Iwa international Specialist Conference

Results

The total phosphate stream available from Dutch sewage treatment plants each year is equal to approximately 12.5% of the annual phosphate requirement of Thermphos International B.V. Measured against the quality criteria, the two existing streams (A and B above) present obstacles which may prevent reuse. The water content is much too high (stream A, some processes stream B), the P₂O₅ content is too low (stream A), the iron content is too high (streams A and B) and the copper and zinc content are too high (streams A and B).

Processes that precipitate phosphate as relatively pure product from liquid streams (stream C) within the sewage treatment plant, can produce materials that are eminently suitable for reuse in the phosphorus production process. Practical experience with the Crystallactor® at Geestmerambacht sewage works in The Netherlands shows that a suitable recovered material (calcium phosphate) can be obtained which poses none of the mentioned problems, and indeed Thermphos International B.V. is currently using the total output from this reactor.

Phosphate precipitation processes are most likely to be installed in sewage works operating biological phosphorus removal, as these offer side streams with high phosphate concentrations which are suitable for operation of such reactors, and because P-recovery can improve their operation. An inventory was made of plants in Holland already operating a biological phosphorus removal stage, and plants susceptible to be adapted for biological phosphorus removal. On this basis, it was estimated that phosphate precipitation reactors could realistically generate a stream of recovered phosphates appropriate for reuse by Thermphos International B.V. that would cover approximately 5% of the phosphate requirement.

Future work
Although the Crystallactor® is a proven concept for phosphorus recovery at sewage treatment plants with biological phosphorus removal, the economic feasibility of this system is poor. Therefore it is necessary to study the technical and economic feasibility of developing and installing alternative phosphate precipitation reactors (for Thermphos only calcium or aluminium phosphates) in appropriate side streams of bio-P sewage treatment plants. One issue in this study is if every bio-P sewage treatment plant should be equipped with a P-recovery facility or that some centralisation can be introduced, e.g. on the central location for the final sludge treatment. (Preliminary calculations will be given in the full paper).
This Dutch analysis was based on the quality criteria for the phosphorus industry of Thermphos International B.V. Since the Second Conference on phosphorus recovery also the Dutch fertilizer industry became interested in the recovery of phosphates. Because the fertilizer industry has less strict requirements concerning the quality of the recovered phosphate, more options for P-recovery come available. (These possibilities will be discussed in the full paper).

69.

Recovered Iron Phosphates, Quality and Use

Jesper Holmquist, M.Sc
Kemira Kemi AB

P.O. Box 902

SE-251 09 Helsingborg

Sweden

1. Introduction

When a resource is being regarded as a limited resource, it normally means that it is limited as to its existing amount or to the amount that is considered to be economically feasible to win. There is however also a third aspect to this expression, which is applicable in the case of phosphorus: biological life as we know it can not exist without the supply of pure phosphorus.

The best way of preventing further addition of toxins into the food chain ought to be to cut off the connection between the two, that is, to stop using the mines in which the most polluted phosphate ores are found. Instead, we can purify the phosphorus already in circulation by treating it in processes like KREPRO, where phosphorus in sewage is separated from pollutants in a controlled way, allowing us to purify the P cycle.

2. Recovery

KREPRO is a process for treating sewage sludge, preferrably digested and precipitated with iron containing coagulants. The treatment comprises a thermal hydrolysis of the sludge at low pH. Then an organic fibre fraction can be separated, reaching a dry solid content of between 45 and 55 percent depending on separation equipment and the conditions during hydrolysis and dewatering. The fluid from the dewatering of the fibre fraction contains two valent iron, orto-phosphate, dissolved COD and ammonium, but only very low concentrations of heavy metals and organic toxins. At a low pH, ferric phosphate can be selectively precipitated from the fluid, without co-precipitation of unwanted compounds such as heavy metals. An oxidising agent is added, which oxidises ferrous iron into ferric, making ferric phosphate precipitate within seconds. The small particle precipitate is separated with centrifuges and obtained as a paste holding 35 percent dry solids. After drying, the end product contains 10 percent phosphorus.

3. Phosphate quality

Based on element analyses, ion balances and equilibrium curves, a chemical formulation of the dried product can be obtained. It reflects an average formulation of the produced test materials, which all have been analysed thoroughly. The formula is
Fe(SO₄)_0,17(PO₄)_0,72(OH)_0,5  1,43 H₂O + 0,02 MgO + 0,11 CaSO₄
In this formla, inorganic carbon compounds have been left out.

The contents of heavy metals in the phosphate product has been closely monitored during the development of KREPRO, due to strict regulations for sperading on farmland. The concentrations are on the same level as the best commercial fertilisers on the market.

Also concentrations of organic compounds have been analysed. It shows that pollutants are eliminated in the process or separated together with the organic fibre fraction before reaching the ferric phosphate precipitation.

The sludge entering the thermal hydrolysis in KREPRO often contains high concentrations of pathogens. Because of that, sludge spreading on farm land is a debated issue. The hydrolysis eliminates all detectable pathogens. Furthermore, the downstream KREPRO process offers no chance of re-infection of the process fluids.

4. Plant availability

A number of growth tests has been performed with many different types of plants in soils from all over Europe. The results from these tests show that ferric phosphate does give a fertilising effect, and that the effect is about 50 percent of super phosphate in the first harvest after fertilising. The long-term results however suggested that the effect of ferric phosphate increases with time. After some weeks, certain plants fertilised with ferric phosphate grew even faster than those treated with super phosphate.

5. Application

The possibilities for developing a market for ferric phosphate has been investigated. Three main fields of application have been found.
5.1. Forest fertilising

The ferric ion binds phosphate in soils that are low in pH and mineral concentration, for instance peat soils in Finland. These soils are subject to substantial P leaching if treated with conventional fertiliser, but with ferric phosphate long term fertilising is possible as well as no leaching of P to the surrounding waterways.

5.2. Agriculture

Numerous tests on agriculture crops have shown that the ferric phosphate maintains an adequate level of Olsen-P in the soil for most crops. On certain soils in Scandinavia with a low pH, the effect is enhanced when the P treatment is combined with lime spreading.

5.3. Industry

Ferric phosphate can replace other phosphate sources in some industrial processes. Due to patent issues, the processes can not be revealed at this moment. In at least one case, the ferric phosphate worked better than the phosphate chemical used today.