Water Pollution XIV

Water Pollution XIV

WIT Press publishes leading books in Science and Technology. Visit our website for the current list of titles. www.witpress.com

WITeLibrary Home of the Transactions of the Wessex Institute. Papers contained in this volume are archived in the WIT eLibrary in volume 228 of WIT Transactions on Ecology and the Environment (ISSN 1743-3541). The WIT eLibrary provides the international scientific community with immediate and permanent access to individual papers presented at WIT conferences. Visit the WIT eLibrary at https://www.witpress.com/elibrary

FOURTEENTH INTERNATIONAL CONFERENCE ON MONITORING, MODELLING AND MANAGEMENT OF WATER POLLUTION

Water Pollution 2018 Conference Chairmen S. Hernández University of A Coruña, Spain Member of WIT Board of Directors S. Mambretti Polytechnic of Milan, Italy Member of WIT Board of Directors

International Scientific Advisory Committee J. Anta Alvarez D. Bonotto Z. Boukalova E. Christoffels J. Gumbo T. Ishikawa R. Mahler A. Pop I. Rukavishnikova A. Slobodov I. Strelnikova G. Zappala

Organised by Wessex Institute, UK University of A Coruña, Spain

Sponsored by WIT Transactions on Ecology and the Environment International Journal of Environmental Impacts

WIT Transactions Wessex Institute Ashurst Lodge, Ashurst Southampton SO40 7AA, UK

Senior Editors H. Al-Kayiem

Universiti Teknologi PETRONAS, Malaysia

G. M. Carlomagno

University of Naples Federico II, Italy

A. H-D. Cheng

University of Mississippi, USA

J. J. Connor

Massachusetts Institute of Technology, USA

J. Th M. De Hosson

University of Groningen, Netherlands

P. De Wilde

Vrije Universiteit Brussel, Belgium

N. A. Dumont

PUC-Rio, Brazil

A. Galiano-Garrigos

University of Alicante, Spain

F. Garzia

University of Rome “La Sapienza”, Italy

M. Hadfield

University of Bournemouth, UK

S. Hernández

University of A Coruña, Spain

J. T. Katsikadelis

National Technical University of Athens, Greece

J. W. S. Longhurst

University of the West of England, UK

E. Magaril

Ural Federal University, Russia

S. Mambretti

Politecnico di Milano, Italy

W. J. Mansur

Federal University of Rio de Janeiro, Brazil

J. L. Miralles i Garcia

Universitat Politècnica de València, Spain

G. Passerini

Università Politecnica delle Marche, Italy

F. D. Pineda

Complutense University, Spain

D. Poljak

University of Split, Croatia

F. Polonara

Università Politecnia delle Marche, Italy

D. Proverbs

Birmingham City University, UK

T. Rang

Tallinn Technical University, Estonia

G. Rzevski

The Open University, UK

P. Skerget

University of Maribor, Slovenia

B. Sundén

Lund University, Sweden

Y. Villacampa Esteve

Universidad de Alicante, Spain

P. Vorobieff

University of New Mexico, USA

S. S. Zubir

Universiti Teknologi Mara, Malaysia

Editorial Board B. Abersek University of Maribor, Slovenia

F. Butera Politecnico di Milano, Italy

Y. N. Abousleiman University of Oklahoma, USA

W. Cantwell Liverpool University, UK

G. Alfaro Degan Università Roma Tre, Italy

C. Capilla Universidad Politecnica de Valencia, Spain

K. S. Al Jabri Sultan Qaboos University, Oman D. Almorza Gomar University of Cadiz, Spain J. A. C. Ambrosio IDMEC, Portugal A. M. Amer Cairo University, Egypt S. A. Anagnostopoulos University of Patras, Greece E. Angelino A.R.P.A. Lombardia, Italy H. Antes Technische Universitat Braunschweig, Germany M. A. Atherton South Bank University, UK A. G. Atkins University of Reading, UK D. Aubry Ecole Centrale de Paris, France H. Azegami Toyohashi University of Technology, Japan J. M. Baldasano Universitat Politecnica de Catalunya, Spain J. Barnes University of the West of England, UK J. G. Bartzis Institute of Nuclear Technology, Greece S. Basbas Aristotle University of Thessaloniki, Greece A. Bejan Duke University, USA M. P. Bekakos Democritus University of Thrace, Greece G. Belingardi Politecnico di Torino, Italy R. Belmans Katholieke Universiteit Leuven, Belgium D. E. Beskos University of Patras, Greece

D. J. Cartwright Bucknell University, USA P. G. Carydis National Technical University of Athens, Greece J. J. Casares Long Universidad de Santiago de Compostela, Spain A. Chakrabarti Indian Institute of Science, India F. Chejne National University, Colombia J-T. Chen National Taiwan Ocean University, Taiwan J. Chilton University of Lincoln, UK C-L. Chiu University of Pittsburgh, USA H. Choi Kangnung National University, Korea A. Cieslak Technical University of Lodz, Poland C. Clark Wessex Institute, UK S. Clement Transport System Centre, Australia M. C. Constantinou State University of New York at Buffalo, USA M. da C Cunha University of Coimbra, Portugal W. Czyczula Krakow University of Technology, Poland L. D’Acierno Federico II University of Naples, Italy M. Davis Temple University, USA A. B. de Almeida Instituto Superior Tecnico, Portugal L. De Biase University of Milan, Italy

S. K. Bhattacharyya Indian Institute of Technology, India

R. de Borst Delft University of Technology, Netherlands

H. Bjornlund University of South Australia, Australia

G. De Mey University of Ghent, Belgium A. De Naeyer Universiteit Ghent, Belgium

E. Blums Latvian Academy of Sciences, Latvia

N. De Temmerman Vrijie Universiteit Brussel, Belgium

J. Boarder Cartref Consulting Systems, UK B. Bobee Institut National de la Recherche Scientifique, Canada H. Boileau ESIGEC, France M. Bonnet Ecole Polytechnique, France C. A. Borrego University of Aveiro, Portugal A. R. Bretones University of Granada, Spain F-G. Buchholz Universitat Gesanthochschule Paderborn, Germany

D. De Wrachien State University of Milan, Italy L. Debnath University of Texas-Pan American, USA G. Degrande Katholieke Universiteit Leuven, Belgium S. del Giudice University of Udine, Italy M. Domaszewski Universite de Technologie de Belfort-Montbeliard, France

K. Dorow Pacific Northwest National Laboratory, USA

O. T. Gudmestad University of Stavanger, Norway

W. Dover University College London, UK

R. C. Gupta National University of Singapore, Singapore

C. Dowlen South Bank University, UK

J. M. Hale University of Newcastle, UK

J. P. du Plessis University of Stellenbosch, South Africa

K. Hameyer Katholieke Universiteit Leuven, Belgium

R. Duffell University of Hertfordshire, UK

C. Hanke Danish Technical University, Denmark

A. Ebel University of Cologne, Germany

Y. Hayashi Nagoya University, Japan

V. Echarri University of Alicante, Spain

L. Haydock Newage International Limited, UK

K. M. Elawadly Alexandria University, Egypt

A. H. Hendrickx Free University of Brussels, Belgium

D. Elms University of Canterbury, New Zealand M. E. M El-Sayed Kettering University, USA D. M. Elsom Oxford Brookes University, UK F. Erdogan Lehigh University, USA J. W. Everett Rowan University, USA M. Faghri University of Rhode Island, USA R. A. Falconer Cardiff University, UK

C. Herman John Hopkins University, USA I. Hideaki Nagoya University, Japan W. F. Huebner Southwest Research Institute, USA M. Y. Hussaini Florida State University, USA W. Hutchinson Edith Cowan University, Australia T. H. Hyde University of Nottingham, UK

M. N. Fardis University of Patras, Greece

M. Iguchi Science University of Tokyo, Japan

A. Fayvisovich Admiral Ushakov Maritime State University, Russia

L. Int Panis VITO Expertisecentrum IMS, Belgium N. Ishikawa National Defence Academy, Japan

H. J. S. Fernando Arizona State University, USA

H. Itoh University of Nagoya, Japan

W. F. Florez-Escobar Universidad Pontifica Bolivariana, South America

W. Jager Technical University of Dresden, Germany

E. M. M. Fonseca Instituto Politécnico do Porto, Instituto Superior de Engenharia do Porto, Portugal

Y. Jaluria Rutgers University, USA

D. M. Fraser University of Cape Town, South Africa G. Gambolati Universita di Padova, Italy C. J. Gantes National Technical University of Athens, Greece L. Gaul Universitat Stuttgart, Germany N. Georgantzis Universitat Jaume I, Spain L. M. C. Godinho University of Coimbra, Portugal F. Gomez Universidad Politecnica de Valencia, Spain A. Gonzales Aviles University of Alicante, Spain D. Goulias University of Maryland, USA K. G. Goulias Pennsylvania State University, USA W. E. Grant Texas A & M University, USA S. Grilli University of Rhode Island, USA R. H. J. Grimshaw Loughborough University, UK D. Gross Technische Hochschule Darmstadt, Germany R. Grundmann Technische Universitat Dresden, Germany

D. R. H. Jones University of Cambridge, UK N. Jones University of Liverpool, UK D. Kaliampakos National Technical University of Athens, Greece D. L. Karabalis University of Patras, Greece A. Karageorghis University of Cyprus T. Katayama Doshisha University, Japan K. L. Katsifarakis Aristotle University of Thessaloniki, Greece E. Kausel Massachusetts Institute of Technology, USA H. Kawashima The University of Tokyo, Japan B. A. Kazimee Washington State University, USA F. Khoshnaw Koya University, Iraq S. Kim University of Wisconsin-Madison, USA D. Kirkland Nicholas Grimshaw & Partners Ltd, UK E. Kita Nagoya University, Japan A. S. Kobayashi University of Washington, USA D. Koga Saga University, Japan S. Kotake University of Tokyo, Japan

A. N. Kounadis National Technical University of Athens, Greece

C. A. Mitchell University of Sydney, Australia

W. B. Kratzig Ruhr Universitat Bochum, Germany

A. Miyamoto Yamaguchi University, Japan

T. Krauthammer Penn State University, USA R. Laing Robert Gordon University, UK M. Langseth Norwegian University of Science and Technology, Norway B. S. Larsen Technical University of Denmark, Denmark F. Lattarulo Politecnico di Bari, Italy A. Lebedev Moscow State University, Russia D. Lesnic University of Leeds, UK D. Lewis Mississippi State University, USA K-C. Lin University of New Brunswick, Canada A. A. Liolios Democritus University of Thrace, Greece D. Lippiello Università degli Studi Roma Tre, Italy

K. Miura Kajima Corporation, Japan T. Miyoshi Kobe University, Japan G. Molinari University of Genoa, Italy F. Mondragon Antioquin University, Colombia T. B. Moodie University of Alberta, Canada D. B. Murray Trinity College Dublin, Ireland M. B. Neace Mercer University, USA D. Necsulescu University of Ottawa, Canada B. Ning Beijing Jiatong University, China S-I. Nishida Saga University, Japan H. Nisitani Kyushu Sangyo University, Japan B. Notaros University of Massachusetts, USA P. O’Donoghue University College Dublin, Ireland R. O. O’Neill Oak Ridge National Laboratory, USA

S. Lomov Katholieke Universiteit Leuven, Belgium

M. Ohkusu Kyushu University, Japan

J. E. Luco University of California at San Diego, USA

G. Oliveto Universitá di Catania, Italy R. Olsen Camp Dresser & McKee Inc., USA

L. Lundqvist Division of Transport and Location Analysis, Sweden

E. Oñate Universitat Politecnica de Catalunya, Spain

T. Lyons Murdoch University, Australia

K. Onishi Ibaraki University, Japan

L. Mahdjoubi University of the West of England, UK

P. H. Oosthuizen Queens University, Canada

Y-W. Mai University of Sydney, Australia

O. Ozcevik Istanbul Technical University, Turkey

M. Majowiecki University of Bologna, Italy

E. Outa Waseda University, Japan

G. Manara University of Pisa, Italy

A. S. Papageorgiou Rensselaer Polytechnic Institute, USA

B. N. Mandal Indian Statistical Institute, India

J. Park Seoul National University, Korea

Ü. Mander University of Tartu, Estonia

F. Patania Universitá di Catania, Italy

H. A. Mang Technische Universitat Wien, Austria

B. C. Patten University of Georgia, USA

G. D. Manolis Aristotle University of Thessaloniki, Greece

G. Pelosi University of Florence, Italy

N. Marchettini University of Siena, Italy J. D. M. Marsh Griffith University, Australia J. F. Martin-Duque Universidad Complutense, Spain T. Matsui Nagoya University, Japan G. Mattrisch DaimlerChrysler AG, Germany F. M. Mazzolani University of Naples “Federico II”, Italy

G. G. Penelis Aristotle University of Thessaloniki, Greece W. Perrie Bedford Institute of Oceanography, Canada M. F. Platzer Naval Postgraduate School, USA D. Prandle Proudman Oceanographic Laboratory, UK R. Pulselli University of Siena, Italy

K. McManis University of New Orleans, USA

I. S. Putra Institute of Technology Bandung, Indonesia

A. C. Mendes Universidade de Beira Interior, Portugal

Y. A. Pykh Russian Academy of Sciences, Russia A. Rabasa University Miguel Hernandez, Spain

J. Mera Polytechnic University of Madrid, Spain

F. Rachidi EMC Group, Switzerland

J. Mikielewicz Polish Academy of Sciences, Poland

K. R. Rajagopal Texas A & M University, USA

R. A. W. Mines University of Liverpool, UK

J. Ravnik University of Maribor, Slovenia

A. M. Reinhorn State University of New York at Buffalo, USA

G. C. Sih Lehigh University, USA

G. Reniers Universiteit Antwerpen, Belgium

A. C. Singhal Arizona State University, USA

A. D. Rey McGill University, Canada D. N. Riahi University of Illinois at UrbanaChampaign, USA B. Ribas Spanish National Centre for Environmental Health, Spain K. Richter Graz University of Technology, Austria S. Rinaldi Politecnico di Milano, Italy F. Robuste Universitat Politecnica de Catalunya, Spain A. C. Rodrigues Universidade Nova de Lisboa, Portugal G. R. Rodríguez Universidad de Las Palmas de Gran Canaria, Spain C. W. Roeder University of Washington, USA J. M. Roesset Texas A & M University, USA W. Roetzel Universitaet der Bundeswehr Hamburg, Germany V. Roje University of Split, Croatia R. Rosset Laboratoire d’Aerologie, France J. L. Rubio Centro de Investigaciones sobre Desertificacion, Spain T. J. Rudolphi Iowa State University, USA S. Russenchuck Magnet Group, Switzerland H. Ryssel Fraunhofer Institut Integrierte Schaltungen, Germany S. G. Saad American University in Cairo, Egypt M. Saiidi University of Nevada-Reno, USA R. San Jose Technical University of Madrid, Spain F. J. Sanchez-Sesma Instituto Mexicano del Petroleo, Mexico B. Sarler Nova Gorica Polytechnic, Slovenia S. A. Savidis Technische Universitat Berlin, Germany A. Savini Universita de Pavia, Italy G. Schleyer University of Liverpool, UK R. Schmidt RWTH Aachen, Germany B. Scholtes Universitaet of Kassel, Germany A. P. S. Selvadurai McGill University, Canada J. J. Sendra University of Seville, Spain S. M. Şener Istanbul Technical University, Turkey J. J. Sharp Memorial University of Newfoundland, Canada Q. Shen Massachusetts Institute of Technology, USA

L. C. Simoes University of Coimbra, Portugal J. Sladek Slovak Academy of Sciences, Slovakia V Sladek Slovak Academy of Sciences, Slovakia A. C. M. Sousa University of New Brunswick, Canada H. Sozer Illinois Institute of Technology, USA P. D. Spanos Rice University, USA T. Speck Albert-Ludwigs-Universitaet Freiburg, Germany C. C. Spyrakos National Technical University of Athens, Greece G. E. Swaters University of Alberta, Canada S. Syngellakis Wessex Institute, UK J. Szmyd University of Mining and Metallurgy, Poland H. Takemiya Okayama University, Japan I. Takewaki Kyoto University, Japan C-L. Tan Carleton University, Canada E. Taniguchi Kyoto University, Japan S. Tanimura Aichi University of Technology, Japan J. L. Tassoulas University of Texas at Austin, USA M. A. P. Taylor University of South Australia, Australia A. Terranova Politecnico di Milano, Italy T. Tirabassi National Research Council, Italy S. Tkachenko Otto-von-Guericke-University, Germany N. Tomii Chiba Institute of Technology, Japan T. Tran-Cong University of Southern Queensland, Australia R. Tremblay Ecole Polytechnique, Canada I. Tsukrov University of New Hampshire, USA R. Turra CINECA Interuniversity Computing Centre, Italy S. G. Tushinski Moscow State University, Russia R. van der Heijden Radboud University, Netherlands R. van Duin Delft University of Technology, Netherlands P. Vas University of Aberdeen, UK R. Verhoeven Ghent University, Belgium A. Viguri Universitat Jaume I, Spain S. P. Walker Imperial College, UK G. Walters University of Exeter, UK B. Weiss University of Vienna, Austria

T. W. Wu University of Kentucky, USA S. Yanniotis Agricultural University of Athens, Greece A. Yeh University of Hong Kong, China B. W. Yeigh University of Washington, USA

M. Zador Technical University of Budapest, Hungary R. Zainal Abidin Infrastructure University Kuala Lumpur, Malaysia K. Zakrzewski Politechnika Lodzka, Poland

K. Yoshizato Hiroshima University, Japan

M. Zamir University of Western Ontario, Canada

T. X. Yu Hong Kong University of Science & Technology, Hong Kong

R. Zarnic University of Ljubljana, Slovenia

G. Zappalà National Research Council, Italy

Water Pollution XIV

Editors S. Hernández University of A Coruña, Spain Member of WIT Board of Directors S. Mambretti Polytechnic of Milan, Italy Member of WIT Board of Directors

Editors: S. Hernández University of A Coruña, Spain Member of WIT Board of Directors S. Mambretti Polytechnic of Milan, Italy Member of WIT Board of Directors

Published by WIT Press Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK Tel: 44 (0) 238 029 3223; Fax: 44 (0) 238 029 2853 E-Mail: [email protected] http://www.witpress.com For USA, Canada and Mexico Computational Mechanics International Inc 25 Bridge Street, Billerica, MA 01821, USA Tel: 978 667 5841; Fax: 978 667 7582 E-Mail: [email protected] http://www.witpress.com British Library Cataloguing-in-Publication Data

A Catalogue record for this book is available from the British Library

ISBN: 978-1-78466-261-5 eISBN: 978-1-78466-262-2 ISSN: 1746-448X (print) ISSN: 1743-3541 (on-line) The texts of the papers in this volume were set individually by the authors or under their supervision. Only minor corrections to the text may have been carried out by the publisher. No responsibility is assumed by the Publisher, the Editors and Authors for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. The Publisher does not necessarily endorse the ideas held, or views expressed by the Editors or Authors of the material contained in its publications. © WIT Press 2018 Printed in Great Britain by Lightning Source, UK. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the Publisher.

