Phosphate Recovery
Home | Introduction | The phosphates file | Current research projects | Bibliography | Contact us
Phosphate removal & recovery from wastewaters
The impact of the phosphate industry on the environment is not solely felt when phosphates are processed, but also when they are used. New methods for recovery of phosphates from wastewaters would lead to an improved level of sustainability in the phosphate cycle. But what are the practicalities?
While the phosphate industry has a major impact on the environment at the point of production because of the wastes generated in the production and processing of phosphate rock for phosphoric acid manufacture (for details see Phosphorus & Potassium 209 and 211) that is not the end of the story. Those phosphates are used to grow the crops that feed both animals and man, and also to make non-food products such as animal feed phosphates, technical phosphates and detergents. As a result of their consumption they finish up back in the environment through agricultural waste, groundwater run-off, urban wastewaters and sewage. Viewed in this way, the result of current human endeavour is to take mineral phosphate from discrete deposits around the world (located chiefly, but not exclusively, in the USA, North Africa, the Middle East and the former Soviet Union), and to disseminate it to all corners of the globe where, after primary use, it has the potential to cause eutrophication of inland bodies of water and algal blooms in the seas.
Steps have been taken in many countries to eliminate the worst excesses in agricultural practices that release phosphates back into the environment, and for many years environmentalists have campaigned hard over specific uses, such as detergent phosphates, that were thought to be the cause of the most visible damage. But the major elements of this phosphate pathway still remain substantially in place. Removal and recycle of phosphorus values from all main waterborne sources would lead to far greater sustainability for the worlds phosphate cycle, especially as producers are having to adapt their technology to deal with the falling grades of the mineral deposits that remain. This is the view propounded in a recent report undertaken for the Centre Européen dÉtudes des Polyphosphates E V (CEEP).1
Origins of phosphate in waterborne sources
Apart from the phosphates that are
leached from natural mineral deposits, phosphates enter groundwater and
wastewaters from two basic sources:
l Agriculture
2
Liquid urban waste, sewage disposal
Those relating to agriculture (from leaching of fertiliser nutrients applied to crops, as well as manures produced by livestock rearing), are largely localised within areas of intensive agricultural activity. These have created environmental problems when water catchments drain into standing bodies of water, such as eutrophication in lakes. In such environments, phosphate is usually the nutrient in most limited supply, and, therefore, its increased availability leads to marked increases in the growth of algae and phytoplankton.
Generally, total phosphorus concentrations in excess of 100 µg P/litre provide sufficient nutrient enrichment in lakes for there to be a probability of eutrophication, and there are many freshwater streams in rural areas in Europe that contain concentrations in excess of 200 µg/litre.2 In other words, phosphate losses from agriculture, while they may be small in agricultural terms, can have a significant effect on the environment.
Despite the fact that urban sewage is still considered to be the dominant source of phosphorus impacting freshwater resources on an overall basis, in locations where lakes are draining intensive agricultural areas, agricultural practices can be the predominant source of phosphorus. It is for this reason that the fertiliser and farming community have begun to address the issue of phosphate management at the farm. Modern farm systems, especially those using intensive methods, have seen the need to maintain nutrient balances that include P, and to develop systems to control phosphorus surpluses at the farm level. The term "precision agriculture" has become the new buzzword to describe such technically advanced farm management. But nutrient management planning must take account of animal manures as well as nutrients supplied as fertiliser, whether animals are reared in confinement or at grazing, if the goal of sustainable agriculture is to be fully achieved.
Generally, however, phosphates arising as potential pollutants from agricultural sources are relatively small and isolated compared to the phosphate loadings contained in liquid urban wastes. By contrast with agricultural wastes and soil run-off, sewage and urban wastewaters arise in large conurbations and in far greater quantity; and the potential for adverse environmental impact from the phosphates they contain are by far the most serious. Estimates vary, but urban residents discharge 2-3g P per capita per day into the sewerage system; that is 2-3 tonnes P equivalent per day per million of population. (In rural localities where there is no sewerage collection system, these phosphate loadings are still discharged back into the environment via septic tanks. They find their way back into the surrounding soil and natural drainage systems, making their contribution to the pollution of lakes and streams too!).
According to estimates made for the UK, an average of 1.2 g P/capita/day enters urban sewage systems from the diet, with a further 1.3-1.8 gP/capita/day from other household activities, including contributions from household detergents.2 Recent reductions in the phosphate content of household laundry detergents have been partially offset by increases from the use of domestic dishwasher detergents. It would seem unlikely that domestic discharge of phosphates into sewage will ever fall far below 2g/capita/day in developed countries such as those of the European Union.
