A

ctictEpia©

K

1

svmscssa

iEa1

0d

Available online at www.sciencedirect.com

Resources, Conservation and Recycling 52 (2008) 931–940

Environmental performance of wastewater treatment plantsfor small populations

Alejandro Gallego ∗, Almudena Hospido, Maria Teresa Moreira, Gumersindo FeijooDepartment of Chemical Engineering, School of Engineering, University of Santiago de Compostela, Rua Lope Gomez de Marzoa,

Campus Universitario Sur, 15782 - Santiago de Compostela, Spain

Received 13 February 2007; received in revised form 19 December 2007; accepted 4 February 2008Available online 18 March 2008

bstract

Life cycle assessment (LCA) is an environmental tool which allows the calculation of all the environmental loads related to a pro-ess/product/service. In the present work, LCA was applied to analyze the environmental impact of different technologies for wastewaterreatment in small populations. In this study, 13 wastewater treatment plants (WWTPs) of less than 20,000 population equivalent (p.e.) locatedn Galicia (NW Spain) were inventoried. The results of the evaluation of the environmental impact are expressed in terms of diverse impactategories. Normalization identified eutrophication, mainly as P-PO4

3−, N-NH4+ and organic load as chemical oxygen demand (COD) in the

reated effluent, and terrestrial ecotoxicity, due to the heavy metals content in the sludge, as the most significant categories for all WWTPs.lectricity use plays an important role in five of seven impact categories and presents the highest importance in four of them. When com-

aring technologies, secondary treatment technologies such as biodenipho and aerobic–anoxic treatment resulted in a lesser environmentalmpact than extended aeration. The operation of the plants has large contribution on the impact, especially for those that make use of extendederation.

2008 Elsevier B.V. All rights reserved.

cation

wpms

ewcatia(

eywords: LCA; Life cycle assessment; Wastewater treatment plants; Eutrophi

. Introduction

Nowadays, society demands that all processes, product orervices must be also analyzed from an environmental point ofiew, including those that are considered for emissions treat-ent, such as wastewater treatment plants (WWTPs). In these

ystems a variety of processes take place: energy is consumed,hemical reagents are used, and sludge and environmental emis-ions are derived. Therefore, it is necessary to analyze theystem to determine the overall pollution associated to thesectivities.

One of the main challenges for European water authoritiess wastewater treatment policies in small villages. In fact, the

uropean Directive 91/271 states that by 1st January 2006,ll generated emissions from populations between 10,000 and5,000 population equivalents (p.e.) must be treated in facilities

∗ Corresponding author. Tel.: +34 981563100x16020; fax: +34 981547168.E-mail address: agallego@usc.es (A. Gallego).

mclmtt

921-3449/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.resconrec.2008.02.001

; Terrestrial ecotoxicity; Electricity use

ith secondary treatment, except in coastal zones, where onlyrimary treatment is compulsory. This fact justifies the environ-ental study of different applied technologies in villages with

mall numbers of inhabitants.Life cycle assessment (LCA) is a technique for assessing

nvironmental parameters and the potential impact associatedith a product/process/service by means of three actions: the

ompilation of an inventory of relevant inputs and outputs ofproduct system; the evaluation of the potential environmen-

al impact associated with those inputs and outputs; and, thenterpretation of the results of the inventory analysis and impactssessment phases in relation to the objectives of the studyBaumann and Tillman, 2004).

There are several previous LCA studies on wastewater treat-ent. Lassaux et al. (2007) recently studied the anthropogenic

ycle of water from pumping stations to WWTPs in the Wal-

on Region (Belgium). The chosen functional unit was a cubic

eter of water. They concluded that the more wastewater treatedhe less the global impact. Along the cycle, the stages that con-ributed most to the impact were the following: water discharge,

9 vatio

Ws

ILw1timnwto

attuCttgs

eidEEnecfio

atTia

2

am

2

hthpiW

reasao

2

sfflo2

(rn

ip1

2

ttatamWf

dmstE

•

•

32 A. Gallego et al. / Resources, Conser

WTP operation and, to a lesser extent, construction of theewage system network.

Some references are available in relation to small city centers.n Sweden, Tillman et al. (1998) and Lundin et al. (2000) appliedCA for the evaluation of traditional wastewater treatments asell as possible alternatives in small towns (between 200 and2,600 p.e.). In both cases, the functional unit selected was thereatment of wastewater associated with 1 p.e./year. Both stud-es concluded that the impact associated with construction was

inor, relative to the associated operation. When different tech-ologies were compared, only one of the proposed alternativesas environmentally safer in all impact categories considered; in

he others, the recommended options depended on the categoriesf interest.

On the other hand, Karrman and Jonsson (2001) presentedmethod to standardize the environmental impact of four sys-

ems of wastewater treatment, quantifying their contribution tohe total impact of the Swedish society. The functional unitsed was wastewater treatment associated with 20,000 p.e./year.onstruction of the facilities was not considered. Wastewater

reatment was found to contribute with more than 10% of theotal impact regarding specific aspects such as discharge of nitro-en, cadmium, lead and mercury and heavy metal uptake inoil.

In Spain, the first approach to this subject was made by Vidalt al. (2002), who compared biologically active sludge treatmentn a WWTP of 22,167 p.e. with two configurations specificallyesigned for nitrogen removal (oxidation channel and Ludzack-ttinger). The functional unit was one ton of treated wastewater.utrophication was largely improved when the proposed alter-atives were considered. However, both of them implied largernergy requirements, which led to greater impact in the rest ofategories. Even so, the global impact related to these two con-gurations was smaller than that of activated sludge, with thexidation channel being the best alternative.

The aim of the study presented here is to assess differentlternatives for small WWTPs in order to establish which one ishe most adequate in terms of minimum environmental impact.he paper will be presented according to the four main phases

n LCA: goal and scope definition, inventory analysis, impactssessment and interpretation of the results.

