White Paper for LOWWIND Options. R. Chris Owen OAQPS/AQMG 9/25/2017

Transcription

1 White Paper for LOWWIND Options R. Chris Owen OAQPS/AQMG 9/25/2017 1

2 Original LOWWIND Options LOWWIND1 Minimum σ v value of 0.5 m/s Eliminates the horizontal meander component of lateral dispersion Eliminates upwind dispersion. LOWWIND2 Minimum σ v value of 0.3 m/s An upper limit of 0.95 on FRAN, the horizontal meander component. Adjustments to the horizontal meander, i.e., 12 hrs is used for the BIGT parameter LOWWIND3 Minimum σ v value of 0.3 m/s Non-default FASTALL approach that matches the centerline concentration for the LOWWIND2 option, based on an effective σ y. Eliminates upwind dispersion 2

3 LOWWIND3 and Appendix W Rulemaking EPA proposed LOWWIND3 as a regulatory option Based on public comment and reassessment, EPA ultimately determined that there may be a possibility for model under predictions EPA decided to defer promulgating the LOWWIND3 option Continued engagement with the modeling community on additional research 3

4 Next Steps with LOWWIND Options The combination of several adjustments to the underlying model science in a packaged LOWWIND option makes it difficult to isolate the impact of each adjustment on overall model predictions Isolating these influences is essential to diagnosing the science and addressing performance issues EPA intends to remove all three existing low wind options and replace them with a new general (and more flexible) LOW_WIND alpha keyword 4

5 Options allowed under LOW_WIND Keyword Minimum σ v value. The default value in AERMOD is 0.2 m/s. LOWWIND1 used a value of 0.5 m/s, LOWWIND2 and LOWWIND3 used a value of 0.3 m/s. Plume meander/upper limit of FRAN. The default upper limit in AERMOD is 1.0, while LOWWIND2 set this value at Minimum wind speed. The default value in AERMOD is m/s, consistent with the default applied in previous versions based on SQRT(2*SVmin*SVmin) with SVmin=0.2. While this value was not adjusted in any of the LOWWIND packages, the minimum wind speed can be adjusted under the existing LOW_WIND keyword. 5

6 Options not to be included in the LOW_WIND Keyword Elimination of upwind dispersion Elimination of horizontal meander Adjustments to FRAN will mimic this Adjustment to the BIGT parameter FASTALL approach to match plume centerline concentrations FASTALL is still an option in AERMOD and can be used independently of the LOW_WIND keyword 6

7 Minimum sigma v (σ v ) σ σ = σ + σ σ 2 vt 2 vc 2 vm.6* = 3 u 2 vc.35* = 0 w 2 * 2 * 2 vm Total lateral turbulence function of the convective and mechanical turbulence Mechanical turbulence a function of the surface friction velocity (u * ) Convective turbulence a function of the convective velocity scale (w * ) 7

8 Hannah et al., 1985 Development and evaluation of OCD Field data collected from 4 ship-based campaigns off the coast of CA σ v = σ θ *U 8

9 Plume meander/upper limit of FRAN C = C 2 2 (1 σ / σ ) + C ( σ / σ 2 2 C, S Ch r h R r h ) Total concentration C C,S Concentration from the coherent plume C Ch Concentration from the random (aka, pancake) plume C R 2 2 FRAN = ( σ r / σ h ) Default FRANMAX = 1.0, adjustable with LOW_WIND 9

10 Accompanying LOWWIND White Papers Minimum value for lateral turbulence (aka, minimum σ v ) Issues Related to Plume Meander in the AERMOD System FRAN BIGT Upwind dispersion 10

11 White Paper An Overview of Recent Building Downwash Research at EPA/ORD David Heist 1, Steve Perry 1, Elaine Monbureau 2, Lydia Brouwer 2, Hosein Foroutan 3 and Wei Tang 4 1 U.S. Environmental Protection Agency (EPA), Office of Research and Development, National Exposure Research Laboratory, Research Triangle Park, NC, USA 2 Jacobs Technology Inc., Research Triangle Park, NC, USA 3 Virginia Tech, Blacksburg, VA 4 Applied Research Associates, Inc., Research Triangle Park, NC, USA Office of Research and Development National Exposure Research Laboratory / CED. Sept 25, 2017

12 Building downwash The limitations of AERMOD s building downwash algorithm were highlighted by introduction of revised 1-hr SO 2 and NO 2 standards OAQPS requested that we (ORD) research the issues and interact with AWMA on their PRIME2 committee Pressing issues include: Long, low buildings Oblique wind directions Features of AERMOD s building downwash algorithm: Wake and Cavity boundaries based on building shape Accounts for plume rise, streamline deflections and enhance turbulence Partitions plume mass between the primary plume, the cavity, and the far wake Cavity material is well-mixed; then material is re-emitted into wake Primary plume disperses within wake (streamline deflections and enhance turbulence) Currently accounts for wind direction by redefining the rectangular building dimensions. 2