Impact of detergent phosphates
For almost 20 years, environmentalists
have labelled detergent phosphates as the major cause of eutrophication
in lakes and other bodies of freshwater, and legislation has been introduced
in many countries to prevent their use in detergent formulations.
Much of the adverse opinion expressed
against the use of phosphates and polyphosphates in detergents, however,
has now been called into question by serious scientific research from reputable
sources.
In 1994, the Netherlands Organization
for Applied Scientific Research (TNO), Institute of Environmental Sciences,
published the results of research showing that the use of phosphate-free
detergents gives no improvement in surface water quality, and leads to
eutrophication that is in many cases worse than that experienced when phosphate-based
detergents are used. The main reason for this is that the alternative non-phosphate
detergent ingredients were found to be so toxic that they kill the vital
zooplankton that feeds on algae and phytoplankton in naturally balanced
freshwater systems. TNOs research concluded that there is "no environmental
reason to exclude STPP-based detergents from a green product nomination".3
In 1995, a three-year study in Englands
Lake District was reported demonstrating that phosphates could be used
beneficially in the rehabilitation of lakes that had been "killed" by acid
rain. Phosphates have well-known buffering properties, and additions of
sodium phosphate were found to be far more effective in controlling the
pH of acidified lakes than the more widespread practice of using lime neutralisation.4
Phosphates from wastewaters as a potential resource
Whatever the rights and wrongs, or
the "pollution" arguments of environmentalists, it is clear that urban
wasterwaters and sewage contain concentrations of phosphate that would
represent significant tonnages of valuable plant nutrient if they could
be recovered economically. At average loadings of 2 gP equivalent/capita/day
(dietary and non-dietary components combined), around 5 tonnes of P2O5
equivalent is discharged each day to municipal sewage systems per million
of population. Thus, for the European Union as a whole, the total might
amount to more than 2,000 tP2O5/day. Even if all non-dietary phosphate
were eliminated, the remaining phosphate loadings in sewage would probably
still be greater than half this figure.
CEEP argues that, apart from Finland,
there are no commercial phosphate rock deposits in the European Union,
and the phosphate content of urban wastewaters could represent a valuable
resource. Furthermore, EU legislation on water quality the EU Urban Wastewater
Treatment Directive (91/27/EEC) which sets out limitations to both phosphorus
and nitrogen concentrations in wastewaters, will also prohibit disposal
of sewage sludges at sea after 31 December 1998. Other directives also
set standards for alternative disposal methods for sewage sludge, which
include landfill and incineration (see Table 1).
The practice of recycling sewage
sludge to farmland appears to be in decline, and the use of secondary treatment
of sewage sludge to reduce its biological oxygen demand actually results
in practically all the phosphate contained in the original sewage being
solubilised in the wastewater. But wastewaters containing more than 1-2
mgP/litre cannot be discharged into "sensitive waters." It is against this
background that the CEEPs study into phosphate removal and recovery was
undertaken.
The study is a comprehensive review
of available technology spanning the whole ambit of phosphate removal and
recovery from wastewaters (from the biological processing of sewage sludge
for recycle on the land), right through to processes that yield isolated
phosphate chemicals for recycle or re-use.
Sewage treatment and phosphate removal
The treatment of sewage can consist merely of the primary sedimentation of the solids, but nowadays usually also involves secondary treatment (aeration), so that the biological oxygen demand is reduced sufficiently for wastewaters to be discharged into natural bodies of water. The sedimented sludge from primary treatment retains much of the phosphorus in organic form, and can be usefully applied to the land if contaminants, such as heavy metals, are not present at too high a level. The aeration process, however, transforms practically all the phosphorus into the water soluble orthophosphate form. Typical wastewaters from such treatments can contain 10-25 mgP/litre, which is far in excess of the EU wastewater directive for sensitive waters.
Phosphates in solution can be precipitated by the addition of flocculating reagents, such as iron and aluminium salts. The degree of phosphate removal depends on how these are added (see Fig. 1 and Table 2).
