1151 © IWA Publishing 2015 Water Science & Technology | 71.

8 | 2015

Fat, oil and grease waste from municipal wastewater:

characterization, activation and sustainable conversion
into biofuel
Carlo Pastore, Michele Pagano, Antonio Lopez, Giuseppe Mininni
and Giuseppe Mascolo

ABSTRACT
Carlo Pastore (corresponding author)
Fat, oil and grease (FOG) recovered by the oil/water separator of a wastewater treatment plant
Michele Pagano
(WWTP) were sampled, characterized, activated and converted into biofuel. Free acids (50–55%) and Giuseppe Mascolo
CNR-IRSA,
fatty soaps (26–32%) not only composed the main components, but they were also easily separable via De Blasio 5,
70132,
from the starting waste. The respective free fatty acid profiles were gas-chromatographically Bari,
Italy
evaluated, interestingly verifying that free acids had a different profile (mainly oleic acid) with respect
E-mail: carlo.pastore@ba.irsa.cnr.it
to the soapy fraction (saturated fatty acids were dominant). The inorganic composition was also
Antonio Lopez
determined for soaps, confirming that calcium is the most commonly present metal. The chemical Giuseppe Mininni
CNR-IRSA,
activation of this fatty waste was made possible by converting the starting soaps into the respective via Salaria km 29,300,
00015,
free fatty acids by using formic acid as activator, coproducing the relevant formates. The activated Montelibretti,
Roma,
fatty matter was then converted into biofuel through direct esterification under very mild conditions
Italy
(345 K, atmospheric pressure) and obtaining thermodynamic conversion in less than 2 h. The process
was easily scaled up, isolating at the end pure biodiesel (purity > 96%) through distillation under
vacuum, providing a final product conformed to commercial purposes.
Key words | biodiesel, calcium soaps, FOG, WWTP-waste valorization

INTRODUCTION

Fat, oil and grease contained in urban wastewater (UW- amount of UW-FOG harvestable from WWTPs. As a
FOG) are known to represent a problem for clogging pipe- rough estimation, considering that the oil concentration in
lines during sewage transportation, potentially causing urban wastewater is usually ranged between 50 and
sanitary sewer overflows (Williams et al. ; Mattsson 150 ppm (Passino ), from a WWTP sized for 100,000
et al. ). Conversely, they also negatively affect the man- of equivalent inhabitants, about 1,600–3,800 kg of UW-
agement costs of wastewater treatment plants (WWTPs). FOG per day might be collected. Such an amount not
Specifically, when grease is separated in the influent as an only cannot be ignored, but if efficiently separated could
upper emulsion through a flotation step, it becomes a represent an exploitable resource for biofuel production as
special waste (European Waste Code (EWC) 190809) well as having the further benefit of, lightening the succes-
whose disposal may account for up to 10% of the total sive WWTP oxidation steps. The valorization of UW-FOG
sludge disposal cost. Because of this, it is a quite common has been studied only recently with the aim of producing
procedure in the management of WWTPs to avoid the pre- methane under mesophilic conditions (Martín-González
liminary separation of UW-FOG. However, this et al. ; Wang et al. ). By contrast, even though it
technological solution causes a hampering of successive has been mentioned several times as a possible starting
processes, because lipids have a detrimental effect on raw material for producing biodiesel, UW-FOG has only
oxygen transfer, thereby causing a severe overcharge of sparingly been studied in practice (Pastore et al. ;
the oxygen supply during aerobic secondary treatments Stacy et al. ), probably due to its over-complex
(Chipasa & Medrzycka ). Such a situation also makes composition. In this paper, the detailed characterization
it really complicated to correctly estimate the effective of the raw oily waste is presented, together with the
doi: 10.2166/wst.2015.084
 1152 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