. Goal and scope definition

During this phase, the purpose of the LCA study is definednd specifications on the LCA model and procedure are deter-ined (Baumann and Tillman, 2004).

.1. Objectives

Galicia (NW of Spain) is characterized by the existence of auge amount of low-populated locations; in fact, according tohe last census (year 2006) more of 93% of the municipalities

ave less than 20,000 inhabitants, stating for almost 50% of theopulation of the region (INE, 2007). This distribution madet the right place to evaluate the environmental performance of

WTPs in small villages (less than 20,000 p.e.).

n and Recycling 52 (2008) 931–940

The most relevant factors contributing to the overall envi-onmental impact will be identified. The analysis of differentxisting treatment technologies (biodenipho, aerobic–anoxicnd extended aeration) will help to establish which technologieshould be considered as the most adequate to implement fromn environmental point of view. The influence of the operationn the environmental impact will be also evaluated.

.2. Functional unit

The functional unit expresses the function of the product orervice under evaluation in quantitative terms and serves as basisor the calculations. It is the reference flow to which all otherows in the LCA model are referred. It also serves as a unitf comparison in comparative studies (Baumann and Tillman,004).

The main function of a WWTP is the treatment of an influentwith the objective of organic load, nutrient and suspended solidseduction) so as to reach satisfactory values before release inatural water courses.

In order to refer to the functional unit as the volume of thenfluent and its associated load, this factor was defined in terms.e., in the same line that other research works (Tillman et al.,998; Lundin et al., 2000; Karrman and Jonsson, 2001).

.3. System

Each WWTP was divided in 4 subsystems (Fig. 1): pre-reatment and primary treatment (subsystem 1), secondaryreatment (subsystem 2), sludge line (subsystem 3) and transportnd sludge use (subsystem 4). All subsystems include consump-ion of electricity and chemicals, that is to say, their productionnd transportation upstream, as well as the transport and treat-ent downstream of the sludge and other waste generated in theWTPs. Table 1 gives a detailed description of the subsystems

or each WWTP.From the technological point of view, the most remarkable

ifference is the type of technologies used for secondary treat-ent (Fig. 2), and to a lesser extend the type of units used for

ludge dehydratation (centrifuges or filter band). In particular,hree different secondary treatments are presented (Metcalf andddy, 2001):

Extended aeration: the reactor is equipped with mechanicalaeration and mixing devices. Screened wastewater enters thereactor and is combined with the return activated sludge. Thetank configuration and aeration and mixing devices provideunidirectional reactor flow, so that the energy used for aerationis sufficient to provide mixing in a system with a relativelylong hydraulic retention time.Biodenipho: influent wastewater is initially mixed with thereturn activated sludge in the anaerobic selector prior to beingdirected to two oxidation ditches placed in series. The oper-

ating sequence of the ditches and operation of the aerationand anoxic zones is varied. Submerged mixers are installedin the ditches so that for some operating phases, the basin isonly mixed and not aerated. The basin continues to receive

A. Gallego et al. / Resources, Conservation and Recycling 52 (2008) 931–940 933

gram

•

2

t2ct

teKwcttwst

3

TC

1

2

3

4

Fig. 1. Flow dia

wastewater and operates as a preanoxic zone. Nitrate is avail-able from a previous aerobic nitrification operation. Besidesdenitrification in the preanoxic zones, nitrate reduction is alsopossible during the aerobic operation depending on the dis-solved oxygen concentration level.Aerobic–anoxic: an aerobic tank where nitrification occurs(for example in the activated-sludge process) is combinedwith a mixed anoxic tank in order to accomplish nitrogenremoval.

.4. System boundaries

This term is defined as the interface between the process sys-

em and the environment or other product systems (ISO 14044,006). According to Lundin et al. (2000), very different choicesan be made for system boundaries in wastewater systems andhese choices will inevitably affect results.

sa

able 1onfiguration per subsystems of the facilities under study

W 1 W 2 W 3 W 4

Input of raw water√ √ √ √

Pre-treatment√ √ √ √

Primary treatment√ √ √ √

Discharge of partially treated waterTransportation and treatment of waste

√ √ √ √

Secondary treatment (extended aeration)√ √ √

Secondary treatment (biodenipho)√

Secondary treatment (aerobic–anoxic)Discharge of treated water

√ √ √ √

Thickening of secondary sludge√ √ √ √

Dewatering of sludge with centrifugation√ √ √ √

Dewatering of sludge with filter bandProduction and transport of chemicals

√ √ √ √

Storage of filter cake√ √ √ √

Transport of treated sludge to farms√ √ √ √

Application of sludge to land√ √ √ √

of the system.

On the one hand, the impact associated with the construc-ion phase of WWTP is lower than that corresponding to thexploitation phase (Tillman et al., 1998; Lundin et al., 2000;arrman and Jonsson, 2001). On the other hand, the impact ofastewater use and its pumping to the treatment plants were not

onsidered, as these actions are not affected by WWTP opera-ion. Therefore, the influent entering the WWTP was consideredhe starting point of the system. This approach is in agreementith the report by Lassaux et al. (2007), which indicates that the

tages before the tap contribute less to the environmental impacthan the stage afterwards.

. Inventory analysis

During this phase, the LCA model is built according topecifications determined in the goal and scope definition, datare collected and calculations carried out to indicate the envi-

W 5 W 6 W 7 W 8 W 9 W 10 W 11 W 12 W 13√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √

√√ √ √ √ √ √ √ √ √

√ √ √ √ √ √√

√ √√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √

√ √ √ √ √√ √ √ √√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √√ √ √ √ √ √ √ √ √

934 A. Gallego et al. / Resources, Conservatio

rT

3

wEeetw2oc

pTto

bc

3

toaE2

t

•

•

•

•

•

•

Fig. 2. Secondary treatments available at the WWTPs under study.

onmental load of the product/process/service (Baumann andillman, 2004).