If the reagents in a two stage sewage treatment system, involving both primary and secondary treatment, are added prior to primary sedimentation, typically 70-90% of the phosphorus is removed into the sludge. When added at several points in the treatment system, the removal can be 80-95% or more. The phosphate thus removed is separated along with other precipitated impurities and is normally amalgamated with the rest of the sewage sludge from the treatment process, for disposal. The sludge mass can increase up to 50% and its volume up to 150% depending on the chemicals used and the nature of the whole range of impurities present in the water-soluble phase.
Such methods can consume considerable
volumes of chemical reagents (precipitants and flocculants) and, owing
to the cost, have been superseded to a large extent in recent years by
biological phosphorus removal systems (BPRs), which are cheaper to operate
but also remove the phosphate to the sludge. Such methods rely on the fact
that activated sewage sludge that has been subjected to vigorous aeration
can take up phosphate beyond that required for its microbial activity.
In a conventional activated sludge
plant, phosphorus removal to the sludge can be 20-40% of that present.
In BPR processes, the increase in phosphorus removal depends on the biodegradable
chemical oxygen demand in the sludge (see Fig. 2).
A number of variations on the basic
technology exist, and combinations involving both chemical and biological
treatments are possible. Some have been developed with retrofitting in
mind, while others involve combinations of bioreactor systems to replace
existing sewage treatment operations.
Processes that use advanced bioreactor
systems, including fluid-bed bio-reactors, have also been developed, as
well as integrated processes such as the Hypro concept, involving hydrolysis
of sludge with sulphuric acid as a key processing step. Invariably, such
processes give rise to a wastewater that is substantially phosphate-free,
together with a phosphate- enriched sludge, which may or may not have a
demonstrably significant agricultural value.
Treatment of sewage sludge
In contrast to past practices, CEEP
says that comparatively little sewage sludge is recovered and recycled
to farmland nowadays, and that the current trend is for less rather than
more sludge to be disposed of in this way. Indeed, CEEPs report observes
that effective methods for returning the valuable nutrients contained in
urban sewage to the land have largely remained elusive.
Nevertheless, the report does examine
technology available for improving the properties of sludge to enhance
its handling characteristics as well as its agricultural value, such as
the Simon N-Viro composting process and the Swiss Combi drum dryer process.
Figure 3 shows the N-Viro process, which dates back to the 1970s, and involves controlled composting at high temperature (50°C/122°F plus) with alkaline additives (e.g. cement dust and lime), in order to pasteurise the sludge and inactivate pathogens and micro-organisms. A dry "soil" that can be used as a growing medium when mixed with normal soil is the product of the process.
The process is claimed to have the
advantage of binding heavy metals present in sewage sludge in an insoluble
form, so that they do not leach out to be taken up by crops. However, disadvantages
include odour problems (chiefly ammonia) that need to be controlled, as
well as uncertain nutrient values. N-Viro has been demonstrated in the
USA and was used in the past by Southern Water in the UK, however, this
water company has now substituted a different composting process.
The Swiss Combi drum dryer process,
shown in Fig. 4, was developed by W. Kunz AG in the 1980s and has
been adopted by Wessex Water in the UK in order to comply with EU directives
and UK legislation banning the disposal of sewage sludge at sea. Essentially,
the process involves drum drying and pelleting of dewatered sewage sludge
at 150°C (302°F). Combustion gases at 450°C (842°F), which
may be derived from adjacent biogas production, are used to provide heat
for the dryer, but it is claimed that up to 70% of the energy used can
be recycled, and electrical energy exported to the local grid.
Wessex Water claims that if the
process is fired on biogas derived from sewage digesters, then typical
operating costs are £50-60 ($80-100)/tonne of dry solid. The dried
product sells for £17-20 ($27-32)/tonne, but Wessex Waters Avonmouth
plant is also able to sell electricity and hot water off-site.
The main objectives of such processes are to sanitise the sludge and improve its handling characteristics, transforming it at low cost into a saleable product. The economics depend crucially on local circumstances, and to date no process of this nature is in widespread use. To recycle phosphorus back to the land on a broader scale, it would appear that other solutions are needed.
Phosphorus recovery
In Japan, limited landfill capacity
and stringent regulations governing agricultural re-use of sewage waste
have combined to make the incineration of sewage sludge the preferred method
of disposal in many cases. Accordingly, this has presented the challenge
of recovering phosphorus values from the incinerator ash, and laboratory
scale work at Chuo University has shown this to be feasible.