consequentially planned chemical activation and successive Activation of UW-FOG and its conversion into FAMEs
conversion into biofuel. Calcium soaps represent an impor- under AlCl3·6H2O catalysis
tant component of the oily phase recovered. Then, in order
to convert this oily waste into fatty acid methyl esters To about 500 g of UW-FOG, 27.0 g HCOOH (which rep-
(FAMEs), a preliminary activation of this soapy component resent the stoichiometric amounts with respect to the
by formic acid was required to produce the relevant fatty soapy components) were added and mixed at room tempera-
acids. Together with free fatty acids (FFAs), Ca(HCOO)2 ture in a 2 L beaker for 2 h. Then, 500 mL of solvent
was also obtained using only the stoichiometric amount of (hexane, heptane) were added to dissolve and separate the
the acid. Then, aluminium chloride hexahydrate was effi- oily phase from the new solid obtained, which was filtered
ciently used as catalyst, resulting in the same time and and washed with new fresh solvent (100 mL × 2). The acti-
being equally active and safer than conventional mineral vated UW-FOG (AUW-FOG), ready to be used in direct
acids, and producing at the end of the process a potentially esterification, was finally recovered as residue by distilling
(re)usable residual aqueous stream (Gregory & Duan ). away the solvent under vacuum (450–480 g, with a FFAs
content >82%). On isolated solids (10–4%), elemental ana-
lyses of metals were carried out after their mineralization.
MATERIALS AND METHODS The isolated AUW-FOG was completely dissolved into
434 g of methanol and the limpid solution was kept under
All solvents, reagents and standards were RP Aldrich pro- stirring and heated to 345.15 K. After a preliminary control
ducts. HCl and KOH 0.1 mol/L used for acids and soaps of the starting composition in terms of FFAs, 4 mL of a
determinations were Carlo Erba Normex solutions. The methanol solution of AlCl3·6H2O (prepared dissolving
standard solutions used for the metals determination were 371 g of it in methanol into a 20 mL calibrated flask) were
purchased from Flucka. directly added into the closed reactor (through the septum
Infrared spectra were recorded with a Perkin-Elmer using a syringe). In this way, the final ratio of FFAs:MeOH
Fourier transform infrared spectroscopy (FTIR) Spectrum was about 1:10, while the concentration of AlCl3·6H2O
BX instrument using KBr cells. referred to FFAs was 2% molar. The reaction time started
at the moment of the addition of the catalyst. After 2 h,
Preparation and characterization of the raw grease the agitation was stopped and the oily phase was easily
from the raw waste recovered from the heavier phase of the obtained
bilayered-solution. In fact, in such a heavier phase, most of
About 25 L fresh samples of EWC 190809 were taken the FAMEs produced (94 ± 1%) were accumulated. For
directly from the Bari West WWTP’s oil–water separator. improving the conversion, a further cycle with fresh metha-
The sample was dewatered for 18 h in a tank obtaining a nol and AlCl3·6H2O was carried out on the later isolated
total solids (TS) reduction from 12.5 ± 0.8 to 22.4 ± 1.4%. phase, obtaining at the end a final product with a very low
The final wet sludge was heated up to 343.15 K and centri- residual FFAs content. Residual methanol was then evapor-
fuged at 3,000 rpm for 5 min with a Rotofix 32 Hettige ated and very pure biodiesel was distilled at 443 K under
Centrifuge. The centrifugation of this heated liquid mass vacuum at 50 mmHg (yield >75% referred to the starting
produced a three layered system, whose top part was an grease compound), by using an appropriate distillator.
oily phase (12% of the raw material, 53.5% of the starting A simpler procedure was also tested by using directly
TS) that was recovered and transferred into brown-glass bot- MeOH (434 g) for carrying out the separation between
tles, then stored in a fridge (273 K). Already at room AUW-FOG and calcium formates. The limpid solution
temperature, this oily phase assumed a creamy consistency recovered by filtration was reacted as described before, con-
allowing its easy manipulation (UW-FOG). firming the same reactivity data in terms of final FFAs
Mono-, di- and triglycerides were determined by the conversion.
EN14105 method for biodiesel, whereas waxes were deter-
mined using the COI/T.20/ Doc. no.18/ Rev. 2 method. Separation and characterization of the hexane-soluble
Finally, sterols and aliphatic alcohols were also analysed fraction from fatty soaps
using published procedures (Pastore et al. ). Aliphatic
alcohols (0.5%), waxes (0.3%), sterols (0.2%) and only To about 5 g UW-FOG 20 mL hexane were directly added and
traces of glycerides were found. stirred in a centrifuge tube of 50 mL. After centrifugation at
 1153 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

