TP and TSS credits and guidance for manufactured treatment devices (mtds)

This page provides total phosphorus (TP) and total suspended solid (TSS) removal credits (stormwater credit) for manufactured treatment devices (mtds). The page includes supporting information, guidance and recommendations, and links to additional information. Credits refer to the amount of pollutant reduced by treatment with a stormwater best management practice (BMP). For more information on credits, link here.

Definition and acronyms

A manufactured treatment device (mtd) is a pre-fabricated stormwater treatment structure utilizing settling (sedimentation), filtration, absorptive/adsorptive materials, vortex separation, vegetative components, and/or other appropriate technology to remove pollutants from stormwater runoff (New Jersey Department of Environmental Protection). MTDs are typically proprietary devices.

Acronyms

TP – total phosphorus
DP – dissolved phosphorus
OP – orthophosphorus
TSS – total suspended solids
TAPE - Technology Assessment Program – Ecology
TER – Technical Evaluation Report
GULD - General Use Level Designation
LCL – Lower confidence limit of the mean (usually the 95% LCL)
MTD (mtd) – manufactured treatment device

What manufactured treatment devices are credited?

Only Washington State Technology Assessment Protocol - Ecology (TAPE) devices that received General Use Level Designation (GULD) are credited for pollutant reductions. These devices typically are permanent structural practices providing primary treatment for pollutants. They are not considered pretreatment practices. Many devices approved through TAPE would be considered pretreatment practices in Minnesota and are not credited for pollutant removal in Minnesota since NPDES stormwater permits require pretreatment for infiltration and filtration primary treatment practices (see sections 16.6 and 17.5 of the Construction Stormwater general permit). For more information on pretreatment devices, link here.

Pollutant removal credits are provided for the following devices.

Device name	Manufacturer	TAPE approval date	Total suspended solids	Total phosphorus
BayFilter w/EMC Media	BaySaver Technologies, Inc.	7/10/19	Basic	Basic
BioPod Biofilter with Curb Inlet	Oldcastle Infrastructure, Inc.	10/28/19	Basic	Basic
ecoStorm plus	Watertectonics, Inc.	1/9/13	Basic
Filterra Bioscape	CONTECH Engineered Solutions, LLC.	9/16/19	Basic	Basic
Filterra System	CONTECH Engineered Solutions, LLC.	6/11/20	Basic	Tiered
FloGard Perk Filter	Oldcastle Infrastructure, Inc.	8/9/18	Basic	Basic
Media Filtration System	CONTECH Engineered Solutions, LLC.	11/15/16	Basic
MWS-Linear Modular Wetland	Bio Clean Environmental Services, Inc. (A Forterra Company)	12/16/19	Basic	Tiered
StormFilter using PhosphoSorb Media	CONTECH Engineered Solutions, LLC.	1/2/20	Basic	Tiered
Stormfilter using ZPG Media	CONTECH Engineered Solutions, LLC.	4/14/17	Basic
StormGarden Modular Stormwater Bio-filtration System	Environmental Solutions	8/28/19	Basic	Basic
The Kraken	Bio Clean Environmental Services, Inc. (A Forterra Company)	12/16/19	Basic	Basic
Up-Flo Filter w/Filter Ribbons	Hydro International	3/5/19	Basic	Basic

TP and TSS credits

To derive credits for GULD-approved mtds, data were compiled from Technical Evaluation Reports (TERs) for Washington State’s Technology Assessment Program – Ecology (TAPE). We compiled data for all storm events for each device. Not all data are included in TAPE’s analysis. For example, for TAPE analysis, inflow concentrations for TP must be 0.1 mg/L or greater. In our assessment, we included an analysis of all data and compared this with analysis of data used for TAPE analysis. A summary of the analysis is provided in the adjacent table. The lower confidence limit (LCL) was derived using Washington State’s bootstrap calculator. Detailed discussions for each practice are provided below.