Sludge obtained from a so-called Enhanced
Biological Phosphorus Removal (EBPR) system, containing on average 7-10%
P on a dry weight basis, is typically dried and then incinerated at 670°C
(1238°F). It was found that the phosphorus values contained in the
incinerator ash could be leached with water and precipitated using ferric
chloride to yield a mixture of ortho- and poly-phosphates, depending on
the precise temperature (in the range 30-55°C, 86-131°F) at which
precipitation takes place. Although incineration is a potentially high-cost
disposal method for sewage sludge, such a phosphate recovery method may
prove to be of interest where no alternative to sludge incineration can
be considered.
Other options focus on the prospect
for recovering the phosphorus values from sewage treatment wastewaters
which, depending on the sewage treatment method, may often contain the
bulk of the phosphorus originally present in the raw sewage. Adsorbents
such as activated alumina, partially burned dolomite (calcium/magnesium
carbonate), and red mud (a residue from bauxite refining in the aluminium
industry), have all been shown to be suitable for removing phosphates from
wastewaters without the need for other reagents to overcome alkalinity.
Aluminium phosphate compounds have dubious value as phosphate fertilisers,
while magnesium is a secondary nutrient often added to fertiliser products
in any case.
Such processes have the advantage
that the phosphate removal step does not lead to additional tonnages of
sludge to process or dispose of, since the adsorbed material is not added
back to the sludge. However, further processing would probably be necessary
to develop a product of sufficient fertiliser value to have good potential
for recycle of phosphate to the land.
Useful phosphate products from waste water
Perhaps the best prospect for recycle
of phosphates from urban sewage wastewater lies in processes that precipitate
discrete crystalline products of reasonable purity, that could be used
directly as fertiliser materials or as a substitute raw material in standard
phosphate fertiliser production technology.
Several processes of this nature
have been developed. Main products are usually either (insoluble) hydroxy-apatite,
42% P2O5 when pure, or (soluble) struvite, magnesium ammonium phosphate,
nominally 8:41:25 N:P2O5:MgO if dehydrated, depending on the processing
conditions. The analysis of the material being processed, whether (nitrogen
containing) sewage sludge or (substantially nitrogen-free) wastewater,
for example, its pH and other factors including the precipitants used (i.e.
magnesium compounds, caustic soda or lime) all have an effect on the most
appropriate product to aim for. Most of the processes do not result in
the production of any additional quantities of sludge.
Crystalactor process
Developed by Dutch consulting engineers DHV in the late 1970s, the Crystalactor process has been used commercially in water softening and other water purification applications for some years. For phosphate removal, the process aims to produce calcium phosphate (hydroxyapatite) in a granular form using a fluidised-bed reactor/crystalliser. The phosphate crystallises on seed grains, usually sand (see Fig. 5).
The process involves preliminary degasification (carbon dioxide removal) of the phosphate laden wastewaters with 96% H2SO4 sulphuric acid and then crystallisation of calcium phosphate by the controlled addition of lime or caustic soda. It has been found that dosing with caustic soda (OH ions) is often more effective than dosing with lime (Ca ions), since it increases the driving force for the conversion of the phosphate to tricalcium phosphate. In practice, dosing with lime is only necessary in "soft" wastewaters.
A demonstration plant was built at
Westerbork, Netherlands, in 1988, where the phosphorus loading (including
organic phosphorus) in raw sewage of 16 mgP/litre is first reduced by conventional
sewage treatment to 9 mgP/litre. Here, the Crystalactor process reduced
the phosphorus loading in the final effluent to less than 0.5 mgP/litre
while producing up to 40 tonnes of 2 mm diameter calcium phosphate pellets
per year. These typically consist of 40-50% calcium phosphate, 30-40% sand,
and up to 10% calcium carbonate. Employing ground pellets as seed materials
eventually reduce the sand content, so that a product with higher calcium
phosphate content is obtained.
However, with the reduction in use of phosphate-based detergents in the Netherlands, the economics of the basic process have become more questionable. In the mid-1980s, for example, typical P concentrations in raw sewage averaged 15 mg/litre with peaks up to 23 mgP/litre, whereas today the respective figures are 10 mgP/litre and 15 mgP/litre. Installing an additional phosphate enrichment step (such as DHVs Phostrip) can produce enriched wastewater streams containing up to 50 mgP/litre, which results in reduced equipment sizes and better economics for the phosphate removal step (see Table 3).
