SUSTAINING BIOSOLIDS RECYCLING UNDER PHOSPHORUS-BASED NUTRIENT MANAGEMENT Material Matters, Inc

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

SUSTAINING BIOSOLIDS RECYCLING UNDER PHOSPHORUS-BASED NUTRIENT MANAGEMENT Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Robin C. Brandt and Herschel A. Elliott The Pennsylvania State University Department of Agricultural and Biological Engineering University Park, PA 16802-1908

ABSTRACT

Forty-nine states have developed site assessment tools (Phosphorus (P) indices) to rank agricultural fields based on vulnerability to P loss. Because long-term biosolids application causes accumulation of soil P above concentrations needed for optimum crop yields, land-based recycling programs may be adversely affected by deployment of P management policies. Ninety-five active biosolids recycling fields were evaluated using the Pennsylvania P index. Without accounting for biosolids P solubility, only 22 of 95 fields qualify for standard nitrogen (N)-based biosolids application rates. Using biosolids-specific P source coefficients, based on Material Matters, Inc. laboratory testing,Material 76 fieldsMatters, coul dInc. be managed by applying biosolidsMaterial to satisfy Matters,crop N needs. Inc. Changes in crop management practices, biosolids application methods, and variables affecting crop / conservation practices were also evaluated as options for decreasing the overall P index ratings for the fields. A combination of measures could be used to qualify nearly all fields for N- based biosolids application. Adapting this approach to other state P site assessment tools is essential for sustaining land application as a biosolids recycling option.

KEYWORDS

Biosolids, phosphorus nutrient management, P index, sensitivity analysis, land-based recycling

INTRODUCTION

Material Matters, Inc. Each year overMaterial 7 million Matters, dry tons Inc. of municipal wastewater biosolids Materialare produced Matters, by Inc. approximately 17,000 municipal wastewater treatment facilities operating in the U.S. (Epstein, 2003). About half of the biosolids produced are land applied as soil amendment and/or fertilizer products. This recycled material contains tremendous amounts of nitrogen (N) and phosphorus (P), and represents a significant nutrient source for agriculture. Ironically, the major challenge to sustaining biosolids recycling may arise not from heavy metals, emerging pathogens or exotic organic chemicals, but from nutrient issues - routinely championed as a major benefit of land- based biosolids application.

In agricultural recycling, biosolids are typically applied at rates designed to satisfy crop N requirements while avoiding nitrate leaching to groundwater. Because of the mismatch between biosolids nutrient content and crop fertility requirements, P is normally supplied in excess of that needed for optimum crop growth. Off-site migration of P to aquatic systems is a concern because dissolved P concentrations in runoff can exceed critical levels that trigger eutrophic effects in surface waters. Eutrophication has been identified by the EPA as a leading surface water quality Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

430 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

problem in the U.S., thus considerable attention is being directed to P-based nutrient management. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Forty-nine of fifty states have adopted a P index prescribed nutrient management strategy to address the issue. The original P index (Lemunyon and Gilbert, 1993) was a field-scale scoring matrix using readily available information to identify vulnerable sites for targeting P-loss reduction efforts. Individual states have customized the original index to address local climate, hydrology, and geomorphology conditions. The current Pennsylvania index (Table 1) contains six source factors and five transport factors. Source factors (e.g., soil test P and fertilizer application rate) describe the amount of P available for runoff and leaching, while transport factors (e.g., runoff potential and erosion rate) describe the mechanisms and pathways of movement of P from fields to receiving streams. Various weighting factors and arithmetic operations are used in P indices to reflect the interrelationships and relative importance of input factors to P loss.

States have established interpretive categories that define nutrient management practices based on a site’s P index numerical score. In Pennsylvania, four management categories are defined. Material Matters, Inc. For the “low”Material and “medium” Matters, categories, Inc. soil application rates for organicMaterial by-products Matters, are Inc. determined using N-based budgets. This is the procedure currently in place for biosolids, as the federal rules (CFR 40, Part 503) require application at the “agronomic rate to provide the amount of nitrogen needed by the food crop, feed crop, fiber crop, cover crop, or vegetation grown on the land”. The “high” category dictates careful management of P, with soil application rates of organic amendments restricted to just meeting the P requirement of the crop (P-based management). This has important ramifications for biosolids management, since typical P concentrations in biosolids translate into application rates that are one-fifth to one-tenth of the current practice. The “very high” category dictates that no additional P (i.e., no biosolids) should be applied to the site (Prohibited category).

Inevitably P-based nutrient management will be mandated for biosolids recycling. Across the US, many versions of the P index are used to support nutrient management policies and regulations (Sharpley et al., 2003). Because biosolids have not been systematically addressed in Material Matters, Inc. most state indices,Material mandated Matters, P-based Inc. nutrient management will be a Materialdifficult proposition Matters, Inc.for many utilities, forcing major changes or even abandonment of biosolids recycling programs. Anticipating Arkansas regulations requiring P-dictated agronomic rates, the city of Fayetteville curtailed its liquid biosolids application program (Scanlan et al., 2002). To sustain land application programs, it is crucial that biosolids managers understand how the P index score is determined in their state and the relative sensitivity of the various input factors. Information should be gathered to calculate P index scores for existing land application fields and the P index status for the overall land application program determined. With this information, the biosolids manager can identify critical or emerging situations that may limit recycling efforts, and evaluate various strategies to sustain recycling programs.

