INTERNATIONAL CONFERENCE ON PHOSPHORUS RECOVERY FROM SEWAGE AND ANIMAL WASTES
Warwick University, UK. 6th & 7th May 1998
Summary of conclusions and discussions
by Chris Thornton, CEEP
Summary of conclusions
Summary of discussions
(1) Boundary conditions for phosphorus recovery
(2) Phosphorus recovery and the water industry
(3) Phosphorus recovery from animal waste
(4) Calcium phosphate recovery
(5) Struvite formation as a route for phosphorus recovery
(6) Economic feasibility of phosphorus recovery
Summary of Conclusions
World wide interest
in Recycling Phosphates from Sewage
Partnership for the future
The first international conference
on phosphate recycling, organised by the Centre Européen
d'Etudes des Polyphosphates (CEEP), brought together more than 100
industry and water company decision-makers from across the world,
along with scientists and researchers, regulators from environment
and agricultural authorities and the European Commission.
Political will is needed to exploit the potential for phosphate recovery from sewage and animal wastes, the conference at Warwick University concluded. "This will require a partnership involving the water and animal waste industries, government authorities and the phosphate industry" affirmed John Driver, chairman of the conference. "The phosphate industry is taking the lead in this. We are convinced that the future lies in phosphate recycling and consider that within a decade up to 25% of phosphates used in detergents and other high-grade applications could be recovered from sewage and animal wastes".
Potential for P recovery
The quantities of phosphorus present in sewage and
animal waste are significant compared with the needs of the detergent
and high grade phosphate industry. EC regulations and local environmental
objectives are making phosphorus removal from waste waters increasingly
widespread. These two facts mean that in the future phosphorus recovery
for recycling could become economically viable. The conference heard
that in the UK around 40 million tonnes of domestic sewage* is produced
each year as well as some 150 million tonnes of farm animal wastes.
A further 150,000 plus tonnes is excreted directly onto fields.
* tonnage of organic wastes present in sewage before dilution with tap and rain water.
These wastes can be estimated to contain around 45,000 and 200,000 tonnes, respectively, of phosphorus (as P). Together this represents approximately six times the consumption of phosphate products in the UK (detergent phosphates and other high grade applications such as metal cleaning, pharmaceutical and food uses).
Recycling -- a viable option
It is likely that a significant proportion of this
P would be available for recovery. Logistical reasons, such as the
cost of transport and the scale of installations, will make recovery
for recycling likely to be an economic option only in the case of
large, geographically concentrated waste streams (sewage from urban
areas, intensive animal production). In rural areas, agricultural
sludge or manure spreading will remain the best option for recycling
Only part of the P in each waste stream can feasibly be recovered. In sewage treatment, for example, recovery is most efficient in side stream processes and in this case a significant proportion (probably 40 50%) of the P will be "lost" into sludges. A higher proportion could be recovered by hydrolysing the sludges to release trapped organic phosphates.
However, in the UK, even a conservative estimate of the potential for P-recovery and recycling (50% recovery applicable to 25% of sewage and to 15% of animal wastes) represents half of industrial phosphate demand. Fertiliser and animal feed, however, use 5 - 6 times more phosphates than industry.
The obligation to remove P from waste waters
The European Urban Waste Water Treatment Directive
91/271/EEC (21st May 1991) requires P removal to be installed at
all, except very small, sewage works in eutrophication sensitive
areas by December 31, 1998. Eutrophication sensitive areas are designated
by national governments of member states. These are all lakes, rivers,
estuaries and coastal waters which are already eutrophic, or may
be vulnerable unless protective measures are taken. Within these
sensitive areas (by Article 5.2 of the Directive) P-removal must
be installed at all sewage works serving conurbations of more than
10,000 pe (person equivalents). This treatment (Annex I.2.b of the
Directive) must remove at least 80% of phosphorus and reduce concentrations
in the outflow to below 2 mg/l (1 mg/l if > 100,000 pe). 70-80%
of dissolved nitrogen must also be removed.
Although this Directive is not yet universally applied, and the 1998 deadline will not be fully respected in some Member States, it is clear that P removal will increase in sewage treatment across Europe in the near future.
The national definition of sensitive areas is subject to regular revision. A tighter definition would bring more waste water treatment plants into the remit of the Directive.
P removal versus P recovery
Where P is removed from waste waters, it is transferred
to sludge, either in an organic form -- biological P removal, or
as a chemical precipitate --chemical P removal.
The majority of sewage works equipped with P removal in Europe use chemical precipitation, often simultaneous with secondary biological treatment with the result that the chemical precipitate is mixed into the organic sludge. Iron compounds are the most widely used chemical for P removal, although aluminium compounds can also be used. This means that sludge volumes are increased not only by the addition of phosphorus as phosphates, iron or aluminium are also added.
Effective P removal requires higher concentrations of precipitation chemicals than will
actually combine with the available P, further increasing sludge volumes. John Upton, Severn Trent Water, UK, indicated that P removal using iron precipitation increases sludge volumes by 40% on average.
This method also results in increased chloride concentrations in outflows. Iron precipitated phosphorus is not readily available to crops. The excess of iron salts in sewage sludges can fix existing P in soils further reducing its availability to crops. This reduces the value of these sludges for agricultural spreading. Where sludge is incinerated, the phosphorus compounds are non combustible, reducing the calorific value of the sludge and increasing ash volumes needing to be landfilled. Recovery of P for recycling, rather than its transfer into sewage sludges, offers economic and environmental paybacks for the water industry. These benefits must be compared with the investment and running costs of P recovery installations.
Economic and technical feasibility
The conference examined the technical pathways currently
under development as well as the economic boundaries which would
make P recovery a viable option. These discussions are detailed
below and summarised here.
Economic viability will depend to some extend on the price the phosphate industry is prepared to pay for the recovered product, the costs for the water industry of Precovery compared to P removal processes, and predominately on sludge disposal routes and costs.
The phosphate industry might reasonably be expected to pay a small premium for
recovered P, compared with mined phosphate rock, as it contains lower levels of heavy metals and other problematic contaminants than most natural rocks.
Transport costs, however, will be significant for phosphates recovered from sewage
works spread over the country. A medium-sized sewage works, of 100,000 pe, might
produce 1-2 lorry loads of recovered phosphates per month.
