Application of phosphorus in industrialbiosolids to agricultural soils
(UKWIR Project SL02 B)
Final report prepared for UKWIR, Defra and EA
by
P.J.A. Withers
ADAS Catchment Management Group, ADAS Gleadthorpe, Meden Vale, Mansfield, Nottinghamshire NG20 9PF, UK
and
N.J. Flynn
Aquatic Environments Research Centre, Department of Geography, Whiteknights, University of Reading, P O Box 227, Reading RG6 6AB, UK
ContentsPage Number
1Introduction
2Methodology
2.1Review of current practices and knowledge
2.2A framework methodology for phosphorus loss risk assessment
2.3Effects of biosolids on soil P sorption and release
3Results
3.1Laboratory incubation study
4Effects of biosolids on runoff risk
4.1Linking soil P sorption to runoff risk – a literature study
4.2Defining biosolid application thresholds
5Development and application of operational guidelines
5.1Development of guidelines for sustainable application rates
5.2Development of guidelines for P loss risk assessment
5.3Defining management response
6Conclusions and further R&D
6.1Conclusions
6.2Further R&D
7Recommendations to the agricultural industry
8Acknowledgements
9References
Appendix 1. Changes in Olsen-extractable P after addition of biosolids
Appendix 2. Soil P sorption isotherms with and without native (Olsen) P for the Bedford biosolid.
Appendix 3. Soil P sorption isotherms with and without native (Olsen) P for the Cambridge biosolid.
Appendix 4. Soil P sorption isotherms with and without native (Olsen) P for the Grimsby biosolid.
Appendix 5. Soil P sorption isotherms with and without native (Olsen) P for the Kings Lynn biosolid.
Appendix 6. Soil P sorption isotherms with and without native (Olsen) P for the Broadholme biosolid.
Appendix 7. Soil P sorption isotherms with and without native (Olsen) P for the Davyhulme biosolid.
Appendix 8. Soil P sorption isotherms with and without native (Olsen) P for the Blackburn biosolid.
Appendix 9. Soil P sorption isotherms with and without native (Olsen) P for the TSP fertiliser.
Appendix 10 - Site assessment guidelines for assessing the risk of phosphorus loss in runoff following application of biosolids to agricultural land.
1Introduction
In 2001, the UK Sewage Sludge Survey showed that an average of 1,072,000 tonnes of dry solids was produced each year between 1998 and 2000 (Defra, 2002). This amount is likely to rise over the next 10 years as a greater proportion of sewage is treated, and higher treatment standards are applied, under the phased implementation of the Urban Waste Water Treatment Directive (UWWTD). This Directive banned the disposal of sewage sludge at sea in 1998, and introduced a requirement that sludge should be ‘re-used wherever appropriate and that disposal should minimise any adverse effects on the environment’. Of the three main alternative disposal routes for sewage sludge, incineration and landfill do not generate the same environmental benefits, and are more expensive, than application to agricultural land. Additionally, access to landfill sites is being increasingly restricted due to other environmental legislation. For these reasons, the UK government supports the recycling of sludge liquids and biosolids (hereafter referred to as just biosolids) to agricultural land as the Best Practicable Environmental Option (BPEO) in most circumstances, and for the added benefits their application brings such as savings in fertiliser costs, improving soil structure, workability and its water holding capacity. Currently 50% of the sludge produced in the UK is recycled to agricultural land (DEFRA, 2002) and the nutrients they contain (N, P, Mg, S and trace elements) can provide savings to agriculture of several million pounds in fertiliser costs (DETR, 2002). The range of biosolids applied to land has also changed in recent years due to concern over pathogen transfers. A major proportion are now improved by treating with calcium (Ca), iron (Fe) or aluminium (Al), or thermally dried and differ markedly from conventionally-treated biosolids in their physical, chemical and biological characteristics, and hence agricultural value.