Preface

This volume contains the papers presented at the 14th International Conference on Monitoring, Modelling and Management of Water Pollution, which took place at A Coruña, Spain, organised by the Wessex Institute. The availability of unlimited water resources cannot any longer be taken for granted as the needs of a growing world population, demanding better standards of living, continues to increase. Prominent among those problems is water quality that, due to the increase of pollutant loads discharged into natural water bodies, requires better tools for assessment and the formation of a framework for regulation and control. Contamination of water resources comes from very different sources, including industrial, agricultural and residential users. This diversity of usage results in the need to understand better the complex physio-chemical process involved. Moreover, environmental problems are essentially interdisciplinary. Engineers and scientists working in this field must be familiar with a wide range of issues including the physical processes of mixing and dilution, chemical and biological processes, mathematical modelling, data acquisition and measurement, to name but a few. Furthermore, water quality can have dramatic effects on human health, not only due to heavy metals and other well-known agents, but also a wide range of emerging chemical and pharmaceutical products whose effects are poorly understood. In view of the scarcity of available data, it is important that experiences are shared on an international basis. Thus, a continuous exchange of information between scientists from different countries is essential. The papers in this book make a significant contribution to the solution of some of these issues. These papers, like others presented at Wessex Institute conferences, are referenced by CrossRef and appear regularly in suitable reviews, publications and databases, including referencing and abstracting services. They are also archived online in the WIT eLibrary (http://www.witpress.com/ elibrary) where they are permanently available in Open Access format to the international scientific community. The Editors would like to thank the authors for their contributions, as well as the member of the International Scientific Advisory Community of the Conference for their invaluable help in reviewing the papers. The Editors A Coruña, 2018

This page intentionally left blank

Contents SEDUNIT Project: study of the accumulation, erosion and sediment transport of cohesive solids in combined sewer systems Jose Anta, Joaquín Suárez, Alfredo Jácome, Manuel Regueiro-Picallo, Jerónimo Puertas, Juan Naves & Montserrat Recarey ......................................................... 1 Laboratory experiment on generation of anaerobic gas and scum from organic sludge in urban rivers Shin Miura, Tadaharu Ishikawa & Tetsuo Hotta .................................................................. 9 Impact of industrial and municipal waste-load on Skinnerspruit in Gauteng Province, South Africa David O. Omole, Badejo A. Adekunle, Julius M. Ndambuki, Adebanji S. Ogbiye, Olumuyiwa O. Onakunle & Praisegod C. Emenike ............................ 21 Real-time monitoring of water quality using smart sensors: feedback of large Pilot Lab tests Amani Abdallah, Isam Shahrour & Marwan Sadek ............................................................ 29 Groundwater quality and its distribution in Siloam Village, Limpopo Province, South Africa Rachel Makungo & John O. Odiyo ...................................................................................... 35 Column leaching heavy metal from tailings following simulated climate change in the Arctic area of Norway Shuai Fu & Jinmei Lu .......................................................................................................... 45 Phytoremediation of wastewater with Thalia geniculata in constructed wetlands: basic pollutants distribution Gaspar López-Ocaña, Raúl Germán Bautista-Margulis, Sergio Ramos-Herrera, Carlos Alberto Torres-Balcazar, Rocío López-Vidal & Liliana Pampillón-González ............................................................. 53 Glyphosate in runoff from urban, mixed-use and agricultural watersheds in Hawaii, USA Steven R. Spengler, Marvin D. Heskett & Samuel C. Spengler ........................................... 65 Path modelling analysis of pollution sources and environmental consequences in river basins António Fernandes, Ana Ferreira, Luís Sanches Fernandes, Rui Cortes & Fernando Pacheco ........................................................................................ 79

Re-evaluating hydrochemical data from aquifers occurring in the Rio Claro city region, São Paulo State, Brazil Raquel Curtolo Quirino & Daniel Marcus Bonotto ............................................................ 89 Arsenic adsorption into the fixed bed column from drinking groundwater Vasile Minzatu, Adina Negrea, Corneliu Mircea Davidescu, Corina Seiman Duda, Mihaela Ciopec, Narcis Duţeanu, Petru Negrea, Daniel Duda Seiman & Bogdan Ioan Pascu .............................................. 101 A new adsorbent for arsenic removal from water Mihaela Ciopec, Iosif Hulka, Narcis Duţeanu, Adina Negrea, Oana Grad, Petru Negrea, Vasile Minzatu & Cristina Ardean ............................................................. 111 Assessment of silver metal released into wastewater after using a silver deodorant Lebogang L. M. Modika, Lufuno Matsheketsheke & Jabulani Ray Gumbo ...................... 121 An experiment on simultaneous operation of nitrification and denitrification of municipal landfill leachate in a single reaction tank Kohji Michioku, Kenji Tanaka, Hiroya Tanaka, Kosuke Inoue, Tamihiro Nakamichi, Masahiro Yagi & Nariaki Wada ..................................................... 131 Effect of internal recirculation on reactor models in wastewater treatment Tamas Karches .................................................................................................................. 145 Monitoring micropollutants in surface and subsurface runoff in the Swist river basin, Germany Ekkehard Christoffels ........................................................................................................ 155 Water quality checks on River Atuwara, South-West Nigeria David O. Omole, Adebanji S. Ogbiye, Ezechiel O. Longe, Ife K. Adewumi, Olugbenga O. Elemile & Theophilus I. Tenebe ................................................................. 165 Water supply, sanitation and health risk in a tropical sub-Saharan region Samuel A. Ogbiye, Oladotun A. Coker & Daniel I. Diwa.................................................. 175 Quantifying the spatio-temporal variability of water quality in an urbanizing perennial Mediterranean river: the case of the Beirut River Sania El-Nakib, Ibrahim Alameddine, Majdi Abou Najm & May Massoud ...................... 187 Dredging works monitoring in the port of Civitavecchia, Rome, Italy: sedimentological and geochemical investigations Daniele Piazzolla, Sergio Scanu, Simone Bonamano, Francesco Paladini de Mendoza, Riccardo Martellucci & Giuseppe Zappala .............................................................................................................. 199 Improvement of the methodology for assessing domestic wastewater treatment quality using benchmarking tools Irina Rukavishnikova, Andrey Kiselev, Maria Berezyuk & Iulia Ashirova ....................... 209

Let’s get our priorities straight Glenn Browning................................................................................................................. 221 Cyberinfrastructure supporting watershed health monitoring and management Tony B. Szwilski, Jack Smith, Justin Chapman & Mark Lewis .......................................... 245 Salinity modelling and management of the lower lakes of the Murray–Darling basin, Australia Jianli Liu, Muttucumaru Sivakumar, Shuqing Yang & Brian G. Jones ............................. 257 Water temperature monitoring in Eastern Canada: a case study for network optimization André St-Hilaire, Claudine Boyer, Normand Bergeron & Anik Daigle ............................ 269 Testing water pollution based on wireless sensor networks and stochastic approximation method: the case of Flint, Michigan, USA Nahed A. Alnahash & Mohamed A. Zohdy ........................................................................ 277 Urban acupuncture: producing water in informal settlements Camilo Cerro ..................................................................................................................... 285 Methodology development for designing flood channels as recreational waterways Cagri Cimen & Anil Olgac ................................................................................................ 295 The importance of considering pollution along the coast from heavily modified water bodies under the Water Framework Directive Regina Temino-Boes, Inmaculada Romero, Remedios Martinez-Guijarro & Maria Pachés..................................................................................................................... 307 Physico-chemical and surface characterisation of a renewable low-cost biosorbent for the uptake of heavy metal ions from aqueous solution John O. Odiyo & Joshua N. Edokpayi ............................................................................... 317 Author index .................................................................................................................... 329


Water Pollution XIV

1

SEDUNIT PROJECT: STUDY OF THE ACCUMULATION, EROSION AND SEDIMENT TRANSPORT OF COHESIVE SOLIDS IN COMBINED SEWER SYSTEMS JOSE ANTA, JOAQUÍN SUÁREZ, ALFREDO JÁCOME, MANUEL REGUEIRO-PICALLO, JERÓNIMO PUERTAS, JUAN NAVES & MONTSERRAT RECAREY Universidade da Coruña, Water and Environmental Engineering Group (GEAMA), A Coruña, Spain

ABSTRACT The SEDUNIT Project is a Spanish National Research focused on the study of the sediment transport in combined sewer systems. The main objective of this study is to advance in the understanding of the sediment accumulation and erosion processes in combined sewer systems and the optimization of the operation and maintenance practices in sewer networks. The research is divided into three tasks corresponding to different scales in order to approach the objectives of the Project: laboratory tests, field work and numerical modelling. A flume test facility is utilized for the laboratory scale. In this facility, a series of experiments are performed in order to study the accumulation and erosion processes. For the second task, a real combined sewer is being monitored in order to validate laboratory results. The sediment properties and the volume of the accumulated solids in the sewer are the main analysed parameters. The last task consists in a review of the existing sediment transport models and their applicability to assess how parameters such as the cohesion of solids affect the resuspension processes and how they should be included in the existing models. Keywords: combined sewer, flume test, sediment transport, sewer dynamics, sewer processes, sewer sediments, urban drainage, wastewater pipeline.

1 INTRODUCTION This Project deals with new sustainable management practices for urban drainage systems. The actual policies must integrate social and economic sustainability criteria and must be more resilient against climate change. For that, strategies based on Low Impact Development (LID) or Water Sensitivity Urban Design (WSUD) should be considered [1]. The purpose of this Project is to understand the in-sewer processes, particularly the accumulation and erosion of bed deposits. In combined sewer systems, the deposits of sediments reduce the hydraulic capacity of the conduits (increasing flooding and pollution spill risks), increase odour and gas production (methane and sulphide mainly) and operating problems. Thus, the first flush after a long dry period mobilize a huge amount of solids which can damage pumping stations at Wastewater Treatment Plants (WWTP) [2]. To avoid or minimize such problems is essential to have review existing sewer design criteria and optimize the operation and maintenance practices accounting for sediment accumulation and erosion processes. Past decades studies were focused in measurements in real sewers or studies of sewer hydraulics in full scale models in order to approach the knowledge of sediment accumulation and transport processes in sewers [3]. Furthermore, most of sediment transport formulas for sewer systems have been developed from studies with non-cohesive and uniform sediments [4]. Current studies are focused on understanding the sediment cohesion, with a variety of particle sizes, and integrating other phenomena such as organic matter transformations [5]– [8]. Thus, working with real wastewater in the most controlled conditions is crucial (in terms of hydraulic and pollution aspects). In the SEDUNIT Project the above conditions are met. The planned activities and the proposed methodologies guarantee an advance in the state of knowledge. A key aspect that ensures obtaining new results is the availability of a Scientific Platform for Pipe flow measurements with real wastewater. This platform allows to work in

WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line) doi:10.2495/WP180011

2 Water Pollution XIV

very close to reality conditions but with a capacity to monitor, measure and characterize water and sediment flows and sediment that are not found in a field campaign inside a sewer system. 2 RESEARCH BACKGROUND The SEDUNIT Project is being developed together with two more Projects called OvalPipe I and II. These projects are headed by a local company that is developing a new plastic eggshaped pipe system with the aim of improving the sewer pipelines efficiency. For that, the OvalPipe I (from 2013 to 2015) was the first approach in order to define and to test the hydraulic and mechanical characteristics. The hydraulic research was performed in a metallic egg-shaped prototype placed at the R&D Centre of Technological Innovation in Building and Civil Engineering (CITEEC) of the University of a Coruña. The results of a series of experiments were numerically modelled and compared with an equivalent-area circular pipe [9]. As a part of the OvalPipe I Project, a flume test facility was also built inside the WWTP at A Coruña to compare the self-cleaning efficiency between a 315 mm circular pipe and a plastic egg-shaped prototype with real wastewater conditions [10]. This first plastic prototype was made with an injection molding process and its weaknesses were the continuity of the geometry and the pipe joints. New building processes are being developing in the OvalPipe II Project (from 2016 to 2018), which is focused on achieving a better fabrication process of pipes and joints between egg-shaped pipes and also between egg-shaped and circular pipes. The resulting prototypes are planned to be also tested in the flume test facility. In this work we present the experimental campaigns at the WWTP experimental flume focused in studying the accumulation and erosion processes as a part of the SEDUNIT Project. 3 LABORATORY CAMPAIGN The laboratory approach to the sediment transport of cohesive solids was performed in a flume test facility fed with real wastewater (Fig. 1). This flume is placed at the pretreatment facility of the WWTP at A Coruña (600,000 inhabitants). Thus, wastewater is driven through a pumping system to an inlet chamber that feeds the flume. The metallic bench presents a length of 10 m, a width of 0.8 m and an adjustable slope from 0% to 2%. Over this bench, two configurations of sewer pipes were placed. In the first campaign, a circular 315 mm PVC pipe was set together with an equivalent area egg-shaped pipe. In the second series of tests, two circular PVC pipes of 315 and 400 mm outer diameters were studied. Downstream, a tailgate is automatically controlled in order to fix boundary conditions. The flume is equipped with different hydraulic and wastewater load sensors. The wastewater characterization at the inlet chamber was monitored with turbidity (SOLITAX) and UV absorption (UVAS) probes. The signal of these sensors was calibrated with the

Figure 1: General view of the flume test facility.

WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)

Water Pollution XIV

3

Figure 2: Weekly variation of TSS input to the flume according to the weather conditions. analysis of the Total Suspended Solids (TSS) and the Chemical Oxygen Demand (COD) from wastewater samples taken every 6 hours with an autosampler. The resulting average TSS and COD concentrations were 224 ± 65 mg/L and 414 ± 80 mg/L respectively. The TSS analysis from inlet samples showed a smooth daily pattern and a negligible influence of the dry–wet weather conditions in the TSS loads (Fig. 2). Two experimental procedures were designed during the laboratory campaign. On one hand, short and long term accumulation tests were performed to analyze the sediment accumulation under different hydraulic conditions: slopes ranges of 1‰, 2‰ and 5‰ and flowrates ranging from 3 L/s to 4.6 L/s. In the cases where the deposition occurred, the focus was on how the hydrodynamics, the bed deposit heights, and the physicochemical properties of sediments evolved. On the other hand, erosion tests were executed with the same initial conditions but different consolidation times for the bed deposits after increasing the flowrate in each pipe up to 12 L/s. The objective of the erosion tests was to observe the influence of the biological transformations of the sediments in sediment erosion processes. A detailed description of the experimental procedures and conditions can be consulted in RegueiroPicallo et al. [10], [11]. The sediment deposits were daily recorded. For that, the flowrate was stopped and the pipes were carefully emptied to avoid the bed erosion. Once the pipelines were drained, ultrasonic sensors and imaging techniques were applied to obtain the deposit heights [10], [11]. Furthermore, manual samples of the bed deposits were grabbed from the same measurement apertures for the physicochemical analysis [12]–[14]. Fig. 3 presents the a boxwhisker plot with the main relevant sediment properties recorded in the tests: sediment density and mean diameter (d50), Total Solids (TS), ratio between Total and Volatile Solids (TS/VS), COD and Oxygen Uptake Rate (OUR). In the sediment accumulation tests, bed deposits occurred with mean flow velocities below 0.26 m/s in the 315 mm and 400 mm circular pipes while no presence of deposit were observed in the egg-shaped pipe (Fig. 4). Growth rates between 1.4 and 3.8 mm/d were obtained from the daily sediment height recordings in the circular geometries, depending on the mixing conditions in the head tank in order to resuspend the particles (see RegueiroPicallo et al. [10]). A value of 2.85 mm/d was obtained in past studies with similar conditions in a 300 mm inner diameter pipe [15]. Besides the hydraulics and the solid loads conditions, the bed deposits accumulation also depends on the wastewater hydrodynamics. The progressively growth of sediments in the circular pipes slightly increases the centreline velocity profiles, which were measured with a WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)


Nortek Vectrino© Acoustic Doppler Velocimeter. Fig. 5 shows the log-scale velocity acceleration for a 7 days accumulation test whereas the Reynolds shear stress remain quasiconstant. In this case, the shear stress is not influenced by the sediment accumulation and a steady deposition ratio can therefore be expected.