The calcium phosphate Crystalactor® plant at Geestmerambacht, Holland, produces
200 300 tonnes/year of phosphate pellets from a 230 000 pe sewage works.
Hideo Katsuura, Unitika Ltd, Japan, indicated that phosphates recovered as struvite by his company are currently sold to the fertiliser industry at around $150/tonne (13.5% P).
The conference looked at the costs of P recovery processes. Several pilot, and full scale plants, have been tested or are already operating in different countries, recovering phosphates from waste water streams through calcium phosphate or struvite formation (see below).
Both pathways can produce pellet-like solids which are readily drained to below 5-10% water and can offer 5 15% phosphorus content. For comparison, mined phosphate rock currently used by industry typically has a 20% phosphorus content.
Other techniques for P recovery exist at the lab or experimental scale, including ion
exchange and recovery from sludge incineration ash.
- pilot plants at sewage works :
- Darmstadt, Germany (Karlsruhe University pilot financed by CEEP, 1997-1998), fluid bed reactor 270l/h ;
- Chelmsford, UK (Essex and Suffolk Water [Compagnie Générale des Eaux]/DHV, 1997-1998), DHV Crystallactor® 700 l/hs ;
- Warriewood, Australia (Sydney Water/Australian Water Technologies fluidised bed demonstration plant, April 1995 March 1996. 50,000 pe and 120,000 l/day) ;
- Westerbork, the Netherlands (DHV Crystalactor®, 1988).
- full scale plants :
- Japan, Shimane Prefecture (Unitika Ltd, struvite production from sewage), 45,000 m3/day, commissioning September 1998;
- Japan, UBE Industries Sakai plant (Unitika Ltd, struvite production), 48m3/day, running since 1995 ;
- the Netherlands, Geestmerambacht (Uitwaterende Sluizen Waterboard Edam, 230,000 pe.) and Heemstede (35,000 pe), DHV Crystalactors® forming calcium phosphate from sewage ;
- Avebe, the Netherlands (struvite from food industry wastes, DHV Crystalactor®, 150m3/h) ;
- Putten, the Netherlands (precipitation of potassium struvite from 700,000 tonnes/year of calf manure, Geochem Research and Delft University) ;
- Gaggenau, Germany (precipitation of calcium phosphate from industrial waste waters, Mercedes Automobiles/Karlsruhe Research Centre), 160 m3/h.
Other full scale plants are known to be functioning using a calcium phosphate recovery process developed by Kurita Water Industries Ltd but precise details are not currently available.
The economics of these operations were examined. Costs are very difficult to assess as these plants have been designed as experimental or demonstration processes and
design and running costs should significantly reduce with engineering and operating
Paybacks for the water industry
P recovery offers the potential to significantly
reduce sludge production compared with established P-removal processes.
The economic payback will depend largely on the sludge disposal
Several participants underlined the tendency to limit agricultural spreading of sludge,
because of local nutrient excesses, risks of heavy metal or pathogen contamination,
local public resistance (odour problems) or because of problems of food industry
acceptance. Quality charters defined by supermarket chains, oil mills or dairy
industries, in many countries, increasingly proscribe sewage spreading.
P recovery can also relieve the water industry of certain operating problems. Struvite
precipitation in pumps and pipes will be prevented downstream of P recovery. Heavy
metals may have a natural tendency to be trapped in recovered phosphates, reducing contamination of outflows or sludges. The phosphate industry is well equipped to deal with these metals with in existing processes.
P recovery is generally readily compatible with nitrate removal, and can even directly
contributes to this, if P is recovered as struvite.
Conditions for P recovery
It is probable that P recovery for recycling will
be viable only in certain situations:
- sewage works subject to the obligation to remove phosphorus ;
- large or concentrated waste streams, ie sewage works in urban areas or intensive animal production, where transport of the recovered P can be rationalised and where recovery installations would be of an economic scale ;
- areas where the low cost option of agricultural spreading of sewage sludge or manure is not available.
Although the proportion of wastes concerned by these criteria is difficult to assess, the conference concluded that it is probably significant and certainly will increase over coming years.
P recycling and the phosphate industry
Phosphates used by industry and in fertiliser manufacture are currently extracted from mined phosphate rock, a non renewable resource. Most of Europe's needs are imported, in particular from Morocco, South Africa and Russia.
A limited resource
Recovery of phosphates from wastes therefore enables
industry to replace an imported non-renewable resource with a sustainable
recycled raw material.
The conference heard from Ingrid Steen, Kemira Kemi Sweden and CEEP Chairperson,
that world reserves of phosphate rock are currently estimated at around 100 years,
depending essentially on the rate of intensification of agriculture in developing
countries. This time scale, based on commercially exploitable reserves using current
technology, is around twice that currently estimated for the depletion of petroleum
Current phosphate rock reserves mean there is no short, or medium, term pressure on
industry to turn to recovered phosphates. The phosphate industry underlined to the
conference, however, that the quality of rock available to industry is already
deteriorating with a lower P content and higher contaminant content (heavy metals, iron, aluminium).
John Driver, Albright and Wilson, UK, indicated that the P content of rock available to
industry has been falling by an average of around 1% per decade.
Sustainable raw material
Recovery of phosphates will, in most processes,
concentrate much of the heavy metal content out of waste waters,
in particular sewage, into the recovered phosphate product. This
may have significant advantages for the sewage works operator. Even
so, the recovered phosphate products will contain levels of heavy
metals -- one or two orders of magnitude lower than those found
in phosphate rock.
The phosphate industry is well equipped to remove heavy metals in its processes, although at a cost. Recovered phosphates offer both a quality premium to thephosphate industry and an opportunity for the water industry to eliminate problematic contaminants.
Cadmium, in particular, has been the focus of much attention in phosphates imported for fertilisers. There is an expectation that standards will tighten significantly in this area. The EC, it was noted, is financing a project to remove cadmium from phosphates mined in Morocco, on site, to avoid their transfer to Europe.
Redefining industry structure
The use of recovered phosphates, produced in appropriate
chemical forms, should not present a processing problem for the
phosphate industry. They would however necessitate a major redefinition
of the industry's structure and logistics.
