Transcript ppt 3.2mb
Jennie Henthorn Henthorn Environmental Services LLC Permit limits are in terms of total recoverable aluminum Water quality criteria are in terms of dissolved aluminum Applying the dissolved Al criteria as a total Al effluent limits can be overly restrictive A translator is an attempt to fix the disparity between the dissolved criteria and total recoverable effluent limits “...the fraction of total recoverable metal in the downstream water that is dissolved” The Metals Translators Guidance Allows for higher permit limits that better reflect potential toxicity (usually, only a fraction of the total Al will be dissolved) Warm water (end of pipe) Trout water (end of pipe) Ave Mon (mg/l) Max Daily (mg/l) Ave Mon (mg/l) Max Daily (mg/l) 0.433 0.750 0.083 0.143 The translator is a fraction: Fd = Cd/Ct The effluent limits are divided by the translator to calculate the new limits Translators require a study with multiple samples The spreadsheets used to calculate translators are changing (and hopefully will continue to do so…) Study must be well designed Decisions on how to proceed can be evasive. Translator must be protective of critical flow this is not the necessarily the same as 7Q10 flow The amount of dissolved Al is dependent on parameters such as TSS, pH, hardness sampling should occur when the combination of parameters makes it so there is the most dissolved Al in stream this is “critical flow” or “design flow” But… no way to determine what these conditions are without a separate study – and results can be less than clear DEP has determined that the default “critical” flow is when precipitation induced discharges are not occurring. ◦ Low flow usually has low TSS = higher dissolved Al ◦ Simplifies preparation and review of a study plan Time/season of sampling Low flow is usually July- October in WV DEP now requires 20 samples collected no more frequently than once per two weeks. The time to collect samples can easily stretch to a year or more due to precipitation requirements. EPA guidance suggest sampling should occur at edge or beyond mixing zone But…in most circumstances for mining permits, no mixing zone is usually available (EOP limits) Often, the nearest downstream monitoring point used for instream sampling is used. Watershed approach may be more appropriate Take a sample or two – make some informed decisions Prepare a sampling plan and get DEP’s approval Remember, translators are not just for aluminum! Do you need to consider other parameters? Choose your path carefully, or you may fall flat on your face (or worse) Dissolved Al Total Al TSS Precipitation Hardness Others “A complete data set allows for more complete understanding of the environmental fate and transport processes and may result in a more accurate permit limit because of reduced variability and uncertainties.” EPA guidance Recall that, translator is determined by dividing the dissolved Al by the total Al this gives the translator for each sample ◦ fD=CD/CT This means that you have 10-20 different translators- how do you decide which one to use for determining permit limits??? Determine if translators are lognormally distributed (Shapiro-Wilks) Determine if translators are TSS-dependent using regression analysis ◦ If translators are lognormally distributed and are NOT TSS-dependent, use the geometric mean for permit limit caculations But what if the translators are not log-normal???? Attempt to make translators lognormal by transforming the data (arcsine square root) ◦ If translators are lognormal transformed and NOT TSS-dependent, use the transformed geometric mean for permit limit calculations OR Use the 95th percentile of the translators But what if the translators are TSS-dependent? Although sampled at low flow, translators will be used over a wide variety of flows and TSS concentrations Translators can be developed relating dissolved Al to other parameters, such as pH or hardness TSS usually has the largest influence and therefore is the focus of EPA guidance Develop regression equation relating TSS and translator Use “critical TSS” in regression equation to determine appropriate translator ◦ The critical TSS is representative of a TSS that would give the highest dissolved Al in the stream ◦ The EPA guidance gives little information on how to determine critical TSS OR Use the 95th percentile of the translators Use DEP’s current spreadsheet to calculate effluent limits! Requires 20 samples exactly Determines if data is lognormal using Shapiro-Wilks Determines if data are TSS-Dependent Requires rain data to determine critical flow Only accepts data taken during critical flow ◦ If a TSS-dependent equation is being developed, critical flow should not be used ◦ Uses only one method for determining “critical” flow – other methods include TSS concentration, and 25th percentile of low-flow Does not have a regression equation for TSSdependent translators ◦ If translators are TSS-dependent, it defaults to using the 95th percentile If data are not lognormal, spreadsheet does not determine if data are lognormal transformed The combination of these problems amounts to using the 95th percentile as a default, when other values may be equally defensible (regression equation, transformed geometric mean). Addition of lognormal transformed calculation Addition of regression analysis for TSSdependent data sets Potentially the addition of using the arithmetic mean when data are normally distributed (and not TSS-dependent) These changes will determine translators that more accurately represent the stream!