13 Wind Tunnel Study Building Geometry H = 15 cm H X W = 1 X 2 1 X 4 1 X 8 H L W (L=H) Model Scale 1:150 Source Height and Location Wind Angle WIND WIND h s = 1.2H, 1.5H, 2H, 3H, 4H Θ Θ = 0, 15, 30, 45, 60 Perry et al., 2016, Atmospheric Environment, 142,

14 Vertical concentration profiles No Building Perpendicular Winds 1x2 x = 3H With 1x2 Building concentration Source height 1.5H 1.2H concentration 4

15 Changes in Plume Parameters with x Perpendicular winds H X W = 1 X 2 1 X 8 H L W σ z h eff 5

16 Large Eddy Simulations (LES) Embedded LES Approach LES zone subgrid-scale model WALE (Wall-Adapting Local Eddy viscosity) RANS zone: Shear-Stress Transport k-ω model Schmidt number = 0.7 Interface treatment: Vortex method 3D Computational Mesh (x-z slice) ~ 6 million grid cells in domain 5H 20H RANS ZONE 8H LES ZONE 3H H 6

17 Wind Tunnel & LES Wind Tunnel 1x2 FLUENT (ELES) 7

18 Building Downwash in AERMOD How PRIME works: Defines near & far wake boundaries Far wake Near wake (cavity) 8

19 Building Downwash in AERMOD How PRIME works: Defines near & far wake boundaries Deflects the plume centerline based on building-affected streamlines 9

20 Building Downwash in AERMOD How PRIME works: Defines near & far wake boundaries Deflects the plume centerline based on building-affected streamlines Partitions the plume: Primary, Cavity & Re-emitted Primary Cavity Re-emitted 10

21 Building Downwash in AERMOD How PRIME works: Defines near & far wake boundaries Deflects the plume centerline based on building-affected streamlines Partitions the plume: Primary, Cavity & Re-emitted Adjusts plume growth rate in wake based on wake turbulence ddσσ zz dddd ~ σσ ww UU Wake turbulence intensity relative to ambient values x/h 11

22 Building Downwash in AERMOD Ground-level concentration for a 1x2x1 building and a 1.5H stack Wind tunnel LES AERMOD 12

23 Building Downwash in AERMOD σσ zz Cavity Re-emitted x/h Three proposed model enhancements: Fix mismatch in plume width at transition between cavity and far wake x/h 13

24 Building Downwash in AERMOD σσ zz Cavity Re-emitted x/h Three proposed model enhancements: Fix mismatch in plume width at transition between cavity and far wake x/h 14

25 Building Downwash in AERMOD σσ zz Three proposed model enhancements: Fix mismatch in plume width at transition between cavity and far wake Cavity Re-emitted x/h Use effective wind speed for primary plume (currently using stack height wind sp) + U eff x/h 15

26 Building Downwash in AERMOD σσ zz Three proposed model enhancements: Fix mismatch in plume width at transition between cavity and far wake Cavity Re-emitted x/h Use effective wind speed for primary plume (currently using stack height wind sp) Adjust cap on ambient turbulence level + U eff x/h 16

27 Building Downwash in AERMOD Ground-level concentration for a 1x2x1 building and a 1.5H stack Incorporating enhancements Wind tunnel AERMOD LES 17

28 Building Downwash in AERMOD Wind direction 0 Original AERMOD With Enhancements Stack location: o: DM0, : UM0; : UC-; : DC+. Measurement location: Open symbols: xx = 3HH; closed xx = 10HH. Bldg size (symbol size): Small 1:1:2; medium 1:1:4; large 1:1:8 18

29 Building Downwash in AERMOD 4 th Enhancement for non-perpendicular winds BPIP AERMOD s building pre-processor Long building at 45 to wind Wind BPIP creates substitute building based on extremities of footprint Test alternative based on along-wind length of building cross section 19

30 Building Downwash in AERMOD Wind direction 45 Original AERMOD With Enhancements Stack location: o: DM0, : UM0; : UC-; : DC+. Measurement location: Open symbols: xx = 3HH; closed xx = 10HH. Bldg size (symbol size): Small 1:1:2; medium 1:1:4; large 1:1:8 20

31 Building Downwash Summary Wind tunnel study Analysis & comparison with AERMOD/PRIME Computational Fluid Dynamics (LES) PRIME algorithm investigations Plume width matching for component parts of plume Effective wind speed Turbulence levels BPIP building characterization Other wind direction effects 21

32 PRIME2 Committee PRIME2 Committee Advisory Subcommittee Formed by AWMA s Atmospheric Modeling and Meteorology Subcommittee in 2016 Activities: Assembling existing databases Additional wind tunnel testing Modifying & testing algorithms Focuses: Velocity deficit in wake Turbulence in wake region Streamlines over building Solid, lattice, and streamlined structures 22