Modeling of NonPoint Source Needs

In the 1996 CWNS, EPA used a model to estimate the needs for controlling certain categories of NPS pollution. While a large part of the nation's water quality problem is attributable to NPS pollution, few States have systematically documented their NPS needs. EPA prepared an NPS needs model to help build a more complete picture of the total water quality needs.

What Is NPS Pollution?

NPS pollution is caused by rainfall or snowmelt moving over and through the ground. As the runoff moves over or through the soil, it picks up and carries away natural pollutants and pollutants resulting from human activity, eventually depositing them into lakes, rivers, wetlands, coastal waters, and ground waters. In addition, habitat alteration and hydrologic modification (changes in hydrologic characteristics) can have adverse effects on the biological and physical integrity of surface waters. NPS pollution is not regulated by NPDES permits.

Sources of NPS pollution include agriculture, silviculture, atmospheric deposition, channelization, construction, contaminated sediments, contaminated ground water, runoff from highways, hydrologic and habitat modification, land development, land disposal, marinas, onsite disposal systems, recreational activities, removal of riparian vegetation, resource extraction, shoreline modification, and streambank destabilization. Agriculture is the leading source of impairment in the nation's rivers, affecting 60 percent of the impaired river miles assessed in the United States, according to EPA's 1994 305(b) report. Other sources of NPS pollution affecting impaired river miles include urban runoff (12 percent of impaired river miles), hydromodification or habitat alteration (17 percent), mining or resource extraction (11 percent), and silviculture (9 percent).

The distinction between NPS and diffuse point sources is sometimes unclear. Although diffuse runoff is usually treated as NPS pollution, runoff that enters and is discharged from storm sewer systems is treated as a point source discharge and hence is subject to the CWA permit requirements. In contrast, NPS dischargers are not subject to Federal permit requirements. Under Section 6217 of the Coastal Zone Act Reauthorization Amendments (CZARA), municipal and commercial SW dischargers in the coastal zone that are not covered by the Phase I SW Program must comply with the requirements of the CZARA. States are encouraged to develop consistent approaches in dealing with urban SW runoff.

What Are the Goals for NPS in the 1996 CWNS?

The 1987 CWA Amendments allow States to use their SRF to fund selected non-Federal NPS control activities that are contained in approved Section 319 NPS Management Plans. Section 319 of the CWA addresses NPS pollution by providing grant funds to help solve NPS problems and by requiring States to develop NPS Management Plans to address these problems. The NPS goals of the 1996 CWNS are to capture as many NPS needs as feasible through reporting those needs documented by the States and to supplement these needs through modeling.

What Are Eligible NPS Needs?

The SRF can fund projects designed to alleviate pollution caused by a wide variety of diffuse sources, including agriculture, silviculture, and abandoned mines. Eligible NPS control projects include virtually any activity that a State has identified in its NPS Management Plan. Such activities include projects to control runoff from agricultural land; conservation tillage and other projects to address soil erosion; development of streambank buffer zones; and wetlands protection and restoration. Documented NPS needs greater than $0.5 million for agriculture and silviculture were reported by 10 States in the 1996 CWNS as shown in Table A-3 in Appendix A. Since few States have developed comprehensive estimates for NPS control, EPA developed a model to estimate national costs for agricultural and silvicultural controls. This model is similar to the model used in the 1992 CWNS, with the differences noted below.

What Is Included or Excluded from the Modeled NPS Estimates?

• The modeled estimates include activities to develop and implement NPS management programs to control runoff from:
• Agriculture (cropland, pastureland, and rangeland);
• Confined animal facilities with fewer than 1,000 animal units (AU); and
• Silviculture.

Because of the lack of nationwide information, EPA did not develop a modeled needs estimate for other sources of NPS pollution. The following areas were excluded from the needs modeling due to insufficient data:

• Abandoned mines;
• Inappropriate land disposal of wastes;
• Channelization and hydromodification;
• Atmospheric deposition;
• Construction;
• Marinas;
• Runoff from roads, highways, and bridges;
• Urban/suburban areas not covered by Phase I SW Program permits; and
• Remediation of contaminated sediments causing a water quality problem.

These sources of NPS pollution could represent significant needs for SRF funding. In addition, other NPS costs that were excluded are recurring O&M costs as well as technical assistance, engineering, and related services that are often provided to farmers or others free of charge by Federal and State agencies.

What Is the Modeled Estimate for NPS Control Needs?

The modeled estimate of the NPS control needs is $9.4 billion. Table 6 summarizes the estimates by category and compares them to the 1992 modeled needs (converted to 1996 dollars). Table A-5 in Appendix A contains State-by-State modeling estimates for each of the NPS categories modeled.

