TCMA Chloride TMDL - TMDL Development TCMA Chloride TMDL - TMDL Development

TMDL Development

This section presents the methodology used to develop the TCMA chloride TMDLs and the resulting load capacity and various components of the TMDL, including load allocations (LA), wasteload allocations (WLA), MOS, seasonal variation, and future growth/reserve capacity.

Chloride TMDL Methodology

The TMDLs were developed for each of the lakes, wetlands and streams in the TCMA impaired by chloride. A TMDL quantifies the allowable pollutant loading to a lake or stream that will result in water quality standards being attained. The WQT for the TMDLs was set to the chronic water quality criterion for chloride of 230 mg/L. The total allowable load, or the TMDL, is allocated to the various sources contributing chloride as well as a MOS and, in general, a RC. The TMDL equation can be written as:

TMDL = WLA + LA + MOS + RC

Where:

WLA = wasteload allocation for permitted sources, including MS4s and treatment facilities

LA = load allocation for natural background and other non-permitted sources (mainly, runoff from rural and non-permitted areas)

MOS = margin of safety

RC = reserve capacity

Several approaches were considered for developing the TMDLs. A simple 0-dimensional, steady-state modeling approached was selected through consultation with the Technical Advisory Committee for calculating the allowable load from runoff, including the permitted MS4 areas and non-permitted areas. This approach assumes that chloride from winter maintenance activities and all other sources eventually makes its way to surface waterbodies through runoff. This approach was chosen for the following reasons: 1) chloride is a conservative substance and is in the dissolved phase in the water environment; therefore, complex fate and transport assessments are not needed; 2) determining the time for a system to respond to reduced chloride loads was not necessary to inform the TMDL or the CMP; and 3) the large number of lakes and streams needing a TMDL and the limited data available for a significant portion of them prohibited a more complex approach. This approach assumes eventual complete flushing in an impaired waterbody over the long-term.

The WQT for the waterbodies included in this TMDL is Minnesota’s chronic water quality standard for chloride, 230 mg/L. On this basis, the TMDL components were calculated as described below, with additional discussion following later in this section. It should be noted that the “WLA for MS4 areas” and “LA for runoff from non-permitted areas” are exclusive and do not overlap.

Total allowable runoff load = runoff volumeTOTAL x WQT

Margin of Safety (MOS) = 10% of the total allowable runoff load (both MS4 and non-permitted areas)

= 10% x runoff volume x WQT

LA for natural background sources (LABG) = runoff volumeTOTAL x natural background concentration

WLA for MS4 areas (WLAMS4) = runoff volumeMS4 x WQT - LABG - MOS

WLA for WWTPs (WLAWWTP) = WWTP design flow x WQT

LA for runoff from non-permitted areas (LAnon-permitted )= runoff volumenon-permitted x WQT - LABG - MOS

Reserve Capacity (RC) = set to zero for this TMDL

In light of the above, the Metro Area chloride TMDL is more explicitly expressed as below:

TMDL = WLAMS4 + WLAWWTP + LAnon-permitted + LABG + MOS

A simple schematic of the modeling approach for lakes and wetlands is shown in the Model schematic for allowable runoff load for stream TMDLs in the TCMA chart and for streams in the Model schematic for allowable runoff load for stream TMDLs in the TCMA chart. The primary differences between the approaches for lakes and streams was that annual runoff was considered for lakes because of the longer retention times in lakes and subsequent mixing, whereas only winter-season runoff was considered for streams. Since the winter-season runoff was considered for the streams, the runoff coefficients were set to 0.98 to account for frozen ground conditions based on best professional judgment. The basic premise of this approach is to constrain runoff from having greater than 230 mg/L of chloride on average throughout the year in an impaired lake, and throughout the winter and spring snow melt season in an impaired stream. To express the TMDLs on an average daily basis, the annual lake and seasonal stream allowable loadings are divided by 365 days per year and by 151 days per winter season (November-March), respectively (see below).

Model schematic for allowable runoff load for lake and wetland TMDLs in the TCMA

Model schematic for allowable runoff load for stream TMDLs in the TCMA

Modeling approach – Lakes and Wetlands

The 0-dimensional modeling approach for lakes and wetlands takes into account the total tributary watershed area, percentage of impervious surface within the watershed area, and average annual precipitation. Based on these variables, an average annual runoff was calculated using the Simple Method (Schueler, 1987). The allowable runoff load was then calculated by multiplying the average annual runoff by the chronic water quality standard for chloride (230 mg/L).