The European phosphate industry's structure has increasingly moved towards a small
number of large production plants. There is only one remaining phosphorus furnace site in Western Europe (Vlissingen, Holland) and less than a dozen phosphoric acid plants (phosphoric acid is also imported from production plants situated alongside the
phosphate rock mines).
In the long term, the use of small, diffusely produced quantities of recovered phosphates would necessitate restructuring the phosphate industry. Small local production units and appropriate stocking, grouping and transport systems may be needed.
In the short term, to get recovery started, it may be more appropriate for recovered
products to be recycled locally in agriculture.
The conference heard that, despite these challenges,
the European phosphate industry is convinced, at top strategic management
level, that P recovery represents a feasible and positive future.
The organisation of this conference by CEEP, the joint research fund of the European phosphate industry and sector Group of CEFIC (European Chemical Industry Federation), confirms this conviction.
The overall objective of the conference was to look at how to move towards this sustainable vision for the future.
The conference requested that the phosphate industry establish a clear description of its wishes, and requirements, regarding the chemical and physical form of P recovery products - including the impact on the value of such products of different factors.
It was concluded that phosphate recovery and recycling
must involve the high grade
phosphate industry (represented by CEEP, the organisers of the conference), but also the fertiliser and animal feed industries : all of whom use phosphate rock as a raw material.
Recovered phosphates will be most appropriately recycled through one or more industries, depending on the site of production, local markets and the recovered product type.
Recovery for recycling will be logistically, ecologically
and economically viable if it is
developed for waste streams from municipal and industrial waste waters, as well as from agriculture, and recycled back to the consuming industry offering the best local solution.
The conference clearly indicated that from the waste operator's, or regulator's, point of view, there is no reason to distinguish recycling back into agriculture from that into high grade phosphate products as the phosphate industry is perceived externally as a "whole", including all these sectors.
The phosphate industry representatives present (CEEP) agreed to enter into
discussions regarding phosphate recovery with both the animal feed industry and the
mineral fertiliser industry.
Joint research and development
The conference concluded that joint research and development is necessary between the water and waste industries, the phosphate industry and environmental regulators. The objectives should be to improve the economic efficiency and to generalise the technical know-how of phosphate recovery for recycling, both into fertilisers and high-grade industrial phosphates for detergents and industrial uses.
The political and financial support of regulators (environment, industry and agricultural authorities), and in particular the EC Commission, will be important to enable this joint research and development to be launched. Industry called for political leadership and EC research support in this area.
Areas for research
The conference identified the following areas where
scientific research and development are necessary to give a better
understanding of the feasibility and potential for P recovery :
- economics of P recovery : costs, payback for industry, paybacks for the water industry and waste operators, and the way these economics may change over coming years ;
- economic study of existing P recovery plants and estimation of cost improvements possible through wider application and experience ;
- establishment of an economic model of costs and paybacks for P recovery in different situations
- study the chemistry of the two currently developed P recovery pathways (calcium phosphate and struvite formation) : complexities of crystallisation / precipitation of the different complex molecules of these products, their formation conditions and reaction kinetics ;
- lab scale experimentation of calcium phosphate and struvite formation to obtain a better practical understanding of their application in a waste water context : investigation of chemical and physical conditions for phosphate product formation in different processes and waste water media ;
- effects on phosphate product formation of potential inhibitors likely to be present in waste waters : organic molecules, metals;
- potential of phosphate formation for P recovery as a route for removing heavy metals from waste water streams ;
- research into other P recovery routes, in particular ion exchange ;
- technical and economic feasibility of the use of different recovered phosphate products by the phosphate and fertiliser industries, in existing or modified process routes;
- use of struvite and potassium struvite as a fertiliser and by the phosphate industry
- construction of demonstration scale, mobile P recovery installations, for both calcium phosphate and struvite recovery, designed for testing P recovery performance in different conditions (varying operating criteria, different waste water streams ) ;
The conference provided the first contact between a significant number of organisations and individuals who confirmed their intention to put together programmes and initiate research. The diffusion of this document should extend these links to the hundred plus other organisations that expressed interest but were unable to attend.
Summary of Discussions
(1) Boundary conditions for P recovery
Phosphate recovery for recycling will only be technically
and economically viable in certain situations. For municipal waste
water treatment plants, the conference identified the following
criteria which will define this:
- P removal installation :
P recovery will only be viable where P removal is installed. Application of the EC Directive 91/271 will significantly increase the number of sewage works concerned in the coming few years ;
- sewage sludge disposal options and costs, including investment and running costs for thickeners and digesters :
where agricultural spreading is possible, this is usually the cheapest method of nutrient recycling ;
- size of sewage works :
scale of installations and collection costs for recovered phosphates mean that P-recovery is most likely to be attractive in medium to large sewage works ;
- sewage works typology and current state of modernisation :
P removal is more easily integrated into biological P-removal systems or in sewage works where P removal has not yet been installed ;
- N removal :
P recovery can facilitate N removal, via struvite, and is more easily integrated into plants fitted with N removal ;
- quality of recovered P : handling properties, P content, concentrations of impurities:
the phosphate industry may be prepared to pay a premium for a high P product ; levels of most contaminants will be considerably lower than those found in currently exploited mined phosphate rock ;
- presence and options for management of heavy metals :
P recovery can offer the water industry a removal and disposal route for such
- capital costs : critical size/capacity ratio ;
- running costs for P recovery, chemical costs for P recovery compared with P removal by precipitation ;
- impact on sewage works operating problems :
P recovery may facilitate operation by avoiding phenomena such as accidental struvite precipitation in pipes and on pumps ;
- operator know how and "culture" :
P recovery is a relatively complex process to pilot compared with chemical P precipitation but is comparable with the operator control know-how developed in modern sewage works
using biological P removal or nitrogen removal ;
- P content of waste water stream :
higher P contents render P recovery more economic ;
- logistics of recycling of recovered P : transport costs, structure of the phosphate
industry, possibilities for the development of local processing units : the phosphate industry will have to evolve and modify its structure in order to establish a
stable market for recovered phosphate products ;
- public opinion :
the public needs to understand the contribution of P recovery and recycling to sustainable development. The risk of public concern over the "health aspects" of recycling constituents of sewage needs to be addressed.
Participants estimated that in the UK, for example,
P removal is currently installed in
23 sewage works, and will rise to 80 by the end of 1999, and 200 by the year 2000.