.1. Data collection

Material and energy flows of the WWTPs (foreground data)ere supplied by the company in charge of their operation.xtraordinary conditions, such as flooding of the plant byxtremely intense rain or unexpected stoppage of units, werexcluded, as these situations were considered exceptional andherefore, did not represent normal operation. Inventory dataere obtained from 4 years (WWTP 1 and 11), 3 years (WWTP–7, 9 and 13) and 2 years (WWTP 8, 10 and 12). A selectionf the most important parameters subjected to analysis and theapacity (stated as p.e.) of each WWTP is presented in Table 2.

For example, to quantify the amount of electricity consumed

er functional unit, annual use was provided by the company.wo options were considered for distribution among the subsys-

ems: measurement of the electricity per unit and the number ofperating hours (WWTP 1, 5, 10, 11, and 12) or, when unfeasi-

•

n and Recycling 52 (2008) 931–940

le, application of a theoretical distribution model used by theompany (WWTP 2–4, 6–9 and 13).

.2. Data quality and simplifications

SimaPro 6.0 software (http://www.pre.nl/simapro) was usedo determine the environmental load of the subsystems. It is onef the most used software tools for LCA studies and includesgreat number of databases (Buwal 250, Ecoinvent v. 1.2,

TH-ESU 96 . . .) and methodologies (Eco-Indicator 99, CMLbaseline 2000, etc.).The databases available in SimaPro 6.0 were used to describe

he system as follows:

Electricity: for electricity production, the database used wasthe IDEMAT (2001) database, updated with the productionprofile for 2004 in Spain (IDAE, 2005).Treatment of solid waste: residues generated in all WWTPswere transported to an incineration plant with energy recov-ery. Since the IDEMAT (2001) database has no data for wasteincineration, data from Ecoinvent (1996) was considered dueto its high similarity to the solid waste coming from the gridremoval at the plants. These data were modified by addingthe recovered energy, specifically calculated according to thelocal conditions for 2005 (Consello de Contas de Galicia,2005).Chemicals: polyelectrolyte for sludge conditioning andhypochlorite for cleaning purposes, used as main chemicals,were referenced to acrylonitrile I (IDEMAT, 2001) and tosodium hypochlorite (Ecoinvent, 1996), respectively.Transport of solid wastes, chemicals and sludge: data from theIDEMAT (2001) database were chosen due to their similaritywith the load capacity of trucks.Fertilizers avoided: application of sludge to soil as fertilizerreduces the need for the use of synthetic fertilizers contain-ing nitrogen and phosphorous, which results in environmentalbenefit if the content of heavy metals content in the sludgeare kept within admissible values. The substitutability wasassumed to be 70% for phosphorus and 50% for nitrogen(Bengtsson et al., 1997). Entries from the IDEMAT (2001)database were used (N-based and P-based fertilizers) asgeneric fertilizers preferable represent the avoided fertilizer.

In addition, other data were obtained from the bibliography:

Bioavailability of metals in sludge: a study performed withsoils of the region was considered (Monterroso et al., 2003),where data regarding the percentage of metal extracted withAEDT were reported as a measurement of bioavailability. Thecorresponding value for Cu is 58.5%; for Pb, 7.8%; for Cr,4.1%; for Ni, 7.5% and for Zn, 18.0%. Due to the absenceof data, for other metals (i.e. Hg, Cd and Fe), the worst-casescenario was considered, with 100% bioavailability.

Methane and nitrogen compounds (N2, NOx and NH3) emis-sions: data from a variety of sources were used to calculateemissions resulting from sludge application for agriculturalpurposes (Hobson, 2000; Lundin et al., 2000; Mossier, 1993).

A.G

allegoetal./R

esources,Conservation

andR

ecycling52

(2008)931–940

935

Table 2Relevant parameters of the inventory data and capacity (stated as population equivalents) from 2000–2003 (WWTP 1 and 11), 2001–2003 (WWTP 2–7, 9 and 13) and 2002–2003 (WWTP 8, 10 and 12)

Parameter WWTP 1 WWTP 2 WWTP 3 WWTP 4 WWTP 5 WWTP 6 WWTP 7 WWTP 8 WWTP 9 WWTP 10 WWTP 11 WWTP 12 WWTP 13

Population equivalents 17,069 11,372 11,357 18,809 15,574 3170 6959 9349 7720 1659 2396 3083 4248

InputsIncoming water

kg COD 49.61 43.67 60.91 66.48 135.24 89.17 50.45 36.03 81.01 54.01 52.74 40.55 51.80kg P-PO4

3− 0.39 0.76 0.16 0.35 0.44 0.32 1.13 0.83 0.66 0.51 0.39 0.78 0.82kg N-NH4

+ 4.80 4.90 1.36 2.73 3.13 3.15 4.06 4.79 4.43 3.40 4.17 2.63 4.33

From background systemkWh 64.79 44.19 35.23 29.29 28.48 105.99 101.70 32.62 43.49 63.59 64.71 112.07 118.84kWh (aeration) 26.80 40.90 33.08 19.92 19.94 74.19 60.10 27.72 26.34 33.11 52.85 72.76 108.0