Summary statistics for data collected for Washington State TAPE GULD certification. NOTE: These data are not to be applied to crediting for mtds in Minnesota. For mtd credits, link here
Link to this table

Manufactured treatment device	Median inflow TP (mg/L)	Median inflow TSS (mg/L)	TP Removal (%)	TSS removal (%)
			Number samples	Data with influent > 0.05 mg/L TP	TAPE analysis	All data	TAPE analysis
			Data with influent > 0.05 mg/L TP	Meeting TAPE criteria	Median	Mean	LCL	TAPE median	TAPE LCL	Median	Mean	LCL	TAPE median
Filterra System	0.090	43.5	17	10	69.0	66.8	59.0	78.5	69.2	89.5	86.2	82.8	89.3
Filterra Bioscape - see Filterra System
StormFilter using PhosphoSorb Media	0.280	389	17	16	81.5	77.4	69.9	78.7	63.6	91.6	88.1	84.8	89.0
Up-Flo Filter w/Filter Ribbons	0.109	29	20	13	50.3	41.5	31.0	57.1	49.9	75	67.6	59.9	77.5
MWS-Linear Modular Wetland	0.103	44	24	16	61.3	60.6	52.5	63.3	54.0	81.9	76.6	71.2	82.9
BayFilter w/EMC Media¹	0.160	53	12	11	68.6	64.0	59.8	68.6	58.5	89.3	86.1	81.1²	89.3
StormGarden Modular Stormwater Bio-filtration System	0.078	29	20	15	58.9	57.4	52.9	58.9	43.2	89.2	86.0	82.8	91.9
BioPod Biofilter	0.104	40	13	7	71.7	66.7	59.4	75.4	62.4	83.3	86.5	82.1	91.9
BoxlessBioPod Biofilter - see BioPod Biofilter
FloGard Perk Filter	0.123	62.5	21	15	66.5	62.3	53.0; 58.6 without outlier	70.7	48.8 (60.6 without outlier)	80.6	74.7	68.5	81.1
The Kraken	0.142	53	14	10	74.5	71.9	63.9	83.6	71.3	90.6	87.4	77.5	90.7
Jellyfish Filter	0.338	134	23	18	75.2	73.6	68.2	71.0	66.8	86.2	80.0	74.1	78.4
StormTree biofiltration	0.118	38	27	29	62	57	51	68	61	93	92	90	92
Media Filtration System	Though this device is GULD certified, it is not used in Minnesota. We therefore did not analyze the data.
Stormfilter using ZPG Media	Though this device is GULD certified, it is not used in Minnesota. We therefore did not analyze the data.
ecoStorm plus	We have not received data for this device
¹ Only data meeting TAPE inflow requirements was available. ² Six outflow concentrations were at the reporting level of 2.5 mg/L; we used half the reporting limit in our analysis

TAPE-approved practices receive 50 percent credit for TP and 80 percent credit for TSS. However, some devices appear capable of removing higher percentages of these pollutants. In Minnesota, the typical target for TP removal is 60 percent. We therefore propose a tiered credit system in which higher TP removal credits may be given provided certain conditions are met. A tiered approach to crediting allows flexibility in selecting practices and associated conditions that may affect performance, such as influent water quality and operation and maintenance. Data from the TERs indicate differing performance levels for each device. It is therefore not possible to develop uniform credits across all devices. It is not desirable to conduct an in depth analysis of each device. We therefore chose the following approach to crediting.

