I UNITED STATES ENVIRONMENTAL PROTECTION AGENCY REGION 8 999 18w STREET SUITE 300 DEPNER, CO 80202-2406 h~:llwwwlwwwe~.govlregion08
-
Ref 8EPR-EP
.
June 4. 2004
Mark T.P a e r , Director Water Quality Control Division Colorado Depa&iiiii-t oT Public Ht&lh-6a &-=mknt 4300 Chary Creek Drive South Denver, Colorado 80246-15 3 0
RE: -
Dear Mr. Pifher:
,
I
-
-
~P;IDL~ p p r o v a-l ~ i t r h & South Platte River, Segment 14
We have completed our review ofthe total maximum daily load (TMDL,] as submitted by your o a c e for the South Platte River, Segment 14. T$e TMDL is included in the document entitled; Total Maximum Daily Load Assessment NiMte. South Platte River. Segment 14. Bowles Ave.. to the Burlington Ditch Diversion. Arapahoe and Denver Counties. Colorado, (CoIorado Department of Public Health and ~ n v i r o d n tApril , 2 1,20,04). This document was submitted to us for review and approval in correspondence dated April 22,2004 and signed by you. In accordance with the Clean Water Act (33 U.S.C. 1251 et. seq.), we approve all aspects of the TMDL as developed for the water qudity limited waterbody as described in Section 303($)(1). Enclosure 1 to this letter provides a mammy of the elements of the TMDL and Encl2 provides details of our review of the TMDL. r
1
Based on our review, we feel the s e i h e TMDL elements listed in Enclosure2 adequately address the pollutant of con=& taking into consideration seasonal variation and a margin of safety. In approving t h i s TMDL, EPA a E m s that the TMDL has been established at a 1eveI necessary to attain and maintainthe applicable water quality standards and has the necessary components of an approvable W L .
I
Thank you for your submittal. If you have any questions concerning this approval, feel free to contact Kathryn Hernandez of my staff at 303/3112-6101.
Max H. Dodson Assistant Regional Administrator Office of posyjtems Protection and Remediation
TOTAL MAXIMUM DAILY LOAD ASSESSMENT
Nitrate South Platte River, Segment 14 Bowles Ave. to the Burlington Ditch Diversion Arapahoe and Denver Counties, Colorado April 21, 2004 TMDL SUMMARY Waterbody Name/Segment Number
Mainstem of the South Platte River from Bowles Ave. to the Burlington Ditch Headgate COSPUS14
Pollutant/Condition Addressed
Nitrate (NO3- ) (protection of water supply use)
Affected Portion of Segment
Points of attainment – Allen Diversion and the Burlington Ditch Headgate
Use Classification/Waterbody Designation
Water Supply, Agriculture Aquatic life Warm 1, Recreation 1a
Waterbody Antidegradation Designation
undesignated (reviewable)
Water Quality Target
Assure the nitrate concentrations at the points of attainment do not exceed 10 mg/L through implementation of controls on various nitrogen constituents
TMDL Goal
Attain Colorado water quality nitrate standards at the Allen Diversion and the Burlington Ditch Headgate
EXECUTIVE SUMMARY Segment 14 of the South Platte River has been identified as water-quality limited for nitrate based on predictive modeling. Low-flow modeling indicates that municipal wastewater treatment facilities are the primary point-source dischargers of nitrate to Segment 14. This TMDL derives wasteload allocations for nitrogen series pollutants that will ensure attainment of Colorado water quality standards for nitrate at the Allen Diversion and Burlington Ditch Headgate. Stormwater runoff from nonpoint sources does not contribute significantly to the nitrate impairment. Some localized groundwater may be affecting the overall water quality. Water quality monitoring is necessary to verify that the TMDL requirements result in attainment of the standards. This TMDL should be reviewed when municipal dischargers propose plant expansions beyond the conditions utilized in this modeling effort, new wastewater treatment plants are proposed on the main stem or tributaries, or when assumptions included in this TMDL assessment are shown to be no longer appropriate.
So. Platte seg 14 Nitrate TMDL
Final
TABLE OF CONTENTS I. II. III. IV. V.
VI.
VII.
VIII. IX.
X.
XI.
INTRODUCTION........................................................................................................1 WATER QUALITY STANDARDS ...........................................................................3 PROBLEM IDENTIFICATION ................................................................................3 WATER-QUALITY GOALS .....................................................................................5 ANALYSIS OF POLLUTANT SOURCES ...............................................................5 Sampling Relevant to the TMDL ..........................................................................5 Identification of Sources .......................................................................................6 TECHNICAL ANALYSIS ..........................................................................................9 Hydrology ..........................................................................................................10 Nitrification .........................................................................................................11 Denitrification .....................................................................................................14 Source Assessment and Modeling Methodology ................................................14 TMDL ALLOCATION .............................................................................................17 Allocation Methodology ....................................................................................17 TMDL Scenario Assumptions ............................................................................17 TMDL and Allocations .......................................................................................19 Waste Load Allocation for Sources to be Controlled .........................................21 Load Allocation Among Sources .......................................................................22 Margin of Safety .................................................................................................22 Implementation ...................................................................................................23 Post-Implementation Monitoring ........................................................................23 PUBLIC INVOLVEMENT .......................................................................................24 RESPONSIVENESS SUMMARY - First Public Comment Draft ........................25 Post-Public Notice Process .................................................................................25 Response to Technical Comments and Explanation of Changes to the Model ..................................................................................25 Response to Legal Comments .............................................................................30 RESPONSIVENESS SUMMARY - Second Public Comment Draft ....................31 Post-Public Notice Process: December 2001 - May 2003 ..................................31 Response to Comments and Explanation of Changes to the Model ..................................................................................31 RESPONSIVENESS SUMMARY - Third Public Comment Draft ......................38 Response to Comments and Explanation of Changes to the Model ...................38 Alternative TMDLs - Alternate Effluent Flow for L/E: 42 and 34 MGD ..........42 LIST OF TABLES
Table 1. Table 2. Table 3. Table 4. Table 5.
Operating conditions for major point-source dischargers. .............................................7 Ammonia and nitrate in water sources...........................................................................9 Acute (1-day) low flows and modeled flows for selected locations on Segment 14 ...........................................................................................13 Empirically determined rates (day-1, base e, first order) of ammonia loss and nitrate loss for two reaches of the South Platte River .......................................14 Ungaged surface water sources for Segment 14. .........................................................15
Colorado WQCD
i
April 21 2004
So. Platte seg 14 Nitrate TMDL
Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. Table 12 Table 13. Table 14. Table 15 Table 16 Table 17. Table 18 Table 19. Table 20. Table 21
Final
Gages used in the analysis of ungaged flows in Segments 6c and 14. ........................15 Reaches used in the residual flow analysis .................................................................16 Effect of hypothetical changes in nitrate concentrations of seepage ...........................16 L/E at 50 MGD: Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia. ................................................................................................19 L/E at 50 MGD: Summary of flows and concentrations of nitrate N to attain water quality targets ..................................................................................................20 L/E at 50 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN ................................................................................................................21 Comparison between Measured Conditions and Modeled Conditions .......................29 Littleton/Englewood WWTP Daily influent flow, nitrate+nitrite and mean daily flow at the Denver gage for the days with the highest daily influent flow ..............34 Littleton/Englewood WWTP Daily influent flow, nitrate+nitrite and mean daily flow at the Denver gage for the days with the highest nitrate+nitrite .............35 Critical Low Flow (1E3) and Historic Lowest Flows in cfs at the Denver Gage (USGS No 06714000)........................................................................41 For L/E at 34 MGD: Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia. .........................................................................................42 L/E at 34 MGD: Summary of flows and concentrations of nitrate N to attain water quality targets as predicted by the TMDL model. ..........................................43 L/E at 34 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN for sources using TMDL modeling assumption. .......................................44 For L/E at 42 MGD: Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia. ................................................................................................45 L/E at 42 MGD: Summary of flows and concentrations of nitrate N to attain water quality targets as predicted by the TMDL model..................................45 L/E at 42 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN for sources using TMDL modeling assumption. ...................................46 LIST OF FIGURES
Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10
Map of Segment 14, Segment 6 and downstream Segment 15......................................2 South Platte River Segment 14, Arapahoe and Denver Counties, Colorado. ................4 Burlington Ditch Headgate – A point of diversion on Segment 14 ...............................5 South Platte River downstream of the Littleton/Englewood Wastewater Treatment Plant ..........................................................................................6 Centennial Water and Sanitation District Treatment Facility ........................................7 City of Glendale Wastewater Treatment Facility ..........................................................8 Graphic representation of low flows (reset on, off and modified) for Segment 14. .........................................................................................................12 Frequency of discharged ammonia for Littleton Englewood WWTP ........................18 Littleton/Englewood WWTP Flow and Nitrate ...........................................................34 Nitrate levels at the Burlington Headgate and Flows at the Denver Gage, Sept. 1998 to Sept 2003. ..............................................................................................40
Colorado WQCD
ii
April 21 2004
So. Platte seg 14 Nitrate TMDL
I.
Final
INTRODUCTION
Section 303(d) of the federal Clean Water Act requires states to identify waterbodies or stream segments where existing effluent limitations are not stringent enough to implement applicable water quality standards. Water-quality limited segments currently identified in Colorado are included on the 1998 303(d) list. Water-quality limited segments are those in which one or more classification or standard is not, or may not be fully achieved. The State is required to develop a Total Maximum Daily Load (“TMDL”) assessment for every segment and parameter that is listed. The TMDL describes loading limits that will ensure attainment of the stream standard, with specific reference to a quantification of the amount of pollutants that a segment can assimilate without exceeding water quality standards. The TMDL document also apportions the allowable pollutant load among multiple pollutant sources. The TMDL is comprised of the Load Allocation (LA), which is that portion of the pollutant load attributed to background or to nonpoint sources, the Waste Load Allocation (WLA) that is that portion of the pollutant load associated with point-source discharges, and the Margin of Safety (MOS). The TMDL may also recognize actions other than waste treatment that may help achieve attainment of standards where there is assurance of implementation. The Urban South Platte Watershed is part of the South Platte Middle Basin - Hydrologic Unit Codes (HUC) 10190002 and 10190003. South Platte River Segment 14, designated as COSPUS14, is located in Arapahoe and Denver Counties, within the Urban South Platte Watershed (Figure 1). This TMDL assessment was accomplished, in part, through the use of a water-quality model. The modeling includes Segment 6c below Chatfield Dam and all of Segment 14. Segment 6c of the South Platte River extends from below Chatfield Reservoir to Bowles Avenue. Segment 6c below Chatfield Dam is classified by the Urban Drainage and Flood Control District's (UDFCD) Geomorphic Assessment Survey (Stevens, 1996) as rural, suburban, and engineered. The reach of Segment 6c from Chatfield Dam to a point near Highway C-470 is classified by UDFCD as rural. The River meanders through the floodplain and is free to move laterally. The River here is wide and shallow with sandy substrate as well as gravel or cobble bars that are vegetated with pioneer weeds, grasses, and willows. Wildlife is abundant in this portion of the segment. Segment 6c from a point near Highway C-470 to the old Brown Ditch diversion is considered suburban. The suburban river reflects human intervention in the construction of bridges, weirs, grade-control structures, intakes and storm sewer outfalls. Some of the river meander loops have been cut off. Eroding banks have been stabilized with rubble or riprap and low banks have been filled with excavated materials. There are fewer bars and less native vegetation than in the rural area, but wildlife is present. The River is wide and shallow in some places and has adjacent levees created by gravel mining.
Colorado WQCD
1
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Figure 1. Map of Segment 14 and the upstream adjoining portion of Segment 6 below Chatfield Dam and downstream Segment 15.
From the old Brown Ditch diversion to Oxford Avenue, the Segment is considered engineered. The River channel has been realigned and straightened with riprap along the banks. The bed is flat and has only a few low bars. The width is mostly uniform. Pioneer vegetation is slowly reappearing in this reach along the few bars and in the riprap. Segment 14 begins at Bowles Avenue in Arapahoe County and flows north. The UDFCD considers the reach between Oxford and the Burlington Ditch as urban. The urban section of the reach has been straightened and realigned; it has a narrow and deep channel. Except during low flow, there are few exposed bars. The riparian vegetation is much less extensive than in the suburban and rural areas. Many bank and channel structures have been constructed in Segment 14 to manage the changing elevation of the riverbed. The River flows by Englewood and Overland golf courses prior to entering commercial and highly industrial areas of metropolitan Denver. Bear Creek, which enters adjacent to the Englewood Golf Course, and Cherry Creek, which enters in the vicinity of 14th Street, influence the segment. Water diversions take place at the Allen Diversion, the Xcel Arapahoe power plant (previously known as Public Service Arapahoe power plant), Overland Golf Course, the Farmers and Gardeners Ditch, and the Burlington Ditch (see Figure 2). The Allen Diversion and the Burlington Ditch are the only municipal water supply intakes in the segment. The Burlington Ditch Headgate marks the end of Segment 14.
Colorado WQCD
2
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Land use in the Urban South Platte Watershed is approximately 85% residential and open space; 15% is commercial and industrial. Industrial and commercial areas are adjacent to the river. The impervious nature of the commercial and industrial zone is not relevant for this TMDL, because stormwater dilutes rather than increases the concentrations of nitrate in Segment 14. II.
WATER QUALITY STANDARDS
The Colorado Water Quality Control Commission (“WQCC”) has adopted a water-quality standard for nitrate of 10 mg/L (as nitrogen) at the point of diversion to a municipal drinking water supply. The standard is a one-day standard. The diversion points on Segment 14 are the Allen Diversion, where the river is diverted for municipal supply use, and the Burlington Ditch Headgate, where the river is diverted for agricultural and municipal water supply uses. Surface waters classified for agricultural use are suitable or intended to become suitable for irrigation of crops usually grown in Colorado and must not pose a hazard as drinking water for livestock. The nitrate standard for agriculture is 100 mg/L (as nitrogen). There are no nitrate standards in place for aquatic life or recreation uses. III.