OutputsDischarges to soil

kg dry sludge 15.51 15.81 18.2 7.60 2.82 13.40 3.89 8.33 8.26 19.82 23.91 6.95 16.53Cr (mg) 113.49 70.73 168.39 65.57 21.56 72.69 18.14 38.48 35.60 94.13 120.75 44.78 40.17Hg (mg) 14.51 24.30 15.56 7.15 4.06 10.58 7.20 4.06 1.80 18.73 12.83 8.12 43.87Cu (g) 2.16 4.89 3.19 0.79 1.17 12.43 2.46 1.20 1.30 3.37 4.61 2.26 3.83Cd (mg) 23.3 14.8 52.7 10.6 7.20 83.27 3.90 12.50 4.50 39.60 47.80 24.30 25.10Ni (mg) 59.11 26.7 136.50 36.60 15.70 97.72 6.80 16.80 20.30 35.0 71.90 25.60 18.90Pb (mg) 152.07 254.2 139.80 37.80 54.30 193.71 45.10 72.90 50.70 88.10 111.90 25.50 137.30Zn (g) 1.66 2.67 2.64 1.00 0.55 4.22 0.33 0.69 0.48 1.89 1.16 0.37 1.20

Discharges to waterkg COD 3.96 6.48 2.60 2.73 5.66 23.77 9.05 4.93 10.95 4.50 2.86 7.61 4.98COD reduction (%) 92.0 85.2 95.7 95.9 95.8 73.4 82.1 86.3 86.5 90.0 94.6 81.2 90.4kg P-PO4

3− 0.17 0.47 0.16 0.27 0.10 0.49 1.29 0.86 0.27 0.13 0.23 0.80 0.63P-PO4

3− reduction (%) 57.1 38.7 0.5 22.3 76.1 0 0 0 58.6 75.1 40.4 0 23.1kg N-NH4

+ 0.59 2.66 0.01 1.70 1.01 2.72 1.14 0.68 1.70 0.26 2.26 2.90 1.10N-NH4

+ reduction (%) 87.7 45.8 99.8 37.5 67.8 13.7 71.8 85.8 61.5 92.3 45.9 0 74.6

Emissions to airCH4 (g) 305.3 309.7 368.3 133.6 137.6 268.8 77.4 161.5 166.3 395.3 450.5 136.5 239.9N2O (g) 8.2 8.7 8.2 5.2 4.0 4.3 1.6 3.8 5.0 14.0 15.2 4.4 8.5NH3 (g) 204.8 217.6 217.0 107.3 99.6 107.3 50.1 147.4 126.2 349.3 380.9 111.1 198.1

To further treatmentkg solid waste 29.40 3.97 2.20 9.69 27.38 31.12 9.45 8.76 27.38 27.37 27.37 24.08 3.91

All data are presented per p.e. year.

936 A. Gallego et al. / Resources, Conservation and Recycling 52 (2008) 931–940

Table 3Normalization results per p.e. year (*1010) using normalization factors for West Europe (year 1995)

Parameter WWTP 1 WWTP 2 WWTP 3 WWTP 4 WWTP 5 WWTP 6 WWTP 7

Eutrophication 0.90 2.25 0.65 1.32 0.85 2.56 4.07Terrestrial ecotoxicity 0.37 0.44 0.47 0.19 0.09 0.28 0.11Acidification 0.19 0.17 0.17 0.10 0.09 0.18 0.14Abiotic depletion 0.17 0.11 0.09 0.08 0.07 0.27 0.27Global warming 0.09 0.07 0.06 0.05 0.05 0.16 0.13Photochemical oxidation 0.011 0.008 0.007 0.006 0.005 0.017 0.015Ozone layer depletion 0.001 0.001 0.001 0.001 0.001 0.001 0.001

Parameter WWTP 8 WWTP 9 WWTP 10 WWTP 11 WWTP 12 WWTP 13

Eutrophication 2.58 1.65 0.66 1.56 3.17 2.14Terrestrial ecotoxicity 0.14 0.17 0.36 0.32 0.22 0.65Acidification 0.12 0.12 0.27 0.29 0.19 0.25Abiotic depletion 0.08 0.11 0.17 0.17 0.28 0.32G 0.1P 0.0O 0.0

4

a(oUttdtcn2wat

ftm

a

5

adT

soJrTbr

o2

Ttitiio(

odcoitFnfttcipally CFC-11, CFC-12 and Halon 1301) is minimal, mainlyassociated with the production of chemicals and electricity withoil. In the case of acidification, the impact is mainly produced by

Table 4Relative contribution (%) of electric consumption to all the impact categories

Impact category Average Standard deviation

Abiotic depletion 96.4 1.0Photochemical oxidation 79.9 8.6Global warming 74.5 7.3

lobal warming 0.05 0.06hotochemical oxidation 0.005 0.007zone layer depletion 0.001 0.001

. Impact assessment

This phase of LCA consists of the assessment and evalu-tion of the environmental impact of the system under studyBaumann and Tillman, 2004). Among the available method-logies, the CML 2 baseline 2000 developed by Leidenniversity was used (Guinee et al., 2001) and in particular,

he following impact categories were considered: abiotic deple-ion, global warming, acidification, eutrophication, ozone layerepletion and photochemical oxidation. Although the charac-erization models for the definition of toxicological impactategories are still lacking consensus among the LCA commu-ity (Aboussouan et al., 2004; Guinee et al., 2004; Larsen et al.,004; Pennington, 2001), they are important when dealing withater systems (Lundie et al., 2004). In particular, and taking into

ccount the data available (heavy metals content in the sludge),errestrial ecotoxicity impact category (TET) was considered.

West Europe values were selected as normalization factorsor all impact categories. The reference information used referso data from 1995, this being, to the best of our knowledge, the

ost recent list available (Huijbregts et al., 2003).The normalization data (Table 3) are dimensionless, hence

llowing for the comparison between impact categories.

. Interpretation of results

In this phase of an LCA study, results from the inventorynalysis and/or impact assessment are evaluated in order toraw conclusions and make recommendations (Baumann andillman, 2004).