These are mainly the country's largest sewage works, so that by this latter date,
around 50% of the UK's municipal waste water flow will undergo P removal.
Increasing costs for sludge disposal, in particular the obligation to incinerate sludge, renders P recovery more attractive. This is due to a reduction in sludge volumes resulting from P removal (elimination of phosphatesand avoidance of the addition of larger than stoichiometric quantities of iron or other salts for P precipitation).
It was affirmed that there is no health risk from industrial recycling of phosphates from sewage. No pathogen canconceivably survive the pH and temperature conditions of the chemical processes used by the phosphate industry. This does not
eliminate the possibility of adverse public opinion associated with the concept of recovering ingredients from sewage and animal wastes for uses such as detergents, food additives, animal food supplements.
(2) P recovery and the Water Industry
Future sewage sludge disposal options
A number of questions remain concerning
the future for sludge disposal options :
- agricultural spreading of sewage sludge is increasingly under pressure, for reasons of nutrient saturation of soils, distance from urban waste water treatment works to sufficient farmland, contamination of soils with heavy metals, pathogens or other
pollutants, local opposition because of nuisance problems and image problems ;
- landfill and sea dumping options are disappearing through environmental legislation ;
- alternatives to spreading include biogas generation, composting and hydrolysis : however, the "markets" for nutrients in the resulting products (and in particular for phosphates) are limited ;
- finally, incineration is a very costly route for sewage sludge disposal. Taken together, these different factors mean that the reduction in sludge quantities
resulting from P recovery can be a very significant economic factor.
Agricultural sludge spreading
There was a general view that the disposal
of sewage sludges to farmlands will
increasingly be limited and expensive in the future.
On the one hand, the areas of land available for spreading are decreasing (for the
reasons given above). On the other hand, the effective application of the EC Urban
Waste Water Treatment Directive will increase sludge volumes. Further tightening of
urban waste water treatment obligations are likely in the future and will accentuate this.
The option of agricultural spreading of sewage will
depend on the local context.
In France, for example, sewage from urban areas is increasingly incinerated :
geographical concentration and suburban development means that there is insufficient farmland locally available for spreading and leads to opposition from inhabitants.
However, in rural areas sewage is generally spread on the land. France is a large
country and small towns away from the large cities have large areas of farmland
accessible. However, even in rural areas, agricultural spreading is being questioned and limited for reasons of contamination from various pollutants.
For animal wastes regulations regarding spreading
on land are also becoming much
tighter, and therefore increasingly costly. Also, the tendency towards intensification of animal husbandry leads to localised surpluses of farm wastes in excess of the capacity of nearby farmland to accept manure.
Farmers and waste producers will have to solve nutrient
management problems if they are to stay in business. P recovery
may provide a solution in some circumstances but there currently
seems to be a significant knowledge gap between industry and waste
providers relating to scale, economics and logistics, etc..
New legislation needs monitoring and influencing so that it is directed towards real
environmental goals. Soils should not be used as sinks for unwanted materials.
Quotas for sludge spreading are often limited by
the nitrogen content, not the P content. This can lead to excess
levels of phosphates in soils where sludges have high P contents
(P removal fitted in sewage works). P recovery could improve this
situation. Bob Foy, of the Northern Ireland Department of Agriculture,
indicated various problems with the agricultural use of sewage sludge.
Increasingly, supermarket chains require that land used to supply
quality assurance produce should not be treated with sludge.
He suggests that sludge from sewage works equipped with chemical P precipitation does not provide available phosphorus for crops. On the contrary, sludge rich in iron or aluminium may deliberately be used as a means of fixing phosphorus and reducing
leaching and runoff.
Sludge disposal and nutrient management
Although P recovery from sludges was seen as making a valuable contribution to
environmental management, the need to solve all sludge disposal problems was also
raised. Bengt Hultman, of the Swedish Royal Institute of Technology, underlined the need to recycle all the materials in waste waters -- nutrients, energy/carbon content, metals
Recovering phosphorus would reduce the value of sewage sludge for farmers if carried
out in isolation.
With the development of P recovery, the problem
of residual sludge disposal still
remains. This has, to some extent, been addressed in Scandinavia : in experimental
processes, sludge is treated with acid, or steam, to hydrolyse it, thus releasing more of the nutrient value. The economic and technical feasibility of such processes remains to be proved, however, as they involve significant energy and chemical input, as well as high investment costs.
Bengt Hultman suggested that hydrolysis of sludge would enable 90%, rather than
40-50%, of P to be recovered for recycling.
Composting is not a viable solution for sludge treatment and nutrient recycling as it is
too energy intensive and has a very limited market due to competition from other
P removal by chemical precipitation
Many European sewage works performing P removal to comply with environmental
outflow regulations currently use chemical precipitation using iron (Fe) or aluminium (Al)
salts. This produces phosphate products which cannot be recycled by industry (see
below) and have little or no value as fertilisers as the phosphorus is poorly, or
non-available, to crops.
Precipitation for P recovery
The need for P recovery technology was widely recognised
as an alternative to
precipitation with Fe and Al salts.
Nic Booker, CSIRO Australia, suggested that more than double stoichiometric
concentrations of iron salts must be added to waste waters to achieve P precipitation down to outflow concentrations of 0,5 mg P/l. Significant quantities of available iron would be present in this sludge. Where this is spread on fields, not only is the phosphorus in the sludge poorly available to crops, but the available iron would further reduce availability of existing soil phosphorus.
P recovery technologies and P recycling in the phosphate industry
Chemical P removal by traditional methods
will generally preclude P recovery for
recycling by the phosphate industry because the phosphorus is precipitated out using
iron or aluminium salts. The resulting iron or aluminium phosphate compounds are
incompatible with technologies currently used in the phosphate industry. They either
requiring excessive energy input, to separate the phosphates from the added
precipitation chemicals, or interfere with the industrial process. John Driver, Albright and Wilson, UK, indicated that approximately 2% iron is the maximum which can be dealt with in the "wet acid" phosphate industry processing route.
Dees Lijmbach, Thermphos, Holland, indicated that 1-2% iron, and approximately 2%
aluminium, are the maximum levels admissible with existing technology in the thermal
Various technologies enable P recovery in a chemical
form that can be recycled by the phosphate and/or fertiliser industries.
