Chemosphere 91 (2013) 963–969

Contents lists available at SciVerse ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Extracting phosphorous from incinerated sewage sludge ash rich in ironor aluminum

0045-6535/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.chemosphere.2013.01.101

⇑ Corresponding author. Tel.: +45 45252260; fax: +45 45883282.E-mail address: lo@byg.dtu.dk (L.M. Ottosen).

Lisbeth M. Ottosen ⇑, Gunvor M. Kirkelund, Pernille E. JensenDepartment of Civil Engineering, Building 118, Technical University of Denmark, 2800 Lyngby, Denmark

h i g h l i g h t s

" Two sewage sludge ashes rich in Fe or Al are characterized and acid washed." Fe for P-precipitation during waste water treatment ease P recovery from the ash." Heavy metals are extracted simultaneous to P." Extraction in H2SO4 result in significant formation of gypsum crystals.

a r t i c l e i n f o

Article history:Received 9 July 2012Received in revised form 19 January 2013Accepted 25 January 2013Available online 9 March 2013

Keywords:Sewage sludge ashCharacterizationPhosphorousHeavy metalsGypsum

a b s t r a c t

Ashes from mono-incineration of sewage sludge (ISSA) generally contain high concentrations of phospho-rous (P) and can be regarded as secondary P resources. ISSA has no direct value as fertilizer as P is notplant available. The present paper experimentally compares P extraction in acid from two different ISSAs;one rich in Al (67 g/kg) and the other in Fe (58 g/kg). The difference related to P precipitation at the wastewater treatment facilities. Another major difference between the ashes was that flue gas purificationproducts were mixed into the first ash and it contained about 5% activated carbon. The Al rich ash hada significantly higher buffering capacity and required more acid for extraction of P. When acid extractionof P from ISSA is the method for recovery, it is thus beneficial to go back to the waste water treatmentfacility and e.g. choose Fe for P precipitation rather than Al.

Formation of a high amount of gypsum crystals in both ashes after extraction in H2SO4 was seen bySEM–EDX. H2SO4 is the cheapest mineral ash, but the gypsum formation must be taken into accountwhen either finding possibility for using the remaining ash in e.g. construction materials or if the choiceis deposition, as the gypsum increases the volume significantly.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Phosphorus is an essential and irreplaceable, yet limited re-source. This is why efforts increase to recycle phosphorous con-tained in wastewater (Cornel and Schaum, 2009). Sewage sludgefrom waste water treatment generally contains appreciableamounts of N and P, but during water and sludge treatment, Nmay escape as gaseous compounds whereas the P is entirely re-tained in the sludge (Huang and Shenker, 2004). Sewage sludgehas been widely applied to agricultural land, but limit values forpollutants in both EU (Council Directive 86/278/EEC) and at na-tional levels e.g. in Denmark (BEK 1650, 2006) restrict this directuse.

Sewage sludge incineration is a well proven technology withhundreds of plants worldwide (Werther and Ogada, 1999). Inciner-

ation results in a mass reduction, but also in a concentration ofheavy metals. On a dry matter basis the heavy metal content inthe ash following incineration increase considerably compared tothe feed sludge (Lapa et al., 2007; Donatello et al., 2010c). The con-centration of P in ISSA is generally 5–10% per weight (Cyr et al.,2007). ISSA is thus a potential secondary resource for P, and thispaper focus on P recovery. Considerable literature is published onthe use of ISSA in construction materials (e.g. in cementitiousmaterial, bricks or lightweight aggregates) (see Donatello, 2009for review), but none of these applications make use of the valu-able P resource in the ISSA.

The P in ISSA is not plant available. The gain in crop yield whenspreading the ash directly is consequently limited. Development ofdifferent techniques for P recovery from ISSA is ongoing. Recoveryby particle separation is not possible (Franz, 2008), and dissolutionfrom ash particles is necessary. The techniques for this are eitherthermal separation or wet chemical treatment followed by a sepa-ration step (Cornel and Schaum, 2009).