Coale et al. (2002) noted that P site indices will be deployed before their predicative capabilities are objectively validated. Few studies have rigorously assessed the impact of P index implementation on the viability of agricultural enterprises (Kogelmann et al., 2006), and the effect on biosolids recycling is also highly uncertain. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

431 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Table 1. Pennsylvania P Index, V2 (after Beegle et al., 2006) NRSA Base Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.Condition Material Matters, Inc. PART B: SOURCE FACTORS Example SOIL TEST Mehlich-3 Soil Test P (ppm P) 168 Soil Test Rating = 0.20* Mehlich-3 Soil Test P (ppm P) 34

FERTILIZER-P RATE Fertilizer-P (lb P2O5/acre) 10 P Applied from multiple fertilizer applications, if any. 0 0.2 0.4 0.6 0.8 1.0

Copyright ©2008 WaterEnvironmentFederation.All RightsReserved FERTILIZER APPLICATION Placed or injected 2" or Incorporated <1 week following Incorporated > 1 week or not Incorporated >1 week or not Surface applied to frozen or 0.2 METHOD more deep application incorporated following incorporated following snow covered soil application in April - October application in Nov. - March

Fertilizer Rating = Fertilizer Rate x Fertilizer Application Method 2

BIOSOLIDS-P RATE Biosolids P (lb P2O5/acre) 767 P Applied from multiple organic‐P applications, if any. 0 0.2 0.4 0.6 0.8 1.0 Residuals andBiosolids2008 BIOSOLIDS APPLICATION Placed or injected 2" or Incorporated <1 week following Incorporated > 1 week or not Incorporated >1 week or not Surface applied to frozen or 0.4 METHOD more deep application incorporated following incorporated following snow covered soil application in April - October application in Nov. - March

432 P SOURCE COEFFICIENT BPR biosolids (0.8) and all other biosolids (0.4)…....OR determined using Water Extraction Testing 0.4 Material Matters, Inc. Biosolids-PMaterial Rating = Matters, Biosolids-P Rate Inc. x Biosolids Application Method x Biosolids-P AvailabilityMaterial Matters, Inc. 123 Material Matters, Inc. Source Factor Sum 158 Example PART B: TRANSPORT FACTORS Field EROSION Soil Loss (ton/A/yr) 3 0 2 4 6 8 RUNOFF POTENTIAL 4 Excessively Somewhat Excessively Well/Moderately Well Somewhat Poorly Poorly/Very Poorly 0 1 2 * SUBSURFACE DRAINAGE 1 None Random Patterened 6 0 2 4 9 ‡ CONTRIBUTING DISTANCE 100 to 199 ft. OR 2 > 500 ft. 350 to 500 ft. 200 to 349 ft. < 100 ft. < 100 ft. with 35 ft. buffer Transport Sum = Erosion+Runoff Potential+Subsurface Drainage+Contributing Distance 10 0.85 1.1 50 ft. Riparian Buffer 1.0 MODIFIED CONNECTIVITY Direct Connection APPLIES 1.0 APPLIES TO DIST Grassed Waterway or None TO DIST > 100 FT < 100 FT * OR rapid permeability soil near a stream Transport Sum x Modified Connectivity/24 0.42 ‡ "9" factor does not apply to fields with a 35 ft. buffer receiving manure. P Index Value = 2 x Source x Transport 132

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

The purpose of this study was, therefore, to evaluate 95 active biosolids-recycling fields using the Pennsylvania P index and ascertain the impact of P index prescribed management on Material Matters, Inc. program sustainability.Material Matters, Inc. Material Matters, Inc.

With existing site and crop management practices serving as a baseline, the impacts of various management changes on the overall P index ratings have been quantified. Although results are specific to the landscape conditions, biosolids sources, management practices, and nature of the site assessment tool, discussion is provided for extrapolating this approach to other states. Such analysis is essential to sustaining biosolids recycling under pending P-based nutrient management.

METHODOLOGY

Successful land application programs from two Pennsylvania municipalities, with >20-years experience (each) served as the basis for this study. These programs included biosolids from four different wastewater treatment plants (WWTPs). Biosolids characteristics shown in Table 2 are Material Matters, Inc. based on oneMaterial representative Matters, composite Inc. biosolids sample (each) analyzedMaterial at the Matters,Penn State Inc. Agricultural Analytical Services Laboratory. All four WWTPs employ activated sludge treatment processes. In Program #1, the BE#1 and NH WWTPs add Fe salts for P removal, while plant BE#2 uses biological P removal (BPR). This facility also uses supplemental Fe-salt addition to enhance P removal. Plant BE#1 employs both aerobic and anaerobic digestion, but on the whole, solids from this facility are classified as aerobic digestion biosolids for land application purposes, which results in lower agronomic rates than would be allowed with anaerobic digestion (USEPA, 1995). Autothermal thermophylic aerobic digestion (ATAD) stabilization is used at the BE#2 WWTP and the NH facility employs lime-post-treatment stabilization for undigested solids (Brandt, 2003). Program #2 manages solids from three WWTPs, with similar anaerobic digestion biosolids cake products. For this study, only biosolids from the NE plant were used in the P index evaluation (Brandt and Weaver, 2003). Biosolids water extractable P (WEP) was determined using a 1:200 solid: solution ratio and the P index phosphorus source coefficient (PSC) was calculated as: Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. (Elliott et al., 2006)

[Note: The calculated PSC value is limited to values between 0.10 and 1.0 for the Pennsylvania P index.]

The Pennsylvania P index (Beegle, et al., 2006) rating was determined for 95 fields used for biosolids recycling on >700 acres dispersed among 15 privately owned/operated farm operations. All farm sites were located in south-central Pennsylvania where the vast majority of biosolids recycling occurs in the state. Fields typically had a history of biosolids application ranging from 8 to 12 years prior to this study. Biosolids farm standards in Pennsylvania require an active soil conservation program, isolation distances, accounting and accommodation for livestock manure spreading (where appropriate), and a variety of management practices (PA Title 25, Subchapter J). Accordingly, biosolids fields are often among the best managed areas, remote from livestock housing facilities. These factors, combined with the relatively small size of many Pennsylvania family farms, results in the need to recruit many farms for a successful land application program, for medium to large WWTPs. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

433 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

The municipality managing Program #1(recycling ~1000 dry tons per year) maintains a land Material Matters, Inc. base of approximatelyMaterial Matters,900 acres (~90Inc. permitted fields) on 19 farms. Material Only about Matters, ½ of the Inc. acreage is actually applied in any given year. In this study, 49 fields were targeted for land application by the biosolids program manager, who was a qualified nutrient management planner, trained in the use of the Pennsylvania P index. Biosolids Program #2 typically land applies >20,000 dry tons per year on over 40 farms (>700 fields). For this study, 46 fields (230 acres) were selected a reasonably representative sampling for Program #2. Biosolids agronomic rate prescriptions and P index inputs were determined by a qualified nutrient management planner (and certified soil scientist) who was involved in farm conservation planning for the sites and thoroughly familiar with land application operations.