One of the main benefits of biosolid application to land is its long-term sustainability as a source of phosphorus (P). An adequate supply of P, and maintenance of soil P fertility, is essential for early crop development and the efficient utilisation of other nutrients, especially nitrogen (N). The manufacture of inorganic P fertilisers will become increasingly expensive as high-quality, accessible phosphate rock deposits become depleted. Biosolids can therefore usefully supply a renewable source of P to meet crop requirements provided any adverse environmental impacts of biosolid-P on water quality can be controlled. The eutrophication of surface waters due to P (and N) enrichment is a major concern worldwide and agriculture has been identified as a significant P source (Foy and Withers, 1995; Sharpley and Rekolainen, 1997). As biosolid products often have a lower N:P ratio than other organic manures, their application to land at optimum or maximum recommended rates of N application may lead to a rapid build-up of surplus P in the soil (Pierzynski, 1994; Withers and Smith, 1995; Maguire et al., 2000a&b; Penn and Sims, 2002). Although biosolids are applied to circa 1% of the UK land area, repeated application to the same area could therefore pose a locally significant environmental hazard in adjacent watercourses due to the well established link between soil P build-up and accelerated P loss in land runoff (Heckrath et al., 1995; Pote et al., 1996; Maguire et al., 2005). Additional ‘incidental’ losses of P may also arise when rain follows soon after biosolid application to fields with preferential flow pathways and rapid connectivity to the watercourse (Withers et al., 2001b, 2003a). The sediment to which P is attached is also a diffuse pollutant in its own right causing siltation of fish spawning gravels, clogging of tributaries and increased flood risk (Owens et al., 2005).
Sustainable management of biosolid P must therefore take account of both the risk of mobilisation of soil and incidental P loss, and the risk of rapid delivery to either groundwaters or surface waters. Where the risk of P transport (ie mobilisation and delivery) is high, then the management of biosolids on the farm may need to be adjusted to reduce this risk (Shober and Sims, 2003). This adjustment may be in the amount of biosolid applied, the method of application or in the way the soil is cultivated or in the way the field is managed. At present, there are no specific controls on the application of biosolid P to agricultural land in the UK, and controls are variable elsewhere (eg Shober and Sims, 2003). Current codes of practice (MAFF, 1998) recommend that total inputs of P in organic manures to soils at P index 3 and above (>25 mg l-1 Olsen-extractable P) should not exceed the amounts of P removed in the crop rotation. This initial guidance makes good economic sense and was introduced to help reduce the long-term eutrophication risk associated with soil P build-up, but severely reduces the availability of suitable land for biosolid applications in some regions.
There are drawbacks using Olsen soil test P (STP) as an indicator of P loss risk because soil P loss occurs largely in particulate form, and the amounts of P lost in runoff from different soils are independent of the STP level (Quinton et al., 2003). Landscape (rainfall, soil type, slope) and land management factors (land use, cultivation practices, management of P inputs, presence of underdrains) are other important factors influencing P loss, and are likely to be site specific. Biosolids also vary in their P availability depending on the type and degree of treatment, and will increase STP by varying amounts (Withers and Smith, 1995; Maguire et al., 2000a; Penn and Sims, 2002). Hence, applications of biosolids outside the current guidelines may be acceptable at sites where it can be demonstrated that the risk of P loss is not increased, or where soil and land management can be modified to safely reduce the risk. It is therefore necessary to assess the risk of P loss from biosolid amended land on a site specific basis in order to help safeguard land application application as a sustainable disposal route.
This project aims to provide simple and practical operational guidelines on the safe application of P in biosolids to agricultural land without increasing the risk of eutrophication of adjacent watercourses. The guidelines will be based on current practices in respect of the different types of biosolid now applied to agricultural land, our current knowledge of P loss risk as influenced by landscape and land management factors and the range in potential management options available to help control P loss to acceptable levels. The work builds on Defra’s R&D programme on phosphorus loss from agriculture, UKWIR’s previous R&D project (SL-02) on biosolid P availability to crops and to the environment and the recent development of risk assessment and modelling approaches to help target mitigation options at those areas which contribute the most P loss.
