Assessment of groundwater recharge in urban areas

ADVERTIMENT . La consul a d’aques a esi queda condicionada a l’accep ació de les següen s condicions d'ús: La di usió d’aques a esi pe mi jà del se ei TDX (www. esisenxa xa.ne ) ha es a au o i zada pels i ula s dels d e s de p opie a in el·lec ual únicamen pe a usos p i a s emma ca s en ac i i a s d’in es igació i docència. No s’au o i za la se a ep oducció amb inali a s de luc e ni la se a di usió i posada a disposició des d’un lloc aliè al se ei TDX. No s’au o i za la p esen ació del seu con ingu en una ines a o ma c aliè a TDX ( aming). Aques a ese a de d e s a ec a an al esum de p esen ació de la esi com als seus con ingu s. En la u ili zació o ci a de pa s de la esi és obliga indica el nom de la pe sona au o a. ADVERTENCIA. La consul a de es a esis queda condicionada a la acep ación de las siguien es condiciones de uso: La di usión de es a esis po medio del se icio TDR (www. esisen ed.ne ) ha sido au o izada po los i ula es de los de echos de p opiedad in elec ual únicamen e pa a usos p i ados enma cados en ac i idades de in es igación y docencia. No se au o iza su ep oducción con inalidades de luc o ni su di usión y pues a a disposición desde un si io ajeno al se icio TDR. No se au o iza la p esen ación de su con enido en una en ana o ma co ajeno a TDR ( aming). Es a ese a de de echos a ec a an o al esumen de p esen ación de la esis como a sus con enidos. En la u ilización o ci a de pa es de la esis es obligado indica el nomb e de la pe sona au o a. WARNING. On ha ing consul ed his hesis you’ e accep ing he ollowing use condi ions: Sp eading his hesis by he TDX (www. esisenxa xa.ne ) se ice has been au ho ized by he i ula o he in ellec ual p ope y igh s only o p i a e uses placed in in es iga ion and eaching ac i i ies. Rep oduc ion wi h luc a i e aims is no au ho ized nei he i s sp eading and a ailabili y om a si e o eign o he TDX se ice. In oducing i s con en in a window o ame o eign o he TDX se ice is no au ho ized ( aming). This igh s a ec o he p esen a ion summa y o he hesis as well as o i s con en s. In he using o ci a ion o pa s o he hesis i ’s obliged o indica e he name o he au ho PhD Thesis Assessmen o g oundwa e echa ge in u ban a eas Hyd ogeology G oup (GHS) Ins i u e o En i onmen al Assessmen and Wa e Resea ch (IDAEA) Spanish Resea ch Council (CSIC) Depa men o Geo echnical Enginee ing and Geosciences, Uni e si a Poli ècnica de Ca alunya (UPC) Ba celonaTech Au ho : Isabel Tubau Fe nández Ad iso s: D . En ic Vázquez Suñé D . Jesús Ca e a No embe , 2015 The au ho wish o acknowledge he inancial suppo o he Minis y o Educa ion and Science o Spain, he Technical Uni e si y o Ca alonian (UPC), he Ca alan wa e au ho i y (ACA), he company o d inking wa e Te -Llob ega (ATLL) and he Ba celona me opoli an company o hyd aulic se ices and was e ea men (EMSHTR). Thanks a e o CLABSA, AGBAR, CETAQUA, ci y council o Ba celona, San Ad ià del Besòs, Badalona and San a Coloma de G amane , EUSAB (Es acionamen s U bans de San Ad ià del Besòs) and o companies and indi idual owne s o wells sampled in his s udy. i I. Abs ac In he ci y o Ba celona and he ownships si ua ed in he del aic a ea o he Besòs Ri e , g oundwa e is ecognized as an aspec o u ban wa e cycle managemen . The e is g oundwa e seepage in o in as uc u e and buildings ha in u n equi e d ainage, a pa o which is in oduced in o sewage sys ems. Sensi ized by his issue, ci y councils ha e had hyd ologic s udies ca ied ou o analyze i s causes and o plan al e na i e ways o using hese esou ces while educing ea men cos s, and minimizing he impac s ha g oundwa e and d ainage ha e on in as uc u e and ice e sa. Based on hese s udies, and s a ing in 2000 some o his has been used o municipal p ac ices (ga den i iga ion and s ee cleaning). Subsequen ly, due o a d ough ha a ec ed he egion be ween Ap il 2007 and Janua y 2009, he go e nmen en i y in cha ge o Ca alonian wa e sou ces (ACA) aims o ge g ea e use ou o u ban g oundwa e in his a ea. This equi es a comp ehensi e unde s anding o he hyd ogeological cha ac e is ics o he en i onmen , he a ailabili y and quali y o hese esou ces, and he app op ia e ools o p ope assessmen and managemen . This has been he amewo k o his hesis. Recha ge is one o he mos impo an aspec s o wa e balance in aqui e sys ems in u ban a eas. U baniza ion o he a ea al e s he na u al hyd ological cycle and di ec wa e echa ge in o aqui e s by in il a ion is educed. In u n, e apo anspi a ion dec eases and new echa ge sou ces appea . This hesis p esen s, i s , a me hodology ha allows o he quan i ying o a iabili y in space and ime o he echa ge in u ban a eas. Po en ial sou ces o echa ge ha ha e been conside ed a e: (1) di ec in il a ion om ain and u ban uno , (2) losses om he sewe sys em, (3) losses om he wa e supply sys em, and (4) o he speci ic sou ces o echa ge (i.e. i e in il a ion, seawa e in usion, e c.). Recha ge calcula ions a e ini ially pe o med by applying analy ical equa ions unde a ious hypo heses. These esul s ha e been e alua ed in he hyd ogeological con ex h ough a nume ical model o low and anspo in he whole a ea o s udy. Secondly, and also h oughou he s udy a ea, de ec ion o some subs ances in g oundwa e has been analyzed. These subs ances and hei deg ada ion p oduc s, po en ially p esen in echa ging sou ces associa ed wi h u ban en i onmen s, belong o so-called 'eme ging o ganic pollu an s'. These a e o ganic chemical compounds used in he o mula ion o daily p oduc s o widesp ead use, such as pha maceu icals, emulsions, ca e p oduc s and pe sonal hygiene, household and indus ial de e gen s, plas ics, pes icides and he bicides, among o he s. The e is ela i ely li le known abou hese subs ances’ in oduc ion o and e olu ion in he aqui e s. O ganic pollu an s whose e olu ion and beha io in he g oundwa e is be e unde s ood can be deg aded by na u al p ocesses con olled mainly by oxida ion- educ ion p ocesses. E en in si u s imula ion echniques ha e been de eloped which a e able o accele a e hese p ocesses o na u al a enua ion o con amina ed en i onmen s. Thi dly, in a p elimina y way o a enua ion p ocesses de ec ion, a me hodology is p esen ed using an end-membe mixing analysis (EMMA) and MIX o compu e mixing a ios and iden i y hyd ochemical eac ions. The me hodology consis s o (1) iden i ying he po en ial sou ces o echa ge, (2) cha ac e ising echa ge sou ces and mixed wa e samples using hyd ogeochemis y, (3) selec ing chemical species o be used in he analysis and (4) calcula ing mixing a ios and when depa u es om he mixing line exis , iden i ying hyd ochemical p ocesses. This app oach has been applied in a pilo a ea in he Besòs Ri e Del a. ii II. Resumen En la ciudad de Ba celona y municipios ubicados sob e los e enos del aicos del ío Besòs, el agua sub e ánea es un ac o que se iene en cuen a en la ges ión u bana del ciclo del agua. La il ación de aguas sub e áneas en edi icios e in aes uc u as obliga el d enaje de es as aguas, pa e de las cuáles son e idas a la ed de alcan a illado. Sensibilizados po es a p oblemá ica los ayun amien os implicados enca ga on es udios hid ogeológicos pa a analiza las causas y plan ea al e na i as pa a el ap o echamien o de es os ecu sos, educi los cos es del a amien o, así como minimiza los impac os que las aguas sub e áneas y d enajes puedan causa a las in aes uc u as y ice e sa. A aíz de es os es udios y desde el año 2000, pa e de és as se des inan a usos municipales ( iego de ja dines y limpieza de calles). Pos e io men e y debido a un pe íodo de sequía que a ec ó desde ab il de 2007 a ene o de 2009, la en idad ges o a de los ecu sos híd icos de Ca aluña (Agència Ca alana de l’Aigua, ACA) plan eó un mayo ap o echamien o de los ecu sos de agua sub e ánea en es e en o no u bano. Es o equie e un buen conocimien o de las ca ac e ís icas hid ogeológicas del medio, de la disponibilidad y calidad de es os ecu sos, así como de dispone de las he amien as adecuadas pa a su e aluación y co ec a ges ión. Es a esis se enma ca en es e con ex o. La eca ga es uno de los aspec os más ele an es del balance híd ico en los sis emas acuí e os de zonas u banas. La u banización del e i o io al e a el ciclo hid ogeológico na u al, la eca ga di ec a de agua a los acuí e os po in il ación se educe, disminuye la e apo anspi ación y apa ecen nue as uen es de eca ga. En es a esis se p esen a, en p ime luga , una me odología pa a cuan i ica la a iabilidad espacial y empo al de la eca ga en zona u bana. Las uen es po enciales de eca ga que se han conside ado son: (1) in il ación di ec a de llu ia y de agua de esco en ía, (2) pé didas del sis ema de alcan a illado, (3) pé didas de la ed de abas ecimien o y (4) o as uen es especí icas de eca ga como la in il ación de agua de ío o la in usión ma ina, en e o as. El cálculo de la eca ga se ha ealizado inicialmen e de iniendo y aplicando ecuaciones analí icas bajo di e sas hipó esis. Es os esul ados se han e aluado en el con ex o hid ogeológico median e un modelo numé ico de lujo y anspo e en oda la zona de es udio. En segundo luga , ambién en oda la zona de es udio, se analiza la p esencia en las aguas sub e áneas de algunas sus ancias po encialmen e p esen es en uen es de eca gas asociadas a en o nos u banos, y de las que se iene ela i o poco conocimien o de su in oducción y e olución en los acuí e os. Es as subs ancias o sus p oduc os de deg adación pe enecen a los denominados ‘con aminan es o gánicos eme gen es’; se a an de compues os químicos o gánicos empleados en la o mulación de p oduc os de uso co idiano y muy ex endido, como son p oduc os a macéu icos, emulsiones, p oduc os pa a el cuidado e higiene pe sonal, de e gen es domés icos e indus iales, plás icos, pes icidas y he bicidas, en e o os. Con aminan es o gánicos cuya e olución y compo amien o en las aguas sub e áneas es más conocido pueden deg ada se median e p ocesos na u ales con olados p incipalmen e po p ocesos de oxidación- educción. Incluso se han desa ollado écnicas de es imulación in-si u capaces de acele a es os p ocesos de a enuación na u al sob e medios con aminados. En una línea p elimina de de ección de p ocesos de a enuación se p esen a, en e ce luga , una me odología pa a la e aluación de po cen ajes de mezcla e iden i icación de p ocesos hid oquímicos Pa a ello se emplean p ocedimien os y he amien as es adís icas. Consis e en (1) iden i icación de las po enciales uen es de eca ga, (2) ca ac e ización de las uen es de eca ga y las mezclas de agua usando especies químicas, (3) selección de las especies químicas que an a se usadas en el análisis y (4) cálculo de los po cen ajes de mezcla e iden i icación de p ocesos hid oquímicos cuando és os se sepa an de la línea de mezcla. Es a me odología se ha aplicado a las aguas del ío Besòs y las aguas sub e áneas de una zona pilo o en el Del a del Besòs. iii III. Resum En la ciu a de Ba celona i municipis ubica s sob e els e enys del aics del iu Besòs, l’aigua sub e ània és un ac o que es é en comp e en la ges ió u bana del cicle de l’aigua. La il ació d’aigües sub e ànies en edi icis i in aes uc u es obliguen el seu d ena ge, una pa de les quals s’aboquen a la xa xa de cla egue am. Sensibili za s pe aques a p oblemà ica els ajun amen s implica s an enca ega es udis hid ogeològics pe anali za les causes i plan eja al e na i es pe l’ap o i amen d’aques s ecu sos, edui els cos os del ac amen , així com minimi za els impac es que les aigües sub e ànies i d ena ges poden causa a les in aes uc u es i ice e sa. A a el d’aques es udis i des de l’any 2000, pa d’aques es aigües es des inen a usos municipals ( eg de ja dins i ne eja de ca e s). Pos e io men i degu a un pe íode de seque a que a a ec a des de l’ab il de 2007 ins gene de 2009, l’en i a ges o a dels ecu sos híd ics de Ca alunya (Agència Ca alana de l’Aigua, ACA) a plan eja un majo ap o i amen d’aques s ecu sos d’aigua sub e ània en aques en o n u bà. Això eque eix un bon coneixemen del medi hid ogeològic, de la disponibili a i quali a , així com de les eines adequades pe a la se a co ec a a aluació i ges ió. Aques és el con ex en que emma ca aques a esis. La eca ega és un dels aspec es més elle an s del balanç híd ic en els sis emes aqüí e s de zones u banes. La u bani zació del e i o i al e a el cicle hid ogeològic na u al, la ecà ega di ec a d’aigua als aqüí e s pe in il ació es edueix, disminueix l’e apo anspi ació i apa eixen no es on s de ecà ega. En aques a esis es p esen a, en p ime lloc, una me odologia pe a quan i ica la a iabili a espaial i empo al de la ecà ega en zona u bana. Les on s po encials de ecà ega que s’han conside a són: (1) in il ació di ec a de pluja i aigua d’esco en ia, (2) pè dues del sis ema de cla egue am, (3) pè dues de la xa xa d’abas amen i (4) al es on s especí iques de ecà ega com in il ació d’aigua del iu o in usió ma ina, en e al es. El càlcul de la ecà ega s’ha eali za inicialmen de inin i aplican equacions analí iques so a di e ses hipò esis. Aques s esul a s s’han a alua en el con ex hid ogeològic mi jançan un model numè ic de lux i anspo en o a la zona d’es udi. En segon lloc, ambé en o a la zona d’es udi, s’anali za la p esència en les aigües sub e ànies d’algunes subs àncies po encialmen p esen s en on s de ecà ega associades a en o ns u bans, i de les que es é ela i amen poc coneixemen de la se a in oducció i e olució en els aqüí e s. Aques es subs àncies o els seus p oduc es de deg adació pe anyen als denomina s ‘con aminan s o gànics eme gen s’; es ac en de compos os químics o gànics emp a s en la o mulació de p oduc es d’ús quo idià i mol es ès, com són p oduc es a macèu ics, emulsions, p oduc es pe la cu a e higiene pe sonal, de e gen s domès ics i indus ials, plàs ics, pes icides y he bicides, en e al es. Con aminan s o gànics, l’e olució i compo amen dels quals en les aigües sub e ànies és més conegu , poden deg ada -se mi jançan p ocessos na u als con ola s p incipalmen pe p ocessos d’oxidació- educció. Inclús s’han desen olupa ècniques d’es imulació in-si u capaces d’accele a aques s p ocessos d’a enuació na u al sob e medis con amina s. En una línea p elimina de de ecció de p ocessos d’a enuació es p esen a, en e ce lloc, una me odologia que pe me l’a aluació de pe cen a ges de mescla e iden i icació de p ocessos hid oquímics. Aques a emp a p ocedimen s i eines es adís iques. Consis eix en (1) iden i icació de las po encials on s de ecà ega, (2) ca ac e i zació de les on s de ecà ega i les mescles d’aigua emp an espècies químiques, (3) selecció de les especies químiques que an a se emp ades en l’anàlisi i (4) càlcul dels pe cen a ges de mescla i iden i icació de p ocessos hid oquímics d’aquells que es sepa en de la línea de mescla. Aques a me odologia s’ha aplica a les aigües del iu Besòs i les aigües sub e ànies d’una zona pilo en el Del a del Besòs. i IV. Ag adecimien os Tengo muchos mo i os pa a es a ag adecida a muchas pe sonas con las que he enido el p i ilegio de coincidi a lo la go del la go pe iodo desde el inicio has a el inal de es a esis (algunos dicen que lo impo an e no es llega al des ino sino el camino,…). Como no quisie a deja me a nadie quie o ex ende mi ag adecimien o a odas ellas en es e pá a o. Especial ag adecimien o a los di ec o es de es a esis En ic Vázquez y Jesús Ca e a, ha sido un g an hono ap ende de oso os. En ic g àcies pe o el eu ecolçamen i con iança. A Xa ie Sánchez-Vila po apoya es a esis. A odos los compañe os y p o eso es del G upo de Hid ología Sub e ánea C is ina Valhondo, Manuela Ba bie i, Jo di Fon , Anna Ju ado, Viole a Velasco, Héc o de Buen, Ch is ina Scho nbe g, Me i xell G an, Ad ià Ga iga, Radu Gogu, , Ma ía Pool, Jo di Sánchez (es a ez no me ol ido), Sil ia A anda, Leona do Donado, Lui Jan Sloo en, Te esa Tolosana, Alejand o Giménez, Vanessa Zabala, Paolo T inche o, Es anis Pujades, Daniele Ped e i, Ma co Ba ahona, Lu des Ma ínez, Emilio Cus odio, Ca les Ayo a, Daniel Fe nández, Xa ie Sánchez-Vila, Maa en Saal ink, Albe Sole , ... A los compañe os de despacho C is, Leo, Zhi eng Zhan, Giuseppe, Claudia, ... Al g upo de química ambien al del IDAE Susana González, Mi en López de Alda, Damià Ba celó, Mi a Pe o ic y C is ina Pos igo. A Mi eia Iglesias, Xa i Ca e as y Josep Ma ia Niñe ola del ACA y al pe sonal del labo a o io de ATLL en Ab e a. A Joan Casas, Joan Acos a, y An onio Iglesias y en gene al al pe sonal de l’ajun amen de San Ad ià del Besòs, de EUSAB y del Pa c del Besòs. A Aline C., J.Piña, Ma a G., Glò ia M., Neus Q., J.Pinyol, Ped o, ... Muy muy especialmen e a Da id, Gonzali a, Susana y An onio. In ii a g a i ud a mi amilia, a Joan Ca les y a su amilia, que ambién son mi amilia. Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 4 2. Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s. This chap e is based on he pape Tubau I, Vázquez-Suñé E, Ca e a J, Valhondo, C. Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s. Science o he To al En i onmen , submi ed. 2.1. In oduc ion G oundwa e , wi h inc easing equency, is conside ed a key ac o in he managemen o u ban a eas (Le ne , 1996; Chil on e al., 1997; Eyles, 1997; Chil on, 1999; Ellis, 1999; Howa d & Is a ilo , 2002; Cus odio, 1997, 2004 and lis o s udies compiled by Naik e al., 2008). Fo one, some ci ies pe o m in ensi e pumping o g oundwa e o a ious uses (wa e supply, indus y, e c.), which can p oduce a numbe o undesi able consequences such as g ound subsidence o de e io a ion o he quali y o wa e as a esul o a ious pa icula causes (e.g. seawa e in usion). Mo eo e , he pumping o g oundwa e in many ci ies ends o be educed o abandoned due o pollu ion and / o changes in land use (e.g. eloca ion o indus ies). In he la e case, he decline in g oundwa e ex ac ion has caused a eco e y o piezome ic le els, gene a ing loods and causing damage o unde g ound in as uc u e in many u ban a eas. Such is he case o Ba celona and i s u ban en i onmen . To con ol hese s uc u al p oblems i is usually necessa y o pump g oundwa e in speci ic a eas wi hin he u ban a ea (unde g ound pa king a eas, building basemen s, unnels, e c.). I is impo an o emembe ha g oundwa e ex ac ion may be necessa y o use ul o co e a ious aspec s o he g owing demand o u ban wa e . I is also due o become a s a egic esou ce ha can mee demand in momen s when he usual supply sou ces may be su e ing p oblems (d ough , acciden s, e c.). The la e conside a ion leads us o wonde whe he u ban g oundwa e can be used sa ely. U ban aqui e s can su e om e y di e se pollu ion sou ces, which include echa ging om he losses o sewe age, pollu ed u ban uno , pollu ed i e s o o he su ace wa e , seawa e in usion, e c. As a esul , a la ge numbe and a ie y o con aminan s can be ound in u ban aqui e s (Howa d e al., 1996; Ba e e al., 1999; Pi , 2001; Le ne , 2002; Mo is e al., 2005; Vázquez-Suñé e al., 2005; Wakida & Le ne , 2005 ; Musol , 2009). Besides "common" pollu an s in he u ban wa e en i onmen , he e a e nume ous o ganic pollu an s o a ious o igins, including a li le unde s ood and poo ly legisla ed Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 5 g oup know as eme ging o ganic con aminan s (EOCs), which a e o pa icula conce n o se e al easons (Ellis, 2006; Tubau e al., 2010; Ju ado e al., 2012a and 2012b; Pe ie e al., 2015). To add ess hese p oblems, we mus unde s and he p ocesses ha de e mine he biological and chemical quali y o u ban g oundwa e , which in u n makes i necessa y o quan i y he g oundwa e balance and especially he echa ge mechanisms (o igin, quan i y and quali y). Ob iously, g oundwa e managemen in u ban a eas equi es a de ailed knowledge o he hyd ogeological sys em as well as he adequa e ools o p edic ing he amoun and wa e quali y e olu ion. In his sense, expe ience in u ban g oundwa e is much lowe han in na u al a eas. A key di e ence lies in echa ge e alua ion. A la ge numbe o s udies ha e been published since he 90s ha e alua e echa ge in u ban a eas (see lis o s udies compiled by Naik e al., 2008), wi h no speci ic me hodology. Howe e , i is es ablished ha new u ban g oundwa e echa ge sou ces appea wi hin supply ne wo k losses and sewe losses. Fu he mo e, e apo anspi a ion dec eases due o he sealing su ace. The e o e, he e a e gene ally highe a es o echa ge in u ban han in he na u al se ings (Le ne 2002, Vazquez-Suñe e al., 2005). Con ibu ion e alua ion o each echa ge sou ce has been add essed (Le ne e al., 2002). Gi en hese unce ain ies, mixing a ios can be used o be e es ima e he ela i e impo ance o di e en sou ces o echa ge (Vazquez-suñe e al., 2010, Tubau e al., 2014). Fu he mo e, g oundwa e modeling may p o e a good ool o o al echa ge e alua ion (Vázquez-Suñé & Sánchez-Vila, 1999; Yang e al., 1999; Le ne , 2002 Hussein & Schwa z 2003, Dahan e al., 2004; T owsdale & Le ne 2003 and 2007; Cox e al., 2007). Howe e , his ype o modeling equi es a speci ic app oach o iden i y he con ibu ion o each pa icula sou ce o he o al echa ge. To add ess hese speci ied unce ain ies, a me hodology is p esen ed o pe o m g oundwa e modeling in u ban a eas ha ake in o accoun low and solu e anspo . Howe e , applica ion o eal cases wi h nume ous echa ge sou ces, such as u ban echa ge, has been hinde ed by se e al issues. The i s issue is associa ed wi h limi ed expe ience wi h he iden i ica ion and selec ion o app op ia e solu es. The second issue is he a ailabili y o long head and concen a ion eco ds. Small and spa se eco ds (in ime and space) could cause he nume ical solu ion o be non-iden i iable o uns able. The objec i e o his pape is o p esen a me hodology o help o e come some o he issues abo e. This me hodology will allow us o quan i y he a iabili y in space and ime o he echa ge in o aqui e s in u ban a eas. In his case, he conside ed po en ial sou ces o echa ge a e: (1) di ec ain and u ban uno in il a ion, (2) losses om he sewe sys em, (3) losses om he wa e supply sys em, and (4) o he speci ic sou ces o echa ge (i.e. i e in il a ion, seawa e in usion, e c.). Recha ge calcula ions a e ini ially Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 6 pe o med by de ining and applying some analy ical equa ions and alida ion has been assessed based on g oundwa e low and solu e anspo modeling. This allows u ban g oundwa e managemen , by aking in o accoun quan i y and quali y ac o s, o e alua e he ela i e con ibu ion o di e en echa ge sou ces by means o low and solu e anspo modeling app oach a a ci y scale. The me hodology is applied o he assessmen o echa ge sou ces in he Ba celona ci y aqui e s. 2.2. Me hodology The p oposed me hod is applicable when g oundwa e da a a e spa se. This me hod can be conside ed a a ia ion o he wa e - able luc ua ion mass balance me hod (i.e. Healy and Cook, 2002; De V ies & Simme s, 2002). In essence, i is based on calib a ing a la ge scale model while using app op ia ely pa ame e ized a eal echa ge unc ions. In p ac ice, i consis s o he ollowing s eps: 1) Concep ual model de ini ion, 2) Implemen a ion in o a nume ical model, 3) Pa ame e iza ion o a eal echa ge, 4) Calib a ion and 5) I e a ion. These s eps a e de ined as ollows. 1. Concep ual model De ini ion. This is achie ed using con en ional hyd ogeological me hods. I equi es unde s anding geology, hyd ogeology, me eo ology, hyd ology, soil uses, pumping a es, hyd ochemis y, e c. Emphasis mus be placed on e alua ing no only o al inpu s and ou pu s, bu also hei e olu ion o e ime. 2. Implemen a ion in o a nume ical model. Once a concep ual model has been de ined, a nume ical model is buil . This equi es speci ying all pa ame e s as well as he spa ial and empo al disc e iza ion and bounda y condi ions (la e al inpu s, echa ge, pumping a es, e c). 3. Pa ame e iza ion o a eal echa ge. U ban echa ge is de ined as a p esc ibed echa ge in low as ollows: ( ) ( ) 1 , , β = = ∑ N iRi i x x Whe e (x, ) is he echa ge a e (he e exp essed pe a ea uni and ime uni ), N is he numbe o e ms, β i a e unknown u ban echa ge pa ame e s o be es ima ed la e and Ri (x, ) a e unc ions ep esen ing he ime a ia ion o echa ge. Ideally, hese unc ions ep esen di e en modes o u ban echa ge. The ac ha hey a e le in e ms o unknown ac o s β i ough o simpli y hei de ini ion, which mus be p esc ibed by he modele . As men ioned in he in oduc ion, a de ailed accoun o all p ocesses in ol ed can be qui e di icul . Fac o s a ec ing hese may include: (1) echa ge sou ces ( ain all, wa e supply, sewage sys em); (2) pe cen age o u banized a ea; (3) es ima ion o pe cen age o wa e dis ibu ed o losses o mains and sewe s; (4) daily wa e [1] Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 7 consump ion pe pe son; (5) popula ion densi y; among o he s. In he u ban a ea all hese ac o s can a y in ime and space. The e o e, i is be e o use simpli ied unc ions ha cap u e he essence o he abo e, ye can be easily de ined wi hou he need o complex models. Some possible unc ions a e de ined below: Recha ge unc ion Ri (x, ) is implemen ed as a echa ge a e in each echa ge a ea, o a gi en ime in e al, as he sum o i e echa ge sou ces [2]: [2]R  , =R  +R  +R  +R  +R  Disagg ega ed, hey a e: [3]R  , =        R  =  ·WP·1−p #$ %+ R  =  ·WI·1−p #$ %+ R  =  ·WS+ R  =  ·[WS·(1−  )·(1−p * )+WS  ·(1−p * + )+WO] R  + =  + ·WP·p #$ ·p #   - Thus, concen a ion o a speci ic compound (Ci) which p o ides he echa ging in each echa ge a ea can be e alua ed as [4]: [4]C 0,1,2 =C 0 345 ·R  +C 0 645 ·R  +C 0 789 ·R  +C 0 788 ·R  +C 0 388+ ·R  + R  , Ci Pnu , Ci Inu , Ci Sws , Ci Sww and Ci Pww’ a e, espec i ely, he speci ic compound concen a ions o each echa ge sou ce e m. In he ollowing sec ions each o hese a iables will be de ined. Recha ge sou ces concep ualiza ion The concep ual model o asses g oundwa e echa ge in u ban a eas (Fig. 2.1 and Table 2.1) has been conside ed based on all a ailable wa e s suscep ible o being echa ged in unde g ound en i onmen s h ough elemen s ha can only p o ide echa ging. This pa icula ly means po en ial wa e s om elemen s, bo h na u al and a i icial, ha can also discha ge g oundwa e ha e no been included, e.g. i e s, aqui e bounda ies, sea bo de s, e c., wi h was ewa e pipes below he wa e able being a special case. He e we ha e conside ed ha he sewage sys em is uni a y (bo h, plu ial and u ban was e wa e is cap u ed by he sys em). Wa e s conside ed a e classi ied in o i e ypes acco ding o chemical cha ac e is ics, po en ial ways o en e g ound, and o igin. The a ing alues a e: (1) [WS] supply wa e lowing h ough he ne wo k; (2) [WS’] supply wa e no lowing h ough he ne wo k, e.g. om pumping g oundwa e ; (3) [WP] ainwa e ; Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 8 (4) [WI] wa e in il a ed ia he su ace du ing pe iods o no p ecipi a ion, e.g. di e en wa e sou ces used o i iga ion in a eas no wa e p oo ed; and (5) [WO] o he wa e discha ged by he sewe sys em ha is nei he ainwa e no wa e supply, e.g. g oundwa e d ains om basemen s. Figu e 2-1 Concep ual model o asses g oundwa e echa ge in u ban a eas (see also Table 1). Table 2-1 Syn hesis o e ms and pa ame e s o concep ual model o asses g oundwa e echa ge in u ban a eas (see also Figu e 1). Ne e heless, only a pe cen age o o al olume o hese wa e s echa ges aqui e s. Pe cen ages [ nu , ws , ww and ww’ ] ha e been classi ied in o h ee e ms acco ding o how po en ially wa e can po en ially en e he aqui e : (1) di ec in il a ion h ough unpa ed a eas [ nu ]; (2) losses o om supply mains [ ws ], whe e wa e lows unde p essu e; and (3) losses om sewe [ ww and ww’ ], whe e wa e usually is no unde p essu e. G oup Pa ame e Desc ip ion WS wa e supply lowing h ough he supply ne wo k sys em WS' wa e supply no lowing h ough he supply ne wo k WP ain all WI wa e o i iga ion WO o he wa e discha ged by he sewe sys em ws pe cen age echa ged om losses in he supply ne wo k nu pe cen age echa ged due o in il a ion om su ace ww pe cen age echa ged om sewe losses ww' pe cen age echa ged om sewe losses in ainy season p pa pa ed a ea pe cen age p pww ain all a io lowing in he sewe age sys ems p wsc a io o wa e supply (WS) consumed o e ained and no e u ned o he sewe p wsc' a io o wa e supply (WS') consumed o e ained and no e u ned o he sewe Wa e olume en e ing o u ban sys em G oundwa e echa ge a ios (exp essed as a decimal) Fac o s adap ed o spa ial and empo al disc e iza ion Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 9 U ban pa ed su ace is in oduced as a pe cen age o pa ed a ea [p pa ]. This can change wi h ime, p pww is ain all pe cen age cap u ed by he u ban d ainage sys em o sewe age sys em, p wsc and p wsc’ a e pe cen ages o wa e supply consumed o e ained, which he e o e do no e u n o he sewe sys em. One ad an age o his kind o concep ual model is ha i allows o balances wi h he global u ban wa e cycle by aking in o accoun con ibu ions no speci ically hyd ogeological (also see: Eiswi h e al 2004; Thomas & Tellam; 2006; Mo is e al., 2007; Vizin ina a al. 2009, F eni e al. 2009). Each o i e echa ge sou ce e ms is de ined as ollows (Fig. 2.1 and Table 2.1): Rain all echa ge in non-u ban a ea, R Pnu [3.1]: [3.1]R  =  ·WP·1−p #$ % WP is ain all o al olume in an a ea o a gi en ime in e al; p pa is unpa ed a ea pe cen age (exp essed as a decimal), which can be use ul in a eas whe e u banized su ace has changed o e ime, aking alue 0 o na u al a ea and 1 o u ban; nu is he a io o wa e ha echa ges aqui e s due o su ace in il a ion. In sho , R Pnu co esponds o he na u al echa ge and he e o e can be e alua ed using common wa e balance calcula ions in he soil and aking in o accoun pa ame e s such as h eshold uno , e apo anspi a ion a e, unsa u a ed zone hickness, and soil mois u e, among o he s. I iga ion wa e in non-u ban soil, R Inu [3.2]: [3.2]R  =  ·WI·1−p #$ % R Inu is in ended o ag icul u al a eas included in an u ban a ea, al hough i can also be conside ed i iga ion o g een a eas in an u ban con ex i his is assumed signi ican , ei he om a ocus on quali y o quan i y. WI is i iga ion wa e olume in an a ea o a gi en ime in e al, which could come om ei he p e iously eco ded wa e olumes in he o al balance o a ailable wa e , e.g. om wa e supply (WS), o om new con ibu ions o balance, e.g. pumping wells o i e s. p pa and nu a e he same pa ame e s desc ibed in [3.1]. In ac , al hough [3.1] and [3.2] ha e been sepa a ed, R Pnu and R Inu should be conside ed in conjunc ion wi h nu e alua ion. Rain all echa ge in u ban su ace, R Pww’ [3.3]: [3.3]R  =  ·WP·p #$ ·p # Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 10 WP and p pa a e desc ibed in [3.1]; p pww is ain all pe cen age cap u ed by he u ban d ainage sys em o sewe age sys em in an a ea; ww is pe cen age echa ged om losses om he sewe sys em in he ainy season. Recha ge om wa e supply ne wo k losses, R Sws [3.4]: [3.4]R  =  ·WS WS is olume o wa e supply lowing h ough he supply ne wo k in a gi en a ea and ws is he pe cen age (as a decimal) o his wa e eaching he aqui e om supply ne wo k losses. Se e al s udies ha e es ima ed his pe cen age om he di e ence be ween wa e supply deli e ed, wa e billed and wa e disappoin ed and some wo ks ake da a om ne wo k sec o s in o de o ake in o accoun aud and losses om la ice de ec s (e.g.Valdés & Cas elló, 2002). Recha ge om seepage losses, (R Sww ) [3.5]: [3.5]R  =  ·[WS·(1−  )·(1−p * )+WS  ·(1−p * )+WO] WS and ws a e de ined in [3.4]; WS’ is wa e olume in an a ea o a gi en ime in e al o consump ion ha is no coming om he gene al supply ne wo k (i.e. om p i a e wells o o he sou ces o supply wa e ), p wsc and p wsc’ a e he espec i e pe cen ages (exp essed as decimals) o wa e supply (gene al o p i a e) consumed o e ained he e o e does no e u n o he sewe sys em; WO is wa e olume in ended o consump ion (no om ne wo k supply) ha is discha ged in o he sewe sys em - o ins ance wa e om g oundwa e d ains o keep an unde g ound s uc u e such as a unnel o building ounda ion dewa e ed; ww is a pe cen age (as decimal) o wa e eaching he aqui e om sewe losses. Chisala and Le ne (2008) explain case s udies and app oaches o quan i y his pa ame e . Pa ame e s o his concep ual model do no explici ly ake in o accoun he p ocesses o he non-sa u a ed zone, an issue in u ban a eas ha is add essed by F iedel (2006). 4. Calib a ion. Calib a ion consis s o adjus ing u ban echa ge pa ame e s β i oge he wi h he emaining model pa ame e s, so as o i obse a ions while ensu ing ha bo h es ima ed pa ame e s and he esul ing model a e consis en wi h p io in o ma ion. The opic is e y b oad and a ull desc ip ion alls beyond he scope o his pape (see Ca e a e al., 2005). The ac ha echa ge pa ame e s a e es ima ed does no equi e any special a en ion. The only issue is ha , in p ac ice, one needs o use in e sion codes. He e we used he Maximum Likelihood Me hod (Ca e a and Neuman, 1986) as implemen ed in VisualTRANSIN (GHS, 2004; Medina and Ca e a, 1996), which leads o minimizing an objec i e unc ion. Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 11 λ λ λ = + + ∑ ∑ ∑ h h c c p p F F F F Whe e ( ) 1 * ( *) − = − − h h F h h C h h is he head objec i e unc ion, which penalizes depa u es be ween compu ed and measu ed heads (h and h*, espec i ely) wi h a weigh ing ma ix 1 − h C . Simila de ini ions hold o F c and F p concen a ions and model pa ame e objec i e unc ions, espec i ely. App op ia e weigh ing be ween hese h ee e ms can be ob ained wi h he pa ame e s λ h , λ c and λ p . 