Figure 3: Physicochemical analysis of the sediment samples collected in the pipe inverts.

Figure 4:

Daily accumulation growth in the circular and egg-shaped pipes with different head tank mixing conditions. (Source: Regueiro-Picallo et al. [10], [11].)

Figure 5:

Evolution of centreline velocity profiles: (a) and Reynolds shear stress; (b) in a 7 days accumulation test. (Source: Adapted from Regueiro-Picallo [10], [11].)


Water Pollution XIV

Figure 6:

5

Evolution of the erosion rate: (a) and the organic matter parameters of the initial bed deposits; (b) for different consolidation periods. (Source: Regueiro-Picallo et al. [11].)

In the erosion tests, although the same inlet conditions different erosion rates were obtained due to the sediment consolidation time applied (3, 7 and 20 days), the erosion rates were calculated as the difference between the initial and the final bed deposits. For that, a photogrammetric technique called Structure from Motion (SFM) was utilised [11]. The initial sediment heights were set to an averaged value of 7.3 mm and 14.3 mm in the 315 and 400 mm pipe respectively. At the end of these erosion tests, it was observed that the eroded mass decreased as the sediment accumulation time in the pipe invert increases (Fig. 6(a)). This trend was also connected with the decrease of the organic matter indicators of the sediment samples; VS/TS, COD and OUR taken before the tests (Fig. 6(b)). Therefore, these relationships show that the bed deposits erodibility is affected by the biological activity of the sediment. The decrease of the organic content concentration and the sample degradation implies a higher cohesiveness of the bed deposits in long-term deposition. In contrast, ‘fresh’ sediment conditions were more sensitive to be resuspended. 4 FUTURE CHALLENGES The following steps of the SEDUNIT Project will focus on the validation of the results obtained in the laboratory test in a real sewer catchment and the analysis of the existing sewer sediment transport models. A field campaign is being developed in the metropolitan area of A Coruña in a series of pipelines with sediment accumulation problems. In these sections the sediment characteristics and the accumulation and erosion ratios are being measured and will be compared with the values obtained at Bens WWTP facility. The objective of this task is based on understanding the evolution and distribution of sediments in urban sewer systems. This knowledge would lead to better management practices and thus less maintenance procedures, which may cost from 300€/km/year to 10,000€/km/year in some particular cleaning activities [3], [15]. Concerning the in-sewer sediment transport models, different formulations are being revised in the literature in order to describe the sediment transport processes. Some authors reported reviews of the principal sediment transport equations applied in the field of sewer systems [4], [16], [17]. Classical formulas such as Meyer-Peter and Müller, Einstein Brown,



van Rijn or Novak and Nalluri describe the bed transport of alluvial non-cohesive particles. In addition, Rouse, Einstein or van Rijn researches, among others, defined formulas for the transport of suspended sediments. Furthermore, there are some models such as AckersWhite, May or Velikanov that determine the total sediment transport. Some of these formulations are implemented in numerical codes, for example in Infoworks ICM software. Most of these equations were developed using river sediments with a uniform distribution. This causes an important limitation in order to describe sewer sediments. The sediments that can be found in sewer systems present different origins and size classes. Thus, in-sewer processes such as the bed deposits cohesiveness or the erosion of the organic superficial layers could not be reproduced by many of these models. Some exceptions, for example the Skipworth model, derived from laboratory tests with synthetic cohesive particles [18]. This model associates the bed shear stress and the eroded bed height. Later, this model was modified after using the results of laboratory studies with organic and inorganic sediments [6]. More recent studies are focused on the erosion rate of cohesive and real sediments in a laboratory scale under different aeration conditions and accumulation periods [19]–[22]. The previous sediment transport models show that the bed shear stress is a key parameter in order to determine the volume of solids transported in sewer pipelines. Therefore, a methodology needs to be applied to obtain its value. For that, the shear stress contribution of the flow and the pipe contour should be considered [23]. Additionally, some studies included the contribution of the roughness introduced by the bed forms [7]. 5 CONCLUSIONS The SEDUNIT Project tried to advance knowledge of the sediment deposition inside the combined sewers. From the laboratory facility installed in a WWTP, different pipe geometries and experimental procedures were performed. First, an egg-shaped pipe was parallel tested with a commercial circular pipe with an equivalent area. The experiments performed suggest that the egg-shaped pipe presents a higher sediment transport capacity, due to the higher velocity and shear stress under low-flow conditions. As the combined sewer pipelines are subjected to dry weather conditions most of the time, the egg-shaped pipes become a solution in order to reduce the sediment accumulation. This kind of geometries are well-known for big collectors but the fabrication processes for small and plastic pipes is still under research. The accumulation experiments in circular pipes of 315 and 400 mm outer diameters showed a linear accumulation within the first 7–11 days of accumulation. The accumulation rate was ranged from 1.4 and 3.8 mm/d. This quasi-constant evolution of the bed-deposits during the first days affects to the centreline velocity profiles. The velocity profiles were accelerated as the bed deposits increased, while the Reynolds shear stress kept constant. Furthermore, erosion tests were performed after a certain deposition period. Although the initial sediment height and flowrate conditions were identical in all cases, the eroded mass decreased as the consolidation time increased. The same trend was observed in the analysis of the organic content parameters from sediment samples taken before the erosion test beginning. Therefore, the erosion experiments proved that the bed strength is connected with the biological processes of the sediments. The ‘aging’ of sediments is a parameter that should be included in the sediment transport models. The eroded mass and therefore the release of pollutant loads in the wastewater depend on this parameter. Classical formulations were derived from alluvial sediment experiments without considering cohesive particles. Current sewer sediment equations were calibrated from cohesive and non-cohesive particles experiments or from new approaches where the bed shear stress is the main parameter. Further steps in sewer sediments models


Water Pollution XIV

7

should consider the evolution of the bed deposits properties in order to design new pipelines, maintenance strategies or estimate the impacts in the receiving media. ACKNOWLEDGEMENTS The SEDUNIT Project was supported by the MINECO and FEDER funding (Ref. CGL201569094-R). The construction of the flume test facility and part of the assistance in the experimental campaign was funded by the Projects OvalPipe I (Ref. ITC-20133052) and OvalPipe II (Ref. RTC-2016-4987-5) powered by the company ABN Pipe Systems S.L.U. and with the support of EMALCSA and EDAR Bens S.A. (MINECO/FEDER, EU). The research work of Mr. Juan Naves was financed by the Spanish Government grant (FPU14/01778). [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

REFERENCES Suarez, J., Puertas, J., Anta, J., Jácome, A. & Álvarez-Campana, J.M., Gestión integrada de los recursos hídricos en el sistema agua urbana: Desarrollo Urbano Sensible al Agua como enfoque estratégico. Ingeniería Agua, 18, pp. 111–123, 2014. Suarez, J. & Puertas, J., Determination of COD, BOD, and suspended solids loads during Combined Sewer Overflow (CSO) events in some combined catchments in Spain. Ecological Engineering, 24(3), pp. 199–217, 2005. Ashley, R., Bertrand-Krajewski, J.L., Hvitved-Jacobsen, T. & Verbanck, M., Solids in sewers. Scientific & Technical Report No. 14. IWA Publishing, London, UK, 2004. Bertrand-Krajewski, J.L., Modelling of Sewer Solids Production and Transport. Cours de DEA “Hydrologie Urbaine”, Transport. INSA de Lyon, Lyon, France, 2006. Vollertsen, J. & Hvitved-Jacobsen, T., Resuspension and oxygen uptake of sediments in combined sewers. Urban Water, 2(1), pp. 21–27, 2000. Rushforth, P.J., Tait, S.J. & Saul, A.J., Modeling the erosion of mixtures of organic and granular in-sewer sediments. Journal of Hydraulic Engineering, 129(4), pp. 308– 315, 2003. Banasiak, R., & Verhoeven, R., Transport of sand and partly cohesive sediments in a circular pipe run partially full. Journal of Hydraulic Engineering, 134(2), pp. 216–224, 2008. Seco, I., In-sewer organic sediment transport. Study of the release of sediments during wet-weather from combined sewer systems in the Mediterranean region in Spain. PhD thesis, UPC, Barcelona, Spain, 2014. Regueiro-Picallo, M., Naves, J., Anta, J., Puertas, J. & Suárez, J., Experimental and numerical analysis of egg-shaped sewer pipes flow performance. Water, 8(12), pp. 587–596, 2016. Regueiro-Picallo, M., Naves, J., Anta, J., Suárez, J. & Puertas, J., Monitoring accumulation sediment characteristics in full scale sewer physical model with urban wastewater. Water Science and Technology, 76(1), pp. 115–123, 2017. Regueiro-Picallo, M., Anta, J., Suárez, J., Puertas, J., Jácome, A. & Naves, J., Characterization of sediments during transport of solids in circular sewer pipes. Water Science and Technology, in press, 2018. APHA, AWWA & WEF, Standard Methods for the Examination of Water and Wastewater. 20th ed, American Public Health Association/American Water Works Association/Water Environment Federation, Washington DC, USA, 1998. ISO, ISO 2591-1:1988. Test Sieving – Part 1: Methods using test sieves of woven wire cloth and perforated metal plate. International Organization for Standardization, Geneva, Switzerland, 1988.



[14] ISO, ISO 13320:2009. Particle size analysis. Laser diffraction methods. International Organization for Standardization, Geneva, Switzerland, 2009. [15] Lange, R.L. & Wichern, M., Sedimentation dynamics in combined sewer systems. Water Science and Technology, 68(4), pp. 756–762, 2013. [16] De Sutter, R., Rushforth, P., Tait, S., Huygens, M., Verhoeven, R. & Saul, A., Validation of existing bed load transport formulas using in-sewer sediment. Journal of Hydraulic Engineering, 129(4), pp. 325–333, 2003. [17] Banasiak, R. & Tait, S., The reliability of sediment transport predictions in sewers: influence of hydraulic and morphological uncertainties. Water Science and Technology, 57(9), pp. 1317–1327, 2008. [18] Skipworth, P.J., Tait, S.J. & Saul, A.J., Erosion of sediment beds in sewers: Model development. Journal of Environmental Engineering, 125(6), pp. 566–573, 1999. [19] Tait, S.J., Marion, A. & Camuffo, G., Effect of environmental conditions on the erosional resistance of cohesive sediment deposits in sewers. Water Science and Technology, 47(4), pp. 27–34, 2003. [20] Banasiak, R., Verhoeven, R., De Sutter, R. & Tait, S., The erosion behaviour of biologically active sewer sediment deposits: observations from a laboratory study. Water Research, 39(20), pp. 5221–5231, 2005. [21] Schellart, A., et al., Detailed observation and measurement of sewer sediment erosion under aerobic and anaerobic conditions. Water Science and Technology, 52(3), pp. 137–146, 2005. [22] Seco, I., Valentín, M. G., Schellart, A. & Tait, S., Erosion resistance and behaviour of highly organic in-sewer sediment. Water Science and Technology, 69(3), pp. 672–679, 2014. [23] Ackers, J., Butler, D. & May, R.W.P., Design of sewers to control sediment problems. CIRIA. Report R141. Construction Industry Research and Information Association, London, UK, pp. 1–181, 1996.


Water Pollution XIV

9

LABORATORY EXPERIMENT ON GENERATION OF ANAEROBIC GAS AND SCUM FROM ORGANIC SLUDGE IN URBAN RIVERS SHIN MIURA1, TADAHARU ISHIKAWA2 & TETSUO HOTTA1 1 CTI Engineering Co., Ltd., Tokyo, Japan 2 Tokyo Institute of Technology, Japan

ABSTRACT Although the area served by sewers in Tokyo’s wards has reached 100%, 80% of this is combined sewer systems that were built before 1980, and when there is a storm runoff, organic sludge is discharged into rivers. Scums with malodor often appear in the brackish water reach of urban rivers due to the buoyancy caused by anaerobic gas produced in organic sludge deposition. In this study, a series of laboratory experiments were carried out using organic sludge collected from combined sewer systems to investigate the generation of anaerobic gas and scum. Water temperature, salinity, sediment ignition loss, and amount of sedimentation were controlled, and the gas components were analysed. The range of experimental conditions was determined from field observations of the Nomi River, which flows through the southern part of the Tokyo ward area. Gas analyses indicated that two-thirds of the anaerobic gas is methane and that hydrogen sulphide increases under high-salinity conditions. Based on the measurements of the rate of gas generation, a practical empirical formula was prepared that includes the effect of control factors. Observations of scum generation in the experiments indicated that generation time is inversely proportional to gas generation rate, and the amount of sediment formed is proportional to the initial volume of sediment accumulation. Keywords: urban river estuary, organic sludge, scum with malodor, anaerobic gas emission.

1 INTRODUCTION While the area served by sewers in the Tokyo ward area has reached 100%, 80% of this is combined sewer systems that were built before 1980 [1]. In rainy weather, organic sludge that have accumulated in pipes and culverts, along with fine-grained soil and sand are discharged into small and medium-sized urban rivers. The water quality in these rivers temporarily deteriorates due to these organic pollutants [2]. In particular, in downstream brackish-water reaches in which flow is stagnant, deposited organic sludge floats to the surface after rainfall as scum giving off a stench and causes environmental problems in surrounding areas. Based on site investigations in past research [3]–[5], the mechanism for scum generation is thought to be as follows (Fig. 1): (a) Rain water collected by the combined sewer system washes out the organic sludge: The rainwater collected by the combined sewer system washes out the organic sludge deposited in pipes during intense rain run-off. (b) Deposition of sludge: Sludge settles in the deep downstream reach due to flow velocity reduction. (c) Stratification and Anoxic: The bottom river water layer becomes anaerobic due to the oxygen consumption of the organic pollutants and density stratification. (d) Generation of anaerobic gases: Anaerobic gases such as methane are generated within the deposited organic sludge. (e) Surfacing of scum: Organic pollutants rise to the surface as scum with malodor due to buoyancy caused by the anaerobic gases.



Figure 1: The mechanism for scum generation. Hamada et al. [6] indicated that scum generated in enclosed bays is organic mud that sticks to itself due to the content of highly sticky oils and fats. In addition, Hotta et al. [7] suggested an empirical formula for the rate of anaerobic gas generation from experiments and field observations in canals, and Ushikubo et al. [8] performed gas generation rate experiments using samples collected in coastal areas. However, the organic matter content (ignition loss) of bottom mud used in this previous research was 30% or less, which is much lower than in post-rainfall sediment in urban rivers. In addition, anaerobic gas production rates determined in research on reducing greenhouse gas emissions vary widely in studies of methane gas output from marsh areas and the like [9]. However, these were results of measurements of emissions in a variety of fields, and the effect on water temperature, salinity, and other such influencing factors is not quantitatively understood. Therefore, in this research, we first gained an understanding of the water quality conditions in which scum forms based on records of observations that Ota City office performed on the Nomi River. In addition, basic laboratory experiments were carried out that simulated anaerobic gas generation and the rise of scum from organic sludge collected in the combined sewer system of the catchment basin to garner a quantitative understanding of its characteristics and generation rates. 2 FIELD OBSERVATIONS 2.1 Research location Fig. 2(a) shows the Nomi River catchment basin, and (b) shows a maximum depth of riverbed lateral profile. The water depth gradually increases from the brackish-water upper reaches near 5.7 KP to 3.8 KP, and the water depth is nearly constant from 3.8 KP to the


Water Pollution XIV

11

mouth of the river. Because of this, the organic sludge that flows out during rainfall is thought to readily deposit in the vicinity of 3.8 KP. In fact, although the section in which scum floats ranges from 5.0 KP to 3.0 KP, river patrols observed floating scum in the neighbourhood of 3.8 KP. 2.2 Method The Ota City Office Environmental Measures Division installed an automatic water quality measuring instrument (Horiba W-22XD) at a site at 3.8 KP on the Nomi River, which flows within the ward. It also recorded and classified the state of scum generation from visual observations during river patrols carried out once daily on weekdays as shown in Table 1. 2.3 Results Fig. 3(a) shows the daily rainfall and daily maximum hourly rain fall at the AMeDAS observation station in Setagaya City. Figs 3(b) and (c) show continuous observation records on a 50 cm river bed at 3.8 KP from April to October 2015, and Fig. 3(d) shows the state of scum generation. In (d), red dotted vertical lines were added on days in which a medium amount of scum developed (no “high amounts” of scum were generated in this time period). In many cases, rain had fallen 1 to 2 days prior to scum generation, and this rainfall is marked in red in (a). In addition, rainfall events after which scum did not develop within 2 days are marked in blue, and blue dotted vertical lines were added 1 day after such rains. From these data the following trends can be seen. The continuous observation record shows that at 3.8 KP the bottom layer salinity in normal periods is 10 to 30%. The bottom layer water DO was almost zero, and the ORP was –330 mV or lower, which is the level at which anaerobic gases are generated by methanogenic bacteria. However, when there is flooding, the salinity declines somewhat, and

Figure 2: (a) Nomi River catchment basin; (b) River-bed lateral profile.