Data used to calculate P index scores was collected from several sources: soil P from soil analyses results, crop management information from cooperating farmers, soil data from NRCS soil surveys, nutrient data from biosolids analysis reports, and land application prescriptions/ field scheduling from biosolids program management personnel (Brandt, 2003; Brandt and Weaver, 2003). Table 3 summarizes P index input factors for the 95 fields. Because the P index Material Matters, Inc. is calculatedMaterial for the P loadingMatters, that Inc. accompanies the biosolids applicationMaterial rate necessary Matters, to Inc. satisfy the N needs of the crop, Table 3 also provides the plant-available N (PAN) prescriptions for the fields based on typical yield goals and N uptake values for the crops (corn, wheat, barley, alfalfa, orchard-grass) to be harvested.

After determining P index scores for all 95 fields, based on practices normally employed by the respective municipal programs, alternative strategies to improve (reduce) P index scores were investigated. In all, eight alternative management scenarios were modeled, subsequent P index scores determined, and P-index-prescribed categories were tabulated (i.e. N-based, P-based, or Prohibited from P application). Selection of alternative management practices was based on the desire to: (1) minimize (or avoid) biosolids application rate reductions that inflate land-base requirements, (2) minimize (or avoid) crop/ conservation program changes, which could provoke farm operator refusal, and (3) maximize use of practices that WWTPs can implement inside-the- fence to reduce the P index score, without the need for on-farm operation changes. In Material Matters, Inc. Pennsylvania,Material the vast Matters,majority of Inc. land application farms are privatelyMaterial owned andMatters, imposition Inc. of excessive requirements for participation in municipal biosolids recycling programs can cause participating farm operators to withdraw. Table 5 provides a summary of alternative P index mitigation scenarios considered in this study to reduce P index scores (reduce P loss risk) and increase the number of fields qualifying for N-based nutrient management.

In addition to the aforementioned trial-run P index scenarios, a nominal range sensitivity analysis (NRSA) of the Pennsylvania P index was performed using the 95 field data set to establish base condition and domain limits. The NRSA technique examines output changes resulting from input perturbations across a range of values, while holding all other inputs at their base condition levels. For this reason, NRSA is often used as a local screening tool for complex models to identify high impact variables warranting more detailed analysis (Saltelli et al., 2000).

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

434 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Table 2. Selected properties of the biosolids.

Material Matters, Inc. Material Matters, Inc. Biosolids Source Material Matters, Inc. Material Matters, Inc. Program #1 Program #2 BE#1 Cake BE#2 Cake NH Cake NE Cake Parameter (Lab ID E5399) (Lab ID E5397) (typical) (Lab ID M02912) Total Solids (%) 19.4 19.6 19.0 32.6 Copyright ©2008 WaterEnvironmentFederation.All RightsReserved Volatile Solids (% of TS) 54.7 59.1 52.0 50.5 pH (1) 8.30 8.10 12.0 7.71 Water extractable P (2) All results expressed on dry weight basis. -1 0.38 1:200 WEP (g kg ) 0.33 3.42 0.13 Residuals andBiosolids2008 1.65 1:200 PWEP (% of PT) 0.84 7.15 0.50 (3) 0.10

435 P Source Coefficient (PSC) 0.10 0.34 0.10 Material Matters,Nitrogen Inc. series (4) Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Total nitrogen (g kg-1) 49.2 50.9 50.0 38.8 Ammonium N (g kg-1) 11.9 11.0 10.0 9.0 Calculated organic N (g kg-1) 37.3 39.9 40.0 29.8 Total elemental contents (5) Total phosphorus (g kg-1) 39.5 47.9 23.0 26.1 Total aluminum (g kg-1) 11.1 9.6 9.0 15.0 Total iron (g kg-1) 81.1 42.2 40.0 81.6 Total calcium (g kg-1) 59.4 56.3 130 20.3 Total potassium (g kg-1) 2.33.42.02.2 Material Matters,Test Methods: Inc. (1) 1:1 soil:water pH; (2) solid:solutionMaterial ratio = Matters, 1:200; (3) PSC=0.102(WEP200)^0.99Inc. per Elliott et al.,Material 2006; ( 4 Matters,) Kjeldahl-N Inc. by Standard Material Matters, Inc. Methods; (5) Digestion by EPA 3051 followed by extract analysis via ICP (EPA 6010B). Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Table 3. Pennsylvania P index input conditions found in the 95 field biosolids study and used in the NRSA analysis. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Pensylvania P Index V2 95 Field Biosolids Application Study Values Used in Nominal Range Sensitivity Analysis

N-BASED PERSCRIPTION (1) Units Mean Median Mode Min Max Stdev (2) Base Domain Condition (5) Plant Available N (PAN) lb PAN/A 126 135 60.0 50.0 300 47.2 Min (6) Max (7)

Source Factors

Soil Test P ppm P (3) 182 168 150 24.0 500 91.9 168 0.0 500

(8) Fertilizer-P Source Rate lb P2O5/A 2.56 0.00 0.00 0.00 10.0 4.32 10 0.0 100 Residuals andBiosolids2008 Fertilizer-P Source Application Method No Units (4) 0.20 0.20 0.20 0.20 0.20 0.00 0.2 0.2 1.00