3,000 rpm for 2 min, it was possible to separate an organic sol- and methyl red as indicator. The respective mass percen-
ution from a white-grey powder. The grey solid was then tages were calculated by using the average molecular
washed again two more times with fresh 20 mL, and the sol- weight, gas-chromatographically determined.
uble fractions collected together. From both the solution
and the wet solid, the hexane was completely evaporated Metals analysis
obtaining a residue of a liquid oily phase (3.5 ± 0.5 g) and a
white-grey powder (1.5 ± 0.4 g), respectively. The solid frac- In total, 100 mg solid sample were mixed with 10 mL of 65%
tion was further re-suspended in hexane and passed through HNO3 and heated up to formation of red-brown fumes. Then,
a 200 μm sieve, separating a fine powder fraction (90–95%) further HNO3 (5 mL) was added and the solution was heated
from impurities. Both the soluble fraction and the fine solid again. These operations were repeated until no oxidation was
were analysed to determine the respective fatty acid profiles apparent. The sample was then reduced to a volume of 2–
and their FTIR spectra were recorded. 3 mL. Finally, they were diluted using distilled water. Samples
were analysed using an Agilent 7700× Inductively Coupled
Fatty acid profile and average molecular weight Plasma–Mass spectrometer (ICP-MS) equipped with a col-
determinations lision cell used in He mode (APHA : 3120 B).
Mineralized sample was also analysed through atomic absorp-
The fatty acid profile was determined weighing 20 mg of tion spectrometry using a SpectrAA-250 Plus from Varian
sample into a glass reactor of 5 mL, together with toluene, (Santa Clara, CA, USA) (APHA : 3111B). Titrimetric
methanol and concentrated H2SO4 (2:2:0.01 v:v:v). The determination of calcium with EDTA was also carried out,
tube was closed and then placed into an ultrasonic bath adopting the standard APHA procedure: 3500 Ca-D.
and kept at 343 K for 2 h. Then, 1 μL of the upper phase
was injected into a Varian 3800 Gas Chromatograph– Analysis of formates
Flame Ionization Detector (GC-FID) equipped with a
MDN-5S capillary column (15 m; 0.32 mm, 0.25 μm film). Formates were determined as HCOOH on hydrolysed
Qualitative attribution of signals to the different fatty samples, by using a GS50 chromatography system
acids was carried out through gas chromatography–mass (Dionex-Thermo Fisher Scientific, Sunnyvale, CA, USA)
spectrometry (GC-MS) analyses with a Clarus 500 gas chro- equipped with an AS50 autosampler, an ED50 conductivity
matograph (Perkin-Elmer) equipped with a VF-5MS detector and an AMMS-ICE MicroMembrane™. The starting
capillary column (20 m; 0.32 mm, 0.25 μm film) interfaced solid samples were freely dissolved in a 0.05 mol/L HCl aqu-
with a Clarus 500 Mass Spectrometer, and comparing the eous solution, and injected via a 25 μL loop, into an IonPac
retention times and the mass profiles with those obtained ICE-AS6 Analytical column (250 mm, 9 mm; Dionex) using
from pure standards. a 1 mmol/L heptafluorobutyric acid (PFBA) solution, at a
Average molecular weight was calculated using the GC- flow rate of 1.0 mL min1. Quantitative and qualitative
FID analysis data and applying the following equation: determinations were carried out, by injecting standard sol-
utions of formic acids.
P
A MWi
AMW ¼ Pi (1)
Ai
RESULTS AND DISCUSSION
where Ai and MWi are the area (gas-chromatographically
detected) and the molecular weight of an identified FFA, Fresh samples of the oily wastes, identified with the EWC
respectively. 190809, named ‘grease and oil mixture from oil/water separ-
ation containing only edible oil and fats’, produced by Bari
FFAs and soaps determinations west WWTP (south of Italy, population equivalent of
242,000, annual production of EWC 190809 of 100 metric
FFAs were determined through a base titration using a tonnes) were sampled and processed as already described,
0.1 mol/L KOH solution, in a diethyl-ether:ethanol 1:1 sol- obtaining UW-FOG. Through titrimetric analysis, this fatty
ution using phenolphthalein as indicator. Fatty soaps were mixture showed a resulting composition of 50–55% FFAs
determined through a titration carried out in organic solvent and 26–32% soaps. The presence of these two classes of com-
(diethyl-ether:ethanol 1:1) using a 0.1 mol/L HCl solution pounds was confirmed by FTIR analysis for the presence of
 1154 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

