Environmental Impact Assessment Case study: Delimara Thermal Power Station, Malta Sensitivity analysis The assessment results depend on the the quality, range, coverage, completeness and consistency of all input data in an inclusive sense; the (usually simplifying and often arbitrary) assumptions made for the analysis; the models used and their (usually implicit) bias. The following "inputs" (data and assumptions) are analysied in terms of their impact on the output (assessment of compliance): Meteorology: choice of year, choice of monitoring stations vs. NCEP FNS re-analysis data, AERMET (mcip generated) vs, direct use of station data; Basic assumptions: comparison of tier 1 assumption (NOx = NO2) and alternative representations of NO2, using AERMOD/OLM (with different assumptions on ozone background levels, CAMx and full (CB05) photochemistry; Emissions: basic first order error analysis vs all compliance criteria, exploring ranges around the default emission values; Comparison of (assumed) constant emissions, vs. actual dynamic emisisons. Background concentrations: exploring a range of different scenarios for background concentrations (other major sources, traffic) in teir effect and contribution on compliance. Model comparison: comparison of results from AERMOD, CAMx, and PUFF results for selected critical hours/episodes. Model resolution for the basic regulatory model (AERMOD), the effect of receptor grid resolution (from 25 m to 1 km), flagstaff height; comparison of AERMOD results at receptor points *arbitrary spacing along logarithmic gradients) vs spatially averaged grid cell results; Meteorology AERMET 2010: Delimara 450 MW, 150m stack, 112.5 g/3 NOx; Meteorology: AERMET 2010, (extracted from MM5 re-analysis runs for the station: Zejtun) Resolution: 100m, 225,000 receptor points simulated, 8,760 hourly runs. Annual average: 7.69 µg/m3 Hourly maxium: 622.77 µg/m3   7120 instances (0.004 %)   distinct times: 1408 (16.07 %); distinct locations: 454 (2.018 %) Compliance: (simulated NOx against NO2 standards) annual: compliance hourly: violations As expected, values for average and maxima, violations extent and timing, vary between the three years and with the meteorological data sets used. Indicator theat are normalized (such as % violation, % of the total grid points in violation etc. stay more or less constant. met data annual avg. hourly max. hours violations % locations % AERMET 2008 4.32 436.05 52 0.002 10.29 AERMET 2009 9.60 372.00 1493 0.004 1.68 AERMET 2010 7.69 622.77 1408 0.004 2.02 Zejtun 2008 15.4 535.52 596 0.001 2.16 Zejtun 2009 18.8 702.72 513 0.001 1.37 Zejtun 2010 17.6 468.03 458 0.001 1.24 Model resolution 100 m: Delimara 450 MW, 150m stack, 112.5 g/3 NOx; Meteorology: AERMET 2008, (extracted from MM5 re-analysis runs for the station: Zejtun) Resolution: 100m, 8760 hourly runs. Annual average: 4.30 µg/m3 Hourly maxium: 436µg/m3   3007 instances (0.002 %)   distinct times: 52 (0.59 %); distinct locations: 2315 (10.29 %) Compliance: (simulated NOx against NO2 standards) annual: compliance hourly: violations As expected, average and maximum concentration values increase slightly with model grid resolution; the Gaussian solution provide a value for a specific point; since the concentrations decrease (somewhat simplified) exponentially with distance from the source, higher resolution means that receptor points and source get "closer", hence higher values. Indicator theat are normalized (such as % violation, % of the total grid points in violation etc. stay more or less constant. Small irregaularities are due to the fact that all resolutions tested share the same grid (point of origin), but the resolution are not perfectlu modulo 2, so that the relative source-receptor distances may change. resolution annual avg. hourly max. hours violations % locations % 50 m 4.32 438.75 54 0.002 10.35 100 m 4.30 436.05 52 0.002 10.29 250 m 4.28 422.74 47 0.002 10.44 500 m 3.94 371.48 32 0.001 9.89 1000 m 3.46 269.23 18 0.002 11.56 OLM: ozone background concentration AERMOD/OLM, 75µg/m3 background: Delimara 450 MW, 150m stack, 112.5 g/3 NOx emissions; Meteorology: AERMET 2009, (extracted from MM5 re-analysis runs for the station: Zejtun) Resolution: 100m, 8760 hourly runs. Annual average: 6.69 µg/m3 NO2 Hourly maxium: 109.1 µg/m3 NO2   0 instances (0.000 %)   distinct times: 0 (0.00 %); distinct locations: 0 (0.00 %) Compliance: (simulated NO2 against NO2 standards) annual: compliance hourly: compliance None of the OLM runs for NO2 predict violations of either annual or hourly NO2 standards. However, values vary with the assumed ozone background concentration, both the annual mean and the hourly maximum increasing mononously with the ozone background. O3 background annual avg. hourly max. O3 background annual avg. hourly max. 40 µg/m3 4.57 75.53 45 µg/m3 4.93 80.33 50 µg/m3 5.27 85.12 55 µg/m3 5.59 89.91 60 µg/m3 5.89 94.70 65 µg/m3 6.17 99.49 70 µg/m3 6.44 104.28 75 µg/m3 6.69 109.08 80 µg/m3 6.94 113.87 85 µg/m3 7.17 118.66 90 µg/m3 7.39 123.45 95 µg/m3 7.60 128.24

Delimara Power plant

Data availability

Assessment models

EIA methodology