PROBLEM IDENTIFICATION
Nitrate is a constituent of concern in the South Platte Urban Watershed. At low-flow conditions the concentrations of nitrate at the Burlington Ditch Headgate have exceeded the standard. Predictive modeling conducted under Phase II of the DRCOG TMDL Steering Committee’s analysis of water quality issues identified nitrate standard exceedances at the Burlington Ditch Headgate under some low flow and high wastewater treatment plant (“WWTP”) discharge scenarios. Because of these exceedances, Segment 14 is on Colorado’s 2002 303d List for nitrate. It is identified as a High priority for Water Quality Control Division (“Division”) resources and was targeted for early completion. Segment 14 is also included on the 2002 303(d) list for E. coli and fecal coliform. These pollutants will be assessed under separate evaluations. During high flow and even moderate flows, monitoring and modeling have shown that nitrate concentrations in the river are well below the nitrate standard; this includes both storm events and higher flows due to releases from the upstream reservoirs. During low flows, the nonpoint sources, groundwater seepage, and other ungaged water sources are not major contributors to the nitrate loading in this segment, but are significant in the dilution of nitrate from point-sources and must be considered in the low-flow modeling. Influences on water quality for Segment 14 include all water draining to the South Platte between Chatfield Dam and the Burlington Ditch Headgate. Segment 6c of the South Platte River is immediately upstream of Segment 14, and therefore is relevant to Segment 14. The major tributaries to Segment 14, Bear Creek and Cherry Creek, also are important for the purposes of this nitrate TMDL assessment. Measured levels of nitrate in these tributaries are very low.
Colorado WQCD
3
April 21 2004
So. Platte seg 14 Nitrate TMDL
Figure 2.
Colorado WQCD
Final
South Platte River Segment 14, Arapahoe and Denver Counties, Colorado.
4
April 21 2004
So. Platte seg 14 Nitrate TMDL
IV.
Final
WATER-QUALITY GOALS
The goal of this TMDL assessment is to assure attainment of the Colorado water quality standard for nitrate of 10 mg/L at the diversions to drinking water supplies in Segment 14. Process controls at municipal wastewater treatment facilities, enforced through effluent permit limits, should result in attainment of standards.
Figure 3. Burlington Ditch Headgate – A point of diversion on Segment 14
V.
ANALYSIS OF POLLUTANT SOURCES
As part of the TMDL assessment, all pollutant sources must be considered and contributions from significant sources must be quantified. From this information, the TMDL model can show how much pollutant loading must be reduced in order to meet water quality standards under specific conditions. Pollutant sources include discrete discharges of a pollutant from a pipe or other discharge structure (point-sources), ungaged surface discharges through a pipe (dry weather stormwater discharges), or diffuse discharges across a broad reach of stream (nonpoint sources). Sampling Relevant to the TMDL An intensive sampling effort was initiated in the fall of 1998 in support of calibration and verification of the nitrate model. The South Platte Coalition for Urban River Evaluation (“SP CURE”) commenced a sampling program focusing on nitrate, ammonia, and other relevant water-quality variables in the reach of the South Platte River extending from Chatfield Dam to the City of Brighton. The cooperative efforts of eight sampling teams and six laboratories provided coordinated data at a frequency exceeding that of previous monitoring efforts. This coordinated program provided the modeling team with a biweekly assessment of the South Platte and its tributaries. SP CURE members in support of this TMDL assessment sampled the following sites: SPR upstream of Centennial Centennial Effluent
Colorado WQCD
SPR at C-470 SPR at Mineral Ave.
5
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
SPR at Union Ave. SPR upstream of Bear Creek Bear Creek at gage Bear Creek at the mouth Little Dry Creek Harvard Gulch SPR at Dartmouth
L/E Effluent SPR at Evans PSC Arapahoe Outfall Big Dry Creek Glendale Effluent Cherry Creek at Champa Above the Burlington Ditch
Special studies were undertaken by South Platte CURE as part of the TMDL assessment for use in the model. These studies included: 1) Special 24- and 48-hour studies (Centennial and Littleton/Englewood Wastewater Treatment Plant [L/E WWTP]) 2) Assessment of sample-site homogeneity (Centennial and L/E WWTP) 3) Estimation of time of travel (Metro District) 4) Estimation of ungaged flows (Centennial, L/E WWTP and DDEH) 5) Waste treatment plant effluent study (L/E WWTP) 6) Measurement of river temperature (Centennial and L/E WWTP) 7) Monitoring of groundwater quality (DDEH). Identification of Sources Nitrate enters the South Platte River from a wide variety of sources. Sources vary greatly in magnitude; the largest sources are wastewater treatment facilities. Concentrations of nitrate and ammonia, as well as nitrification rates and temperature, are important considerations for nitrate modeling. Significant Point-Source Dischargers to Segment 14 - The primary source of nitrate at the Allen Diversion is the Centennial Wastewater Treatment Plant. At the Burlington Ditch Headgate, the effluent from the L/E WWTP is the primary source of nitrate. The current operating conditions for the major point-source dischargers are outlined in Table 1.
Figure 4. South Platte River downstream of the Littleton/Englewood Wastewater Treatment Plant
Colorado WQCD
6
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Other Permitted Dischargers to Segment 14 - Many permitted facilities are currently discharging into Segment 14, but the nitrate loads from most of these sources do not significantly affect the concentration of nitrate in the river. These are predominantly facilities engaged in construction dewatering or other similar small quantity discharges. Contributions from Dischargers outside Segment 14 – The main contributors of nitrate are wastewater treatment facilities. Centennial WWTP discharges to Marcy Gulch a tributary to Segment 6c approximately six miles upstream of the Allen Diversion. The contribution of ammonia and nitrate from Centennial can affect the concentration of nitrate at the Allen Diversion, which is located in Segment 14. (While the discharge from Centennial does not currently cause an exceedence of the nitrate standard at the Allen Diversion, there is some possibility for such an exceedence under future conditions if there were no limits for ammonia or nitrate for the Centennial discharge.) Centennial’s effluent has only a minor effect on the concentration of nitrate at the Burlington Ditch Headgate. Glendale WWTP discharges to Cherry Creek (Cherry Creek, Segment 3), about five miles upstream of the confluence with the South Platte River and approximately nine miles upstream of the Burlington Ditch Headgate. As a smaller discharger, Glendale has a minimal effect on the concentration of nitrate at the Burlington Ditch Headgate. Wastewater Treatment Facility Centennial Littleton/Englewood Glendale 1
Current Monthly Average Discharge (MGD) 5.0 28.3 0.8
Maximum Monthly Permitted Design Capacity (MGD) 8.5 36.3 [50.0]1 2.0
Littleton/Englewood plans to expand to 50.0 MGD
Table 1. Operating conditions for major point-source dischargers.
Figure 5. Centennial Water and Sanitation District Treatment Facility
Colorado WQCD
7
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Figure 6. City of Glendale Wastewater Treatment Facility
Contributions from Other Water Sources - Modeling requires specific assignment of ammonia and nitrate concentrations at low flow for all water sources. Flow residuals were calculated by the difference between simultaneous flow measurements at pairs of stream flow gages. Flow residuals include both seepage and ungaged flows from small tributaries. Dry weather contributions of small tributaries were estimated from field data and a relationship between drainage area and flow. The contributions from the small tributaries enter the South Platte at discreet locations and are included in the model in the appropriate order. The residual flow that remains after adjusting for the contributions from ungaged tributaries is called “seepage” and is assumed to be primarily ground water. Seepage inflow is distributed evenly in the model along the appropriate reach. In addition, there are five specific point sources that contribute water as well as nitrate, as outlined in Table 2. Xcel Arapahoe withdraws water from the river for pass-through cooling purposes only. The model accounts for the withdrawal and return flow. The volume of water returned to the river that is reduced by 50% during cooling; the nitrogen load is unchanged. Contaminated Groundwater Sites - The seepage of groundwater during low flow has a measurable effect on the concentration of nitrate in Segment 14. Over most of Segment 14, background groundwater nitrate concentrations are estimated to average between 2 and 3 mg/L of nitrate (as nitrogen). The data suggest that there are localized sources of groundwater contamination where nitrate concentrations range from 10 -100 mg/L; with a median concentration of approximately 12 mg/L. The areas of contamination have not been fully characterized, but contributions have been estimated in order to address concentrations and flow rates. As demonstrated by the monitoring data, these sources are small in volume and currently do not significantly influence the nitrate concentration at the points of attainment. Stormwater - The stormwater contribution to nitrate in Segment 14 is not significant based on concentration. Stormwater runoff typically contains less than 1 mg/L nitrate (as nitrogen). The increase in stream flow during a storm event decreases the overall nitrate concentration in the river, thereby holding nitrate concentrations well below the stream standard. Recent data indicates potential exceedence of the nitrate standard is a low-flow concern. Discharge from
Colorado WQCD
8
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
stormwater outfalls during dry-weather can be groundwater seepage due to infiltration or from sumps, drainage tiles, culverts and gulches that have been enclosed, or other urban contributions. For the purpose of this TMDL, these flows are being treated as ungaged flows. Atmospheric Deposition - Atmospheric deposition of nitrate is an uncontrollable source of contamination. The loading resulting from atmospheric deposition is not significant for this TMDL assessment, and is inherent in the background concentrations of nitrate for the modeling on Segment 14. Ammonia-Nitrogen mg/L Source
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Chatfield Res. Bear Creek Res.
0.00 0.08
0.00 0.12
0.00 0.08
0.00 0.10
0.00 0.03
0.00 0.01
0.00 0.03
0.00 0.05
0.00 0.06
0.00 0.16
0.00 0.11
0.00 0.05
Bear Creek at Mouth
0.22
0.18
0.12
0.16
0.20
0.31
0.14
0.33
0.17
0.15
0.11
0.14
Little Dry Creek
0.15
0.18
0.24
0.24
0.23
0.30
0.19
0.25
0.20
0.16
0.13
0.13
Cherry Creek Res.
0.04
0.04
0.04
0.05
0.05
0.05
0.12
0.12
0.12
0.02
0.02
0.02
Xcel Arapahoe
1.15
0.30
0.34
0.35
0.36
0.32
0.38
0.35
0.45
0.50
0.80
Source Chatfield Res.
Jan 0.10
Feb 0.09
Mar 0.02
Apr 0.01
May 0.12
Jun 0.07
Jul 0.01
Aug 0.03
Sep 0.03
Oct 0.00
Nov 0.00
Dec 0.04
Bear Creek Res.
0.92
0.88
0.51
0.55
0.50
0.18
0.26
0.44
0.17
0.48
0.64
0.95
Bear Creek at Mouth
1.25
1.28
1.33
0.89
0.73
0.46
0.36
0.33
0.52
1.04
1.13
0.94
Little Dry Creek
1.78
1.35
1.10
1.10
1.04
1.08
0.95
1.07
1.24
1.33
1.69
1.65
Cherry Creek Res.
0.05
0.05
0.05
0.01
0.01
0.01
0.05
0.05
0.05
0.80
0.80
0.80
Xcel Arapahoe*
6.2
6.7
5.1
4.0
5.0
2.9
4.1
4.5
0.40
Nitrate-Nitrogen mg/L
2.9
6.2
6.6
5.8
*Pass-through cooling with discharge just below Littleton/Englewood WWTP- values from SP CURE monitoring.
Table 2. Ammonia and nitrate in water sources.
VI.
TECHNICAL ANALYSIS
The TMDL assessment accounts for all influences on loads and concentrations of nitrate based on monitoring data, effluent characteristics, or other pertinent information. The goal of the modeling is to identify the point source allocations that would ensure attainment of the nitrate standard at the Burlington Ditch Headgate and Allen Diversion. Nitrate and ammonia concentrations, flows, and temperature must be evaluated for each source. Time of travel, nitrification (ammonia loss rate) and denitrification (nitrate loss rate) must be evaluated for the river itself. Nitrite (NO2-) concentrations were evaluated and proven to be inconsequential for this assessment. Because nitrite concentrations are very low throughout Segment 14, nitrite was not included in the TMDL assessment.
Colorado WQCD
9
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
The initial modeling effort was part of the TMDL screening process initiated under Phase II of the DRCOG TMDL Steering Committee. Much of the technical information detailed in this section is extracted from the modeling report entitled "Total Maximum Daily Load (TMDL) for Nitrate in the Urban Region of the South Platte River, with Emphasis on Segment 14," prepared by Lewis and Saunders, March 3, 2000, revised October 8, 2001 and June 12, 2003. The report is included as an attachment. Hydrology Low-Flow Analysis: Critical nitrate levels in Segment 14 occur during low-flow conditions when there is little river water to dilute the sources of nitrogen. The Division uses DFLOW4 software to develop critical low-flow estimates for permitting purposes. The acute critical low flow is the empirically based 1-day low flow with an average 1-in-3 year recurrence interval. DFLOW4 was used to develop low-flow estimates for the South Platte River below Chatfield, South Platte River above L/E WWTP, South Platte River above Xcel Arapahoe and South Platte River above Xcel Zuni. For each of these locations, the monthly acute DFLOW values were obtained using the gage records for the interval October 1, 1990 to September 30, 2000. The low-flow values for the tributaries and diversions were estimated by the method of differences, to match the DFLOW values in the mainstem in terms of water balance. Table 3 gives the acute low-flow values developed for this analysis. Flow Accumulation/Reset: Issuance of a CDPS permit for a discharger typically involves the assumption that the facility being permitted will be discharging at design capacity. This assumption is problematic for modeling of multiple dischargers in the sense that growth occurs at different rates within different facilities’ service areas, and therefore it is unlikely that all facilities will reach capacity at the same time. In addition, the accumulation of full capacity discharges from all points in the downstream direction carries the underlying assumption that none of this additional water would be removed for consumptive use. CDPS permitting for a discharger also involves the assumption that the available dilution at a point just above the discharge will be determined by DFLOW analysis based on historical flows. These historical flows do not include the hypothetical full discharge capacity of all upstream dischargers. Thus the two CPDS permitting practices (full capacity discharge and upstream historical low flow) are conflicting when they are implemented in a multi-discharger model for TMDL purposes. The assumption of historical low flows above each discharger can be incorporated into the model by resetting the flow of the river to the historical low flow above each discharger, thus removing the increment between historic and capacity discharges just above each new discharge. This approach removes constituent mass as well as water. (The “reset” scenario will produce different results than the accumulation of design flows in the downstream directions.) Hypothetically, either approach could produce more extreme results than the other, depending on the constituent under consideration and the characteristics of the river, such as travel time, and the discharges. A third approach, the “Modified Reset”, adjusts flow of the river back to the historical low flow
Colorado WQCD
10
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
only above the L/E discharge point. Figure 7 shows the modeled flows along the river for January under the three scenarios: Reset On, Modified Reset, and Reset Off. Table 3, lowest panel, provides a numeric comparison of the three flow scenarios. Comparison of the resulting flow scenarios and the modeling results shows that the Modified Reset option provides results between the Off and On approaches and has a more realistic flow scenario. The Reset On option removes flow (and nitrogen) just above L/E’s discharge, then removes much of L/E’s flow at Public Service Arapahoe just 0.4 mile down stream of L/E’s discharge. Using the modified reset approach allows L/E’s full design capacity to travel down the river to the Burlington headgate. The Modified Reset approach was used in the TMDL calculations. Nitrification Nitrification is the process by which ammonia, in the presence of oxygen, is converted by microbes to nitrate. This process occurs naturally in streams. Rates of nitrification vary, however, according to the amount of ammonia, availability of substrate for attachment of the specialized nitrifying bacteria, amount of organic matter, temperature, and other factors. Because nitrification reduces the concentration of ammonia and increases the concentration of nitrate, its rate is important for the modeling of either nitrate or ammonia. In Segment 14, the rate of nitrification can be estimated from the disappearance of ammonia. Other processes, including ammonia uptake by algae and ammonia regeneration from decomposition, are comparatively small in relation to nitrification rates when the ammonia concentration is high. Nitrification rates were estimated for the main stem on the basis of change in ammonia load between upstream and downstream locations. The rate of nitrification in the South Platte between the outfalls of the Centennial and L/E WWTPs cannot be determined at this time. Precise estimation of the current rate was difficult because measured ammonia concentrations were so low. Since the modeled future conditions include much higher ambient ammonia concentrations, it is reasonable to expect that nitrification rates in the future will increase. The Division chose to use the Colorado Ammonia Model default nitrification rate of 6.0 day-1 for reaches of the system where there is insufficient data to generate a site-specific rate. Between the L/E outfall and the Burlington Headgate, the rate of nitrification could be characterized because of the relatively high concentrations of ammonia. In addition, the nitrification rate showed a distinct relationship with flow rates; at low flow, the conversion to nitrate is lower; at higher flows, the conversion rate is higher. This relationship is incorporated in the model. Table 4 summarizes the nitrification and denitrification rates for the South Platte River between L/E and the Burlington Ditch Headgate.