the typical bands located at 1,700–1,710 cm1, attributable to As for the soaps, the nature of the inorganic component
the asymmetric stretching of carbonyl functionality of FFAs, was qualitatively and quantitatively determined on the
and those at 1,580–1,530 cm1 relative to the stretching of mineralized samples, by using three different techniques:
carbonyl of soaps (see Figure 1(a)). By applying Equation ICP-MS, atomic absorption and titrimetric analysis (see
(2), adapted from the equation used by Poulenat et al. Table 1) (APHA : Method 3500- Ca–B, C and D).
(), on the FTIR spectra recorded on these samples, it From these analyses, the preliminary data obtained by
was possible to confirm a soap content of 26 ± 3%. FTIR concerning the high presence of calcium were con-
firmed. Calcium quantification resulted already effective by
Soaps Absorbance titrimetric analysis using EDTA.
%Soaps ¼ 100 (2)
Soaps Absorbance þ Acids Absorbance To define more precisely the nature of the organic part
for the two fractions obtained by the separation with
Interestingly, an easy separation/isolation of these two hexane, the determination of the fatty acid profile was car-
classes of compounds was possible through the use of ried out through GC-FID analysis on the respective
hexane. By weighting the respective residual masses, it was esterified samples (Figure 2).
confirmed again the value of 30 ± 3% for the soaps content Figure 2 shows the fatty acid profile related to UW-FOG,
into starting UW-FOG. Figure 1(b) and 1(c) show the FTIR FFAs and calcium soaps. It is worth noting the quite differ-
spectra recorded on these isolated fractions. Easily recogniz- ent composition of soaps and FFAs. Through this
able in the two spectra are again the aforementioned signals determination and the use of methylheptadecanoate, it
typical of FFAs and soaps. For the latter, the signals centred was possible to estimate not only the qualitative profile of
at 1,576 and 1,538, related to carbonyl stretching are better acids contained in the two different components, but also
resolved and can be attributed to calcium fatty soaps (Poule- to evaluate their fatty acid percentage (80–83% and 69%,
nat et al. ). respectively).

Figure 1 | (a) FTIR spectra of UW-FOG (neat compound, KBr cells); (b) FTIR spectra of extracted FFAs (neat compound, KBr cells); (c) FTIR spectra of calcium soaps (nujol mull, KBr cells); (d)
FTIR spectra of calcium formate (nujol mull, KBr cells).
 1155 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