CSIR process
In South Africa, CSIR has also developed a fluidised bed crystallisation column at laboratory scale for removal of phosphate from a variety of influent streams. Phosphate can be removed either as hydroxyapatite or struvite, according to the make-up of the feed stream, and retention times in the range of 3-10 minutes result in over 90% P removal at pH controlled between 8.0 and 9.5.
Struvite crystallisation can occur naturally at sewage treatment works if the right molecular ratio of P:N:Mg exists (roughly 1:1:1). This can cause scaling problems in pipes. CSIR has therefore proposed a process that combines sewage treatment with phosphate crystallisation, such that both hydroxyapatite and struvite can be obtained as crystalline products (see Fig. 6).
The process claims to be capable of reducing phosphate loadings to less than 0.1 mgP/litre from raw sewage intake containing 30-80 mgP/litre. Conditions for struvite precipitation are optimum at pH above 8.0, while those for hydroxyapatite are best at pH above 9.5.
Kurita process
Using similar chemistry to that in the DHV and CSIR processes already described, Kurita Water Industries, Japan, has developed a process based around a fixed bed crystallisation column. The Kurita process is designed to remove phosphate from the secondary effluent of sewage treatment works using phosphate rock as seed material.
The fixed bed column reactor is packed with 0.5-1.0 mm phosphate rock particles. The secondary effluent passes upward through the column, having been conditioned with the addition of calcium chloride and caustic soda. Hydroxyapatite is precipitated. It is claimed that the residual phosphorus content of the effluent lies below 1.0 mgP/litre. However, it is not clear how the crystals are removed from the reactor.
OFMSW/BNR
A three-stage process that integrates
anaerobic digestion of the organic fraction of municipal solid waste (OFMSW),
biological nutrient removal (BNR) and phosphate crystallisation, has been
developed in a joint research effort by three institutions in Italy and
Spain Department of Environmental Sciences, Venice, Department of Materials
and Land Science, Ancona, and Department of Engineering, Barcelona.
Solid municipal (food) waste collected
from shops and institutional catering establishments, is pulverised and
added to primary sewage sludge and co-digested. Anaerobic fermentation
of the OFMSW provides fatty acids and other biodegradable compounds that
act as a carbon source for efficient BNR. The BNR step operates on the
liquid phase, while the solid phase is processed in further anaerobic digesters,
receiving the phosphate-enriched liquor from the BNR.
Effluent from this combined treatment
has high concentrations of Ca, Mg, PO4 and NH4 ions, in a stoichiometry
suitable for precipitation of either hydroxyapatite or struvite. Seed grains
of sand or quartz assist nucleation of the crystals in either a packed
bed or fluidised bed reactor. Typical phosphate concentrations entering
the reactor range from 28-81 mgP/litre, and phosphorus removal efficiencies
of 52-87% are claimed. At this stage, the integrated technology
has only been demonstrated at laboratory scale.
RIM-NUT Ion-Exchange
The RIM-NUT ion-exchange/ precipitation
process, developed by Italys Water Research Institute and the University
of Bari, removes both ammonia and phosphate ions from wastewater obtained
in tertiary sewage treatment. Research was motivated by stringent Italian
discharge laws that specify maximum admissible concentrations of 15 mgNH4/litre
(10 mg total N), and 10 mgP/litre for sea outfall or 0.5 mg/P/litre for
lakes.
Small scale pilot plant tests were
undertaken in 1981/2, after which a 10 m3/hr demonstration plant was operated
in 1983. The process involves a two-stage ion exchange, using two columns
of cation exchange resin and two columns of anion exchange resin, followed
by precipitation of the nutrients. The cation exchange removes ammonium
ions from the water, while the anion exchange removes the phosphate. Sodium
chloride (NaCl) is used to regenerate both the anionic and cationic resins.
The regeneration eluates are fed
to a precipitation/settling tank where the Mg:N:P ratio is adjusted to
1:1:1 if necessary by the addition of further phosphate and magnesium salts.
Caustic soda or soda ash is then added to precipitate struvite, MgNH4PO4.6H2O,
which is settled and separated. Phosphate removal efficiencies greater
than 95% were achieved in "closed loop" operation, in line with the level
of removal necessary to meet the discharge regulations for lakes. As used,
the RIM-NUT process, shown in Fig. 7, was only ever required to
process a fraction of the total effluent from the sewage
treatment facility at West Bari.
The magnesium ammonium phosphate product consists of more than 93% inorganic material and corresponds to a fertiliser grade of 5:27:15 (N:P2O5:MgO). Dehydration at 90°C (194°F) gives an anhydrous product analysing 8:41:25.