Colorado WQCD
11
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Segment 14 - January Centennial
Bear Cr L/E
Cherry Cr
175
Flow, cfs
150 125 100 75 50 25 0 0
3
6
9
12
15
18
21
River Miles below Chatfield Reservoir
Reset On
Segment 14 - January Centennial
Bear Cr L/E
Cherry Cr 10
175
Flow, cfs
150
8
125 100
6
75
4
50
2
25 0
0 0
3
6
9
12
15
18
21
River Miles below Chatfield Reservoir
Reset Off
Segment 14 - January
Flow, cfs
Centennial
Bear Cr L/E
Cherry Cr
175 150 125 100 75 50 25 0
10 8 6 4 2 0 0
3
Modified Reset
6
9
12
15
18
21
River Miles below Chatfield Reservoir
Figure 7. Graphic comparison January flows (reset on, off and modified)
Colorado WQCD
12
April 21 2004
So. Platte seg 14 Nitrate TMDL
Location
Final
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Acute low flows for tributaries and withdrawals (by difference), cfs DOW Fish Unit release
1.0
0.3
0.9
3.0
8.0
5.9
8.0
4.6
0.6
1.8
0.3
0.9
Allen Plant withdrawal
7.4
7.7
8.3
9.0
17.0
14.0
21.0
12.0
8.0
7.6
6.0
8.0
Bear Creek at mouth
19.0
17.8
15.0
17.7
28.0
13.0
7.0
10.0
12.2
17.9
18.6
24.6
Farmers & Gardeners withdrawal
13.5
13.5
16.6
13.8
13.7
14.0
13.0
11.0
7.0
7.0
11.0
12.4
Cherry Creek at mouth
10.9
8.2
8.2
8.0
21.0
13.0
13.0
13.0
11.0
12.0
8.0
11.0
S.P. abv Centennial
1.8
0.8
1.6
4.2
14.5
9.3
10.0
7.3
2.9
3.8
1.7
1.9
S.P. abv Bi-City (L/E WWTP)
26.0
27.0
25.0
28.0
60.0
58.0
30.0
33.0
20.0
27.0
31.0
39.0
S.P. abv Xcel Arapahoe
59.0
60.0
64.0
59.0
95.0
102.0
67.0
71.0
55.3
63.0
67.0
74.0
S.P. abv Xcel Zuni
62.0
63.0
65.0
61.2
97.0
106.0
71.0
76.0
61.2
67.0
69.0
76.0
Bear Creek Res. Release
12.0
10.8
9.0
10.7
24.0
8.0
2.0
3.0
4.2
8.9
10.6
17.5
Cherry Cr Res. Release
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
S.P. at Denver Gage
67.2
65.5
64.3
63.2
112.6
114.6
81.1
88.8
75.3
81.3
74.0
82.3
S.P. at Burlington
69.1
67.6
66.3
65.3
115.7
119.7
86.9
95.8
81.2
85.9
76.0
84.3
Acute low flows, (from DFLOW) cfs
1
Model flows - reset on (cfs)
Model flows 1 - reset off, flows accumulated (cfs) S.P. at Denver Gage
138.8
130.4
125.2
135.2
164.8
145.0
135.2
153.2
156.6
162.9
147.4
148.7
S.P. at Burlington
140.7
132.5
127.1
137.3
167.9
150.1
141.0
160.2
162.5
167.5
149.8
150.7
115.0
113.5
108.1
114.1
161.4
156.0
130.8
138.6
125.0
129.8
120.5
129.1
S.P. at Burlington 116.9 -------------------------v Littleton/Englewood at 50.0 MGD
115.6
110.1
116.3
164.4
161.1
136.6
145.6
130.9
134.4
122.8
131.1
1
Model flows - modified reset (cfs) S.P. at Denver Gage
Table 3. Acute (1-day) low flows and modeled flows for selected locations on Segment 14.
Colorado WQCD
13
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Denitrification Denitrification is the process by which microbes convert nitrate to nitrogen gas. This process occurs naturally whenever nitrate is present and oxygen is absent. In the South Platte River, denitrification takes place in the sand and gravel zone just below the surface (hyporheic zone) because this zone is often free of oxygen. Rates of denitrification can be estimated from the disappearance of nitrate, if nitrification is taken into account. The rate of denitrification is measurable in both reaches, but was substantially higher in the reach from L/E to the Burlington Ditch Headgate. For the upper reach, a median rate, with a temperature correction appropriate for each month, is used in modeling. For the lower reach, a flow / nitrate loss-rate relationship was developed and used in the model.
Month
Chatfield-Englewood Loss Rates2 Temperature1 Ammonia Nitrate
Englewood-Burlington Loss Rates3 Temperature1 Ammonia Nitrate
Jan Feb Mar
4.5 4.8 7.5
1.82 1.86 2.29
0.65 0.66 0.74
2.5 2.5 6.2
1.55 1.55 2.07
0.60 0.60 0.71
Apr May Jun Jul Aug Sep Oct Nov Dec
11.2 14.8 15.5 20.6 20.1 18.9 13.6 9.5 4.5
3.05 4.02 4.24 6.28 6.05 5.51 3.67 2.67 1.82
0.88 1.03 1.06 1.32 1.30 1.23 0.97 0.81 0.65
8.6 12.5 14.5 18.5 19.0 17.3 10.5 7.1 3.5
2.49 3.36 3.91 5.33 5.53 4.86 2.88 2.21 1.68
0.79 0.94 1.02 1.22 1.25 1.16 0.86 0.74 0.63
6.00 1.08
1.29 1.045
5.98 1.08
1.30 1.045
Rate at 20°C Theta 1
Temperatures shown here are for the upper end of each reach.
2
These rates also were applied to Cherry Creek and Bear Creek.
3
Using a flow rate of 120 cfs at the Denver Gage.
Table 4. Empirically determined rates (day-1, base e, first order) of ammonia loss and nitrate loss for two reaches of the South Platte River from Chatfield Reservoir to the Burlington Ditch Headgate, as determined by model calibration. Rates for the lower reach are flow dependent.
Source Assessment and Modeling Methodology All sources of nitrate are categorized for the purpose of modeling. Some sources have no significant effect on nitrate concentrations at the Burlington Ditch Headgate. The sources that were used in the model are as follows: Colorado WQCD
14
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Point-Source Dischargers to Segment 14 - The major municipal dischargers are treated as independent variables. Other Permitted Point-Source Dischargers to Segment 14 - The other point-source dischargers were not included in the model because their nitrate loads and/or the quantity of the discharge were minimal. Sensitivity analysis showed that these sources did not affect the final nitrate level at the diversions. Contributions from Tributaries and Upstream - Modeling requires assignment of ammonia and nitrate concentrations at low flow for all water sources, including tributaries. There are five sitespecific sources included in the model. The median values for all ungaged flows, including the portion that is dry-weather ungaged tributary flows, are included in Table 5.
Location
Discharge cfs
Ammonia N mg/L
--1
Little Dry Creek
3.1
West Harvard Gulch
0.3
0.014
2.55
Harvard Gulch
0.6
0.016
0.13
Sanderson Gulch
2.4
0.027
1.94
Lakewood Gulch All Others 1
--
Nitrate N mg/L
1
6.7
0.051
1.17
17.3
0.025
1.69
Monthly values are available for Little Cry Creek (see Table 2)
Table 5. Ungaged surface water sources for Segment 14.
Seepage - An analysis of discharge at paired gages, taking into account diversions, point-source discharges, and tributary flows support an estimate of the contributions from ungaged flows in distinct reaches of the river. Table 6 details the gages considered for the modeling of ungaged flows. Table 7 summarizes the reaches of Segments 6c and 14 and the intervening flows and diversions considered in modeling the contribution of ungaged flows.
Gage
Miles below Chatfield
Gage Number
0
PLACHACO
Above Union Avenue (Above Allen Plant)
6.79
06710245
Below Union Avenue (Below Allen Plant)
6.88
06710247
Englewood (Above L/E WWTP)
9.73
06711565
Denver (Above Burlington)
17.51
06714000
SP below Chatfield Reservoir
Table 6. Gages used in the analysis of ungaged flows in Segments 6c and 14.
Colorado WQCD
15
April 21 2004
So. Platte seg 14 Nitrate TMDL
Reach Boundaries
Final
Length of Length of Reach Record (miles)
Intervening Flows and Diversions
Flow Threshold (cfs)
Chatfield to above Union
6.79
4/12/892/5/96
Fish Hatchery, Ensor Wellfield
200
Chatfield to below Union
6.88
2/7/969/30/00
Fish Hatchery, Ensor Wellfield, Allen Filter Plant
120
Above Union to Englewood
3.08
2/1/83-
Allen Filter Plant, Bear Creek, Little Dry Creek
400
Below Union to Englewood
2.87
2/7/969/30/00
Bear Creek, Little Dry Creek
200
Englewood to Denver
7.78
2/1/839/30/00
L/E Discharge, PSC Arapahoe and Zuni, Cherry Cr., Farmers and Gardeners Ditch
400
2/5/96
Table 7. Reaches used in the residual flow analysis leading to estimates of ungaged flows in Segments 6c and 14.
Groundwater with High Concentrations of Nitrate - A secondary modeling exercise was undertaken to evaluate the potential effect of hypothetical groundwater sources. Some data collected by the Denver Department of Environmental Health (DDEH) identified pockets of high concentrations of nitrate in the groundwater. The data however, are variable and can support only very general assignment of contributions from these sites. The secondary modeling showed that concentrations of nitrate in groundwater of 10 to 100 mg/L could have a significant effect at low flows. Table 8 summarizes this modeling exercise. Hypothetical Concentration in Groundwater (Seepage) mg/L Nitrate-Nitrogen
Modeled Concentration in South Platte Downstream** mg/L Nitrate-Nitrogen
3*
14.4
10
14.7
20
15.0
50
16.1
100
17.8
* Characteristic values for the lower half of Segment 14. ** Immediately downstream of hypothetical groundwater source, under fully mixed conditions.
Table 8. Effect of hypothetical changes in nitrate concentrations of seepage from the DDEH study area on concentrations of nitrate in the South Platte River at acute low flow.
The areas of contaminated groundwater have not been fully characterized but contributions have been estimated in order to address concentrations and flow rates. The two most significant sites were included in the modeling, one below Harvard Gulch and another just above Mississippi Avenue. Median values of 10.3 and 17.6 mg/L nitrate, respectively, were used in the model. Even though the concentrations are relatively high, the mass loading is small because the flow is small. These contributions are represented in Tables 10 and 11 as Denver Contaminated Sites.
Colorado WQCD
16
April 21 2004
So. Platte seg 14 Nitrate TMDL
VII.
