputs. Back in the days when these calculations were done on slide rules, these calculations generally had to be performed by hand, and the limitations had to be tolerated. As computing capabilities improved over time, researchers addressed these limitations, developing what I will refer to as ENHANCED Gaussian models Enhanced Gaussian models can address these inabilities. Puff model Segmented plume models Enhanced Gaussian models generally address both diffusion and transport 2009 National Radiological Emergency Planning Conference

Enhanced Gaussian Model Structure
Processing algorithm that: Divides the calculation domain into equal time steps Assign meteorological and release data to each time step Processes each time step individually, integrating the calculation results Use of time steps allow model to reflect temporal changes A rectangular two-dimensional (x,y) wind field Each cell is assigned a wind vector for each time step The wind vector assigns the initial conditions for that time step Processing algorithm that: Divides the calculation domain into equal time steps. Assign meteorological and release data to each time step. Processes each time step individually, integrating the calculation results. Use of time steps allows model to reflect temporal changes. Often, a rectangular two-dimensional (x,y) wind field. Each cell is assigned a wind vector for each time step. The wind vector assigns the initial conditions for that time step. 2009 National Radiological Emergency Planning Conference

Gaussian Model Enhancements
Segmented Plume Plume is approximated by a series of straight-line Gaussian plumes, each estimated on the parameter values applicable to that time step Puff Model Plume is approximated by a series of puffs, the diffusion and transport of each is estimated on the parameter values applicable to that time step Two models in frequent use are: Segmented Plume Puff Plume is approximated by a series of straight-line Gaussian plumes, each estimated on the parameter values applicable to that time step. Puff Model Plume is approximated by a series of puffs, the diffusion and transport of each is estimated on the parameter values applicable to that time step. 2009 National Radiological Emergency Planning Conference

Segmented Plume Model This is an example of the isopleth output from a typical commercially available segmented plume model Please note that my use of this plot does not constitute a NRC recommendation for the particular model nor does the exclusion of other models constitute a recommendation against them. We can see from the plot that the wind direction just about doubled back. We can also readily see this plot wasn’t generated by a straighline Gaussian model. This plot illustrates TEDE dose and includes inhalation. The downwind hot spot was due to the terrain rising as a hill. People on this hill inhale a higher concentration than do those at lower elevations. Screen capture from MIDAStm software by PLG, Inc. 2009 National Radiological Emergency Planning Conference

Advanced Diffusion Models We are now going to jump ahead to more advanced models, which because of the amazing advances in computer processing, are moving out of research facilities and coming into use. 2009 National Radiological Emergency Planning Conference

Advanced Models In many Gaussian models, terrain height is addressed only in determining the effective plume height The impact of terrain on plume transport is not addressed Straight-line models can not “curve” a plume around mountains or follow a river valley Advanced models can address terrain impact on plume transport In many Gaussian models, terrain height is addressed only in determining the effective plume height. The impact of terrain on plume transport is not addressed. Straight-line models can not “curve” a plume around mountains or follow a river valley. Advanced models can address terrain impact on plume transport. 2009 National Radiological Emergency Planning Conference

Advanced Models In a particle-in-cell (PIC) model, the wind field is three-dimensional Wind vectors have x,y,z components Calculated for each time step using modified potential flow algorithms Requires wind speed and direction data for multiple elevations Terrain displaces affected wind field cells; wind vectors for these cells = 0 We’re going to skim over one particular type of advanced model, the particle-in-cell model. There are other formulations. PIC is one that I am familiar with. In a particle-in-cell (PIC) model, the wind field is three-dimensional. Wind vectors have x,y,z components. Calculated for each time step using modified potential flow algorithms. Requires wind speed and direction data for multiple elevations. Can use wind speed and direction data from multiple meteorological towers Terrain displaces affected wind field cells; wind vectors for these cells = 0. 2009 National Radiological Emergency Planning Conference

Dispersion in Advanced Models
In lieu of Gaussian formulation, PIC models use large numbers of virtual particles, each of which are tagged with: Time step of injection into wind field Radionuclide characterization Three-axis diffusion coefficient A group of particles is created for each time step, and injected into the wind field in sequence In lieu of Gaussian formulation, PIC models use large numbers of virtual particles, each of which are tagged with: Time step of injection into wind field Radionuclide characterization Isotopic magnitude Particulate or gaseous Three-axis diffusion coefficient A group of particles is created for each time step, and injected into the wind field in sequence. 2009 National Radiological Emergency Planning Conference