Table 1 | Qualitative and quantitative metals analysis of the soapy fraction In any case, as far as our purposes of valorization are
concerned, this starting composition of UW-FOG needs a
AA ICP-MS Titrimetry
APHA 3500-B APHA 3500-C APHA 3500-D
preliminary activation of the calcium soap component
Metal %wt %wt %wt through an acid in order to obtain FFAs. Such components
Na 0.14 ± 0.03 0.08 ± 0.01 – are, in turn, easily convertible into FAMEs through direct
K 0.05 ± 0.01 0.03 ± 0.01 – esterification. Mineral acids, such as HCl, H2SO4 and
Mg 0.03 ± 0.01 0.05 ± 0.01 –
H3PO4, are normally involved for such a step. However,
the use of these acids may cause a complication in terms
Ca 6.75 ± 0.05 6.80 ± 0.05 6.9 ± 0.2
of final management and treatment of the relevant residual
Fe 0.04 ± 0.01 0.03 ± 0.01 –
aqueous streams. For this reason, the activity of the strongest
Al Nd 0.03 ± 0.01 –
organic acid was tested, namely HCOOH (pKa ¼ 3.77) in
Mn Nd Nd –
the calcium soaps activation process (Equation (3)). In this
Cu Nd Nd – way, at the end of the activation, besides the formation of
Zn Nd 0.01 ± 0.01 – new FFAs, also calcium formates would be produced
which could have further possible application.

(RCOO)2 Ca þ 2HCOOH ⇌ RCOOH þ Ca(HCOO)2 (3)

The reaction was directly tested on UW-FOG by adding

the stoichiometric amount of HCOOH without the use of
any solvent and keeping the overall mass, under stirring at
room temperature. After 2 h, the starting creamy mixture
assumed a more liquid consistency, with a powdery white
solid dispersed in it. The extraction with an organic solvent
(hexane or heptane or methanol) operated on this complex
semi-liquid mixture allowed a new solid to be isolated (the
FTIR spectrum is depicted in Figure 1(d)). Experimental
content of formates on this new white-powder resulted at
Figure 2 | Fatty acid profiles recorded for soaps, FFAs and UW-FOG. Error bars indicate
69.1 ± 0.6%, while calcium content resulted at 31 ± 0.3%,
the deviation in the readings between 10 experiments. in good agreement with the theoretical values of Ca
(HCOO)2, which are 69.2% and 30.8%, respectively. This
Matching all the experimental data, it was also possible means that the reaction was already completed by using stoi-
to characterize the isolated solids as calcium soaps, having chiometric formic acid. Considering the slight difference of
the formula Ca(RCOO)1,5. Indeed, the same composition acidities between FFAs and HCOOH (ΔpKa∼1), it was an
was found in primary sludge from the same WWTP (Pastore unexpected conversion, probably due to the concomitant
et al. ), and both look very similar to the soapy fraction production of calcium formate that further stabilized the
directly analysed on human faeces (Sammons & Wiggs overall system.
; Owen et al. ). In addition, the composition of Interestingly, the direct use of the right amount of
the soapy fraction in terms of saturated fatty acids, looks methanol for the separation of calcium formates from
very similar to the FOG deposits reported by Iasmin et al. AUW-FOG made possible the following direct esterification
(). These analogies open a new possible scenario, for (Equation (4)), without the use of other intermediate steps
explaining the phenomena involved in the clogging of pipe- (and solvents). Through a simple filtration of calcium for-
lines, suggesting that there could be a significant mates, the limpid reddish solution was directly heated up
contribution of the soapy components that come from to 345 K and allowed to react for 2 h by adding AlCl3·6H2O.
human faeces, together the hydrolysis and saponification
of lipids occurring during sewage transportation (He et al.
AlCl3 6H2 O
; He et al. ; Iasmin et al. ). Further investigations 345K, 2h
→
have been carried out about such an issue. RCOOH þ MeOH ← RCOOMe þ H2 O (4)
 1156 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