Final
TMDL ALLOCATION
Allocation Methodology The sources contributing to the nitrate load in Segment 14 have been identified and categorized for the purpose of allocation. Wastewater treatment plants and groundwater are the two major types of sources. While there may be opportunities to reduce nitrate contributions where there are localized areas of high nitrate in the groundwater, the implementation of reductions was considered too uncertain to incorporate as an assumption. Therefore, it was assumed that reduction in nitrogen species discharged by municipal wastewater treatment plants will have to assure attainment with the nitrate standard. Modeling established the basis for discussions among the municipal dischargers that were expected to have allocations (L/E, Centennial, and Glendale). Allocations to these dischargers were not based on a whole-segment allocation strategy, but instead were developed as a reasonable strategy considering the level of influence at the water supply diversions and reasonable attainability in terms of concentration for each facility. TMDL Scenario Assumptions Specific assumptions were made to initiate the TMDL modeling. These include using the reset option as discussed above and setting L/E WWTP discharge capacity at 50 MGD. the 50 MGD capacity was selected since the L/E plant is operating at near capacity and the Cities of Littleton and Englewood are actively planning for expansion. Specific assumptions were also necessary about how much of the total nitrogen load would be discharged by each facility as ammonia, as well as what target levels to use. Target levels: The nitrate target is set at 9.8 mg/L at the Allen Treatment Plant diversion and 10.0 mg/L at the Burlington Ditch headgate. The levels are not the same for each target since the point at which state waters ends and the water system begins is not the same for each system. For the Allen Treatment Plant, the diversion from state waters occurs at the intake structure on the South Platte River itself. In this case, 9.8 mg/L has been established as the target. Thornton’s intake is from the Burlington Ditch, approximately three miles down stream of the headgate. The Ditch is also state waters. There are no additional sources of nitrogen between the headgate and Thornton’s diversion and any significant nitrogen transformations would result in net loss to the system. In this case 10.0 mg/L nitrate has been established as the target. Discharged Ammonia: Since nitrogen transformations play such a large role in determining the concentration of nitrate at the target point, the TMDL to protect the water supply use in Segment 14 could be developed that specified either the daily Total Inorganic Nitrogen (“TIN”) load or the daily maximum ammonia and nitrate for each month. The TIN load must be a level that is protective regardless of whether the nitrogen is discharged as ammonia or nitrate. A TIN TMLD would provide more flexibility for the discharger. The Ammonia/Nitrate TMDL would provide slightly higher total nitrogen load, but the added burden of compliance with two limitations. The Division has chosen to present the draft TMDL in terms of a TIN load that is protective, rather than specifying both nitrogen terms. Other scenarios that are presented during the public comment period will also be considered. Colorado WQCD
17
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
In order to establish the TMDL for TIN, conservative assumptions were made regarding the concentration of ammonia discharged by the WWTPs based on historic ammonia discharge concentrations. From January 1993 through August 2000, the Littleton Englewood WWTP sampled ammonia concentration in its discharge 2800 times. A frequency histogram of this data is provided in Figure 8. Approximately 2360 samples (or 84%) had ammonia concentrations at or below 5 mg/L. An assessment of the monthly occurrences shows that values less than 1.0 mg/L occurred in every month, while values equal to or greater than 15 mg/l occurred only in January, February, July and October. Initial scenarios of 0.0 mg/L and 15.0 mg/L ammonia were modeled. The Division compared modeled TIN results for each month for two scenarios: all WWTPs discharging 0.0 mg/L ammonia, and all WWTPs discharging 15.0 mg/L ammonia. The results are summarized in Table 9. Each month, for each WWTP, the ammonia discharge scenario was selected that met the target levels with the lowest TIN. As could be expected, because the conditions in the River change from month to month, the selected scenario changes depending on the month. In the winter and spring months, with low flow (lower travel times and lower loss rates) and lower temperatures (lower loss rates), 0.0 mg/L ammonia (which means discharging all the TIN as Nitrate) was the most conservative approach. In the summer and fall, with higher flows and higher temperatures, 15.0 mg/L ammonia was the most conservative approach.
Frequency Histogram - Ammonia in Effluent Littleton/Englewood (1/93 - 8/00)
Frequency
400
300
200
100
0 0
10
20
Ammonia (mg/L)
Figure 8. Frequency of discharged ammonia for Littleton Englewood WWTP (January 1993 to August 2000)
Colorado WQCD
18
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Effluent TIN, mg/L assuming 0.0 mg/L Ammonia Mar Apr May Jun Jul Aug Sep Oct
Source
Jan
Feb
Centennial Littleton/Englewood Glendale*
29.7 23.4 40.0
27.7 23.2 40.0
Centennial Littleton/Englewood Glendale*
30.5 24.0 40.0
27.8 23.7 40.0
Allen Plant Burlington
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
0.0
0.0
15.0
15.0
0.0
15.0
15.0
15.0
15.0
15.0
15.0
28.8 31.8 44.8 43.1 49.4 48.0 43.3 40.8 22.6 25.4 39.5 43.0 40.7 43.8 36.2 32.1 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Effluent TIN, mg/L assuming 15.0 mg/L Ammonia 29.2 32.4 46.5 43.4 48.3 46.7 42.1 40.6 23.0 24.7 36.3 39.1 36.9 39.9 32.6 29.4 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Target Points, Nitrate-Nitrogen (mg/L)
Nov
Dec
36.0 26.9 40.0
31.9 26.6 40.0
35.5 25.7 40.0
32.4 26.5 40.0
Selected Discharged Ammonia Concentrations (mg/L) 0.0 0.0 0.0 0.0 15.0 15.0 15.0 Centennial 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 Littleton/Englewood 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 Glendale* * Glendale WWTP TIN set at 40 mg/L, ammonia discharge set equal to L/E Boxes identify which scenario (0.0 or 15.0) is protective for each month for each plant.
Table 9. L/E at 50 MGD: Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia.
TMDL and Allocations Table 10 presents the modeled flows (in cfs) and concentrations of nitrate (in mg/L) at selected locations in Segment 14 that would be necessary to attain the nitrate targets at the Allen Treatment Plant Diversion and the Burlington Ditch Headgate. Table 11 presents the TMDL in terms of the monthly TIN load in kg/d.
Colorado WQCD
19
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
L/E at 50 MGD Month Location
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
0
Discharge (cfs) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 50 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
0
0
0
1
5
2
0
1
0
0
0
1
0
1
3
8
6
8
5
1
2
0
1
2
1
2
4
15
10
10
8
3
4
2
2
13
13
13
13
13
13
13
13
13
13
13
13
30
27
28
31
46
42
45
44
40
39
34
31
26
22
23
27
32
31
26
35
36
37
34
29
19
18
15
18
28
13
7
10
12
18
19
25
26
27
25
28
60
58
30
33
20
27
31
39
77
77
77
77
77
77
77
77
77
77
77
77
0
0
0
0
0
0
1
1
1
0
0
0
118
119
117
120
154
157
130
136
120
125
123
131
5
2
2
6
11
17
16
13
16
10
8
5
3
3
3
3
3
3
3
3
3
3
3
3
11
8
8
8
21
13
13
13
11
12
8
11
117
116
110
116
164
161
137
146
131
134
123
131
0
Nitrate Concentration (mg/L) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 50 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
1
1
1
30
28
29
32
45
43
33
32
27
26
21
32
10
10
10
10
10
10
10
10
10
10
10
10
8
8
8
8
8
9
9
9
9
8
8
8
1
1
1
1
1
0
0
0
1
1
1
1
4
4
4
4
4
5
5
5
5
5
5
4
23
23
23
10
40
42
22
25
18
14
11
12
-*
12
-*
12
12
12
12
12
12
12
12
-*
11
11
11
12
13
13
13
14
13
12
12
11
2
2
2
2
3
3
3
3
3
3
3
2
40
40
40
40
25
25
25
25
25
25
40
40
10
12
13
7
6
6
6
6
6
6
7
8
10
10
10
10
10
10
10
10
10
10
10
10
* These values are flow-weighted averages of two locations. When there is no flow, an average cannot be calculated
Table 10. L/E at 50 MGD: Summary of flows and concentrations of nitrate N to attain water quality targets as predicted by the TMDL model.
Colorado WQCD
20
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
L/E at 50 MGD Month Source
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
WASTE LOAD ALLOCATIONS (kg/d TIN) WASTEWATER TREATMENT PLANTS
Centennial Littleton/Englewood Glendale
956
892
927
1024
1443
1388
1555
1504
1356
1307
1143
1027
4432
4394
4294
4678
6875
7405
6989
7557
6174
5568
4868
5019
303
303
303
303
303
303
303
303
303
303
303
303
LOAD ALLOCATIONS (kg/d TIN) GROUNDWATER SOURCES
Seepage South Platte
76
56
58
78
112
161
176
227
217
198
133
101
Cherry Creek
88
51
48
131
171
208
200
178
209
171
120
91
Bear Creek
31
31
25
31
13
19
19
31
36
42
36
31
0
1
0
1
5
14
18
23
18
12
2
0
Contaminated Sites
UNUGAGED SURFACE WATER SOURCES
Little Dry Creek
15
12
10
10
10
10
9
10
11
11
14
14
Sanderson Gulch
12
12
12
12
12
12
12
12
12
12
12
12
Lakewood Gulch
20
20
20
20
20
20
20
20
20
20
20
20
W Harvard Gulch
2
2
2
2
2
2
2
2
2
2
2
2
73
73
73
73
73
73
73
73
73
73
73
73
All others UPSTREAM SOURCES
SP blw Chatfield
0
.1
0
0
1.6
.3
0
0
0
0
0
0
Bear Cr blw Res
29
26
13
17
31
4
1
4
2
14
19
43
0
0
0
0
0
0
0
0
0
0
0
0
Cherry Cr blw Res.
Table 11. L/E at 50 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN for sources using TMDL modeling assumption.
Waste Load Allocation for Sources to be Controlled Centennial: The allocation for Centennial is established to protect the Allen Diversion. Centennial’s TIN daily concentration allocation ranges from 27.7 mg/L (892 kg/day) in February to 48.3 mg/L (1388 kg/day) in July. It is anticipated that with the newly constructed facilities at Centennial, the concentration of nitrate discharged from Centennial should normally be below 15 mg/L, keeping their TIN discharge low. The interaction between Centennial and Littleton/ Englewood WWTP effluents was explored in the modeling of nitrate concentrations at the Burlington Ditch Headgate. Modeling revealed that the concentration of nitrate at the Burlington Ditch Headgate was not affected substantially by changes in nitrate at Centennial. Littleton/Englewood: The L/E treatment facility is downstream from the Allen Diversion. It is the discharger with the largest effect on nitrate concentration at the Burlington Ditch Headgate. The TIN allocation for L/E ranges from 22.6 mg/L (4280 kg/day) in March to 39.9 mg/L (7557 kg/day) in August. The allocation of 22.6 mg/L for TIN will require construction of denitrification facilities at the L/E WWTP. Colorado WQCD
21
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Glendale: Glendale discharges to Cherry Creek five miles upstream of the confluence with Segment 14. The effluent from the Glendale WWTP has only a small effect on the nitrate reaching the Burlington Ditch Headgate. Glendale’s effluent is subject to considerable dilution by seepage before reaching the mouth of Cherry Creek, even under low-flow conditions. L/E and Centennial have both agreed to give Glendale a TIN allocation of 40 mg/L for each month. This allocation for Glendale will not require any modifications to the facility. Other Point sources: No allocations were included for other point sources. Sensitivity analysis of the model indicates that when the cumulative load from new or expanding facilities reaches 75 kg /day TIN (approximately equivalent to 20 mg/L TIN at 1 MGD), this TMDL and the modeling should be reviewed to determine the effect of the discharges. Stormwater outfalls that have nitrate concentrations exceeding 20 mg/L should be analyzed to assure that there are no illegal connections to a sanitary sewer or industrial source. If the concentrations of nitrate are due to contaminated groundwater infiltration, then this issue should be handled under the procedure outlined above for groundwater contamination sites. Load Allocation Among Sources Table 11 presents the monthly TIN load allocation for groundwater sources, ungaged surface water sources and upstream sources. These values were generated using information regarding the monthly flow, nitrate and ammonia concentration values used in the TMDL model that represent the current loading from these sources. No reduction for these sources is included in the TMDL. Contaminated Sites: A specific plan for correcting contaminated groundwater cannot be achieved at this time. Localized high concentrations of nitrate in groundwater entering Segment 14 are a concern and potentially contribute to the total nitrate loading at the Burlington Headgate. The TMDL modeling shows that they contribute up to 23 kg/d to the nitrogen loading of the South Platte River. Margin of Safety The margin of safety is the TMDL component that accounts for unknowns or uncertainties in the development of the TMDL. The margin of safety may be explicit (a separate value in the TMDL) or implicit (included in factors determining the TMDL). The calibrated nitrate model used in this assessment uses implicit mechanisms for managing the margin of safety. The magnitude of the margin of safety is reduced by much of the work performed in preparation of this TMDL. Some of the tasks were identified as the result of earlier modeling efforts and some were performed as ongoing data collection efforts. The combination improved the quality of the data available. Specific work includes: • •
Period of flow record - The number of gages and the period or record provide a sound basis for low flow analysis. Nitrate water quality data - The coordinated water-quality monitoring program provides a sound basis for water quality modeling.
Colorado WQCD
22
April 21 2004
So. Platte seg 14 Nitrate TMDL
• • • •
Final
Nitrate analysis procedures - The analytical methods for nitrate are accurate. Behavior of nitrate in the environment - Nitrate fate is well understood and can be measured with accurate analytical methods. Dependability (acceptance) of the model - The model is well accepted, calibrated and validated. Special studies performed - See Section VI.
The model incorporates several worst case and conservative assumptions, including: • All facilities discharging at maximum design flow • All facilities discharging at maximum permitted concentration • Concurrent critical low flow and maximum discharge The above conservative assumptions and operating realities result in a substantial implicit margin of safety assuming implementation of the new discharge permit limits based on this TMDL assessment. An explicit margin of safety of 0.2 mg/L nitrate is included for the Allen treatment plant diversion. Implementation Implementation of the TMDL may require that load reduction be apportioned among all significant pollutant sources. The allocations described above will be considered when discharge permits are developed. At the time of permit issuance, the reasonable potential for each source to exceed the allocation will be evaluated and appropriate limitations and monitoring requirements will be imposed Site evaluation and remediation of groundwater contamination could be accomplished through remediation actions triggered under other federal, state, and local requirements related to contaminated sites. This TMDL is not intended to address or imply any groundwater clean-up standards, which must be addressed separately. This TMDL is also not intended, in and of itself, to require clean up of any particular sites. The purposes for including contaminated groundwater in this TMDL are to: (a) recognize that groundwater affects nitrate load in Segment 14; (b) acknowledge the need to require sites undergoing cleanup to evaluate effects on Segment 14; and (c) ensure that the groundwater contributing to Segment 14 surface water exceedence is dealt with appropriately. Reduction options would include; treatment based on economic viability; detention during periods of low flows; best management practices; or a proportionate increase of point-source treatment (to be paid by the responsible party) to offset the groundwater contribution. It is anticipated that the evaluation of the groundwater and any implementation would be undertaken by the responsible party and/or lead agency. Post-Implementation Monitoring Water-quality monitoring and modeling will determine the extent to which the implementation plans achieve the compliance goals. Monthly sampling at critical locations on the South Platte River and its tributaries will demonstrate the effectiveness of the wasteload allocations. The Division intends to work cooperatively with SP CURE and other agencies monitoring water quality, as necessary, to assess the success of the nitrate TMDL.