Table 1: Classification of the amount of scum.

High

Medium

Low

None

Figure 3: Results of field observations. DO increases slightly. Correspondingly, ORP does increase, but then frequently falls to −330 mV or less 1 to 2 days after small rainfalls (daily amounts of 50 mm or less). Medium-level amounts of scum generation frequently occur 1 to 2 days after a relatively small rainfall of 10 mm or more. Because the lag time for flood flow to arrive at the Nomi River is about 1 h, this 1- to 2-day time difference is mainly thought to be the time required to develop sufficient buoyancy, resulting from gas generation within sludge after its deposition. In addition, the frequency of scum generation increases in the summer (June to August) and decreases in the spring and autumn even with the same level of rainfall. Seasonal differences in water temperature are thought to play a role in this. On the other hand, there are many cases in which scum does not develop after daily rainfall amounts of 50 mm or more. The reason for this is believed to be that organic sludge deposition is more difficult


Water Pollution XIV

13

due to increased flow velocity, so no sufficiently buoyant sludge layer develops that could generate anaerobic gases. In addition, the salinity in May and October is frequently 20% or higher, and it is also possible that methane formation is inhibited by high salinity, so scum does not readily develop. Basic experiments were carried out related to anaerobic gas and scum generation on sedimentary sludge collected from within the sewer system in the Nomi River catchment basin, taking into consideration the actual conditions above. 3 LABORATORY EXPERIMENTS 3.1 Experimental method 3.1.1 Experimental apparatus A fixed amount of organic sludge and salinity-adjusted river water was poured into a 45 mm inner diameter, 250 mm high syringe and stirred, and the salinity within the syringe was measured. Next, a piston was inserted, and after discharging air from the injection hole, the syringe was blocked closed and left to stand upright with the piston side up in a thermostatic chamber to track changes inside the syringe (Fig. 4). 3.1.2 Test samples Sediment from under manholes was collected in two places in the middle of the Nomi River catchment basin drainage district. Table 2 presents the results of the analysis of these samples. It can be seen that the samples had a large amount of organic content, were slightly acidic, and had become anaerobic. In addition, electric conductivity was about the same as for fresh water. These samples were used in experiments untreated. 3.1.3 Experimental conditions From the site conditions described in 2, experiments were carried out varying the four factors presented in Table 3. A 5 cm buoyant sludge layer, 25°C water temperature, 10% salinity, and 70% loss on ignition were considered to be standard conditions. To investigate the influence of each factor, each factor was varied in the range shown in the right column. However, because the buoyant sludge layer thickness changes due to consolidation, the values in the chart are just guidelines, and the final result was adjusted according to the sludge input per unit area (mg/cm2). The water temperature was controlled by a temperature setting on the thermostatic chamber, and the salinity was adjusted by diluting sea water collected at the mouth of the Nomi River. In addition, the time to scum generation at the site was taken into account, and the test period was set to 2.5 days (60 h). Note that because of the limitations of the experimental equipment, measurements of DO and ORP in the water column could not be carried out. However, because the height of the experimental water column was about 1/50 of the site water depth (approximately 3 m), the anaerobic gas in the syringe was thought to have developed in an hour or less. 3.1.4 Items to be measured and analysed A camera was installed laterally, and conditions within the syringe were photographed every 10 min. The buoyant sludge layer thickness (Hsed), the height of the column of water (Hliq), the scum thickness (Hscum), and the height of gas (Hgas) were measured. The frictional resistance of the piston was confirmed to be very small (0.02 atm or less) in preliminary testing. In addition, the gases generated in 3 cases of differing salinity (with other factors at



Initial

Hgas

Gas generation Floating scum

Hscum Hliq

Swelling of the organic sludge

H

Hsed

Figure 4: Experimental apparatus. Table 2: Organic sludge samples.

Sample 1 Sample 2

COD [mg/g-dry]

Ignition loss [%]

pH [-]

ORP [mV]

EC [mS/cm]

286 275

70 79

6.64 6.71

-123 -68

0.37 0.22

Table 3: Experimental conditions.

Mud thickness Water temperature Salinity Ignition loss

Standard condition

Range

5 cm 25°C 10 70%

2.2–8.1 cm 15–30℃ 0–30 70%, 79%

standard condition) were analysed for methane (CH4) and carbon dioxide (CO2) content by TCD (Thermal Conductivity Detector) and for hydrogen sulfide (H2S) content by FPD (Flame Photometric Detector). 3.2 Results Fig. 5 shows the changes within the syringe for standard conditions using sample 1. From the start of the test to t=2 hours, the buoyant sludge layer contracted slightly because of consolidation. At t=2 to 3 hours, the piston started to rise as the sludge expanded. That is, it is thought that gas was generated in the sludge layer, and voids increased. At t=3 hours, twothirds of the sludge layer separated to become scum and rose to the bottom surface of the piston. This is thought to be because more buoyancy was created by fine bubbles generated in the sludge layer. At t=6 h, the rate of rise of the piston having reached a constant rate. It is thought that gas generation rate had reached steady state at this point. From t=12 h onward, a gas layer emerged between the scum and the bottom of the piston, and a free water surface formed. In conjunction with this, the scum gradually settled, and the sludge layer grew


Water Pollution XIV

15

thicker. This is mainly thought to be because the fine bubbles within the scum coalesced and escaped upward. In addition, fine bubbles rising from the sludge layer were observed visually. Once the scum settled, it did not float again. In all test cases, much the same course was seen as described above, so the average rate of rise of the piston from t=12 h to 60 h, in which the piston rise was stable, was taken to be the gas generation rate per unit area. Changes in the gas generation rate caused by changing the test conditions were clarified and resolved below. 4 DISCUSSION 4.1 Changes in gas generation rate with varied sludge input Fig. 6 presents gas generation rates when only sludge input was varied. From this it can be understood that the amount of gas generated is by and large proportional to the sludge input. That is, it is inferred that the generation of anaerobic gas takes place roughly uniformly in the entire buoyant sludge layer. The relationship in the figure is approximated by the following equation, where P0: anaerobic gas generation rate at standard conditions [cm3N/cm2/day], and W: sludge input per unit area [mg-dry/cm2]: (1)

𝑃𝑃0 = 0.048𝑊𝑊 0 hour (start)

2 hours 3 hours 5 hours (floating 30 hours (consolidation) (swelling) scum) (increasing gas)

60 hours (scum disappearance)

16 14

Surfacing of scum

12

Height[cm]

10

Gas

Scum

8 6

Water

4 2 0

Organic sludge 0

6

12

18

24

30

36

42

48

Elapsed T ime[hour]

Figure 5: Changes within the syringe (standard case).


54

60


Incidentally, Hotta et al. [7] assumed from field observations in canals that the anaerobic gas generation rate and buoyant sludge layer thickness generally had a proportional relationship, and while similar to the results of this research, the proportionality coefficient is different. This point is considered more later on. 4.2 Changes in gas generation rate with varying water temperatures Fig. 7 presents the results of experiments carried out when only the water temperature was varied, taking θT to be the ratio of change from standard conditions (25°C). The rate of anaerobic gas generation is sensitive to water temperature, and there is a fourfold or greater difference between 20°C and 30°C. In addition, at 10°C and below, almost no gas is generated. One can argue that the fact that scum mainly develops in the summertime is due to this temperature dependence. In biogas plant experiments [10], a temperature dependence was exhibited, however, the difference between 20°C and 30°C was just under two-fold, and the effect was small compared to these experimental results. These facts indicate that temperature dependence differs with experimental materials and environmental conditions and suggest the need to confirm gas generation rates depending on the conditions of the intended field. These experimental results can be approximated by the following equation where θT: temperature correction factor, and T: water temperature [°C]: (2)

𝜃𝜃𝑇𝑇 = 0.0058(𝑇𝑇 − 10)1.9

4.3 Changes in gas generation rate with varying salinity Fig. 8 presents the results of experiments carried out in which only the salinity varied, taking θS to be the ratio of change from standard conditions (salinity 10%). Gas generation rate decreases rapidly with increased salinity, and when salinity exceeds 25%, the rate drops to almost zero. The reason for this is thought to be that with an increase in sulfide ions, the activity of sulfate-reducing bacteria that produce H2S dominates that of methanogenic bacteria that produce CH4, and the solubility of H2S is much higher than that of CH4 (80 times greater at 20°C). Consequently, it is believed that the relatively small 2.0

temperature correction factor (25 ℃)θT [-]

Anaerobic gas generation rate P0 [cm3N/cm2/day]

3.0 2.5

P0 = 0.048W

2.0 1.5 1.0 0.5 0.0

1.5

1.0

0.5

0.0

0

20

40

60

θT = 0.0058(T-10)1.9

10

15

20

25

35

30

water temperature T[℃]

Imput of organic sludge W [mg-dry/cm2]

Figure 6: Relationship of anaerobic Figure 7: Relationship of anaerobic gas

gas generation rate to input.

generation rate temperature.


to

water

Water Pollution XIV

17

amount of scum generated in the summer and autumn of 2015 as shown in Fig. 2 is related to both low water temperature and the bottom layer of water being high in salinity. These experimental results can be approximated by the following equation, where θs: salinity correction factor, and S: salinity [%]: 𝜃𝜃𝑠𝑠 = −0.08𝑆𝑆 + 1.8

(3)

Hanna et al. [9] arranged the amounts of methane gas generated in a large number of lakes and marshes. While the amount of methane generated was shown to be smaller when salinity was high, their definition of typical salinity and their method for measuring the methane generation rate in the various lakes and marshes was not standardized, so this remains a qualitative discussion. 4.4 Changes in gas generation rate as loss on ignition of bottom sediment varies The rate of anaerobic gas generation is thought to be highly related to the amount of source organic material (ignition loss). Fig. 9 presents the results of these experiments and the field observations of Hotta et al. [7], in which the average value of the experiments in samples having 70% loss on ignition were treated as the standard. It can be seen that the increase in gas generation rate relative to standard conditions θIL is mainly linear with respect to loss on ignition. These results are approximated by the following equation where θIL: ignition loss correction factor, and IL: ignition loss [%]:

4.5 Components of anaerobic gas

𝜃𝜃𝐼𝐼𝐼𝐼 = 0.014𝐼𝐼𝐼𝐼

(4)

Table 4 presents the components in the gas generated. A dash in the table indicates a component that was not analysed. The bottom row lists examples of reports from anaerobic fermentation of livestock excreta, edible oil residue, and the like at 10 biogas plant facilities [10] (referred to subsequently as biogas) for comparison. Approximately two-thirds of the gas generated in these experiments is CH4, which is about the same as biogas. Conversely, this gas contains less carbon dioxide than biogas. The reason for this is thought to be that, perhaps in addition to the difference in material to be decomposed, the difference in fermentation times may have an effect (20 days in biogas plants, 2.5 days in these experiments). Interestingly, when salinity reaches 20%, H2S exceeds CO2. Because H2S is extremely poisonous, its effect on the human body and aquatic organisms even at minute amounts cannot be ignored. 4.6 Suitability of empirical formulas From the results of eqns (2)–(5), the anaerobic gas generation rate Pa [cm3N/cm2/day] can be expressed by the following equation. Here, P0 is the gas generation rate at standard conditions in eqn (1), and θ is the correction factor for each influencing effect determined by eqns (2), (3), and (4). Fig. 10 shows the agreement of the above equation and the experimental results. 𝑃𝑃𝑎𝑎 = 𝜃𝜃𝑇𝑇 𝜃𝜃𝑆𝑆 𝜃𝜃𝐼𝐼𝐼𝐼 𝑃𝑃0 WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)

(5)


4.7 Scum generation characteristics As shown in Fig. 5, when the buoyancy associated with gas generation exceeds a threshold value, the organic sludge floats to the surface as scum. The relationship of the gas generation 1.6

sample1

1.4

sample2

2.0

1.5

ignition loss correction factor θIL [-]

salinity correction factor (10 salt ‰) θS[-]

2.5

θS = -0.08S + 1.8

1.0

0.5

0.0

1.2 1.0

0.6 0.4

5

10

15

20

25

sample1 sample2

0.2

observation results of Hotta et al. 0.0

0

θIL = 0.014IL

0.8

0

20

generation rate to salinity.

60

CH4 66.3 64.7 62.0 48–65

generation rate to loss on ignition.

CO2 11.7 7.4 4.8 31–42

N2 3–16

H2S 0.02 1.02 5.96 -

O2 0–4.3

Gas generation rate (simulation)[cm3N/cm2 /day]

4

3

2

1 sample1 sample2 0

0

1

100

Figure 9: Relationship of anaerobic gas

Table 4: Components of anaerobic gas. Salinity 0 Salinity 10 Salinity 20 Biogas

80

ignition loss IL[%]

salitiny S[‰]

Figure 8: Relationship of anaerobic gas

40

2

3

4

Gas generation rate (experoment)[cm3N/cm2/day]

Figure 10: Relation of empirical formula to experimental results.


Water Pollution XIV

19

rate and the time from test start to scum generation is plotted in Fig. 11. Although there is a lot of scatter in the data, there is an inversely proportional trend. It is considered that scum develops when the total amount of anaerobic gas generated exceeds the gas retention capacity of the sludge, so it is necessary to consider the gas retention capacity of the sludge in regard to sludge rise. In addition, the relationship of the risen scum thickness to sludge input exhibits a largely proportional relationship as shown in Fig. 12. 4.8 Scum generation characteristics As shown in Fig. 5, when the buoyancy associated with gas generation exceeds a threshold value, the organic sludge floats to the surface as scum. The relationship of the gas generation rate and the time from test start to scum generation is plotted in Fig. 11. Although there is a lot of scatter in the data, there is an inversely proportional trend. It is considered that scum develops when the total amount of anaerobic gas generated exceeds the gas retention capacity of the sludge, so it is necessary to consider the gas retention capacity of the sludge in regard to sludge rise. In addition, the relationship of the risen scum thickness to sludge input exhibits a largely proportional relationship as shown in Fig. 12. 5 CONCLUSION In this investigation, based on observations of scum generation during patrols of the Nomi River performed by the Ota City office, basic experiments were performed that simulated anaerobic gas and scum generation from temporary sedimentation of organic sludge associated with flooding. The main experimental findings are as follows: The generation of anaerobic gas started comparatively quickly after the water directly above the bottom mud became anaerobic, and a few hours later scum rapidly floated to the surface. In addition, after scum lost buoyancy and settled, gas was generated at a generally constant rate. The gas generation rate was by and large proportional to the sludge input.

•

12

10

8

maximum scum thickness[cm]

scum generation time [hour]

sample1 sample2

6

4

2

0

0

1

2

3

4

gas generation rate Pa [cm3N/cm2 /day]

Figure 11: Relationship of gas generation rate to scum generation time.

sample1 sample2

10 8 6 4 2 0

0

20

40

60

sludge input W [mg-dry/cm2]

Figure 12: Relationship of sludge input to scum thickness.



•

•

•

The anaerobic gas generation rate increased rapidly with increased water temperature, and there was an approximately four-fold difference between 30°C and 20°C. In addition, there was a negative correlation with salinity, and at salinities of 25% and higher the rate was almost zero. CH4 makes up approximately two-thirds of the anaerobic gas, followed by CO2. However, when salinity reaches 20%, H2S exceeds CO2. While the amount is not necessarily high, because H2S is very poisonous and causes milky turbid water [2], it is thought to be important in environmental predictions and assessments of urban rivers. When anaerobic gas generation rates increased, scum developed quickly, and scum thickness was positively correlated with sludge input.

