Roughing trickling filters are used to reduce organic loadings ahead of the activated sludge process at a Merck & Co. Inc. manufacturing site. The site is subject to the Pharmaceutical MACT (Maximum Achievable Control Technology) Rule, and must demonstrate required reductions of hazardous air pollutants (HAPs). TOXCHEM+ predictive fate software is used at the site to estimate the reduction and removal of the HAPs during wastewater treatment. Sitespecific biodegradation rates of the HAPs can be demonstrated as part of an overall strategy to devise the best approach for achieving compliance with the MACT Rule. An experimental sampling program of filter influent and effluent and exhaust air was completed over 2 days to assess the fate of the HAPs in the process, and to estimate site-specific biodegradation rate coefficients for the HAPs for use in the fate modeling software.In each of 6 tests, mass balances for the organic compounds were constructed around the trickling filter. The difference between calculated mass inputs and outputs was considered to be the mass of compound removed by biodegradation. Following tabulation of the removal of the compounds by the principal fate mechanisms of stripping, effluent pass-through and biodegradation, biodegradation rate coefficients were estimated using TOXCHEM+ for each test. The estimated biorate coefficients were compared to those used in the software database. The biorate coefficients of certain of the compounds were very similar in magnitude to the model defaults (ethanol, acetonitrile, methyl ethyl ketone, hexane), while methanol was substantially lower, and toluene, triethylamine and i-propanol generally higher. The redox environment of the trickling filter may be one reason why some estimated biorate coefficients deviate from the default aerobic biorate coefficient in the software database.Using TOXCHEM+, the emission rates from the wastewater facility using only a single default biodegradation rate for each compound, and using two biodegradation rate coefficients (calibrated coefficient for the roughing filter, model default coefficient for the activated sludge unit) were compared. The results showed that estimated emission rates were substantially lower (e.g. 73 kg/d (161 lb/d)) using the dual biodegradation rate coefficient model, mostly due to reduced emissions of toluene. The model also showed that the majority of the compound load was removed by the trickling filters, before the activated sludge tanks had any opportunity for removal. The study indicated that model accuracy can be improved by calibration of biodegradation rates for HAPs and VOCs, and in using the dual biodegradation rate coefficient model for estimating emissions. Greater accuracy of emissions predictions makes the model a WEFTEC 2002 better tool for wastewater treatment facility emission evaluations and for decision-making in response to the Pharmaceutical MACT regulations.
KEY WORDSTrickling filter, Emissions, HAPs, Pharmaceutical, Model BACKGROUND A sampling program, consisting of six test...