964 L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969

The present paper deals with acid extraction for P recovery. Thecharacteristics of ISSA vary depending on the socio-economic nat-ure of the catchment area, type of sludge incinerated and sludgepretreatment process (Donatello, 2009). During waste water treat-ment different precipitating agents are used to retain P to avoideutrophication of surface waters. These precipitating agents are of-ten Al-, Fe or Ca-rich compounds, reacting with P to precipitate sta-ble phosphates in the sludge. This paper focuses on the influence ofAl and Fe used in the waste water treatment process on the char-acteristics of the ISSA and on extraction of P by acid. Experimentalwork is conducted with two ashes. The obtained knowledge maybe used when operating waste water treatment plants in order toproduce ISSA from which the P recovery consumes least acid, thusbenefitting the recovery.

2. Experimental

2.1. Sewage sludge ashes for experiments

ISSAs from two Danish mono-incineration plants were used. Inboth the sludge was incinerated in a fluidized bed combustor atabout 850 �C. The sewage sludge incinerated was from urbanwastewater in both cases. The two ashes differ in the chemicalsoriginally used for P precipitation during waste water treatmentand in the ash collection systems at the incineration plants.

� MAl: Mølleåværket Renseanlæg Lundtofte incinerates sludgefrom wastewater treatment from 1,35,000 PE. P was precipi-tated with Al. The ash was a mixture of ash from a multicyklonand flue gas purification products. The flue gas purification sys-tem was treatment with sodiumbicarbonate and activatedcarbon.� AFe: Spildevandscenter Avedøre, Hvidovre, incinerates waste-

water sludge from about 3,30,000 PE. P was precipitated withFe. The ash from electrofilters was mixed with flue gas neutral-ization products (lime) from bag filters.

After sampling, the ashes were stored in closed plastic contain-ers at room temperature.

2.2. Ash characterization

Characterization and extraction experiments were made withdried ash (105 �C, 24 h). Concentrations of P, Al, Fe, Na, Cu, Cd,Ni, Pb and Zn were measured after pre-treatment of the ash inaccordance to DS259; 1.0 g ash and 20.0 ml (1:1) HNO3 was heatedat 200 kPa (120 �C) for 30 min. Filtration through 0.45 lm filterand the concentrations were measured with ICP–OES. Ash pHwas measured by suspending 10.0 g ash in 25 ml distilled water.After 1 h agitation pH was measured directly in the suspension.Loss on ignition (LoI) was found after 30 min at 550 �C. Water con-tent was measured as weight loss after 24 h at 105 �C (calculatedas weight loss over the weight of the wet sample). Five duplicatesof each of these analyses were made. Grain size distribution wasdetermined with laser diffractometry. Solubility in water was eval-uated: 50.0 g ash suspended in 500 ml distilled water and agitatedfor 1 min. After settling the water was decanted. New 500 ml dis-tilled water added. This was repeated so the ash was washed threetimes. Finally the suspension was filtered and the ash dried andweighed. The gas production after suspension in 10% HNO3 wasmeasured with a volumetric calcimeter method.

Morphology and element distribution analysis was performedby SEM–EDX. No pre-treatment of the samples was made. Theaccelerating voltage of the SEM was 20–25 kV with large field

detector and X-ray cone. Major element content was determinedby X-ray fluorescence (XRF).

2.3. Acid extraction

Extractions were made in HNO3 and H2SO4 with different aims.Extractions in HNO3 show the direct pH dependent extraction.H2SO4 is cheapest mineral ash and thus relevant for P extractionin industrial scale. In H2SO4 extractions, however, there is a riskfor precipitation of different sulfates and the actual extraction asa result of lowering pH of the elements forming sulfates cannotbe evaluated.

2.3.1. pH dependent extraction in HNO3 of various elementsTo examine the pH dependent extraction of P, Al, Fe, Na, Cu, Ni,

Pb and Zn from the ashes, the following procedure was used: 10.0 gdry ash and 50 ml HNO3 in various concentrations (from 0.01 M to1.5 M) were suspended for 1 week. The suspensions were filtered(0.45 lm). The concentrations were measured in the liquid phasewith ICP–OES. Each extraction was made in duplicate. Extractionsin distilled water were made as reference.

