Document DvymK961nym44MOQNvK2YV2kQ

AR226-2322 ^y Charles J Zarzecki 04/03/2003 09:31 AM To: cc: Subject: Catherine A Barton/AE/DuPont Re: Beychok Article Q Cathie, First, by "Monday's meeting", I assume you mean April 14? Regarding CW, base on my schedule and yours, I would say a minimum of 2 weeks. However, I was at the CW wastewater partitioning team meeting yesterday, and it looks like they will only have a sampling plan proposed by Monday. Then they will have to implement it, get the results, and then we can move ahead with the TOXCHEM modeling. After that, we will have an emission rate for the wastewater. Looks like it will be awhile before we can do any ambient air dispersion modeling to get definitive results, anyway. As a first cut, I guess we can run the stack numbers to see what they look like. Charlie Z. Catherine A Barton ^ Catherine A Barton 04/02/2003 04:57 PM To: cc: Subject: Charles J Zarzecki/AE/DuPontQDuPont Re: Beychok Article Q Charlie, First, thanks for taking the time to go over Beychok's article thoroughly. I have the same instinctive feeling that you do: to lambast (yes, its' a great word) the model is a losing battle. We have to develop a more positive strategy that the agency can live with. I like your alternative strategy of fine-tuning our inputs. I want to contact the site guys who did the estimates to see how much better they think they can get (I'll do that tomorrow). For Monday's meeting, I want to put together a graphic that shows them some options. Let me soak on it overnight and run something by you manana. Incidentally, I met with the Cworks strategy team and they are happy with our modeling plan. Ann Masse wants me to get the emissions information from the site guys by 4/11 so we can start cranking numbers. Help me with this: when do you think we can reallistically have the first round of ambient air numbers? I have to convert my flow diagram into text for Ann, and she wants me to put in real dates for deliverables (pending her boys get us the emissions data by 4/11). I'll be in the office tomorrow, but I think you're at home. No matter...we'll connect. thanks again, EID747739 c Charles J Zarzecki ^ Charles J Zarzecki 04/02/2003 03:14 PM To: cc: Subject: Catherine A Barton/AE/DuPont Beychok Article Cathie, I finally was able to review Ihe Beychok article on error propagation in dispersion modeling. This could add fuel to Dave Rurak's fire. It's common knowledge that Gaussian models overpredict by a factor of 2, however, due to error propagation, he says as high as BOX. Also, as far as 1-hour concentrations are concerned, the actual averaging lime for Pasquill's dispersion coefficients range from 3 minutes to 30 minutes, depending who you talk to. I always thought they were 15-minute averages. This short-term to 1-hour assumption can be shown (by Beychok) to result in a 2.5X overprediction. However, in my opinion, since the 1-hour period is the basic lime-step in the ISC (and other) model, an annual average concentration at any given receptor is 8,7601-hour concentrations divided by 8,760 hours. If it predicts poorty for each 1-hour period, then it predicts poorly for the whole year. Dave has to understand that the model is a screening tool. The more accurate information you put into a model, the greater the accuracy of the prediction. Instead of lambasting (is that a real word?) the model's 1-hour prediction, we should investigate how to fine-tune .the input data and model options (e.g. particle settling, hours of operation, hourly emission rates, more accurate model (AERMOD), etc...). Or, go out and do some sampling to get the "real thing". What do you think? Beychok references about a half-dozen other publications that discuss the shortcomings of Gaussian models. Some of them go way back. As you can see, nobody really took them seriously. What I am getting at: the regulatory agencies (OEPA, WVDEQ) are not going to want to hear about how bad ISC is for particular time periods. They may be open to other "accepted" models and to sharpening our pencils. Regards, Charlie Z. EID747740