TSS credit: 80 percent. If the LCL exceeds 80 percent, the LCL is used but the credit cannot exceed 85 percent.
Tier 1 - TP
- Credit: 50 percent reduction
- Conditions for receiving credit: Inspect manufactured treatment device a minimum of twice during year 1. Conduct inspections following manufacturer’s instructions/recommendations. Use maintenance indicators provided by manufacturers to determine if maintenance is needed. Maintenance procedures should follow manufacturer’s guidelines. After year 1, follow manufacturers inspection, operation, and maintenance schedule and procedures unless otherwise required or unless year 1 inspections indicate a need for increased inspection and/or operation and maintenance.
Tier 2 - TP
- Credit: based on an assessment of 95% lower confidence limits (LCLs) for the device (see discussion below)
- Conditions for receiving credit:
  - Must meet Tier 1 conditions.
  - DP should be 25 percent or less of TP. If site specific monitoring data do not exist to verify this, acceptable land uses include commercial land use, industrial land use, transportation land use, highly urban and ultra-urban environments (ultra-urban), or combinations of these. The credit should not be applied in residential land use, to roof runoff, areas with extensive tree canopy, or other areas where dissolved phosphorus is likely to exceed 25 percent of total phosphorus unless site specific monitoring data, collected using appropriate protocols, shows DP:TP ratios are 0.25 or less (link here for discussion of dissolved phosphorus ratios).
  - Inspect manufactured treatment device monthly during the frost free season for the first year of operation after construction. Conduct inspections following manufacturer’s instructions/recommendations. Use maintenance indicators provided by manufacturers to determine if maintenance is needed. Maintenance procedures should follow manufacturer’s guidelines. Determine future inspection and maintenance schedules from findings during the first year of operation.
Tier 3 - TP
- Credit: based on monitoring for a specific device, except for devices with assigned Tier 3 credit values
- Conditions for receiving credit:
  - Must meet Tier 1 and Tier 2 conditions
  - Unless a specific Tier 3 numeric credit is established, demonstrate, via water quality monitoring, that the device meets the recommended removal credit. Monitoring data must be collected using appropriate protocols

The adjacent table provides TSS and TP removal credits for mtds. This credit only applies to the water treated by the device. To calculate annual pollutant mass removal, the removal credit must be multiplied by the volume treated. Link here for information on determining annual volume to apply credit.

Caution: Total phosphorus credits only apply to the volume of water being treated by a device

Information: Credits listed in the following table, and conditions associated for receiving the credit, are based on analysis of data from TAPE TERs. The credits may therefore be adjusted if additional data is collected, analyzed, and supports modification of the credits.

Credits for manufactured treatment devices. The credits are expressed as a percent removal.
Link to this table

Device	Manufacturer	Total phosphorus (%)			Total suspended solids (%)
Device	Manufacturer	Tier 1 (Basic)	Tier 2	Tier 3	Total suspended solids (%)
BayFilter w/EMC Media	BaySaver Technologies, Inc	50	60		85
BioPod Biofilter with Curb Inlet	Oldcastle Infrastructure, Inc.	50	60		82
ecoStorm plus	Watertectonics, Inc.	50
Filterra Bioscape	CONTECH Engineered Solutions, LLC.	50	60		82
Filterra System	CONTECH Engineered Solutions, LLC.	50	60		82
FloGard Perk Filter	Oldcastle Infrastructure, Inc.	50	53		80
Media Filtration System	CONTECH Engineered Solutions, LLC.	50
MWS-Linear Modular Wetland	Bio Clean Environmental Services, Inc. (A Forterra Company)	50	54		80
StormFilter using PhosphoSorb Media	CONTECH Engineered Solutions, LLC.	50	60	70	85
Stormfilter using ZPG Media	CONTECH Engineered Solutions, LLC.	50			80
StormGarden Modular Stormwater Bio-filtration System	Environmental Solutions	50	53		83
The Kraken	CONTECH Engineered Solutions, LLC.	50	58		80
Up-Flo Filter w/Filter Ribbons	Hydro International	50			80
Jellyfish	Contech	50	56		80
StormTree biofiltration	StormTree	50	60		85

Derivation of credits

The discussion in this section provides an overview of the methodology for deriving credits and a discussion of factors that potentially affect pollutant removal for mtds.

Data analysis

As discussed above, data were compiled from Technical Evaluation Reports (TERs) for Washington State’s Technology Assessment Program – Ecology (TAPE). The data can be found at this link - File:Water quality data for MTDs Feb 4.xlsx. Removal credits only apply to water treated by a device.

TSS removal is 80 percent, the LCL if greater than 80 percent, but cannot exceed 85 percent.
Tier 1 applies a 50 percent reduction for TP. This is the TAPE-approved value.
Tier 2 is based on an assessment of 95% lower confidence limits (LCLs) for the device. Since the LCL varies when all data are considered compared to just data used by TAPE, we typically select the more conservative value, though this may be adjusted based on additional analysis and professional judgement.
Tier 3 requires monitoring for a specific device unless a numeric credit, determined from data analysis and professional judgement, is given to the device.