Dispersion in Advanced Models
The particles disperse through the wind field Movement depends on three-axis diffusion coefficients and three-axis wind field vectors On encountering a terrain face, particles follow vectors into adjacent cells, moving around or over the terrain Particle positions are periodically recorded and integrated over time When dispersion is complete, the integrated particles in each cell are converted to concentrations and then dose Inhalation dose assigned only for particles in cell layer adjacent to the ground Ground deposition can only occur from the cell layer adjacent to the ground The particles disperse through the wind field. Movement depends on three-axis diffusion coefficients and three-axis wind field vectors. On encountering a terrain face, particles follow vectors into adjacent cells, moving around or over the terrain. Particle positions are periodically recorded and integrated over time. When dispersion is complete, the integrated particles in each cell are converted to concentrations and then dose. These are generally finite plume dose models as the cells have known volume and location. Inhalation dose assigned only for particles in cell layer adjacent to the ground Ground deposition can only occur from the cell layer adjacent to the ground. 2009 National Radiological Emergency Planning Conference

Particle Tracking Model
During projection, the wind shifted around from down-river (left), CCW to upriver This is a representation of a particle distribution at a river valley site. In this particular case, the plume was initially heading downriver to the left. A wind direction shift moved it up river In this plot, the particle color changes with elevation. If you look at the lower left of the plot, the particles are darker—consistent with their climb over the river valley walls. This is a programmer’s diagnostic plot—commercial models may not display this plot. Screen capture from MIDAStm software by PLG, Inc. 2009 National Radiological Emergency Planning Conference

Particle Tracking Model
This is the dose isopleth that resulted from the previous particle distribution. The complexity of this plot is obvious when compared to a straight-line Gaussian model. A particle tracking model with a modified potential wind field is very compute intensive. At the utility, we ran this model on a Digital VAX, which had a 14 MHZ processor with a 64 bit data bus and 2 Mb ram. A one hour release could take nearly an hour to calculate. In today’s quad core 3.2 ghz processors performance would obviously be better. As noted earlier, the advantages of such a model may be realized only if adequate meteorological data can be provided The Interagency Monitoring and Assessment Advisory Center (IMAAC) at Los Alamos maintains a similar capability. Screen capture from MIDAStm software by PLG, Inc. 2009 National Radiological Emergency Planning Conference

Inputs to Dispersion Models
Wind speed Wind direction Ambient temperature Release height Rainfall Stability class Moving away from Diffusion models, lets recap the information needed to drive typical dispersion models. Wind speed is measured by anemometers. These instruments have a starting speed related to bearing friction and usually cannot measure wind speeds less than 0.5 mph. There is typically one sensor at 10 meter elevation and additional sensors at higher elevations. Wind direction is measured by wind vanes. By convention, wind direction is reported as the direction FROM WHICH the wind is blowing. There is typically one sensor at 10 meter elevation and additional sensors at higher elevations. Ambient temperature is measured by resistance temperature detectors. There is typically one RTD at 10 meters and at least one other elevation. The 10 meter RTD output is used a ambient temperature and is sometimes used as input to plume rise models. A delta-T reading is obtained by subtracting the 10 meter temperature from the higher elevation temperature. Release height is generally the physical height of the stack, but some models may calculate plume rise based on ambient temperature and stack flow velocity Rainfall can be used as an input in some models to determine wet deposition velocity Lastly, Stability class is based on the delta-T reading. 2009 National Radiological Emergency Planning Conference

Inputs to Dispersion Models
A straight-line Gaussian model can generally be driven by a single set of meteorological data Non-temporal; non spatial Wind field models can often be driven by multiple meteorological stations Temporal and spatial Improves modeling in complex terrain A wind field model driven by a single meteorological tower may not provide results any better than those from a straight-line Gaussian model Highly dependent on surrounding terrain and meteorological regimes Additional meteorological towers may be necessary to adequately model sea breeze sites A straight-line Gaussian model can generally be driven by a single set of meteorological data Non-temporal; non spatial Wind field models can often be driven by multiple meteorological stations Temporal and spatial Improves modeling in complex terrain A wind field model driven by a single meteorological tower may not provide results any better than those from a straight-line Gaussian model Highly dependent on surrounding terrain and meteorological regimes Additional meteorological towers may be necessary to adequately model sea breeze sites 2009 National Radiological Emergency Planning Conference

Elevated vs Ground Level Plumes
Although we have referred to elevated plumes earlier, I included this slide here to show the relationship between elevated and ground level plumes as regard to ground concentrations, and hence doses. In the elevated case, the plume is aloft and the ground concentration is low. As the plume diffuses vertically, the ground concentration increases, reaching a peak some distance away from the plant. A caution: radiation shine from the aloft plume can cause ground level doses higher than those calculated from the ground concentration alone. For the ground level release, the plume concentration is maximum at the release location and decreases with increasing distance. Ground Level Concentration Ground Level Concentration 2009 National Radiological Emergency Planning Conference

Terrain and Building Impacts on
Diffusion Models In this next section, we will discuss the impact that terrain, meteorological regimes and buildings can have on diffusion and transport. We’ll look at terrain effects first. 2009 National Radiological Emergency Planning Conference