2.3.2. Extraction in H2SO4 for P recoveryExtractions were made in different acid concentrations and L:S

ratios to create a range of different pH environments in which thesolubility of P and other elements can be monitored. Extractions inH2SO4 were made for 2 h or 1 week (all as double determinations)in 0.19 M H2SO4 at liquid to solid ratios (L:S) of 2, 5, 10 and 20.Extractions were also made in H2SO4 for 2 h in 0.08 M H2SO4 and0.28 M H2SO4 at L:S of 2, 5, 10 and 20. Conductivity and pH weremeasured in the suspensions. The suspensions were filtered(0.45 lm) and the filtrate saved for ICP–OES analysis (Cu and P).The ash at the filter paper was flushed with distilled water undervacuum. The water was discharged and the ash dried and savedfor SEM–EDX analyses.

3. Results and discussion

3.1. Chemical characterization of ashes

Results from the ash characterization are shown in Table 1. Bothashes are secondary resources for P as they contain 7% and 10% perweight. A much higher Fe concentration was found in AFe than inMAl whereas the Al concentration was highest in MAl. The mutualdistribution underlines the use of chemicals during waste watertreatment. The Na concentration was much higher in MAl than inAFe, which was related to use of sodium bicarbonate to neutraliza-tion the flue gas.

The LoI was very low in AFe (0.2%) revealing good incineration.This LoI was lower than in seven English ISSA (between 0.9% and2.1%), which were characterized in (Donatello, 2009). The LoI inMAl (5.7%) was on the contrary higher than in these ashes. Duringextraction experiments with MAl small, black particles were float-ing. These contributed to about 6% of the total weight, correspond-ing well to the LoI. These black particles may be activated carbonfrom the flue gas purification process.

Both ashes were alkaline (pH about 10). Apart from this, theywere reacting differently when suspended in water or acid. MAlwas most reactive. The solubility in water of MAl was almost 10times higher. The conductivity measured in an ash suspension inwater was likewise higher (about 20 times). Further the productionof gas when reacting with acid was about 10 times higher fromMAl than AFe.

Table 1Characteristics of the two ashes.

Parameter MAl AFe Danishlimitingvalues

Physical andchemicalcharacteristics

pH (H2O) 10.0 ± 0.02 10.3 ± 0.3Conductivity(mS/cm)

32.4 ± 1.1 1.69 ± 0.04

Loss on ignition(550 �C)

5.7 ± 0.06 0.19 ± 0.02

Solubility inwater (%)

21.5 2.7

Gas productionin acid (ml gas/g)

23.9 ± 0.5 2.0 ± 0.2

Grainsize d50

(lm)110 105

Macro elements P (g/kg) 69.9 ± 2.7 99.0 ± 1.3Na (g/kg) 69.0 ± 1.9 3.7 ± 0.3Al (g/kg) 67.0 ± 3.0 19.7 ± 0.6Fe (g/kg) 15.7 ± 0.9 58.1 ± 1.3

Trace elements Zn (mg/kg) 1700 ± 110 2000 ± 120 4000Cu (mg/kg) 540 ± 8 690 ± 22 1000Pb (mg/kg) 85 ± 1.4 112 ± 1 120Ni (mg/kg) 29 ± 0.8 72 ± 1 30

L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969 965

3.2. Heavy metal concentrations in the ashes

The concentrations of Zn, Cu, Pb, and Ni were little higher in AFethan in MAl (Table 1), but the concentrations were generally low inboth compared to the concentrations in other ISSAs reported in lit-erature and collected by Cyr et al. (2007). The Ni concentration inboth ashes was less than the minimum concentration (79 mg Ni/kg(Cyr et al., 2007)) reported. The concentrations of Cu and Zn hadthe same level as reported for seven ISSAs (Donatello, 2009). ThePb concentration was 4–5 times lower in the two present ashes.Different industries are responsible for high concentrations of hea-vy metals in the sludge and subsequent in the ISSA (Donatello,2009). No major industries discharge waste water to the presenttreatment facilities, thus the low heavy metal content.