The estimated needs for controlling runoff from cropland, pastureland, and rangeland is $3.8 billion. The 1996 needs for agricultural land of $3.8 billion are slightly less than the needs modeled in the 1992 CWNS of approximately $4.2 billion (inflated to 1996 dollars). The area of cropland reported in the 1992 Natural Resources Inventory (NRI) was approximately 10 percent, or 40 million acres less than the 1987 NRI value used in the 1992 CWNS. This drop may be attributed to the fact that conservation reserve program (CRP) land was not considered as cropland in the 1992 NRI, but was considered as such in the 1987 NRI.

The 1996 estimated needs for confined animal facilities of $2.1 billion are lower than the 1992 needs of $3.1 billion (adjusted to 1996 dollars). The lower needs are primarily due to the reduced number of animal feedlots containing fewer than 1,000 AU. This reduction is attributable to the trend in the animal industry toward larger operations. Operations with more than 1,000 AU are considered to be point sources and therefore are ineligible for funding as NPS projects. Compared to the 1987 Census of Agriculture, the total number of farms reported in the 1992 Census were 13.4 percent fewer for broilers, 39.2 percent fewer for layers, and approximately 22 percent fewer for beef, dairy, and swine. Similarly, within each livestock category, the number of animals also decreased for a majority of the farm sizes.

The annual national cost for implementing forestry BMPs was estimated to be $316 million. The 20-year period cost was determined to be approximately $3.5 billion. The 1996 CWNS estimate for silviculture of $3.5 billion is larger than the 1992 estimate of $2.7 billion (adjusted to 1996 dollars). The difference between the 1992 and 1996 needs estimates can be attributed partially to the increase in total area of timberland harvested in the United States, but mostly to the increase in average BMP implementation cost per acre of timberland harvested in each State. The number of States that have a formal regulatory structure for forestry operations increased from 1992 to 1996, thereby increasing BMP implementation costs.

The methodologies used to develop the modeled estimates are presented in the paragraphs that follow. These methodologies are similar to those used in the 1992 CWNS, with some refinements, and use of the most recent data.

How Were Cropland, Pastureland, and Rangeland Modeled?

Runoff from crop production and grazing land carries primarily sediments, salts, nutrients, and pesticides to the downstream receiving waters. Sediments generally result from erosion of cropland and grazing land. Excessive application of chemical fertilizer or animal manure on land frequently results in high concentrations of nitrogen and phosphorus in runoff or in leaching of nitrogen to ground water. Pesticide applications on cropland and pastures can introduce toxic pollutants into both surface water and ground water.

The cost estimating methodology addressed the control of erosion and pollutant export from cropland and grazing land. This methodology is based on applying a best management system (BMS), which combines soil conservation practices and other management measures that, when applied, will achieve NPS pollution control through the reduced transport of sediments, nutrients, and chemicals into surface and ground water.

Erosion control measures in the model were based on the implementation of soil conservation practice groups identified by the U.S. Department of Agriculture's (USDA's) Farm Service Agency. Water quality management was incorporated into the model by applying additional control measures, such as nutrient management, pesticide management, and irrigation water management.

The model estimated costs by determining an appropriate set of BMPs and estimating the implementation costs. This was accomplished as follows:

• Review NRI data. This national database provides data on area of farm land, crop type, soil erosion rate, soil loss tolerance, slope, and conservation practices in use in 1992. (The NRI is compiled by the USDA every five years. The 1996 CWNS used the latest available 1992 NRI data. The 1992 CWNS used the 1987 NRI data.)
• Develop a best management system. If land required erosion control, conservation practice groups were selected to reduce soil erosion to the soil loss tolerance level specified for that land. Additional measures to provide water quality management were also selected to complete the BMS.
• Determine needs for cropland, pastureland, and rangeland. Total capital costs of erosion control and water quality management were computed for cropland, pastureland, and rangeland in each State.

How Were Confined Animal Facilities Modeled?

A confined animal facility is a lot or facility where animals have been, are, or will be stabled or confined and fed or maintained for a total of 45 days or more in any 12-month period. Crops, vegetation forage growth, or post-harvest residues are not sustained in the normal growing season over any portion of the lot or facility.

Runoff from confined animal facilities may contain nutrients, oxygen-demanding substances, organic solids, salts, pathogens, and sediments. Runoff includes process-generated wastewater and precipitation that comes into contact with manure, litter, or other material used in or resulting from the production of animals.

For the purposes of the 1996 CWNS, costs were estimated only for confined animal operations with fewer than 1,000 AU. Confined animal operations (feedlots) with 1,000 AU or more are considered point sources, and are not eligible for funding under the Section 319 program. Therefore, the large operations cannot receive SRF loans. Under certain circumstances, facilities with less than 1,000 AU can be regulated as point sources, which could result in an overestimation of these needs.

The methodology used to develop EPA's estimate was based on the use of model feedlot facilities, which were intended to represent typical facility sizes within each livestock category. Livestock categories considered were beef feedlots, dairies, swine feedlots, and broiler and layer houses. The approach was similar to that used in the economic analysis for the CZARA, and cost data from that analysis were used in developing the 1996 CWNS cost estimates.