5. I e a ion The las wo s eps may be epea ed using di e en echa ge unc ions, o combina ions o hem, un il a sa is ac o y i is ound. 2.3. Applica ion o he Ba celona u ban aqui e s 2.3.1. Concep ual model De ini ion Ba celona and i s a ea o in luence include he ci y and some municipali ies in he me opoli an a ea (Fig. 2.2) wi h a popula ion o o e 3 million people. The a ea is bounded by he Collse ola (no h), he Medi e anean Sea and Besòs del a (no h-eas ) and he Llob ega aqui e s (sou hwes ). The clima e is Medi e anean, wi h an a e age ain all o abou 600 mm / yea . Figu e 2-2 Schema ic desc ip ion o he hyd ogeology o Ba celona. The le main hyd ogeological uni s. Righ piezome ic map (Fo a be e unde s anding ha e been ep esen ed as cu es equidis ance di e en le el. The ed lines ep esen low lines. G oundwa e le els in Ba celona (Fig. 2.2) ha e been ising, which c ea es nume ous p oblems o unde g ound s uc u es (subway, unde g ound pa king lo s, basemen s, e c). In o de o emedia e his p oblem, Ba celona has de eloped he i s ph ea ic wa e 0 2000 4000 6000 Me e s Piezome ic su ace Isolines: 0 - 25 5 m 25 - 250 25m [5] Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 12 seconda y dis ibu ion ne wo k in he wo ld. Wa e om his seconda y ne wo k is used o many pu poses in ga den i iga ion and s ee cleaning. Recen ly, he Wa e au ho i ies (Ca alan Wa e agency, ACA) aim is o ge g ea e use o u ban g oundwa e in his a ea, e en o wa e supply pu poses. All o hese possible uses a e condi ioned by p ocesses con olling he wa e ’s quali y. These p ocesses mus be unde s ood o u he de elopmen o he seconda y ne wo k and possible supply uses. To his end, wo k will be pe o med h oughou he whole ci y. One o he key issues he e is he e alua ion o he di e en sou ces o echa ge (Vazquez-Suñé e al., 2005, and 2010). These include: (1) The Besòs Ri e , which la gely con ains wa e om seconda y e luen WWTP (Was ewa e T ea men Plan s), especially du ing he summe ; (2) he in il a ion o ain all in he highlands o he ci y ha a e no buil on (conside ed in e ms o na u al wa e echa ge as "clean"); (3) wa e supply ne wo k losses in he No h-Eas a ea. This wa e is impo ed om he i e Te ; (4) wa e supply ne wo k losses in he Sou h-Wes a ea. This wa e is impo ed om he i e Llob ega ; (5) he was ewa e gene a ed in he No h-Eas sec o ( om he i e Te ); (6) was ewa e sec o Sou h-Wes ( om he i e Llob ega ); (7) u ban uno ainwa e ha leaches in o much o he a mosphe ic deposi ion u ban a ea (se e al kinds o "u ban poles" emissions o ehicles and indus ies, e c.) and echa ge aqui e s h ough losses o di ec in il a ion o sewage; and (8) he in usion o seawa e . Each echa ge sou ce o g oundwa e in u ban a eas in oduces speci ic compounds and con aminan s in o he soil and g oundwa e . A ci y-wide geological model including low and anspo s p ocesses is al eady a ailable. This will be e ined a he pilo a eas, which will be selec ed based on he a ailabili y o exis ing in o ma ion. A comple e Geological and Hyd ogeological Da abase does his need o be capi alized (geology and p e ious hyd aulic pa ame e iza ion, g oundwa e heads e olu ion, g oundwa e hyd ochemis y e olu ion, hyd aulic es , pumping a es, e c.) has been used o build he model o hese aqui e s. 2.3.2. Flow and anspo modeling A huge nume ical low and solu e anspo model o he Ba celona aqui e s has been conduc ed o wa e managemen pu poses. The model is pe o med using ini e elemen Visual T ansin CODE (GHS, 2004). The model s uc u e has been conside ed as consis ing o h ee laye s: he op laye ep esen s he supe icial aqui e sys ems o Ba celona, Badalona and Besòs Del a, he middle laye ep esen s he Main Besòs Del a aqui e and he bo om laye ep esen s he deepes Besòs Del a aqui e . This o e lay ne wo k con ains all a ailable da a (heads and chlo ide concen a ions), limi s and bounda y condi ions o he domain, hyd aulic pa ame e s o aqui e s (i.e: conduc i i y, Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 13 s o age coe icien , po osi y, e c.), g oundwa e ex ac ion and echa ge ha e also been included. The pe iod co e s modeling om 1916 o 2006. In his model, calib a ion echniques ha e been used o an in e se p oblem. Tha is, all da a en e ed ha e been compiled and he p og am e u ns hyd aulic pa ame e s consis en wi h he p elimina y in o ma ion a ailable, which p o ide a be e i wi h he le els and concen a ions measu ed. The adjus men is conside ed e y good a he ci y scale. Ano he impo an aspec o he calib a ion consis ency in he mass balance (concep ual model) is in ega ds o g oundwa e in lows and ou lows in he whole sys em and hei empo al a iabili y. The model is conside ed o accu a ely ep oduce he empo al e olu ion o inpu s and ou pu s o he a ious elemen s o he sys em (Fig. 2.3). Un il eaching he inal calib a ion e sion o he model, quan i ica ion o echa ge sou ces p o ide he mos unce ain esul s. To assess a con iden quan i ica ion o echa ge sou ces he me hodology desc ibed in he nex sec ion was applied. Figu e 2-3 Mass balance esul ing o nume ical model. 2.3.3. Simula ion o u ban echa ge Calcula ing wi h he echa ge model c ea ed a unc ion o ime [Eq. 6] ha b ings oge he di e en ypes o echa ge wa e depending on he ype and sou ce o he o iginal wa e be o e in il a ing he g ound, which a e: (1) ainwa e in il a ion om he na u al e ain (o land no build on), which quali ies as a "na u al echa ge", R Pnu ; (2) ap wa e om losses in he supply ne wo k, R Sws , (3) sewe wa e om losses in he sewe sys em, R Sww and R Pww’ and (4) echa ge om i e s o s eams, which has been ea ed di e en ly by ano he ime unc ion. URBAN RECHARGE 27.02 NON-URBAN RECHARGE 2.86 BESÒS RIVER CONTRIBUTION 8.61 CREEKS CONTRIBUTION -3.32 SUBWAY DRAINAGE -3.03 TOTAL ABSTRACTION -34.75 LA LLAGOSTA CONTRIBUTION 6.72 LLOBREGAT AQUIFER CONTRIBUTION -3.39 SEA TOTAL CONTRIBUTION -0.81 STORAGE -0.10 GLOBAL MASS BALANCE (hm 3 /y) SUMMARY TABLE, AVERAGE 1916-2006 ALLUVIAL BOUND. SEA CREEKS BESÒS RIVER BESÒS RIVER LLOBREGAT AQUIFER ABSTRACTIONS RECHARGE ∆S SUBWAY MASS BALANCE INPUT/OUTPUT BESÒS MAIN AQUIFER (LAYER 2) BESÒS LOWER AQUIFER (LAYER 3) NON-URBAN ZONE URBAN ZONE BESÒS ALLUVIAL BESÒS SHALLOW AQUIFER Chap e 2: Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s 20 2.5. Conclusions A eal echa ge may be he mos impo an inpu in u ban aqui e sys ems. Ye , accu a e e alua ion o his inpu is di icul . We p esen ed a me hodology o help es ima e he a eal echa ge o u ban aqui e s. The p oposed me hod is essen ially simila o he wa e - able luc ua ion mass balance me hod, applicable when g oundwa e head da a a e spa se, so ha lack o spa ial and empo al co e age is au oma ically in e pola ed by he model. I consis s in (1) de ini ion o he concep ual model; (2) implemen a ion in o a nume ical model; (3) pa ame e iza ion o u ban echa ge; (4) calib a ion and (5) i e a ion. The wo las s eps need o be epea ed un il inding he bes i is ound o he ange o unce ain y is na owed. This me hodology is applicable o complex sys ems by aking in o accoun empo al a iabili y o all wa e sou ces. A ansien (1916-2006) g oundwa e low and solu e anspo model o he Ba celona u ban aqui e s was buil as a ool o helping in de ining he managemen p og am. This model was calib a ed agains heads and chlo ide concen a ion da a ob aining e y sa is ac o y i s o bo h a iables. Mass balance and hyd aulic pa ame e s we e also consis en wi h p io calcula ions. This applica ion e lec s he impo ance o echa ge e olu ion on he aqui e s mass balance. They a e one o he mos impo an inpu s. We belie e ha he si ua ion ound in he Ba celona u ban aqui e may be ypical o many u ban aqui e sys ems. A eal echa ge in u ban a eas may be a e y signi ican po ion o o al inpu s. He e i was he i s . Based on gene al c i e ia in he calib a ion p ocess o echa ge we p o ide ha : − Hypo hesis o assess echa ge in u ban a eas ha e been p oposed based on pe cen ages o wa e mains and sewe ne wo k suscep ible o losses and echa ge he g oundwa e scheme. These wa e s a e classi ied acco ding o hei chemical p ope ies, po en ial ways o en e he g ound, and o igin. Fo his pa icula s udy case, he spa ial and empo al disc e iza ion echa ge, in absence o mo e di ec da a, uses popula ion densi y, pe capi a alloca ion, and land use. − Calib a ion o a ious echa ge scena ios equi es ha o he pa ame e s o be calib a ed o ha e he minimum amoun o de ia ion possible. − A good head se ing may no be su icien o quan i y echa ge, especially in hose a eas in luenced by bounda ies; in which case hyd ochemical measu es a e use ul o alida e o ejec he esul ing concep ual model. Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 21 3. Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. This chap e is based on he pape Tubau I, Vázquez-Suñé E, Ca e a J, González S, Pe o ic M, López de Alda MJ, Ba celó D. Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. Jou nal o Hyd ology; 383(2010):102–110. doi:10.1016/j.hyd ol.2009.11.030. 3.1. In oduc ion High popula ion g ow h inc eases he demand o wa e and he gene a ion o was ewa e . g oundwa e is an impo an esou ce o wa e supply in he wo ld and aqui e s in u ban zones a e ulne able o con amina ion by a a ie y o con aminan s due o u ban ac i i ies. App oxima ely 90 % o he wa e o supply comes om impo ed su ace wa e in he ci y o Ba celona, which is he s udy a ea. Since supply sou ces a e subjec o pe iodic d ough s, he Ca alan Wa e Agency (ACA) is seeking al e na i e wa e esou ces. The e a e a numbe o aqui e s wi h a ying li hologies and ages below he ci y (Váquez-Suñé e al., 2005). The e o e, g oundwa e mus be conside ed as an op ion. The ques ion is whe he u ban g oundwa e can be used sa ely. U ban aqui e s may su e pollu ion om di e en sou ces, including losses om wa e supply ne wo ks, leakage om sewe s, seepage om i e s o o he su ace wa e bodies, seawa e in usion, e c. As a esul , a la ge numbe and a ie y o con aminan s may be ound in u ban aqui e s (Howa d e al., 1996; Ba e e al., 1999; Le ne , 2002; Mo is e al., 2005; Váquez-Suñé e al., 2000; Wakida and Le ne , 2005; Naik e al., 2008). Some o hese con aminan s a e subjec o en i onmen al quali y s anda ds (EQS) in d inking wa e (Council o he Eu opean Union, 1998), g oundwa e (Eu opean Pa liamen and Council o he Eu opean Union, 2006), and/o su ace wa e (Eu opean Pa liamen and Council o he Eu opean Union, 2008). O he was ewa e subs ances, also known as eme ging con aminan s, whose oxici y, occu ence and pe sis ence a e less well known, a e Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 22 subjec o e iew o possible iden i ica ion as p io i y subs ances o p io i y haza dous subs ances (Eu opean Pa liamen and Council o he Eu opean Union, 2008). The lis o p io i y subs ances in he ield o wa e policy, adop ed o he i s ime in 2004 and e ised in 2008, includes a o al o 33 subs ances, some o which had ne e been subjec o egula ion in wa e be o e. This is he case o he alkylphenol polye hoxyla es (APEOs) deg ada ion p oduc s (DPs) nonylphenol (NP) and oc ylphenol (OP). APEOs a e nonionic su ac an s ex ensi ely used by indus y, and in he pas also in household p oduc s. Scien i ic e idence on he occu ence and oxici y in wa e o hei deg ada ion p oduc s mo i a ed he iden i ica ion o NP and OP as p io i y haza dous subs ances and he es ablishmen o EQS o hem in su ace wa e s, which ansla ed in annual a e age (AA) concen a ions o 0.3 and 0.01–0.1 lg/L o NP and OP, espec i ely, and a maximum allowable concen a ion (MAC) o 2 lg/L o NP. These compounds as well as o he APEOs DPs, such as alkylphenol ca boxyla es (APECs), a e o en ound in su ace wa e s, a concen a ions some imes su passing he es ablished EQS as e idenced, o ins ance, in a b oad su ey ca ied ou a he Eu opean le el in sea ch o he occu ence o 35 selec ed compounds in i e wa e samples om o e 100 Eu opean i e s (Loos e al., 2009). O he s udies, such as an ex ensi e su ey ca ied ou in 2000 in he USA o e alua e he p esence o 65 o ganic was ewa e con aminan s in 47 g oundwa e si es om 18 s a es, ha e e idenced also hei p esence in aqui e s (Ba nes e al., 2008). In ac , APEOs DPs we e among he sea ched con aminan s mos abundan and mos equen ly de ec ed in bo h o he abo e s udies. In he s udy a ea, González e al. (2004) obse ed he p esence o APEOs DPs and LAS a hund eds o lg/L concen a ions a he en y poin s o was ewa e ea men plan s (WWTPs). This s udy is ocused on deg ada ion p oduc s o nonionic su ac an s APEOs and he ionic su ac an s LAS. LAS and hei deg ada ion p oduc s (sul ophenyl ca boxyla es, SPC) a e no ega ded as p io i y subs ances al hough hey a e oday he mos common domes ic su ac an s in he a ea. Su ac an s a e used in a a ie y o indus ial and domes ic p oduc s such as cleaning p oduc s, deg ease s and de e gen s. They a e no mally di ided in o ou main g oups: (1) anionics, (2) nonionics, (3) ca ionics and (4) ampho e ics (Ying, 2006). Anionic and nonionic su ac an s cu en ly accoun o 90% (50% and 40% espec i ely) o he su ac an p oduc ion in wes e n Eu ope, i.e. 2.5 million ons in 2002. LAS accoun o 45% o he anionics, whe eas a y alcohol e hoxyla es (AEs) and APEOs ep esen 60% and 10% o he nonionics, espec i ely (Eu opean Commi ee o O ganic Su ac an s and hei In e media es, 2007). The mos widely s udied APEOs a e nonylphenol polye hoxyla es (NPEOs) and oc ylphenol polye hoxyla es (OPEOs), (see Fig. 3.1 o nomencla u e). Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 23 Figu e 3-1 Ac onyms and chemical s uc u es o alkylphenol e hoxyla es (APEOs) and some o hei deg ada ions p oduc s (DPs). Comme cially p oduced APEOs con ain be ween 6 and 12 e hoxyla e g oups (APnEO, n = 6, …, 12). These unde go apid p ima y biodeg ada ion (Gige e al., 1984; Ahel e al.,1994a; Di Co cia e al., 1998; Jonke s e al., 2001). This ini ial bio ans o ma ion o APEOs p oduces deg ada ion p oduc s wi h sho ened e hoxy chains, mainly AP2Es and AP1Es (AP2EOs, AP1EOs and he mo e hyd ophilic AP2ECs and AP1ECs), Fig. 3.2. Ca boxyla ed e hoxy chains AP2ECs and AP1ECs ha e been obse ed in ae obic biodeg ada ion (Ahel e al., 1994b; Lu e al., 2008) bu no unde anae obic condi ions (Lu e al., 2008). Anae obically ea ed sewage sludge a o s he o ma ion o APs (Gige e al., 1984; Ahel e al., 1994a; Ejle sson e al., 1999). Figu e 3-2 Bio ans o ma ion pa hways o alkylphenol polye hoxyla es. Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 24 The oxici y o APEOs is inc eased wi h dec easing e hoxyla e chain leng h (S aples e al., 2004; Ba be e al., 2007). Alkylphenol (AP) is hei sho es chain and he mos oxic (En i onmen Canada, 1999). Because o hei oxici y and pe sis ence, APEOs ha e been eplaced by LAS and AEs since he mid 1960s. LAS and AEs a e mo e biodeg adable and hei biodeg ada ion p oduc s a e less oxic (Kime le and Swishe , 1977). LAS is ae obically well deg aded and he o ma ion o me aboli es (mainly SPCs) well known (Yada e al., 2001; Scheleheck e al., 2007) and display high le els o biodeg ada ion in some WWTPs (B unne e al., 1988; Schöbe l and Ma l, 1989; Thiele e al., 1997). Howe e , LAS a e belie ed no o deg ade signi ican ly unde s ic ly anae obic condi ions in labo a o y es s o anae obic diges e o sewage sludge (Be na e al., 2007). LAS, APEOs and hei deg ada ion p oduc s ha e been de ec ed in he en i onmen : es ua ine and coas al zones (Vi ian, 1986; González e al., 2004; Zolle , 2006), i e s (Ahel e al., 1994b; Céspedes e al., 2005; Loos e al., 2009), g oundwa e wi h apid in il a ion o un ea ed and ea ed was ewa e (Ba be e al., 1988; Field e al., 1992; Leenhee e al., 2001; La o e e al., 2003; Swa z e al., 2006). Mo eo e , he beha io o LAS and APEOs has been s udied unde labo a o y condi ions (La son and Games, 1981; Manzano e al., 1998; Pe ales e al., 1999; Jonke s e al., 2001; Ying e al., 2008). O e all, i is no known o wha ex en do biodeg ada ion pa hways s udied in he labo a o y and was ewa e plan s explain wha occu s in ppb concen a ion in he na u al en i onmen , especially in g oundwa e whe e he esidence ime is long and whe e wa e may unde go a sequence o edox condi ions. In ac , K uege e al. (1998) compa ed mass emo al a es o LAS in labo a o y column expe imen s and ield ace es s. Biodeg ada ion was 2–3 imes highe in he labo a o y han in he ield unde simila concen a ions o dissol ed oxygen. Ying e al. (2008) compa ed he biodeg ada ion o NPs and OPs unde ae obic and anoxic condi ions in he labo a o y using wa e and soil om aqui e s. The e ha e been ewe s udies on he a e o APEOs in u ban g oundwa e , e.g. deg ada ion o APEOs in g oundwa e nea i e s (Ahel e al., 1996) and LAS in con amina ion plumes (Thu man e al., 1986; Ba be e al., 1988; La son e al., 1989). This s udy deals wi h he a e APEOs DPs and LAS su ac an in u ban g oundwa e a ield scale. To his end, he occu ence o APEOs DPs in g oundwa e was compa ed wi h: (1) sou ces o echa ge o /and pollu ion con aining hese subs ances, (2) g oundwa e edox condi ions, (3) occu ence o linea alkylbenzene sul ona e (LAS) which a e nowadays he mos common domes ic su ac an s in he a ea and (4) o he common con aminan s in u ban g oundwa e in he ci y o Ba celona. The APEOs DPs analyzed we e NP2EC, NP1EC, NP, OP2EC, OP1EC and OP. In his s udy, only he o iginal LAS p oduc was analyzed. Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 25 3.2.1. Si e desc ip ion The s udy a ea occupies 136 km2 and includes Ba celona and some owns o he me opoli an a ea (Fig. 3.3), wi h a popula ion o 2.2 million. Clima e is Medi e anean wi h an a e age ain all o some 600 mm/yea . The a ea is bounded by he Collse ola ange (Ca alan coas al anges, maximum ele a ion 512 m.a.s.l.), he Medi e anean sea and Besòs del a ( o he no heas ) and he Llob ega del a (sou hwes ). Aqui e s wi h di e en li hologies and ages ha e been mapped unde he ci y. g oundwa e lows gene ally seawa ds om he moun ain ange in all o hem (Fig. 3.3). Se e al echa ge sou ces in he aqui e s o Ba celona ha e been iden i ied (Váquez-Suñé e al., 2000). These include ain all in il a ion in he non-u banized a eas a he highe pa s o he ci y (Fig. 3.3 and 3.4). They a e ega ded as ‘‘clean” na u al echa ge wa e . The e a e wo po en ial in lows om su ace wa e bodies: seawa e in usion and wa e om Besòs Ri e . Figu e 3-3 Schema ic desc ip ion o s udy a ea. Bo om, loca ion and cha ac e is ics g oundwa e sampling poin s and piezome ic map, no e ha con ou s in e al a e 5 m o heads below 25 m (con inuous lines) and 25 m abo e (dashed). No e he change in g adien a he Paleozoic and Qua e na y con ac s. Figu e 3-4 Es ima ed annual a e age g oundwa e balance in he pe iod 2000–2006 exp essed as pe cen age o inpu s (modi ied a e Vázquez-Suñé e al., 2005). Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 26 The la e con ains a la ge p opo ion o e luen s om seconda y WWTPs, especially du ing he summe . Addi ional sou ces o g oundwa e echa ge a ibu able o an h opic ac i i ies include: (1) losses om he wa e supply ne wo k (supply wa e comes om he Llob ega Ri e in sou he n qua e s and om he Te Ri e in no he n qua e s), (2) in pa ed a eas, uno wa e , which washes away he u ban su ace a mosphe ic deposi ion and echa ges he aqui e s h ough di ec in il a ion o sewe seepage, and (3) losses om sewage wa e . Acco dingly, we iden i ied up o eigh di e en echa ge sou ces in he ci y o Ba celona: Ri e Besòs, (2) ain all echa ge in no he n non-u ban a ea, Te Ri e wa e supply, (4) Llob ega Ri e wa e supply, (5) Te Ri e sewage wa e , (6) Llob ega Ri e sewage wa e , (7) Ci y uno and (8) Sea wa e in usion. 