The Danish EPA (Miljøstyrelsen) has two sets of limiting valuesfor heavy metals when spreading waste at agricultural land; con-centrations related to the dry matter and concentrations relatedto phosphorous. There were absolutely no problems in meetingthe P-related limiting values for the two ISSA’s. The dry mass re-lated concentrations also met the limiting values, except from Niin AFe (Table 1). Both ashes meet the first set of limiting valuesand could be spread at agricultural land, however, pretreatmentis needed to get plant available P.

3.3. Morphology and mineral composition

X-ray fluorescence analysis showed that with the exception ofoxygen, the major elements present (with a concentration higherthan 2%) in the two ISSAs were:

MAl : Ca ð10%Þ > Al ð8%Þ � Sið8%Þ > Pð7%Þ > Nað6%Þ> Sð4%Þ > Feð2%Þ

AFe : Ca ð17%Þ > Feð12%Þ > Pð9%Þ > Sið8%Þ > Alð2%Þ

In comparison to the concentration ranges found by XRF and gi-ven in Donatello et al. (2010b) for seven different ISSAs, the majordifferences are: lower Si content in the present ashes (12.4% Si wasthe lowest in Donatello et al. (2010b)), high concentration of Ca inAFe and Na in MAl (compared to 13.5% and 1.6% as highest

Donatello et al. (2010b)). The Fe concentration in MAl was lower,and the Al concentration in AFe was lower (4.7% Fe and 6.3% Alwere lowest concentrations in Donatello et al. (2010b)).

The concentrations found with XRF were generally in goodagreement with the concentrations found using DS259 followedby ICP analysis (Table 1) except from Fe in AFe, where the concen-tration is two times higher in the XRF analysis. The differenceshows that the DS259 standard does not give total concentrationsfor Fe and the reason is that there is a small mineral residue, whichis removed before ICP analysis and this residue contain insolubleFe compounds.

The grain size distributions of the two ashes were similar. d50

were 110 lm and 105 lm, which are in the normal range to whatwas reported in literature (Biswas et al., 2009; Donatello et al.,2010a, b, c). Examples of SEM images of the two ISSAs are shownat Fig. 1a and b. The ash particles generally had irregular shapes.The larger particles contained of smaller particles sintered togetherprobably by an amorphous phase, which seemed to have an openstructure. The morphology of these ISSA particles was similar towhat was previously reported e.g. (Donatello, 2009). Spherical par-ticles were noticed in both ashes.

From an SEM/EDX elemental mapping (not shown) it was seenthat O, P, Fe and Al were distributed all over the particle surfaces inboth ashes. In addition, in AFe also K, Na, Mg and Ca were similarlydistributed over the surfaces. Si, S, Na and Cl were on the contraryfound in refined areas in both ashes. In MAl Na and S were showedoverlapping maps to some extent. This is an indication of Na2SO4

which is known to form when using sodium bicarbonate in the fluegas treatment. In MAl K and Ca were found both in refined areaswith high concentrations and sparsely distributed over largerareas. As P was not found in refined particles, it is not possible toseparate phosphorous by particle separation, supporting the find-ing of Cyr et al. (2007).

3.4. Extraction in HNO3 – the pH dependency

The pH in suspensions of the ashes in different concentrationsof HNO3 (Fig. 2) show that the buffering capacity of MAl was high-est as pH in the suspension with same acid concentration washigher. This corresponded well to the higher reactivity of thisash. The activated carbon may also contribute to the bufferingcapacity.

The percentages extracted from the ashes as a function of pHare shown in Fig. 2. Approximately 100% Na was extracted regard-less pH, supporting the presence of Na2SO4 seen by SEM–EDX. Thetotal Na concentration was much lower in AFe and the extractionwas pH dependent (increasing with decreasing pH). The Na con-taining particles were thus generally different in the ashes.