Colorado WQCD
23
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
The Division has proposed trend-monitoring sites for evaluation of nitrate and additional pollutants throughout Segment 14 on a monthly basis. The domestic water supply diversion at the Burlington Ditch Headgate will be included for assessment of attainment of the stream standard. A number of sites correspond with USGS gaging stations, for which accurate flow characterization is possible. These trend-monitoring sites currently proposed: South Platte River (SPR) upstream of the Allen Filter Plant SPR at the 19th Street gage SPR at the Henderson gage Bear Creek at the gage Cherry Creek at Champa Clear Creek at the confluence Sand Creek at the FRICO siphon SPR above the Burlington Ditch Headgate VIII. PUBLIC INVOLVEMENT The issue of nitrate impairment on Segment 14 has been discussed with interested members of the community at various forums over the past seven years. In 1994, the South Platte Urban Watershed TMDL Phase I was initiated. The results of Phase I presented a proposed approach to stakeholders, developed an organizational framework, evaluated technical issues, identified major water quality concerns and devised a phased work plan for conducting a South Platte Urban Watershed TMDL study. The Phase I Steering Committee consisted of the Division, the Denver Regional Council of Governments Urban Drainage and Flood Control District and US EPA Region VIII. Phase II of the TMDL project, which concluded in 1998, focused on narrowing the scope of the analysis and setting priorities for further work. One goal was the selection of appropriate watershed and receiving water models. A second goal was to evaluate nitrate and copper as the first constituents for TMDL modeling. Data gaps were identified and monitoring needs were identified to fill those gaps. Phase II also had a broad-based Steering Committee. In 1999 South Platte CURE was formed to concentrate on actual development of individual TMDLs. SP CURE members include the cities of Aurora, Brighton, Denver, Glendale, Golden, and Thornton. Centennial, L/E, Denver, South Adams County, and Metro Wastewater represent wastewater utility providers. Representatives from industries include Conoco, Coors Brewing and Public Service companies. The Big Dry Creek Watershed Association and the Farmers and Reservoirs Irrigation Company are interested local partners who have joined SP CURE's efforts to monitor and improve water quality in the South Platte Urban Watershed. SP CURE members worked to share water quality information as it became available and have participated in DRCOG committee meetings and meetings with the Division and EPA. A public meeting was held in Denver on January 25, 2000 to provide an update on the status of several TMDLs and to answer questions from the public. Approximately 50 people attended the meeting. The Segment 14 Nitrate TMDL was included in that meeting. Formal notice of the Division’s intent to finalize this TMDL was published on April 3, 2000. Colorado WQCD
24
April 21 2004
So. Platte seg 14 Nitrate TMDL
IX.
Final
RESPONSIVENESS SUMMARY - First Public Comment Draft
Post-Public Notice Process Issues were raised during the public comment period for this TMDL regarding the adequacy of the public involvement process. Rather than continue discussions about the process, the Division decided to suspend the formal process for the TMDL and to hold a series of meetings with interested parties to assure that they understand the numerical modeling upon which the TMDL is based. Three public meetings were held with attendance ranging from 7 to 14 external parties. A general meeting was held on June 22, 2000, to explain and discuss the model. A meeting on June 29, 2000, focused on questions collected after the previous meeting. A final meeting was held on July 21, 2000. The Division again published notice of the TMDL on August 4, 2000. Formal comments were accepted until September 18, 2000. Extensive comments were received from the City of Thornton. As a result of these comments, the Division revised the model and the TMDL. The following section contains the Division’s response to the technical and legal comments. Response to Technical Comments and Explanation of Changes to the Model The following technical comments were received either as formal comments or in the informal modeling meetings held in June and July, 2000. Why was this type of model used rather than a model such as QUAL2E or WASP5, commonly used EPA models? The South Platte Segment 14 Nitrate TMDL model is a multi-page spreadsheet using MicroSoft Excel. It is a mechanistic, steady state deterministic model that handles non-conservative substances by mass-balance principles. This type of model has been used in Colorado for many years and its foundation algorithms come from EPA water quality modeling, specifically STREAMDO. This model is similar to the Colorado Ammonia Model, which is used by the Division and others to develop permit effluent limits. Also, this type of model has been used in modeling of South Platte segment 15, just downstream of segment 14. The model choice was made in 1997, by the South Platte Urban Watershed Total Maximum Daily Load Steering Committee, which was organized by Denver Regional Council of Governments. The choice of models is discussed in the committee’s final report entitled Phase II: Regulatory Processes, Model Development, Screening Criteria and Copper and Nitrate Evaluation (DRCOG, 1998). QUAL2E was reviewed by the Steering Committee, but not chosen because the spreadsheet model approach was considered to be more transparent and flexible. The model was developed without regard to standard engineering practices. The commenter identifies four critical steps (problem identification, model selection, data collection, model calibration and validation) as the basis of “standard engineering practices,” and objects to the modeling on the basis that not all steps were followed.
Colorado WQCD
25
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
The Division believes that the assessment and model adequately addressed these steps. The initial phase of the Nitrate TMDL began in 1993 with scoping of Denver Metropolitan area water quality problems by a contractor retained by the Division and EPA Region VIII. In 1994, a TMDL stakeholders group consisting of EPA, WQCD, DRCOG, and many other Denver urbanarea parties was formed. This Steering Committee identified and characterized water quality problems in the Denver urban area watershed and then selected an appropriate model for developing TMDLs. The South Platte Urban Watershed TMDL Steering Committee completed Phase II of the Nitrate TMDL in 1998. To augment the large amount of data already available in the urban area, an extensive and coordinated, TMDL specific, monitoring effort was started in the late 1990’s by the South Platte Coalition for Urban River Evaluation. This continued well into 2000 and the data were utilized in calibration and validation of the current TMDL model. Model calibration and validation are discussed below. The model uses unrealistic flow conditions and fails to reflect actual diversions. Two issues are raised by the commenter: 1) computational errors in the original hydrologic analysis, and 2) the computational strategy used for estimating ungaged flows, both of which have been corrected in the revised hydrologic analysis. The placement of the Farmers and Gardners Ditch relative to the confluence of Cherry Creek (the Ditch diverts water above Cherry Creek) was corrected. Correcting the location of the ditch made it necessary to recalculate withdrawals from and additions to the mainstem that are estimated by the method of differences. In addition, the original hydrologic analysis used incorrect information for the outflow from Chatfield dam to Segment 6c of the South Platte. This information was obtained from the Bureau of Reclamation (“BOR”), and is reported as “Outflow from Chatfield Dam.” The outflow, however, flows into a manifold where it can be separated according to destination. Some of the release is diverted within the manifold and does not reach the South Platte directly. Thus the use of the BOR outflow data led to an overestimation of the Chatfield releases to Segment 6c. This in turn affected the estimates of seepage for the reaches involving Chatfield and downstream gaging stations. Flow released directly from Chatfield into Segment 6c is measured at a gage on the River maintained by the State Engineer's Office on the River immediately below the dam. Seepage estimates were then recalculated. The result is an upward revision of the seepage estimates for the South Platte between Chatfield and Union. The commenter expressed concern about the comparability of seepage estimates based on different periods of record, and recommended various more complicated approaches, for instance by weighting gage records according to the amount of data available. The division does not think that these additional complexities are justified. Because the residuals are small numbers obtained by subtractions, the seepage estimates are subject to error. This error has been minimized to the fullest extent possible by using all the gage data that seem applicable and using smoothing methods and medians to remove undue influence of random measurement error. The Division does acknowledge, however that the record for the gage at 50th Avenue is short; residuals based on this gage have been excluded from the revised hydrologic analysis Because some time had elapsed since inputs were determined for the original model, and because Colorado WQCD
26
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
some revisions were necessary, the Division elected to broaden the revision by extending the hydrologic record through water year 2000. In addition to extending the hydrologic record, data were added for calculating ungaged flows (seepage) in Bear Creek. The model is based on a “reset” approach, which removes nitrogen load from the river. An assessment of this nature, with multiple discharges and contributions, poses a challenge in regards to low-flow assumptions. Discharge permits must be written for the “critical” flow condition, which for acute conditions such as nitrate is generally the 1E3 or the empiricallybased 1-day low flow with an average 1-in-3-year recurrence interval, as determined from historical records. Discharge permits also must be written for the maximum effluent discharge that might be expected during the permit term, which is generally the design capacity. For multiple dischargers, if an upstream plant is modeled as discharging at capacity (a condition that is not included in the historical record), the flow predicted at the downstream discharge point will exceed the empirically-based low flows. The flow “reset” approach allows the model to remove the increment of flow attributed to the difference between present rate of wastewater discharge and the capacity discharge so that the river does not accumulate all of the capacity of the discharges. The model was developed with the ability to toggle the “reset” on and off, giving the Division the opportunity to assess which scenario is more appropriate for setting limits. The original TMDL was developed with the reset in place. The revised modeling report provides tables of results with the reset both On and Off. In the revised model, the comparison of the results with the reset On and Off, shows that the 1E3 low flow conditions (reset On) is not the most critical for nitrate delivery in segment 14 of the South Platte River. For that reason, the revised TMDL relies upon modeling without the reset approach. The model contains a fundamental mathematical error in the nitrate transformation rate. The Division acknowledges that there was a mathematical error in the spreadsheet used for calculating nitrogen transformation rates. These rates have been recalculated and are discussed below. The error was restricted to an ancillary spreadsheet and was not present in the mathematical representation of nitrate transformation in the TMDL model. The model oversimplifies the biological processes and nitrogen kinetics in the river and the selected transformation rates are excessive. The Division believes that the model adequately reflects the nitrogen kinetics in the River. The representation of biological processes is consistent with previous experience in characterizing loss rates for ammonia and nitrate in the South Platte (e.g., TMDL for oxygen in Segment 15, South Platte River). Ammonia loss rates are interpreted for the purposes of modeling as nitrification rates, but other process also may play a role. Adsorption of the ammonium ion, volatilization of ammonia and uptake of ammonia by autotrophs or bacteria could contribute to the measured loss rate but these process are not expected to be significant when compared to Colorado WQCD
27
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
nitrification in the Segment 14. Similarly, nitrate loss rates are treated as denitrification rates. Uptake by autotrophs could contribute to the measured nitrate loss rate, but it is unlikely to be significant when compared to denitrification as long as ammonia is present. A potential role for other minor processes contribution to the disappearance of ammonia or nitrate is not problematic to the modeling provided that these other processes continue to operate in the future as they have in the past, which is highly likely. These other processes are accounted for, or included in , the empirically-derived loss rates. Therefore, “simplification” is justified by the dominant roles played by the processes of nitrification and denitrification, and the present approach is consistent with the available data. Transformation rates (ammonia loss rates and nitrate loss rates) were not selected from the literature, they were estimated on the basis of actual ammonia and nitrate data from the River. Because of the delay in promulgating the TMDL, more data sets were available for use when the model was recalibrated in 2001, and some of these sets could be used for validation. Calibration (estimation of the transformation rates for ammonia and nitrate) is discussed in the modeling report beginning on page 23. Forty-five data sets were available for determination of the loss rates. Before screening the datasets, ten data sets were selected at random and reserved for validation. The remaining sets were screened for completeness, and sets were discarded that were missing too much information. Data sets were also discarded for which the flows did not balance between the gages (typically due to wet conditions involving storm flows from small tributaries). Some data sets remaining after this screening had additional deficiencies (missing data, etc.) and minimal substitution and interpolation was done in order to fill out these data sets so that they could be used. The calibration procedure leads to estimation of loss rates for each date on which information is available in the calibration data set. Rates were estimated separately for an upstream reach (above Dartmouth) and a lower reach (below Dartmouth). A rate for nitrification was obtained on each date for each reach through an adjustment of the ammonia loss rate sufficient to produce a match between the modeled and observed ammonia concentrations at the downstream end of each of the two reaches. With the nitrification rate in hand, the nitrate disappearance rate (denitrification) was estimated, assuming that the nitrification rate represented progressive addition of new nitrate within each reach at a known rate. The magnitude of the loss rates estimated in the calibration procedure is consistent with estimates from other Colorado streams (e.g., CDPS permit analyses on Segment 15, Boulder Creek). How was the model validated? Model validation is discussed in the modeling report beginning at page 26. Ten of the original 45 data sets were reserved for validation. Data sets were discarded that had too much missing information, or for which the flows did not balance between the gages (typically due to wet conditions involving storm flows from small tributaries). For each data set remaining, flow data for each date and reach were inserted into the model and the model was run for each date. The predicted nitrate levels were compared with the actual nitrate levels. Differences are small between observations in the validation data sets and predictions by the calibrated model, and the standard errors are less than 5 percent of the nitrate standard (10 mg/L). The Division concludes that the model performed well on the validations data sets. Colorado WQCD
28
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
How well do modeled conditions simulate or account for current conditions in Segment 14? Model validation is discussed above, and was done at the discharge and flow values that were measured on the sampling dates. The TMDL modeling is of a future or predicted condition, one that is very different from current conditions. The following table compares the statistics for the period of September 1998 – June 2000 with the TMDL conditions. Since the TMDL is based upon increased effluent flow, and the simultaneous occurrence of critical conditions, measured conditions are not comparable to the modeled conditions.. Measured Conditions 1 (9/98 – 6/00)
Modeled Conditions TMDL
median flow median flow (min – max) mg/L range (min – max) mg/L range Centennial WWTP 12.3 26.1 (8.8 - 18.0) 4 - 5.6 MGD (23.5 - 28.6) 8.5 MGD TIN TIN SPR at Allen WTP 1.3 7.3 Intake (0.1 - 3.8) not reported (5.9 - 9.4) 26.5-45.7 2 NO3 NO3 cfs L/E WWTP 31.8 32 (18.4 - 46.8) 22.4 - 34.2 MGD (32 - 32) 36.3 MGD TIN TIN Glendale WWTP 11.0 36.0 (1.1 - 15.4) 0.69 - 0.96 MGD (34.1 - 40.6) 2 MGD TIN TIN SPR at Burlington 4.0 7.8 Headgate - Intake (0.02 - 7.4) 3 112 - 2540 cfs (5.4 - 9.4 ) 105.9-146.7 2 cfs NO3 NO3 1 from SP CURE data sets; 2 range in monthly flow, 3 a measurement of 13.7 was discarded as invalid since it far exceeded the other measurements on the same day. TIN means Total Inorganic Nitrogen or Ammonia + Nitrate
Table 12. Comparison between Measured Conditions and Modeled Conditions (Prepared for Oct 2001)
How sensitive is the model to changes in input levels? The largest concentrated nitrogen additions to the South Platte River are the wastewater treatment plants. Under modeled conditions (design capacity) the combined discharge of Centennial, L/E and Glendale is 46.8 MGD (or 72.45 cfs). If they each were to discharge 35 mg/L nitrate+ammonia nitrogen, the load to the system would be approximately 13,700 lbs/day; L/E WWTP alone would contribute 10,600 lbs/day. Ungaged flows, on the other hand are approximately 3 cfs per mile (see Table 4 of the Modeling Report) for 20.7 mile (or 62 cfs), and estimated ammonia+nitrate concentration is approximately 3 mg/L (see page 15 of the Modeling Report). Using these estimates, this source contributes to approximately 1000 lbs/day to the system. The model is much less sensitive to changes in seepage rates or seepage concentration than to changes in waste water treatment plant effluent and discharge because of their much lower contribution. The commenter objects to the use of such a low nitrate concentration in the ground water seepage. Since the ungaged flow rate is substantial, an order of magnitude increase in the Colorado WQCD
29
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
concentration could result in a considerable increase in overall loading. The commenter suggests that information from wells along segment 14 that was incorrectly rejected because the well water appeared to be influenced by water from the River. The Division believes that this information was treated properly. Water in alluvial wells does not reflect a new source of nitrate, but nitrate diverted from the River into the alluvium. The seepage concentrations used in the model are appropriate The model’s sensitivity to changes in the WWTP discharge flow and concentration is discussed beginning on page 30 of the modeling report. Glendale has a large influence on the nitrate concentration at the mouth of Cherry Creek but a minor influence on the nitrate concentration at the Burlington Headgate. Centennial has a major influence on the nitrate concentration at the Allen WTP intake, but a minor influence on the concentration at the Burlington Headgate. The most influential source in the concentration of nitrate at the Burlington Headgate is the concentration and discharge rate of the L/E WWTP. The other influential factor is the nitrate loss rate, which is discussed above. Response to Legal Comments One commenter also submitted "legal arguments." These arguments claim that development of this TMDL is procedurally flawed because of public and/or stake-holder participation problems. The commenter doesn't assert that there was not public/stakeholder participation; nor does it assert that it was not able to participate. Rather, the commenter asserts that the kind and level of its participation was insufficient under the Division's Continuing Planning Process (“CPP”). A dispute exists as to why the commenter did not have greater and/or different participation in the development of the South Platte Segment 14 TMDL. The CPP is a document that the Division is required by the federal Clean Water Act to develop and submit to EPA. It describes the various processes that are followed in Colorado with regard to water quality protection. The CPP is not a regulatory document, and thus does not mandate any particular process. With regard to TMDLs, the CPP only summarizes and describes the different processes that may be followed in the development of a TMDL. The CPP, therefore, vests in the Division the discretion to determine how best to develop a TMDL. Assertions to the contrary notwithstanding, there is no specific level of public or stakeholder participation specified or required in the CPP or elsewhere. Although the CPP refers to both stakeholder and public involvement, the CPP does not further distinguish between the two. There is, therefore, no legally established levels of participation for public participation versus stakeholder participation, and for good reason: since every TMDL is unique, the kind and level of public/stakeholder participation will necessarily vary. The CPP also provides that a rigid approach - one that, for example, imposes different requirements on public versus stakeholder involvement - is inappropriate. In any event, as indicated by the subsequent actions of the Division described above, any deficiencies in the public participation process were rectified when the Division suspended the TMDL development process, held a series of public meetings, re-noticed the TMDL, solicited Colorado WQCD
30
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
public comments, and then revised the TMDL as a result of the public comments received. Importantly, the TMDL was revised based upon comments received from the commenter. Thus, the public participation process followed in the development of this TMDL accomplished the objective of the CPP and the commenter: substantive comments on, and any appropriate changes to, the TMDL. X.