Table 3 shows that eutrophication is the category with greaterignificance in all WWTPs. This result agrees with other LCAsn wastewater treatment (Roeleveld et al., 1997; Karrman andonsson, 2001; Vidal et al., 2002; Lassaux et al., 2007). The ter-

estrial ecotoxicity is the second impact category in importance.he important significance of the terrestrial toxicity producedy heavy metals in the sludge is well documented in the bibliog-aphy, and concretely, has been studied in detail in other LCAs

OATE

0 0.10 0.16 0.1512 0.012 0.017 0.01902 0.001 0.001 0.002

f WWTPs (Hospido et al., 2004, 2005; Karrman and Jonsson,001).

The results were also interpreted by a dominance analysis.his analysis is normally used to investigate which parts of

he life cycle give rise to the greatest (dominant) environmentalmpact (Baumann and Tillman, 2004). In this study, it was usedo identify the element of the inventory that contributes mostn the different impact categories. The consumption of electric-ty was pointed out as the dominant element, being the elementf higher impact in four of the impact categories under studyTable 4).

If all the impact categories are analyzed in detail, it can bebserved that in the case of abiotic depletion, the electricity pro-uction is the main process of impact because of the extraction ofarbon, gas natural and coal as raw materials. For photochemicalxidation, the emissions associated to the production of electric-ty are also the major contributors, having a lower importancehose produced by the use of sludge for agricultural purposes.or climate change, CO2 emissions from the combustion of coal,atural gas and oil are again the main contributors. The emissionsrom the sludge applied (CH4 and N2O) and from the incinera-ion of waste (CO2, CH4 and N2O) have also a certain share tohis impact. Emission of gases that reduce the ozone layer (prin-

zone layer depletion 63.5 15.0cidification 44.0 17.6errestrial ecotoxicity 12.5 9.7utrophication 0.4 0.2

A. Gallego et al. / Resources, Conservation and Recycling 52 (2008) 931–940 937

, P-P

tp

cw

5

cctaPwt3

om

Wr

•

•

•

5

stasludge at the WWTPs depends on two opposite aspects: theimplementation of secondary treatment and the existence ofanaerobic digestion (Hospido et al., 2008). If the former occurs,higher production takes place. If the latter happens, a great

Table 5Input and output of P-PO4

3− (kg of P-PO43/p.e. year)

Plant kg P-PO43− in (average

± standard deviation)kg P-PO4

3− out (average± standard deviation)

WWTP 1 0.39 ± 1.0 0.17 ± 0.04WWTP 2 0.76 ± 0.14 0.47 ± 0.11WWTP 3 0.16 ± 0.08 0.16 ± 0.08WWTP 4 0.35 ± 0.1 0.27 ± 0.08WWTP 5 0.44 ± 0.23 0.10 ± 0.05WWTP 6 0.32 ± 0.31 0.49 ± 0.48WWTP 7 1.13 ± 0.56 1.29 ± 0.70WWTP 8 0.83 ± 0.77 0.86 ± 0.28WWTP 9 0.66 ± 0.15 0.27 ± 0.15

Fig. 3. Relative contribution of N-NH4+

he air emissions of NH3 when the sludge is used for agriculturalurposes.

Due to their importance, eutrophication, terrestrial eutrophi-ation and electricity use (that controls the remaining categories)ere analyzed in more detail.

.1. Eutrophication

In this category, the main impact was attributed to the dis-harge of the treated effluent. To interpret this category, aontribution analysis was made to identify which environmen-al loads contribute most to the environmental impact (Baumannnd Tillman, 2004). As can be seen in Fig. 3, the discharges of-PO4

3− and N-NH4+ and, to a lower extent, COD in the treated

ater were the most relevant factors contributing to eutrophica-ion (with a minimal contribution of almost 64%, for WWTP).

It is important to note that reduction in the eutrophying loadf the effluent was attained in all WWTPs such that the effluenteets the standards defined by existing regulations.An interesting aspect to analyze is the question of which

WTPs can provide higher levels of treatment, in terms of theeduction of COD, P-PO4

3− and N-NH4+:

Table 2 shows that regardless of the technology considered,treatment percentages for COD were greater than 73% in allWWTPs and often exceeded 90%.The negative values reported at WWTPs 6–8 and 12 for theremoval of P-PO4

3− can probably be attributed to the lack ofmeasurements and to the fact that concentration values werevery close to the detection limit of the equipment (Table 5).In those cases, it was assumed that no P-PO4

3− removaloccurs. For the WWTPs with biodenipho treatment (WWTP

1 and 10), the percentages of reduction were 57.1 and 75.1%,respectively (Table 2). The WWTPs with aerobic–anoxictreatment (WWTP 5 and 9) resulted in percentages of reduc-tion of 76.1 and 58.6%, respectively (Table 2).

WWWW

O43−, COD releases to eutrophication.

In the case of N-NH4+, biodenipho and aerobic–anoxic treat-

ments were the most efficient technologies, with valuesbetween 61.5 and 92.3% (Table 2). WWTPs with extendedaeration obtained results with important disparity, varyingfrom 10.3 to 99.8% (Table 2). WWTPs 6 and 12 had theworst treatment efficiencies for N-NH4

+, COD and P-PO43−.

In this case, the poor environmental performance is likely todepend not only on the technology but also on the operationof the technology.

.2. Terrestrial ecotoxicity

Terrestrial ecotoxicity is dominated by the application ofludge with heavy metals on agricultural land, being Hg and Crhe main contributors accounting, on average, for 51.0 ± 13.0%nd 31.2 ± 11.0% of the impact, respectively. The amount of

WTP 10 0.51 ± 0.16 0.13 ± 0.05WTP 11 0.39 ± 0.44 0.23 ± 0.13WTP 12 0.78 ± 0.64 0.80 ± 0.63WTP 13 0.82 ± 0.82 0.63 ± 0.27

9 vation and Recycling 52 (2008) 931–940

rsa

tgu(

5

oitiweaoGicot

atIafviordIlr

mtnts2

otoibwfo7A

ofhe

avy

met

als

insl

udge

used

for

agri

cultu

ralp

urpo

ses

(mg/

kgdr

ym

atte

r)

itva

lues

aL

imit

valu

esb

W1

W2

W3

W4

W5

W6

W7

W8

W9

W10

W11

W12

W13

6–25

100.