ACKNOWLEDGEMENT The authors would like to thank the Ota City office, Tokyo Metropolitan Government for their support of the field survey and data collection. REFERENCES Bureau of Sewerage Tokyo Metropolitan Government, Sewerage in Tokyo 2016, p. 9, 2016. [2] Miura, S., Hotta, T., Negishi, H., Tsuruta, Y. & Ishikawa, T., Modelling of colloidal sulfur in salinity-stratified urban streams. Proceedings of the 37th IAHR World Congress August 13 – 18, 2017, Kuala Lumpur, Malaysia, pp. 3442–3449, 2017. [3] Yamazaki, M. & Tsukui, K., Study on the Formation of Scum in Rivers. (1) Study on the origin of scum, Annual Report of the Tokyo Metropolitan Research Institute for Environmental Protection 1991, pp. 171–179, 1991. [4] Yamazaki, M. & Tsukui, K., Study on the Formation of Scum in Rivers. (2) Results of Investigation for Sedimental Conditions in Shiratoribashi-Iidabashi Area of Kanda River. Annual Report of the Tokyo Metropolitan Research Institute for Environmental Protection 1991–1992, pp. 182–184, 1991. [5] Yamazaki, M. & Tsukui, K., Study on the Production of Scum in Rivers. (3) Change in the Sediment Characteristics by a Rainfall. Annual Report of the Tokyo Metropolitan Research Institute for Environmental Protection 1992, pp. 167–171, 1992. [6] Hamada, Y., Tanabe, H., Shimizu, N., Yoshioka, I., Mito, Y., Saitoh, T. & Hibino, T., Designing of mud treatment method for sludge in cove using GCA. Journal of JSCE B3 (Ocean Engineering), 68(2), I_1151–I_1156, 2012. [7] Hotta, T., Amano, M., Yamashita, Y., Chen, F.Y. & Shoji, H., Observation on malodor gas generation in coastal area. Annual Journal of Coastal Engineering, JSCE, 49, pp. 1101–1105, 2002. [8] Ushikubo, A., Takeshima, S. & Takai, Y., Sulfate reducing and methane fermentation in a brackish water ecosystem: a modeling experiment. Bulletin of the Society of Sea Water Science, Japan, 47(1), pp. 19–23, 1993. [9] Hanna, J.P., Brian, A.N. & Patrick, J.M., Salinity influence on methane emissions from tidal marshes. Wetlands, 31, pp. 831–842., 2011. [10] Schulz, H. & Eder, B., Biogas-Praxis. Grundlagen, Planung, Anlagenbau, Beispiele, Ökobuch, 2001.

[1]


Water Pollution XIV

21

IMPACT OF INDUSTRIAL AND MUNICIPAL WASTE-LOAD ON SKINNERSPRUIT IN GAUTENG PROVINCE, SOUTH AFRICA DAVID O. OMOLE1, BADEJO A. ADEKUNLE2,3, JULIUS M. NDAMBUKI3, ADEBANJI S. OGBIYE1, OLUMUYIWA O. ONAKUNLE1 & PRAISEGOD C. EMENIKE1 1 Department of Civil Engineering, Covenant University, Nigeria 2 Department of Civil Engineering, Federal University of Agriculture, Nigeria 3 Department of Civil Engineering, Tshwane University of Technology, South Africa

ABSTRACT South Africa’s semi-arid climate makes surface water a highly valued resource for the country. Typical of many developing countries, however, surface water quality is often lowered because of effluents which are discharged into nearby rivers, streams and lakes. In Gauteng Province, South Africa, Skinnerspruit is an important water body which flows eastwards from Hartbeerspoort (a lake in West Pretoria), and flows approximately parallel to the Magalies Freeway, passing through both residential and industrial estates in the capital city. It therefore serves as a sink for industrial and municipal effluent discharges which take place daily. Field sampling studies conducted on Skinnerspruit (as well as two canals that deliver effluents from industrial and residential estates into Skinnerspruit) in January and June focused on faecal coliform, dissolved oxygen (DO), and nitrate/nitrite as nitrogen. Other tested parameters included chemical oxygen demand (COD), pH, total suspended solid, and temperature. Some of the results showed that faecal coliform in the river had mean values of 111,444 upstream (U/S) and an attenuated mean value of 3,607 cfu/100 ml downstream (D/S). However, DO had mean concentration of 7.24 mg/l U/S but an improved value of 7.75 mg/l downstream. pH improved marginally from its alkaline state of 8.18 U/S to 8.13 D/S. Temperature also improved marginally from 17.55oC U/S to 17.04oC D/S. Nitrate worsened from to 1.31 to 3.19 mg/l D/S. COD improved from 27 to 20.11 mg/l. These results indicate that Skinnerspruit is heavily polluted, especially from faecal coliform. However, the river is responding positively through natural attenuation processes. Keywords: river, pollution, municipal waste, effluent, attenuation, dissolved oxygen.

1 INTRODUCTION Rivers serve diverse important functions for society, inclusive of water supply, transportation, recreation, food production, flood conduit, and sinks for effluent discharge from municipal and industrial activities [1], [2]. These functions of river systems are invaluable contributions to the sustenance to modern society and civilization. The river ecosystem, however, needs to be monitored and protected to ensure it continuous to serve its purpose sustainably. A river has capacity to take up waste matter and break it down over time, before returning to its original state [3]–[5]. However, this capacity is contingent on a variety of physical factors, including the need to monitor and control waste-loads, so that the natural capacity of the stream to break down foreign matter is not exceeded [3]. However, this is not always the case as rivers are relentlessly loaded with waste matter as a consequence of modernization, industrialization and national development [6]. South Africa has a rich network of surface water bodies, which includes several rivers. These surface water bodies constitute valuable natural water resources to a water-stressed nation [7]. South Africa is the 29th driest country out of 193 countries ranked with respect to total available fresh water resources per capita [8]. South Africa’s freshwater resources serve a variety of purposes with 62% going into agriculture; 27% to domestic, municipal, and industrial uses; and 3% to ecological conservation purposes [9]. The foregoing facts makes



it necessary to protect every available freshwater resource in the country. Therefore, the aim of the current research was to measure and monitor the concentration of pollutants being discharged into Skinnerspruit, a river that runs through commercial, industrial and residential districts of the City of Tshwane Metropolitan City [10], in order to determine the possible impacts of the pollutants on the river. 2 METHOD AND MATERIALS 2.1 Study area Skinnerspruit originates from Hartbeerspoort and travels nearly parallel to Magalies Freeway till it merges with Apies river, approximately 30km (as the crow flies) at Pretoria central (Fig. 1). Pretoria is a major industrial centre with pockets of industrial estates situated around the city [10]. Skinnerspruit path crosses industrial and residential areas, thus making it susceptible to waste discharges from both areas. The Kwagga (S1) and Transoranje (S2) areas consists of several industries as well as residential areas, while Rebeccastraat (S4) is mostly residential. Zeilerstreet Canal (S3) carries residential wastes which empties into Skinnerspruit (Fig. 1). Shortly after S5, Skinnerspruit merges with Apies river near Pretoria Central (Fig. 1). Water samples were taken along Skinnerspruit at S2, S4, and S5, while run-off draining from Kwagga into a canal was sampled at S1 and residential effluent draining through a canal into Skinnerspruit was sampled at S3 (Fig. 1). All samples were obtained between 0800 and 1000 hours. Grab samples were collected at each sampling point in January and June. South Africa has four seasons, Summer, Autumn, Winter, and Spring [11]. The samplings were designed to occur in the extreme seasons (Winter which peaks in June and Summer/rainy season which peaks in January). The summer is characterized by slight rainfall while winter is characterized by zero precipitation. Temperature ranges between 21oC to 27oC in the rainy season (Autumn and Summer or August and January) and between 8oC to 15oC in June [7]. 2.2 Sampling and sample analysis Physical parameters such as dissolved oxygen (DO), total suspended solids, pH, and temperature were determined in-situ using a hand-held HACH HQ40d portable meter which has an IntelliCAL probe. Chemical and biological contaminants were however determined by analyses of grab water samples which were transported to Daasport Wastewater Treatment Plant (DWTP) laboratory. Two batches of water samples were obtained for the laboratory analyses. The samples for chemical analyses were collected into 2-litre polyethylene bottles. The samples for determination of faecal coliform, however, were collected in 250 ml bottles and hermetically sealed. All samples were transported to the laboratory within an hour of sampling and analyses done upon arrival at the laboratory. COD was determined using titrimetric method and nitrate was determined using Hach UV screening model DR 6000. Faecal coliform was determined using spread plate method. 3 RESULTS AND DISCUSSION The data for January and June are presented in Tables 1 and 2. The data in Table 1 shows that a high concentration of faecal coliform is discharged from the storm canal at Kwagga road (S1) and the Zeiler street canal (S3) into Skinnerspruit. This faecal coliform load however became greatly attenuated by 87.94% at S1 and 99.53% at S3 in June at the peak of winter (Table 2). This reduction is attributable to the ambient temperature drop which is unfavourable to the bacteria, which is in tandem with the study of Lonsane et al. [12].



Water Pollution XIV

Figure 1: Outline of Skinnerspruit and sampling locations. (Source: edited Google map.)

23


Table 1: Result of analysis for January. Sampling station 1 2 3 4 5

COD DO Faecal coliform (mg/l) (mg/l) (cfu/100ml)

Description Storm water canal from PMP next to Kwagga Road Skinnerspruit at Transoranje road Zeilerstreet Canal before Skinnerspruit Skinnersruit at Rebeccastraat in Vom Hagen Street Skinnerspruit at Premos before Apies Confluence

(NO2+NO3) -N (mg/l)

TSS (mg/l)

pH

Temp (oC)

20

4.35

34,000

0.76

7.4

7.69

22.8

16

7.19

7,000

0.98

12.2

8.15

22.2

28

7.85

320,000

0.96

11.8

8.51

24.5

14

7.61

4,000

1.59

10.2

8.23

24.5

40

7.43

34,000

1.32

5.4

8.2

21.3

(NO2+NO3)N (mg/l) (mg/l)

TSS (mg/l)

pH

Temp (oC)

Table 2: Result of analysis for June. Sampling station 1 2 3 4 5

Description Storm water canal from PMP next to Kwagga Road Skinnersruit at Transoranje road Zeilerstreet Canal before Skinnerspruit Skinnersruit at Rebeccastraat in Vom Hagen Street Skinnerspruit at Premos before Apies Confluence

COD DO Faecal coliform (mg/l) (mg/l) (cfu/100ml) 16

5.82

4,100

0.98

12.4

8.96

13.41

26

8.37

100

0.64

2

8.16

8.84

32

7.73

1,500

1

2.4

8.22

14.73

42

9.02

7,800

1.79

3.6

8.22

12.14

24

7.34

37,000

1.47

0.4

8.2

8

Similarly, dissolved oxygen content of the river improved in June by 33.79% at S1 and 1.53% at S3. This is also in agreement with the study carried out by Omole et al. [3], [9] which shows that DO level of running streams improves with lowered temperature, among other factors. Conversely, Nitrate and Nitrite content of the effluent samples (measured as Nitrogen, N) increased during the winter in June (which is characterized by lowered temperature) by 28.95% at S1 and 4.17% at S3. According to Kim et al., [13] the oxidation rate of ammonia and nitrate increases with temperature rise. This means that when temperature is lowered, oxidation rate is reduced, and less nitrogen is produced, thereby leading to accumulation of nitrite and nitrate in the canals. The pH readings also indicate that both the effluent from the canals and water from the river are alkaline (Tables 1 and 2). Akaline water provides enabling environment for some alkaphilic bacteria and may make other bacteria grow some resistance [14], thereby altering the ecological landscape of the area. Temperature naturally responded to the seasons (winter and summer with generally lowered temperatures in June (Tables 1 and 2).


Water Pollution XIV

25

3.1 3.1 DO, nitrate, nitrogen, and TSS trend along Skinnerspruit Like what took place in the canals, DO level of the river increased in June above the DO level in January (Figs 2 and 3). Also, the river continued to be attenuated with respect to DO level, despite waste-loads from both canals (S1 and S2). The DO level increased but took a downturn after S4 (Fig. 3). This may be connected to increased pollutant load in the stream from other unidentified point and non-point sources of pollution along the river. The total suspended solids (TSS) was gradually and progressively attenuated between S2 and S5 in January as well as June. Nitrate and nitrite levels in the river was lower in winter (June) than in summer (January). This is the reverse of what occurred in the canals (Tables 1 and 2). This means that the complete oxidation of nitrogen compounds in the river was higher than in the canal. The complete oxidation of nitrite in the river could have been aided by velocity of the stream [15]. When the stream body velocity is increased, turbulence is introduced, thereby aiding oxidation. Moreover, the capacity of water to absorb DO is inversely proportional to temperature increase [15]. 3.2 Faecal coliform trend along Skinnerspruit The faecal pollution load on Skinnerspruit had a similar trend of progressive increment, both in January and June (Fig. 4). However, there seems to be no correlation between the faecal coliform load in the canal (which is being emptied into the river) and the faecal loading pattern in the river itself. In January, the faecal load in the river at S2 is lower than that at S1 (canal) (Table 1). This shows that the pollution had been abated between S1 and S2. Similarly, in January, the faecal load from the canal at S3 had been abated at S4. The sudden rise in faecal load at S5 therefore suggests that although the stream naturally attenuates the waste-loads, other unidentified sources continue to pollute the river downstream of the canals, between S4 and S5 (Table 1). A similar situation is repeated in June, thus confirming there is an independent pollution source along Skinnerspruit. A study of the map of Pretoria shows that at Pretoria Central, several peri-urban settlements surround Skinnerspruit. Such settlements contain markets, abattoirs and shanties where uncontrolled pollution activities might emanate.

Figure 2: Trend of DO, (NO3+NO2)-N, and TSS in January.



Figure 3: Trend of DO, (NO3+NO2)-N, and TSS in June.

Figure 4: Trend of faecal coliform along Skinnerspruit in January and June. 3.3 COD trend along Skinnerspruit Chemical oxygen demand (COD) is a measure of inorganic pollutants that might be present in effluents [16]. Data (Tables 1 and 2; Fig. 5) indicates that the highest levels of COD were found in portions of Skinnerspruit which traverses Pretoria Central area (S4 and S5). This, again, points to the fact that commercial activities and pollution from shanty dwellers in the area might be responsible for the heightened COD loading in the river. 4 CONCLUSION AND RECOMMENDATION The current study considered the pollution trends along Skinnerspruit in both summer and winter seasons vis-a-vis waste-loads into the river from two major canals in the city of Pretoria. It was observed that although the canals contributed high concentrations of


Water Pollution XIV

27

Figure 5: Trend of COD along Skinnerspruit in January and June. pollutants into the river, the river responded well to the pollutants from the river. However, pollutants at the latter portions of the studied segments along Skinnerspruit suggested the presence of other unidentified contributors of pollution to the river. It is therefore recommended that regulatory authorities investigate the activities taking place between Vom Hagen and Premos in Pretoria Central, with the aim of arresting any pollution trend that could jeopardize environmental integrity and public health. ACKNOWLEDGEMENT The authors appreciate the management of Covenant University for supporting this research and for sponsorship to the WITS conference. [1] [2] [3] [4]

[5]

REFERENCES Baron, J.S. et al., Meeting ecological and societal needs for freshwater. Ecological Applications, 12(5), pp. 1247–1260, doi: 10.2307/3099968, 2002. Ngene, B.U., Tenebe, I.T., Emenike, P.C. & Airiofolo, R.I., Statistical evaluation of hydro-meteorological data: A case study of Ishiagu in South-East Zone Nigeria. ARPN Journal of Engineering and Applied Sciences, 10(18), pp. 8192–8199, 2015. Omole, D.O., Longe, E.O. & Musa, A.G., An approach to reaeration coefficient modeling in local surface water quality monitoring. Environmental Modeling & Assessment, 18(1), pp. 85–94, doi: 10.1007/s10666-012-9328-0, 2013. Isiorho, S.A., Omole, D.O., Ogbiye, S.A., Olukanni, D.O., Ede, A.N. & Akinwumi, I.I., Study of reed-bed of an urban wastewater in a Nigerian community. Proceedings of the IASTED International Conference on Environmental Management and Engineering, EME 2014, doi: 10.2316/P.2014.821-00110.2316/P.2014.821-001, 2014. Tenebe, I.T., Ogbiye, A., Omole, D.O. & Emenike, P.C., Estimation of longitudinal dispersion co-efficient: A review. Cogent Engineering, 3(1), pp. 1–16, doi: 10.1080/23311916.2016.1216244, 2016.