Biosolids-P Source Rate lb P2O5/A 677 767 344 226 1320 264 767 0.0 1320 436 Biosolids-P Source Application Method No Units (4) 0.50 0.40 0.40 0.40 0.80 0.14 0.4 0.2 0.8 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Biosolids-P Source Coefficient No Units (4) 0.45 0.40 0.40 0.40 0.80 0.13 0.4 0.1 1.0

Transport Factors

Soil Erosion tons/A/yr 3.16 3.00 3.00 2.00 5.00 0.75 3.0 0.0 5.0

Runoff Potential No Units (4) 3.22 4.00 4.00 2.00 6.00 1.09 4.0 0.0 8.0

Sub-Surface Drainage No Units (4) 0.04 0.00 0.00 0.00 1.00 0.20 1.0 0.0 2.0

Contributing Distance No Units (4) 2.40 2.00 0.00 0.00 6.00 2.56 2.0 0.0 8.0

Modified Connectivity No Units (4) 1.00 1.00 1.00 1.00 1.00 0.00 1.0 1.0 1.1

(1) N-based perscription based on realistic crop yield goals and crop N uptake (5) Median, or most commonly reported, non-zero value from 95 field study (2) Calculated sample standard deviation (6) Minimum possible value was selected for all factors (3) Soil test by Mehlich-3 method (7) Maximum plausible value dictated by categorical options in P Index or 95 field study (4) Discrete categorical value provided in PA P Index (V2, Beegle et al., 2006) (8) Maximum plausible fertilizer P2O5 value based crop uptake for corn silage Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Notably, NRSA is a simple and straight-forward technique that can be used to rank-order key input sensitivities (Frey and Patil, 2001). For this study, an Excel 2007 sensitivity analysis Add- Material Matters, Inc. In tool, SensITMaterial 1.31a (Middleton, Matters, Inc. 2007), was used to generate tornadoMaterial and spider Matters, plot graphics Inc. showing dominant impact P index variables.

RESULTS AND DISCUSSION

Characterization of Biosolids and Application Fields

Table 2 shows the results of characterization analyses for the biosolids considered in the study, which are generally typical of biosolids produced nationally. While originating from the same municipality, the WEP results for BE#1 and BE#2 biosolids indicate that these two materials behave quite differently when it comes to environmentally relevant P. Plant BE#1, which employs an oxidation ditch operating in the extended aeration mode, with Fe-salt addition for P removal, shows significantly lower WEP. This finding is consistent with previous research (Brandt et al., 2004), but even lower WEP results were expected from the extremely high Fe Material Matters, Inc. content (>80Material g WEP kg Matters,-1). The slightly Inc. elevated WEP here may be due,Material in part, Matters, to some level Inc. of BPR occurring in the oxidation ditch at plant BE#1 (typical Fe content in unamended biosolids is ~10g WEP kg-1). Plant BE#2 uses BPR wet-side treatment, which routinely yields the greatest biosolids WEP levels (Brandt et al., 2004). Hence, results from this analysis are again consistent with expectations for this type of treatment, although the Al+Fe content in this material (>50 g kg-1) very likely tends to minimize the WEP. Without the elevated Al+Fe concentration, one would expect even higher WEP results for biosolids from BE#2. The NE and NH cake biosolids likewise show elevated Fe contents. The NE biosolids are noteworthy for the exceptionally high Fe level (~80 g kg-1) resulting from co-disposal of Fe-based water treatment residuals.

All land application sites are located in south-central Pennsylvania (Lancaster and Lebanon Counties) and dominated by highly productive medium to fine textured residuum soils. The mean/median values shown in Table 3 provide a useful description of the typical field conditions for long-term biosolids application sites in Pennsylvania. Soil test P (STP, Mehlich-3) levels Material Matters, Inc. range from aMaterial low of 23- Matters, ppm P to Inc. 500-ppm P, with a median value ofMaterial 168-ppm Matters, P. This is Inc. noteworthy, as optimum STP levels in PA soils range from 20 to 50-ppm P. Elevated STP levels accompany repeated manure and/or biosolids applications, and are reason for concern relative to P-based management. Biosolids-P rates shown in Table 3 are derived from N-based agronomic rates prescribed for each field in consideration of biosolids N content and PAN crop requirements, after considering all N input sources (i.e. carry-over N from previous biosolids or manure applications, legume-N carry-over, and chemical fertilizer use). Biosolids-P rates ranged -1 -1 from 226 to 1320 lb P2O5 A , with a median rate of 767 lb P2O5 A . Since crops typically -1 remove 50 to 100 lb P2O5 A , the challenge facing biosolids land-appliers with P-based nutrient management is readily apparent. The necessity to avoid fields with excessive erosion, and in close proximity to streams (100-ft isolation distance required in Pennsylvania), are factors that help to mitigate P export risk from biosolids fields. Since all biosolids in this study have been dewatered, immediate (direct) injection is not practiced. Summer application and incorporation within one week is typical for these programs, but not always practiced. It is also noteworthy that

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

437 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

starter fertilizer (chemical P fertilizer) is used only on corn. Interestingly, only 12 of 69 corn fields were identified as using starter fertilizer. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Table 4 provides a summary of crops grown in the study area, which are primarily cropped for animal feed. As shown, corn is grown in 69 of the 95 fields (six fields are double-cropped with barley). Small grains (barley, wheat) are grown in 19 fields (including double-crop fields) and forage crops (alfalfa and orchard grass) are grown in the remaining 14 fields. Overall, the mixture of crops is reasonably representative of farms typically involved in biosolids recycling programs in Pennsylvania.

Table 4. Crops grown in the 95 field study area.