91.

640.

941.

030.

570.

790.

600.

650.

940.

950.

541.

173.

490–

1750

1000

9.50

62.0

123

123

39.0

70.3

60.5

60.0

56.0

47.5

50.5

64.5

41.5

0–17

5010

0013

956

430

819

527

827

499

024

726

929

133

055

652

10–

4010

1.5

1.0

3.0

1.5

1.0

1.7

1.0

1.5

1.7

2.0

2.0

3.5

2.0

0–40

030

03.

824

.010

270

.329

.082

.523

.326

.832

.723

.540

.049

.020

.00–

1200

750

9.8

220

101

70.0

97.0

93.0

148

112

78.7

57.0

60.0

47.0

140

0–40

0025

0010

710

0082

780

442

917

4747

646

332

353

127

129

953

0

eses

tabl

ishe

dby

the

Eur

opea

nD

irec

tive

(Eur

opea

nE

cono

mic

Com

mun

ity,1

986)

.es

prop

osed

byth

eE

urop

ean

Com

mis

sion

(Eur

opea

nC

omm

issi

on,2

000)

.

38 A. Gallego et al. / Resources, Conser

eduction of sludge is achieved. All the WWTPs studied haveecondary treatment and none anaerobic digestion, and thereforell have similar values of production of sludge (Table 2).

From a general point of view, the level of heavy metals inhe sludge in the 13 WWTPs indicate that this impact is not areat threat because they are far under the more demanding val-es of the legislation or proposed by the European CommissionTable 6).

.3. Electricity use

As mentioned, electricity consumption is a key element in theverall environmental performance of a WWTP. If we were tognore the emissions associated to the processes that are requiredo produce the electricity we consume, we would then be ignor-ng around 30% of CO2 emissions and the associated globalarming impact (Azapagic et al., 2004). On the other hand, the

nergy consumption is a major contributor to the operationalnd maintenance costs of a WWTP, representing 50% of theperational costs of treatment plants of less than 10,000 p.e. inreece (Tsagarakis et al., 2000, 2003). Therefore, its reduction

s not only an environmental challenge but also an economicalhallenge. In fact, energy use has been already identified as onef the major sustainable development indicators for wastewaterreatment systems (Palme et al., 2005).

In this study, the distribution per subsystem was considerednd subsystem 2 represented an average consumption of 69% ofhe total amount, while subsystem 1 and 3 stood for 14 and 17%.n particular, Table 2 shows the electricity use associated to theeration systems within subsystem 2 for all the WWTPs. Fiveacilities (WWTPs 6–7 and 11–13) stand out because of theirery high rates of consumption (73.58 ± 21.20 kWh/p.e.), whichs clear when compared to other facilities with the same technol-gy (extended aeration), WWTPs 2–4 and 8, with much lowerates of consumption (29.1 ± 11.0 kWh/p.e.). Even more, theseifferences do not entail better treatment efficiencies (Table 2).n this sense, Tsagarakis et al. (2003) propose the use of on-ine control of the aeration systems as an effective way to attainemarkable savings in the electrical consumption.

The use of filter band or centrifuge in the sludge treat-ent was another of the main technological differences between

he WWTPs (Table 1). The use of one system or another didot influence the environmental impact of the WWTP, sincehe contribution to the global electrical consumption was verymall, with similar values: 3.91 ± 2.20% (centrifuges), and.86 ± 0.47% (filter band).

In the case of electricity use, the electricity production profilef Spain was considered. This profile is clearly dominated byhe use of fossil fuels, so a sensitive analysis was performed inrder to check the influence of the electricity production profilen the results (Table 7). Considering the plant more influencedy electricity consumption (WWTP 13), normalization resultsere calculated using the Portuguese profile (principally fossil

uels production: 80.8%), Spanish (used in this study, 57.6%f fossil fuel production), French (mainly nuclear production:2.9%) and Norwegian (mostly renewable production: 99.2%).s it can be observed in Table 7, the production by different Ta

ble

6C

once

ntra

tion

Lim

Hg

1C

r10

0C

u10

0C

d2

Ni

30Pb

75Z

n25

0

aL

imit

valu

bL

imit

valu

A. Gallego et al. / Resources, Conservation and Recycling 52 (2008) 931–940 939

Table 7Relative distribution of the production of energy by sources for Portugal, Spain, France and Norway and normalization values for WWTP 13 with these distributions

Country Production of energy by sourcesa Normalization values for WWTP 13b

Fossil (%) Nuclear (%) Renewable (%) EU TE AC AD GW PO OD

Portugal 80.8 2.6 16.6 2.14 0.67 0.44 0.45 0.22 0.04 0.006Spain 57.6 23 19.4 2.14 0.65 0.25 0.32 0.16 0.02 0.002France 11.4 72.9 15.7 2.12 0.61 0.15 0.04 0.05 0.01 0.001Norway 0.5 0.3 99.2 2.12 0.60 0.12 0.004 0.02 0.002 0.001

tic ded

sigetp

6

6

opo5sgttiafiv

6

atacLao

dceoweM

tl

•

•

7

1c

•

•

•

•

•

a IDAE (2005) and Buwal 250 (1996).b EU, eutrophication; TE, terrestrial ecotoxicity; AC, acidification; AD, abioepletion.

ources in each country can affect the absolute value of somempact categories (especially acidification, abiotic depletion andlobal warming) but the general environmental performance ofach technology will be very similar because the relative impor-ance of eutrophication and terrestrial ecotoxicity will remainractically equal.