Even though a 50 times bigger system was used with consist in maximizing resources recovery from sludge, in
respect to previous works (Pastore et al. ), the resulting which the conversion of lipid of primary sludge into FAMEs
reactions were equally effective. In fact, the thermodynamic is also included (Pastore et al. ), and the anaerobic diges-
conversion was already obtained. In addition, the recovery tion is limited to the residual phase (Olkiewicz et al. ).
of FAMEs was simplified by obtaining a final biphasic Such a very innovative approach could definitively turn
system in which the bottom phase contained most of the WWTPs into biorefineries.
synthesized product (94%). The composition of the upper
phase, instead, was mainly due to methanol, residual FFAs
and most of the catalyst used. Such a distribution allowed
CONCLUSIONS
the upper phase to be reused with new AUW-FOG for a
new cycle of reaction, satisfying the criteria of resources
A very sustainable valorization of EWC190809 was studied.
optimization. This recyclability was positively tested for
The chemical process was based on the starting waste
three times. At the end, aluminium could be anyway recov-
composition, following a preliminary detailed characteriz-
ered and employed as a flocculant, after its basification, in
ation. Interestingly, the soapy fraction resulted in a different
sedimentary processes involved in WWTPs.
fatty acids profile, with respect to those revealed on the over-
The final product obtained by direct esterification, after
all UW-FOG, indicating its direct human origin. The
evaporation of residual methanol, was distilled under
activation of this soapy fraction was already efficiently
vacuum (424–434 K, 50 mTorr) for a final yield of 72% of
obtained by the addition of the stoichiometric amount of
FAMEs referred to UW-FOG. The distillate was analysed
formic acid, coproducing calcium formates as the only by-
and found to be absolutely pure (title >96%) as well as com-
product, which easily could find a proper application. Con-
pliant to EN14214 and so ready to be introduced into the
versely, final obtainment of biofuel was completed through
current biofuels market. This easy up-scalable efficient and
a direct esterification mediated by AlCl3·6H2O and the very
fast procedure (yield of about 70%, in 3–4 h) could be con-
pure product was isolated by distillation under vacuum
sidered highly competitive with respect to an anaerobic
process.
digestion (which usually needs a co-feed of urban waste
and at least 15–20 days of time) and even easy to be
placed into a WWTP context, allowing the waste produced
to be immediately processed in situ. ACKNOWLEDGEMENTS
The economic convenience of the proposed process could
be easily quantified: per each kg of starting FFAs, around 1 kg This study was financially supported by CNR through ‘Bio-
of final refined product could be obtained, whose market value raffineria di terza generazione integrata con il territorio e
is about 1 Euro (0.85 Euro/L). The respective main costs of biocombustibili’ Progetto Premiale. The authors would like
production are limited to the reacting methanol (125 g to acknowledge Giuseppe Bagnuolo, Ruggero Ciannarella,
quoted 0.10 Euro; Pokoo-Aikins et al. ) and formic acid Regina Del Sole and Francesco Greco for their experimental
(50 g, 0.05 Euro). As for the thermal balance, it was already support.
demonstrated that one tenth of the final product covers all
the energy required to sustain the overall thermal needs (Pas-
tore et al. ). The difference between market value of final REFERENCES
product and respective costs of production, not only makes
possible an easy recovery of the starting financial investment, APHA  Standard Methods for the Examination of Water and
Wastewater, 21st edn. American Public Health Association,
but it can also compete with present biodiesel production-sys-
Washington, DC, USA.
tems. On the other hand, from the same amount of FFAs, the Chipasa, K. B. & Medrzycka, K.  Behavior of lipids in
maximum obtainable bio-methane from anaerobic digestion is biological wastewater treatment processes. J. Ind. Microbiol.
around 1 Nm3, whose market value is about 0.25 Euro. To be Biotechnol. 33, 635–645.
competitive with natural gas, very strict conditions need to be Gregory, J. & Duan, J.  Hydrolyzing metal salts as coagulants.
Pure Appl. Chem. 73, 2017–2026.
verified (particular stakeholder contexts, subsidies, etc.). At
He, X., Iasmin, M., Dean, L. O., Lappi, S. E., Ducoste, J. J. & de los
the end, the process reported in this work could be well Reyes, III, F. L.  Evidence for fat, oil, and grease (FOG)
inserted into a new and more complex and integrated waste- deposit formation mechanisms in sewer lines. Environ. Sci.
water sludge management, whose main objective would Technol. 45 (10), 4385–4391.
 1157 C. Pastore et al. | Fat, oil and grease waste from municipal wastewater Water Science & Technology | 71.8 | 2015