Runoff coefficient

𝑅_𝑣=0.05+0.9×𝐼𝐼𝑎𝑎

Where: I_a = impervious fraction

R_v = runoff coefficient

Average annual runoff

𝑅_v = 𝑃 × 𝑃_𝑗 × 𝑅_𝑣

Where: P = Annual rainfall (inches/yr) (30.6 inches/year for the TCMA)

P_j = Fraction of annual rainfall events that produce runoff (usually 0.9)

R_v = runoff coefficient

R = Average annual runoff (inches)

Average annual runoff volume

𝑄_𝑣 = 𝐴 × 𝑅 / 12

Where: R = Annual runoff (inches/yr)

A = total tributary watershed area (acres)

Q_v = average annual runoff volume (ac-ft/yr)

Allowable Annual Runoff Load (L, lbs/year)

𝐿 = 𝑄_𝑣 × 𝐶 × 2.72

Where: Qv = average annual runoff volume (ac-ft/yr)

C = chronic water quality standard (230 mg/L)

2.72 = conversion factor

Modeling approach - Streams

A slightly modified approach was taken for the streams. The streams tend to exhibit the highest chloride concentrations during the spring snowmelt, which is then flushed through the system. The approach was modified to account for frozen ground conditions and seasonal runoff volume. The runoff coefficient was adjusted to 0.98 over the entire tributary watershed area to account for frozen ground conditions. The seasonal runoff volume is considered to be the total precipitation equivalent for the period from November 1 through March 31 (season is 151 days per year). This period is typically when salt is being applied and is expected to accumulate and run off during the spring snowmelt (as well as occasional winter melts). A seasonal precipitation equivalent for the period of November 1 through

March 31 was determined to be 6.29 inches based on the UMN climate data for the period of record from 1981-2010.

This approach was used to determine the loading capacity for streams. The calculation is shown below.

Runoff coefficient

R_v = 0.98 (frozen ground conditions)

Average seasonal runoff

𝑅 = 𝑃 × 𝑅_𝑣

Where: P = Seasonal precipitation water equivalent (6.29 inches/season for the TCMA)

R_v = 0.98 (frozen ground conditions)

R = Average seasonal runoff (inches)

Average seasonal runoff volume

𝑄_𝑣= 𝐴 × 𝑅 / 12

Where: R = Seasonal runoff (inches/season)

A = total tributary watershed area (acres)

Q_v = average seasonal runoff volume (ac-ft/season)

Allowable Seasonal Runoff Load (L, lbs/yr)

𝐿=𝑄_𝑣 ×𝐶 × 2.72

Where: Q_v = average seasonal runoff volume (ac-ft/season)

C = chronic water quality standard (230 mg/L)

2.72 = conversion factor

Wasteload Allocation Methodology

MS4 Wasteload Allocation - Runoff

A categorical WLA has been established for the permitted MS4s within each impaired watershed. The runoff loading capacity L, after deducting the natural background and 10% MOS, was split between the permitted and non-permitted parts of the watershed in simple proportion to their runoff volumes.

Wasteload Allocation – Wastewater Sources

Wastewater source discharges were included in the WLA where applicable. The allowable load for treatment facilities within an impaired watershed was set to the expected discharge or design flow of the facility multiplied by 230 mg/L of chloride and a units conversion factor as follows: WLA = 0.012 x Q x C

Where: WLA = Wasteload Allocation (lbs/day)

Q = mean discharge (gpm)

C = chronic water quality standard (230 mg/L)

0.012 = conversion factor

An alum (aluminum sulfate) treatment plant for stormwater can be considered a wastewater source discharge. However, alum treatment does not contribute chloride to the system. The alum treatment facility is a pass-through for stormwater that already contains chloride. Since the chloride source is the MS4, the WLA has already been assigned to the MS4 and the alum treatment facility does not require an individual chlorideWLA. Use of aluminum chloride or ferric chloride for treatment of lake sediments or stormwater should be avoided in waters and watersheds with chloride impairments.

Other

The WLAs for regulated construction stormwater (MNR10001) were not developed since chloride is not a typical pollutant from construction sites.