. Discussion

.1. Influence of the operation

One of the main objectives of this research work was tobserve the influence of the operation on the global impact of thelants. In the case of the plants that used the biodenipho technol-gy (WWTPs 1 and 10) and aerobic–anoxic treatment (WWTPsand 9), the influence of the operation was not significant, with

imilar values obtained for the consumption, purification andlobal environmental impact (Tables 2, 3 and 5). However, inhe case of those using an extended aeration, the operation ofhe plants clearly influenced their environmental impact. Thiss illustrated by WWTP 11 and WWTP 12, both with extendederation but WWTP 11 consuming much less electricity perunctional unit (29.29 kWh versus 112.1 kWh), while maintain-ng greater removal of COD (96% versus 81%), N-NH4

+ (22%ersus 0%) and P-PO4

3− (38% versus 0%).

.2. Applicability of the results

The technologies described in this study are worldwidepplied in order to purify wastewaters from villages, however,he usefulness of the results here presented can be questioned asll the facilities analyzed are located in a specific region (Gali-ia, NW Spain). To the best of our knowledge, this is the firstCA applied to biodenipho, aerobic–anoxic and extended aer-tion technologies, so no specific comparison with results fromther locations is possible.

According to our findings, three are the main factors thateterminate the environmental performance of the WWTPs: dis-harges of COD, P-PO4

3− and N-NH4+, disposal of sludge and

lectricity use. These factors are likely to be the main focus

f the environmental performance study in other parts of theorld and in fact they have already been reported as the main

nvironmental sustainability indicators for WWTPs (Lundin andorrison, 2002). However, the extrapolation of the results needs

pletion; GW, global warming; PO, photochemical oxidation; OD, ozone layer

o be carefully considered as it will be affected by the followingocal conditions:

Wastewater quality, especially the amount of nutrients(eutrophication).Final disposal of the sludge. There is a wide literature on thistopic; for instance, Hospido et al. (2004) concluded that thetreatment of sewage through thermal processes (incinerationor pyrolysis) entails important reductions in terrestrial eco-toxicity compared with land application of digested sludge,although it has less beneficial results for eutrophication and,to a lesser extend, for global warming and acidification.

. Conclusions

To fulfill the objectives, a detailed environmental analysis of3 WWTPs of less than 20,000 p.e. was carried out. The mainonclusions achieved are:

Eutrophication, and to a lesser extend terrestrial ecotoxicitywere identified as the main impact categories in the WWTPsanalyzed.Electricity use, emission of nutrients to the water and appli-cation of sludge for agricultural purposes were identified asthe main contributors to the global impact of all the facilitiesstudied.Regarding the comparison of technologies, biodenipho(WWTPs 1 and 10) and aerobic–anoxic treatments (WWTPs5 and 9) were found to be the less damaging options forsecondary treatment, as they attained higher N-NH4

+ and P-PO4

3− removal efficiencies than extended aeration (WWTPs2–4, 6–8 and 11–13) by using low-middle values of electricityand producing similar amounts of sludge.The comparison of different sludge dewatering units (filterband, centrifuges) did not find relevant differences from theenvironmental point of view.The degree of influence of WWTP operation on the globalimpact was observed; in the case of the plants that used the bio-denipho technology (WWTPs 1 and 10) and aerobic–anoxictreatment (WWTPs 5 and 9), the influence of the operation of

the plants was not significant; similar values of consumption,purification and environmental impact were obtained. In thecase of those using extended aeration (WWTPs 2–4, 6–8 and11–13), operation influenced the environmental performance,

9 vatio

•

A

eSNla

R

A

A

B

B

B

C

E

E

E

G

G

H

H

H

H

H

I

I

I

I

K

L

L

L

L

L

M

M

M

P

P

R

T

T

Ttion of municipal wastewater treatment systems. Water Air Soil Pollut

40 A. Gallego et al. / Resources, Conser

with variations in electricity consumption and purificationlevels affecting the associated environmental impacts.In order to extrapolate the results obtained to other locations,special attention should be paid to wastewater quality (nutri-ents) and final disposal of the sludge. If the study is focused inglobal warming, abiotic depletion or acidification, attentionmust be paid to the electricity production profile.

cknowledgements

This work has been partially financed by the Galician Gov-rnment (Project Reference-PGIDIT04TAL269003PR) and thepanish Ministry of Education and Science (Project Reference-OVEDAR-Consolider CSD 2007-00055). A. Gallego would

ike to express his gratitute to the Galician Ministry of Innovationnd Industry for its financial support.

eferences

boussouan L, van de Meent D, Schonnenbeck M, Hauschild M, Delbeke K,Struijs J, et al. Declaration of Apeldoorn on LCIA of non-ferrous metals.The Netherlands: Apeldoorn; 2004.

zapagic A, Perdan S, Clift R. Sustainable development in practice. In: Casestudies for engineers and scientists. West Sussex, England: John Wiley &Sons, Ltd; 2004.

aumann H, Tillman AM. The Hitch Hiker’s Guide to LCA. An orientationin life cycle assessment methodology and application. Lund, Sweden: Stu-dentlitteratur; 2004.

engtsson M, Lundin M, Molander S. Life cycle assessment of wastewatersystems: case studies of conventional treatment, urine sorting and liquid com-posting in three Swedish municipalities. Report 1997:9. Chalmers Universityof Technology, Gothenburg, Sweden; 1997.

uwal 250. Data Base. Swiss Federal Environmental Protection Agency. Bern,Switzerland; 1996.

onsello de Contas de Galicia. Informe de fiscalizacion del la Sociedade Galegade Medio Ambiente S.A. Santiago de Compostela, Spain; 2005 [in Spanish].

coinvent. DataBase. The Swiss Centre for Life Cycle Inventories. Zurich,Switzerland; 1996.