[6] [7]

[8] [9] [10] [11] [12] [13] [14] [15] [16]

Kahoush, M., Bio-Fenton and Bio-electro-Fenton as sustainable methods for degrading organic pollutants in wastewater. Process Biochemistry, pp. 0–1, doi: 10.1016/j.procbio.2017.10.003, October, 2017. Abia, A.L.K., Ubomba-Jaswa, E. & Momba, M.N.B., Impact of seasonal variation on Escherichia coli concentrations in the riverbed sediments in the Apies River, South Africa. Science of the Total Environment, 537, pp. 462–469, doi: 10.1016/ j.scitotenv.2015.07.132, 2015. Gulati, M., Jacobs, I., Jooste, A., Naidoo, D. & Fakir, S., The water–energy–food security nexus: Challenges and opportunities for food security in South Africa. Aquatic Procedia, 1, pp. 150–164, doi: 10.1016/j.aqpro.2013.07.013, 2013. Omole, D.O., Badejo, A.A., Ndambuki, J.M., Musa, A.G. & Kupolati, W.K., Analysis of auto-purification response of the Apies River, Gauteng, South Africa, to treated wastewater effluent. Water SA, 42(2), p. 225, doi: 10.4314/wsa.v42i2.6, 2016. Pretoria population 2017. World Population Review, pp. 1–5, 2017. Mhlongo, S. & Mativenga, P.T., Water quality in a mining and water-stressed region. Journal of Cleaner Production, 171, pp. 446–456, doi: 10.1016/J.JCLEPRO.2017. 10.030, 2018. Lonsane, B., Parhad, N. & Rao, N., Effect of storage temperature and time on the coliforms in water samples. Water Research, 1(4), pp. 309–316, doi: 10.1016/00431354(67)90006-1, 1967. Kim, J.-H., Guo, X. & Park, H.-S., Comparison study of the effects of temperature and free ammonia concentration on nitrification and nitrite accumulation. Process Biochemistry, 43(2), pp. 154–160, 10.1016/J.PROCBIO.2007.11.005, 2008. Padan, E., Bibi, E., Ito, M. & Krulwich, T.A., Alkaline pH homeostasis in bacteria: New insights. Biochimica et Biophysica Acta (BBA) – Biomembranes, 1717(2), pp. 67–88, doi: 10.1016/j.bbamem.2005.09.010, 2005. Omole, D.O., Badejo, A.A., Ndambuki, J.M., Musa, A.G. & Kupolati, W.K., Analysis of auto-purification response of the Apies River, Gauteng, South Africa, to treated wastewater effluent. Water SA, 42(2), p. 225, 10.4314/wsa.v42i2.06, 2016. Adewumi, I. & Ogbiye, A.S., Using water hyacinth (Eichhornia crassipes) to treat wastewater of a residential institution. Toxicological & Environmental Chemistry, 91(5), pp. 891–903, doi: 10.1080/02772240802614648, 2009.


Water Pollution XIV

29

REAL-TIME MONITORING OF WATER QUALITY USING SMART SENSORS: FEEDBACK OF LARGE PILOT LAB TESTS AMANI ABDALLAH, ISAM SHAHROUR & MARWAN SADEK Civil and Geo-Environment Engineering Laboratory (LGCgE), Lille University, France

ABSTRACT The control of the drinking water quality is a major concern for public health. This control is generally conducted in laboratory, which requires long time. This type of control is not adapted for accidental or malicious pollutions, which can have serious consequences to the population health. Therefore, accurate real-time control of the water quality is required for ensuring a safe water supply. This paper presents results obtained within the European project SmartWater4Europe for the analysis of the capacity of an optical-based device (EventLab, Optiqua) to detect water contamination by chemical substances. Analyses were conducted in a large-scale Pilot Lab, which allows the injection of chemical substances in a water circuit under controlled conditions. Tests conducted with 4 chemical products (cadmium chloride, mercury dichloride, sodium hypochlorite and glyphosate) at different concentrations showed the capacity of EventLab device to detect the injection of concentrations exceeding 5 mg/L. Keywords: real-time, monitoring, contaminant, water, refractive index, Optiqua.

1 INTRODUCTION Access to safe drinking water is crucial for human health [1], [2]. Water contamination causes more than 14,000 deaths per day. In addition, drinking water infrastructure could be subjected to malicious acts by voluntary introduction of contaminants [3].Consequently, the real-time monitoring of the water quality is required to ensure both good water quality supply and to take early mitigation measurements in case of any water contamination [4]–[7]. The real-time detection of water contamination constitutes a complex task [8], because of the wide variety of contaminants (chemical, biological, radiological...) and the inherent difficulties to an early detection of each contaminant. Within the European Project SmartWater4Europe, we conducted a preliminary study using a large-scale Pilot Lab to explore the capacity of innovative devices to detect different kinds of water contamination. This paper presents the Pilot Lab and the results obtained with EventLab device, which is based on the measurement of the refractive index.

2 EXPERIMENTAL DISPOSITIVE AND PROCEDURE 2.1 Presentation of the Pilot Lab The design of the Pilot Lab aimed at the construction of an experimental device, which allows to inject “contaminated solution” under controlled condition in a water circuit and to track the response of water quality devices to this injection. Figs 1 and 2 show the Pilot. It is composed of 16 mm opaque double layer pipes, water tanks for filling and emptying, a system for injecting chemical or biological products, pumps, valves, manometers and flow sensors. The total length of water pipes is equal to 61 m. An injection system is used to introduce chemical or biological products in the water circuit. Sensors are used to track the pressure at different points of the water circuit. Water quality devices are connected at 41 m from the “contaminated solution” injection section. WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line) doi:10.2495/WP180041


Figure 1: Laboratory-scale network.

Figure 2: Experimental set-up at the sensor lab. 2.2 EventLab (Optiqua) device EventLab (Optiqua) is an optical-based device that measures the change in the water refractive index, using the Mach Zehnder Interferometry (MZI) principle [9], [10]. This Index is a good indicator of the water quality; because the presence of a substance in the water leads to a change in this index. EventLab operates at a sensitivity level of 10-7 in the refractive index. In addition to the MZI chip, EventLab includes electronics, software, data algorithms, and data communication [10]. The output signal of the sensor is the phase shift in the light passing through the sensor due to the presence of a substance. The phase shift variation is related to the variation in the refractive index. Since the water temperature affects the water refraction index [11], recorded data are adjusted in order to take into consideration the effect of temperature.


Water Pollution XIV

31

2.3 Experimental procedure Experiments were conducted as follows. First, the water circuit was cleaned for 30 minutes, then a baseline output was established by the injection of tap water for 60 minutes. After stabilization of the baseline, the “contaminated solution” was injected in the water circuit. By the end of the test, the water circuit was cleaned, disinfected and dried with compressed air to prevent the formation of bio films. 3 RESULTS Fig. 3 shows the results of an experiment conducted by successive injection of mercury chloride (HgCl2). It illustrates the variation in the compensated phase during the test. The time interval between two successive injections is 20 minutes, while the injection time is

Figure 3: Optiqua EventLab response as a function of the cumulative concentration level of HgCl2.

(a) Delta phase variation during injection test.

Added concentration [mg/L] 0 5 10 20 30 40 50 60

Maximum ΔΦ 0 0.17075 0.17097 0.20245 0.44088 0.49246 0.48220 0.56646

(b) Variation of the maximum delta phase with the solution concentration.

Figure 4: EventLab response to the injection of a solution with HgCl2.



equal to 3 minutes. The figure shows clearly the response of EventLab to the injection events and the increase in the amplitude of the response with the increase in the solution concentration. Experiments were also conducted with cadmium chloride, sodium hypochlorite and glyphosate. Results of these tests are summarized in Figs 5–7. These results show that EventLab detects the introduction of any of these substances at a concentration exceeding 5 mg/L. The relationship between the solution concentration and the phase increment is quasilinear with a regression coefficient higher than 0.8. Added Concentration [mg/L] 0 10 20 40 50 60 70 (a) Delta phase variation during injection test.

Maximum ΔΦ 0 0.03467 0.06450 0.09491 0.10000 0.08802 0.11785


Figure 5: EventLab response to the injection of a solution with CdCl2. Added Concentration [mg/L] 0 5 10 20 30 40 50 60 (a) Delta phase variation during injection test.

Maximum ΔΦ 0 0 0.042273 0.04300 0.101707 0.085288 0.086481 0.134593


Figure 6: EventLab response to the injection of a solution with NaClO.


Water Pollution XIV

Added concentration 0 5 10 20 30 40

(a) Delta phase variation during injection test.

33

Maximum ΔΦ 0 0.531639 1.069637 1.578406 2.178423 2.198035


Figure 7: EventLab response to the injection of a solution with C3H8NO5P. 4 CONCLUSION This paper included the presentation of an experimental investigation of the capacity of an optical-based device (EventLab) to detect in real-time water chemical contamination. Tests were conducted in a Pilot Lab within the European project SmartWater4Europe. They included the injection of different chemical substances (Cadmium chloride, Mercury chloride, Sodium hypochlorite and Glyphosate) at different concentrations. Tests showed that EventLab detected the injection of these substances at a concentration exceeding 5 mg/L. The relationship between the substance concentration and maximum delta phase is quasilinear with a regression coefficient higher than 0.8. This study shows that EventLab is efficient for a real-time detection of water chemical contamination. [1] [2] [3] [4] [5] [6] [7]

REFERENCES Helbling, D.E. & VanBriesen, J.M., Continuous monitoring of residual chlorine concentrations in response to controlled microbial intrusions in a laboratory-scale distribution system. Water Research, 42(12), pp. 3162–3172, 2008. Payment, P. & Hartemann, P., Les contaminants de l’eau et leurs effets sur la santé. Revue des sciences de l’eau. Journal of Water Science, 11, pp. 199–210, 1998. Guepie, F.C.B.K. et al., Vigires’ eau: système de surveillance en temps réel de la qualité de l’eau potable d’un réseau de distribution en vue de la détection d’intrusion, 2013. Jain, S. & McLean, C.R., An integrating framework for modeling and simulation for incident management. Journal of Homeland Security and Emergency Management, 3(1), 2006. Panguluri, S. et al., Distribution system water quality monitoring: Sensor technology evaluation methodology and results. US Environ. Protection Agency, Washington, DC, USA, Tech. Rep. EPA/600/R-09/076, 2772, 2009. Yang, Y.J. et al., Adaptive monitoring to enhance water sensor capabilities for chemical and biological contaminant detection in drinking water systems, 2006. van Wijlen, M. et al., Innovative sensor technology for effective online water quality monitoring. Proceedings of the 4th Singapore International Water Week, 2011.



[8]

Kemmerer, R.A. & Vigna, G., Intrusion detection: a brief history and overview. Computer, 35(4), pp. 27–30, 2002. [9] Heideman, R.G. & Lambeck, P.V., Remote opto-chemical sensing with extreme sensitivity: design, fabrication and performance of a pigtailed integrated optical phasemodulated Mach–Zehnder interferometer system. Sensors and Actuators B, Chemical, 61(1), (3), pp. 100–127, 1999. [10] Tangena, B. et al., A novel approach for early warning of drinking water contamination events. Proceedings of the 4th International Conference on Water Contamination Emergencies: Monitoring, Understanding, Acting, 11–13 Oct. 2010, Mullheim, 2011. [11] Thormählen, I., Straub, J. & Grigull, U., Refractive index of water and its dependence on wavelength, temperature, and density. Journal of Physical and Chemical Reference Data, 14(4), pp. 933–945, 1985.


Water Pollution XIV

35

GROUNDWATER QUALITY AND ITS DISTRIBUTION IN SILOAM VILLAGE, LIMPOPO PROVINCE, SOUTH AFRICA RACHEL MAKUNGO & JOHN O. ODIYO Department of Hydrology and Water Resources, University of Venda, South Africa

ABSTRACT Small scale subsistence agriculture aimed at ensuring food security at household level and lack of proper sanitary facilities make groundwater aquifers to be vulnerable to contamination. This is worse in rural communities that reside in semi-arid areas and are dependent on groundwater. Fertilisers from small scale subsistence agriculture and faecal matter from pit latrines contaminate groundwater aquifers. This study assessed the quality of groundwater from selected boreholes in Nzhelele area. A number of residents within Nzhelele area practice small scale subsistence agriculture and lack proper sanitary facilities. EC and pH and turbidity were measured using Cyberscan PC510 benchtop meter and Eutech TN 100 turbidity meter, respectively. 850 Professional Ion Chromatography and atomic absorption spectroscopy were used to analyse non-metals (nitrate, chloride, sulphate) and metals (copper, manganese, zinc, calcium, potassium, magnesium, and iron), respectively. EC, turbidity, chloride, nitrate and iron exceeded the recommended guidelines in some of the boreholes. Turbidity and nitrates had negative effects on human health. A borehole located within the vicinity of a small scale agricultural farm had elevated EC, turbidity, sulphate, magnesium, calcium, and copper. It is therefore crucial to urgently derive solutions aimed at minimising groundwater contamination and treatment of groundwater from the study area as it is the main source of domestic water supply. Spatial distribution maps and water quality classification based on nitrates and fluorides indicated class 4 groundwater quality which is dangerous and totally unsuitable for human consumption. These maps are therefore essential for interactive simple interpretation of water quality status, are useful as decision making tools even at locations where monitoring have not been done, and are useful in advising the residents on water quality parameters which may require treatment. Keywords: groundwater contamination, pit latrines, subsistence agriculture, water quality.

1 INTRODUCTION Limpopo is one of the poorest provinces in South Africa where groundwater is mostly used as a source of domestic water supply. This accounts for almost 70% of rural domestic water supply in Limpopo Province [1]. Due to its paramount importance as a source of water supply, there is a need to continuously monitor and know its water quality status. This assists in developing solutions for managing and preventing groundwater contamination and potential health effects. Maherry et al. [2] identified Limpopo Province as one of the areas of priority research and implementation of remediation technologies since communities rely on untested groundwater as the main source of drinking. Small scale subsistence agriculture is practiced to alleviate poverty and there is wide spread use of pit latrines in rural areas of Limpopo Province. This makes groundwater aquifers to be vulnerable to contamination. The need to increase agricultural productivity to sustain livelihoods in rural areas encourages intensive application of fertilisers in small scale agriculture. Farmers are still encouraged to use fertillisers to increase crop productivity and ensure long term food security [3]. For example, Odiyo et al. [4] found elevated turbidity, nitrate, calcium, magnesium, sulphate and chloride linked to agricultural activities or excessive use of fertilizers in the Soutpansberg area, South Africa. Thus, application of fertilisers need to be managed to avoid environmental problems such as groundwater pollution. Increased use of pit latrines may cause human and ecological health impacts



associated with microbiological and chemical contamination of groundwater [5]. Lack of safe drinking water and adequate sanitation measures have caused diseases such as cholera, dysentery, salmonellosis and typhoid, claiming millions of lives every year in developing countries [6]. The problem of contaminated groundwater is typical in most developing countries. Jeyaruba and Thushyanthy [7] found that 81% of the wells in a case study area in Sri Lanka were not suited for drinking due to the nitrate-N concentration. Kanyerere et al. [8] reported that most groundwater sources were not potable for domestic use in rural areas of Malawi and identified possible factors that contaminate water at specific sites. This included location of pit latrines upslope from water sources, groundwater-surface water interaction in lowlying areas, distance between water points and pit latrines, depths of water sources and topography. Arsenic and fluoride contamination of groundwater as well as poor sanitation facilities posed high health risks in rural areas of India [9]. Zamxaka et al. [6] considered poor sanitation and hygiene conditions and lack of, or little environmental awareness among the people in rural areas as the major causes of source water contamination in selected rural communities of the Eastern Cape Province, South Africa. Edokpayi et al. [10] reported that 87.5% of boreholes sampled in Limpopo Province, South Africa were not suitable for human consumption and posed carcinogenic risk. Odiyo and Makungo [11] evaluated the overall quality status of groundwater from private boreholes, implications for domestic use and possible sources of contamination in Siloam Village in Limpopo Province, South Africa. The study only used descriptive statistics (minimum, maximum, mean, and standard deviation) to describe overall water quality status of Siloam Village, and thus did not capture the water quality variation from one borehole to another. Following the latter study, this study conducted a close examination of water quality of each individual borehole to identify their water quality status and specific water quality problems. This assists in identifying boreholes with deviating trends which are likely to be indications of specific water quality problems. In addition, the study generated detailed geographical information system (GIS) maps indicating spatial distribution of groundwater quality in Siloam Village. This was aimed at integrating water quality results to enable simple interpretation of groundwater quality by the community, in addition to prediction of water quality parameters at unmonitored locations. The spatial distribution maps are based on colour coded classification system to enhance interactive and simplified interpretation of water quality status and identification of parameters of concern by communities. Spatial interpolation methods are frequently used to estimate values of physical or chemical constituents in locations where they are not measured [12]. This is because it is practically impossible to sample at all locations due to lack of access or cost implications. Spatial interpolation using GIS assists in ensuring comprehensive monitoring of groundwater quality. 2 STUDY AREA AND METHODS The study area is located in in the northern region of Limpopo Province in South Africa. Data from 11 boreholes (BH1-BH11) from Odiyo and Makungo [11] were used in this study (Fig. 1). Most residents are dependent on groundwater for domestic uses and small-scale irrigation of crops. Methods for groundwater sampling, quality control and analysis are provided in Odiyo and Makungo [11]. EC and pH, and turbidity were measured using Cyberscan PC510 benchtop meter and Eutech TN 100 turbidity meter, respectively. 850 Professional Ion Chromatography and atomic absorption spectroscopy were used to analyse non-metals (nitrate, fluoride, chloride, sulphate) and metals (copper, manganese, zinc, calcium, potassium, magnesium and iron). The sampling period was from August 2013 to January


Water Pollution XIV

Figure 1:

37

Location of boreholes in the study area. (Source: modified from Odiyo and Makungo [11].)

