PRIME2: Consequence Analysis and Model Evaluation
1 like329 views
The document discusses the Prime2 model for wind effects on buildings, detailing its features, including improved theories for building wake effects and downwash predictions. It presents the evaluation of the model's predictions against field data and addresses its overprediction tendencies compared to prior models. The conclusion highlights the need for ongoing collaboration to enhance the model and address its shortcomings.
PRIME2: Consequence Analysis and Model Evaluation
- 1. PRIME2 Consequence Analysis and Model Evaluation Pacific Northwest International Section of the A&WMA 2017 Annual Conference Boise, ID. Sergio A. Guerra, PhD Ron Petersen, PhD, CCM November 3, 20171
- 2. Outline 1. Background on PRIME2 2. Consequence Analysis 3. Field Evaluation 4. Case studies 2 PRIME2 Consequence Analysis and Model Evaluation
- 3. 3 Key Features of PRIME2 • Building wake effects decay rapidly back to ambient levels above the top of the building versus the current theory that has these effects extending up to 3 building heights. • Lateral dispersion enhancement in the wake is less than vertical dispersion enhancement (current PRIME has them identical). • The approach turbulence and wind speed is calculated at a more appropriate height versus the current theory where half the wake height at 15 building heights downwind of the building is used. • Wake effects for streamlined structures are reduced. • Wake effects decrease as approach roughness increases. PRIME2 Consequence Analysis and Model Evaluation
- 4. 4 Project Summary • Wind tunnel testing was performed to evaluate downwash effects from rectangular and streamlined structures. • CPP developed equations for predicting wind speed and turbulence in building wakes for rectangular and streamlined structures based on wind tunnel observations. • CPP’s updates were compiled into a new AERMOD executable (PRIME2). • Field versus model comparisons show that PRIME2 predictions are generally within a factor of two of field observations but have a overprediction tendency. Predictions also tend to be higher values than with PRIME. • Other theoretical problems have been identified. Correcting these may alleviate the current overprediction tendency in PRIME2. PRIME2 Consequence Analysis and Model Evaluation
- 5. Implementation Process CPP and ORD Submittals to EPA OAQPS Journal Articles Published OAQPS Codes CPP and ORD Enhancements EPA releases New PRIME as Alpha option EPA releases PRIME 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; and 5-A protocol on methods and procedures to be followed has been established 5 PRIME2 Consequence Analysis and Model Evaluation
- 6. Consequence Analysis PRIME2 Consequence Analysis and Model Evaluation6
- 7. PRIME2 Consequence Analysis and Model Evaluation7 PRIME2_v17234a Evaluation 1:10:10 BDG with MakeMet Hs=1.2Hb PRIME2_v17234a2 AERMOD_v16216r zo=2cm zo=25cm zo=100cm Max=173.0 ug/m3 Max=73.5 ug/m3 Max=132.9 ug/m3 Max=72.2 ug/m3 Max=79.0 ug/m3 Max=70.9 ug/m3
- 8. PRIME2 Consequence Analysis and Model Evaluation8 PRIME2_v17234a Evaluation 1:10:10 BDG with MakeMet Hs=1.5Hb PRIME2_v17234a2 AERMOD_v16216r zo=2cm zo=25cm zo=100cm Max=130.1 ug/m3 Max=97.0 ug/m3 Max=53.2 ug/m3 Max=68.8 ug/m3 Max=69.1 ug/m3 Max=66.8 ug/m3
