Phosphate Recovery

Home | Introduction | The phosphates file | Current research projects | Bibliography | Contact us

BACKGROUND:

Nutrient removal from wastewater discharges is an increasing challenge for water authorities, as regulatory authorities tighten discharge standards to avoid eutrophication problems in receiving waters. Significant costs are associated with the extra treatment processes required to meet these new discharge standards. The most widely used technologies for nutrient removal include biological nitrification/denitrification for nitrogen removal and metal salt precipitation for phosphorus removal. Both approaches result in the nutrient being made unrecoverable as a fertiliser. An alternative to these conventional technologies which can provide for recovery of the nutrient as a commercial fertiliser could be the production of struvite.

Nitrogen and phosphorus are normally present in domestic sewage at concentrations of around 40 and 10 mg/l respectively. While very dilute, the enormous flows of sewage mean that significant quantities of N and P enter the natural environment each year via sewage flows. For example, the Melbourne Western treatment plant at Werribee processes about 600,000 m³/day of effluent, with a total phosphate flow of around 18 tonnes per day or nearly 6,600 tonnes per annum. At Australian market prices for super phosphate fertiliser, this phosphate discharge is theoretically worth ~ A$4.9 million per annum! The fundamental problem with realising any value for this phosphorus is the difficulty of concentrating it out of a dilute solution into a sufficiently concentrated form where it can be marketed as a useful product. Assuming that some process can be devised for recovering and concentrating the nutrients in sewage into a useable form, what is the potential value of these nutrients and to what extent can they offset the cost of wastewater treatment? Assuming a value for anhydrous ammonia of A$300 per tonne (82% N) and A$500 per tonne for wet process phosphoric acid (75% PO₄) and concentrations of NH3 and P in sewage of 40 and 10 mg/l respectively, then the value of the nutrients in a kilolitre of sewage is 3.5 cents. With wastewater treatment operating costs being around 10 c/kl, recovery of the nutrients as a fertiliser is obviously not a direct economic proposition. However, it could be a good way to offset some of the unavoidable costs associated with treating the wastewater before discharge and final disposal of the sludge.

CURRENT NUTRIENT REMOVAL TECHNOLOGIES

Current approaches to nutrient removal from sewage and other wastewaters are summarised in the following table.

Biological phosphorus removal

Biological processes for the removal of phosphorus have been developed during the last twenty years and are now beginning to compete with the more conventional physico-chemical approach of precipitation with metal salts. Biological phosphorus removal (BPR) plants commonly involve a number of different stages of treatment, including anaerobic, aerobic and anoxic zones and require liquid residence times of around 24 hours. They are thus large plants with a significant capital and operating cost. On average and with good operational control, BPR plants can reduce phosphorus concentrations to less than 1 mg/l, although some plants with insufficient readily assimilable carbon in the feedwater have difficulty in consistently achieving this level.

The phosphorus in the raw sewage accumulates in the sludge as polyphosphate crystals, with P concentrations in the sludge being around 5g/l. If this sludge goes anaerobic at any stage, the phosphorus is released back into solution as orthophosphate and consequently most BPR plants avoid anaerobic digestion of the sludge before disposal. The undigested sludge is unstable and odorous but can be stabilised by aerobic digestion without incurring phosphorus release, although it is always possible for the sludge to go anaerobic at a later stage during storage eg in a land-fill. Sludges from extended aeration plants should be at least partially stabilised. Another approach is to treat the sludge with lime to bind up the phosphorus as the insoluble calcium phosphate.

At those plants where the sludge is digested anaerobically, both ammonia and soluble orthophosphate appear in solution at concentrations between 500 and 1,000 mg/l. This solution is usually separated from the sludge during further processing, such as belt pressing, and must then be dealt with separately. In most cases, this involves lime addition to tie up the phosphorus and recycle of the ammonia rich liquors back to the inlet of the plant where the ammonia must be nitrified. However, the ammonia and phosphate in the sludge liquors could be used in the production of a solid fertiliser such as struvite.
 

Physico-chemical phosphorus removal

Phosphorus can also be removed by precipitation with metal salts such as aluminium, iron and calcium, with final P concentrations below 1 mg/l being achievable with this approach. To achieve lower phosphate concentrations, increasingly high doses of metal salts are required, often in excess of two times stoichiometry. Because of its relative simplicity and reliability, metal salt precipitation has been used more widely than BPR to remove phosphorus. However, the high chemical cost and significantly increased sludge volumes associated with metal salt precipitation are definite disadvantages and have led to the increasing use of BPR. Another disadvantage is that metal salt precipitation ties up the available phosphorus and thus makes it unavailable as a nutrient.