RESPONSIVENESS SUMMARY - Second Public Comment Draft
The Division’s intent to finalize the TMDL was published on November 2, 2001. Comments were received from the City of Thornton, The Cities of Littleton and Englewood, The Farmers Reservoir and Irrigation Company, and South Platte Cooperative for Urban River Evaluation (“SP CURE”). Post-Public Notice Process: December 2001 - May 2003 A series of meetings and conversations were held with technical representatives of the commenters in early 2002. Specifically, the focus of the interaction was to discuss how to achieve consistency between how ammonia limits for the permits are calculated to protect the aquatic life use and how nitrate assimilative capacity for the TMDL is calculated to protect the water supply use. Namely: • a common ammonia loss rate to be used for the permit and the TMDL in Marcy Gulch and the South Platte River above L/E’s discharge location. • a common ammonia loss rate to be used for the permit and the TMDL in the South Platte River below L/E’s discharge location • ammonia loss rate for Cherry Creek below Glendale WWTP • daily ammonia discharge rates. Separate meetings were also held with legal, policy and technical representatives of Thornton and L/E in late February and early March 2002. A second set of meetings and conversations was held with technical representatives in April and May 2003. At the meeting, the TMDL modeling contractor presented an overview of the changes that had been made to the model. The major focus of the meeting and subsequent conversations was to select daily ammonia discharge rates. Because nitrogen is not conservative (it changes form as it travels down the river), the modeling provides a different TIN number depending on what assumptions are made about how much of the total nitrogen is released as ammonia. If relatively more nitrate is released, more nitrogen is lost from the system through denitrification. If relatively more ammonia is released, the ammonia-nitrogen must first be converted to nitrate before denitrification can remove it from the system, hence, less nitrogen is lost. Dischargers were asked to select an initial daily ammonia discharge rate so that a starting point could be selected. Response to Comments and Explanation of Changes to the Model Significant substantive written comments were received from the parties. Some of the comments resulted in changes in the model, modeling assumptions or input parameters. The issues are addressed below. Colorado WQCD
31
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Consistency between the TMDL model and CAM: The Division uses the Colorado Ammonia Model to develop permit limits to protect aquatic life from acute and chronic ammonia toxicity. It is important that the assumptions that are used in the TMDL model are the same as those made in the CAM model to enable the TMDL results to be subsequently incorporated into one or more discharge permits. After discussion with the interested parties, the Division has decided to use the CAM default value of 6.0 day-1 for the ammonia loss rate for reaches of the system where we do not have data to generate a site-specific loss rate. Even though the S P CURE site-specific data showed a zero loss rate for the South Platte River below Marcy Gulch, that appears to be because the upstream ammonia was not detectable. It is more reasonable to assume that if ammonia was present, some would be lost in that reach of the River. It has been the Division’s practice to use the default value (6.0 day-1) for development of ammonia limits in cases like this. On this basis, the Division has decided to use this value for the South Platte River between Marcy Gulch and L/E and for Cherry Creek. Margin of Safety and the Conservative Nature of the Assumptions: Comments were received that stressed both that the model was too conservative and that it was not conservative enough. Among the factors cited to support one commenter’s conclusion that the TMDL is unduly conservative and onerous and unreasonable are: the flow reset option should be used; the target nitrate value at Burlington should be 10.0 mg/L not 9.4 mg/L; the nitrate concentration in the seepage should not be assumed to be at the top of the range; critical low-flow conditions should not be assumed to exist every day throughout the year. Factors cited to support another commenter’s conclusion that the TMDL is not conservative enough are: daily attainment of the nitrate standard is not assured using monthly averages for nitrate discharges, especially when the daily flows can exceed the “design capacity”; a single nitrate loss rate is inappropriate since the rate in the lower reach appears to vary as the flow varies, and the treatment of ungaged flows is unrealistic. Flow reset option should be used: The flow-reset concept is described in the TMDL and modeling report. It allows the model to remove the increment of flow attributed to the difference between present rate of wastewater discharge and the capacity discharge so that the river does not accumulate all of the capacity of the discharges. The reset option provides an opportunity to understand the River conditions at “critical” low flows for which permits must be written. The model was developed with the ability to toggle the “reset” on and off, giving the Division the opportunity to assess which scenario is more appropriate for setting limits, since standards must be attained at all flow conditions. The revised modeling report provides tables of results with the reset both On and Off. In the revised model, the comparison of the results with the reset On and Off, shows that the 1E3 low flow conditions (reset On) is the most critical for nitrate delivery in segment 14 of the South Platte River. For that reason, the current TMDL relies upon modeling with the reset approach. Target at the Burlington Ditch headgate should be 10.0 mg/L not 9.4 mg/L: The nitrate targets in the previous draft of the TMDL were 9.4 mg/L nitrate at both the Burlington Ditch headgate and the Allen Treatment Plant diversion. This value was chosen as a conservative target. The Division agrees that other factors provide adequate conservativism and has changed the targets for both diversion points. The current target is set at 9.8 mg/L at the Allen Treatment Plant diversion and 10.0 mg/L at the Burlington Ditch headgate.
Colorado WQCD
32
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
The rationale behind changing the two targets independently is that the point at which “waters of the state” ends and the water system begins is not the same for each system. For the Allen Treatment Plant, the diversion from state waters occurs at the intake structure on the South Platte River itself. In this case, 9.8 mg/L has been established as the target. Thornton’s intake is from the Burlington Ditch, approximately three miles down stream of the headgate. The Ditch is also state waters. There are no additional sources of nitrogen between the headgate and Thornton’s diversion and the only significant nitrogen transformations that are possible would result in net loss to the system. While the transformations in the Burlington Ditch were not modeled, the Division is confident that if the target of 10.0 is met at the headgate, the water supply use will be protected at Thornton’s diversion point. Seepage Quantity Estimates are inappropriate: In the estimation of ungaged flows (both seepage and dry weather surface flows), gage records were analyzed to determine the flow residuals at five locations. Flow residuals are the left over flow that can not be explained by upstream stream flow, diversions or tributary inflows. Each month was analyzed separately and the analysis was limited to the lower end of the flow range. Each date that was analyzed generated a residual value. The median of the residuals for each month was used as the estimation of ungaged flow for that month. (The estimates were smoothed, using a three-point moving average across the months.) Given the limitations on the amount of data, the Division continues to support this approach. Seepage Quality Estimates are inappropriate: In the time between the previous and current drafts of the TMDL, additional information was obtained concerning the nitrate concentration in the seepage component. Nitrate concentrations were adjusted based on an assessment of the new information. Samples collected by USGS NAWQA as part of the urban land use study indicate that nitrate concentrations along the river were 0 to 5 mg/L and there was not apparent spatial pattern to the concentrations. Therefore, a concentration of 1.5 mg/L, which is near the median of the values along the mainstem, was applied to both reaches. Along Cherry Creek, the NAWQA data supports the use of 3 mg/L. Allocations should be provided for each month: The previous version of the TMDL used the allocations established in the most critical month (February) and extended that to the entire year. The revised TMDL provides an allocation developed for each month based on the monthly low flow and river conditions. Daily ammonia can exceed the chronic ammonia limit. The previous version of the TMDL set a WWTP’s contribution of ammonia to the monthly average (chronic) values that have been set to protect aquatic life, rather than a daily maximum value. In some cases, discharge monitoring reports show that the daily ammonia concentrations significantly exceed the monthly average limit. To address this the Division asked dischargers to select a daily ammonia discharge rate to include in the model. The dischargers chose 15 mg/L, except when the acute value to aquatic life was lower than 15. In these cases, the acute value was used. This scenario was included in the modeling report. Daily discharges can exceed “design capacity”: Daily peaking flow from a waste water treatment plant can exceed design capacity. To evaluate the effect this might have on the nitrate Colorado WQCD
33
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
TMDL, the Division examined daily flow and nitrate discharge records for L/E from January 1993 to August 2000. The information for the 520 days with both measurements is plotted in Figure 9. The Division found that peak flow and high nitrate did not occur at the same time. Relationship between Effluent flow and Nitrate Littleton/Englewood (1/93-8/00) 38 Design Capacity
36
flow (mgd)
34 32 30 28 26 24 22 20 0
10
20
30
40
50
nitrate (mg/L)
Figure 9. Littleton/Englewood WWTP Flow and Nitrate
Further, the Division found that generally, the peak flows occurred during wet weather events when infiltration and inflow dilute the effluent and the River was carrying additional runoff water. Tables 13 presents effluent flow, nitrate+nitrite, and flow at the Denver gage for the dates with the highest nitrate+nitrite. Table 14 presents effluent flow, nitrate+nitrite, and flow at the Denver gage for the dates with the highest effluent flow. Date
WWTP Flow (MGD)
NO3+NO2 (mg/L)
28-Apr-98 12-May-98 17-Aug-00 19-Oct-99 16-Jun-99 30-May-95 17-Jul-00 28-Jul-98 4-Aug-98 27-Apr-99 4-May-99
31.0 31.2 31.3 31.5 32.1 32.2 32.4 32.5 33.2 34.8 37.5
23.25 19.28 24.10 21.60 17.80 16.10 26.22 20.63 26.00 18.00 17.40
Daily Flow Denver gage (cfs) 1550 2120 1010 248 2050 2170 1880 736 1420 446 2300
Table 13. Littleton/Englewood WWTP Daily influent flow, nitrate+nitrite and mean daily flow at the Denver gage for the same date, for the days with the highest daily influent flow (Jan 1993 - August 2000)
Colorado WQCD
34
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Date
WWTP Flow (MGD)
NO3+NO2 (mg/L)
3-Nov-98 20-Apr-00 13-Apr-00 29-Sep-98 16-Jun-98 2-Jun-98 7-May-00 14-Jun-00 3-May-00 20-Oct-98 18-May-00 6-May-00 5-May-00 4-May-00 15-Jun-00 27-Apr-00 20-Jul-00 17-May-00 30-Jun-98
26.4 24.0 23.0 25.9 28.3 28.4 22.8 24.6 22.8 25.8 23.0 22.5 22.0 22.5 25.1 23.1 27.1 22.4 27.7
30.00 30.10 30.12 30.40 30.83 31.00 31.02 31.50 31.84 32.00 32.10 32.17 33.63 36.79 37.00 37.51 37.60 39.80 45.75
Daily Flow Denver gage (cfs) 164 279 255 138 403 822 511 286 419 252 763 525 530 415 244 251 267 806 260
Table 14. Littleton/Englewood WWTP Daily influent flow, nitrate+nitrite and mean daily flow at the Denver gage for the same date, for the days with the highest nitrate+nitrite (Jan 1993 - August 2000)
Nitrate loss rate is not characteristic of modeled flow conditions: The nitrate loss rates used in the previous version of the model and TMDL were developed using the median of the individual loss rates calculated from measured data (1.77 day-1). However, the individually calculated nitrate loss rates below L/E show a strong relationship with flow. At low flows, the nitrate loss rate is lower. The median rate (from dates when the flows at the Denver gage ranged from 123 to 2570 cfs) is not characteristic of the conditions when flows are lower. The flows for the modeled condition range from 65 to 120 cfs. Inclusion of the 2002 data sets in the analysis for the current draft of the model and TMDL allowed a more focused assessment of low flow conditions. The Modeling Report, beginning on page 37 (see also Figures 7 and 8), discusses the derivation of flow-dependent nitrate and ammonia loss rates for the South Platte River between L/E and the Burlington Ditch headgate. The nitrate loss rates for the modeled flow conditions range from 0.61 day-1 to 1.3 day-1. Cost/Benefit Analysis: One commenter asserted that the Division has not prepared a cost/benefit analysis as required by §25-8-102(5) of the Colorado Water Quality Control Act (“Act”). Section 25-8-102(5) states: (5) It is further declared that the general assembly intends that this article shall be construed to require the development of a water quality program in which the water quality benefits of the pollution control measures utilized have a reasonable relationship to the economic, environmental, energy, and Colorado WQCD
35
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
public health costs and impacts of such measures, and that before any final action is taken, with the exception of any enforcement action, consideration be given to the economic reasonableness of the action. Such consideration shall include evaluation of the benefits derived from achieving the goals of this article and the economic, environmental, public health, and energy impacts to the public and affected persons.