Unitika Phosnix process
The Phosnix process developed by Unitika Ltd, Japan, is a tertiary phosphate removal system and is based on the crystallisation of struvite from an enriched phosphate stream.
Secondary sewage digester effluent is fed to the base of an air-agitated reaction tower, where caustic soda and magnesium chloride are added to arrive at the correct molar ratios and pH for the magnesium ammonium phosphate to precipitate at ambient temperature. Blown air fluidises the reactor contents so that the crystals have time to grow to a point where they sink as pellets to the bottom of the reactor for removal. Dewatering leaves a product that has a residual moisture content of about 10%, and analyses 5.3:13.3:11.5 (N:P:Mg), compared to 5.7:12.6:9.9 for pure struvite, MgNH4PO4.6H2O.
Phosphate removal efficiencies in the range 88-97% are claimed, on effluent streams ranging from 30 to 905 mgP/litre. Unitika also claim that the sales of the struvite product as fertiliser cover the cost of operating the process.
Sydney Water Board
A phosphate removal system, claiming
to produce a crystalline calcium phosphate product from a variety of wastewaters,
was developed in the early 1990s by Sydney Water Board, Australia. This
process, which is still understood to be at the laboratory scale, also
involves crystallisation in an agitated bed reactor.
Wastewater is first decarbonated
(if necessary) by aeration in the presence of added sulphuric acid. Additional
calcium ions are then added in the form of gypsum. A moving bed of fine
magnesia particles in the reactor serves to seed the crystallisation,
so that product size crystals can be grown in a retention time of 0.5-2.0
hours.
The dosage of calcium ions added
as gypsum is controlled in order to achieve a phosphorus concentration
in the process effluent of less than 0.3 mgP/litre, whatever the phosphate
loading of the influent wastewater may be. The process is said to be equally
suitable for treating effluents from primary, secondary or tertiary sewage
treatment, and is thus able to deal with wastewaters containing "several
hundred" mgP/litre (see
Fig. 8).
What now?
For such phosphate recovery processes
to succeed on a wide enough scale to affect the "sustainability" of the
industrial/agricultural phosphate cycle, both legislative and economic
incentives must act in a positive way. The economics of the processes is
critical. Cost indicators such as the phosphate recovery levels achievable,
the consumption and cost of reagents, the practical need for "seeding"
materials at full industrial scale, and the ultimate fate of heavy metals
contained in municipal waste will all have an influence on the relative
cost and value of the products obtained.
Many of the advanced processes discussed
are still at the development stage and it is likely that some practical
problems remain to be solved. Furthermore, it has been argued that some
of these processes involve what is in effect "alien" technology to conventional
sewage treatment practice (rather similar to the "alien" nature of fuel
cell technology in power generation, which was discovered over 100 years
ago and is only now beginning to be taken seriously! See Nitrogen
magazine).
CEEP says that the barriers to the
introduction of such techniques are not fundamentally technical, but stresses
that imagination and leadership are required for them to be accepted widely.
Who is likely to provide these? For the phosphate fertiliser industry,
what incentives are there to use these recovered phosphate products as
they come available, and how will they be marketed for their plant nutrient
value?
It is also possible that current
legislative pressure to reduce phosphates progressively in wastewaters
is actually working against the economics of recovery of what is left behind.
Is there possibly a case for returning to the use of phosphates in detergents,
and phasing out phosphate precipitation in sludge, so that the phosphate
loadings in wastewaters are increased to the point that they are economically
recoverable? Is it just possible that an incentive of this kind may bring
the imagination and leadership that CEEP refers to? n
References
1 "Phosphorus removal and recovery technologies", produced by Environmental and Water Resource Engineering section, Imperial College of Science, Technology and Medicine, London, for Centre Européen dÉtudes des Polyphosphates EV, Selper Publications, London, 1997.
2 Foy, R.H., Withers, P.J.A.: "The Contribution of Agricultural Phosphorus to Eutrophication", Fertiliser Society Proceedings, No. 365, April 1995. (Described in brief detail Phosphorus & Potassium 198)
3 "STPP the Green Choice?", Phosphorus & Potassium, 195, pp. 23-26, (Jan/Feb 1995).
4 "More good press for phosphates", Phosphorus & Potassium, 200, pp. 2, (Nov/Dec 1995).