- 9. PRIME2 Consequence Analysis and Model Evaluation9 PRIME2_v17234a Evaluation 1:10:10 BDG with MakeMet Hs=2.5Hb PRIME2_v17234a2 AERMOD_v16216r zo=2cm zo=25cm zo=100cm Max=37.2 ug/m3 Max=26.6 ug/m3 Max=23.1 ug/m3 Max=37.8 ug/m3 Max=35.6 ug/m3 Max=37.9 ug/m3
- 10. PRIME2 Consequence Analysis and Model Evaluation10 PRIME2_v17234a Evaluation Bowline Point BDGs with Bowline Met Data PRIME2_v17234a2 AERMOD_v16216r Hs=1.33Hb=87.8m Hs=1.80Hb=117.4m Hs=2.50Hb=163.1m Max=1013.1 ug/m3 Max=638.3 ug/m3 Max=279.9 ug/m3 Max=279.9 ug/m3 Max=192.2 ug/m3 Max=192.2 ug/m3 Max Observed = 823.5 ug/m3
- 11. Field Evaluation PRIME2 Consequence Analysis and Model Evaluation11
- 12. Bowline Point Field Evaluation for Receptors 1 and 3 Q-Q Plot of Predicted vs. Observed Concs. with BPIP Values Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias R1&3 AERMODv16216r (ug/m3) 422.17 447.71 1.06 0.06 R1&3 PRIME2v17234a (ug/m3) 422.17 684.51 1.62 0.47 PRIME2 Consequence Analysis and Model Evaluation12
- 13. Refined BPIP Method Example Bowline Point PRIME2 Consequence Analysis and Model Evaluation13
- 14. Refined BPIP Method Example: Bowline Point Merged Tiers Building Dimensions BPIP (m) Updated BPIP (m) Building Height(Hb) 65.23 65.23 Building Width (W) 121.95 121.95 Building Length (L) 109.93 35.98 XBADJ -127.62 -97.20 YBADJ -2.47 -2.5 Assumption 1: Tallest tiers combine (green bdg) Assumption 2: BDG WIDTH (W) is crosswind width of merged tier. Assumption 3: XBADJ starts at the upwind edge of the merged tier Assumption 4: BDG LENGTH (L) is calculated by dividing the area of the merged tier by W PRIME2 Consequence Analysis and Model Evaluation14
- 15. Refined BPIP Method Example: Bowline Point Unmerged Tiers Building Dimensions BPIP (m) Updated BPIP (m) Building Height(Hb) 65.23 65.23 Building Width (W) 94.57 49.9 Building Length (L) 130.27 27.65 XBADJ -132.56 -127.90 YBADJ -27.17 -4.0 Assumption 1: Tallest tiers do not combine (green bdg) Assumption 2: BDG WIDTH (W) is crosswind width of unmerged tier. Assumption 3: XBADJ starts at the upwind edge of the tallest tier Assumption 4: BDG LENGTH (L) is calculated by dividing the area of the tallest tier by W PRIME2 Consequence Analysis and Model Evaluation15
- 16. Bowline Point Field Evaluation for Receptors 1 and 3 Q-Q Plot of Predicted vs. Observed Concs. with BPIP Values Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias R1&3 AERMODv16216r (ug/m3) 422.17 447.71 1.06 0.06 R1&3 PRIME2v17234a (ug/m3) 422.17 684.51 1.62 0.47 PRIME2 Consequence Analysis and Model Evaluation16
- 17. Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias R1&3 AERMODv16216r (ug/m3) 422.17 237.67 0.56 -0.56 R1&3 PRIME2v17234a (ug/m3) 422.17 535.01 1.27 0.24 Bowline Point Field Evaluation for Receptors 1 and 3 Q-Q Plot of Predicted vs. Observed Concs. with Modified BPIP Values PRIME2 Consequence Analysis and Model Evaluation17
- 18. Refined BPIP Method Example Alaska North Slope PRIME2 Consequence Analysis and Model Evaluation18
- 19. Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias AERMODv16216r 3.13 3.59 1.15 0.137 PRIME2_17234 3.13 7.58 2.42 0.829 Alaska North Slope Field Evaluation Q-Q Plot of Predicted vs. Observed Concs. with BPIP Values PRIME2 Consequence Analysis and Model Evaluation19