uropean Commission. Working documents on sludge 3rd draft:http://ec.europa.eu/environment/waste/sludge/pdf/sludge en.pdf; 2000.

uropean Economic Community. Council Directive 86/278/EEC of 12 June1986 on the protection of the environment, and in particular of the soil,when sewage sludge is used in agriculture. Official Journal of the EuropeanCommunities, L 181/1986; 1986.

uinee J, Koning A, Pennington D, Rosenbaum R, Hauschild M, Olsen S, etal. Bringing science and pragmatism together—a tiered approach for mod-eling toxicological impacts in LCA. Int J Life Cycle Assess 2004;9(5):320–6.

uinee J, Gorree M, Heijungs R, Huppes G, Kleijn R, de Konnig A, et al. Lifecycle assessment: an operational guide to the ISO standards. The Nether-lands: Ministry of Housing, Spatial Planning and Environment; 2001.

obson J. CH4 and N2O emissions from waste water handling. In: Good practiceguidance and uncertainty management in National Greenhouse Gas Inven-tories. Geneve, Switzerland: Intergovernmental Panel on Climate Change(IPCC) Publications; 2000.

ospido A, Moreira MT, Feijoo G. A comparison of municipal wastewatertreatment plants for big centres of population in Galicia (Spain). Int J LifeCycle Assess 2008;13(1):57–64.

ospido A, Moreira MT, Martın M, Rigola M, Feijoo G. Environmental eval-uation of different treatment processes for sludge from urban wastewater

V

n and Recycling 52 (2008) 931–940

treatments: anaerobic digestion versus thermal processes. Int J Life CycleAssess 2005;10(5):336–45.

ospido A, Moreira MT, Fernandez-Couto M, Feijoo G. Environmental perfor-mance of a municipal wastewater treatment plant. Int J Life Cycle Assess2004;9(4):261–71.

uijbregts MAJ, Breedveld L, Huppes G, de Koning A, Van Oers L, Suh S.Normalisation figures for environmental life-cycle assessment: The Nether-lands (1997/1998), Western Europe (1995) and the world (1990 and 1995).J Clean Prod 2003;11:737–48.

DAE. Instituto para la Diversificacion y Ahorro de la Energıa. Boletın 7: Efi-ciencia energetica y energıas renovables. Madrid, Spain; 2005 [in Spanish].

DEMAT. Inventory Data of Materials. Faculty of Design, Engineering andProduction, Delft University of Technology, Delft, the Netherlands; 2001.

NE. Instituto Nacional de Estadıstica (Nacional Institute ofStadistics). Population figures and Demographic Censuses:http://www.ine.es/inebmenu/mnu cifraspob.htm; 2007.

SO, International, Organization y Standardization. ISO 14044. Gestion ambi-ental. Analisis del Ciclo de Vida. Requisitos y directrices. Madrid, Spain;2006 [in Spanish].

arrman E, Jonsson H. Normalising impacts in an environmental systems anal-ysis of wastewater systems. Water Sci Technol 2001;43(5):203–300.

arsen HF, Birkved M, Hauschild M, Pennington D, Guinee J. Evaluation ofselection methods for toxicological impacts in LCA—recommendations forOMNIITOX. Int J Life Cycle Assess 2004;9(5):307–19.

assaux S, Renzoni R, Germain A. Life cycle assessment of water: from thepumping station to the wastewater treatment plant. Int J Life Cycle Assess2007;12(2):118–26.

undie S, Peteres GM, Beavis PC. Life cycle assessment for sustainablemetropolitan water systems planning. Environ Sci Technol 2004;38:3465–73.

undin M, Bengtsson M, Molander S. Life cycle assessment of wastewater sys-tems: influence of system boundaries and scale on calculated environmentalloads. Environ Sci Technol 2000;34(1):180–6.

undin M, Morrison GM. A life cycle assessment based procedure for devel-opment of environmental sustainability indicators for urban water systems.Urban Water 2002;4:145–52.

etcalf and Eddy Inc. Wastewater Engineering. Treatment and Reuse. 4th edi-tion revised by G. Tchobanoglous, F.L. Burton, H.D. Stensel. McGraw-Hill,New York, USA; 2001.

onterroso C, Kidd P, Macias F. Bioavailability of heavy metals in sewagesludge-amended soils as affected by soil parent material. In: 7th InternationalConference on the Biogeochemistry of Trace Elements; 2003.

ossier AR. Nitrous oxide emissions from agricultural soils. In: van Amstel AR,editor. IPCC Workshop Proceedings: Methane and Nitrous Oxide, Methodsin National Emissions Inventories and Options for Control; 1993.

alme U, Lundin M, Tillman AM, Molander S. Sustainable development indica-tors for wastewater systems-researchers and indicator users in a co-operativecase study. Resour Conserv Recycl 2005;43(3):293–311.

ennington DW. Current issues in the characterisation of toxicological impacts.Int J Life Cycle Assess 2001;6(2):89–95.

oeleveld PJ, Klapwijk A, Eggels PG, Rulkens WH, van Starkenburg W.Sustainability of municipal wastewater treatment. Water Sci Technol1997;35(10):221–8.

illman AM, Svingby M, Lundstrom H. Life cycle assessment of municipalwaste water systems. Int J Life Cycle Assess 1998;3(3):145–57.

sagarakis KP, Mara DD, Horan NJ, Angelakis AN. Small municipal wastewatertreatment plants in Greece. Water Sci Technol 2000;41(1):41–8.

sagarakis KP, Mara DD, Angelakis AN. Application of cost criteria for selec-

2003;142(1–4):187–210.idal N, Poch M, Martı E, Rodrıguez-Roda I. Evaluation on the environmental

implications to include structural changes in a wastewater treatment plant. JChem Technol Biotechnol 2002;77:1206–11.