The paper compares the anaerobic digestion of biological and chemical phosphorus enriched sludges. The research has used sequential extraction and internal standards to determine the speciation of P and metals in both laboratory and full-scale digestion. Bioavailable iron and P were found to be very low in CPR digested sludge, compared to the controls. Some of the biologically removed P was remobilised during digestion but most was precipitated as calcium and calcium magnesium complexes, including struvite. There was no evidence of deficiencies of important metal cofactors or increased solubility of toxic metals.
Keywords: Anaerobic digestion, speciating trace metals, struvite, biological and chemical phosphorus removal.
Phosphorus removal during wastewater treatment will change the nature of the sludges fed to anaerobic digesters. Previous research on chemical phosphorus removal (CPR) has shown adverse effects from the inorganic enrichment . The phosphate from CPR is retained within the digester and reduces the biodegradability of volatile solids . The digestibility of biological phosphorus removal (BPR) sludges are unaffected but between 20 and 50% of the phosphorus is resolubilised during anaerobic digestion  and this may generate inorganic deposits within the digester or scale in downstream equipment. The phosphate enriched precipates may also be suitable for recovery. There may be other problems caused by the mobilisation of potentially toxic metals or the precipitation of important cofactor trace metals. This research reports on changes in phosphorus and metal speciation that occur in digesters treating BPR and CPR sludges and its effects on digestion.
Sequential chemical extraction methods based on those of Stover et al  and Uhlmann et al  were developed to compare metal and phosphorus fractionation of anaerobically digested BPR and CPR sludges. The extraction steps are shown in Tables 1 and 2.
The metal and phosphate fractions were analysed by plasma emission spectrophotometry (Thermo Jarrell Atomscan 16). Colorimetric analysis (ascorbic acid method) was necessary to analyse soluble reactive phosphate . Attempts to use ion chromatography to measure soluble reactive phosphate were unsuccessful due to the strong extractant solutions. A number of standard analar compounds were used as internal standards to validate and adjust the methods. (Table 3). Laboratory scale research was used to determine the effects of phosphorus sludges on digestion under well controlled conditions and samples from full-scale digesters used to investigate the influence of complex conditions on metal and phosphorus speciation. Sewage Treatment Works (STW) A was a control digester with no phosphorus removal, STW B used BPR and STW C CPR (pre-precipitation with ferric sulphate) (table 4). A more extensive methodology and results are in the thesis of the work (7).
RESULTS AND DISCUSSION
Relative total average P concentrations were 31g kg-1 (of dried sludge) in the CPR digester, 26g kg-1 in the BPR digester and 16g kg -1 in the control digester. The average iron content was 72g kg-1 in the CPR digester, 13g kg-1 in the BPR digester and 12g kg-1 in the control digester (fig 1). The soluble P in the CPR digester was less than 0.1mg L-1 compared to 13mg L-1 prior to iron dosing. Soluble P in the control digester was 28mg L-1 and 107mg L-1 in the BPR digester (fig 2). Analysis of the speciation of P in the CPR digester showed that it remained bound to iron (represented by the NaOH fraction in fig 3 and 4) but in the reduced ferrous form (represented by the EDTA fraction in fig 5 and 6. This confirmed laboratory-scale results, in which ferric phosphate was found to be briefly solubilised under anaerobic conditions and re-precipitated as ferrous phosphate. Iron is normally dosed at a 2:1 molar ratio with phosphorus  therefore sufficient iron is present for re-precipitation of all solubilised phosphorus as Fe3(PO4)2, but if iron is dosed at lower concentrations than 1.5:1, soluble phosphate could be released during digestion . Ferrous phosphate is not bioavailable , therefore P limitation in a CPR digester could reduce digester performance [7, 10].
Loss of digester performance may also be caused by excesses or deficiencies of cofactor metals. Copper and chromium for example are reported as causing toxicity problems [11, 12, 13]. Metal bioavailability and toxicity will decrease as the metal becomes more difficult to extract. Metal fractionation results showed that copper in both the BPR and CPR digesters shifted from the residual to the HNO3 fraction (fig 7 and 8). The internal standards suggest that the residual fraction is CuS and CuFeS2. This may, therefore, represent a decrease in the formation of chalcopyrite (CuFeS2) because the P enrichment leaves less soluble iron. Copper speciation was most markedly changed in the CPR sludge digester, the residual fraction decreasing from 26 % pre-CPR to 6 % post-CPR and the Na4P2O7 fraction correspondingly increasing from 12 % pre-CPR to 25 % post-CPR, indicating increased copper bioavailability in this digester. The trend towards increased solubility of metals in the phosphorus-rich digesters was more pronounced with chromium (figs 9 and 10). In the control digester, 58% was in the residual and HNO3 fractions whereas in the P enriched digesters only 35 % (BPR) and 26 % (CPR) of the chromium was represented by these fractions.