The routes for the formation of calcium phosphate, or struvite, are already well
developed, and both demonstration and full scale plants are running. These routes were discussed in workshops at the conference, summarised below. Experimental installations have also been studied using ion exchange, magnetic precipitation and other techniques. These merit further investigation, but their application in a waste water context may prove problematic.
Nic Booker particularly underlined the interest of the "RimNut" ion exchange system
because of its applicability to large volume streams of waste water with high organic
concentrations. In contrast most existing P recovery technologies are best adapted to a relatively "clean" side stream.
Sludge incinerator ash can contain a high level of phosphates, similar to incinerated
chicken litter. The latter may be more appropriate for P recovery. Sewage sludge
incinerator ash is very variable rendering difficulties in the design of industrial recovery processes. Also, if the sludge is produced via chemical precipitation of P by iron or aluminium, these chemicals will prevent P recovery.
P recovery and biological P removal
Biological P removal may offer interesting opportunities
for P recovery. Marc van
Loosdrecht, Delft Technical University, Holland, explained that bio P removal know-how has significantly improved over recent years. Plants have become reliable and offer very efficient P removal. Well managed modern plants generally enable outflow
concentration to remain within regulatory limits without further chemical P precipitation.
Bio P removal plants include a side stream, or return stream, with a relatively high P
concentration in lower volumes than the main stream -- and usually with relatively stable conditions. This is particularly appropriate for P recovery. However, only around 50% of the input P will be available for recovery in such a side stream, the remainder being fixed in the organic sludges.
One of the difficulties is that the P is fixed into aerobic sludge. This is susceptible to
releasing the P if it becomes anaerobic at any stage before effective disposal (such as through spreading or incineration), and in particular, during thickening processes. P
recovery can alleviate this problem by transferring much of the P into recovered
products, rather than into sludge.
Nic Booker, CSIRO Australia, underlined the problems
of process control for bio P
removal. Outflow concentrations of < 0.5 mgP/l are obtained when bio P plants are run by engineers, but often not by waste water operators. The main factor influencing bio P removal performance is usually the control of organic substrate quality and concentration.
It is noted that the kind of engineering know how and control necessary to effectively run bio P removal plants is comparable to that necessary for P recovery technologies, which are also relatively sensitive.
(3) Phosphorus Recovery from Animal Waste
Dees Lijmbach, Thermphos, Holland,
reviewed the presentations at this workshop by
Dr. Haygarth (UK Institute of Grassland and Environmental Research) and Leo van
Ruiten (van Ruiten Adviesbureau, Holland).
Manures contain differing amounts of P. Manure production in Holland is quite
different from the UK in that it is concentrated in three provinces. There are
regulations limiting the level of P application to farmland. For these reasons, the
manure needs to be moved within Holland, exported or processed.
Participants requested better information regarding the quantities and forms of
manure. Some forms may be advantageous for land application while others are
Optimising manure as fertiliser
There was some discussion of reducing the phosphorus
content in non ruminants'
manure by added phytase. It was pointed out by a feed supplier that this has to be done carefully to avoid associated health risks.
It is preferable to optimise the use of manure as fertiliser before looking at other
options. It seems illogical to take the nutrient resource off the farmland and import
chemical fertilisers. This will require improved farming practice : farmers will have to
track the P cycle on their farm, noting incoming feed and fertiliser and outgoing manure and crop products. Farmers need to know how much of the P in the manure is available to crops.
When applying manure to land nitrogen content and odour should also be considered.
There are concerns of storage problems and potential leaching and transfer to water
after spreading. Strategies need to be established to minimise nutrient transfer to
Land spreading may be the best solution in the UK since livestock is not as
concentrated as in Holland but alternatives still need to be developed. Transfer of
manure to neighbouring regions should be considered but poses transport problems and costs.
Incineration is viewed as the next best option to land spreading.
There were concerns that legislation might change these suggested use options. There are economics to consider and these will vary by region and animal type.
Stabilisation of the manure -- to make it available when the farmer needs it and improve handling and transported -- is an important area for research. Production of biogas from manure prior to spreading can reduce odour, eliminate weed seeds and ensure stabilisation and drying -- therefore reducing transport costs. It was pointed out that incineration also addresses these issues.
Phosphorus may be too inexpensive a commodity to make recovery and sale attractive today. But as a valuable and limited resource, the price can be expected to increase over the long term as reserves are depleted -- although there are no trend in this direction can yet be identified.
The role of regulators
Legislation may be the only way forward. Common
Agricultural Policy rules will
change the way we farm. Except in certain cases, such as very large industrial
production units and chicken litter, P recovery from manure is only likely to be
adopted if and when better nutrient management requirements are imposed by
The phosphate industry believes, however, that it is important to initiate
research and development into recovery from animal wastes because they
represent a concentrated and relatively homogenous source of phosphorus
and because a tightening of regulations is highly likely, even if the time scale
is not currently known.
The average farmer does not want to be involved
with the P cycle and would sooner let the experts handle his manure
problem. However, farmers are unwilling to pay the costs of processing
manures. This has led centralised processing factories to fail for
There is no clear competent body to take a role as a player for P recovery alongside the phosphate industry. Waste operators are rarely structured -- no equivalent of the water industry exists -- and farmers' organisations are not generally motivated on this subject.
The phosphate industry will have to address regulators to make progress on P recovery from animal wastes.
Leo van Ruiten's report, produced for CEEP (a draft
of which was presented at the
conference), concludes that for Holland, despite the large quantities of calf and pig
manures, it is improbable that P recovery from these sources will be economically viable in the near or medium term future.
Recovery of P from incinerated chicken litter is the most likely route for P recovery.
Chicken litter is produced in large, concentrated farms, is readily incinerated (high dry
matter and energy content), producing ash high in phosphate but difficult to use for
Potential of chicken litter for P recovery
Fibrowatt in the UK is currently incinerating chicken
litter to produce energy and a
The company pays farmers for their litter if it meets set moisture specifications -- ie
does not have too high a water content.