He, X., de los Reyes, III, F. L., Leming, M. L., Dean, L. O., Lappi, Pastore, C., Lopez, A. & Mascolo, G.  Efficient conversion
S. E. & Ducoste, J. J.  Mechanisms of fat, oil and grease of brown grease produced by municipal wastewater
(FOG) deposit formation in sewer lines. Water Res. 47, 4451– treatment plant into biofuel using aluminium chloride
4459. hexahydrate under very mild conditions. Bioresour.
Iasmin, M., Dean, L. O., Lappi, S. E. & Ducoste, J. J.  Factors Technol. 155, 91–97.
that influence properties of FOG deposits and their Pokoo-Aikins, G., Heath, A., Mentzer, R. A., Mannan, M. S.,
formation in sewer collection systems. Water Res. 49, 92–102. Rogers, W. J. & El-Halwagi, M. M.  A multi-criteria
Martín-González, L., Castro, R., Pereira, M. A., Alves, M. M., Font, approach to screening alternatives for converting sewage
X. & Vicent, T.  Thermophilic co-digestion of organic sludge to biodiesel. J. Loss Prevent. Proc. 23, 412–420.
fraction of municipal solid wastes with FOG wastes from a Poulenat, G., Sentenac, S. & Mouloungui, Z.  Fourier
sewage treatment plant: reactor performance and microbial transform infrared spectra of fatty acid salts: kinetics of high-
community monitoring. Bioresour. Technol. 102, 4734–4741. oleic sunflower oil saponification. J. Surfact. Deter. 6 (4), 305–
Mattsson, J., Hedström, A., Viklander, M. & Blecken, G.-T.  310.
Fat, oil, and grease accumulation in sewer systems: Sammons, H. G. & Wiggs, S. M.  The separation, estimation
comprehensive survey of experiences of Scandinavian and analysis of calcium soaps in human faeces. Clinica
municipalities. J. Environ. Eng. 140, 04014003. Chimica Acta 5, 141–145.
Olkiewicz, M., Caporgno, M. P., Fortuny, A., Stüber, F., Fabregat, Stacy, C. J., Melick, C. A. & Cairncross, R. A.  Esterification of
A., Font, J. & Bengoa, C.  Direct liquid–liquid extraction free fatty acids to fatty acid alkyl esters in a bubble column
of lipid from municipal sewage sludge for biodiesel reactor for use as biodiesel. Fuel Process. Technol. 124, 70–
production. Fuel Process. Technol. 128, 331–338. 77.
Owen, R. W., Weisgerber, U. M., Carr, J. & Harrison, M. H.  Wang, L., Aziz, T. N. & de los Reyes, III, F. L.  Determining
Analysis of calcium-lipid complexes in faeces. Eur. J. Cancer the limits of anaerobic co-digestion of thickened waste
Prev. 4 (3), 247–255. activated sludge with grease interceptor waste. Water Res. 47,
Passino, R.  Manuale di conduzione degli impianti di 3835–3844.
depurazione delle acque, 3rd edn. Zanichelli/ESAC, Roma. Williams, J. B., Clarkson, C., Mant, C., Drinkwater, A. & May, E.
Pastore, C., Lopez, A., Lotito, V. & Mascolo, G.  Biodiesel  Fat, oil and grease deposits in sewers: characterisation
from dewatered wastewater sludge: a two-step process for a of deposits and formation mechanisms. Water Res. 46,
more advantageous production. Chemosphere 92, 667–673. 6319–6328.

First received 14 October 2014; accepted in revised form 9 February 2015. Available online 20 February 2015