PHOSPHATE RECOVERY AS STRUVITE

In domestic sewage the molar ratio of nitrogen to phosphorus is around 8 to 1 with the phosphorus normally being present as the soluble orthophosphate. This molar imbalance and the relatively low concentration of phosphorus in sewage make the recovery of phosphorus in a useable form a significant challenge. A better source of phosphate for recovery as a fertiliser is the phosphate released into solution when the sludge from a BPR plant undergoes anaerobic digestion. In this case, phosphorus concentrations of around 500 mg/l can be reached in the supernatant solution.

As the orthophosphate ion is not volatile (unlike ammonia) and is of similar molecular size to other ions in sewage, the only viable approach to its recovery in a concentrated form is precipitation as an insoluble salt. This approach is, of course, the basis of the most widely used method for phosphate removal from wastewaters, namely precipitation as the metal salt eg FePO₄. However, such salts completely tie up the phosphate thus making it unavailable as a nutrient. The range of salts called metal ammonium phosphates have the propitious properties of being not only insoluble enough to achieve significant phosphorus removal but also of being able to make the phosphorus and ammonia available to plants by a biologically based slow release mechanism (Bridger et. al., 1962). The most well known example of this type of salt is "struvite" or MgNH₄PO₄, which is commonly formed in anaerobic digesters when significant levels of Mg occur in the raw sewage. The excellent fertilising properties of struvite have been well studied and reported in a number of publications (Lunt et. al., 1964). Salutsky et. al. (1972) have clearly demonstrated the efficient precipitation of phosphorus from anaerobic digester effluents as MgNH₄PO₄, with P levels being reduced from around 100 mg/l to 2-3 mg/l. Simultaneous ammonia removal was also achieved and elemental analysis found the precipitate to be largely struvite.

The recovery of both ammonia and phosphate from sewage and the subsequent production of struvite has been well studied by Liberti et al (1981). In this study, both ammonia and phosphate anions were recovered from secondary sewage effluent via ion exchange and the eluants from resin regeneration used to produce both magnesium ammonium phosphate (MgNH₄PO₄) and ammonium nitrate (NH₄NO₃). While an economic analysis of this nutrient recovery process appeared favourable, it was not adopted commercially, the main reasons appearing to be the high cost of chemical inputs (MgCL₂, NaOH and brine solutions) and the failure to establish a firm market for the fertiliser produced.
 

Uses for struvite

Studies by Bridger et. al. (1962) have confirmed the excellent agronomic properties of MgNH₄PO₄. While only slightly soluble in water and soil solutions, struvite was found to be a highly effective source of phosphorus, nitrogen and magnesium for plants through both foliar and soil application. The release of nutrients appeared to be enhanced by a biological nitrification mechanism, with the nutrients being released at a controlled rate over an extended period of time. When properly granulated, it can be applied to soil at rates greatly exceeding those of conventional fertilisers without danger of burning plant roots.
 

Economics

Despite such attractive agronomic properties struvite is not widely used in the fertiliser industry, the main reason appearing to be its high cost of production from the raw chemicals. As described above, Liberti et. al. (1991) had to use both MgCl₂ and NaOH in addition to ammonia and a phosphate source. Other studies (Schulze-Rettmer, 1991) have also highlighted the significant chemical cost associated with this approach.

In a wastewater treatment plant which is required to remove both nitrogen and phosphorus, the cost of supplying N and P is just the incremental cost associated with changing the treatment plant design as discussed in previous sections. The supernatant from the anaerobic digester of a BPR plant is an excellent source of both N and P for struvite production, but this still leaves the cost of both MgCl₂ and NaOH supply. However, a good understanding of the process chemistry, combined with clever design, has the potential to produce a cost effective process for struvite production.
 
 

References:

Bridger, G L, Salutsky, M L, Starostka, R W (1962) "Metal Ammonium Phosphates as Fertilisers", Agricultural and Food Chemistry, 10, 3, 181-188

Liberti, L, Gianfranco, B, Petruzzelli, D, Passino, R (1981) "Nutrient Removal and Recovery from Wastewater by Ion Exchange", Water Research, 15, 337-342

Lunt, O R, Kofranek, A M and Clark, S B (1964) "Availability of Minerals from Magnesium Ammonium Phosphate", Agricultural and Food Chemistry, 12, 6, 497-504

Schulze-Rettmer, R (1991) "The Simultaneous Chemical Precipitation of Ammonium and Phosphate in the Form of Magnesium Ammonium Phosphate", Water Science & Technology, 23, 659-667