The resulting fertiliser, produced by combining the recovered fly ash and solids, is
sold to farmers at the same price as mineral fertilisers. Transported up to 60 70
km, this incineration product is currently only listed as fertiliser in the UK, not in the
A slow release product, the fertiliser has gained significant repeat business. As
additional plants come on stream, tow more are under construction, there may be a
need to find other uses.
A chicken litter incinerator costs around £70 million to build, including gas cleansing
and scrubbing equipment. Regulations require them to use at least 85% chicken litter
The volumes are currently too small for industrial phosphate producers. Fibrowatt
asked whether the phosphate industry would be prepared to pay higher prices for this
material, than for phosphate rock, as it contains lower levels of heavy metals.
(4) Calcium phosphate recovery
The calcium phosphate P recovery pathway has been developed to full-scale plant level. Operating, and laboratory, experience are available from a number of plants :
These plants produce calcium phosphate "pellets"
with a P content of 5 to 15%.
The calcium phosphates would appear to be deposited by amorphous precipitation
around the seed material, rather than true crystallisation. This is a complex chemical
proces and neither the chemical nor the physical parameters are fully understood.
The chemistry is made more complex because calcium phosphate is not one molecule
but potentially a number of different compounds - calcium hydroxyapatite, dicalcium
phosphate dihydrate, octacalcium phosphate, tricalcium phosphate with different
hydration complexes. The solubility and crystallisation properties of these different
molecules vary and the balance between them will modify the overall behaviour of a
Speciation of phosphates
Dr Alan House, UK Institute of Freshwater Ecology,
presented a paper outlining the
different factors controlling speciation of phosphates and their solubility.
The chemical processes occurring in calcium phosphate reactors are separated in space and time. The dynamics of precipitation proceed through nucleation, growth and termination steps throughout the system. Understanding the dynamics of this could lead to improved ways of handling the reaction system.
Current processes involve a pH of 9, or higher, in order to drive the precipitation. This
implies high chemical costs for lime or other bases. Carbon dioxide stripping may also
be necessary, and is installed in most of the pilot and full scale plants, in order to avoid calcium carbonate precipitation at these high pHs.
Calcium and phosphate ions are often above saturation
levels in the sewage effluent at operating pH, or at least in the
streams entering P recovery reactors : so why does precipitation
not occur without the driving force of high pH ? Answers to this
question could offer possibilities for reducing the amount of calcium,
or base, added to induce precipitation.
One factor may be the presence of precipitation inhibitors such as organic molecules,
fatty acids, polyelectrolytes, magnesium ions, trace metals. Eva Valsami Jones,
Natural History Museum, UK, underlined the variety of organic molecules present in
waste waters, from small organic acids to large humic and fulvic molecules. Some may enhance mineral solubility, increasing the super-saturation necessary for precipitation ; others may passivate mineral surfaces, assisting the preservation of formed phosphates. Identification of these different organic molecules could be achieved by concentration onto an absorbent, followed by stripping with an organic solvent. If organic molecules were identified as inhibitors, then an organic removal step could be envisaged, either by non-selective -- such as carbon treatment or peroxide treatment, or selective -- for those types of organic species that could be shown to inhibit precipitation.
There was also considerable discussion of the effects
of different seeding materials for calcium phosphate precipitation.
The DHV Crystallactor® plant at Geesmerambacht, Netherlands, uses sand as a seed
material. This reduces the P content of the produced pellets, but is acceptable in their current use in chicken feed. The produced pellets contain around 11% phosphorus. Sand is also used at the Essex and Suffolk Water pilot at Chelmsford, UK. Other experiments have used phosphate molecules (hydroxyapatite, DCP) as seed
material to obtain a higher P content in the produced pellets, important where
transportation of the recovered material cannot be avoided. The Warriewood demonstration plant, Australia, used magnesium oxide powder to raise
pH and facilitate phosphate formation, but this proved expensive.
Dr Dietrfied Donnert, Karlsruhe Research Centre, Germany, suggested that calcite
(CaCO3) may have advantages as a seed crystal through allowing calcium phosphate
formation to take place at lower pHs -- reducing the amounts of chemicals, lime or
magnesium oxide used for raising pH -- and reducing carbonate precipitation. This
avoids the need to strip CO2 from input water. It is not clear how these effects are
achieved, particularly once the calcite seed material has been coated with deposited
Full-scale phosphate precipitation units have been developed for industrial waste
waters, using calcite seed material in stirred reactors, at a starch factory and at the
Mercedes car factory at Gaggenau, Germany. The resulting amorphous precipitants
contained 6 to 10 % phosphorus.
Application of ultrasound was also suggested as a means of improving calcium
phosphate formation and lowering operating pH.
Further research areas regarding calcium phosphate recovery
The workshop suggested that research
is needed into :
- chemical reactor conditions, in particular how to minimise the high pH apparently necessary to drive precipitation and how to manage varying phosphate concentrations;
- physical reactor conditions, in particular kinetics (flow rates, turbulence), which affect precipitation and agglomeration and inter-particle friction (formation of pellets rather than fines);
- these conditions need to be researched in different waste water operating environments. Determine and report on what parts of the periodic table are present in the water under precipitation conditions at various points in the reactor. What solids are also present? Determine and report on the organic impurities in the water and their concentration;
- role of different "seed" materials (sand, calcite ). Increased purity levels of the calcium phosphate product may be achieved: for example the product as it is now could be better utilised by industry if the silicate and carbonate levels were lowered;
- interference of other ions and of sewage-derived organic molecules likely to be present in sewage;
- possibilities for improving calcium phosphate formation using ultrasound;
- solubility phases of different calcium phosphate compounds.
This research could best be achieved using laboratory equipment -- to obtain better
knowledge of the physical chemistry of the products and reactions involved followed by the use of existing full-scale reactors, or pilot plants, to field test the different hypotheses in waste water operating conditions.
The overall objective is to optimise the economics of the P recovery process in a sewage works context. This could be achieved by reducing the quantities of chemicals -- used for CO2 stripping and to raise the pH -- and by improving the efficiency of P recovery. Reduced capital costs for new calcium phosphate removal plants could result from a simplification of the process.
Comparison of the results between the different existing and pilot plants should highlight areas for improvement. It would also be useful to research the effectiveness of calcium phosphate recovery in removing heavy metals from the sewage stream as this would provide a value added service to the water industry. The phosphate industry is equipped to deal with such contaminants.