Iron, nickel and cobalt are methanogenic cofactors . The concentrations of iron in the control and BPR digester were similar (fig 5 and 6) but the control digester was the only digester with easily available iron (iron in the solubles KNO3 and KF extracts or adsorbed iron), the CPR digester having the least available iron. The results also show other changes. In the control digester iron was also present in the HNO3 and residual fractions, whereas the EDTA fraction (representing iron phosphate) was increasingly important in the BPR and CPR digesters.
The BPR and control digesters had very similar nickel profiles (figs 11 and 12). The CPR sludge contained a higher proportion of EDTA-extractable (52 %) and adsorbed (11 %) nickel than either the control and BPR digesters, however, this was not due to CPR as a similar fractionation profile for Digester C was recorded prior to CPR. The total concentration of nickel in the CPR digester, like the chrome, was much higher than in either the control or BPR digesters. The CPR plant does include metal processing industrial effluent.
The cobalt fractionation results were different for each digester (figs 13 and 14). All the digesters contained similar concentrations of soluble cobalt but there was less cobalt in the other labile fractions of the BPR and CPR digesters. The control digester contained the highest proportion of cobalt in the soluble, KNO3 and KF extracts, i.e. 25 % in the control 18 % in the BPR digester and 10 % in the CPR digester.
Anaerobic treatment of BPR sludge is known to release soluble phosphate, magnesium and potassium according to the molar ratios of their uptake during BPR . Phosphate and magnesium are rapidly re-precipitated but the potassium remains in solution. This is illustrated in our research (figure 15) the BPR digester contains soluble magnesium at 12 mg L-1 and potassium at 218mg L-1 compared to 35 mg L-1 soluble Mg and 57 mg L-1 soluble K in the control and 45mg -1 L soluble Mg and 60mg L-1 K in the CPR sludge digester. Using the soluble potassium concentrations in a BPR digester, it is possible to predict the total concentrations of magnesium and phosphate released from the BPR sludge . This is shown in fig 15, which indicates that the theoretical concentrations of phosphate and magnesium released into the BPR digester were 590mg L-1 and 120mg L-1, respectively. Thus, only 10 % of the released soluble magnesium and 20 % of the soluble phosphorus remained in solution indicating precipitation of both magnesium and phosphorus in this digester. In spite of this, the residual soluble P concentration in the BPR digester was still ten times greater than that of the control and 100 times greater than the CPR digester.
Previous work has indicated that struvite  could be a problematic sink for some of this released phosphorus. Analysis of the speciation results described here indicates that the largest P fraction was calcium phosphate in both the BPR and control plant (extracted in the HCl fraction) (figures 3 and 4). In the CPR digester, on the other hand, most of the P is in the NaOH extract as soluble reactive phosphate (ferrous phosphate). This was confirmed by the internal standards, the laboratory work and an analysis of the iron fractionation. Model compound testing showed that magnesium from struvite was recovered in the pyrophosphate (Na4P207 fraction), and the BPR digester does show the highest percentage of magnesium in this fraction (20 %) (fig 16 and 17). These results are supported by Jardin and Pőpel  who calculated that 20% of the P released from BPR sludge formed struvite and Wild et al  who also noted that 15% of the re-solubilised P formed struvite and 33% calcium phosphate. The relatively high proportions of magnesium extracted in the EDTA fractions of both the BPR and CPR digesters (29 and 40 % in comparison to 15 % in the control digester) was attributed to precipitation of magnesium-calcium phosphate in the BPR digester.
Laboratory experiments indicated that soluble P concentrations needed to be above 500mg L-1 or the pH to be more alkaline than the neutral for struvite to become more important. Borgerding  found similar results but also noted struvite precipitated on cooling down from 35ºC. The struvite formed in the BPR digester is twice that in the control digester but it is suggested below these critical thresholds (fig 16 and 17).
There was little struvite formation in the CPR reactor (fig 16 and 17) which helps confirm Mamais et al  suggestions that iron dosing can prevent struvite formation since ferrous phosphate is formed preferentially to struvite. Water hardness may also be important since calcium phosphate complexes are also precipitated more readily.
CONCLUSION AND RECOMMENDATIONS
Phosphorus in CPR digesters is precipitated as ferrous phosphate whereas in BPR digesters calcium is the major sink.
CPR causes inorganic enrichment of anaerobic biomass and may lead to decreased digester performance. There is less available iron and phosphate and this needs further research to confirm its importance compared to encapsulation of the volatile solids.