Phosphorus analysis

Phosphorus may be divided into particulate (PP) and dissolved (DP) forms. Dissolved phosphorus is identified as passing through a 0.45 micron filter. For more information on the forms of phosphorus in water, link here. OP is often measured instead of DP. Using data from Capitol Region Watershed District, we estimate OP is about 70 percent of DP.

Some mtds, such as those utilizing only filtration or settling as the treatment mechanism, treat only PP, while some devices may utilize adsorption (e.g. with an amendment), biological uptake (e.g. from vegetation), or other mechanisms that can also treat DP. For devices that do not treat DP, the treatment effectiveness, as a percent removal rate, will decrease as the DP fraction of TP increases. This is illustrated in the adjacent figure for two mtds (Note the figure illustrates OP:TP ratios rather than DP:TP ratios).

TP removal as a function of OP:TP ratio for two mtds. Data are from TERs for the two devices.

The adjacent table summarizes data from two mtds, from the Minnesota Stormwater Manual, from Capitol Region Watershed District (CRWD) in Minnesota, and from a study by Fairbairn. The data show that DP:TP ratios vary from 0.23 for the CRWD data to about 0.4 for residential land use. The data also show that DP concentrations in the data collected for mtds is low compared to Minnesota data.

We observed that TP removal efficiency decreased as inflow TP concentrations decreased below about 0.15 mg/L. This is illustrated in the adjacent plot. The adjacent table indicates that typical TP concentrations in Minnesota runoff are above this 0.15 mg/L threshold.

TP removal as a function of inflow TP concentration

Concentrations of TP, OP, and DP for four mtds and Minnesota runoff
Link to this table

Data source	Rainfall type	TP (mg/L)	OP (mg/L)	DP (mg/L)	DP:TP¹
Filterra	I	0.09	0.01		0.16
UpFlo	I	0.11	0.01		0.13
Modular wetland	I	ND
Phosphosorb	II	0.28
Manual - residential	II	0.325		0.20	0.43
Manual - commercial	II	0.200	0.07		0.25
CRWD	II	0.304	0.042	0.053	0.23
Fairbairn	II	0.16	0.68	0.78	0.43

¹When only OP data was available, we estimated DP assuming OP is 70 percent of DP

Predicted (equation 1) vs observed TP removal for two MTDs.

For devices that utilize media that does not leach phosphorus, does not retain dissolved phosphorus, and does not selectively remove larger particle sizes, retention of total phosphorus should correlate with removal of total suspended solids and the fraction of phosphorus that is in particulate form. For example, if a device retains 85% of TSS and does not selectively remove larger particles, and if particulate phosphorus is 75% of total phosphorus in runoff, the total phosphorus removal should be 0.85*0.75=0.6375 or 64%. The formula for TP removal as a percent of TP in runoff (TP_R) is thus given by

\( TP_R = PP_F * TSS_R [Eq 1] \)

where PP_F is the fraction of TP in particulate form and TSS_R is the removal percent for TSS.

Research indicates that clay size particles sorb more phosphorus than larger particles, though clay particles also desorb phosphorus more readily. To determine the applicability of equation 1, we conducted the following.

We plotted expected TP removal based on equation 1 versus observed TP removal. Equation 1 is applicable if the resulting plot is linear and close to 1:1.
We evaluated the TERs to determine if the device selectively removed coarser particles.

Filterra and UpFlo were the only two devices for which we had orthophosphate data, which allows us to calculate PP. For each device we used Equation 1 to calculate the expected TP removal (TPR) and compared it to the observed removal. The resulting plots are shown in the adjacent figure. For Filterra the regression was significant at the 0.01 level (p = 1.9 X 10^-14) with an R² of 0.975. The intercept was fixed at 0, giving a slope of 1.10. For the UpFlo device, the regression was significant at the 0.01 level (p = 1.2 X 10^-10) with an R² of 0.848. The intercept was fixed at 0, giving a slope of 0.977.