2014. Results on fluoride were not included in this study as they have already been discussed in Odiyo and Makungo [11]. The results were also compared with DWAF [13] guidelines to determine potential health effects and identify boreholes with excessive concentrations of water quality parameters and links to agricultural activities and pit latrines. Colour coded maps showing spatial distribution of groundwater quality were generated by interpolating mean groundwater quality parameters for the period August 2013 to January 2014 using inverse distance method in Quantum Geographical Information System (QGIS) version 3.0.1 software. The colour coding followed DWAF et al. [14] classification system (Table 1) which was aimed at interactive and simplified identification of groundwater quality and its suitability for domestic use. Table 1: Water quality classification. (Source: DWAF et al. [14].) Class Class 0 Class 1

Description Blue (ideal water quality) Green (ideal water quality)

Class 2

Yellow (marginal water quality) Red (poor water quality)

Class 3 Class 4

Purple (dangerous water quality)

Effects Suitable for life time use Suitable for use, rare instances of negative effects Conditionally acceptable. Negative effects may occur in some sensitive groups Unsuitable for use without treatment, chronic effects may occur Totally unsuitable for use. Acute effects may occur



3 RESULTS AND DISCUSSION Figs 2–4 show the water quality parameters in BH1-BH11 throughout the study period. pH values for BH1 were slightly above DWAF [13] recommended guideline of 6–9 in the months of December 2013 and January 2014, with pH values of 9.28 and 9.29, respectively (Fig. 2). DWAF et al. [14] noted that pH values between 9 and 9.5 do not have any potential health effects but only have a slightly soapy taste and insignificant effects on bathing. EC values for most boreholes (BH2, BH4, BH5, BH7, BH9 and BH11) were above the recommended guideline of 70 mS/m. However, it was only BH9 which had extremely high EC values, exceeding 400 mS/m. Water with EC values exceeding 370 mS/m is slightly corrosive, extremely salty and bitter, and has possible health risks [13]. Water from the rest of the boreholes with EC values from 70–150 mS/m had insignificant health effects on sensitive groups. Turbidity values were higher than the recommended guideline of 1 NTU in BH2 (Jan-14), BH4 (Aug-13), BH9 (Aug-13 and Jan-14). Turbidity levels for BH2 and BH4 are associated with slight chance of adverse aesthetic effects and infectious disease transmission exists since they fell in the range of 1–5 NTU. Water in BH9 had turbidity levels >10 NTU, which is associated with severe aesthetic effects and a chance of disease transmission at epidemic level due to infectious disease agents and chemicals adsorbed onto particulate matter as specified in DWAF [13]. BH9 was the only borehole with sulphate concentration above the recommended guideline of 200 mg/L (Fig. 3). Water with sulphate concentration above the guideline has noticeable slight taste and can cause diarrhoea in sensitive and some non-adapted individuals [13].

Figure 2: Physical parameters in groundwater. WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)

Water Pollution XIV

Figure 3: Non-metals in groundwater.

Figure 4: Magnesium, calcium, zinc and potassium in groundwater.


39


Chloride concentrations exceeded the recommended guideline of 100 mg/L for more than 3 months of the sampling period in BH2, BH7, BH10 and BH11. Chloride concentration ranging from 100–200 mg/L has no aesthetic or health effects but possibly result in increased corrosion rate in domestic appliances. The highest chloride concentration of 1177.4 mg/L, associated with objectionable salty taste, not slaking thirst and likelihood of rapid corrosion in domestic appliances, was recorded in BH9 in November 2013. BH3, BH4, BH5, BH6, BH7, BH9, BH10 and BH11 had nitrate concentrations that exceeded the recommended limit of 6 mg/L. Most of the concentrations exceeded 20 mg/L which DWAF [13] associates with methaemoglobinaemia and mucous membrane irritation in infants and adults, respectively. Odiyo and Makungo [11] linked high nitrate concentrations in BH3, BH4, BH6, BH7, BH8, and BH11 to groundwater contamination by faecal matter from pit latrines which were within the vicinity of the boreholes. The pit latrines were at maximum distance of 45 m, which was less than the recommended distance of >50m. Studies by Mudau [15] and Odiyo and Makungo [11] found microbial water quality indicators above recommended guidelines indicating high risk of infectious disease transmission in Nzhelele area. This showed that groundwater in most boreholes in Nzhelele area has been contaminated by human waste from pit latrines. The results show that BH9 dominantly had high EC and turbidity levels, and concentrations of sulphate, magnesium, calcium and copper as compared to the rest of the boreholes, though some of them were within recommended guidelines. Chloride, copper and zinc are some of the micro-nutrients that are required for plant growth [15]. Sulphate, magnesium and calcium are secondary plant nutrients. This indicates that fertilisers that are applied are likely to contain these parameters which are then leached into groundwater. BH9 is located in a small scale subsistence farm where fertilisers are applied to enhance crop growth. Thus, proper application and management of fertilisers is required to prevent elevated concentrations in groundwater in the future. The physical parameters and concentrations of most of the chemical parameters mostly increased during the rainfall months (Oct-13, Nov-13, Dec-13 and Jan-14) in BH9 while they were mostly comparable for the rest of the boreholes throughout the sampling period. Most crops are planted during the rainfall season and this is an indication that fertilisers contributed to elevated physical and chemical water quality parameters in BH9. Guo et al. [17] also noted that groundwater quality deteriorated in the wet season and this was linked to application of fertilisers in this season. Magnesium, calcium, zinc, potassium and copper concentrations were within the limits in all boreholes during the sampling period indicating no potential health effects (Figs 4 and 5). Iron concentrations in BH2 and BH8, in Jan-14 and Dec-13, respectively, exceeded 0.1 mg/L. Iron concentration within the range of 0.1-0.3 has very slight effects on taste and marginal other aesthetic effects but does not affect human health as stated in DWAF [13]. The results indicate that turbidity and nitrates had negative effects on human health. Studies by Odiyo and Makungo [11], [18] have already indicated that there is high fluoride in Siloam Village which is likely to originate from the geological formations of the study area. Thus, solutions aimed at minimising or reducing turbidity, nitrates and fluorides in groundwater from the study area are urgently required as groundwater is the main source of domestic water supply. Examples of spatial distribution maps showing distribution of EC, nitrates and fluorides are shown in Figs 6–8. EC indicated class 4 groundwater quality which is dangerous and totally unsuitable for human consumption around BH9. Nitrates and fluorides also dominantly indicated a similar case as that of EC, though this was throughout most of the study area for these anions. In Siloam Village, drilling of private boreholes is increasing at


Water Pollution XIV

Figure 5: Copper and iron in groundwater.

Figure 6: Spatial distribution of EC.

Figure 7: Spatial distribution of nitrates.


41


Figure 8: Spatial distribution of fluoride. an alarming rate as the community copes with water stress which has been exacerbated by the recent drought spells. The maps can therefore aid as decision making tools to assist in determining the status of groundwater quality at locations where monitoring have not been done. They are also useful in advising the residents on water quality parameters which may require treatment, even at unmonitored locations. For example, households where BH1, BH2 and BH8 are located or used can be advised to purchase filters that specifically treat fluoride since it was the only parameter with excessive concentration. In addition, households whose water sources have not been tested will be able to have an idea of specific water quality parameters that require treatment to make water suitable for human consumption. 4 CONCLUSION The study examined water quality parameters of individual boreholes in Siloam Village to determine their water quality status and identify point specific water quality problems. EC, turbidity, chloride, nitrate and iron exceeded the recommended guidelines in some of the boreholes. Turbidity and nitrates had negative effects on human health. BH9 dominantly had high EC, turbidity, sulphate, magnesium, calcium and copper as compared to the rest of the boreholes. This indicated that fertilisers applied within the vicinity of BH9 are potentially leached into groundwater resulting to contamination. High nitrates were linked to groundwater contamination by faecal matter from pit latrines. This in addition to elevated fluoride concentrations as reported from earlier studies makes residents of Siloam Village vulnerable to potential health risks. It is therefore crucial to urgently derive solutions aimed at minimising contamination and reducing turbidity, nitrates and fluorides in groundwater from the study area as it is the main source of domestic water supply. Spatial distribution maps and water quality classification based on nitrates and fluorides indicated class 4 groundwater quality which is dangerous and totally unsuitable for human consumption.


Water Pollution XIV

43

These maps are essential for interactive simple interpretation of water quality status by the community and can therefore be used as decision making tools even at locations where monitoring have not been done. They are also useful in advising the residents on water quality parameters which may require treatment. [1] [2] [3] [4]

[5] [6]

[7] [8] [9] [10]

[11] [12] [13] [14]

REFERENCES du Toit, W., Holland, M., Weidemann, R. & Botha, F., Can groundwater be successfully implemented as a bulk water resource within rural Limpopo Province? Analysis based on GRIP datasets. Water SA, 38(3), pp. 391–398, 2011. Maherry, A., Tredoux, G., Clarke, S. & Engelbrecht, P., State of nitrate pollution in groundwater in South Africa. Presented at CSIR 3rd Biennial Conference 2010, Pretoria, South Africa, 2010. Baiphethi, M.N. & Peter, J., The contribution of subsistence farming to food security in South Africa. Agrekon, 48(4), pp. 465–482, 2009. Odiyo, J.O., Makungo, R. & Muhlarhi, T.G., Investigating the impacts of geochemistry and agricultural activities on groundwater quality in the Soutpansberg fractured rock aquifers. WIT Transactions on Ecology and the Environment, vol. 182, WIT Press: Southampton and Boston, pp. 121–132, 2014. Tredoux, G., Talma, A.S. & Engelbrecht, J.F.P., The increasing nitrate hazard in groundwater in the rural areas. Presented at WISA 2000 Biennial Conference, Sun City, South Africa, 2000. Zamxaka, M., Pironcheva, G. & Muyima, N.Y.O., Microbiological and physicochemical assessment of the quality of domestic water sources in selected rural communities of the Eastern Cape Province, South Africa. Water SA, 30(3), pp. 333– 340. Jeyaruba, T. & Thushyanthy, M., The effect of agriculture on quality of groundwater: a case study., Middle-East J. Sci. Res, 4(2), pp. 110–114, 2009. Kanyerere, T., Levy, J., Yongxin, X. & Saka, J., Assessment of microbial contamination of groundwater in upper Limphasa River catchment, located in a rural area of northern Malawi. Water SA, 38(4), pp. 581–596, 2012. Mishra, D.S., Safe drinking water status in the state of Bihar, India: challenges ahead, Reviewed paper 202. 34th WEDC International Conference, Addis Ababa, Ethopia, 2009. Edokpayi, J.N., Enitan, A.M., Mutileni, N. & Odiyo, J.O., Evaluation of water quality and human risk assessment due to heavy metals in groundwater around Muledane area of Vhembe District, Limpopo Province, South Africa. Chem. Cent. J., 12(2), https://doi.org/10.1186/s13065-017-0369-y, 2018. Odiyo, J.O. & Makungo, R., Chemical and microbial quality of groundwater in Siloam Village, implications to human health and sources of contamination. IJERPH, 15(317), doi:10.3390/ijerph15020317, 2018. Murphy, R.R., Curriero, F.C., & Ball, W.P., Comparison of spatial interpolation methods for water quality evaluation in the Chesapeake Bay. J. Environ Eng., 136(2), 160171, 2010. DWAF, South African Water Quality Guidelines, Vol. 1: Domestic Water Use, 2nd ed., DWAF: Pretoria, South Africa, p. 216, 1996. DWAF, DoH & WRC, Quality of Domestic Water Supplies, Vol. 1: Assessment Guide, 2nd ed., Water Research Commission No: TT 101/98; DWAF: Pretoria, South Africa, p. 104, 1998.



[15] Mudau, T.C., Effects of pit latrines location on groundwater of Nzhelele Village within Limpopo Province. Honours dissertation, University of Venda, Thohoyandou, South Africa, 2011. [16] van Straaten, P., Rocks for Crops: Agrominerals of Sub-Saharan Africa, ICRAF: Nairobi, Kenya, p. 338, 2002. [17] Guo, X., Zuo, R., Meng, L., Wang, J., Ten, Y. & Chen, M., Seasonal and spatial variability of natural and anthropogenic and natural factors influencing groundwater quality based on source apportionment. IJERPH, 15(279). doi: 10.3390/ ijerph15020279, 2018. [18] Odiyo, J.O. & Makungo, R., Fluoride concentrations in groundwater and human health impact in Siloam Village, Limpopo Province, South Africa. Water SA, 38(5), pp. 731– 736, 2012.


Water Pollution XIV

45

COLUMN LEACHING HEAVY METAL FROM TAILINGS FOLLOWING SIMULATED CLIMATE CHANGE IN THE ARCTIC AREA OF NORWAY SHUAI FU1,2 & JINMEI LU1 Department of Engineering and Safety, UiT The Arctic University of Norway, Norway 2 Key Laboratory of Poyang Lake Environment and Resource Utilization, Ministry of Education, School of Resources Environmental & Chemical Engineering, Nanchang University, China 1

ABSTRACT This study aimed to assess how the current climate change perspective, with various air temperature (4°C, 10°C, 14°C and 18°C) affected metal releasing from tailings. Heavy metals pollution from tailings leaching are of increasing concern. Column leaching experiment was conducted for 15 weeks to a series of tailings with 20 mm/week water leaching four temperature situations. Leachate chemical physics properties and concentrations of Fe, Ni, Mn and Zn in leachates measured at each cycle. Multivariate statistical approaches to evaluate potential risk variations in leachate quality and identify temperature effect on heavy metals leaching in the Arctic area. Results showed higher temperature encourage oxidation and sulfuration in tailings that promoted heavy metal release from tailings through runoff and erosion. Ni, Zn and Mn have the similar resource from tailings and positive correlation in the leaching activity. The leaching of Fe was closely related to temperature change and affect the leaching of other metals. Temperature, however, increased risk by heavy metal leaching from tailings by temperature change should be caught more attention. Keywords: leaching, heavy metals, temperature, Arctic.

1 INTRODUCTION Tailings are a dominant component in mining waste and act as source of contaminants, which take serious risk to human health and ecological implications [1]. As tailings surround areas are densely polluted recent years, it has been established that tailings operate as an active edaphic compartment which performs a fundamental role in redistribution of metals to ecosystem. In the context, tailings dam has a significant heavy metals leaching contribution to surround environment. It is extremely important biogeochemical zones with the capability of altering the leachate of materials from tailings. Besides the natural processes such as weathering of tailings, considerable amount of metals generated by solution like acid mine drainage, rainfall leaching, etc. enter into deep soils and groundwater [2]. There are many factors affected heavy metals releasing and transporting [3], [4]. Some affected heavy metals form change and some affected acid mine drainage generation. Heavy metals leaching from tailings and acid mine drainage is produced when sulphide-bearing material is exposed to oxygen and water [5]. Many heavy metals leached from tailings when the acid mine drainage generated. Although this process occurs naturally, mining and climate change can promote acid mine drainage generation and tailings leaching simply through increasing the quantity of sulphide expose and reaction rate. There are many factors influence acid mine drainage generation and heavy metals leaching from tailings, such as temperature, precipitant, pH, salinity, conductivity and so on [6]. The degree of environmental pollution by tailings leaching is dependent on its composition, climate change and biochemical reaction, which in turn way vary depending on the geology of the tailings or sources, and surround environment. Temperature played an important role in heavy metals leaching from tailings, especially in the Arctic area. The Arctic has undergone dramatic change during the past decade. And temperature changed twice or more hence than the inland area. Which led



difference of tailings oxidation and sulfuration between Arctic area and inland. High temperature will accelerated oxidization and sulfidation of tailings, which promote acid mine drainage generation and enhance heavy metals release [7]–[10]. Treatment of tailings leachate which is composed of several dissolved toxic metals is too complex and expensive. If tailings leachate is not managed properly, it causes considerable environmental degradation, water and soil contamination, severe health impact on nearby communities, biodiversity loss and aquatic ecosystem [11]. The aim of this study is to identify leaching characteristics of heavy metals from tailings at different temperature in Arctic area. Knowing contribution of different temperature to heavy metals leaching, the employment of multivariate statistical techniques is benefit for studying their relationships. Which is good for establishing proper management strategies and a decision support system based on risk assessment criteria for improving the sustainability and safety of tailings leaching activities. 2 MATERIALS AND METHODS 2.1 Characteristic of study area There is no active mining of massive sulphide deposits in Norway today; but the operations have left behind tailings, waste rocks and adits that in many cases discharge low-pH, metalladen waste streams. As an important mining area in northern Norway and serious tailings deposit by open pit and underground mine, Ballangen faced the risk of metal release from tailings [12]. 7 million tons of tailings deposited in Ballangen, covering an area of 500,000m2 [13]. A large landfill was located in the coastal zone and is built with pond walls. The deposit took place in the years 1988–2002. A total of 8,537,468 tons of nickel ore was collected with an average content of 0.52% nickel. On top of the masses is a thin layer of soil, approx. 20cm. This layer is too thin to prevent air and water from coming into contact with the exhaust masses. Many heavy metals such as iron, copper, zinc, cadmium and nickel had a high content in the tailings and surround soil and water. All surface drainage from the mining area flows into the fjord, surround was noticeably affected by pollution from the mining area, and mainly affected by the heavy metals. It is also worth mentioning that residents in the surround, drinking water source until 2007, have been affected by cancer to a significantly greater extent than the national average. The average temperature for Ballangen municipality was used in the assessment to determine the temperature the samples should be stored in. The temperature is between 12°C and 17.1°C. In laboratory experiments, leaching activity can’t occur when the temperature below 0°C. Therefore, temperature range 5–18°C was chosen in this experiment. Highest temperature was chosen as 18°C, as the average temperature is expected to increase in the future as a consequence of climate change. The mean annual temperature and precipitation of Ballangen were 4.1°C and 1420 mm in 2016 (Ballangen metreological station located at 68°25′20′N, 17°27′28′E, eklima.met.no). 2.2 Experiment and chemical analysis A column experiment was conducted in the greenhouse to investigate the impacts of temperature change on heavy metals leaching from mine tailings. Four temperature degrees was set in the experiment: 5, 10, 14 and 18. Each treatment was established with a repetition. 8 columns were filled with mine tailings (Table 1) from Ballangen and then sent to 4 incubators to keep each at steady temperature, 600ml water (80 mm/month precipitation) were added in each column every two weeks to leach. Leachate collected each two weeks,


