Document V3N7OV3kgoNk32aqXDB65xrZZ

Air Dispersion M odeling Verification Study Oct 22,2003 g tNoa toJ AR226-2346 EXP015790 O b jective Compare results from ambient air sampling and air dispersion modeling to understand how modeling predictions relate to measured ambient air concentrations to0^too1\ to EXP015791 Verification Study Scope Air Sampling Program Analytical Methods Modeling Analysis of Results oNm two EXP015792 Air Sampling Locations Simultaneous coverage of all 4 major wind directions (N, S, E, W) Capture point of predicted maximum concentration from modeling Include co-located samplers at predicted maximum Total of six locations IaoMm to EXP015793 Air Sampling Loc ASH026725 EXP015794 ASH026726 EXP015795 | /* Air Sampling Strategy Measure ambient air levels i 24-hour sampling duration Seven (7) sampling events scheduled over 6 weeks Vary day of week to capture variations in daily operations (none expected) and variations in meteorology Record site-specific meteorological and operational data EXP015796 Monitoring Strategy: Samples taken every 6 days for 6 weeks. MONTH! 2 3 4 5 6 8 9 10 11 12 1 4 15 16 17 1 8 0 )2 0 21 22 23 2 4 0 ) 2 6 27 28 29 3 d _______ MONTH 2 12 3 4 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 EXP015797 Monitoring Strategy: Samples taken every 6 days for 6 weeks, On site training For internal use: potential schedule S M OTctW/NTohv F S -- *28 29Qc>)31 1 2 3 4 6 7 8 9 1 0 ^ ^ 1 2 13 14 15 16 @ 1 8 19 20 24 25 26 2f 2 8 @ ) 30 December 12 3 g 7 8 9 10 11 12 13 SoB 14 15 16 17 18 19 20 to MVO 21 22 23 24 25 26 27 28 29 30 31 EXP015798 Total Samples (1) field blank per event (5) primary samples per event (1) duplicate sample per event*_____________ (7) samples x (7) sampling events = 49 total samples ffl oN> WO * Sample location 3 is co-located with location 2 and serves as a duplicate sample EXP015799 Equipment "Sorbent" sampling method to capture PFOA on ion-exchange media OSHA Versatile Sampling tube (OVS) Captures particles and vapors Remote power source will be provided Weatherproof stations for sampling | equipment (<Wft EXP015800 OVS Tube / ASH026732 EXP015801 Analytical Method Validated method being used Acetone extraction followed by liquid chromatography/mass spectrometry (LC/MS) Detection limit of 0.1 ug/m3 gaonmttooj> EXP015802 i Modeling Program Using U.S. EPAISCST3 Model Model will simulate actual field conditions - Site-specific surface meteorological data (wind speed and direction, temp, stability) i - O ff-site upper air meteorological data - Em issions data Site specific operational data w ill be used to relate production data to emissions levels using "capacity factor" Capacity factor based on prior stack testing and mass balance I - 24-hour averaging time - Predictions at actual height and location o f air monitoring i intake EXP015803 Comparison o f Results Comparing 35 pairs of measured and modeled data - (5) primary sampling points x (7) sampling events Will use statistical tools to evaluate correlation/differences between measured and modeled data Internal? Questions we are trying to answer Are modeled and measured data wellcorrelated? (+/- X%) What is the probability or confidence level that the model predictions will be equal to or greater than the measured value? Does model "accuracy" change at different oat0uoJ\ concentration levels? EXP015805 Statistical Analyses Plot of measured vs. modeled data to test linear correlation Statistical goodness o f fit criteria (R -squared) Simple sign test for residuals - Calculate (measured - modeled) - Trend in residuals: positive, negative, neutral Plot of residuals vs. modeled data to test goodness of fit at different concentrations T est sp ecific hypotheses; exam ples - Model predicts within 20% o f measured value. - Model over predicts 50% o f the time. EXP015806 Project Timing remaining internal questions Are there site communications we need to think about prior to sampling? otcuon>3J EXP015808 EXTRA SLIDE Feasibility o f Deposition M odeling Wet and dry deposition rats are small - maximum rate 0.18 gram/m2-yr for year 2000 Direct field measurement on 24-hr basis immeasurable, so direct comparison with short term modeling not feasible Use model verification to add degree o f confidence Oroo> o EXP015809 EXTRA SLIDE Deposition M odeling Both wet and dry deposition algorithms depend on the ability of the model to calculate ambient air concentrations - Dry deposition flux = Ambient hourly concentration x deposition velocity i - Wet deposition flux = f(scavenging coefficient x ambient concentration integrated in the vertical) Verification of ambient air model will help | confirm basis for deposition calculations W EXP015810