Topic 10 TMDLs

1
Topic 10 - TMDLs

• Need?
• Section 303(d) CWA
• TMDL?
• Litigation
• EPA Actions
• Overview of TMDL Regulations
• Example Neuse River, NC

http//www.epa.gov/owow/tmdl/overviewfs.html
2
1. The Need

• gt40 assessed waters below wq std
• gt20,000 indvd river segments, lakes, estuaries
  impaired
• 300,000 mi rivers shorelines
• 5 million ac lakes
• Sediments, excess nutrients, harmful
  microorganisms
• 218 million people lt10 mi from impaired waters

3
2. Section 303(d) of 1972 CWA

• States, territories, authorized tribes develop
  lists of impaired waters
• gt10 of samples violate WQ std
• No specs on samples
• Representative of true WQ distribution?
• Sampling uncertainty
• Establish priority rankings for listings
• Develop TMDLs for these waters

4
3. TMDL

• Max amt of pollutant waterbody can receive
  still meet WQ std
• Allocates pollutant loadings among pt non-pt
  sources
• EPA approval/oversight
• EPA regulations 1985, 1992
• Implement section 303(d) CWA TMDL provisions

5
4. Litigation

• TMDLs required since CWA 1972
• Few developed to date
• Legal action from citizen groups to implement
  TMDLs
• 40 legal actions in 38 states
• EPA under court order to establish TMDLs

6
5. EPA Actions to Implement TMDLs

• 1996 FAC
• 20 reps from ag, forestry, env advocacy,
  industry, state-, local-, tribal govts
• 1998 recommendations
• Public comment 100s of meetings, 34,000 formal
  comments
• Results in 2000 ruling current TMDL program

7
6. Current TMDL Program

• States, territories, tribes
• Impaired waters on 303(d) list until WQ stds
  attained
• Listing April 1, even years
• Based on monitoring evaluated data
• EPA has 30 days to approve listing TMDL
• Expeditious scheduling 8-13 yrs

8
Oregon 303(d) List
http//www.deq.state.or.us/wq/303dlist/303dpage.ht
m
9
7. Example Neuse River, NC

• TMDLs based on WQ models
• Most models fail to include uncertainty in
  predictions unsuitable for TMDLs
• TMDL devp requires predicting future compliance
  after a pollutant-load reduction (model)

10
Neuse continued

• Most models deterministic
• Outputs determined by inputs predictions
• Single value in space time
• Uncertainty of predictions!!!
• Natural variation
• Mis-specified conditions
• Measurement error
• Most models under-represent dynamics
• Miss high low extremes key for TMDL

11
Building a TMDL Model

• Mass balance model for a pollutant
• Accumulation Loading Outflow Reaction
• Loading all inputs (mass/time)
• Outflow all outputs (mass/time)
• Reaction all internal processes that add/remove
  pollutant (mass/time)

12
Building a TMDL Model

• Accumulation Loading Outflow Reaction
• V dc/dt W(t) Qc kVc
• where, c response variable (mass/vol)
• W(t) indp input variable
• V, Q, k coefficients of model

outputs
internal removal/addition
inputs
13
Building a TMDL Model

• Accumulation Loading Outflow Reaction
• V dc/dt W(t) Qc kVc
• c g (X, ß) ?
• where g math function(s) relating c, conc to
• X set of all input variables
• ß model parameters
• ? error term (accommodates predicted vs
  observed discrepancies)

14
(No Transcript)
15
(No Transcript)
16
(No Transcript)