The following information was collected from the TERs.

Stormwater Management StormFilter with PhosphoSorb media removed 78-83 percent of clay- and silt-sized particles, compared to 88 percent overall
The mean particle size for the UpFlo device was silt-sized (11 microns)
The mean particle size for the Modular Wetland was silt-sized
Approximately 60 percent of particles for the Filterra device were silt sized or smaller

Conclusions: Analysis of the phosphorus data for the mtds indicates the following may affect pollutant removal.

Elevated concentrations of DP relative to TP. Using equation 1 and assuming 80 percent TSS removal and 60 percent TP removal, the particulate fraction is 75 percent assuming no retention or loss of DP by the device.
TP inflow concentrations below about 0.15 mg/L may result in reduced TP removal. This concentration is typically exceeded in Minnesota runoff.
Phosphorus removal did not appear to be affected by particle size for the mtds analyzed, though literature suggests runoff with a high percentage of clay-sized particles may result in reduced TP removal. This may account for the observed lower removal for the UpFlo device.

Recommendations

Tier 2 and Tier 3 credits should not be given if the DP:TP ratio is greater than 0.25. If DP:TP data are not collected, we recommend Tier 2 credit not be given for residential land use, roof runoff, areas with extensive tree canopy, or other areas where dissolved phosphorus is likely to exceed 25 percent of total phosphorus.

Particle size analysis

TAPE testing of mtds includes particle size analysis. TAPE requires a modified ASTM 3977 PSD analysis and requires that runoff have 50% or more silt-size or smaller particles (50 microns or less).

For pollutant removal data from TAPE testing to be applicable to Minnesota, it is important to ensure the particle size distribution in the runoff collected for TAPE certification compares reasonably to Minnesota runoff. We gathered information on particle size distribution of runoff from the literature. The adjacent table summarizes the results. The median particle size from the literature review was 54.5 microns, and the mean was 81.7 microns.

Study	Median particle size (microns)
NURP₅₀ (nationwide)	8
NURP₉₀ (nationwide)	90
Selbig mixed (WI)	95
Selbig parking (WI)	32
Selbig arterial streets and roof (WI)	43
Selbig feeder streets (WI)	80
Selbig residential (WI)	80
Gonclaves and van Seters	13.7 (4.2 – 31.1)
Kim and Sansalone	100 (40-400)
Sansalone et al.	520
MRSC (WA)	120
EPA (nationwide)	8
Kellner ad Hubbart (MO)	59
Filterra	50
Stormwater Management StormFilter with PhosphoSorb media	100
UpFlo	11
Modular wetland	15-45
Median	54.5
Mean	81.7

The results indicate the four mtds fall within acceptable particle size distributions. The PSD for the UpFlo and one of the modular wetland systems had finer PSDs than typically observed. The Stormwater Management StormFilter with PhosphoSorb media had a a coarser PSD but was shown to remove 78-83 percent of silt- and clay-sized fractions.

The following conclusions were drawn from our literature review.

The central tendency for particles is in the silt to very fine sand range (approx 60-90 microns).
There is no specific pattern as to why some studies had smaller average particles compared to others.
Some studies showed larger particles in snowmelt compared to rainfall runoff and larger particles in residential compared to commercial runoff.
Season does not appear to affect PSD.
Land use appears to affect PSD, though the variability in PSDs within a given land use is greater than the variability between different land uses. Residential areas tend to have coarser PSD than parking and commercial areas.
Particle size appears to have played a role in the performance of some mtds. More specifically, at two sites with about 80% silt size or smaller particles, the devices clogged more quickly and TP removal was subsequently reduced.

Conclusion: The PSDs used to assess mtd performance were in an acceptable range for Minnesota stormwater runoff.

Derivation of TP credits for specific mtds

This section provides summaries of TAPE monitoring data for each device.