Little P and Al were extracted in water in MAl – 300 mg P/kgand 280 mg Al/kg (both corresponding to 0.4% of the totals). Thisindicates that these two elements were combined in some ash par-ticles, supporting the finding from the SEM–EDX analysis. Of theremaining elements only Na was water extractable. From bothashes P and Al were extracted at low pH. The pH was about 1.7in the MAl suspension before extraction (corresponding to thehighest acid concentration). At pH 2.2 insignificant extractionswere found, whereas at pH 1.7 about 80% P and 50% Al was ex-tracted. This was in contrast to AFe, where especially P was ex-tracted over a wider pH range. The P extraction started at pH 5and Al around pH 3. At pH of about 1 the extraction approached100%. At pH of about 2, however, less P was extracted from AFethan from MAl, but the acid addition to reach pH 2 was signifi-cantly lower. These findings reveal that P was present in differenttypes of particles in the two ashes. The dissociation constants fororthophosphoric acid (H3PO4) are pK1 2.1, pK2 7.2 and pK3 12.7(Corbridge, 1995) and thus the sharp increase in P desorption

Fig. 1. SEM images of (a) MAl, (b) AFe, (c) MAl after extraction in 0.08 M H2SO4 and L:S 20 m, and (d) MAl after extraction in 0.19 M H2SO4 and L:S 20.

966 L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969

below 2 in MAl may be related to the speciation of phosphates, asat pH below 2 the major species is the uncharged H3PO4.

Even at the lowest pH little Fe was solubilized from the ashes –less than 2% and 14% of the amount measured using DS259 fromMAl and AFe, respectively. Fe2O3 which is often found in ISSAs isinsoluble in acid (Donatello, 2009). As little Fe was extracted incontrast to P, Fe and P were generally not associated in AFe, eventhough Fe was originally used for precipitation of P during wastewater treatment. There is, however, a possibility for re-precipita-tion of solubilized Fe during the extractions and thus this is onlyan indication.

The extraction of heavy metals from the ashes showed steepestcurves for MAl. The percentages of heavy metals extracted in 1.5 MHNO3 from the two ashes were:

� MAl: Cu (66%) > Pb (56%) > Zn (16%) > Ni (11%).� AFe: Cu (80%) > Zn (71%) > Pb (47%) > Ni (35%).

The extraction of heavy metals started at the same or higher pHthan P. It is thus impossible to extract P in acid without simulta-neously extracting heavy metals.

3.5. Duration of extraction in H2SO4

Two extractions in H2SO4 were alkaline after 2 h (MAl, L:S 2 and5). Here a decrease of about 1.5 pH unit was seen after 1 week. Onthe contrary, an increase of up to 0.5 pH unit was seen between 2 hand 1 week for the remaining and acidic suspensions. The tendencywas thus pH changes towards neutrality. The decrease in pH of thealkaline samples could be a result of reaction with CO2 from the air(carbonation). Grunning et al. (2010) did not find evidence of

carbonation of another ISSA and revealed it was because it didnot contain CaO detectable by XRD. From XRF of the present ashes,however, a high content of Ca was found and in case of CaO orCaOH carbonation may take place. Another possibility for thedecrease in pH could be slow reaction kinetics for precipitationof hydroxides. The conductivity measured directly in the ashsuspensions decreased in all samples between 2 h and 1 weeksupporting the latter hypothesis.

Donatello et al. (2010a, b, c) suggested 2 h as optimum forextraction of P in H2SO4, because the major part of P was extractedhere. Biswas et al. (2009) reported that in less than 4 h complete Pextraction was obtained. In the present investigation extraction inH2SO4 was conducted for 2 h and for 1 week for comparison. Theresults are shown in Fig. 3 as percentage extracted of Cu, Al andP. Cu was chosen to represent the heavy metals as Cu was ex-tracted to the highest percentage from both ashes in HNO3 andthus represents worse case. In the two suspensions with alkalinepH (MAl 2 h and MAl 1w at L:S 2.5 and 5) insignificant amountswere extracted (Fig. 3). In every other extraction, P was adsorbedor re-precipitated between 2 h and 1 week. Equilibrium was thusnot reached within 2 h. In the most acidic suspensions (both ashes)more Cu was extracted after 1 week than after 2 h; the same wasseen for Al in the most acidic AFe suspension, but for the mostacidic suspensions with MAl, the extracted Al percentage de-creased. For both ashes the highest P extraction (L:S 20) showedthat the duration of 2 h was better than 1 week as most P and leastCu was soluble after 2 h. So the cleanest P product in relation Cuwas obtained after 2 h compared to 1 week. The present investiga-tion suggests that the extraction duration can be too long. Slowreaction kinetics or adsorption may decrease the extracted Pover time. The system is highly complex and from the present

(a)

(b)

Fig. 2. Extraction (1 weeks) of different elements in HNO3 as a function of pH and the pH as a function of HNO3 concentration (a) MAl and (b) AFe.