Crop Program #1 Program #2 Combined Corn (for grain or silage) 36 27 63 Barley/corn (double crop) 6 0 6 Alfalfa 5 3 8 Material Matters, Inc. Wheat Material Matters, Inc. 1 11 Material Matters,12 Inc. Barley 1 0 1 Orchard grass 0 5 5 Totals 49 46 95

Application of the P Index

The 95 fields were evaluated using the Pennsylvania P index (Table 1). Only 22 of 95 fields had P index scores of less than 80 (low or medium risk categories) indicating they qualify for standard N-based biosolids application rates (Table 5 and Figure 1). Twenty fields were rated as having a high risk of P loss (scores from 80-100) and thus application rates would be restricted to meeting the P need of the crop grown. Fully 53 of the fields had P index scores of >100 indicating that spreading of biosolids (or other P source) would be Prohibited. Greater than three-fourths of the fields would effectively be unusable for biosolids application without Material Matters, Inc. changes in PMaterial index input Matters, factors throughInc. various measures. Material Matters, Inc.

Besides identifying site vulnerability to P loss, an overriding purpose of the P index approach is to serve as a guide to identifying management practices for decreasing off-site P migration risk. For biosolids recycling, the challenge is to reduce the risk of P loss without reducing the biosolids application rate (BAR). Substantial reductions in the conventional N-based application rate result in the need for supplemental N fertilizer to achieve crop yield goals and greater land area (more sites) to accommodate the biosolids from a given facility. Both negatively impact the viability of biosolids recycling programs. Identifying and qualifying farm sites for in biosolids in Pennsylvania is often a labor intensive and costly undertaking. Despite typically lower costs for land application recycling, an abrupt necessity to expand the land-base for biosolids recycling can drive biosolids generators to other options, such as landfill disposal. The decision to abandon a successful land application program is often difficult to reverse as the institutional memory of WWTP personnel fades and participating farmers move on to alternative fertilizer sources. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

438 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Table 5. Management potential for several P Index input factors.

Material Matters, Inc. Material Matters, Program Inc. #1 Program #2 Material Combined Matters, Inc. Material Matters, Inc. STRATEGY N P X N P X N P X

1. BENCHMARK SCENARIO: P Index (V2) with default inputs for PSC (0.40 14 8 27 8 12 26 22 20 53

2. BENCHMARK with elimination of 14 8 27 8 12 26 22 20 53 chemical-P fertilizers on 12 corn fields Residuals andBiosolids2008 3. BENCHMARK with reduction of biosolids- 17 7 25 9 13 24 26 20 49 supplied PAN by 20% on 63 corn fields 439 4. BENCHMARK with reduction in soil Material Matters, Inc. Material Matters,18 Inc.8 23 13 9 Material24 Matters,31 17 Inc. 47 Material Matters, Inc. erosion by 1.0 ton/A/yr on all fields

5. BENCHMARK with reduction in 20 4 25 12 10 24 32 14 49 application method by one category

6. BENCHMARK with WEP -based PSC 200 34 3 12 42 3 1 76 6 13 values (per Elliott et al., 2006)

7. BENCHMARK with PSC = 0.10 for all 45 1 3 42 3 1 87 4 4 biosolids products on all fields

8. Strategy 7 + elimination of double-crop 46 3 0 42 3 1 88 6 1 PAN for corn on three fields Material Matters,N=N-based Inc. mgmt, P=P-based mgmt, X=MaterialProhibited Matters, from biosolids-P Inc. application. Material Matters, Inc. Material Matters, Inc.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Identifying strategies to decrease the overall P index rating of a field without substantially reducing the amount of biosolids applied is not a trivial task. The input parameters which Material Matters, Inc. constitute a MaterialP index vary Matters, widely inInc. the extent and ease with which theyMaterial can be altered.Matters, Alteration Inc. of some input factors can be constrained by economic considerations, wastewater and solids management regulations, cropping practices, landowner preferences, and field conditions. For example, one way to reduce the P index score is to reduce the biosolids-P loading by applying less material than crop PAN requirements. In the case of legumes (e.g. alfalfa), biosolids-N reduction does not necessitate supplemental fertilizer N. However, for non-legumes, reducing biosolids-N application requires a complementary increase in fertilizer-N. In the case of corn that is routinely supplied with starter fertilizer, additional fertilizer-N could be included in the starter, thus avoiding the need to make a separate, costly N application to the field. However, there is clearly a cost associated with adding chemical-N to the starter fertilizer, and this cost must be borne by someone. Some facilities may choose to bear this cost for the farm operator, if this approach helps to preserve the existing land-base.

Implementing Management Options

Material Matters, Inc. Table 4 summarizesMaterial the Matters, Pennsylvania Inc. P index input factor scenarios Materialconsidered Matters, in this study Inc. to reduce P loss risk and preserve N-based management. In this section we will consider the P index score impact accompanying implementation of these changes in program management and contrast each against a BENCHMARK scenario for comparison.

Strategy 1. BENCHMARK Scenario: This strategy is “business-as-usual,” prescribing all biosolids fields at the N-based agronomic rate, as though P-based management were not a concern. Default biosolids PSC values are used, based on the WWTP treatment process employed (PSC = 0.80 for BPR treated biosolids and 0.40 for all other biosolids). As noted earlier, and shown in Table 4 and Figure 1, only 22 of 95 fields qualify for N-based management: 12 of 63 (12/63) corn fields, 1/6 double-crop barley/corn fields, 1/8 alfalfa fields, 6/12 wheat, and 2/5 orchard grass fields. None of the fields receiving BPR biosolids qualified for N-based management. If this scenario cannot be improved dramatically, continued land application recycling for these municipalities would be unlikely. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Strategy 2. BENCHMARK with elimination of chemical-P fertilizers on selected corn fields: In this strategy, starter fertilizer-P has been eliminated from eight corn fields not qualifying for N-based management in the benchmark scenario. This strategy produced no change in the P risk category for any fields, as shown in Figure 1.

Strategy 3. BENCHMARK with reduction of biosolids-supplied PAN by 20% on 63 corn fields: For this strategy, biosolids-N, and therefore biosolids-P loading is reduced by 20%, with the understanding that supplemental N would need to be added to starter fertilizer, which would then be used on all 63 corn fields that did not qualify for N-based management in the benchmark scenario. It is further understood that additional acreage would be required in the land base to make-up for reduced applications on corn fields. This strategy produced four additional N-based fields (up from 22 to 26) and reduced the Prohibited fields from 53 to 49 (Figure 1).