2Methodology
2.1Review of current practices and knowledge
Regional practices
Visits were made to three regional water companies (South West Water, United Utilities and Anglian Water) to review the amounts and types of biosolid materials spread to land, the methods of application being employed, site selection and current risk assessment procedures and to select potential sludges for laboratory study. The amounts of biosolid recycled to agricultural land vary quite considerably between regions but land application is still the most important outlet. Whether it will remain a viable route in the future depends on the land bank available for spreading and the imposition of regulations restricting application rates to meet statutory requirements (NVZs) and adherence to voluntary codes of good agricultural practice. Most biosolid is spread within ca. 25 km of the STW although land bank pressures have forced wider distribution in at least one region. Soil compaction and odour are viewed as the main potential problems during land spreading, although decreasing availability of land due to the build-up of high soil P levels (Index 3 and above, MAFF, 2000) is becoming an important issue. In two of the regions, at least 40-50% of the available land was at P index 3 and above. Deterioration in water quality associated with enriched runoff following land application of biosolid was not considered a major problem. Since the introduction of the Safe Sludge Matrix in 1998 (ADAS, 2001; Hickman et al., 2000), the range in types of biosolid spread onto agricultural land has become narrower and treatment with dry or liquid lime, Fe dosing and composting has become more common. A broad summary of the proportions of the major biosolid types spread onto agricultural land in each region is shown in Table 1.
Table 1. Proportions (% of total spread) of different biosolids recycled to agricultural land in three regions of England, and the range in their P content.
Type of biosolid / RegionSouth-west / North-West / East
Digested cake / 35 / 50 / 35
Digested liquid / - / 20 / -
Lime-stabilised cake / 55 / 30 / 60
Thermally-dried cake / - / - / 5
Compost / 10 / - / -
P content (% in DS) / 0.7 – 3.0 / 0.4 – 3.8 / 0.9 – 3.7
Lime-stabilized products clearly make up a significant, if not dominant, proportion of the biosolids now applied. The range in biosolid P content also varies substantially from 0.4 – 4% in the dry solids (DS). In all three regions, the amount of P-stripping with Fe/Al-dosing is forecast to increase as a result of the Water Framework Directive although the amounts of P-stripped biosolid spread onto agricultural land are currently still relatively low. Most sludge is incorporated into available arable land in autumn (65%) and spring (35%). Smaller amounts of biosolid are spread onto grass in regions where this land use is more dominant, with a significant proportion injected in liquid form.
A number of risk assessment procedures to assess site suitability to stockpiling and spreading biosolids are already in place to minimise the risk of pollution. These assessments include (a) farm survey, (b) field survey and soil analysis, (c) pre-application survey to ensure conditions are suitable and (d) stockpile survey. Site factors routinely taken into account in assessing site suitability include annual rainfall, soil type, slope, land use, flood risk, proximity to a watercourse, conservation status (SSSI, SAC), presence of wells, springs and boreholes and extent of recent underdrainage. Fields which regularly receive biosolid are re-analysed once every 5 or 10 years. In one region, farms receiving biosolid have an independent nutrient management plan conducted to help farmers reduce their additional fertiliser costs and ensure soil nutrient levels do not become excessive in individual fields. Slope assessments are designed to reduce runoff risk from stockpiles and also for historic reasons where liquids used to be applied. Stockpile size is a logistical and economic issue and some materials are liable to slump after rain. Stockpiles are not covered due to the impracticalities but a top layer of compost does help to reduce odour problems. The optimum size of a stockpile is ca. 500-750 tonnes to allow spreading in one day. If the site is risky for stockpiling, then control options (straw bales, soil bunds, divert water away by channels) are implemented. Stockpiles must be sited 130 m away from buildings and 30 m from a surface water where the stockpile might compact the soil causing runoff, or where soil is liable to waterlogging, or within 10 m of a field underdrain (in practice hard to avoid). There is extensive avoidance of flood risk sites. Stockpiles are investigated at least every 6 months. Further details of the regional visits undertaken are given in the first annual report.
2.1.1 Current knowledge on phosphorus loss
Diffuse P loss in land runoff is a natural process which is accelerated by the accumulation of surplus P in the soil and the increased quantities and solubilities of the different P amendments applied to agricultural land (Sharpley and Withers, 1994; Sharpley et al., 2000). Whilst the precise impacts of inorganic and organic P forms in land runoff on aquatic ecology are often difficult to distinguish from other contributory factors or pollutants, there is general consensus that eutrophication problems in freshwaters are linked to concentrations and loads of P (Correll, 1998; Foy 2005; Mainstone and Parr, 2002). In sensitive ‘unimpacted waters’, only small increases in the annual average P concentration are required to cause a range of undesirable symptoms. Agriculture has been implicated as a major source of P in many waters where point source discharges from STWs have been reduced (Foy et al., 1995; Jarvie et al., 2001). In contrast to point sources, diffuse P losses from agriculture are spatially and temporally very variable, and difficult to quantify and trace back to specific fields (Edwards and Withers, 1998). They originate from a number of different areas within the catchment, are transported by a number of different hydrological pathways and arrive in the watercourse at varying times, often when biological activity is low and some distance from their source area.
