The document discusses improvements to the AERMOD plume rise model, specifically the PRIME2 enhancements, which include new calculations for turbulence and wind speed. These advancements aim to correct inaccuracies in predicted concentrations affected by building downwash effects, as supported by collaboration with the EPA and various industry associations. A preliminary evaluation for four case studies highlights the comparative performance of the updated model against previous versions, indicating inconsistencies in predicted concentrations across different scenarios.

1. Sergio Guerra, PhD | GHD
Ron Petersen, PhD, CCM | Petersen Research and Consulting
James Paumier | Consultant
Extended Abstract: 616634
A&WMA’s 112th Annual Conference & Exhibition
Quebec City, Canada
June 27, 2019
PRIME2 Model Evaluation
-AERMOD 18081
-AERMOD D18227_ORD
-AERMOD D18227_PRM2

2. Abstract
The Plume Rise Model Enhancements (PRIME) formulation in AERMOD has been updated based new equations developed from wind tunnel measurements taken downwind of
various solid and streamlined structures. These new equations, along with other building downwash improvements have been included as alpha options in the upcoming new
version of AERMOD.
The PRIME2 options include:
• PRIME2UTurb which enables enhanced calculations of turbulence and wind speed
• PRIME2Ueff which defines the height used to compute effective parameters Ueff, Sweff, Sveff and Tgeff at plume height and at 30 m
• Streamline defines the set of constants for modeling all structures as streamlined. If omitted, rectangular building constants are used.
The ORD Options include:
• PRIMEUeff which controls the heights for which the wind speed is calculated for the main plume concentrations.
• Average between plume height and receptor height recommended in ORD version
• Default is current method in AERMOD, stack height wind speed.
• PRIMETurb which adjusts the vertical turbulence intensity, wiz0 from 0.6 to 0.7.
• PRIMECav modifies the cavity calculations
These improvements aim to address important theoretical issues that significantly affect the accuracy of predicted concentrations subject to downwash effects. This research effort
was funded in part by the American Petroleum Institute, the Electric Power Research Institute, the Corn Refiners Association and the American Forest & Paper Association. As part
of it, the PRIME2 subcommittee under the A&WMA APM committee was formed to: (1) establish a mechanism to review, approve and implement new science into the model for this
and future improvements; and (2) provide a technical review forum to improve the PRIME building downwash algorithms. Collaboration and cooperation from the EPA Office of
Research and Development (ORD) has been on-going during the research project resulting in new alpha options aimed at solving known issues with the treatment of building
downwash effects in AERMOD. The intent is that these experimental options will be tested by the user community to create enough justification to make these beta (approved on a
case-by-case basis) and eventually default options in AERMOD.
A preliminary evaluation for the following four cases will be presented:
• Arconic- Davenport, IA (formerly Alcoa)
• Mirant Potomac River Generating Station- Alexandria, VA
• Basic American Foods- Blackfoot, ID
• Oakley Generating Station- Oakley, CA
The evaluation includes comparing 1-hr, 24-hr and annual averages along with Q-Q plots and isopleths. A discussion related to the results obtained will also be presented.

3. Why a New Downwash Model?
• AERMOD’s PRIME algorithm based on research carried out before
2000
• Original theory based on a limited number of building dimensions
and building types
• Theory is not suitable for porous, streamlined, wide or elongated
structures
• Theory based on theoretical assumptions that can be improved
PRIME2 Model Evaluation

4. Wind Tunnel Testing
Three Rectangular Buildings
Three rectangular buildings Two Streamlined Structures
PRIME2; Phase II Building Downwash Enhancements4
1:4:4

5. New equation documented
https://doi.org/10.1016/j.jweia.2017.11.027 PRIME2 Model Evaluation

6. Update AERMOD
AERMOD Fortran Files Modified
PRIME.f
Where all building downwash
calculations are carried out
Calc1.f
Where approach wind
conditions come from
Modifications
• New Zeff
• New Ueff, SWeff, SVeff
• New U30, SW30 and
SV30 for input to new
equation
Modifications
• New turbulence equation
• New wind speed equation
Recompile and Create
AERMOD/PRIME2
6 PRIME2 Model Evaluation

7. EPA ORD AERMOD/PRIME
Modifications
Current PRIME
Three ORD model enhancements
1. Fix mismatch in plume vertical
spread at transition between
cavity and far wake.
2. Use effective wind speed, Ueff,
for primary plume versus stack
height for concentration
calculations
3. Adjust cap on ambient
turbulence from 0.06 to 0.07.
No effect on PRIME2.
Fix 1
Current PRIME
Ueff at Stack Top
Fix 2
@ (Hp+RecH)/2
RecH) = 0
Hp
PRIME2 Model Evaluation

8. Field Evaluations
Presented on June 28, 2-18 at the A&WMA’s 11th Annual Conference &
Exhibition in Hartford, CT
PRIME2 Model Evaluation

9. Bowline Point
9
29.6
m
65.2
m
86.9
m
Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
STACK 0 - 449.3 86.87 358 - 409 7.9 – 30.9 5.72
9
PRIME2 Model Evaluation

10. Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
STACK 1 39.2 554-584 17-21 3.66
34.0
m
39.2
m
Alaska North Slope Field Study
10

11. Millstone Nuclear Power Station
(Dominion Millstone Power Station)
Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
REAC 1 48.3 291 - 297 4.6 – 8.7 2.12
TURB1 1 29.1 292 - 306 10.5 1.4
TURB2 1 29.1 292 - 306 10.5 1.4
TURB3 1 29.1 292 306 10.5 1.4
41.6
m
44.7
m
27.6
m
29.1
m
48.3
m
11

12. Duane Arnold Energy Center
23.5
m
42.7
m
1
m
23.5
m
45.7
m
Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
STACK 5 1 45.7 293 – 299 7.4 - 40.8 1.4
STACK 4 1 23.5 294 - 300 0.01 2.12
STACK 1 1 1.0 299 - 303 0.01 1.4
12

13. Duane Arnold Energy Center
23.5
m
42.7
m
1
m
23.5
m
45.7
m
Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
STACK 5 1 45.7 293 – 299 7.4 - 40.8 1.4
STACK 4 1 23.5 294 - 300 0.01 2.12
STACK 1 1 1.0 299 - 303 0.01 1.4
13

14. Duane Arnold Energy Center
23.5
m
42.7
m
1
m
23.5
m
45.7
m
Q (g/s) Hs (m) Ts (K) Vs (m/s) Ds (m)
STACK 5 1 45.7 293 – 299 7.4 - 40.8 1.4
STACK 4 1 23.5 294 - 300 0.01 2.12
STACK 1 1 1.0 299 - 303 0.01 1.4
14

15. Case Study Evaluation
Conducted for the following four cases:
• Arconic (formerly Alcoa )- Davenport, IA
• Mirant Potomac River Generating Station- Alexandria, VA
• Basic American Foods- Blackfoot, ID
• Oakley Generating Station- Oakley, CA
The evaluation compares the following averages for each
case:
• 1-hour (H1H)
• 24-hr (H1H)
• Annual
PRIME2 Model Evaluation

16. Arconic- Davenport, IA (formerly Alcoa)
PRIME2 Model Evaluation
Met data: Davenport, IA met data, 2010-2014 from IDNR (http://www.iowadnr.gov/Environmental-Protection/Air-
Quality/Modeling/Dispersion-Modeling/Meteorological-Data).
Q Stack Height Stack Temp. Stack Vel. Stack Diam. Height of BPIP Cont. Bdg.
g/s m K m/s m m
S_349 1.94 21.34 310.93 17.82 2.46 16.7-17.5
Stack

17. Arconic H1H Comparisons for 1-hr, 24-hr and
Annual Averages
172.7
30.0
3.0
166.0
26.8
4.5
95.9
19.7
1.0
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
200.0
1-hr 24-hr Annual
Q(µg/m3)
Arconic 349 w BPIP H1H
18081 D18227-ORD D18227-PRM2
PRIME2 Model Evaluation

18. Arconic: BPIP inputs ran with AERMOD v18081
PRIME2 Model Evaluation

19. Arconic: BPIP inputs ran w AERMOD vD18227_ORD
PRIME2 Model Evaluation

20. Arconic: BPIP inputs ran w AERMOD vD18227_PRM2
PRIME2 Model Evaluation

21. Mirant Potomac River Generating Station,
Alexandria, VA
PRIME2 Model Evaluation
Met data: Reagan Airport, VA original 1992 met data reprocessed with AERMETv18081.
Q Stack Height Stack Temp. Stack Vel. Stack Diam. Height of BPIP Cont. Bdg.
g/s m K m/s m m
BS4 1 48.2 303.71 18.42 2.44 35.3
Stack

22. Mirant H1H Comparisons for 1-hr, 24-hr and
Annual Averages
21.1
6.9
0.6
32.3
11.5
1.1
26.7
7.6
0.3
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
1-hr 24-hr Annual
Q(µg/m3)
Mirant BS4 w BPIP H1H
18081 D18227-ORD D18227-PRM2
PRIME2 Model Evaluation

23. Mirant: BPIP inputs ran with AERMOD v18081
PRIME2 Model Evaluation

24. Mirant: BPIP inputs ran
with AERMOD vD18227_ORD
PRIME2 Model Evaluation

25. Mirant: BPIP inputs ran with
AERMOD vD18227_PRM2
PRIME2 Model Evaluation

26. Basic American Foods, Blackfoot, ID
PRIME2 Model Evaluation
Met data: On-site met data from the Blackfoot Met Tower operated by the Idaho National Laboratory and
supplemented with data from the Pocatello Regional Airport, 2002-2006.