2.3.4. Sampling Fi y- h ee wa e samples we e collec ed om Sep embe 2006 o Janua y 2007 (Fig. 3.3). Two samples om possible echa ge sou ces: one sample om Ri e Besòs and one sample o d inking wa e om a public ap. Fi y-one samples we e collec ed om g oundwa e : h ee samples om na u al sp ings, 23 om obse a ion piezome e s and 25 om ac i e wells belonging o p i a e indi iduals and companies (i.e., cu en ly pumped, no om disused wells). Piezome e and wells a ied in dep h (Fig. 3.3). Piezome e samples we e ob ained by pumping un il a olume ha was a leas h ee imes ha o he piezome e had been ex ac ed. Field pa ame e s measu ed in si u included elec ic conduc i i y, pH, empe a u e, Eh and dissol ed oxygen. They we e measu ed con inuously using a low cell o a oid con ac wi h he ai . The ins umen s we e calib a ed daily by means o s anda d solu ions. Samples we e collec ed a e s abiliza ion o ield pa ame e s and we e no il e ed in he ield. Each wa e sample unde wen chemical analysis in acco dance wi h he Spanish guidelines o d inking wa e (Royal Dec ee 140/ 2003). These guidelines include 53 pa ame e s ( ou adioac i i y pa ame e s we e no analyzed). The discussion o he esul s o his s udy only add esses some o he main ace species o was e wa e such as ni a e, ammonium, bo on, T ichlo e hene and Pe chlo e hene. Su ac an samples we e s o ed in Polye hylene e eph hala e (PET) con aine s ha we e ambe in colo o a oid pho o-deg ada ion. Con aine s we e s o ed in a e ige a o a 4 °C un il anspo o he labo a o y, whe e he samples we e ea ed on a daily basis. Once in he labo a o y, su ac an s samples we e il e ed h ough 0.45 µm memb ane il e and p econcen a ed on LiCh olu C18 (Me ck, Da ms ad , Ge many) solid phase ex ac ion (SPE) ca idges (IST, UK) wi hin 24 h in o de o a oid any deg ada ion o a ge compounds and loss o sample in eg i y. All SPE expe imen s we e pe o med using a Bake SPE 12 g appa a us (J.T. Bake , De en e , Ne he lands) connec ed o a acuum sys em se a -15 psi o he dispensing o samples Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 27 h ough he SPE ca idges. Disposable 3-mL ca idge columns packed wi h 500 mg o LiCh olu C18 so ben om Me ck (Da ms a , Ge many) we e ac i a ed and condi ioned i s wi h 7 mL o me hanol and hen wi h 3 mL o HPLC wa e a a low a e o 1 mL/min. Samples we e loaded a a low a e o 5 mL/min. A e p econcen a ion, he so ben s we e comple ely d ied (30 min) o a oid hyd olysis. A e d ying, he SPE ca idges we e w apped in aluminum oil and kep a -20 °C un il analysis (max. 1 mon h). Ca idges we e elu ed wi h 2 x 4 mL o me hanol. The elua es we e e apo a ed o d yness wi h a gen le s eam o ni ogen and econs i u ed wi h me hanol o a inal olume o 1 mL. The comple e SPE p ocedu e is desc ibed by Pe o ic e al. (2001). 2.3.5. Analy ical me hods All sol en s (wa e , ace oni ile and me hanol) we e high pe o mance liquid ch oma og aphy (HPLC) g ade and we e pu chased om Me ck (Da ms ad , Ge many). The s anda ds used in his s udy we e o he highes pu i y a ailable. High pu i y (98%) 4- e -OP and 4-NP we e ob ained om Ald ich (Milwaukee, WI, USA). OP1EC, OP2EC, NP1EC and NP2EC we e syn hesized acco ding o he me hod desc ibed by Diaz e al. (2002). Comme cial LAS wi h low dialkyl e alinsul ona e con en (<0.5%) we e supplied by Pe oquimica Española S.A. in a single s anda d mix u e wi h he p opo ional composi ion o he ou homologs o : C10: 3.9%, C11: 37.4%, C12: 35.4%, C13: 23.1%. 4-n- NP-d8 used as he in e nal s anda d we e ob ained om D . S. Eh ens o e (Augsbu g, Ge many). S ock solu ions (1 mg/mL) o indi idual s anda ds and s anda d mix u es we e p epa ed by dissol ing p ecise amoun s o pu e s anda ds in me hanol. S anda d wo king solu ions we e ob ained by u he dilu ion o s ock solu ions wi h me hanol. Analyses we e pe o med on a Wa e s 2690 se ies Alliance HPLC (Wa e s, Mil o d, MA) wi h a qua e na y pump equipped wi h a 120 ial capaci y sample managemen sys em. The analy es we e sepa a ed on a 5 µlm, 125 x 2 mm i.d., C18 e e sed-phase column (Pu osphe STAR RP-18), p eceded by a gua d column (4 x 4, 5 lm) o he same packing ma e ial, bo h supplied by Me ck, Da ms ad , Ge many. The sample injec ion olume was se a 10 µL. A bina y mobile phase g adien wi h me hanol (A) and wa e (B) was used o analy e sepa a ion a a low a e o 200 µL/min. The elu ion g adien was linea ly inc eased om 30% A o 80% A in 10 min, hen inc eased o 95% A in 10 min and kep isoc a ic o 5 min. A bench op iple quad upole mass spec ome e Qua o liquid ch oma og aphy (LC) om Mic omass (Manches e , UK) equipped wi h a pneuma ically assis ed elec osp ay p obe and a Z-sp ay in e ace was used o his s udy. Capilla y ol age was se a -2.8 kV, ex ac o lens a 7 V, and RF lens a 0.6 V. The sou ce and desol a ion empe a u es we e 150 and 350 °C, espec i ely. The ni ogen lows we e op imized al 50 L/h o he in cone gas and 540 L/h o Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 28 desol a ion gas. Fo andem mass spec ome y (MS–MS) expe imen s, he a gon collision gas was main ained a he p essu e o 5.8 x 10-3 mba . The de ec ion was pe o med unde nega i e ioniza ion (NI) by LC–MS–MS in mul iple eac ions moni o ing (MRM) mode as desc ibed by Pe o ic e al. (2003) and González e al. (2008). 3.2. Resul s and discussion 3.2.1. Gene ic pollu ion: ammonium, ni a e and pe chlo e hene Resul s o gene ic con aminan s a e shown in Table 3.1 and Fig. 3.5. Based on he esul s o a s a is ical analysis o chemical da a, g oundwa e samples collec ed in he s udy a ea we e g ouped in o h ee ca ego ies (Z1, Z2 and Z3). Z1 a e samples aken in non-u ban zone and he emaining samples was di ided acco ding o he concen a ion o ammonium: samples wi h an ammonium concen a ion lowe han 0.3 mg/L (Z2) and samples wi h a concen a ion highe han 0.3 mg/L (Z3). As i u ns ou , he samples in Z3 we e collec ed in aqui e s nea Ri e Besòs which lows o e del aic aqui e s wi h high ansmissi i y and wa e le els close o he soil su ace. As a esul , mos o he echa ge comes om he i e . Zone Z2 samples displayed ai ly oxidizing condi ions, mainly hose loca ed in he lowe hal o he ci y (Fig. 3.5). Reducing condi ions a e ound in Z2 samples aken om deep piezome e s wi h opening in schis s (Fig. 3.3 and 3.5). Ni a e concen a ions exceeding 50 mg/L we e de ec ed in 22 ou o 33 samples and exceeded 50 lg/L o pe chlo e hene (PCE) in app oxima ely one hi d o he samples. By con as , educed species such as ammonium p e ailed in Z3 samples. Despi e he ac ha Z3 is in a p edominan ly indus ial a ea, PCE and ichlo e hene (TCE) concen a ions we e ela i ely lowe (mean o 5.8 lg/L). G oundwa e in he non-u ban zone (Z1) was ai ly di e se. Shallow samples we e ae obic, while he deep ones we e educing. Ne e heless, o al ni ogen, bo on and PCE we e low, indica ing ha pollu ion was much lowe han in Z2 and Z3. Ni a e, ammonium and bo on a e used as ace s o was e wa e in u ban g oundwa e (Wakida and Le ne , 2005; Ba e e al., 1999). High concen a ions o bo on (and ions such as chlo ide and sodium) we e de ec ed in samples om del aic ma e ials as a esul o ma ine in usion in he lowe pa s o he ci y. The esul s o one sample (‘‘PEAR”, Fig. 3.5) a e included in he s udy al hough hey a e no classi ied in zone Z2 because his sampleappea s o come om an indus ial pollu ion si e and may no be ep esen a i e o Z2 wa e s. Ve y high concen a ions o o al o ganic ca bon we e de ec ed in his sample (186 mg/L, wo o de s o magni ude g ea e han he emaining samples), BTEX (benzene, oluene, e hylbenzene and xylenes) and low Eh. Chap e 3: Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g oundwa e : Ba celona case s udy. 29 Table 3-1 Gene ic con aminan a e age and maximum concen a ions in he s udy a ea g ouped acco ding o he zones o Fig. 3.5. Figu e 3-5 Spa ial dis ibu ion o ammonium, ni a e, pe chlo e hene and bo on. 3.2.1. Su ac an s: LAS and APEOs deg ada ion p oduc s LAS was he su ac an wi h highes concen a ions in po en ial echa ge wa e s (Table 3.2). Despi e he ac ha alkylphenolic su ac an s ha e been banned in domes ic p oduc s in he s udy a ea since he 1990s, APEOs DPs we e de ec ed by González e al. (2004) in WWTPs and also in he i e sample (Table 3.2). No wi hs anding, concen a ions o APEOs DPs we e much lowe han LAS concen a ions in echa ge wa e s. This con as s wi h aqui e samples om Z3 whe e highe concen a ions o APEOs DPs han LAS we e de ec ed (Table 3.3 and Fig. 3.6). The highes concen a ions o all he APEOs DPs we e ound in wells and piezome e s loca ed in Z3. Recha ge in his a ea comes basically om he i e . Ne e heless, measu ed concen a ions in Z3 we e highe han concen a ions in he i e , NP2EC being he maximum. Ahel e al. (1996) Amonium (mg/L) 0.0 (0.24) 0.0 (0.07) 2.9 (8.6) Ni a e (mg/L) 18.4 (49.94) 71.6 (280.74) 10.3 (24.15) Bo on ( µ g/L) 69.9 (154.94) 185.8 (753.53) 378.0 (700.47) Te aclo e hene ( µ g/L) 1.0 (4.59) 53.9 (347.86) 5.8 (24) T iclo e hene (µg/L) 0.1 (0.53) 11.8 (119.6) 2.1 (5.3) a Resul s a e g ouped as in Fig.3.5 . a e age (maximum) Z1 a (8 samples) Z2 a (33 samples) Z2 a (33 samples) PEAR PEAR Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 36 me hodology is applied o he assessmen o i e echa ge wa e s in he Besòs Del a aqui e s in Ba celona, Spain. The p oposed app oach consis s o ou s eps: (1) iden i ying he po en ial sou ces o echa ge, (2) cha ac e ising echa ge sou ces and mixed wa e samples using hyd ogeochemis y, (3) selec ing chemical species ( ace s) o be used in he analysis and (4) calcula ing mixing a ios and iden i ica ion o hyd ochemical eac ions in g oundwa e . Figu e 4-1 Ri e e sus g oundwa e a e age concen a ions in he s udy a ea o se e al o species such as majo and mino ions, me als, edox indica o s, pes icides, among o he s. The i e da a se is collec ed mon hly by he Ca alan Wa e Agency (ACA) 2 km ups eam o he s udy zone and i can be downloaded om h p://www.genca .ca . 4.2. Me hodology and da a ea men 4.2.1. Iden i y he sou ces o echa ge The iden i ica ion o he echa ge sou ces is speci ic in each s udy a ea and i equi es an unde s anding o he sys em. Wa e balances and nume ical models can help in dis ega ding po en ial sou ces o conside ing new ones. Anyway, a p e ious concep ual model is necessa y in o de o iden i y he po en ial end-membe s and o e alua e p elimina y alues o he o e all g oundwa e balance. Flow a e and hyd ochemis y o a echa ge sou ce can a y g ea ly o e ime. The e o e, i is w ong o conside a echa ge sou ce as a single componen in he hyd ochemical balance. The a iabili y o he hyd ochemical composi ion o a echa ge sou ce can be cha ac e ized by he in eg a ion 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 Log concen a ion G oundwa e , ppm Log concen a ion Ri e , ppm Majo ions Gene al indica o s Me als T ihalome hanes Polycyclic a oma ic hyd oca bon Alkylphenols Pes icides 1:1 Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 37 o di e en end-membe s. In addi ion, he a iabili y in he composi ion o a echa ge sou ce can be quan i ied om mixing calcula ions. The same could be applied o g oundwa e and i s di e en echa ge sou ce con ibu ions. 4.2.2. Cha ac e iza ion o he hyd o chemical composi ion o po en ial echa ge sou ces and mixed wa e samples Quali a i e analysis Many o he hyd ochemical species in he mixed wa e s can help in iden i ying he sou ce wa e s, bu a p io concep ual model is equi ed bo h o an assessmen o geochemical condi ions and o selec ing he ace ha migh display he b oades ange o concen a ion among he echa ge sou ces. Mos o he hyd ochemical species conside ed canno be used unc i ically because hei beha io migh depend on local geochemical condi ions (e.g., sulpha e would no be conse a i e in educing en i onmen s). Thus, he selec ion o he app op ia e species is a c i ical s ep in mixing calcula ions (Ba e e al., 1999; Choi e al., 2005; Mo is e al., 2006; Ju ado e al., 2013). We ha e o ake in o accoun ha hei concen a ions should also a y widely om one sou ce o ano he . Again, i is necessa y o a p e ious hyd ochemical concep ual model o selec he mos help ul species. Mixing calcula ions a e ep esen a i e o he analysed samples and inco po a e some deg ee o unce ain y. These unce ain ies a e mainly ela ed o sampling, handling and analy ical e o s; lack o da a; and he occu ence o addi ional p ocesses ( eac ions). Mul i a ia e analysis (EMMA) o selec he end-membe s and he species in he mixing calcula ions As s a ed abo e, an EMMA o e s an app oach whe e eigen ec o and esidual analysis o obse ed wa e chemis y es ima es he numbe o con ibu ing end-membe s. An EMMA aims a iden i ying he minimum numbe o end-membe s equi ed o explain he a iabili y o measu ed concen a ions in ime o space. An EMMA helps o manage he deg ee o unce ain y associa ed wi h he concep ual model. The explana ion o he a iance and he con ibu ion o each species o he mix u e a e ob ained om he analysis o in o ma ion p o ided by he calcula ion o he eigen alues. We p opose he ollowing me hodological guidelines o he selec ion o end-membe s and species: a) Ini ia e EMMA wi h all species and ake u he s eps ha will elimina e he species wi h he highes weigh in he eigen alue. b) A common ule is o e ain all eigen ec o s associa ed wi h eigen alues (n) g ea e o equal o 1, called he “ ule o 1” (Hoope , 2003; Jo eskog e al., 1976). This ule Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 38 consis s o selec ing he eigen alues ha explain a iance g ea e han [1/numbe o specie]. c) Selec ion o species acco ding o hose eigen alues ha we wan o explain. The numbe o end-membe s will be n + 1. d) P ojec ion o he obse a ion poin s in he space de ined by he eigen ec o s associa ed wi h each eigen alue. Wi h his p ojec ion, i is possible o make a p elimina y selec ion o end-membe s and an analysis o he poin s ha all ou side he a ea de ined by hese end-membe s. This ype o g aphic is a isual way o in e p e he mixing o all he species in ol ed and o assess he end-membe s. Hence, we can in e ha sou ce o a combina ion o sou ce e ms o p ocesses may explain he mixing. These s eps may iden i y he species and/o combina ions o species ha be e explain he end-membe s selec ed. In addi ion, we also ha e a p e y good idea o how a we can es ablish he sys em. The selec ed end-membe s and species conside ed in he EMMA analysis a e hose used o ini ia ing he mixing calcula ions. 4.2.3. Compu e mixing a ios and e alua ion o he end-membe composi ion To s udy he in luence o wa e s om di e en sou ces, he code MIX (Ca e a e al., 2004) can be used. This so wa e can be eely downloaded om h p://www.h2ogeo.upc.es/so wa e/MIX_PROGRAM/index.h m. MIX is a maximum likelihood me hod o es ima ing mixing a ios whils acknowledging unce ain y in end-membe concen a ions. Maximising he likelihood o concen a ion measu emen s wi h espec o bo h mixing a ios and end-membe concen a ions leads o a gene al cons ained op imisa ion p oblem. The a iables mus sa is y some condi ions o be sui able in mixing models: hey mus exhibi a conse a i e beha io and hei alues mus be signi ican ly di e en in he ex eme componen s. This me hod also allows us o o mula e a hypo hesis abou he geochemical eac ions esponsible o he non-conse a i e beha io in elemen s. MIX code allows he e alua ion o mixing a ios in case o unce ain end membe s using he concen a ions o mixed samples o educe he unce ain y. In his app oach measu emen unce ain y is quan i ied h ough co a iance ma ices, which equi es de ining he eliabili y o measu emen s and he esul s depend on he assumed s anda d de ia ions. This code was p e iously used in se e al s udies (Vàzquez-Suñé e al., 2010; Jiménez-Ma ínez e al., 2011; Mo ales-Casique, 2012; Cáno as e al., 2012a, 2012b; Gal án e al., 2012; Ju ado e al., 2012, 2013). Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 39 4.2.4. Si e desc ip ion The Besòs Ri e , loca ed nea Ba celona in he no heas o Spain, lows sou h-eas e ly and has a ca chmen a ea o 1038 km 2 . The s udy a ea co esponds o he lowe pa o he ca chmen . I is loca ed in no h-eas e n Ba celona, in he Besòs Ri e Del a, a he own o San Ad ià del Besòs, and i is close o he unde g ound pa king a ea a he Plaça de la Vila si e (Fig. 4.2). Figu e 4-2 Loca ion o he s udy a ea and hyd ogeological concep ual model. The clima e is ypically Medi e anean, wi h ex eme empe a u es in Janua y and Augus and a yea ly a e age empe a u e o 15 °C. Rain all a e ages app oxima ely 600 mm/yea . A e age s eam low is 4 m 3 /s a he San a Coloma gauging s a ion (Fig. 4.2) belonging o he Ca alan Wa e Agency (ACA). The gauging s a ion is loca ed app oxima ely 2.5 km ups eam om he Plaça de la Vila si e and 4.5 km om he sea. Ri e low is measu ed daily, whils i e wa e quali y samples a e analysed mon hly by he Ca alan Wa e Agency (ACA). The Besòs Ri e is a good example o a Medi e anean lu ial sys em, wi h low base low (modal low is 2 m 3 /s), ma ked ebb lood, s ong loods (mo e han 100 m 3 /s) and e y i egula low pa e ns, which a e mainly con olled by he ain all e en s (Fig. 4.3). These e en s a y d ama ically om yea o yea and wi hin he yea . Ri e wa e quali y depends on many ac o s. In he inal pa o he ca chmen (s udy a ea), i e wa e could be conside ed o be a mix o wa e om di e en sou ces such as ain all, aw sewe Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemi eac ions in g oundwa e . wa e , ea ed sewe wa e , wa e s om se e al pa s o he ca chmen , e c.; u he mo e, a ia ion in ime is e y signi ican . Figu e 4-3 On he op, chlo ide concen a ions and Besòs Ri e low om 2005 o 2010. A he bo om, e olu ion o g oundwa e e sus i e le el. No e ha he iangles ep esen he 4 g oundwa e sampling campaigns. The geology o he s udy a ea was ca ied ou by sedimen ological s udy o upda e he geological om ci il wo k p ojec s in he dimensional geological s uc u e and o ede ine a la e al connec ions bo h be ween uni s and he sea ( The aqui e s a e o med wi hin Qua e na y sedimen s ha es disco dan ly ma e ials ha ange om Paleozoic o Pliocene. In lu ial aqui e s, consis ing o accumula ions begins a he end o he Lowe Valley and consis s o ou del aic sedimen a y cycles. The younges cycle displays wedge geome y sands and g a els. These geological bodies consis o Shallow Aqui e , which is an uncon ined Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemi sewe wa e , wa e s om se e al pa s o he ca chmen , e c.; u he mo e, a ia ion in ime is e y signi ican . op, chlo ide concen a ions and Besòs Ri e low om 2005 o 2010. A he bo om, e olu ion o g oundwa e e sus i e le el. No e ha he iangles ep esen he 4 g oundwa e sampling campaigns. The geology o he s udy a ea was ca ied ou by MOP (1966). We pe o med a ho ough sedimen ological s udy o upda e he geological model wi h new in o ma ion ob ained om ci il wo k p ojec s in he del a a ea. This allowed us o de ine in de ail he h ee s uc u e and o ede ine a qui e uni s, hei geome ies, and he la e al connec ions bo h be ween uni s and he sea ( Velasco e al., 2012 ). The aqui e s a e o med wi hin Qua e na y sedimen s ha es disco dan ly ma e ials ha ange om Paleozoic o Pliocene. In he lowe alley, he e a e ypical lu ial aqui e s, consis ing o accumula ions o allu ial sand and g a els. The del a i sel he Lowe Valley and consis s o ou del aic sedimen a y cycles. The younges cycle displays wedge geome y wi h sil s and clays sepa a ing sands and g a els. These geological bodies consis o h ee hyd ogeological uni s: 1) which is an uncon ined aqui e o med by sands; 2)Main Aqui e , Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemi cal 40 sewe wa e , wa e s om se e al pa s o he ca chmen , e c.; u he mo e, op, chlo ide concen a ions and Besòs Ri e low om 2005 o 2010. A he bo om, e olu ion o g oundwa e e sus i e le el. No e ha he iangles ep esen he 4 pe o med a ho ough model wi h new in o ma ion ob ained del a a ea. This allowed us o de ine in de ail he h ee qui e uni s, hei geome ies, and he ). The aqui e s a e o med wi hin Qua e na y sedimen s ha es disco dan ly on op o lowe alley, he e a e ypical o allu ial sand and g a els. The del a i sel he Lowe Valley and consis s o ou del aic sedimen a y cycles. The wi h sil s and clays sepa a ing wo packages o h ee hyd ogeological uni s: 1) aqui e o med by sands; 2)Main Aqui e , which Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 41 is made up o he deep aqui e a he Del a; and 3) Sil wedge (aqui a d), which sepa a es he shallow and deep aqui e s. The Besòs aqui e s ha e su e ed hea y wa e ex ac ion since he end he 19 h Cen u y. Ex ac ion p oduced la ge d awdowns, leading o seawa e in usion. Since he 1970s, u ban p essu es ha e caused many indus ies o mig a e away om he ci y. This has esul ed in a educ ion o pumping and a p og essi e eco e y o heads (Vàzquez-Suñé e al., 2005). The i e is hyd aulically connec ed o he Shallow Aqui e a he Del a. The Besòs Ri e 's i egula low pa e n is e lec ed in he well hyd og aphs and he aqui e s espond o i e loods (Fig. 4.3). Se e al echa ge sou ces ha e been iden i ied in he aqui e s o he Besòs Ri e Del a by Vàzquez-Suñé e al. (2010), including: (1) he Besòs Ri e (RIV), (2) ain all (REC), (3) he Te Ri e wa e supply (TER), (4) he Te Ri e sewage wa e (SW_T), and (5) ci y uno (RUNOFF). Among hese echa ge sou ces, wa e om he pollu ed Besòs Ri e may be conside ed he mos impo an echa ge sou ce in he s udy a ea. G oundwa e gene ally lows om he Besòs Ri e o he unde g ound pa king a ea a he Plaça de la Vila si e because he e is a con inuous pumping o app oxima ely 150–200 L/s o a oid seepage p oblems. Fou d ainage pumping wells and 16 piezome e s we e implemen ed o con ol he seepage p oblems in he pa king a ea and o assess g oundwa e head a ia ions in he su oundings a eas (Fig. 4.2). Fu he mo e, his g oundwa e con ol ne wo k is pe iodically sampled o g oundwa e quali y assessmen . We assessed ha almos all he g oundwa e ex ac ed be ween he i e and he pa king d ainage pumping is echa ged om he i e (Ondi iela e al., 2005). 4.2.5. Sampling and analysis Wa e samples we e collec ed om July 2007 o May 2010 in ou di e en ield campaigns in July 2007, Feb ua y 2008, Oc obe 2008 and May 2010. Fou samples om he Besòs Ri e , conside ed he main echa ge sou ce o he aqui e , and 51 samples om he g oundwa e we e collec ed. All he g oundwa e samples we e ob ained by pumping un il a olume ha was a leas h ee imes ha o he piezome e had been ex ac ed. Field pa ame e s we e measu ed in si u including elec ical conduc i i y (EC), pH, empe a u e, Eh and dissol ed oxygen. They we e measu ed con inuously using a low cell o a oid con ac wi h he ai . The ins umen s we e calib a ed daily by means o s anda d solu ions. Samples we e collec ed a e s abilisa ion o ield pa ame e s and we e no il e ed in he ield. They we e s o ed in a ield e ige a o and aken o he labo a o y a he end o he sampling day. Each wa e sample unde wen chemical Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 42 analysis in acco dance wi h he Spanish guidelines o d inking wa e (Royal Dec ee 140/2003). These guidelines include 53 pa ame e s ( ou adioac i i y pa ame e s we e no analysed). All he samples we e analysed a he ATLL (Aigües Te Llob ega ) labo a o y in Ba celona. Chlo ide, sulpha e and ni a e we e analysed using ion ch oma og aphy (IC). Calcium, magnesium, i on and a senic we e analysed by ICP-MS and p io o he analysis, he solu ion was acidi ied wi h 1% ( / ) HNO3 and cen i uged o 3500 pm. Ammonium was analysed by spec opho ome y (based on indophenol blue me hod) and using a comme cial eac i e ki . Bica bona e was analysed manually by chemical e alua ion wi h sulphu ic acid, accoun ing o he pH o he sample. To al o ganic ca bon (TOC) was analysed using he 680 °C combus ion-in a ed me hod wi h a pla inum ca alys and using a Non-Dispe si e In a-Red (NDIR) de ec o . The amoun o o ganic ca bon can be de e mined by he non-pu geable o ganic ca bon (NPOC). I is based on he emo al o ino ganic ca bon om he sample by pu ging he acidi ied sample o a pH less han 2.0 wi h an ine gas. To al o ganic ca bon (TOC) may be de e mined by means o TC measu ing me hod. Mo eo e , mos o he abo e men ioned hyd ochemical pa ame e s o he Besòs Ri e wa e ha e been measu ed mon hly by he Ca alan Wa e Agency (ACA, 2012) a he San a Coloma gauging s a ion (Fig. 4.3). 4.3. Resul s and discussion 4.3.1. Besòs Ri e 4.3.1.1. E alua ion o he numbe o end-membe s and ep esen a i e chemical species Quali a i e analysis Ri e wa e quali y depends on many ac o s. In he inal pa o he ca chmen (s udy a ea), i e wa e could be conside ed o be a mix o wa e om di e en sou ces, such as ain all, aw sewe wa e , ea ed sewe wa e , wa e s om se e al pa s o he ca chmen , e c.; u he mo e, a ia ion o e ime is e y signi ican . As men ioned be o e, up o i e di e en echa ge sou ces ha e been iden i ied in Besòs Ri e Del a aqui e s (RIV, REC, TER, SW_T and RUNOFF). Howe e , p e ious s udies ha e demons a ed ha Besòs Ri e is by a he la ges con ibu o o g oundwa e echa ge, ep esen ing he 90% (Ju ado e al., 2013). The e o e, RIV has been he only echa ge sou ce ha we ha e ake in o accoun . A mo e de ailed discussion o he minimum numbe o end- membe s can be ound in he ollowing s eps. Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 43 Mul i a ia e analysis (EMMA) o selec he end-membe s and species As men ioned be o e, he Besòs Ri e hyd ochemis y da a ha e been ob ained om he su ace wa e quali y moni o ing ne wo k o he Ca alan Wa e Agency (ACA). The selec ed sampling pe iod was om Janua y 2005 o May 2009 and he samples ha e been collec ed on a mon hly basis. The o al numbe o species was 53, bu no all o he species we e analysed e e y ime. Addi ionally, we collec ed ou samples om he Besòs Ri e du ing he ou ield g oundwa e campaigns in July 2007, Feb ua y and Oc obe o 2008, and May 2010 (C1 o C4). In o al, 56 samples om he Besòs Ri e ha e been used o pe o m he EMMA. The species selec ed in he EMMA we e hose wi h mo e da a a ailable in he ACA da abase and we e mainly majo species. The 11 selec ed species we e chlo ide (Cl), sulpha e (SO4), sodium (Na), bica bona e (HCO3), calcium (Ca), magnesium (Mg), po assium (K), ni a e (NO3), ammonium (NH4), o al ni ogen (N o ) and elec ical conduc i i y (EC). Ri e sample eigen ec o e alua ion was pe o med wi h he EMMA accoun ing o he a o emen ioned 11 species. Because we ha e selec ed 11 species in his analysis, an eigen ec o mus explain mo e han 9.1% (1/11) o he da a a iance using he ule o one. Acco ding o ha ule, 3 eigen ec o s should be e ained and hey explained he 90% o he o al a iance. Fi e di e en analyses we e pe o med (Fig. 4.4). Figu e 4-4 Eigen ec o s 1, 2 and 3 esul ing om p ocessing he Besòs Ri e da a wi h EMMA. The analysis “A” includes he species chlo ide, sulpha e, bica bona e, calcium, sodium, magnesium, po assium, ni a e, ammonium, o al ni ogen and elec ical conduc i i y. The analysis “B” is equal o “A” bu wi hou ni a e; “C” is equal o “A” bu wi hou ni a e and magnesium; “D” is equal o “A” bu wi hou ni a e and po assium and “E” is equal o ”A” bu wi hou ni a e, magnesium and po assium. Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 44 Analysis A The i s eigen ec o explained 68% o he a iance (Fig. 4.4, EG1 om analysis A). All he species excep ni a e con ibu e signi ican ly o ha . The species ha con ibu ed mos ( ela i e weigh s anging om 0.39 o 0.26) we e, in descending o de : EC, SO4, Na, Cl, HCO3, Ca, K, Mg, NH4, and N o . In con as NO3, con ibu ed less and wi h he opposi e sign (−0.05). This i s eigen ec o was associa ed wi h wo ypes o wa e om he Ri e Besòs: (1) dilu ed wa e ha co esponds o we pe iods o wa e supply inpu s and (2) concen a ed wa e s ela ed o d y pe iods. Ni a e concen a ion canno be explained wi h hese wo end-membe s. The second eigen ec o addi ionally explained 13% o he o al a iance, which added o he o al explained a iance o 81%. The la ges con ibu ion was p o ided by NO3 (0.59), ollowed by Mg (0.49), NH4 (−0.38), Ca (0.36), K (−0.25) and N o (−0.23). I is impo an o men ion ha he species ha con ibu ed less o he second eigen ec o (weigh s less han abs (0.1)) we e he ones ha con ibu ed mos o he i s eigen ec o . The hi d eigen ec o accoun ed o 10 % o he a iance and was ela ed o he concen a ions o ni ogen species (weigh s we e N o (0.62), NO3 (0.55), and NH4 (0.46)). The emaining species con ibu ed wi h weigh s lowe han 0.17. Analysis B When NO3 was no included in he EMMA (Fig. 4.4, analysis B), a o al a iance o 87 % was ob ained wi h wo eigen ec o s. The i s eigen ec o co esponded o mo e o less dilu ed wa e and ep esen ed an inc ease om 68 % o 74 % o he o al a iance. The second eigen ec o addi ionally explained 13 % o he o al a iance and was ela ed o ni ogen species (weigh s o 0.56 o NH4 and 0.53 o N o ). Mg and Ca species also con ibu ed signi ican ly bu wi h he opposi e sign (−0.48 and −0.35, espec i ely). As in he p e ious analysis, he emaining species con ibu ed wi h weigh s lowe han 0.17 The hi d eigen ec o explained a a iance lowe han 10 % (100 % /10). The species ha con ibu ed mos o his eigen ec o we e Mg (0.48) and K (−0.45) bu wi h he opposi e sign. Analyses C, D and E When NO3 and Mg (Fig. 4.4, analysis C), NO3 and K (Fig. 4.4, analysis D) NO3, Mg and K (Fig. 4.4, analysis E) we e no included in he EMMA; he esul s we e simila o he p e ious analysis B. The i s wo eigen ec o s explained mo e han 85 % o he a iance o he da a in he h ee analyses. The i s eigen ec o dis inguished be ween i e wa e ob ained du ing we and d y pe iods, whils he second eigen ec o was ela ed o he ni ogen concen a ions (mainly ammonium concen a ion). Consequen ly, based on he Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 45 esul s ob ained, we decided o no explain he ni a e concen a ions. The concep ual model o explain he 87 % o i e composi ion co esponds o h ee end-membe s (n + 1 eigen ec o ): EM1 was an ex eme low concen a ion associa ed o ain all pe iods, EM2 and EM3 we e mo e concen a ed ends, dis inguished mainly by high ammonium and low calcium and magnesium. The selec ion o he end membe s was pe o med by plo ing he eigen ec o s (Fig. 4.5). Figu e 4-5 P ojec ion o he eigen ec o s 1 and 2 o he analysis “B”. No e ha he esul ing end-membe s in mixing calcula ions (MIX_1 and MIX_2, Table 4.1) e alua ed wi h EMMA a e also p ojec ed. The di e ence be ween he analysis and MIX_1 and MIX_2 is he a iances assigned o he end- membe s. In he analysis MIX_1 he a iances assigned o each species o he end- membe s a e mul iplied by a ac o o 100, whe eas in analysis MIX_2 hey a e mul iplied by ac o s o 10, 500 and 100, espec i ely, o he End-Membe sW1, D1 and D2. Figu e 4-6 P ojec ion o he eigen ec o s 1, 2 and 3 o he analysis “A” o “E” (Fig. 4.4) esul ing om p ocessing he Besòs Ri e da a wi h EMMA. Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 52 aqui e is mainly echa ged by he i e . Consequen ly, he h ee end-membe s e ained in he EMMA o he i e wa e ha e been used (W1, D1 and D2). I was necessa y o assess whe he g oundwa e concen a ions can be explained by mixing p ocesses o hese end membe s and o quan i y hei p opo ions. The compu ed composi ion o he i e echa ged wa e o he 20 species is shown in Table 4.2 (Analysis “MIX_GW2”). G oundwa e mixing a io e alua ion has been pe o med; he e o e, i was in ended ha he concen a ions o he obse a ion poin s o g oundwa e would no change he p e ious adjus men de e mined in he i e . The e o e, he s anda d de ia ions assigned o he obse a ion poin s (Table 4.2) we e highe han hose assigned o he end- membe s. Mo eo e , he s anda d de ia ions a he g oundwa e obse a ion poin s we e selec ed depending on whe he species we e mo e o less conse a i e. Table 4-2 Resul s e alua ed using MIX o g oundwa e da a (MIX_1GW1 andMIX_2GW2). No e ha he i e end-membe s we e hose e alua ed in he analysisMIX_2, and he a iances assigned o he i e end-membe s a e lowe han hose assigned o he obse a ion poin s because we assumed ha he aqui e is mainly echa ged by he Besòs Ri e . “Ri: A e age concen a ion in Besòs Ri e obse a ion poin s” and “Gi: A e age concen a ion a G oundwa e obse a ion poin s”. 4.3.2.2. Compu e g oundwa e mixing a ios Among he h ee end-membe s conside ed, D2 was by a he la ges con ibu o o he o al echa ge, ep esen ing he 65 %, ollowed by he we i e end-membe (W1, 26 %). The emaining 9 % co esponded o D1. The con ibu ion o W1 was mo e signi ican Analysis MIX_2GW2 EM1:W1 EM2:D1 EM3:D2 Obs. poin s (exp esed (Ri)) Obs. poin s (exp esed (Gi)) EM: W1,D1 & D2 EM1:W1 EM2:D1 EM3:D2 Chlo ide (mg/L) 43.7 356.2 315.8 0.05·Ri 0.05·Gi 0.04·Ri 36.4 358.3 312.0 Elec ical Conduc i i y (µS/cm) 585.6 1996.5 1891.5 0.07·Ri 0.07·Gi 0.03·Ri 586.4 1991.0 1867.8 Sodium (mg/L) 29.0 280.9 246.6 0.07·Ri 0.07·Gi 0.05·Ri 27.2 286.3 251.6 Sulpha e (mg/L) 52.8 205.9 198.6 0.12·Ri 0.11·Gi 0.04·Ri 55.5 201.8 198.1 Bica bona e (mg/L) 222.7 500.5 470.3 0.11·Ri 0.10·Gi 0.07·Ri 247.8 496.6 481.0 Calcium (mg/L) 62.2 133.9 151.3 0.19·Ri 0.16·Gi 0.06·Ri 65.9 134.0 158.4 Magnesium (mg/L) 13.2 24.2 30.1 0.37·Ri 0.28·Gi 0.10·Ri 14.5 24.5 33.0 Po asium (mg/L) 7.2 51.1 36.4 0.50·Ri 0.79·Gi 0.17·Ri 4.6 49.2 26.6 To al ni ogen (mg/L) 6.2 26.1 5.7 0.43·Ri 1·Gi 0.12·Ri 5.7 25.9 4.3 Ni a e (mg/L) 13.4 7.4 13.2 5.54·Ri 15·Gi 0.47·Ri 12.7 7.1 11.5 Ammonium (mg/L) 4.1 31.5 3.6 0.40·Ri 0.90·Gi 0.12·Ri 4.2 31.6 4.0 Phospho us (mg/L) 0.000 2.305 2.012 7328·Ri 27756·Gi 0.48·Ri 0.000 2.305 2.012 Phospha e (mg/L) 1.904 3.818 3.177 29758·Ri 125434·Gi 0.70·Ri 1.904 3.818 3.177 Fluo ide (mg/L) 0.187 1.244 0.374 14193·Ri 11342·Gi 0.57·Ri 0.187 1.244 0.374 To al o ganic ca bon (mg/L) 6.323 12.427 8.663 3283·Ri 8652·Gi 0.21·Ri 6.323 12.427 8.663 A senic (mg/L) 0.003 0.006 0.008 4085·Ri 1970·Gi 0.31·Ri 0.003 0.006 0.008 B omide (mg/L) 0.059 0.57 0.854 5729·Ri 4304·Gi 0.43·Ri 0.059 0.570 0.854 Bo on (mg/L) 0.034 0.218 0.292 5466·Ri 4631·Gi 0.38·Ri 0.034 0.218 0.292 Dissol ed oxygen (mg/L) 8.736 8.267 10.601 3998·Ri 59779·Gi 0.26·Ri 8.736 8.267 10.601 I on (mg/L) 0.036 0.209 0.232 5788·Ri 3751·Gi 0.35·Ri 0.036 0.209 0.232 Inicial Concen a ions Applied s anda de ia ion End membe esul s Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 53 du ing he las ield campaign (C4) because o he equen ain e en s (Fig. 4.3), being he 41 % o he o al echa ge. Accoun ing o he o he 3 campaigns, W1 ep esen ed, on a e age, only he 20 % o he o al echa ge. On a e age, he d y i e end-membe s (D1 and D2) con ibu ed mo e signi ican ly o he o al echa ge han he we end-membe (W1), ep esen ing he pe cen ages o 74 % and 26 %, espec i ely. These esul s a e consis en wi h he ain all e en s ha occu ed in he Besòs Ri e Del a, which con olled he Besòs Ri e low dynamics. The we end-membe is ela ed o sho bu in ense ain all e en s whils he d y end-membe s a e ela ed o he null o low le els o ain all ha occu he es o he yea . This is especially d ama ic in summe whe e Ri e Besòs low 4.3.3. Iden i ica ion o hyd o chemical p ocesses in g oundwa e The p esen ed me hodology could be use ul o iden i y some hyd ogeochemical p ocesses. When compa ing he a e age concen a ions a he Besòs Ri e wi h he a e age concen a ion in he aqui e o se e al species as majo and mino componen s, me als, polycyclic a oma ic hyd oca bons (PAHs), pes icides and su ac an s, i can obse ed ha he concen a ion o a numbe o he species is highe in he i e han in he aqui e (Fig. 4.1). As an example, oxida ion and educ ion p ocesses migh be deduced om Fig. 4.1, whe e ammonium, ni a e and dissol ed oxygen ha e lowe concen a ions in he aqui e . As shown in Fig. 4.7, some species did no beha e conse a i ely because hey di e ged away om he ideal mixing line. These de ia ions could be a ibu ed o hyd ogeochemical p ocesses ha ake place in he i e aqui e sys em. On he one hand, es ima ed concen a ions o ace s such as magnesium and, o a lesse ex en , calcium ha e gene ally mainly comes om was ewa e ea men plan e luen s, which is ep esen ed by he i e end-membe D1. been lowe han measu ed ones (Fig. 4.7). This ac migh be indica i e ha some dissolu ion p ocesses o ca bona es can occu in he aqui e apa om simple mixing o i e wa e . I is also es ima ed ha g oundwa e has an excess o b omide, luo ide, a senic, i on, and bo on o a lesse ex en han he heo e ical mix o he i e . On he o he hand, po assium, dissol ed oxygen, TOC, phospha es, phospho ous and ni a e p esen ed highe es ima ed concen a ions han measu ed ones. Consequen ly, hyd ogeochemical p ocesses migh be deple ing he a o emen ioned species om he aqui e . Dissol ed oxygen, ni a e and TOC may unde go edox p ocesses whils po assium may be exchanged wi h o he ca ions such as calcium o magnesium. 