A large amount of research has been conducted worldwide over the last 10 years to improve our understanding and knowledge of P loss from agricultural land. Losses of P occur when it rains, the resulting flow rates and flow pathways being determined by inherent landscape type factors (eg rainfall, soil type and slope) and the extent to which these have been, or are, modified by land management (eg land use, presence of underdrains, degree of soil compaction, P inputs). Research suggests that rates of P loss at the field scale where flow is initiated are highly sensitive to land use, land management practices and the management of P inputs. At the catchment scale, flow accumulates in larger quantities and the effects of individual management operations at the field scale become diluted unless a critical density of these operations occurs within the catchment. Past examples of major changes in farming systems which might have increased diffuse P loss include large scale land use change (eg winter cereals where the majority of erosion occurs, Chambers et al., 2000), installation of underdrainage systems or a switch to slurry-based livestock handling systems (Withers et al., 2000). Similarly, the repeated spreading of biosolids within the locality of a STW might constitute a critical farming operation affecting eutrophication due to the widespread build-up of soil P levels within the locality.
Phosphorus loss in surface and sub-surface runoff occurs in dissolved form (arbitrarily defined as the P passing a 0.45 m filter, DP) and in particulate form (PP) in association with eroding soil particles, and aggregates of particles. Soil P is therefore a major source of P release to runoff that has become accelerated in recent decades. Calculations of the P surplus which has accumulated in UK soils from applications of fertilisers and manures indicates that soil P levels have doubled since about the 1930’s (Withers et al., 2001a). Intensification of farming methods has also increased the risk and incidence of erosion (Spiers and Frost, 1985; Evans, 2005; Chambers et al., 2000) and eroded particles are finer–textured and therefore preferentially enriched with P compared to the bulk soil. Site factors influencing the risk of runoff, erosion, soil wash and associated phosphorus in agricultural runoff include rainfall, soil type (texture, dispersivity and permeability), slope, land use and land management. Land uses which are high risk of sediment and P loss include late-sown winter cereals, potatoes, sugar beet, field vegetables, outdoor pigs, grass reseeds, forage maize, outwintering stock and forage crops grazed over winter (Defra, 2005). Land management factors that increase the risk of runoff and erosion are tillage practice, sowing dates, direction of cultivations and soil compaction.
Additional ‘incidental’ P loss may also occur following fresh P additions to the land which have not had time to equilibrate with the soil (Haygarth and Jarvis, 1998). In the broadest sense, they also include P losses originating from permanent manure stores, concrete feedlot areas and farmyards even though they are more point source in nature. Decaying vegetation including crop residues can also release P to runoff, although the relative contribution of this source is not well established. Incidental P loss rates are very variable (<1 to 25% of total P applied) but can become a more dominant source of P than soil depending on the amount of P applied, the P release properties of the materials applied (% P extractable in water), the timing of storm events after application, soil conditions and the amounts of runoff generated (Sharpley and Moyer, 2000; Withers et al., 2003a). Large P applications left on the surface of wet, frozen, compacted and intensively underdrained soils in high rainfall areas are particularly vulnerable to P loss. Concentrations of P in runoff are often greatest during the first storm event following P application, but can remain high for several weeks, or even months after application (Smith et al., 2001; Withers and Bailey, 2003; Withers et al., 2003a).
Soil P, fresh P additions, crop residues, and farm buildings therefore represent the main sources of inorganic and organic P which become mobilised in land runoff by detachment, dissolution and desorption (Fig. 1). Once mobilised, P is delivered to the watercourse either rapidly by overland flow and by-pass flow in field drainage systems, or more slowly by sub-surface lateral flow and soil matrix flow via groundwater. Opportunities for retention of P during the delivery process are dictated by the speed of transfer, the degree of connectivity and the type of landscape (Fig. 1). During transport and on entering the water, the amounts of DP and PP continually change as ‘new’ sources become transported from different areas within the field, farm and catchment.