Section 102(5) was adopted in 1981 as a part of Senate Bill 10. Asserting that this and other provisions of SB10 conflicted with the federal act, and concerned that the state would use such economic considerations to override the requirement of the federal Clean Water Act (“CWA”), the US EPA objected to this language and threatened to revoke Colorado’s authorization to conduct the water quality programs under the CWA. To address EPA’s concerns, the state adopted SP 83 a couple of years later. Among other things, it provided that the Colorado Water Quality Control Commission (“WQCC”) shall “maintain a [water quality] program which does not conflict with the provisions of the federal act....” §25-8-202(6). Thus, the policy declarations in §102 are subordinate to the express commands of §202(6). Section 102 is a legislative declaration of policy and is not self-implementing. The policy is to be implemented in specific circumstances elsewhere in the Act. For example, before the Commission adopts stream standards, it is required to consider economic and other factors §258-204(4). Stream standards therefore, such as the nitrate standard applicable to South Platte Segment 14, have already had economic factors taken into consideration, and any subsequent actions based on the standards, such as a TMDL, are consistent with the policy. Similarly, as a part of its evaluation of discharge permit applications, the Division is required in specified circumstances to conduct a similar economic evaluation of the impact of the permit’s effluent limitations. §25-8-503(8). Accordingly, any changes to a permit as a result of a TMDL (which is driven by standards that have already undergone an economic analysis), may receive a similar evaluation, thus complying with the policy of §102. In fact, the language in §503(8) restates much of the provisions of §102 and provides a specific implementation requirement. Therefore, a TMDL is not the appropriate stage during which to conduct such an evaluation: the evaluation has already been done, and the TMDL will merely allocate the load to attain stream standards. Actual implementation of the standards, as identified by the TMDL, will be accomplished by discharge permits that are also subject to such an analysis [see, for example 5 CCR 1002-31, §31.14(3)] Finally, as is apparent from the language of §102, an economic evaluation under §102 is not intended to change the result of the action being undertaken, but merely to document the consequences, both positive and negative. Accordingly, the Division cannot, for example, issue a permit that will result in a violation of the standards, regardless of the results of the economic evaluation. 5 CCR 1002-62, §61.8(1)(e); §61.8(3)(f) To summarize, §102 is not a substantive provision, but is implemented through other specific provisions of the Act and regulations. Similarly a TMDL in itself does not require any specific actions, and since discharge permits will implement standards already adopted by the Commission after evaluation of all the factors listed in §102, recourse is with the Commission.
Colorado WQCD
36
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Daily Maximum or Monthly Average Limits: Several parties commented on the time-step and averaging periods used in the model and questioned what averaging period would be used in the implementation of the TMDL in discharge permits. The previous version of the TMDL was unclear about how the waste load allocations would be implemented in discharge permits. Conditions in the River change throughout the year: Flow, seepage, temperature and other factors change throughout the year in response to seasonal factors. It is the Division’s general practice to represent these fluctuations on a monthly basis. Monthly medians or averages are appropriately used to characterize the factors that are considered in the model. The water supply standard must be met on a daily basis: Regulation 31 (at 31.16, Table II) specifies that the water supply standard for nitrate is a one-day standard to be met at the point of water supply intake. It is appropriate to implement the TMDL as a daily restriction on the discharge of nitrogen. Storage in Tani Lakes makes Monthly Average Limits more Appropriate: One commenter asserts that since Thornton diverts the water and generally stores it in the Tani Lakes for several months, a daily maximum nitrogen limit is unnecessary, and the cost to comply is unreasonable. The point of intake to Thornton’s water supply system is when it leaves state waters, in other words, at the intake structure located approximately three miles downstream on the Burlington Ditch. The Tani Lakes (Thornton’s storage reservoirs) are not waters of the state. When the water reaches the Tani Lakes, it has already been withdrawn for use and treatment. In addition, the Lakes are lined and there is no hydrologic connection to surface or ground waters. Until and unless the Water Quality Control Commission sets a site-specific nitrate standard that has a different implementation period, the provisions in Regulation 31 must be followed. Post implementation TMDL evaluations should be documented and made available for public review: The Division anticipates that the water quality of the South Platte River in segment 14 will continue to be evaluated at regular intervals. At a minimum, the water quality will be assessed in conjunction with each regular review of the water quality classifications and standards. Additionally, it may be reviewed every two years in each Section 303(d) Listing cycle. The Division is also open to accepting and reviewing water quality data at any point in the post-implementation period that would suggest that the TMDL is not meeting its water quality targets and goals, if Division resources are available for such a review. Wasteload allocations should be incorporated into discharge permits as quickly as possible: The Division anticipates incorporating this TMDL into the Glendale, Centennial and L/E discharge permits at the time of permit renewal. Glendale’s current permit expires on July 31, 2008; Centennial’s expires on December 31, 2003, and L/E’s expires on March 30, 2006. Compliance schedules may be included in permits to provide time for construction of plant modifications necessary to comply with the permit conditions.
Colorado WQCD
37
April 21 2004
So. Platte seg 14 Nitrate TMDL
XI.
Final
Responsiveness Summary - Third Public Comment Draft
The Division’s intent to finalize the TMDL was published in July, 2003. Comments were received from the city of Thornton, The Cities of Littleton and Englewood, the Farmers Reservoir and Irrigation Company and the Centennial Water and Sanitation District. Response to Comments and Explanation of Changes to the Model Substantive written comments were received from the parties. One of the comments resulted in changes in the model. The issues are addressed below. A. Revisiting the Flow Reset Options: One commenter asked that the Flow Reset options be revisited and an intermediate solution be considered, since neither the “on” nor “off” options adequately characterize flows that are realistic or historic. The Division agreed to review the reset conditions. A “Modified Reset” approach was developed that adjusts flow of the river back to the historical low flow only above the L/E discharge point. This is discussed on page 11 of the TMDL and displayed in Figure 7 and Table 3. Comparison of the resulting flow scenarios and the modeling results shows that the Modified Reset option provides results between the Off and On approaches and reflects a more realistic flow scenario. The Reset On option removes flow (and nitrogen) just above L/E’s discharge, then removes much of L/E’s flow at Public Service Arapahoe just 0.4 miles downstream of L/E’s discharge. Using the modified reset approach allows L/E’s full design capacity to travel down the river to the Burlington headgate. The model was changed in the following way to accomplish The Modified Rest approach: 1. 2.
3.
Select the “SouthPlatte” tab in the TMDL model Turn off the Reset at the Public Service Company Arapahoe facility. (This is labeled as “PSC – Arapahoe, linked to diversion” in the TMDL model at cell AG60.) Change cell AG60 from an If Statement to AG59 + X59 + O59, which allows upstream flows to continue to accumulate, and not be reset to the DFLOW value. Turn off the Reset at the Public Service Company Zuni facility. (This is labeled as “PSC – Zuni (Special Calcs)” in the TMDL model at cell AG84.) Change cell AG84 from an If Statement to AG83 + X83 + O83, which allows upstream flows to continue to accumulate, and not be reset to the DFLOW value.
Allowing L/E’s flow to pass the Arapaho and Zuni power plants results in more flow in the river. This has a direct effect on the loss of nitrate, which in the model is a function of flow. More flow translates to more nitrate loss. (It also translates to faster travel time, but that has a more modest effect on the results.) The Modified Reset approach was used in the TMDL calculations. The net effect of the change in the reset approach is to change the most restrictive numeric TIN value for L/E from 20.0 mg/L to 22.6 mg/L for March
Colorado WQCD
38
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
B. Effects on Downstream Water Bodies One commenter requests that the Division provide some level of assurance that the current nitrate TMDL can be re-evaluated and modified on a timely basis as necessary to accommodate future TMDLs or nutrient standards in the downstream receiving waters, including Barr Lake If standards are not attained in a downstream segment (including Barr Lake), all significant sources of the pollutant in question will be evaluated in a TMDL for that segment. If the analysis shows that further reductions in nitrogen are required from the sources that were limited in this TMDL, such reductions would supercede the requirements of this TMDL. The timing on re-evaluation will depend on the relative priorities of the TMDLs that the Division is working on. Priorities are set at the time of the promulgation of the 303(d) List. C. Effect of Developing More Site-specific Information: One commenter requests that language be added to the TMDL allowing Centennial Water and Sanitation District to develop site-specific ammonia removal factors for its permit and not be tied to the total ammonia limit included in the TMDL. Whenever the fundamental assumptions upon which this TMDL are based are changed, the Division will be obligated to assess the effects of the revised assumptions on the results of the TMDL. Nitrogen loss rates are intrinsic to the model and any rates that are substantially different than modeled rates will most likely have an effect on one or both the target points (Allen Treatment Plant Diversion and the Burlington Headgate). The Division can give no such assurance that the TMDL will not have to be revisited when Centennial develops site-specific ammonia removal factors. D. Incorporation into Discharge Permits: One commenter asked what the barriers are to opening the discharge permits and revision effluent limits as soon as the TMDL has been formally accepted and approved. The water quality model is set up to evaluate the decay of both ammonia and nitrate and, as such, the predicted ammonia and nitrate concentrations at a downstream location are dependent on the amount of ammonia and nitrate in the discharge. Therefore, values for both total ammonia and TIN are required to be input into the model. Tables 9,16 and 19 show applicable TIN concentrations for the discharge from the Centennial W&SD WWTP at two assumed total ammonia concentrations (0 mg/l and 15 mg/l). Since the TIN values are linked to their respective total ammonia value in the model, these two parameters must be evaluated together when determining reasonable potential. Since the model is evaluating the downstream total nitrate concentration using the TIN and ammonia values, there is no need to input a nitrate value into the model and a reasonable potential determination for nitrate will not be necessary. A reasonable potential evaluation will be conducted for total ammonia and TIN. The TMDL will be revised to provide clear direction on this point. E. Implementation in Discharge Permits: One commenter asked for the TMDL to include language in that directs permit drafters on how to interpret the values in the TMDL, such that reasonable potential for the TIN values to exceeded should be evaluated separately from total ammonia.
Colorado WQCD
39
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
There is no statutory barrier to reopening a permit and incorporating the provisions of a TMDL. The Division included a schedule of compliance for construction of additional facilities necessary to meet the requirements of the TMDL in the permit for the Littleton/Englewood WWTP. We have their application for approval to construct those facilities before us for approval ahead of the schedule required in their permit. According to their construction schedule the remainder of the engineering and construction work will be moving forward at or ahead of the pace required in the permit. The Division will evaluate the need to include similar schedules in the permits for the Centennial Water and Sanitation District and the City of Glendale. If either entity is determined to be discharging a combination of concentrations of ammonia and TIN that would contribute significantly to exceedances of the nitrate standard at the Allen Treatment Plant intake or at the Burlington Canal headgate, the Division will reopen the permit and include a schedule for meeting the limits consistent with the concentrations identified in the TMDL. F. Recent Nitrate and Flow Levels at the Burlington Headgate: Two commenters requested clarification of the actual status of the nitrate levels at the Burlington Canal. One commenter asked that the recent data be discussed in terms of the recent low flows in the South Platte. The modeling is based on information gathered throughout the study area from September 1998 through June 2000 plus data fro 2002, when flows were low. Since that time, more data has been gathered. The 2002 data was especially helpful in developing flow-related nitrate loss rates for the reaches below L/E. Nitrate: Nitrate levels at the Burlington Headgate and flows at the Denver Gage (approximately 3.2 miles upstream of the Headgate) are presented in Figure 10. The reported nitrate level of 13.7 mg/L from March 17, 1999 was probably a measurement error and was not used in the calibration or validation of the modeling. The five measurements from November 2002 to March 2003 that exceed the 10 mg/L standard occurred when the flow at the Denver Gage was between 54 and 76 cfs.
40
10000
30
1000
20
100
10
10
0
Flow (cfs)
Nitrate (mg/L)
Nitrate and Flow in the South Platte River
1 03 02 18 16 17 17 15 04 05 15 03 02 04 05 06 -S -D -A -F -N -M -A -D -D -N -M -J -J -J -J un un un un eb ep ug ug ec ov ec ec ov ay ar -9 -0 -0 -9 -9 -9 -9 -9 02 01 02 01 00 00 00 9 3 0 9 8 8 9 9
Sample Date Nitrate
Figure 10.
Colorado WQCD
Flow
Nitrate levels at the Burlington Headgate and Flows at the Denver Gage, Sept. 1998 to Sept 2003. 40
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
G. Drought and Flow: Questions have been raised about Colorado’s recent drought and the effect that it has had on South Platte River flows in the metropolitan area. The river and water deliveries are highly managed in this area, and record low precipitation does not necessarily equate to record low stream flow. In order to better understand the flow regime of 2002 and 2003, the Division has assessed historic flows at Denver Gage (USGS gage number 06714000) which is located approximately 3.2 miles upstream of the Burlington Headgate. The Division uses DFLOW4 software to develop critical low-flow estimates for permitting purposes. The acute critical low flow is the empirically based 1-day low flow with an average 1in-3 year recurrence interval (known as the “1E3”). This software does not return the lowest flow in the period of analysis. For the TMDL modeling, DFLOW4 was used to develop low-flow estimates for the South Platte River below Chatfield, the South Platte River above L/E WWTP, the South Platte River above Xcel Arapahoe and the South Platte River above Xcel Zuni. For each of these locations, the monthly acute DFLOW values were obtained using the gage records for the interval October 1, 1990 to September 30, 2000. Table 3 in the TMDL report gives the acute low-flow values that were developed for the model. Table 15 present the results of using DFLOW4 and the records for the Denver Gage for various study periods. 1910 to Present
Study Period 1970 to Present
1983 to Present
94 years 24 cfs
34 years 47 cfs
21 years 50 cfs
Jan 1935 Mar 1951 Apr 1925 May 1963 July 1955 Sep 1954
Mar 1978 Apr 1978 July 2002 Aug 2002
Aug 2002 Sep 2002
Lowest recorded daily flow
8.8 cfs on Mar 3, 1951; (53 other days with flow below 24 cfs, none of which occurred in 2002))
26.5 cfs on Aug 2, 2002; (21 other days in July, Aug and Sept 2002 had flows below 50 cfs)
Comments
The low flows of the 2002-03 drought do not control the 1E3 for this length of record. The 1925, 1935, 1954-55 and 1963 low flows were much lower than the 2002-03 flows
26.5 cfs on Aug 2, 2002; (20 other days in Mar and Apr of 1978 and 1982 and July and Aug of 2002 had flows below 47 cfs) The low flows of the 2002-03 drought do not control the 1E3 for this length of record. The spring flows in the 197779 drought were lower than the spring flows of 2002-03 drought. The fall low flows of the 2002-03 drought were lower than the 77-79 drought
No of Years Critical low flow (DFLOW4) Month and year that critical flow occurred
Table 15.