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

440 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Strategy 4. BENCHMARK with reduction in soil erosion by 1.0 ton/A/yr: The soil erosion loss for all fields not qualifying for N-based management was reduced in this scenario, yielding Material Matters, Inc. nine additionalMaterial N-based Matters, fields (up Inc. from 22 to 31 fields). Prohibited Materialfields likewise Matters, decreased Inc. from 53 to 47 (Figure 1). While this strategy reduces P export risk more than previous approaches, it could require substantial modifications in farm management/conservation practices that would be untenable for many farm operators. Even if on-farm practices were modified to achieve the desired soil erosion reduction, it is unlikely that the gain in N-based fields for biosolids recycling would justify the effort required to implement this strategy.

Strategy 5. BENCHMARK with reduction in application method by one category on selected fields: For this strategy, all fields not qualifying for N-based management, and with biosolids application method (BAM) factors of 0.6 or 0.8, were reduced by 0.2 (one P index category, Table 1). These BAM factors affect surface-applied fields not scheduled for incorporation. As shown in Figure 1, 32 of the 95 fields under consideration qualified for N-based management with this strategy. While a slight improvement over previous strategies, this approach still falls far short of producing sufficient fields to justify continued land application. This approach could also be difficult to implement with farm conservation plan constraints and cropping. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Strategy 6. BENCHMARK with WEP-based PSC values: This is the first strategy considered that can be implemented without farm operator involvement. Here the WEP level is determined for each biosolids product, which is then used to calculate the appropriate PSC value reflecting source-specific P loss risk (Elliott et al., 2006). When appropriate PSC values are input into the benchmark strategy, the outlook for N-based nutrient management dramatically improves. Figure 1 shows that 76 of the 95 fields now qualify for N-based biosolids recycling, leaving only 19 fields as P-based or Prohibited. Closer examination of fields not qualifying for N-based management reveals the following: (1) none of the 12 fields scheduled to receive BPR biosolids (WWTP BR#2) are N-based, (2) five fields have STP levels >200 ppm P (three of these have ≥440ppm P), (3) Three fields are double-cropped, (4) three fields are <200 from a stream, (5) one field is somewhat poorly drained, and (6) many of these high risk fields do not provide for incorporation of biosolids. A number of the 19 fields have multiple high risk factors. However, many of these fields could potentially be qualified using strategies previously discussed, together Material Matters, Inc. with source Materialspecific PSC Matters, values. Inc. Material Matters, Inc.

Strategy 7. BENCHMARK with PSC=0.10 for all biosolids products on all fields: Strategy 6 shows that the PSC is an extremely powerful P index input factor. Accordingly, in this trial, the PSC is set at 0.10 for all biosolids products. Interestingly, this change only impacts the BPR product, as all other biosolids already qualify for this PSC level. Based on previous work (Brandt et al., 2004) it is plausible that the BE#2 BPR biosolids could be amended to achieve a PSC of 0.10. If this scenario yields promising P index score reductions for BE#2 biosolids, bench studies could be performed to investigate appropriate Al+Fe addition rates. Figure 3 shows that use of a PSC of 0.10 on all 95 fields, yields 87 of 95 fields qualifying for N-based management. This is a very promising finding, assuming that WWTP metal salt amendments are practical and affordable for the municipality.

Strategy 8. Strategy 7 with elimination of double cropping PAN for corn on three fields: One last trial was evaluated to see if some of the remaining 8 fields not qualifying for N-based Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

441 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

management in Strategy 7 might be suitable for continued biosolids recycling. In this scenario, the application rate on three double-crop fields was reduced to eliminate corn PAN, which in Material Matters, Inc. turn greatly Materialreduced the Matters, biosolids- Inc.P loading. This measure advancedMaterial two of the Matters, three fields Inc. from Prohibited to P-based and one field from Prohibited to N-based management. Further measures for the seven fields not qualifying for N-based management could potentially result in qualifying all of remaining fields, but the effort required may not justify keeping these fields in the program. All of these fields have drawbacks and it may be advisable to eliminate these from the land application program and replace with new fields recruited in the same general area as the current land base. Since the P index score is determined for each field, and each crop year, borderline fields should probably be phased out as new fields are added to the program.

Sensitivity Analysis

Knowing which P index inputs most influence the final site score can help focus management strategies to increase the number of usable fields under P-based constraints. Determining the power of P index input parameters to influence a field’s rating can be accomplished on a trial- and-error basis, as illustrated in the previous section. If field data are in a spreadsheet format, Material Matters, Inc. determining Materialthe impact Matters, of various Inc. management scenarios is straightforward,Material though Matters, tedious. Inc. Sensitivity analysis is a powerful diagnostic technique for assessing the behavior of a P index. In this study, NRSA was used to determine the effect of input factor perturbations on the final P index score for biosolids recycling in the 95 field study area.

Baseline conditions (P index score = 132) for the sensitivity analysis were established from the 95 field study area. Median, minimum and maximum P index factors in Table 3 were used to establish the base condition and domain limits. The subsurface drainage base value was set at 1.0 to preserve this input in the evaluation (Brandt and Elliott, 2004). The base condition was evaluated to identify the importance of individual input factors on final P index rating and results plotted as tornado and spider plots (Figures 2a and 2b respectively).

Since the P index function is linear with respect to all input variables, the Sensitivity Coefficient for each input is the slope of a plot of P index score versus the percent change in that variable Material Matters, Inc. ComparisonMaterial of factor SensitivityMatters, CoefficientsInc. are best visualized in a Materialso-called “spider”Matters, plot Inc. (Eschenbach, 1992), shown in Figure 2b. This figure shows the biosolids-P source factors (rate, application method, and PSC) superimposed as a single steeply sloping line, indicating that these factors are High-Impact variables. Fertilizer-P rate and application method factors, with relatively flat spider plot lines, are Low-Impact variables with the lowest Sensitivity Coefficients.