(5) Struvite formation as a route for P recovery
This workshop discussed the potential of struvite
(magnesium ammonium phosphate) and potassium struvite (potassium
ammonium phosphate) formation as a route for phosphate recovery
The workshop discussions were summarised by Professor Schuiling, Utrecht Institute
for Earth Sciences, wgo identified two main points :
Struvite as a removal mechanism for phosphate and
(some) ammonia :
It is preferable to remove phosphate from wastes in a beneficial way (e.g. bio-available) regardless of the need for recovery. Thus struvite production is preferable to Fe or Al phosphate precipitation.
The need to find or develop market applications for the recovered struvite :
There seems little possibility that struvite would be a useful raw material for the phosphate (non-fertiliser) industry, because it is incompatible with current industrial processes, but nevertheless struvite recovery and recycling offers significant environmental advantages. It is a recovery route which is definitely chemically and technically feasible and has already been demonstrated in full scale plants (Holland, Japan).
Full-scale struvite recovery processes are already operational or being built in Japan
and the Netherlands :
- the DHV Crystallactor® fluid bed process is used in a full scale struvite recovery installation at the AVEBE potato processing plant, the Netherlands (150m3/h) ;
- the Unitika Ltd (Osaka) struvite precipitation process is already in application at the Ube Industries Sakai plant (industrial waste waters) and is due to be commissioned in September 1998 at the Shimane Prefecture sewage works, Japan (45,000 m3/day) ;
- the Geochem Research/Delft University Earth Sciences stirred precipitation process produces potassium struvite from 700,000
tonnes/year of calf manure at Putten, the Netherlands (early 1998).
The Unitika Ltd. (Osaka, Japan) process, presented by Mr Hideo Katsuura, produces
easily handled granules with a phosphorus content of around 13.5%. The Putten calf
manure plant produces crystals with a phosphorus content of around 12,5%. Neither
product needs drying other than by gravity filtration.
Struvite and the phosphate industry
Doubts were expressed about the suitability of struvite as a pathway for recovery of P for recycling to the (non fertiliser) phosphate industry. There was doubt that the quantities produced in a sewage treatment plant, even a large operation like Slough, would justify putting in a process to recover P for industry. Local availability of Mg in waste water would also be a factor significantly affecting feasibility.
There was also a view that it did not make sense to recover struvite as a source of P
for phosphate industry, thus discarding the Mg and NH4 content, when the whole
molecule could be used to advantage as a fertiliser.
Struvite was considered to be a much less desirable material for recycling to the non
fertiliser phosphate industry than, for example, calcium phosphate. Dirk Van der Ploeg, Thermphos, Holland, indicated that the use of struvite for P recycling by the phosphate industry may necessitate significant changes in processes. For the thermal process route, magnesium would significantly modify slag properties and furnace technology would have to be adapted. Ammonia released during a traditional (oxidative) sintering process would result in increased production of NOx or N20, necessitating scrubbing and posing air pollution or greenhouse gas problems.
For the wet acid process route (sulphuric acid attack), magnesium would exclude the
use of struvite as it does not crystalise out in the way that calcium gypsum does.
Potassium struvite would pose similar problems due to behaviour of potassium
Struvite and the waste water industry
Struvite deposit formation is a natural phenomenon
that is a widespread problem in
waste water treatment operations.
Nic Booker, CSIRO, Australia, termed it "an accident waiting to happen". Deposits
form in pipes where waste waters stand between intermittent flows, but also in areas
of turbulent flow or pressure change. They thus impede pipes, valves and pumps.
There was support for the view that the emphasis should be put on the further
development of struvite technology as a means of solving problems at waste water
processing plants rather than expecting payback through phosphates in recovered
Slough sewage works
Steve Williams, Thames Water, UK, presented the
case of Slough sewage works where struvite deposits have been causing
considerable operating problems by blocking pumps and pipes.
One pipe several hundred yards long requires periodic overnight soaking with sulphuric
acid to clear struvite and prevent blocking. Magnetic devices, similar to those used to reduce carbonate precipitation in hard water, have given variable results.
Recovery of struvite from the waste water upstream of deposition problem areas would avoid these problems and thus give significant operating savings.
Contribution to N removal
Struvite recovery also contributes to nitrogen removal
from the waste water stream,
reducing or avoiding investments necessary to respect soluble nitrogen outflow
regulations. In some regions, animal waste provides a higher source of concentrated P-containing waste streams than sewage.
In the UK, chicken waste is a significant potential resource. In the Netherlands the
largest resources are in pig waste which is generally thought unsuitable for struvite
recovery. Highly viscous, the P is tightly bound and the struvite generated does not
separate well. A full scale struvite recovery plant is now functioning in Holland using
Agricultural value of struvite
It was recognised that struvite is an excellent material for removing P from waste water or sludge, but that its uses are less well known.
Struvite was generally regarded favourably as a fertiliser, although there was
considerable lack of knowledge about how it performs and its value.
Although struvite is relatively insoluble in water, making it valuable as a slow release
fertiliser with limited risks of leaching, it is citrate soluble and hence a good nutrient
source for plants. Struvite as a farm fertiliser, has the advantage over sewage sludge
that it contains no heavy metals and avoids possible overdosing of nutrients.
Little was known about the market for struvite. Farmers in the USA are interested in
using it but they don't want to pay. Struvite recovered from sewage is sold for ¥20 per bag in Japan, which is sufficient to recover the costs of production. Compared with this, the stated price in Australia of A$1700 per tonne seems unrealistic. Recovered struvite could find niche markets. There could be opportunities to make it available to farmers at the local level through direct local processing and distribution by co-operatives or farm suppliers.
Areas of struvite recovery research
The workshop identified the following research areas :
- the need to demonstrate the agricultural value of struvite and study its dissolution behaviour (nutrient release) in soils and accessibility for crops. Existing work in this area needs collating and assessing ;
- the conditions for nucleation and growth under "real" conditions are still incompletely understood. ;
- the ideal conditions for struvite recovery were stated to be high P and NH4 concentrations and low suspended solids. Kinetics was thought to be the most limiting factor. Research work so far has been mainly carried out using fairly "clean" solutions containing low levels of organic contaminants. Further work is needed on struvite formation in waste waters and in biological nutrient removal side streams and sludges. There is a lack of information on nucleation characteristics and how struvite forms ;
- process control and management will also be important to optimise struvite formation and maximise phosphate removal ; interest in further research into the formation conditions for different forms of struvite : magnesium ammonium phosphate, potassium ammonium phosphate, and possible other substitutions
(Mn2+ for Mg2+ etc) ; experimentation and development of pilot struvite recovery installations in order to test different operating conditions ;
- precipitation may not be the best way to proceed :struvite is more easily crystallised on a fluid bed than calcium phosphate.