There were only minor changes in solubility of the potentially toxic metals chrome and copper or the important nutrient metals nickel and cobalt.
Struvite was shown to be one of the by-products of BPR but iron and calcium are antagonistic. Concentrations of soluble P above 500 mg L-1 and a pH greater than 7.8 were necessary to increase the proportion of struvite.
Iron has a high affinity for phosphate in the anaerobic digester and small doses in combination with BPR could be used to avoid any problems with struvite or recycling of soluble P. This needs more research.
An Internationally agreed standard methodology for P and metal speciation is suggested to allow comparison of results.
The work was supported by EPSRC grant GR/K96946. The authors would also like to thank Severn Trent Water for additional financial support, help, information and samples from working digesters.
Yeoman S., Lester J.N. and Perry R., The effects of chemical phosphorus precipitation on anaerobic digestion. Environmental Technology., 11, 709-720 (1990).
Kindzierski W.B. and Hrudey S.E., Effects of phosphorus removal chemicals upon methane production during anaerobic sludge digestion. Canadian Journal of Civil Engineering.,13, 33-38 (1986).
Pitman A.R., Management of biological nutrient removal plant sludges – change the paradigms? Water Research.,33, 1141-1146 (1999).
Stover R.C., Sommers L..E. and Silveira D.J., Evaluation of metals in wastewater sludge. Journal of the Water Pollution Control Federation.,48, 2165-2175 (1976).
Uhlmann D., Roske I., Hupfer M. and Ohms G., A simple method to distinguish between polyphosphate and other phosphate fractions of activated sludge. Water Research.,24, 1355-1360 (1990).
Eaton A.D., Clesceri L.S. and Greenberg A.E., (Ed) Standard methods for the examination of water and wastewater (19th edition). APHA-AWWA-WEF (1995).
Carliell-Marquet C.M., The effect of phosphorus enrichment on fractionation of metals and phosphorus in anaerobically digested sludge. PhD thesis Loughborough (2001).
Upton J., Nutrient removal in the UK – Now and in the future. Paper presented at the International Conference on Phosphorus Recovery (CEEP), Warwick University, UK. May 6-7 (1998).
Frossard E. and Morel J.L., Phosphorus species and availability in FeSO4 treated sludges. In: Proceedings of the 2nd ESA Congress, Warwick University, UK, August, 408-409 (1992).
Alphenaar P.A, Sleyster R., de Reuver P., Ligthart G.J. and Lettinga G., Phosphorus requirement in high-rate anaerobic wastewater treatment. Water Research.,27, 749-756 (1993).
Mosey F.E., Swanwick J.D. and Hughes D.A., Factors affecting the availability of heavy metals to inhibit anaerobic digestion. Water Pollution Control.,70, 668-679 (1971).
Mehrota A., Mehrota I. and Tandon S.N., Role of Cu (II) and Zn (II) species on the anaerobic digestion of synthetic activated sludge. Journal of Chemical Technologyand Biotechnology.,56, 185-190 (1993).
Barber W.P. and Stuckey D.C., Metal bioavailability and trivalent chromium removal in ABR. Journal of Environmental Engineering.,126, 649-656 (2000).
Speece R.E., A survey of municipal anaerobic sludge digesters and diagnostic activity assays. Water Research., 22, 365-372 (1988).
Jardin N. and Pöpel H.J., Phosphate release of sludges from enhanced biological P-removal during digestion. Water Science and Technology.,30, 281-292 (1994).
Sen D. and Randall C.W., Factors controlling the recycle of phosphorus from anaerobic digesters sequencing biological phosphorus removal systems. Hazardous and Industrial Wastes.,20, 286-298 (1988).
Wild D., Kisliakova A. and Siegrist H., Prediction of recycle phosphorus loads from anaerobic digestion. Water Research.,31, 2300-2308 (1997).
Borgerding J., Phosphate deposits in digestion systems. Journal of the Water Pollution Control Federation.,44, 813-819 (1972).
Mamais D., Pitt P.A., Cheng Y.W., Loiacono J. and Jenkins D., Determination of ferric chloride dose to control struvite precipitation in anaerobic sludge digesters. Water Environment Research.,66, 912-918 (1994).
Table 1: Successive stages of the metals extraction method (Stover et al., 1976).
Liquid: solid ratio (ml:g DS)
1 M, (pH 6.5)
0.5 M, (pH 6.5)
0.1 M, (pH 9.95)
0.1 M, (pH 4.63)
1.0 M, (pH 0.25)
Approx. 1 h
Sludge pellet + 1.3 ml conc. HNO3 + 2.7 ml conc. HCl
Table 2: Successive stages of the modified phosphorus extraction method.