27. Basic American Foods, Blackfoot, ID
PRIME2 Model Evaluation
Met data: On-site met data from the Blackfoot Met Tower operated by the Idaho National Laboratory and
supplemented with data from the Pocatello Regional Airport, 2002-2006.
Q Stack Height Stack Temp. Stack Vel. Stack Diam. Height of BPIP Cont. Bdg.
g/s m K m/s m m
EU_24 0.13 12.65 359.26 12.44 0.76
EU_25 0.06 12.65 338.71 5.76 0.91
EU_26 0.13 12.65 359.26 12.44 0.76
EU_31 0.12 12.65 344.26 14.57 1.04
EU_32 0.05 12.60 338.71 10.52 0.79
EU_33 0.05 12.60 327.59 11.34 0.61
Stack
7.0

28. BAF H1H Comparisons for 1-hr, 24-hr and
Annual Averages
451.3
146.0
48.8
611.6
152.7
51.1
350.7
119.1
35.4
0.0
100.0
200.0
300.0
400.0
500.0
600.0
700.0
1-hr 24-hr Annual
Q(µg/m3)
BAF EU24-27& EU31-33 w BPIP H1H
18081 D18227-ORD D18227-PRM2
PRIME2 Model Evaluation

29. BAF: BPIP inputs ran with AERMOD v18081
PRIME2 Model Evaluation

30. BAF: BPIP inputs ran
with AERMOD vD18227_ORD
PRIME2 Model Evaluation

31. BAF: BPIP inputs ran with
AERMOD vD18227_PRM2
PRIME2 Model Evaluation

32. Oakley Generating Station, Oakley, CA
PRIME2 Model Evaluation
Met data: Metro Oakland International Airport, CA met data for years 2009-2013. Downloaded from CARB
(https://www.arb.ca.gov/toxics/harp/metfiles2.htm).
Q Stack Height Stack Temp. Stack Vel. Stack Diam. Height of BPIP Cont. Bdg.
g/s m K m/s m m
EU_31 1 47.41 362.00 22.26 5.59 31.7
Stack

33. Oakley GS H1H Comparisons for 1-hr, 24-hr and
Annual Averages
1.4
0.4
0.1
1.6
0.4
0.1
4.3
0.8
0.1
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
1-hr 24-hr Annual
Q(µg/m3)
Oakley GS Tur1 w BPIP H1H
18081 D18227-ORD D18227-PRM2
PRIME2 Model Evaluation

34. Oakley: BPIP inputs ran with AERMOD v18081
PRIME2 Model Evaluation

35. Oakley: BPIP inputs ran
with AERMOD vD18227_ORD
PRIME2 Model Evaluation

36. Oakley: BPIP inputs ran with
AERMOD vD18227_PRM2
PRIME2 Model Evaluation

37. Results
• Performance of version D18227-PRM2 is inconsistent
• Predicted values are lower than the base case for Arconic and BAF with
small differences for Mirant (except for annual which has PRIME2
significantly lower).
• The base case predictions are lower than the ones from the D18227-
ORD and D18227-PRM2 versions for the Oakley case.
PRIME2 Model Evaluation

38. Results
• In general, version D18227-ORD produces higher concentrations
than AERMOD 18081.
• The primary reason for this is that the ORD version uses the wind speed
at the average between plume centerline height and receptor height to
calculate concentrations versus stack height in AERMOD.
• PRIME2’s theory does not depend on the wake height but rather the
building height, width, length and position relative to the stack.
PRIME2 Model Evaluation

39. Future Areas of Improvements
• PRIME plume rise
• PRIME streamline
• Height used to calculate
turbulence
• BPIP-PRM
• Cavity plume amplification
problem
PRIME2 Model Evaluation

40. Conclusions
• PRIME2 includes a superior theory to account for building
downwash effects for rectangular and streamlined structures.
• Inconsistencies may be due to other parts of the model
• Work from EPA ORD complements the work performed by PRIME2.
• Plan is to continue EPA collaboration to address model
improvements to AERMOD related to building downwash.
PRIME2 Model Evaluation

41. Sergio A. Guerra, PhD Ron Petersen, PhD, CCM
sergio.guerra@ghd.com rpetersen@petersenresearch.com
Office: + 720 974 0935 Mobile:+1 970 690 1344
PRIME2 Model Evaluation

42. www.ghd.com
PRIME2 Model Evaluation

43. Objectives
PRIME2 Subcommittee was
formed to:
• Establish a mechanism to review,
approve and implement new science
into the model for this and future
improvements
• Provide a technical review forum to
improve the PRIME building
downwash algorithms
PRIME2 Model Evaluation

44. Next Steps: Implementation
Model
Improvements
Submitted to
EPA
AERMOD with
PRIME2 and
ORD Switches
EPA
releases
New PRIME2
as Alpha
option
App W
Sec
3.2.2
reqs.
EPA
releases
PRIME2 as
Beta
option
Notice of
proposed
rulemaking
(NPRM)
New PRIME is
released as
default
regulatory
option
Alpha option needs to meet the alternative refined model requirements in App W, Section
3.2.2 before it can become a Beta option. These requirements include:
1-Model has received a scientific peer review;
2-Model can be demonstrated to be applicable to the problem on a theoretical basis;
3-The data bases to perform analysis are available and adequate;
4-Appropriate performance evaluations show model is not biased toward underestimation; &
5-A protocol on methods and procedures to be followed has been established
PRIME2 Model Evaluation