4.4. Conclusions We ha e p esen ed a me hodology o assessing mixing a ios and iden i ying hyd ochemical eac ions in g oundwa e . In his app oach, we use EMMA and MIX Chap e 4: Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . 54 analysis join ly. The me hodology is applied o he assessmen o i e echa ge wa e s in he Besòs Del a aqui e s in Ba celona, Spain. The applica ion o his me hodology in he case o s udy has allowed us o: iden i y he minimum numbe o he i e end-membe s equi ed o explain he a iabili y o measu ed concen a ions. Th ee end-membe s we e inally selec ed: 2 co esponding o d y season (D1 and D2) and one om he we season (W1). The iden i ica ion o end-membe s was used no only o e alua e he i e echa ge in o he aqui e bu also o compu e he expec ed concen a ion o some species, such as phospha es, luo ide, bo on, TOC and b omide. Ri e wa e mixing was he mos ep esen a i e echa ge sou ce in o he aqui e . O e all, d y season end-membe s domina ed o e we season end-membe , in a p opo ion 4:1. Some species did no p ope ly i he mixing line. This sugges s he iden i ica ion o some geochemical p ocesses ha may occu in he i e -aqui e sys em such as edox p ocesses, dissolu ion p ocesses and ion exchange p ocesses. Chap e 5: Conclusions 55 5. Conclusions This chap e includes a summa y o he main con ibu ions o his hesis. The majo indings espond o he o e all aim o gaining a be e unde s anding o he na u e o g oundwa e in u ban a eas. This includes assessing he olume and quali y o aqui e echa ge as well as di e en ia ing wi hin he g oundwa e composi ion be ween concen a ions o compounds om sou ces o echa ge, and concen a ions a ia ions due o hyd ochemical p ocesses. Fi s , a me hodology has been p esen ed o help es ima e he a eal echa ge in u ban aqui e s. I is based on calib a ing a la ge scale model using app op ia ely pa ame e ized a eal echa ge unc ions. These unc ions a e conside ed pe cen ages o wa e suscep ible o echa ging he aqui e s. Wa e s a e classi ied acco ding o hei chemical p ope ies, po en ial ways o en e he g ound, and o igin. This app oach allows g oundwa e managemen o be included in he managemen o o he wa e cycle unc ions in he ci y. A ansien g oundwa e low and solu e anspo model o he Ba celona u ban aqui e s was buil as a ool o help in de ining he managemen p og am. This applica ion has shown he impo ance o echa ge e olu ion on he aqui e s mass balance. Rega ding possible in oduc ion o subs ances o pollu an s om po en ial sou ces o echa ge, a campaign o collec g oundwa e samples h oughou he s udy a ea has been conduc ed. Among o he compounds, subs ances used in he o mula ion o indus ial p oduc s and common domes ic p oduc s we e analyzed. Fo one, ionic su ac an s linea alkylbenzene sul ona e (LAS) ha e been analyzed. LAS and hei deg ada ion p oduc s (sul ophenyl ca boxyla es, SPC) a e no ega ded as p io i y subs ances al hough hey a e oday he mos common domes ic su ac an in he a ea. LAS concen a ions we e de ec ed mo e han one o de o magni ude lowe han expec ed based on echa ge sou ces and in compa ison wi h o he subs ances such as ni a e, ammonium pe chlo e hene o ichlo e hene, in all (oxidizing and educing) aqui e s, which sugges s deg ada ion o hese subs ances unde di e en edox condi ions.. Secondly, nonionic su ac an ex ensi ely used by indus y (and in he pas also in household p oduc s) has been analyzed. Speci ically, alkylphenol polye hoxyla es (APEOs) and hei deg ada ion p oduc s (DPs) nonylphenol polye hoxyla es (NP2EC, NP1EC), oc ylphenol polye hoxyla es (OP2EC, OP1EC), nonylphenol (NP) and oc ylphenol (OP). The oxici y o APEOs is inc eased wi h dec easing e hoxyla e chain leng h. NP and OP ( hei sho es chain) a e conside ed haza dous subs ances. Chap e 5: Conclusions 56 The highes g oundwa e concen a ions o APEOs and DPs we e de ec ed in aqui e s whose majo sou ce o echa ge is a i e ecei ing la ge amoun s o e luen s om seconda y was e wa e ea men plan s (WWTPs). In ac , APEOs DPs concen a ions we e abo e hose in he i e . NP2EC was he compound de ec ed a highes concen a ions. These inc ease wi h ammonium in samples wi h low dissol ed oxygen. These deg ada ion p oduc s we e i ually absen in oxidizing aqui e s whose main sou ce o echa ge is no he i e . In his case, only he ul ima e deg ada ion p oduc (NP) was de ec ed, which sugges s ha pa en compounds ha e deg aded. A me hodology o compu e mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e has also been p esen ed. I is based on using an end-membe mixing analysis (EMMA) and MIX. This app oach has been applied in a Besòs Ri e Del a pilo a ea. Pe o med in he Besòs Ri e he esul s showed ha 3 end-membe s a e equi ed o explain i s empo al a iabili y, accoun ing o he selec ed species. Unde he assump ion ha g oundwa e is a mix o i e wa e , he esul s indica e ha edox p ocesses, ca bona e dissolu ion/p ecipi a ion and ion exchange p ocesses may occu in Besòs Del a aqui e . Also ela i e excesses and de aul s o ce ain compounds ha e been quan i ied in g oundwa e (e.g. b omide, luo ide, bo on, a senic, and TOC, phospho ous, phospha e, espec i ely). Chap e 5: Conclusions 57 O he scien i ic pape s published by in e na ional jou nals ela ed o his hesis a e: • Vàzquez-Suñé E., Ca e a J., Tubau I., Sánchez-Vila X., Sole A., 2010. An app oach o iden i y u ban g oundwa e echa ge. Hyd ology and Ea h Sys em Sciences 14 (2010) 2085-2097. h p://dx.doi.o g/10.5194/hess-14-2085-2010 • Velasco V., Cabello P., Vàzquez-Suñé E., López-Blanco M., Ramos E., Tubau I., 2012. A s a ig aphic sequence based geological model o cons aining hyd ogeological modeling in he u banized a ea o he Qua e na y Besòs Del a (NW Medi e anean Coas , Spain). Geol Ac a 2012;10(4):373–93. h p://dx.doi.o g/10.1344/105.000001757 • Ju ado A., Vàzquez-Suñé E., Sole A., Tubau I., Ca e a J., Pujades E., Anson I., 2013. Applica ion o mul i-iso ope da a (O, D, C and S) o quan i y edox p ocesses in u ban g oundwa e . Applied Geochemis y (2013) 114-125. h p://dx.doi.o g/10.1016/j.apgeochem.2013.02.018 • Velasco V., Tubau, I., Vàzquez-Suñé E., Gogu R. Gai ana u D.. Alca az M., Se ano-Juan A., Fe nández-Ga cía D., Ga ido, T., F aile, J., Sánchez-Vila, X., 2014. GIS-based hyd ogeochemical analysis ools(QUIMET). Compu e & Geosciencies 70 (2014) 164-180. • Ju ado, A., Vázquez-Suñé, E., Ca e a, J., Tubau, I, and Pujades, E. 2015. Quan i ying chemical eac ions by using mixing analysis. Science o he To al En i onmen 502 (2015) 448–456. h p://dx.doi.o g/10.1016/j.sci o en .2014.09.036 The abs ac s o hese pape s a e a ached a he end o his documen , as well as pape s in he chap e s on which his hesis is based: • Tubau I., Vázquez-Suñé E., Ca e a J., González S., Pe o ic M., López de Alda M.J., Ba celó D., 2010. Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g ound wa e : Ba celona case s udy. Jou nal o Hyd ology 383 (2010) 102-110. • Tubau I., Vázquez-Suñé E., Ju ado A., Ca e a J., 2014. Using EMMA and MIX analysis o asses mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . Science o The To al En i onmen 470-471 (2014) 1120-1131. • Tubau I, Vázquez-Suñé E, Ca e a J, Valhondo, C. Quan i ica ion o G oundwa e Recha ge in U ban En i onmen s. Science o he To al En i onmen (submi ed). Chap e 6: Re e ences _______________________________________________________________________________ 58 6. Re e ences Ahel M., Gige W. Ma kus K., 1994a. Beha iou o alkykphenol polye hoxyla e su ac an s in he aqua ic en i onmen -I. Occu ence and ans o ma ion in sewage ea men . Wa e Resea ch 28 (5), 1131–1142. Ahel M., Gige W., Scha ne C., 1994b. Beha iou o alkykphenol polye hoxyla e su ac an s in he aqua ic en i onmen -II. Occu ence and ans o ma ion in i e s. Wa e Resea ch 28 (5), 1143–1152. Ahel M., Gige W., Scha ne C., 1996. Beha iou o alkykphenol polye hoxyla e su ac an s in he aqua ic en i onmen -III. Occu ence and elimina ion o hei pe sis en me aboli es du ing in il a ion o i e wa e o g ound wa e . Wa e Resea ch 30 (1), 37–46. Cus odio E., 1997. G oundwa e quan i y and quali y changes ela ed o land and wa e managemen a ound u ban a eas: blessings and mis o unes. G oundwa e in he U ban En i onmen : P oblems, P ocesses and Managemen . Balkema: 11- 22. Cus odio E., 2004. Hid ogeología u bana: una nue a ama de la ciencia hid ogeológica. Bole ín Geológico y Mine o, 115, Núm. especial: 283-288 ISSN: 0366-0176 283. A andes R., 2004. El plan pa a la eu ilización de las aguas del subsuelo de Ba celona po los se icios municipales: un ema pa a la e lexión y una apues a po la sos enibilidad. El Agua y la Ciudad Sos enible: Hid ogeología U bana. AIH-GE. IGME, Mad id. Se ie H y AG, 11: 45-60. Ba be LB., Thu man EM., Sch oede M.P., 1988. Long- e m a e o o ganic mic opollu an s in sewage-con aminan ed g oundwa e . En i onmen al Science and Technology 22, 205–211. Ba be L.B., Lee K.E., Swackhame D.L., Schoen uss H.L., 2007. Rep oduc i e esponses o male a head minnows exposed o was ewa e ea men plan e luen , e luen ea ed wi h XAD8 esin, and an en i onmen ally ele an mix u e o alkylphenol compounds. Aqua ic Toxiclogy 82, 36–46. Ba nes K.K., Kolpin D.W., Fu long E.T., Zaugg S.D., Meye M.T., Ba be L.B., 2008. A na ional econnaissance o pha maceu icals and o he o ganic was ewa e con aminan s in he Uni ed S a es-(I) g oundwa e . Science o he To al En i onmen 402, 192–200. Ba e M, Hiscock K.M., Pedley S, Le ne D.N., Tellam J.H. and F ench M,J., 1999. Ma ke species o iden i ying u ban g oundwa e echa ge sou ces: a e iew and case s udy in No ingham, UK. Wa e Res 1999;33(14):3083–97. Be na J.L., Cassani G., Hage C.D., Rehman N., López I., Schowanek D., S ebe J., Taege K., Wind T., 2007. Anae obic biodeg ada ion o su ac an s scien i ic e iew. Tenside Su ac an s De e gen s 44 (6), 312–347. <www. sdjou nal.com> Documen Numbe : TS100351. Bickle M, Chadwick A, Huppe HE, Hallwo h M, Lyle S., 2007. Modelling ca bon dioxide accumula ion a Sleipne : implica ions o unde g ound ca bon s o age. Ea h Plane SciLe 2007;255(1–2):164–76. B unne P.H., Cap i S., Ma comini A., Gige W., 1988. Occu ence and beha iou o linea alkylbenzenesulphona es, nonylphenol, nonylphenol mono- and nonylphenol die hoxyla es in sewage and sewage sludge ea men . Wa e Resea ch 22 (12), 1465–1472. Cáno as C.R., Olías M., Sa mien o A.M., Nie o J.M., Gal án L., 2012a. Pollu an anspo p ocesses in he Odiel Ri e (SW Spain) du ing ain e en s. Wa e Resou Res 2012a;48. Cáno as C.R., Olias M., Vàzquez-Suñé E., Ayo a A., Nie o J.M., 2012b. In luences o eleases om a esh wa e ese oi on he hyd ochemis y o he Tin o i e (SW Spain). Sci To alEn i on 2012b;416:418–28. Ca e a J., Alcolea A., Medina A., Hidalgo J. and Sloo en L.J., 2005. In e s p oblem in hyd ogeology. Hyd ogeology Jou nal (2005) 13:206-222. DOI 10.1007/s10040-004-0404-7 Ca e a J., Vàzquez-Suñé E., Cas illo O., Sanchez-Vila X., 2004. A me hodology o compu e mixing a ios wi h unce ain end-membe s. Wa e Resou Res 2004;40. Chap e 6: Re e ences _______________________________________________________________________________ 59 Ca e a J. and Neumann S.P., 1986. Es ima ion o aqui e pa ame e s unde s eady-s a e and ansien condi ions: I. Backg ound and S a is ical amewo k. Wa e Resou ces Resea ch, 22 (2), 199-210. Céspedes R., Laco e S., Raldúa D., Gineb eda A., Ba celó D., Piña B., 2005. Dis ibu ion o endoc ine dis up o s in he Llob ega Ri e basin (Ca alonia, NE Spain). Chemosphe e 61, 1710–1719. Chil on, J. (ed.), 1999. G oundwa e in he U ban En i onmen . Selec ed ci y p o iles. Balkema, 341p. Chil on, J. (ed.), 1997. G oundwa e in he U ban En i onmen . P oblems, P ocesses and Managemen , XXVII Cong ess o he In e na ional Associa ion o Hyd ogeologis s (IAH), No ingham 21-27 Sep embe 1997. Balkema, 682 p. Chisala B. N. and Le ne D. N., 2008. Dis ibu ion o sewe ex il a ion o u ban g oundwa e . Wa e Managemen 161 Issue WM6. Ch is ophe sen N, Hoope R., 1992. Mul i a ia e-analysis o s eam wa e chemical-da a — he use o p incipal componen s-analysis o he end-membe mixing p oblem. Wa e Resou Res 1992;28(1):99–107. Ch is ophe sen N., Neal C., Hoope R., Vog R., Ande sen S., 1990. Modeling s eamwa e chemis y as a mix u e o soilwa e end-membe s — a s ep owa ds 2nd-gene a ion acidi ica ion models. J Hyd ol 1990;116(1–4):307–20. Choi B., Yun S., Yu S., Lee P., Pa k S., Chae G., 2005. Hyd ochemis y o u ban g oundwa e in Seoul, Sou h Ko ea: e ec s o land-use and pollu an echa ge. En i on Geol 2005;48(8):979–90. Council o he Eu opean Union., 1998. Council Di ec i e 98/83/EC o 3 No embe 1998 on he Quali y o Wa e In ended o Human Consump ion. O icial Jou nal o he Eu opean Communi ies L 330/32. h p://eu -lex.eu opa.eu/LexU iSe /LexU iSe .do?u i=OJ:L:1998:330:0032:0054:EN:PDF. Cox M.H., Su G.W. and Cons an z J., 2007. Hea , chlo ide and speci ic conduc ance as g ound wa e ace s nea s eams, G oundWa e , 45(2), 187-195, 2007. Dahan O., McG aw D., Ada E., Pohll G., Bohm B., Thomas J., 2004: Mul i- a iable mixing cell model as a calib a ion and alida ion ool o hyd ogeologic g oundwa e modeling, J. Hyd ol., 293(1–4), 115–136, 2004. De V ies J. and Simme s I., 2002. G oundwa e echa ge: an o e iew o p ocesses and challenges. Hyd ogeology Jou nal 10, 8–15. De Buen H. 2009. Model hid ogeològic en pe il de lux, anspo de solu conse a iu i anspo de calo del iu Besòs a l“alçada de la Plaça de la Vila de San Ad ià de Besòs (G oundwa e modelling o low, conse a i e solu e and hea anspo in he Besòs Ri e Del a aqui e in San Ad ià de Besòs). [Mas e Thesis, 88 pp.]Ba celona, Spain: Tech. Uni . o Ca alonia; 2009. Diaz A., Ven u a F., Galce an M.T., 2002. Simul aneous de e mina ion o es ogenic sho e hoxy chain nonylphenols and hei acidic me aboli es in wa e by an insample de i a iza ion/solidphase mic oex ac ion me hod. Analy ical Chemis y 74 (15), 3869–3876. Di Co cia A., Cons an ino A., C escenzi C., Ma inoni E., Sampe i R., 1998. Cha ac e iza ion o ecalci an in e media es om bio ans o ma ion o he b anched alkyl side chain o nonylphenol e hoxyla e su ac an s. En i onmen al Science and Technology 32 (16), 2401– 2409. Eiswi h M., Wol L., Hö zl H., 2004. Balancing he con aminan inpu in o u ban wa e esou ces. En i onmen al Geology (2004) 46:246–256. Ejle sson J., Nilsson M.L., Kylin H., Be gman K., Ka lson L., Quis M., S ensson BH., 1999. Anae obic deg ada ion o nonylphenol mono- and die hoxyla es in diges o sludge, Land illed municipal solid was e, and Land illed sludge. En i onmen al Science and Technology 33 (2), 301–306. Chap e 6: Re e ences _______________________________________________________________________________ 60 Ellis J.B., 2006. Pha maceu ical and pe sonal ca e p oduc s (PPCPs) in u ban ecei ing wa e . En i onmen al Pollu ion 144 (2006) 184-189. doi:10.1016/j.en pol.2005.12.018 Ellis B., 1999. Impac s o U ban G ouw h on Su ace Wa e and G oundwa e Quali y. IAHS nº259, 437 p. En i onmen Canada, 1999. P io i y Subs ances Lis Assessmen Repo . Nonylphenol and i s E hoxyla es. Canadian En i onmen al P o ec ion Ac , CEPA 1999. <h p://www.hc- sc.gc.ca/ewh-sem /al _ o ma s/hecs-sesc/pd /pubs/con aminan s/psl2- lsp2/nonylphenol/nonylphenol-eng.pd >. EPA, 2001. “Remo al o Endoc ine Dis up o Chemicals Using D inking Wa e T ea men P ocesses”. EPA/625/R-00/015. S a es En i onmen al P o ec ion Agency, O ice o Resea ch and De elopmen . Washing on, DC. h p://www.epa.go . Eu opean Pa liamen and Council o he Eu opean Union, 2008. Di ec i e 2008/105/EC o he Eu opean Pa liamen and o he Council o 16 Decembe 2008 on En i onmen al Quali y S anda ds in he Field o Wa e Policy, Amending and Subsequen ly Repealing Council Di ec i es 82/176/EEC, 83/513/EEC, 84/156/ Eu opean Commi ee o O ganic Su ac an s and hei In e media es (CESIO). 2007. Su ac an S a is ics o Wes e n Eu ope Yea 2002. <www.ce ic.o g>. Eu opean Pa liamen and Council o he Eu opean Union. 2006. Di ec i e 2006/118/EC o he Eu opean Pa liamen and o he Council o 12 Decembe 2006 on he P o ec ion o G oundwa e agains Pollu ion and De e io a ion. O icial Jou nal o he Eu opean Union L 372/19. <h p://eu - lex.eu opa.eu/LexU iSe /LexU iSe .do?u i=OJ:L:2006:372:0019:0031:EN:PDF>. Eu opean Pa liamen and Council o he Eu opean Union. 2000. Di ec i e 2000/60/EC o he Eu opean Pa liamen and o he Council es ablishing a amewo k o he Communi y ac ion in he ield o wa e policy, on 23 Oc obe 2000. (in sho , he EU Wa e F amewo k Di ec i e o WFD). h p://ec.eu opa.eu/en i onmen /wa e /pa icipa ion/no es_en.h m Field J.A., Leenhee J.A., Tho n K.A., Ba be II L.B., Ros ad C., Macalady D.L., Daniel S.R., 1992. Iden i ica ion o pe sis en anionic su ac an -de i ed chemicals in sewage e luen and g oundwa e . Jou nal o Con aminan Hyd ology 9, 55–78. F eni G., Mannina G., Vi iani G., 2009. Unce ain y assessmen o an in eg a ed u ban d ainage model. Jou nal o Hyd ology 373 (2009) 392–404. F iedel, M.J., 2006. Reliabili y in es ima ing u ban g oundwa e echa ge h ough he adose zone. Managing Sus ainable De elopmen in A id and Semia id Regions. J.H. Tellam e al. (eds.), U ban G oundwa e Managemen and Sus ainabili y, 169–182. 2006 Sp inge . Gal án L, Olías M, Cáno as CR, To es E, Ayo a C, Nie o JM., 2012. Re ining he es ima ion o me al loads dissol ed in acid mine d ainage by con inuous moni o ing o speci ic conduc i i y and wa e le el. Appl Geochem 2012;27(10):1932–43. GHS. 2004. Visual T ansin Code. Hyd ogeology G oup (GHS), Depa men o Geo hecnical Enginee ing and Geosciences, Uni e si a Poli ècnica de Ca alunya (UPC),Ba celona. Gige W., B unne P.H., Scha ne C., 1984. 