The WLAs for regulated industrial stormwater were also not developed. Industrial stormwater must receive a WLA only if the pollutant is part of benchmark monitoring for an industrial site in the watershed of an impaired waterbody (as detailed in the MPCA’s June 8, 2001, memo). There are no chloride benchmarks associated with the Industrial Stormwater Permit (MNR050000).

Permitted entities located in more than one chloride impaired nested watershed; therefore, receiving multiple WLAs for chloride will be required to meet the most stringent downstream WLA. This approach assumes that by achieving the most stringent WLA all the others will also be met.

Load Allocation Methodology

Natural background load allocation

Natural background loads of chloride were calculated by multiplying the watershed runoff by 18.7 mg/L, the natural background concentration of chloride in TCMA streams estimated by Stefan et al. (2008).

Non-permitted runoff load allocation

The allowable runoff load from anthropogenic sources was calculated by subtracting the natural background load and the MOS from the allowable runoff load. The allowable load in runoff from anthropogenic sources was then divided between MS4 and non-permitted runoff based on the amount of runoff coming from each associated area within the impaired watershed.

An aggregate LA has been established for the non-permitted watershed runoff sources within each impaired watershed. This consists of townships, cities, counties, and MnDOT outside of the urban boundary and not covered under an MS4 permit. This aggregate LA includes winter maintenance activities in these areas as well as other potential sources including runoff from agricultural lands where fertilizer containing chloride may be applied, and the impact of septic systems on shallow groundwater and recharge.

This LA was calculated by the 0-dimensional modeling equation for the portions of the watershed that are outside the permitted MS4 areas. MOS and LA (natural background) were then subtracted out to get the LA (categorical non-permitted entities).

Margin of Safety

The MOS is the component of the TMDL allocation that accounts for uncertainty within the calculation methods, sample data, or the allocations which will result in attainment of water quality standards. For the purposes of developing the TMDLs for each lake, wetland and stream, an explicit 10% MOS was selected. This MOS was based on best professional judgment considering the potential variability of the monitored parameters from spatial, temporal, and seasonal changes seen within each lake and stream. Also, an explicit 10% is a reasonable estimate consistent with many other TMDLs prepared by the MPCA. It is reflective of the uncertainty in the data and the modeling. Almost all completed TMDLs for lakes in Minnesota make use of a 0-dimensional model and an explicit 10% MOS is typical. Implementation of the TMDL relies on an adaptive management approach that will revisit whether on-going efforts and the TMDL targets are sufficient to restore impaired waters.

Seasonal Variation

The TMDLs developed for lakes, wetlands and streams consider chloride sources from both seasonal sources, such as spring snowmelt and runoff, as well as continuous year-round sources of chloride such as the WWTPs. Historical loadings from salt application to impervious areas present in shallow groundwater may contribute chloride to surface waters throughout the year. See the TCMA Chloride TMDL - Watershed and Waterbody Characterization section of the TCMA Chloride Management Plan for more information about the impacts of chloride to groundwater. The TMDL for lakes assumes lake water quality responds to loadings on an annual or longer term basis. Therefore, the TMDLs for lakes have been developed to achieve an annual average daily load. Some impaired lakes indicate a seasonal trend, with higher chloride concentrations in winter and early spring. The MOS helps to protect for these seasonal variations. Continued monitoring and adaptive management will also be needed to ensure the TMDL is protective of the waterbody.

Chloride loadings to streams vary seasonally. Stream water quality responds to loadings on a seasonal basis and the highest chloride concentrations tend to occur during the spring snowmelt. Therefore, the TMDL has been developed to achieve compliance for the winter and spring snowmelt period.

TMDL Summary

A summary of the TMDLs is presented in the Summary of TMDL and Components for Impaired Lakes and Wetlands in the TCMA table below for lakes and wetlands and the Summary of TMDL (lbs/day) and Components for Impaired Streams in the TCMA table for streams. The TMDL for each waterbody including the individual MS4 and other wastewater source discharges within the WLA are presented in Appendix A-4.