The varying regression line lengths in the spider diagram indicate the range of P index scores associated with each input (Figure 2b). The difference between the low and high P index scores is the Swing associated with each input factor. The impact of many independent input variables can be conveniently summarized in a “tornado” diagram (Eschenbach, 1992). In this diagram (Figure 2a), input variables are arranged in decreasing order, based on the magnitude of their total possible impact on the P index score. The central vertical axis represents the baseline condition P index score (132) and the Swing (in P index units) for any input variable is read on the x-axis.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

442 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Prohibited P-Based N-Based 100 1 4 6 90 13 4

50 Residuals andBiosolids2008 87 88 40 17 14

Number of of Fields Number 76

443 20 20 20 Material Matters,30 Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. 20 31 32 10 22 22 26

0 1 2 3 4 5 6 7 8 BENCHMARK No Starter P Reduce Corn Reduce SE Reduce BAM Use Source Set All PSC=0.10 on Corn PAN by 20% by 1 Ton/A/yr One Cat. Specific PSCs PSCs=0.10 & Reduce Dbl-Crp N Biosolids Program Strategies to Reduce P Loss Risk

Figure 1. Summary of P index prescribed management for 95-field study, employing several alternative strategies to reduce P index risk and maximize fields qualifying for N-based nutrient management. [see Table 5 for descriptions of various strategies shown.]

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

In this study, the tornado diagram is asymmetrical because baseline values were not arbitrarily Material Matters, Inc. selected midwayMaterial between Matters, the high Inc. and low domain extremes, as is sometimesMaterial doneMatters, in NRSA, Inc. but rather based on realistic conditions guided by the 95 field study.

Variables at the top of the tornado diagram have the greatest potential for changing model output, and thus are highly sensitive factors. For example, for a PSC change over its domain (0.1 to 1.0) is associated with a P index score Swing of 230 units (285 – 55 = 230) when all other inputs are fixed at their baseline values (Tables 1 and 3). Thus, for this study, the PSC factor alone has the power to change the P index score by as much as 230 points. Fertilizer P source rate, in contrast, contributes no more than 17 P index units of Swing. Accordingly, the PSC factor is a much more sensitive input than the fertilizer rate variable for biosolids recycling conditions defined by the 95 field study. It is instructive to consider both the Sensitivity Coefficient value and Swing amount for an individual input factor, hence the reason for showing both tornado and spider diagrams.

As shown, P loss vulnerability was found to be relatively insensitive to mineral fertilizer inputs. Material Matters, Inc. Accordingly,Material modification Matters, of this Inc. Low Impact P index variable shouldMaterial have little Matters, effect on Inc. P index scores, which is illustrated in the previous section. Soil erosion is an Intermediate-Impact P-index input. In this study, the Pennsylvania P index was most sensitive to the biosolids PSC, biosolids application rate and biosolids application method, which are High-Impact P index factors (Figure 2). Thus, beyond changing biosolids application rates and methods, the greatest potential to reduce the P index score is in reducing biosolids PSCs. Again, the trial-and-error scenarios presented in the previous section support this conclusion.

Figure 3 shows tornado and spider plot NRSA results for the 95 field study when the biosolids PSC is fixed at 0.10. For this situation, the base condition P index score drops to 55, well under the N-based management category cutoff level, again demonstrating the power of this input. In this scenario, the biosolids PSC is a fixed value and thus is shown as a single point in Figures 3a, and 3b. Interestingly, STP and runoff potential now emerge as High-Impact P index factors along with biosolids application rate and biosolids application method. This suggests that these two Material Matters, Inc. factors now Materialplay a dominant Matters, role Inc.in biosolids field P index scores. UnfortunatelyMaterial Matters, little (or Inc. nothing) can be done to modify these variables once a particular field has been chosen for inclusion in a land application program. Rather, fields with elevated STP and poor drainage should be avoided when choosing biosolids fields.

CONCLUSIONS

Continual land application of biosolids causes a buildup of soil P and increases the potential for P enrichment of surface and groundwater supplies. Wastewater treatment plants practicing land application will eventually need to address the mismatch between the biosolids N:P ratio and crop nutrient needs. This may involve side-stream processes to strip P from digester supernatant, adding N fertilizer to biosolids, or novel selection of cropping systems

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

444 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

a. Biosolids‐P Source Coefficient 0.1 1 Biosolids‐P Source Rate 0 1320 Biosolids‐P Appl. Method 0.2 0.8 Runoff Potential 0 8 Soil Test P 0 500 Contributing Distance 0 6 Soil Erosion 0 5 Subsurface Drainage 0 2 Fertilizer‐P Source Rate 0 100 Modified Connectivity 1 1.1 Fertilizer‐P Source Appl. Method 0.2 1 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc. ‐50 0 50 100 150 200 250 300 350 P Index Base Condition = 132 P Index Score

350 b. 300 P Index Base Condition = 132 250

200 Score 150 Index Biosolids‐P Source Coefficient Biosolids‐P Source Rate

P 100 Material Matters, Inc. Material Matters, BiosolidsInc. ‐P Appl. Method Runoff PotentialMaterial Matters, Inc. 50 Soil Test P Contributing Distance Soil Erosion Subsurface Drainage 0 Fertilizer‐P Source Rate Modified Connectivity Fertilizer‐P Source Appl. Method ‐50 ‐100% 0% 100% 200% 300% 400% 500% 600% 700% 800% 900% 1000% 1100% Input Value as % of Base Case

Figure 2. Nominal range sensitivity analysis results for P Index base condition = 132: (a) tornado diagram, (b) spider plot.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

445 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

a. Soil Test P 0 500 Runoff Potential 0 8 Biosolids‐P Source Rate 0 1320 Biosolids‐P Appl. Method 0.2 0.8 Contributing Distance 0 6 Soil Erosion 0 5 Fertilizer‐P Source Rate 0 100 Subsurface Drainage 0 2 Fertilizer‐P Source Appl. Method 0.2 1 Modified Connectivity 1 1.1 Material Matters, Inc. Biosolids‐MaterialP Source Coefficient Matters, Inc. 0.1 Material Matters, Inc.