4-nonylphenol in sewge sludge: accumula ion o oxic me aboli es om nonionic su ac an s. Science 225 (4662), 623–625. González S., Pe o ic M., Ba celó,D., 2004. Simul aneous ex ac ion and a e o linea alkylbenzene sul ona es, coconu die hanol amides, nonylphenol e hoxyla es and hei deg ada ion p oduc s in was ewa e ea men plan s, ecei ing coas al wa e s and sedimen s in he Ca alonian a ea (NE Spain). Jou nal o Ch oma og aphy A 1052, 111–120. González S., Pe o ic M., Ba celó D., 2008. E alua ion o wo pilo scale memb ane bio eac o s o he elimina ion o selec ed su ac an s om municipal was ewa e s. Jou nal o Hyd ology 356, 46–55. Healy R.W. 2010. Es ima ing g oundwa e echa ge. Camb idge Uni e si y P ess, Camb idge, UK. Healy R.W. and Cook P.G. 2002. Using g oundwa e le els o es ima e echa ge. Hyd ogeology Jou nal (2002) 10:91-109. Chap e 6: Re e ences _______________________________________________________________________________ 61 Hoope R.P., 2003. Diagnos ic ools o mixing models o s eam wa e chemis y. Wa e Resou Res 2003;39(3):1055. h p://dx.doi.o g/10.1029/2002WR001528. Hoope R.P., Ch is ophe sen N., Pe e s N.E., 1990 Modelling s eam wa e chemis y as a mix u e o soil wa e end membe s—an applica ion o he Panola Moun ain Ca chmen , Geo gia, USA. J Hyd ol. 1990;116:321–43. Howa d K.W.F. and Is a ilo R., 2002. Cu en p oblems o hyd ogeology in u ban a eas, u ban agglome a es and indus ial cen es. NATO Science Se ies IV, Ea h and En i onmen al Sciences, Kluwe . Howa d, K.W.F., Eyles, N., Li ings one, S., 1996. Municipal land illing p ac ice and i s impac on g oundwa e esou ces in and a ound u ban To on o, Canada. Hyd ogeology Jou nal 4 (1), 64–79. Hussein, M. and Schwa z, F., 2003. Modeling o low and con aminan anspo in coupled s eam aqui e sys ems. J.Con am.Hydol., 65(1-2), 41-64, 2003. James A.L. and Roule N.T. 2006. In es iga ing he applicabili y o end-membe mixing analisis (EMMA) ac oss scale: a s udy o eigh small, nes ed ca chmen s in a empe a e o es ed wa e shed. Wa e Resou Res 2006;42(8):W08434. h p://dx.doi.o g/10.1016/S0169- 7722(02)00229-2. Jiménez-Ma ínez J, A a ena R, Candela J. 2011. The ole o leaky bo eholes in he con amina ion o a egional con ined aqui e . A case s udy: he Campo de Ca agena Region, Spain. Wa e Ai Soil Pollu 2011;215:311–27. Jones J, Sudicky E, B ook ield A, Pa k Y. 2006. An assessmen o he ace -based app oach o quan i ying g oundwa e con ibu ions o s eam low. Wa e Resou Res 2006;42(2):W02407. h p://dx.doi.o g/10.1029/2005WR004130. Jonke s N., Kneppe T.P., Voog P., 2001. Ae obic biodeg ada ion s udies o nonylphenol e hoxyla es in i e wa e using liquid ch oma og aphy_elec osp ay andem mass spec ome y. En i onmen al Science and Technology 35 (2), 335–340. Jo eskog K.G., Klo an J.E., Reymen R.A. 1976., Geological ac o analysis (me hods in geoma hema ics 1). Ams e dam: Else ie ; 1976 [178 pp.]. Ju ado A, Vàzquez-Suñé E., Sole A., Tubau I., Ca e a J., Pujades E., Anson, I., 2013. Applica ion o mul i-iso ope da a (O, D, C and S) o quan i y edox p ocesses in u ban g oundwa e . Applied Geochemis y (2013) 114-125. h p://dx.doi.o g/10.1016/j.apgeochem.2013.02.018. Ju ado A., Vázquez-Suñé E., Ca e a J., López de Alda M., Pujades E., Ba celó D., 2012a. Eme ging o ganic con aminan s in g oundwa e in Spain: A e iew o sou ces, ecen occu ence and a e in a Eu ope con ex . Science o The To al En i onmen 440 (2012) 82- 94. Ju ado A, Mas oianni N, Vàzquez-Suñé E, Ca e a J, Tubau I, Pujades E., 2012b. D ugs o abuse in u ban g oundwa e . A case s udy: Ba celona. Science o The To al En i onmen (2012) 424: 280–8. Kime le R.A. and Swishe R.D., 1977. Reduc ion o aqua ic oxici y o linea alkylbenzene sul ona e (LAS) by biodeg ada ion. Wa e Resea ch 11, 31–37. K uege C., Radako ich K., Sawye T., Ba be L.B., Smi h R.L., Field J., 1998. Biodeg ada ion o he su ac an linea alkylbenzenesul ona e in sewagecon amina ed g oundwa e : a compa ison o column expe imen s and ield ace es s. En i onmen al Science and Technology 32, 3954–3961. La son R.J. and Games L.M., 1981. Biodeg ada ion o linea alcohol e hoxyla es in na u al wa e s. En i onmen al Science and Technology 15 (12), 1488–1493. Occu ence and a e o alkylphenol polye hoxyla e deg ada ion p oduc s and linea alkylbenzene sul ona e su ac an s in u ban g ound wa e : Ba celona case s udy Isabel Tubau a,* , En ic Vázquez-Suñé b , Jesús Ca e a b , Susana González b,c , Mi a Pe o ic b,d , Ma ía J. López de Alda b , Damià Ba celó b,e a Depa men o Geo echnical Enginee ing and Geosciences, Uni e si a Poli ècnica de Ca alunya (UPC), C/ Jo di Gi ona 1-3, Edi ici D-2, 08034-Ba celona, Spain b Ins i u e o En i onmen al Assessmen and Wa e Resea ch (IDÆA), C/ Jo di Gi ona 18-26, 08034-Ba celona, Spain c Cen o Tecnológico del Agua (CETaqua), Passeig dels Tille s 3, 08034-Ba celona, Spain d Ins i ució Ca alana de Rece ca i Es udis A ança s (ICREA), Passeig Lluis Companys 23, 08010-Ba celona, Spain e Ca alan Ins i u e o Wa e Resea ch (ICRA), C/ Emili G ahi 101, Edi ici H2O, Pa c Cien í ic i Tecnològic de la Uni e si a de Gi ona, 17003-Gi ona, Spain a i c l e i n o Keywo ds: U ban g ound wa e Alkylphenol polye hoxyla es (APEOs) deg ada ion p oduc s (DPs) Linea alkylbenzene sul ona es (LAS) Ci y o Ba celona s u m m a y This s udy in es iga es he a e o alkylphenol polye hoxyla es (APEOs) deg ada ion p oduc s (DPs) and he occu ence o linea alkylbenzene sul ona e (LAS) su ac an s in u ban g ound wa e a ield scale. The occu ence o APEOs DPs in g ound wa e was s udied in connec ion wi h: (1) sou ces o echa ge o /and pollu ion con aining hese subs ances, (2) g ound wa e edox condi ions, (3) occu ence o LAS, which a e cu en ly he domes ic su ac an s mo e used in he s udy a ea and (4) o he common con aminan s in u ban g ound wa e in he ci y o Ba celona. The APEOs DPs analyzed included wo nonylphenol ca - boxyla es (NP2EC, NP1EC), wo oc ylphenol ca boxyla es (OP2EC, OP1EC), nonylphenol (NP) and oc yl- phenol (OP). The highes g oundwa e concen a ions o APEOs DPs we e de ec ed in aqui e s whose majo sou ce o echa ge is a i e ecei ing la ge amoun s o e luen s om seconda y was e wa e ea - men plan s (WWTPs). In ac , APEOs DPs concen a ions we e abo e hose in he i e . NP2EC was he compound de ec ed a highes concen a ions. These inc ease wi h ammonium in samples wi h low dis- sol ed oxygen. These deg ada ion p oduc s we e i ually absen in oxidizing aqui e s whose main sou ce o echa ge is no he i e . In his case, only he ul ima e deg ada ion p oduc (NP) was de ec ed, which sugges s ha pa en compounds ha e deg aded. These esul s indica e ha APEOs a e pe sis en o less deg aded in educing condi ions, whe eas hey a e deg aded when oxidizing condi ions p e ail. By con as , LAS concen a ions we e mo e han one o de o magni ude lowe han expec ed based on echa ge sou ces in all (oxidizing and educing) aqui e s. Ó2009 Else ie B.V. All igh s ese ed. In oduc ion High popula ion g ow h inc eases he demand o wa e and he gene a ion o was ewa e . G ound wa e is an impo an e- sou ce o wa e supply in he wo ld and aqui e s in u ban zones a e ulne able o con amina ion by a a ie y o con aminan s due o u ban ac i i ies. App oxima ely 90% o he wa e o supply comes om impo ed su ace wa e in he ci y o Ba celona, which is he s udy a ea. Since supply sou ces a e subjec o pe iodic d ough s, he Ca alan Wa e Agency (ACA) is seeking al e na i e wa e esou ces. The e a e a numbe o aqui e s wi h a ying li hologies and ages below he ci y (Váquez-Suñé e al., 2005). The e o e, g ound wa e mus be conside ed as an op ion. The ques ion is whe he u ban g ound wa e can be used sa ely. U ban aqui e s may su e pollu ion om di e en sou ces, including losses om wa e supply ne wo ks, leakage om sewe s, seepage om i e s o o he su ace wa e bodies, sea- wa e in usion, e c. As a esul , a la ge numbe and a ie y o con aminan s may be ound in u ban aqui e s (Howa d e al., 1996; Ba e e al., 1999; Le ne , 2002; Mo is e al., 2005; Váquez-Suñé e al., 2000; Wakida and Le ne , 2005; Naik e al., 2008). Some o hese con aminan s a e subjec o en i- onmen al quali y s anda ds (EQS) in d inking wa e (Council o he Eu opean Union, 1998), g oundwa e (Eu opean Pa lia- men and Council o he Eu opean Union, 2006), and/o su ace wa e (Eu opean Pa liamen and Council o he Eu opean Un- ion, 2008). O he was ewa e subs ances, also known as eme g- ing con aminan s, whose oxici y, occu ence and pe sis ence a e less well known, a e subjec o e iew o possible iden i- ica ion as p io i y subs ances o p io i y haza dous subs ances (Eu opean Pa liamen and Council o he Eu opean Union, 2008). 0022-1694/$ - see on ma e Ó2009 Else ie B.V. All igh s ese ed. doi:10.1016/j.jhyd ol.2009.11.030 *Co esponding au ho . Tel.: +34 934011820; ax: +34 934017251. E-mail add ess: [email p o ec ed] (I. Tubau). Jou nal o Hyd ology 383 (2010) 102–110 Con en s lis s a ailable a ScienceDi ec Jou nal o Hyd ology jou nal homepage: www.else ie .com/loca e/jhyd ol Using EMMA and MIX analysis o assess mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e Isabel Tubau a,b , En ic Vàzquez-Suñé c, ⁎, Anna Ju ado a,c , Jesús Ca e a c a GHS, Dep Geo echnical Enginee ing and Geosciences, Uni e si a Poli ecnica de Ca alunya, UPC-Ba celona Tech, Jo di Gi ona 1-3, 08034 Ba celona, Spain b Geological Ins i u e o Ca alonia, Balmes 209-211, 08006 Ba celona, Spain c GHS, Ins i u e o En i onmen al Assessmen & Wa e Resea ch (IDAEA), CSIC, Jo di Gi ona 18-26, 08034 Ba celona, Spain H I G H L I G H T S •App oach using EMMA and MIX o asses mixing a ios in g oundwa e . •Pollu ed i e echa ge in u ban aqui e s. •Th ee end-membe s a e e ained o explain empo al a iabili y o i e wa e . •Depa u es om mixing lines migh ep esen he occu ence o p ocesses. a b s a c a i c l e i n o A icle his o y: Recei ed 23 July 2013 Recei ed in e ised o m 30 Oc obe 2013 Accep ed 30 Oc obe 2013 A ailable online xxxx Keywo ds: Mixing analysis EMMA MIX U ban g oundwa e This s udy p esen s a me hodology using an end-membe mixing analysis (EMMA) and MIX o compu e mixing a ios and o iden i y hyd ochemical eac ions in g oundwa e . The me hodology consis s o (1) iden i ying he po en ial sou ces o echa ge, (2) cha ac e ising echa ge sou ces and mixed wa e samples using hyd ogeochemis y, (3) selec ing chemical species o be used in he analysis and (4) calcula ing mixing a ios and iden ifica ion o hyd ochemical eac ions in g oundwa e . This app oach has been applied in he Besòs Ri e Del a a ea, whe e we ha e collec ed 51 g oundwa e samples and a long da a egis e o he hyd ogeochemis y o he Besòs Ri e c ea ed by he Ca alan Wa e Agency is also a ailable. The EMMA pe - o med in he Besòs Ri e sugges s ha 3 end-membe s a e equi ed o explain i s empo al a iabili y, accoun - ing o he species chlo ide, sulpha e, sodium, bica bona e, calcium, magnesium, po assium, ammonium, o al ni ogen, and elec ical conduc i i y. One i e end-membe is om he we pe iods (W1), and wo a e om d y pe iods (D1 and D2). These end-membe s ha e been used o compu e mixing a ios in g oundwa e samples because he Besòs Ri e is conside ed he main echa ge sou ce o he aqui e . O e all, d y season end-membe s domina ed o e he we season end-membe , in a p opo ion o 4:1. Mo eo e , when depa u es om he mixing line exis , geochemical p ocesses migh be iden ified. Redox p ocesses, ca bona e dissolu ion/p ecipi a ion and ion exchange p ocesses may occu in Besòs Del a aqui e . © 2013 Else ie B.V. All igh s ese ed. 1. In oduc ion The de e io a ion o g oundwa e quali y has become a majo issue in u ban aqui e s. In u ban a eas, a as a ay o con aminan s may be ound because hey a e in oduced in o he aqui e h ough di e en e- cha ge sou ces. Thus, he assessmen o g oundwa e quali y equi es iden i ying no only he possible ocus o con amina ion bu also he p ocesses ha he pollu an s unde go. In ac , his s udy is mo i a ed by he assessmen o g oundwa e quali y a he Besòs Ri e allu ial aqui e unde nea h Ba celona, Spain. G oundwa e quali y in his a ea is significan ly be e han i e wa e quali y, despi e he ac ha pumped g oundwa e is essen ially echa ged om he i e . A mass balance o chemical species is o en used in hyd ology and e- la ed sciences o aid in he e alua ion o wa e balances. I s simples o m consis s o measu ing he concen a ions o any conse a i e spe- cies occu ing in inflowing and ou flowing wa e s and, assuming con i- nui y, es ablishing he mass balance o solu e and wa e . Howe e , he join e alua ion o echa ge a es and quali y can mos likely be bes achie ed h ough flow and solu e anspo modelling when long head and concen a ion eco ds a e a ailable. Because hese models e- qui e he managemen o much in o ma ion, simple me hodologies a e necessa y. Mixing calcula ions a e an al e na i e o complex models and can be used o be e es ima e he ela i e impo ance o di e en sou ces o echa ge. Mixing calcula ions include (1) End-Membe Mixing Analysis Science o he To al En i onmen 470–471 (2014) 1120–1131 ⁎Co esponding au ho a : Jo di Gi ona 18, 08034 Ba celona, Spain. Tel.: +34 934006100; ax: +34 932045904. E-mail add ess: en ic. [email protected] (E. Vàzquez-Suñé). 0048-9697/$ –see on ma e © 2013 Else ie B.V. All igh s ese ed. h p://dx.doi.o g/10.1016/j.sci o en .2013.10.121 Con en s lis s a ailable a ScienceDi ec Science o he To al En i onmen jou nal ho me page : www . else ie . com/ loca e/sc i o en Hyd ol. Ea h Sys . Sci., 14, 2085–2097, 2010 www.hyd ol-ea h-sys -sci.ne /14/2085/2010/ doi:10.5194/hess-14-2085-2010 © Au ho (s) 2010. CC A ibu ion 3.0 License. Hyd ology and Ea h Sys em Sciences An app oach o iden i y u ban g oundwa e echa ge E. V´ azquez-Su˜ n´ e1, J. Ca e a1, I. Tubau1,2, X. S´ anchez-Vila2, and A. Sole 3 1Ins i u e o En i onmen al Assessmen and Wa e S udies, Spanish Na ional Resea ch Council, IDAEA-CSIC, Ba celona, Spain 2Depa men o Geo echnical Enginee ing and Geosciences, Technical Uni e si y o Ca alonia, UPC Ba celona Tech, Ba celona, Spain 3G up de Mine alogia Aplicada i Medi Ambien , Dep. C is allog afia, Mine alog´ ıa i Dip` osi s Mine als, Facul a de Geologia, Uni e si a de Ba celona, Ba celona, Spain Recei ed: 25 Feb ua y 2010 – Published in Hyd ol. Ea h Sys . Sci. Discuss.: 22 Ap il 2010 Re ised: 4 Oc obe 2010 – Accep ed: 8 Oc obe 2010 – Published: 27 Oc obe 2010 Abs ac . E alua ing he p opo ion in which wa e s om di e en o igins a e mixed in a gi en wa e sample is ele an o many hyd ogeological p oblems, such as quan i ying o- al echa ge, assessing g oundwa e pollu ion isks, o man- aging wa e esou ces. Ou wo k is mo i a ed by u ban hy- d ogeology, whe e wa e s wi h di e en chemical signa u e can be iden ified (losses om wa e supply and sewage ne - wo ks, infil a ion om su ace uno and o he wa e bod- ies, la e al aqui e s inflows, ...). The ela i e con ibu ion o di e en sou ces o o al echa ge can be quan ified by means o solu e mass balances, bu applica ion is hinde ed by he la ge numbe o po en ial o igins. Hence, he need o inco - po a e da a om a la ge numbe o conse a i e species, he unce ain y in sou ces concen a ions and measu emen e - o s. We p esen a me hodology o compu e mixing a ios and end-membe s composi ion, which consis s o (i) Iden i- fica ion o po en ial echa ge sou ces, (ii) Selec ion o ac- e s, (iii) Cha ac e iza ion o he hyd ochemical composi ion o po en ial echa ge sou ces and mixed wa e samples, and (i ) Compu a ion o mixing a ios and ee alua ion o end- membe s. The analysis pe o med in a da a se om samples o he Ba celona ci y aqui e s sugges s ha he main con ib- u o s o o al echa ge a e he wa e supply ne wo k losses (22%), he sewage ne wo k losses (30%), ain all, concen- a ed in he non-u banized a eas (17%), om uno infil a- ion (20%), and he Bes` os Ri e (11%). Rega ding species, halogens (chlo ide, fluo ide and b omide), sul a e, o al ni- ogen, and s able iso opes (18O, 2H, and 34S) beha ed qui e conse a i ely. Bo on, esidual alkalini y, EDTA and Zn did no . Ye , including hese species in he compu a ions did no a ec significan ly he p opo ion es ima ions. Co espondence o: E. V´ azquez-Su˜ n´ e (en ic. [email p o ec ed]) 1 In oduc ion G oundwa e is inc easingly ecognized as a ac o in u ban managemen (see e iews by: Le ne , 1996; Chil on e al., 1997; Eyles, 1997; Chil on, 1999; Ellis, 1999; Llamas and Cus odio, 2002; Howa d and Is afilo , 2002; AIH GE, 2004; V´ azquez-Su˜ ne e al., 2005a). On one hand, some ci ies s ill show o e exploi a ion wi h a se ies o unwan ed ou comes such as subsidence, o de e io a ion in wa e quali y due o a numbe o causes including seawa e pollu ion. On he o he hand, g oundwa e pumping has been abandoned in many ci ies because o pollu ion and/o changes in land uses (no- o iously eloca ion o indus ies). This has caused heads o ise, leading o flooding and po en ial s uc u al damage o unde g ound u ban s uc u es. Con olling heads ise may demand g oundwa e pumping in selec ed loca ions. Pump- ing may also be needed o add ess he inc easing demand o wa e in a sus ainable way. Amongs o he hings, his e- qui es unde s anding he p ocesses a ec ing wa e chemical and mic obiological quali y. A p ope assessmen o g ound- wa e quali y in ol es he quan ifica ion o o e all echa ge, as well as he quali y assessmen o he a ious sou ces in- ol ed. E alua ing echa ge in u ban en i onmen s is essen ially di e en om ha in na u al sys ems o wo easons: (1) echa ge sou ces a e adically di e en and (2) less expe- ience is a ailable. Su ace infil a ion is g ea ly educed be- cause a la ge po ion o su ace a ea is i ually impe ious. The e o e, mos ain all becomes uno , which is di e ed o a d ainage sys em. Howe e , educ ions in di ec infil- a ion can be coun e balanced by educ ions in e apo an- spi a ion. Mo eo e , new sou ces o echa ge may a ise in ci ies, such as losses in he sewage sys em (Wol e al., 2004; Blackwood e al., 2005) o in he wa e dis ibu ion sys- em (wo ld-wide losses exceed 50 L/pe son/day, bu epo ed Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union. 