- 20. Building Dimensions BPIP (m) Updated BPIP (m) Building Height(Hb) 34.0 34.0 Building Width (W) 51.26 51.26 Building Length (L) 55.67 25.81 XBADJ -45.24 -43.70 YBADJ 6.58 6.6 Assumption 1: Tallest tier combine (green bdg) Assumption 2: BDG WIDTH (W) is crosswind width of merged tier. Assumption 2: XBADJ starts at the lee edge of the merged tier Assumption 3: YBADJ is calculated by dividing the area of the merged tier by the width of the artificially created building Refined BPIP Method Example: Alaska North Slope Merged Tiers PRIME2 Consequence Analysis and Model Evaluation20
- 21. Building Dimensions BPIP (m) Updated BPIP (m) Building Height(Hb) 34.0 34.0 Building Width (W) 52.98 20.25 Building Length (L) 28.61 25.30 XBADJ -28.7 -28.6 YBADJ -11.79 4.8 Assumption 1: Tallest tier do not combine (green bdg) Assumption 2: XBADJ starts at the lee edge of the merged tier Assumption 3: YBADJ is calculated by dividing the area of the merged tier by the width of the artificially created building Refined BPIP Method Example: Alaska North Slope Unmerged Tiers PRIME2 Consequence Analysis and Model Evaluation21
- 22. Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias AERMODv16216r 3.13 3.59 1.15 0.137 PRIME2_17234 3.13 7.58 2.42 0.829 Alaska North Slope Field Evaluation Q-Q Plot of Predicted vs. Observed Concs. with BPIP Values PRIME2 Consequence Analysis and Model Evaluation22
- 23. Model Version Top 25 Mean X obs Top 25 Mean X predict Top 25 Pre/Obs Fractional Bias AERMODv16216r 3.13 2.78 0.89 -0.120 PRIME2_17234 3.13 6.14 1.96 0.648 Alaska North Slope Field Evaluation Q-Q Plot of Predicted vs. Observed Concs. with Modified BPIP Values PRIME2 Consequence Analysis and Model Evaluation23
- 24. Alaska North Slope: Observed Values Max: 5.29 µg/m3 PRIME2 Consequence Analysis and Model Evaluation24
- 25. Alaska North Slope: AERMODv16216r Max: 4.48 µg/m3 PRIME2 Consequence Analysis and Model Evaluation25
- 26. Alaska North Slope: PRIME2v17234a Max: 13.20 µg/m3 PRIME2 Consequence Analysis and Model Evaluation26
- 27. What Could be Causing Higher PRIME2 Predictions? • BPIP Building Dimensions Inputs: Building length and upwind and downwind faces not defined correctly. • Streamlines: Need to be reviewed and updated as necessary. • The wind speed used to calculate concentrations. Currently PRIME and PRIME2 are using the stack height wind speed. PRIME2 Consequence Analysis and Model Evaluation27
- 28. Case Study Comparison PRIME2 Consequence Analysis and Model Evaluation28
- 29. BPIP Values: AERMODv16216r H1H 24-hr avg = 78.9ug/m3 PRIME2 Consequence Analysis and Model Evaluation29
- 30. BPIP Values: PRIME2_17234a2 PRIME2 Consequence Analysis and Model Evaluation30 H1H 24-hr avg = 41.4ug/m3
- 31. EBD Values: AERMODv16216r PRIME2 Consequence Analysis and Model Evaluation31 H1H 24-hr avg = 39.5ug/m3
- 32. Conclusions • PRIME2 includes a superior theory to account for building downwash effects for rectangular and streamlined structures. • Benefits from improved theory cannot be fully realized due to outstanding issues in the model. • Work from EPA ORD complements the work performed by CPP. • Plan is to continue EPA collaboration to address model improvements to AERMOD related to building downwash. PRIME2 Consequence Analysis and Model Evaluation32
- 33. Sergio A. Guerra, PhD Ron Petersen, PhD, CCM sguerra@cppwind.com rpetersen@cppwind.com Mobile: + 612 584 9595 Mobile:+1 970 690 1344 wwww.SergioAGuerra.com CPP, Inc. 2400 Midpoint Drive, Suite 190 Fort Collins, CO 80525 + 970 221 3371 www.cppwind.com @CPPWindExperts Questions? 33 PRIME2 Consequence Analysis and Model Evaluation