Transport and Diffusion at Coastal Sites
Sea Breeze Also applies to any other large body of water Caused by differences in temperature of the air above water versus that above land after sunrise If the regional wind flow is light, a circulation will be established between the two air masses At night, the land cools faster, and a reverse circulation (weak) may occur Sea Breeze Also applies to any other large body of water. Caused by differences in temperature of the air above water versus that above land after sunrise. Fades as the day progresses The air will be cooler and more stable over water since the water doesn’t heat as fast as the land does If the regional wind flow is light, a circulation will be established between the two air masses. At night, the land cools faster, and a reverse circulation (weak) may occur. 2009 National Radiological Emergency Planning Conference

Impact of Sea Breeze The air over the water is cooler and is stable As the sea breeze forms, the stable air flows over the unstable air mass at the shore The boundary between the stable and unstable air is known as the thermal internal boundary layer (TIBL) The air over the water is cooler and is stable. As the sea breeze forms, the stable air flows over the unstable air mass at the shore. The boundary between the stable and unstable air is known as the thermal internal boundary layer (TIBL). Because the air below the TIBL is unstable, there is turbulence and mixing, drawing the plume to ground level As the day progresses, the TIBL layer moves inland In our figure, one stack is outside the TIBL, the other is inside. As the TIBL moves inland, both stacks will be outside the TIBL Without consideration of seabreeze, the diffusion and transport result could be wrong. Because the air below the TIBL is unstable, there is turbulence and mixing, drawing the plume to ground level As the day progresses, the TIBL layer moves inland 2009 National Radiological Emergency Planning Conference

Impact of Sea Breeze Along coastlines, the illustrated pattern often develops Although the cycling will certainly increase diffusion and result in lower ground level concentrations, the transport of the plume can be affected A dose projection based on the onshore wind direction, will underestimate the spread of the plume up the coast. Estimating dispersion at sea breeze sites can be challenging. In some cases, the inability to fully assess sea breeze will be compensated for by increasing the number of downwind sectors that protective actions will be implemented in. 2009 National Radiological Emergency Planning Conference

Transport and Diffusion at Valley Sites
The site I worked at was located in a valley terrain, so I am very familiar with dispersion in a valley. Most afternoons, you could watch the cooling tower plume swing around Considering the left hand illustration first. We learned earlier that the solar radiation heats the earth—not the air. The floor of a valley starts to heat, creating vertical thermal currents up the walls of the valley. This continues through the day. After sunset, the sun no longer heats the earth. As the earth cools, the air cools, and the denser cool air flows into the valley floor. This is called GRAVITY DRAINAGE At the center of the bottom of the valley, the cool air tends to flow downriver Thus, the location of the meteorological tower may decide what the wind direction is—if the tower is close to the valley walls and is impacted by gravity drainage, the wind will appear to flow cross-valley. 2009 National Radiological Emergency Planning Conference

Transport and Diffusion at Valley Sites
Here is a short section of actual meteorological data. Note the three very different wind directions associated with this valley. Now, ask yourself, which way is the plume moving? What are the downwind sectors affected by the plume? The drawing is a weak attempt to illustrate the data. The valley fills in with cool air through the night. As the sun starts to heat the valley walls and floor, thermal currents flow up the sides of the valley, burning way the cool core The 500 foot reading is above the valley walls and represents the regional flow 150 ft 500 ft 35 ft 2009 National Radiological Emergency Planning Conference

Building Wake In addition to dispersion impacts by terrain, the plant buildings can also impact dispersion. This illustration shows the top and side views of wind flows around and over a building. A ground level plume released between the buildings will undergo significant turbulence, and part of the release may be captured within the building wake cavity. Generally, this situation will not have a significant impact at locations away from the building, such as an offsite receptor. A sharp edged building would create more streamline distortion and turbulence than shown here 2009 National Radiological Emergency Planning Conference

Building Wake Another situation occurs when a release is via a stack near the plant. If the top of the stack is not greater in elevation than the displacement zone, the plume could become entrained within the wake and drawn to ground far sooner than projected by the Gaussian model. The NRC generally requires a stack to 2.5 times that the nearest solid building become treating releases as elevated.. If the stack (plus plume rise) isn’t high enough, the plume could be drawn into the building wake. Although mixing would occur, the plume would be at ground level sooner than projected 2009 National Radiological Emergency Planning Conference

2. Dispersion We’re going to move on to the second major step in doing dose assessment.

Similar presentations

Presentation on theme: "2. Dispersion We’re going to move on to the second major step in doing dose assessment."— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

2. Dispersion We’re going to move on to the second major step in doing dose assessment.

Similar presentations

Presentation on theme: "2. Dispersion We’re going to move on to the second major step in doing dose assessment."— Presentation transcript:

Similar presentations

About project

Feedback