373 G e o l o g i c a A c a , V o l . 1 0 , N º 4 , D e c e m b e 2 0 1 2 , 3 7 3 - 3 9 3 D O I : 1 0 . 1 3 4 4 / 1 0 5 . 0 0 0 0 0 1 7 5 7 A a i l a b l e o n l i n e a w w w . g e o l o g i c a - a c a . c o m A sequence s a ig aphic based geological model o cons aining hyd ogeological modeling in he u banized a ea o he Qua e na y Besòs del a (NW Medi e anean coas , Spain) V. VELASCO P. CABELLO E. VÁZQUEZ-SUÑÉ M. LÓPEZ-BLANCO E. RAMOS I. TUBAU Geomodels Ins i u e. G oup o Geodynamics and Basin Analysis. Depa amen d’Es a ig a ia, Paleon ologia i Geosciències ma ines, Facul a de Geologia, Uni e si a de Ba celona (UB) c/ Ma í i F anquès s/n, 08028, Ba celona, Spain. Cabello E-mail: [email protected] López-Blanco E-mail: [email protected] Ramos E-mail: emilio. [email protected] GHS, Ins i u e o En i omen al Assessmen and Wa e Resea ch (IDAEA), CSIC Ba celona, Spain. E-mail: [email p o ec ed] G up de Hid ologia Sub e ània (GHS), Depa amen o Geo echnical Enginee ing and Geosciences, Uni e si a Poli ècnica de Ca alunya (UPC)-Ba celonaTech Jo di Gi ona 1-3 Building D-2, 08034 Ba celona, Spain. Velasco E-mail: iole a. [email protected] Tubau E-mail: isabel. [email protected] A B S T R A C T The Qua e na y Bes˜s del a is loca ed on he Medi e anean coas in NE Spain. The Bes˜s Del a Complex includes 3 aqui e s cons i u ed by 3 sandy and g a elly bodies, sepa a ed by lu i ic uni s. These aqui e s supply wa e o domes ic and indus ial use in his a ea. Managemen o g oundwa e has been p oblema ic in he Bes˜s del a since he 1960s, and con inues o pose majo p oblems o subsu ace enginee ing wo ks in his highly u banized egion. This s udy seeks o demons a e he ad an ages o de ailed geological cha ac e iza ion and modeling o designing and cons uc ing a hyd ogeological model. A ailable in o ma ion o he subsu ace was compiled, in eg a ed and homogenized in a geospa ial da abase. The in e p e a ion o hese da a enabled us o delimi geological uni s by means o a sequence s a ig aphic subdi ision. A h ee-dimensional acies bel -based model o he Bes˜s del a was buil on he basis o his geological cha ac e i- za ion. This model was used o cons ain he dis ibu ion o hyd aulic pa ame e s and hus o ob ain a consis en hyd ogeological model o he del a, which was calib a ed by da a o wa e managemen and p oduc ion o e he las hund ed yea s. The esul ing hyd ogeological model yielded new insigh s in o wa e on displacemen s in he aqui e du ing he ime-span conside ed, imp o ing p edic ions in an a emp o op imize aqui e managemen . 2 1 G oundwa e . Del a. Sequence s a ig aphy. Geological model. Hyd ogeological model. KEYWORDS 2 2, 3 1 3 2, 3 11 Applica ion o mul i-iso ope da a (O, D, C and S) o quan i y edox p ocesses in u ban g oundwa e Anna Ju ado a,b, ⇑ , En ic Vàzquez-Suñé a , Albe Sole c , Isabel Tubau b,d , Jesus Ca e a a , Es anislao Pujades a,b , Isabel Anson a,b a GHS, Ins i u e o En i onmen al Assessmen & Wa e Resea ch (IDAEA), CSIC, Jo di Gi ona 18-26, 08034 Ba celona, Spain b GHS, Dep . Geo echnical Enginee ing and Geosciences, Uni e si a Poli ecnica de Ca alunya, UPC-Ba celona Tech, Jo di Gi ona 1-3, 08034 Ba celona, Spain c G up de Mine alogia Aplicada i Medi Ambien , Dep . C is allog a ia, Mine alogia i Dipòsi s Mine als, Facul a de Geologia, Uni e si a de Ba celona, Ba celona, Spain d Geological Ins i u e o Ca alonia, Balmes 209-211, 08006 Ba celona, Spain a i c l e i n o A icle his o y: A ailable online 14 Ma ch 2013 a b s a c The di e en echa ge sou ces and hei mixing a ios we e in es iga ed in wa e samples om an u ban aqui e . Two dis inc zones om Ba celona ci y we e compa ed: (1) Poble Sec and (2) Besòs Ri e Del a, which a e loca ed in a densely popula ed a ea wi h a s ong indus ial impac . In his s udy 106 wa e samples we e collec ed om July 2007 o May 2010 o hyd ochemical and iso opic cha ac e isa ion. The applica ion o en i onmen al iso opes coupled wi h hyd ochemis y p o ided he necessa y in o ma- ion o iso opically quan i y g oundwa e echa ge sou ces and e alua e he occu ence o edox p o- cesses. In he Besòs Ri e Del a, a dec ease in dissol ed SO 2ÿ 4 concen a ion and an inc ease in d 34 S SO4 and d 18 O SO4 alues we e obse ed in g oundwa e samples, bo h indica ing SO 2ÿ 4 educ ion. Mo eo e , o he chemical indica o s suppo ed a educing en i onmen , such as low o null le els o dissol ed O 2 and NO ÿ 3 , he p esence o NH þ 4 and an inc ease in dissol ed Fe and As. The educing condi ions we e p ob- ably induced by he o ganic C dissol ed in wa e in il a ing om he Ri e Besòs. In Poble Sec, he ela- ionship be ween d 34 S SO4 and d 18 O SO4 showed a s ong in luence o sewage wa e in il a ion in o he aqui e . Howe e , he aqui e is oxic and he e is no in luence o SO 2ÿ 4 educ ion. Ó2013 Else ie L d. All igh s ese ed. 1. In oduc ion De e io a ion o g oundwa e quali y in u ban aqui e s has be- come a ma e o conce n. In such a eas, g oundwa e can be con- amina ed by a wide ange o pollu an s. Mo eo e , g oundwa e can be an al e na i e sou ce o wa e supply in ci ies wi h an a id o semia id clima e (Tubau e al., 2010). A p ope assessmen o g oundwa e quali y equi es he quan i ica ion o he o al e- cha ge and he composi ion o he a ious sou ces in ol ed. These quan i a i e assessmen s enable o iden i y he o igin and he a e o chemical compounds ( om pollu ion o a enua ion) and also e alua e managemen s a egies. Howe e , e alua ing he echa ge in u ban en i onmen s is essen ially di e en om u al a eas. This is mainly due o he la ge numbe o di e en sou ces, hei a iable composi ion, and he geochemical p ocesses unde gone by he mix u es. Thus, many adi ional mixing a io compu a ions equi e ha he concen a- ions o end-membe s o be di e en and accu a ely known. Cu en ly, he concen a ions o he di e en species in he end-membe s a e no usually known wi h ce ain y. They can be highly a iable in space and, especially, in ime. This could be he eason why s udies o iden i y and quan i y sou ces and p ocesses in ol ed in he wa e quali y o an u ban aqui e a e e y sca ce (see Vàzquez-Suñé e al., 2010 and e e ences he ein, o a e iew). Some common sou ces o g oundwa e in u ban a eas a e sewe leakage and in il a ion om was e wa e ea men plan s. These sou ces p o ide o ganic C o he wa e and p omo e biodeg ada- ion eac ions o a a ie y o pollu an s (NO ÿ 3 , SO 2ÿ 4 , As-bound o Fe, o ganic mic opollu an s). Biodeg ada ion p ocesses a e ela ed o he edox s a e o wa e . The e o e, iden i ying he edox e olu- ion along a low line is a key issue. Unlike physical p ocesses such as dispe sion, mixing and dilu- ion; biodeg ada ion implies a dec ease in he o al dissol ed mass o a solu e and may in ol e a a ia ion o i s iso opic p opo ions. This sugges s ha iso ope s udies can be help ul in iden i ying and quan i ying biodeg ada ion p ocesses. In ac , he applica ion o en i onmen al iso opes has been employed o cha ac e ise and as- sess di e en issues in u ban g oundwa e echa ge. Bu le and Ve hagen (1997) used wa e iso opes in he s udy o he ci y o P e o ia (Sou h A ica) o di e en ia e be ween he local wa e echa ge and wa e supply, which had a di e en iso opic compo- si ion. Choi e al. (2005) complemen ed a hyd ochemical s udy in 0883-2927/$ - see on ma e Ó2013 Else ie L d. All igh s ese ed. h p://dx.doi.o g/10.1016/j.apgeochem.2013.02.018 ⇑Co esponding au ho a : GHS, Ins i u e o En i onmen al Assessmen & Wa e Resea ch (IDAEA), CSIC, Jo di Gi ona 18-26, 08034 Ba celona, Spain. Tel.: +34 934006100. E-mail add ess: [email p o ec ed] (A. Ju ado). Applied Geochemis y 34 (2013) 114–125 Con en s lis s a ailable a SciVe se ScienceDi ec Applied Geochemis y jou nal homepage: www.else ie .com/loca e/apgeochem GIS-based hyd ogeochemical analysis ools (QUIMET) V. Velasco a,b, n , I. Tubau c , E. Vázquez-Suñè a , R. Gogu d , D. Gai ana u d , M. Alca az a,b , A. Se ano-Juan a,b , D. Fe nàndez-Ga cia b , T. Ga ido e , J. F aile e , X. Sanchez-Vila b a GHS, Ins i u e o En i onmen al Assessmen and Wa e Resea ch (IDAEA –CSIC), Jo di Gi ona 18-26, 08034 Ba celona, Spain b GHS, Depa men o Geo echnical Enginee ing and Geosciences, Uni e si a Poli ècnica de Ca alunya (UPC-Ba celonaTech), Jo di Gi ona 1-3, 08034 Ba celona, Spain c ICGC, Ins i u Ca og àfic i Geològic de Ca alunya, Balmes 209-201, Ba celona, Spain d Technical Uni e si y o Ci il Enginee ing, B-dul Lacul Tei, 124, 020296 Bucha es , Romania e ACA, Agència Ca alana de l'Aigua, P o ença 204, 08036 Ba celona, Spain a i c l e i n o A icle his o y: Recei ed 10 June 2013 Recei ed in e ised o m 16 Ap il 2014 Accep ed 17 Ap il 2014 A ailable online 8 May 2014 Keywo ds: GIS G oundwa e quali y Geospa ial da abase Hyd ogeochemical analysis a b s a c A so wa e pla o m (QUIMET) was de eloped o imp o e he so ing, analysis, calcula ions, isualiza- ions, and in e p e a ions o hyd ogeochemical da a in a GIS en i onmen . QUIMET is composed o a geospa ial da abase plus a se o ools specially designed o g aphical and s a is ical analysis o hyd ogeochemical da a. The geospa ial da abase has been designed o include o ganic and ino ganic chemical eco ds, as well as ele an physical pa ame e s ( empe a u e, Eh, elec ical conduc i i y). The ins umen s o analysis co e a wide ange o me hodologies o que ying, in e p e ing, and compa ing g oundwa e quali y da a. They include, among o he s, chemical ime-se ies analysis, ionic balance calcula ions, co ela ion o chemical pa ame e s, and calcula ion o a ious common hyd o- geochemical diag ams (Salini y, Schoelle –Be kalo , Pipe , and S i ). The GIS pla o m allows he gene a ion o maps o he spa ial dis ibu ion o pa ame e s and diag ams. Mo eo e , i allows pe o ming a comple e s a is ical analysis o he da a including desc ip i e s a is ic uni a ia e and bi a ia e analysis, he la e including gene a ion o co ela ion ma ices and g aphics. Finally, QUIMET o e s in e ope abili y wi h o he ex e nal pla o ms. The pla o m is illus a ed wi h a geochemical da a se om he ci y o Badalona, loca ed on he Medi e anean coas in NE Spain. &2014 Else ie L d. All igh s ese ed. 1. In oduc ion Wa e a ailabili y and accessibili y mus emphasize quali y aspec s, such as whe he wa e complies o gi en s anda ds equi ed o any pa icula use (Ba hel e al., 2008). In a gi en si e he quali y o g oundwa e may be ad e sely a ec ed by ac o s such as indus ia- liza ion, i iga ion p ac ices, o u baniza ion (Fos e , 2001; Vázquez- Suñè e al., 2005b; Scanlon e al., 2005; Ke a a e al., 2011). An accu a e assessmen o he nega i e impac s o hese di e en ac i i ies is pa amoun o he p o ec ion o wa e bodies and ecosys ems associa ed (Na a o-O ega e al., 2012). In o de o ensu e compliance wi h s anda d egula o y guidelines (e.g., he Eu opean Wa e F ame Di ec i e, WHO guidelines), con inuous moni o ing, e alua ion, and in e p e a ion o a la ge da abase in ol ing many po en ial geochem- ical pa ame e s is equi ed. This includes: (a) de e mining he o igin o g oundwa e and he p ocesses con olling i s chemical composi ion and he co esponding spa io- empo al dis ibu ion; (b) e alua ing he cu en g oundwa e quali y end wi h e e ence o soil use, as well as geological and hyd ogeological se up; and (c) es ablishing he egional backg ound composi ion o g oundwa e (in Mendizabal and S uy zand, 2009). Accomplishing hese asks is no s aigh o wa d. The majo di ficul ies o be aced a ise om: (i) he need o manipula e la ge da a se s collec ed o e many yea s; (ii) he in eg a ion o da a s emming om di e se sou ces and ga he ed wi h di e en da a access echniques and o ma s; (iii) he managemen o da a wi h a ying quali y s anda ds (and hus di e en associa ed e o s) and empo al and spa ial ex en , wi h po en ial big gaps o in o ma ion a he spa ial o empo al scales in some a eas; (i ) da a-handling de i ed om bo h field da a, analysis o wa e samples a he labo a o y and g oundwa e models (also ep e- sen a i e scales o da a a e qui e di e en ), and ( ) he in eg a ion o g oundwa e quali y in o ma ion wi h o he ele an in o ma- ion such as geological da a. Op imal da a managemen o his as amoun o spa io- empo al in o ma ion canno be eadily handled wi hou he combina ion o a comp ehensi e geospa ial da abase and a numbe o e ficien ech- nologies and me hodologies capable o classi ying, compa ing, sum- ma izing and in e p e ing huge da a se s. Classic g oundwa e analysis Con en s lis s a ailable a ScienceDi ec jou nal homepage: www.else ie .com/loca e/cageo Compu e s & Geosciences h p://dx.doi.o g/10.1016/j.cageo.2014.04.013 0098-3004/&2014 Else ie L d. All igh s ese ed. n Co esponding au ho . E-mail add esses: [email p o ec ed],[email p o ec ed] (V. Velasco). URL: h p://www.h2ogeo.upc.es/English/so wa e/QUIMET_ACADEMIC/index. h m (V. Velasco). Compu e s & Geosciences 70 (2014) 164–180 Quan i ying chemical eac ions by using mixing analysis Anna Ju ado a,b, ⁎, En ic Vázquez-Suñé a , Jesús Ca e a a , Isabel Tubau b , Es anislao Pujades a,b a GHS, Ins i u e o En i onmen al Assessmen & Wa e Resea ch (IDAEA), CSIC, Jo di Gi ona 18-26, 08034 Ba celona, Spain b GHS, Dep . Geo echnical Enginee ing and Geosciences, Technical Uni e si y o Ca alonia, UPC-Ba celona Tech, Jo di Gi ona 1-3, 08034 Ba celona, Spain H I G H L I G H T S •U ban aqui e is infil a ed by a pollu ed i e ha is he main echa ge sou ce. •Th ee end-membe s a e necessa y o explain he empo al a iabili y o i e wa e . •Depa u es om ideal mixing line a e a ibu ed o chemical p ocesses. •Mixing calcula ions a e used o quan i y chemical p ocesses in u ban g oundwa e . •The p ocesses quan ified a e edox p ocesses and ca bona e dissolu ion. a b s a c a i c l e i n o A icle his o y: Recei ed 21 July 2014 Recei ed in e ised o m 11 Sep embe 2014 Accep ed 11 Sep embe 2014 A ailable online xxxx Edi o : D. Ba celo Keywo ds: U ban aqui e Mixing analysis Non-conse a i e species Chemical eac ion quan ifica ion This wo k is mo i a ed by a sound unde s anding o he chemical p ocesses ha a ec he o ganic pollu an s in an u ban aqui e . We p opose an app oach o quan i y such p ocesses using mixing calcula ions. The me hodol- ogy consis s o he ollowing s eps: (1) iden ifica ion o he echa ge sou ces (end-membe s) and selec ion o he species (conse a i e and non-conse a i e) o be used, (2) iden ifica ion o he chemical p ocesses and (3) e al- ua ion o mixing a ios including he chemical p ocesses. This me hodology has been applied in he Besòs Ri e Del a (NE Ba celona, Spain), whe e he Ri e Besòs is he main aqui e echa ge sou ce. A o al numbe o 51 g oundwa e samples we e collec ed om July 2007 o May 2010 du ing ou field campaigns. Th ee i e end-membe s we e necessa y o explain he empo al a iabili y o he Ri e Besòs: one i e end-membe is om he we pe iods (W1) and wo a e om d y pe iods (D1 and D2). This me hodology has p o ed o be use ul no only o compu e he mixing a ios bu also o quan i y p ocesses such as calci e and magnesi e dissolu ion, ae obic espi a ion and deni ifica ion unde gone a each obse a ion poin . © 2014 Else ie B.V. All igh s ese ed. 1. In oduc ion Ensu ing good wa e quali y is a majo challenge in u ban a eas. U ban aqui e s may su e pollu ion om echa ge sou ces such as wa e leakage om sewe and sep ic sys ems, seepage om i e s, sea- wa e in usion, and losses om wa e supply ne wo k e c. As a esul , a wide ange o o ganic pollu an s a e ound in u ban aqui e s. Since hese pollu an s each g oundwa e en i onmen , hei occu ence depends on simul aneous anspo and biogeochemical p ocesses. Fo a quan i- a i e assessmen o g oundwa e quali y a sound unde s anding o all hese p ocesses a ec ing hese pollu an s is essen ial. Howe e , he quan ifica ion o hese p ocesses is no an easy ask. In o de o quan i y such p ocesses, modelling ools ha p o ide ex ensi e biogeochemical capabili ies a e necessa y (P omme e al., 2000). Reac i e anspo models quan i y he spa ial and empo al e o- lu ion o a numbe o chemical species subjec ed o anspo phenom- ena and chemical eac ions. Nume ous codes ha e been de eloped in ecen yea s. They ha e been applied o p oblems such as leacha e a enua ion (Islam e al., 2001; Van B eukelen e al., 2004), pollu ion plumes (B un e al., 2002), mic obially media ed eac ions (Schä e e al., 1998a,b; Hun e e al., 1998; Tebes-S e ens e al., 1999; Wang and Papengu h, 2001), ca bon cycle modelling cases (G eskowiak e al., 2005) and he quan ifica ion o edox p ocesses in aqui e s (Kea ing and Bah , 1998; McGui e e al., 2002; Massmann e al., 2004; Mio linski, 2008) and in he hypo heic zone (Mille e al., 2006; Za ne ske e al., 2011). Howe e , simple app oaches a e equi ed since hese echniques a e edious and ime consuming because o he need o a eliable flow and conse a i e anspo models. Mixing calcula ions cons i u e an al e na i e o eac i e anspo models. Mixing calcula ions ypically in ol ed: (1) End-Membe Mixing Analysis (EMMA) (Hoope e al., 1990; Hoope , 2003) o find he minimum numbe o end-membe s equi ed o explain a se o chemical analyses, and (2) mixing a ios defined as he p opo ion o Science o he To al En i onmen 502 (2015) 448–456 ⁎Co esponding au ho a : Jo di Gi ona 18, 08034 Ba celona, Spain. Tel.: +34 934006100; ax: +34 932045904. E-mail add ess: annaju [email protected] (A. Ju ado). h p://dx.doi.o g/10.1016/j.sci o en .2014.09.036 0048-9697/© 2014 Else ie B.V. All igh s ese ed. Con en s lis s a ailable a ScienceDi ec Science o he To al En i onmen jou nal h o me page : www. e lse i e .c om /loc a e/ s ci o en