Contech Filterra

The 95% boot strap LCL is 52% for all data and 69% for TAPE analyzed data. TAPE did not use data when TP inflow concentrations were < 0.1 mg/L. Of 22 samples collected, 12 had TP concentrations < 0.1 mg/L. TP removal decreases as the inflow concentration decreases below 0.1. However, if we consider only TP values > 0.05 mg/L, the LCL is 58.9 percent. TP concentrations of 0.05 mg/L or less are uncommon in Minnesota. Consequently, the recommended Tier 2 value is 58%. The median removal for all data was 67% and for TAPE-analyzed data was 78.5%. We therefore recommend a Tier 3 value of 60%, which is conservative but meets a commonly employed regulatory threshold.

Contech Stormwater Management StormFilter with PhosphoSorb media

The 95% boot strap LCL is 69.7% for all data and 69% for TAPE analyzed data. The average influent TP concentration was 0.28 mg/L. Because of uncertainty about removal for lower TP concentrations, the recommended Tier 2 value is 65%. The median removal for all data and for TAPE-analyzed data was 81%. Because of uncertainty about removal for lower TP concentrations, the recommended Tier 2 value is 75%.

Hydro International UpFlo

The 95% boot strap LCLs for the UpFlo were less than 50%. Thus, there is no Tier 2 credit for this device. Note the average particle size for this device during TAPE testing was 11 microns, which may have affected the performance of the device.

BioClean MWS-Linear Modular Wetland

The 95% boot strap LCL is 58.8% for all data and 54.1% for TAPE analyzed data. TAPE did not use data when TP inflow concentrations were < 0.1 mg/L. Of 25 samples collected, 11 had TP concentrations < 0.1 mg/L. TP removal was unaffected by inflow concentrations. Consequently, the recommended Tier 2 value is 54%. The median removal for all data was 61.3% and for TAPE-analyzed data was 63.3%. We do not recommend a specific Tier 3 value based on this analysis.

Calculating annual volume treated

The above credits apply to water treated by a device. To determine the quantity of pollutant removed, the volume being treated must be determined.

Pollutant removal is often assumed to represent annual removal based on BMP design. For example, in Minnesota, a device designed to treat the first inch of runoff will typically treat about 90 percent of the annual runoff volume from an impervious surface. Devices tested and approved in Washington State are required to treat 91% or more of the average annual runoff. Because climatic conditions differ between Washington State and Minnesota, the annual average volume to which the credit applies in Minnesota must be determined. The following are acceptable methods for calculating the average annual volume treated by a device.

Translators. Translators are based on a relationship between volume treated in one location to volume treated in another location. For example, 90 percent annual volume treated in Location A may be equivalent to 80 percent volume treated in Location B. In this example, to treat 90 percent of the annual volume in Location B a BMP would have to be sized up from the same BMP in Location A. Translators have not yet been established for Washington State rainfall and runoff data.
Modeling. Using Minnesota climate data, a BMP can be sized in a continuous simulation model to determine the percent of total runoff that is captured and treated by a BMP and the runoff that bypasses the BMP.
Monitoring. Monitoring is based on direct measurement of total runoff volumes and runoff volumes captured and treated by a device. Runoff not captured by a device bypasses the device without treatment. As an example, bypass is determined during TAPE monitoring.
Devices approved in similar climatic conditions. Data for devices that have been tested in climate conditions similar to Minnesota can be used to calculate annual volume captured and treated by the device.
The device is downstream of a retention device that ensures a specified volume is treated by the device. An example would be an upstream retention pond designed to capture 90 percent of the annual runoff volume. If all the runoff captured by the pond is treated by the downstream device, the annual treated volume equals the volume captured by the upstream device. If the upstream device provides treatment of a pollutant, this will reduce the pollutant removal efficiency of the downstream device and the pollutant removal credit must be adjusted accordingly.

Protocol for monitoring

As indicated for Tier 2 and Tier 3 credits, monitoring data may be used to verify mtd performance or to characterize stormwater runoff (e.g. to determine DP:TP ratios or particle size distribution). Data must meet one of the following protocol.

We need to determine this. It could be multiple options, such as meeting TAPE protocol, modified TAPE protocol, ASTM, STEPP, something else. We have this page that may be useful.

References

Selbig
Zhao

Contents