The methodology was based on the assumption that facility runoff was going to be controlled primarily through diversion for runoff containment and channelization of on-site effluent to the ultimate control structures. All runoff collected in these control structures was assumed to be used for irrigation.

The following steps were used to estimate the cost of controlling NPS pollution from feedlot operations:

• Identify model feedlots. Model feedlots were defined to represent typical sizes within each livestock category.
• Develop NPS management plan. NPS runoff control measures were identified and a typical management plan was selected for the model feedlots in each livestock category.
• Estimate needs for confined animal facilities. The number of livestock operations in each model feedlot was obtained from the 1992 Census of Agriculture data for each State (which is the latest data available; the 1992 CWNS used 1987 data). The total cost of implementing the NPS management plan was then estimated using this national database. Estimates for two control options were developed. Option 1 included lined retention ponds and irrigation for ultimate disposal. Option 2 also included irrigation for ultimate disposal but used filter strips in lieu of lined retention ponds. The estimate presented in this report is for Option 1. This is considered by the agricultural community to be the more effective approach, although it has the higher cost of the two options.

There was no change in the cost estimating methodology for the 1996 CWNS. However, the national databases used were updated to reflect the most recent data available.

EPA also modeled the needs for controlling runoff from confined animal facilities based on the USDA's definition of an animal unit, or AU. An AU is a unit of measurement which allows comparison between various animal types (e.g., cattle, dairy, poultry, etc.). EPA's definition of an AU is based on the NPDES permit regulations. The USDA revised these AU definitions based on the weight of the animals. If this revised definition were used, the number of facilities eligible for SRF funding would increase and result in total needs of $2.8 billion for confined animal facilities (compared to the EPA estimate of $2.1 billion reported).

How Was Silviculture Modeled?

Silvicultural activities can degrade water and habitat quality if sufficient care is not taken to prevent adverse effects. Sediment from erosion due to the presence of access roads and harvesting activities, temperature increases due to riparian shade removal, and pesticides and fertilizers used during timber operations are some of the major pollutants exported from timber-harvesting sites to receiving waters.

The methodology used to estimate the costs of controlling NPS pollution from silvicultural activities employed the following components:

• Develop estimates of annual forestland area harvested per State. The area of forestland harvested annually was computed by using the U.S. Forest Service's Forestry Resources of the United States, 1992. The distribution of the timberland area in relation to the type of terrain, however, was developed by considering the geographical characteristics of each State. Only privately owned forestlands were considered (Federal lands are ineligible for SRF loans).
• Identify silvicultural best management practices. Silvicultural BMPs were identified to control erosion from roads built to gain access to harvesting sites, to control the introduction of pesticides into watercourses, to maintain the stability of stream banks, and to ensure the revegetation of harvested sites, among other pur-

poses. BMPs identified were similar to those used for CZARA but were more refined.

• Identify typical comprehensive management plans. Typical comprehensive management plans were identified for controlling pollution and adverse habitat impacts for various site and timber characteristics.
• Develop cost estimates for management plans. Estimates for the cost per acre of implementing BMPs were developed for various types of forest management units (FMUs). These estimates indicated that the greatest variations in BMP implementation costs were caused by the general slope of the FMU and the degree of regulatory control practiced in the State.
• Estimate needs for silvicultural activities. Total costs of managing NPS pollution from silvicultural activities were estimated for each State.

What Are the Limitations of NPS Control Modeling?

The estimates presented in the 1996 CWNS represent EPA's effort to assess needs nationally for selected aspects of NPS control. The estimates represent only a portion of the total expected costs for NPS activities (specifically, agricultural land, confined animal facilities, and silviculture). Estimates will be refined and enhanced in future CWNSs, with documented NPS needs reported on a watershed basis.

Several cautions on the use of this information are appropriate:

• The estimates for confined animal facilities were prepared assuming no controls were in place. Therefore the estimates may overstate the real needs.
• Estimates for NPS BMPs were based on practices and requirements developed under the CZARA. Examples of these practices include appropriate erosion control, nutrient management planning, containment of runoff from confined animal facilities, and preharvest planning for silviculture.
• Some SRF-eligible areas with potentially very high costs, such as NPS runoff from abandoned mines, were not included.
• The estimates for agricultural controls, confined animal controls, and silvicultural controls are for capital investment or initial implementation of NPS controls, not for ongoing costs of O&M. O&M costs are not eligible for SRF funds but nevertheless represent a portion of the costs for NPS control.
• Areas with a large number of confined animal facilities, including permitted facilities with more than 1,000 AU, may have underestimated needs. This is because of the lack of land for spreading manure and the higher cost of transporting the manure to land adequate for spreading. These areas can also have large phosphorus inputs already in the soil, further increasing the cost of spreading manure.

EPA Home | Privacy and Security Notice | Contact Us