(6) Economic feasibility of P recovery
The workshop tried to identify the different factors
that will influence the economic
feasibility of P recovery from municipal waste water.
Cost estimate exercise
A "typical" case of a medium sized sewage works (500,000 pe) with an average
input of 3gP/pe/day and subject to obligatory P removal was examined in order to
compare the overall economics of P recovery with removal by chemical precipitation.
Sludge disposal :
- the sludge generated by iron precipitation would be 7.25 tonnes/day of iron phosphate (2.7 tonnes/day of iron as FeCl), on the basis of a 1:1 molar relationship, increasing sludge mass by 22%. However, water industry participants suggested that iron has to be used in up to double stoichiometric concentrations for effective P removal. John Upton, Severn Trent Water, UK, estimated that P removal using iron precipitation increases sludge volumes in practice by 40% ;
- sludge disposal costs were estimated at $85/t for agricultural spreading, $360 for landfill and $600 for incineration ;
- Glen Daigger, CH2M Hill, USA, estimated that the extra sludge generated would require capital investment in expanded sludge handling facilities of up to $200/t ;
- overall, increased sludge disposal costs were estimated at $1500 to $3000 / day.
Chemical costs :
- the cost of FeCl used for P precipitation was estimated at $1600/day on the basis of stoichiometric 1:1 quantities of iron. This cost may well be doubled in practice as higher concentrations of iron are used.
P recovery using lime-induced fluid-bed crystallisation :
- such a process would involve chemical costs for lime used (estimated at $640/day) and for energy consumption in the fluid bed reactor (estimated at $130/day) ;
- capital investment is estimated at around $30 million ;
- staff time for operating the fluid bed reactor is not taken into account as difficult to estimate but may not be significantly higher than for P removal by precipitation.
When additional sludge handling investment is added in, the overall payback
period is <10 years which, in the US, is generally an acceptable criteria.
The main point to remember here is that this type of cash saving is only applicable
when P stripping is already obligatory to meet legislative requirements. The
incremental saving of about $2200/day compared to P stripping by precipitation
(chemical and sludge cost savings), is what generates the 10 year payback.
Calculation of recovered phosphate value
The quantities of phosphate pellets generated (about 6T/day or 2000T/yr as pellets)
would not be enough to independently support even the simplest industrial phosphate
Feasibility therefore depends on grouping recovered phosphates from a number of sources (several waste water treatment plants, industrial processes, animal wastes) in order to bring together significant quantities without excessive transport and logistics costs.
The current price of phosphate rock at a UK port is approximately US$ 40 50 /
tonne (P content around 20%). This would be a reasonable minimum selling price for
the P recovered from sewage as industry processable pellets. However, a higher
purity of recovered calcium phosphate product might allow a reduction in processing
costs at both the rock digestion and solvent purification stages.
The organic loading of the recovered product may, however, require low temperature
calcination for removal the associated capital, energy and operating costs need to
be added into the overall project economics. One option would be to look at on-site acidulation of this product at the SWT to make triple super-phosphate (TSP) : this could involve participation of the fertiliser industry.
Suggestions for further economic investigation
Currently available figures are too approximate to give more than a rough indication
of viable economics. The following areas need further research into costs and
- economics of recovery in relation to the size and topology of sewage works: mainstream/ sidestream P recovery, P precipitation versus bio-P ;
- capital costs and running costs for P recovery installations. Utilisation of Geestmerambacht and other full scale plant experience : DHV can provide useful insights into design/capital costing. The costs at existing plants, however, are very high because of the experimental or demonstration design of these. The possibilities of cost savings through operating experience and multiplication of plants needs to be examined ;
- potential for developing the triple super-phosphate (TSP) option with the fertiliser industry and of other solutions for local processing or use of recovered phosphates ;
- P content and other qualities of final product, which will affect its market value ;
- need to consider the implications of low P (but high N) sludge on disposal costs.
Where sewage works currently include denitrification,
or where such capacity is to be
installed, bio-P removal or recovery adaptation costs are minimised.
Steve Williams, Thames Water, UK, underlined the energy cost of sludge aeration for N removal. The combination of P recovery and N removal may enable some economies at this level. Recovery of P as struvite implies removal of ammonium and can thus enable economies in N removal investments and running costs.
Heavy metal removal
John Upton, Severn Trent Water, UK, suggested that
the removal of heavy metals from the waste water stream by fixing
them in the P recovery product, would offer a
significant cost payback to sewage operators.
There is increasing pressure to remove heavy metals from outflows (eg. revised
dangerous substances legislation in the UK). Heavy metals in outflows are likely to be
in a soluble form, and may possibly be fixed into calcium phosphate in a P recovery
reactor. Heavy metal removal is increasingly a significant cost for waste water
Conclusions of economics workshop
Given a choice, the water industry will generally adopt the cheapest available option
which can consistently meet regulatory requirements. Phosphate recovery will only
therefore be economically viable where there is a mandatory requirement for P removal. The water industry will only adopt P recovery if it offers a payback of at most ten years compared to P stripping by precipitation or biological methods.
Within the principle of sustainable development should legislation be considered to
promote P recovery in order to get the phosphate and water industries to work
If recovery from sewage and wastes become a major P source, this would necessitate a restructuring of the P industry. There is also a need to involve the fertiliser industry as a potential major outlet for recovered P products from sewage and animal wastes.
The next step will be to put together a team involving DHV and other processengineers (design and operation of P recovery installations), the water industry (Anglian Water, Severn Trent and other interested companies from the UK and Europe), CH2M Hill (economics expertise and computer model), the phosphate industry and if possible the fertiliser and animal feed industries. This team will rework and refine the figures produced at the workshop and develop a proposal for further research into costs and economics.