Colorado WQCD
The low flows of the 2002-03 drought do control the 1E3 generated using the last 20 to 25 years of record.
Critical Low Flow (1E3) and Historic Lowest Flows in cfs at the Denver Gage (USGS No 06714000)
41
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
H. Alternative TMDLs - Alternate Effluent Flow for L/E: Littleton/Englewood WWTP requested that the Division develop the TMDL for variable effluent flow conditions. They requested allocations for effluent discharges of 34 MGD and 42 MGD. in addition to the 50 MGD used in the modeling. The TMDL was re-modeled for these two discharge conditions. 1.
TMDL with L/E at 34 MGD
The TMDL model was used to develop allocations for a 34-MGD discharge by L/E. Centennial and Glendale discharges were not adjusted. The Modified Reset approach as described above was used. The Allen Treatment Plant and Burlington Ditch headgate nitrate targets were set at 9.8 mg/L and 10.0 mg/L respectively. The L/E discharge capacity was set to 34 MGD by selecting the “Flow “tab in the TMDL model and changing the L/E WWTP design flow to 34 MGD, at cell P24. Initial modeling of 0.0 and 15.0 mg/L ammonia were completed. Centennial’s ammonia was left unchanged from the results of the 50-MGD TMDL (see Table 9). Glendale’s ammonia was set the same as L/E’s . Table 16 presents the comparison of TIN for the two ammonia levels
Source
Jan
Feb
Centennial Littleton/Englewood Glendale*
29.7 24.8 40.0
27.7 24.5 40.0
Centennial Littleton/Englewood Glendale*
30.5 26.6 40.0
27.8 26.2 40.0
Allen Plant Burlington
9.8
9.8
10.0
10.0
L/E at 34 MGD Effluent TIN, mg/L assuming 0.0 mg/L Ammonia Mar Apr May Jun Jul Aug Sep Oct 28.8 31.8 44.8 43.1 49.4 48.0 43.3 40.8 23.5 27.1 47.5 51.3 45.7 50.1 39.3 35.1 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Effluent TIN, mg/L assuming 15.0 mg/L Ammonia 29.2 32.4 46.5 43.4 48.3 46.7 42.1 40.6 25.2 27.6 44.8 47.9 42.2 46.5 36.2 32.9 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Target Points, Nitrate-Nitrogen (mg/L)
Nov
Dec
36.0 28.6 40.0
31.9 29.3 40.0
35.5 28.5 40.0
32.4 30.3 40.0
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
15.0
0.0
15.0
0.0
15.0
0.0
Selected Discharged Ammonia Concentrations (mg/L) 0.0 0.0 0.0 0.0 0.0 0.0 15.0 15.0 15.0 15.0 Centennial 0.0 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 Littleton/Englewood 0.0 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 Glendale* * Glendale WWTP TIN set at 40 mg/L, ammonia discharge set equal to L/E Boxes identify which scenario (0.0 or 15.0) is protective for each month for each plant.
Table 16.
Colorado WQCD
For L/E at 34 MGD Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia.
42
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Table 17 presents the modeled flow (in cfs) and concentrations of nitrate (in mg/L) at selected locations in Segment 14. Table 18 presents the 34-MGD TMDL in terms of monthly TIN load in kg/d.
L/E at 34 MGD Month Location
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
1
0
0
0
0
Discharge (cfs) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 34 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
0
0
0
1
5
2
0
1
0
1
3
8
6
8
5
1
2
0
1
2
1
2
4
15
10
10
8
3
4
2
2
13
13
13
13
13
13
13
13
13
13
13
13
30
27
28
31
46
42
45
44
40
39
34
31
26
22
23
27
32
31
26
35
36
37
34
29
19
18
15
18
28
13
7
10
12
18
19
25
26
27
25
28
60
58
30
33
20
27
31
39
53
53
53
53
53
53
53
53
53
53
53
53
0
0
0
0
0
0
1
1
1
0
0
0
93
94
92
95
129
132
106
111
96
100
99
95
5
2
2
6
11
17
16
13
16
10
8
5
3
3
3
3
3
3
3
3
3
3
3
3
11
8
8
8
21
13
13
13
11
12
8
11
92
91
85
92
140
136
112
121
106
110
98
92
0
0
0
0
0
0
0
0
Nitrate Concentration (mg/L) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 34 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
1
1
1
30
28
29
32
45
43
33
32
27
26
21
32
10
10
10
10
10
10
10
10
10
10
10
10
8
8
8
8
8
9
9
9
9
8
8
8
1
1
1
1
1
0
0
0
1
1
1
1
4
4
4
4
4
5
5
5
5
5
5
4
25
24
24
27
30
33
27
32
21
18
14
27
-*
12
-*
12
12
12
12
12
12
12
12
-*
11
11
11
12
13
13
13
14
13
12
11
12
2
2
2
2
3
3
3
3
3
3
3
2
40
40
40
40
25
25
25
25
25
25
25
40
10
12
13
7
6
6
6
6
6
6
7
8
10
10
10
10
10
10
10
10
10
10
10
10
* These values are flow-weighted averages of two locations. When there is no flow, an average cannot be calculated
Table 17.
Colorado WQCD
L/E at 34 MGD: Summary of flows and concentrations of nitrate N to attain water quality targets as predicted by the TMDL model.
43
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
L/E at 34 MGD Month Source
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
WASTE LOAD ALLOCATIONS (kg/d TIN) WASTEWATER TREATMENT PLANTS
Centennial Littleton/Englewood Glendale
956
892
927
1024
1443
1388
1555
1504
1356
1307
1143
1027
3194
3155
3027
3181
5783
6182
5435
5989
4662
4263
3683
3774
303
303
303
303
303
303
303
303
303
303
303
303
LOAD ALLOCATIONS (kg/d TIN) GROUNDWATER SOURCES
Seepage South Platte
76
56
58
78
112
161
176
227
217
198
133
101
Cherry Creek
88
51
48
131
171
208
200
178
209
171
120
91
Bear Creek
31
31
25
31
13
19
19
31
36
42
36
31
0
1
0
1
5
14
18
23
18
12
2
0
Contaminated Sites
UNUGAGED SURFACE WATER SOURCES
Little Dry Creek
15
12
10
10
10
10
9
10
11
11
14
14
Sanderson Gulch
12
12
12
12
12
12
12
12
12
12
12
12
Lakewood Gulch
20
20
20
20
20
20
20
20
20
20
20
20
W Harvard Gulch
2
2
2
2
2
2
2
2
2
2
2
2
73
73
73
73
73
73
73
73
73
73
73
73
All others UPSTREAM SOURCES
SP blw Chatfield
0
.1
0
0
1.6
.3
0
0
0
0
0
0
Bear Cr blw Res
29
26
13
17
31
4
1
4
2
14
19
43
0
0
0
0
0
0
0
0
0
0
0
0
Cherry Cr blw Res.
Table 18.
L/E at 34 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN
for sources using TMDL modeling assumption.
2.
TMDL with L/E at 42 MGD
The TMDL model was used to develop allocations for a 42-MGD discharge by L/E. Centennial and Glendale discharges were not adjusted. The Modified Reset approach as described above was used. The Allen Treatment Plant and Burlington Ditch headgate nitrate targets were set at 9.8 mg/L and 10.0 mg/L respectively. The L/E discharge capacity was set to 42 MGD by selecting the “Flow “tab in the TMDL model and changing the L/E WWTP design flow to 42 MGD, at cell P24. Initial modeling of 0.0 and 15.0 mg/L ammonia were completed. Centennial’s ammonia was left unchanged from the results of the 50-MGD TMDL (see Table 9). Glendale’s ammonia was set the same as L/E’s . Table 19 presents the comparison of TIN for the two ammonia levels.
Colorado WQCD
44
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
L/E at 42 MGD Effluent TIN, mg/L assuming 0.0 mg/L Ammonia Mar Apr May Jun Jul Aug Sep Oct
Source
Jan
Feb
Centennial Littleton/Englewood Glendale*
29.7 23.9 40.0
27.7 23.7 40.0
Centennial Littleton/Englewood Glendale*
30.5 25.3 40.0
27.8 24.9 40.0
Allen Plant Burlington
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
9.8
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
Nov
Dec
36.0 27.5 40.0
31.9 27.7 40.0
35.5 26.9 40.0
32.4 28.3 40.0
9.8
9.8
9.8
10.0
10.0
10.0
28.8 31.8 44.8 43.1 49.4 48.0 43.3 40.8 23 26.1 42.7 46.4 42.7 46.3 37.4 33.2 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Effluent TIN, mg/L assuming 15.0 mg/L Ammonia 29.2 32.4 46.5 43.4 48.3 46.7 42.1 40.6 24.2 26.0 40.0 43.0 39.1 42.6 34.1 30.9 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Target Points, Nitrate-Nitrogen (mg/L)
Selected Discharged Ammonia Concentrations (mg/L) 0.0 0.0 0.0 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 Centennial 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 Littleton/Englewood 0.0 0.0 0.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 Glendale* * Glendale WWTP TIN set at 40 mg/L, ammonia discharge set equal to L/E Boxes identify which scenario (0.0 or 15.0) is protective for each month for each plant.
Table 19.
0.0 0.0 0.0
For L/E at 42 MGD: Comparison of TIN Results Using 0.0 mg/L and 15.0 mg/L Ammonia.
Table 20 presents the modeled flow (in cfs) and concentrations of nitrate (in mg/L) at selected locations in Segment 14. Table 21 presents the 42-MGD TMDL in terms of monthly TIN load in kg/d. L/E at 42 MGD Month Location
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
1
0
0
0
0
Discharge (cfs) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 42 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
Colorado WQCD
0
0
0
1
5
2
0
1
0
1
3
8
6
8
5
1
2
0
1
2
1
2
4
15
10
10
8
3
4
2
2
13
13
13
13
13
13
13
13
13
13
13
13
30
27
28
31
46
42
45
44
40
39
34
31
26
22
23
27
32
31
26
35
36
37
34
29
19
18
15
18
28
13
7
10
12
18
19
25
26
27
25
28
60
58
30
33
20
27
31
39
65
65
65
65
65
65
65
65
65
65
65
65
0
0
0
0
0
0
1
1
1
0
0
0
105
106
104
108
142
144
118
124
108
112
111
118
5
2
2
6
11
17
16
13
16
10
8
5
3
3
3
3
3
3
3
3
3
3
3
3
11
8
8
8
21
13
13
13
11
12
8
11
105
103
98
104
152
148
124
133
118
122
110
119
45
April 21 2004
So. Platte seg 14 Nitrate TMDL
Final
Nitrate Concentration (mg/L) Chatfield DOW Fish Unit SP abv Centennial Centennial WWTP SP abv Allen Plant SP abv Bear Creek Bear Creek mouth SP abv L/E L/E WWTP at 42 MGD Denver Sites SP abv Cherry Cr Cherry Cr abv Glendale Glendale WWTP Cherry Cr mouth SP at Burlington
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
1
1
1
30
28
29
32
45
43
33
32
27
26
21
32
10
10
10
10
10
10
10
10
10
10
10
10
8
8
8
8
8
9
9
9
9
8
8
8
1
1
1
1
1
0
0
0
1
1
1
1
4
4
4
4
4
5
5
5
5
5
5
4
23
23
23
10
40
42
22
25
18
14
11
12
-*
12
-*
12
12
12
12
12
12
12
12
-*
11
11
11
11
12
13
13
14
13
12
11
12
2
2
2
2
3
3
3
3
3
3
3
2
40
40
40
40
25
25
25
25
25
25
40
40
10
12
13
7
6
6
6
6
6
6
7
8
10
10
10
10
10
10
10
10
10
10
10
10
* These values are flow-weighted averages of two locations. When there is no flow, an average cannot be calculated
Table 20. L/E at 42 MGD Summary of flows and concentrations of nitrate N to attain water quality targets as predicted by the TMDL model. L/E at 42 MGD Month Source
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
WASTE LOAD ALLOCATIONS (kg/d TIN) WASTEWATER TREATMENT PLANTS
Centennial Littleton/Englewood Glendale
956
892
927
1024
1443
1388
1555
1504
1356
1307
1143
1027
3802
3771
3659
3930
6364
6841
6221
6777
5425
4916
4280
4407
303
303
303
303
303
303
303
303
303
303
303
303
101
LOAD ALLOCATIONS (kg/d TIN) GROUNDWATER SOURCES
Seepage South Platte
76
56
58
78
112
161
176
227
217
198
133
Cherry Creek
88
51
48
131
171
208
200
178
209
171
120
91
Bear Creek
31
31
25
31
13
19
19
31
36
42
36
31
0
1
0
1
5
14
18
23
18
12
2
0
Contaminated Sites
UNGAGED SURFACE WATER SOURCES
Little Dry Creek
15
12
10
10
10
10
9
10
11
11
14
14
Sanderson Gulch
12
12
12
12
12
12
12
12
12
12
12
12
Lakewood Gulch
20
20
20
20
20
20
20
20
20
20
20
20
W Harvard Gulch
2
2
2
2
2
2
2
2
2
2
2
2
73
73
73
73
73
73
73
73
73
73
73
73
SP blw Chatfield
0
.1
0
0
1.6
.3
0
0
0
0
0
0
Bear Cr blw Res
29
26
13
17
31
4
1
4
2
14
19
43
0
0
0
0
0
0
0
0
0
0
0
0
All others UPSTREAM SOURCES
Cherry Cr blw Res.
Table 21.
L/E at 42 MGD: Monthly Waste Load Allocations and Load Allocation in kg/d TIN
for sources using TMDL modeling assumption. Colorado WQCD
46
April 21 2004