Cu Al P

Fig. 3. Extracted Cu, Al and P after 2 h and after 1 week at different L:S ratios in 0.19 M H2SO4.

L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969 967

investigation does not answer which processes are occurring, how-ever one likely reaction is adsorption of dissolved P to iron oxides.This adsorption occurs to an increasing extent with decreasing pHand the adsorption was not finished after 2 h when dealing withiron oxide tailings (Zeng et al., 2004).

3.6. Influence of H2SO4 concentration and L:S ratio on extractions ofP and heavy metals

Fig. 4 shows percentages of Cu and P extracted in different con-centrations H2SO4 with varying L:S. Almost full P extraction was

(a) (b)

(c) (d)

Fig. 4. Extracted Cu and P after 2 h in H2SO4 of different concentrations and L:S (a) Ash MAl Cu, (b) Ash MAl P, (c) Ash AFe, Cu and (d) Ash AFe P.

968 L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969

obtained from both ashes in 0.19 and 0.38 M H2SO4 at L:S 20 (Littlemore than 100% was seen in some extractions probably related tothe ash not being homogeneous). The acid addition (mole H+ per gash) was the same in the two extractions 0.19 M H2SO4 at L:S 20and 0.38 M H2SO4 at L:S ratio 10, but the extracted percentage ofP was considerably less in the last, revealing that the P extractionwas dependent on the L:S ratio. The Cu extraction on the contrarywas not dependent on L:S. Biswas et al. (2009) experimentallyfound the optimum extraction of an ISSA sample from a Japanesesewage center to be 0.05 M H2SO4 at L:S 150, which is much higherthan in the present experiments (and a higher acid addition). Suchhigh L:S result in a large volume of solution with a low P concen-tration, which is not optimal for the further processing. Donatelloet al. (2010b) suggested that it would be possible to go as low asL:S 2 by using the right mixer, but this needs to be verified.

3.7. Gypsum formation during extraction with H2SO4

Gypsum from reaction between SO2�4 from the acid and Ca2+ re-

leased from the ash was seen from the SEM–EDX analysis and is inaccordance with the XRD data presented by Donatello et al.(2010a). SEM images of MAl washed for 2 h in 0.08 M H2SO4 atL:S 20 and 2 are shown in Fig. 1c and d. The long, needle like crys-tals were gypsum (EDX analysis showed that they constituted ofCa, O and S). Such crystals were not present in the ash beforeextraction (Fig. 1a and b). The ash at Fig. 1d with highest acid addi-tion clearly had a much higher concentration of gypsum crystalsthan the ash with lower acid addition (Fig. 1c). This is probablylinked to differences in pH. At the highest pH the gypsum forma-tion is limited by a limited release of Ca from the ash, whereasthe lower pH caused a much higher Ca extraction and followinga higher gypsum formation. The SEM images showed that thespherical particles were still present after extraction in H2SO4 bothat pH 4.8 and 2.5. The formation of gypsum will influence the reuseoptions of the treated ash in e.g. construction materials (Donatelloet al. (2010a) showed that acid washing of ISSA in H2SO4 signifi-cantly reduced the strength of mortar samples relative to non-acid

washed ash). If the ash residue is to be deposited, the gypsum willsignificantly increase the volume thus requiring a larger deposit.These considerations must be taken into account when choosingthe acid type. The present investigation was not designed as a com-parison between extraction in HNO3 and H2SO4 and due to differ-ent L:S, durations and acid concentrations in the extractions adirect comparison is not possible. Graphs of extracted percentagevs. pH for one weeks extractions in the two acids are rather similar,and there is no clear indication that one of these acids should bebetter for P extraction than the other.