Summary of TMDL and Components for Impaired Lakes and Wetlands in the TCMA
Link to this table

Lake/Wetland	AUID	Watershed Area (ac)	TMDL and Components (all values in lbs/yr of chloride)
			Loading Capacity (TMDL	WLA		LA		Margin of Safety
				MS4 Categorical	Wastewater Sources1	Non-Permitted Aggregate	Natural Background
Battle Creek Lake	82-0091-00	4,326	2,153,699	1,766,033	0	0	172,296	215,370
Brownie Lake	27-0038-00	452	341,418	279,963	0	0	27,313	34,142
Carver Lake	82-0166-00	2,242	1,071,123	878,321	0	0	85,690	107,112
Como Lake	62-0055-00	1,850	994,078	815,144	0	0	79,526	99,408
Diamond Lake	27-0022-00	744	486,017	398,534	0	0	38,881	48,602
Kasota Ponds North	62-0280-00	10	6,234	5,112	0	0	499	623
Kasota Ponds West	62-0281-00	6	5,742	4,708	0	0	459	574
Kohlman Lake	62-0006-00	7,533	4,839,183	3,106,733	1,050,484	0	303,096	378,870
Little Johanna Lake	62-0058-00	1,703	1,224,242	1,003,879	0	0	97,939	122,424
Loring Pond (South Bay)	27-0655-02	34	9,764	8,007	0	0	781	976
Mallard Marsh	62-0259-00	16	9,851	8,077	0	0	788	985
Parkers Lake	27-0107-00	1,064	1,431,262	528,161	787,163	0	51,528	64,410
Peavey Lake	27-0138-00	776	205,995	165,889	3,692	0	16,184	20,230
Pike Lake	62-0069-00	5,735	3,591,268	2,943,971	1,059	0	287,217	359,021
Powderhorn Lake	27-0014-00	332	218,588	179,242	0	0	17,487	21,859
Silver Lake	62-0083-00	655	370,011	303,409	0	0	29,601	37,001
South Long Lake	62-0067-02	114,785	26,334,624	21,534,261	4,030	0	2,106,448	2,633,059
Spring Lake	27-0654-00	39	15,600	12,792	0	0	1,248	1,560
Sweeney Lake	27-0035-01	2,439	1,456,271	1,194,142	0	0	116,502	145,627
Tanners Lake	82-0115-00	1,732	826,520	677,746	0	0	66,122	82,652
Thompson Lake	19-0048-00	178	134,340	110,159	0	0	10,747	13,434
Valentine Lake	62-0071-00	2,404	1,165,072	955,359	0	0	93,206	116,507
Wirth Lake	27-0037-00	426	1,095,000	897,900	0	0	87,600	109,500

¹ WLA=0 in the wastewater sources column means that there is no wastewater discharges in that watershed

Summary of TMDL (lbs/day) and Components for Impaired Streams in the TCMA
Link to this table

Stream	AUID	Watershed Area (ac)	TMDL and Components (all values in lbs/day of chloride)
			Loading Capacity (TMDL)	WLA		LA		Margin of Safety
				MS4 Categorical	Wastewater Sources 1	Non-Permitted Aggregate	Natural Background
Bass Creek	07010206-784	5,434	11,566	9,484	0	0	925	1,157
Bassett Creek	07010206-538	25,209	57,092	43,993	3,442	0	4,292	5,365
Battle Creek	07010206-592	7,246	15,422	12,646	0	0	1,234	1,542
Elm Creek	07010206-508	66,382	141,274	115,145	0	700	11,302	14,127
Judicial Ditch 2	07030005-525	1,587	3,378	2,770	0	0	270	338
Minnehaha Creek	07010206-539	109,151	235,279	189,928	3,537	0	18,584	23,230
Raven Stream	07020012-716	42,750	94,558	2,932	3,576	71,673	7,279	9,098
Raven Stream, East Branch	07020012-543	14,751	34,969	2,928	3,576	22,815	2,511	3,139
Rush Creek, South Fork	07010206-732	13,844	29,521	24,150	58	10	2,357	2,946
Sand Creek (South) - includes 07020012-662	07020012-513	175,578	382,821	29,156	9,154	277,251	29,893	37,367
Unnamed creek (Headwaters to Medicine Lk)	7010206-526	6,447	13,722	11,252	0	0	1,098	1,372
Unnamed creek (Unnamed ditch to wetland)	07010206-718	793	1,688	1,384	0	0	135	169
Unnamed Stream (Unnamed lk 62-0205-00 to Little Lk Johanna)	07010206-909	1,627	3,462	2,839	0	0	277	346

¹WLA=0 in the wastewater sources column means that there is no wastewater discharges in that watershed