10 20 30 40 50 60 70 80 90 100 110 120 130 P Index Base Condition = 55 P Index Score

120 b. 110 100 90 80

Score 70

60 Index Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

P 50 Soil Test P Runoff Potential Biosolids‐P Source Rate Biosolids‐P Appl. Method 40 Contributing Distance Soil Erosion 30 Fertilizer‐P Source Rate Subsurface Drainage Fertilizer‐P Source Appl. Method Modified Connectivity 20 Biosolids‐P Source Coefficient 10 ‐100% 0% 100% 200% 300% 400% 500% 600% 700% 800% 900% 1000% Input Value as % of Base Case

Figure 3. Nominal range sensitivity analysis results for P Index base condition = 55: (a) tornado diagram, (b) spider plot.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

446 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Over the next few years however, biosolids recycling programs will increasingly be impacted by P-based nutrient management policies and regulations. The extent and speed with which pending Material Matters, Inc. P managementMaterial strategies Matters, will influence Inc. biosolids recycling programsMaterial are highly Matters, variable sinceInc. policies are evolving at the state level. While the P indexing concept has widespread adoption, the evolution from concept to field assessment tool has followed several diverse paths. The nature of some P indices may allow biosolids recycling to continue on a “business as usual” basis. However, P-related issues will represent a significant challenge for biosolids recycling programs in some locations. Facilities that apply biosolids in more than one state face additional challenges since states’ P indices can yield dramatically different ratings for identical biosolids and field conditions (Osmond et al., 2006).

REFERENCES

Beegle, D., R. Bryant, W. Gburek, P. Kleinman, A. Sharpley, and J. Weld. 2006. The Pennsylvania phosphorus index, Version 2. Penn State University, University Park, PA.

Material Matters, Inc. Brandt, R.C.Material and H.A. Matters,Elliott. 2005. Inc. Sensitivity analysis of the PennsylvaniaMaterial phosphorus Matters, indexInc. for agricultural recycling of municipal biosolids. J. Soil Water Cons. 60:209-219.

Brandt, R.C., Elliott, H.A., O’Connor, G.A., 2004. Water extractable phosphorus in biosolids: implications for land-based recycling. Water Environment Research 76, 121-129.

Brandt, R.C. 2003. Land application of biosolids under phosphorus-based nutrient management. PhD thesis. The Pennsylvania State University. University Park, Pennsylvania.

Brandt, R.C. and S.R. Weaver. 2003. Agricultural recycling of Philadelphia biosolids: Implications of phosphorus-based nutrient management. Report prepared for the Philadelphia Water Department. Available from Material Matters, Inc., Elizabethtown, Pennsylvania.

Coale, F.J., Sims, J. T., Leytem, A.B., 2002. Accelerated deployment of an agricultural nutrient Material Matters, Inc. managementMaterial tool: Matters, The Maryland Inc. phosphorus site index. JournalMaterial of Environmental Matters, QualityInc. 31, 1471-1476.

Elliott, H.A., Brandt, R.C., Kleinman, P.J.A., Sharpley, A.N., Beegle, D.B., 2006. Estimating source coefficients for phosphorus site indices. Journal of Environmental Quality 35:2195-2201.

Epstein, E., 2003. Land application of sewage sludge and biosolids. Lewis Publishers, Boca Raton, Fl.

Eschenbach, T.G. 1992. Spider plots versus tornado diagrams for sensitivity analysis. Interfaces 22(6):40-46.

Frey, H.C. and S.R. Patil. 2001. Identification and review of sensitivity analysis methods. In: Proceedings of NCSU/USDA workshop on sensitivity analysis. June 11-12, 2001. Raleigh, North Carolina. Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

447 Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

Residuals and Biosolids 2008

Kogelmann, W.J., R.B. Bryant, H.S. Lin, D.B. Beegle, J.L. Weld. 2006. Local assessments of the impact Material Matters, Inc. of phosphorusMaterial index Matters, implementation Inc. in Pennsylvania. J. Soil WaterMaterial Cons. Matters, 61:20-30. Inc.

Lemunyon, J.L., Gilbert, R.G., 1993. The concept and need for a phosphorus assessment tool. Journal of Production Agriculture 6, 483-496.

Middleton, Michael A. 2007. SensIT sensitivity analysis add-in for Excel 2007 (Excel 97-2007) for windows and Macintosh. University of San Francisco. Available online from Decision Toolworks.

Osmond, D., M. Cabrera, S. Feagley, G. Hardee, C. Mitchell, P. Moore, R. Mylavarapu, J. Oldham, J. Stevens, W. Thom, F. Walker, and H. Zhang. 2006. Comparing ratings of the southern phosphorus indices. J. Soil and Water Conservation. 61:325-337.

Scanlan, P.L., Naylor, L., and Bunn, D. 2002. Can TNM derail your beneficial use program? Proc. 16th Annual Residuals and Biosolids Mgmt. Conf. Water Environ. Federation. Austin, TX. March 3-6.

Material Matters, Inc. Sharpley, A.N.,Material Weld, Matters,J.L., Beegle, Inc. D.B., Kleinman, P.J.A., Gburek,Material W.J., Moore, Matters, Jr., P.A., Inc. Mullins, G., 2003. Development of phosphorus indices for nutrient management planning strategies in the United States. Journal of Soil and Water Conservation 58, 137-152.

Saltelli, A., K. Chan, and E.M. Scott. Eds. 2000. Sensitivity analysis. John Wiley and Sons, Ltd. West Sussex, England.

USEPA, 1995. Process Design Manual. Land Application of Sewage Sludge and Municipal Septage. EPA /625/R-95/001, Office of Research and Development, Cincinnati, OH.

Material Matters, Inc. Material Matters, Inc. Material Matters, Inc.

448