4. Conclusions

Acid extraction for P recovery was compared for of two differentashes from mono-incineration of sewage sludge, where either Al orFe originally was used for precipitation P from the waste water.One ash was subsequently rich in Al and the other in Fe. The Al richash contained activated carbon and required significantly moreacid for full extraction of P. This reveals that in order to obtainthe lowest acid consumption for P recovery from ISSA, it is prefer-able to use Fe instead of Al for precipitation of P at the waste watertreatment plant.

The heavy metal concentrations in the two ashes were low.Simultaneously to the release of P by acidification, a fraction ofheavy metals was released as well. In order to produce a high gradeP-product from the extracted solution a separation step is thusneeded. Between 2 h and one week of extraction, changes in theextracted amounts were seen. The P concentration decreasedwhereas the heavy metal concentration generally increased inthe liquid phase. Thus to obtain both the cleanest liquid in relationto heavy metals and a maximum P concentration, the duration ofthe extraction process should not be too long.

Sulfuric acid is the cheapest mineral ash and is thus often sug-gested for P extraction from ISSA. The use of this acid thoughcauses formation of a high quantity of gypsum crystals in theremaining ash (seen as needle shaped crystals by SEM). The gyp-sum must be taken into account when handling the remaining

L.M. Ottosen et al. / Chemosphere 91 (2013) 963–969 969

ash in e.g. construction materials or as an increased volume to bedeposited.

Acknowledgement

Miljøstyrelsen (The Danish EPA) is acknowledged for financialsupport.

References

Biswas, B.K., Inoue, K., Harada, H., Ohto, K., Kawakita, H., 2009. Leaching ofphosphorus from incinerated sewage sludge ash by means of acid extractionfollowed by adsorption on orange waste gel. Journal of Environmental Sciences21, 1753–1760.

Corbridge, D.E.C., 1995. Phosphorous. An outline of its chemistry biochemistry andtechnology, Fifth ed. Elsevier Science.

Cornel, P., Schaum, C., 2009. Phosphorous recovery from wastewater: needs,technologies and costs. Water Science & Technology 59, 1069–1076.

Cyr, M., Coutand, M., Clastres, P., 2007. Technological and environmental behaviorof sewage sludge ash (SSA) in cement-based materials. Cement and ConcreteResearch 37, 1278–1289.

Donatello, S. 2009. Characteristics of incinerated sewage sludge ashes: potential forphosphate extraction and re-use as a pozzolanic material in construction

products. PhD Thesis. Department of Civil and Environmental Engineering,Imperial College. London.

Donatello, S., Freeman-Pask, A., Tyrer, M., Cheeseman, C.R., 2010a. Effect of millingand acid washing on the pozzolanic activity of incinerator sewage sludge ash.Cement and Concrete Composites 32, 54–61.

Donatello, S., Tong, D., Cheeseman, C.R., 2010b. Production of technical gradephosphoric acid from incinerator sewage sludge ash (ISSA). Waste Management30, 1634–1642.

Donatello, S., Tyrer, M., Cheeseman, C.R., 2010c. EU landfill waste acceptancecriteria and EU hazardous waste directive compliance testing of incineratedsewage sludge ash. Waste Management 30, 63–71.

Franz, M., 2008. Phosphate fertilizer from sewage sludge ash (SSA). WasteManagement 28, 1809–1818.

Grunning, P.J., Hills, C.D., Carey, P.J., tul=0?>Grunning et al. (2010). Acceleratedcarbonation treatment of industrial wastes. Waste Management 30, 1081–1090.

Huang, Shenker, 2004. Water-soluble and solid-state speciation of phosphorus instabilized sewage sludge. Journal of Environmental Quality 33, 1895–1903.

Lapa, N., Barbosa, R., Lopes, M.H., Mendes, B., Abelha, P., Boavida, D., Gulyurtlu, I.,Santos Oliveira, J., 2007. Chemical and ecotoxicological characterization ofashes obtained from sewage sludge combustion in a fluidised-bed reactor.Journal of Hazardous Materials 147 (1–2), 175–183.

Werther, J., Ogada, T., 1999. Sewage sludge combustion. Progress in Energy andCombustion Science 25, 55–116.

Zeng, L., Xiaomei, L., Jindun, L., 2004. Adsorptive removal of phosphate fromaqueous solutions using iron oxide tailings. Water Research 38, 1318–1326.