Water Pollution XIV

47

Table 1: Heavy metals content in mine tailings. Elements CaO MgO SiO4 Al2O3 Fe2O3 MnO Co Ni Zn

Units % TS % TS % TS % TS % TS % TS mg/kg TS mg/kg TS mg/kg TS

Tailings 3.18 27 39.5 4.47 17.3 0.165 38 77.8 48.6

pH, PE, TDS, salinity and conductivity of the leachates were measured at once by HI98193 [14]. In order to determine total concentrations of five heavy metals (Fe, Zn, Ni and Mn), tailings samples were subjected to microwave-assisted digestion with concentrated HNO3 according to ASTM 3682. Reference materials (CRMs) (GSS-16) as a control sample added in the digestion experiment was in the certified. Leachate was treated follow EPA 200.8. Heavy metals of leachate and tailings were determined by an inductively coupled plasma atomic emission spectrometry (ICP-AES). 2.3 Statistical analysis Basic statistics of the raw data was carried out by SPSS24.0 software. Correlated analysis were applied to the data set for identifying associations (common origin) between metals. 3 RESULTS AND DISCUSSION 3.1 Heavy metals in tailings In Table 1, the concentrations of heavy metals of tailings in the study area are shown. CaO, MgO, SiO4, Al2O3, Fe2O3 accounted 91.45% of the tailings (Table 1), suggesting a major of tailings dominated by heavy metals oxide. Sulphur showed low in the tailings and there were less sulphide metals. SiO4 (39.5%) was the highest content in the tailings, and the content of Cao (3.18%) and MgO (27%) in the original tailings were high, they were easy to create buffer solution to retard acid mine drainage generation. High content of CaO and MgO is easy to form a solid shell to prevent heavy metals leaching. According to the test, Co, Ni, Mn and Zn had a high content in the tailings and all of them far exceeds the background of Norway [15]. Although, Zn and Mn are essential element of the organism, a too high concentration level can also produce poisoning effect on the human body. Therefore, the total amount of heavy metals in the tailings carried a risk to the surround environment. The tailings used for the column leaching experiment under different scenarios of climate change showed high total concentrations of metals(Table 1). These concentrations were much higher than those of Norway soils background [15], [16], and with high content of metal oxidize by exposing to air for a long time [17]. It is evident that intense redox reaction occur in the tailings deposite [18]. Heavy metals will be activated by oxidized reaction in tailings, and the chemical forms of heavy metals will be changed [11]. Tailings oxidized is benefit for generating acid mine



drainage and heavy metals releasing. Couple with precipitant, many heavy metals will be leached from tailings to surround environment. Many factor will affect the tailings oxidizing, such as temperature, oxygen and precipitation. So, climate change will influence heavy metal storing and transporting in the tailings. Surface water and precipitation will scour and leach tailings, that will accelerate heavy metals release and transport from tailings [19], it take much oxygen to the tailings accelerate the reaction of tailings oxidized. Temperature will promote or restrain oxidized reaction to change heavy metals’ form, so as to change their store forms and transport ability [20]. The ability of heavy metals leaching from tailings various from temperature, heavy metals and forms [7]. 3.2 Characterisation of the leachate The variation of leachate’s physicochemical property presented in Fig. 1. The influence of temperature and time on leachates’ pH, TDS, salinity and SO42- was studied using column leaching experiment at 5, 10, 14 and 18°C. In Fig. 1(a), it is evident that leachates’ pH decreased with temperature rising and leaching cycle on. The highest values of pH showed at 5°C above 7, and lowest showed at 18°C, there was little change between 10°C and 14°C, it was even down to 4 when the temperature climb up to 18°C. Leachates’ pH at 10°C were lower than that at 14°C from first week to the 12th week, and opposite showed from 14th to 20th week. The results from leaching experiment indicated that TDS and salinity changed with the same trend. Both TDS and salinity had high values at the first leaching week, descend to a low value from 1st week to 5th week. From 5th to 15th week, they kept at this value with little variance. There was little change among the values of TDS and salinity at 10°C and 14°C, and their values of 18°C were apparently higher than that at 5°C. Metallic oxide takes up 91.45% of the tailings composition. Many metal oxide react with the leaching water, Mg2+, Na+ and Ca2+ dissolved to the water, which increased the values of salinity and TDS of the leachate. Although there was little metallic sulphide in the tailings, H+ and SO42generated with metallic sulphide and oxide hydrolysis at the same time. So pH values of the leachate were lower and SO42-contents were higher at the beginning of the leaching cycle. Higher temperature is benefit for tailings sulfuration and oxidation [11], which accelerated acid generation in the mine tailings leaching, so highest and lowest pH showed in 5°C and 18°C, respectively. The results of leaching concentrations of leachates from tailings columns as shown in Fig. 2. The highest and lowest leaching concentration of Fe was at 18°C and 5°C, respectively. There are much difference between leaching character Fe and the other measured heavy metals. Fe keep a low leaching concentration at 5°C with small change. Leaching concentration of Fe increased with leaching cycle on at 10 and 14°C, and higher showed at 10°C. Both the leaching concentration of Zn at 5°C and 14°C was low during the leaching experiment, they increased with the leaching cycle on, and the leaching concentration at 14°C was higher than that at 5°C. Higher leaching concentration was got 10°C and 18°C, they had a decreasing trend with leaching time. The highest leaching concentration at 10°C and 18°C got at the first leaching cycle, 7347 µg/L and7909 µg/L, respectively. After the first cycle, the leaching concentration at 10°C was higher than that at 18°C. Leaching concentration of Ni decreased with leaching cycle on at each test temperature. In each cycle, highest and lowest concentrations were at 18°C and 5°C, the leaching concentration at 10°C was higher than that at 14°C. The highest Ni leaching concentration is 609mg/L and 7mg/L, at 18°C and 5°C, respectively.


Water Pollution XIV

49

Figure 1: Variations of chemical physics properties of leachates. At the first cycle, the leaching concentrations of Mn are 574 µg/L, 2964 µg/L, 1470 µg/L and 7548 µg/L at 5°C, 10°C, 14°C and 18°C. Leaching concentration of Mn had a small change from the first cycle to 8th cycle at 5°C, 10°C and 14°C. At 18°C, its leaching concentration decreased with leaching time on. In the leaching test, highest and lowest leaching concentration of Mn were at 18°C and 5°C, and that of 10°C was higher than at 14°C. From the leaching cycle, all the leaching heavy metals had high and low leaching concentration at 18°C and 5°C, leaching concentration at 10°C was higher than that at 14°C. There are many heavy metal release kinetic equation in fitting heavy metal leaching from tailings, such as primary diffusion equation, parabolic equation and Elovich equation on heavy metal leaching tailings. All the leaching heavy metals fitted well with the first-order kinetic equation at each temperature. Cumulative concentration increased with leaching cycle on, and the fastest accumulation of the temperature is 10°C. 3.3 Effects of temperature on heavy metal leaching from tailings The temperature has significant impact on changes in leaching of heavy metals (as shown in Fig. 2). Temperature affect heavy metals leaching from tailings by change metal solubility and biochemical reaction [20]. Lower temperature decreased the metal ions solubility and metal sulphur oxidation reaction, so less heavy metals release to water at 5°C. Metal oxides



Figure 2: Leached concentrations of Fe, Zn, Ni and Mn from tailings. Table 2:

Cumulative equation of heavy metals leaching from tailings at different temperatures. Heavy metals Fe

Zn

Ni

Mn

Temperature

Heavy metals release equation

R2

5 10 14 18 5 10 14 18 5 10 14 18 5 10 14 18

y = 0.2124t + 2.6424 y = 68.285t – 209.63 y = 31.21t – 63.261 y = 209.49t + 146.6 y = 88.132t – 292.02 y = 1073.5t + 6834.6 y = 165.25t – 329.49 y = -259.78t + 11127 y = 0.6583t + 12.906 y = 46.782t + 574.11 y = 12.453t + 167.61 y = 20.196t + 804.66 y = 104.64t + 639,11 y = 1065.5t + 2486 y = 919.76t + 1274.1 y = 1228.7t + 9689.3

R² = 0,6275 R² = 0.8912 R² = 0.9795 R² = 0.9574 R² = 0.8331 R² = 0.9842 R² = 0.921 R² = 0.8105 R² = 0.5193 R² = 0.9162 R² = 0.7101 R² = 0.4518 R² = 0.9715 R² = 0.9927 R² = 0.9898 R² = 0.9212


Water Pollution XIV

Table 3:

T SO4 pH TDS salinity Fe Zn Ni Mn

51

Correlations between heavy metals leaching concentrations and chemical physics properties of leachates. T 1 0,358* -0,472** 0,052 0,045 0,652** 0,127 0,219 0,577**

SO4

pH

1 -0,408* 0,206 0,200 0,068 0,097 0,179 0,215

*

1 -0,021 -0,026 -0,061 -0,337 -0,219 -0,288

TDS

salinity

Fe

Zn

Ni

Mn

1 0,999** -0,048 0,471** 0,687** 0,541**

1 -0,053 0,480** 0,691** 0,542**

1 0,060 0,057 0,433*

1 0,887** 0,638**

1 0,802**

1

* Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).

and sulphides are easier to oxidize and hydrolyze at higher temperature, and higher temperature will rise water solubility of heavy metals ions, so higher leaching concentration showed at 10°C, 14°C and 18°C. There are significant positively correlation showed at T(temperature)-Fe and T-Mn, positively correlation showed at T-Ni and T-Zn. It is indicated temperature have positively effect on heavy metals leaching [21]. Because most of Ni was leached out after 6 weeks at 18°C, so it also had significant positively correlation with temperature. 10°C is the proper temperature of heavy metals oxidize and vulcanize in the leaching, more acid generate promoted heavy metals releasing. These results showed that the temperature had appreciable effect on the Zn, Ni and Mn leaching out at 10°C. 4 CONCLUSION The results definitely demonstrated that temperature change not only resulted in the heavy metal release in tailings but also led to variations of leachate characteristics. In addition to the heavy metal concentrations in tailings, heavy metal leaching was strongly associated with pH, temperature, salinity and TDS. The temperature of the fastest heavy metals accumulation is 10°C in the tailings leaching. Proper increase temperature will accelerate tailings oxidization and sulfidization, promote acid generation and increase TDS, finally promote heavy metals releasing. ACKNOWLEDGEMENT This study was financially supported by the MIN-NORTH project funded, Interreg Nord Program: Development, Evaluation and Optimization of Measures to Reduce the Impact on the Environment from Mining Activities in Northern Regions. [1] [2]

REFERENCES Azapagic, A., Developing a framework for sustainable development indicators for the mining and minerals industry. Journal of Cleaner Production, 12(6), pp. 639–662, 2004. Asa, S.C. et al., Application of sequential leaching, risk indices and multivariate statistics to evaluate heavy metal contamination of estuarine sediments: Dhamara Estuary, East Coast of India. Environmental Monitoring and Assessment, 185(8), pp. 6719–6737, 2013.



[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

[15] [16] [17] [18] [19] [20] [21]

Kozak, K. et al., The role of atmospheric precipitation in introducing contaminants to the surface waters of the Fuglebekken catchment, Spitsbergen. Polar Research, 34, 2015. Dijkstra, J.J., Development of a consistent geochemical modelling approach for leaching and reactive transport prosesses in contaminated materials, 2007. Akcil, A. & Koldas, S., Acid Mine Drainage (AMD): causes, treatment and case studies. Journal of Cleaner Production, 14(12), pp. 1139–1145, 2006. Dijkstra, J.J., Meeussen, J.C. & Comans, R.N., Leaching of heavy metals from contaminated soils: an experimental and modeling study. Environmental Science & Technology, 38(16), pp. 4390–4395, 2004. Visser, A. et al., Climate change impacts on the leaching of a heavy metal contamination in a small lowland catchment. J. Contam. Hydrol., 127(1–4), pp. 47– 64, 2012. Tyagi, R., Meunier, N. & Blais, J., Simultaneous sewage sludge digestion and metal leaching—effect of temperature. Applied Microbiology and Biotechnology, 46(4), pp. 422–431, 1996. Tsai, L.J. et al., Effect of temperature on removal of heavy metals from contaminated river sediments via bioleaching. Water Research, 37(10), pp. 2449–2457, 2003. Guo, Y.-G. et al., Leaching of heavy metals from Dexing copper mine tailings pond. Transactions of Nonferrous Metals Society of China, 23(10), pp. 3068–3075, 2013. Kefeni, K.K., Msagati, T.A. & Mamba, B.B., Acid mine drainage: Prevention, treatment options, and resource recovery: A review. Journal of Cleaner Production, 2017. Segalstad, T.V., Walder, I.F. & Nilssen, S., Mining mitigation in Norway and future improvement possibilities. America Society of Mining and Reclamation (ASMR), 2007. Iversen, E., Oppfølgende undersøkelser etter nedleggelse av gruvedriften ved Nikkel og Olivin AS, Ballangen kommune. Fysisk/kjemiske undersøkelser i gruveområdet i 2002–2007. Yan, Q. et al., Leaching experiments of experimental pollution caused by heavy metals of waste rocks in the copper mine: a cause study of Yaoyuanshan Ore deposit in the Fenghuangshan Copper Ore Field, Anhui. China Acta Geoscientica Sinica, 29(2), pp. 247–252, 2008. Skjelkvåle, B.L. et al., Trace metals in Norwegian surface waters, soils, and lake sediments-relation to atmospheric deposition, 2006. Skjelkvåle, B. et al., Heavy metal surveys in Nordic lakes harmonised data for regional assessment of critical limits, 1999. Iversen, E. & Berge, J., Nikkel og Olivin A/S Utredning av konsekvenser i forbindelse med nytt deponi på Fornes, 2001. Xiaojuan, S., Shulan, Z. & Lian, D., Leaching characteristics of MSW compost heavy metals in soil under different temperatures and simulated acid rain. Chinese Journal of Enviromental Engineering, 6(3), pp. 995–999, 2012. Duo, M., Leaching characteristics and releasing amount evalution of Mo tailing. Liaoning Institute of Technology, 2007. Tao, Y. et al., Precipitation and temperature drive seasonal variation in bioaccumulation of polycyclic aromatic hydrocarbons in the planktonic food webs of a subtropical shallow eutrophic lake in China. Science of the Total Environment, 2017. Chen, A. et al., Alkaline leaching Zn and its concomitant metals from refractory hemimorphite zinc oxide ore. Hydrometallurgy, 97(3–4), pp. 228–232, 2009.


Water Pollution XIV

53

PHYTOREMEDIATION OF WASTEWATER WITH THALIA GENICULATA IN CONSTRUCTED WETLANDS: BASIC POLLUTANTS DISTRIBUTION GASPAR LÓPEZ-OCAÑA, RAÚL GERMÁN BAUTISTA-MARGULIS, SERGIO RAMOS-HERRERA, CARLOS ALBERTO TORRES-BALCAZAR, ROCÍO LÓPEZ-VIDAL & LILIANA PAMPILLÓN-GONZÁLEZ Universidad Juárez Autónoma de Tabasco, México

ABSTRACT Constructed wetlands (CW) are efficient wastewater treatment technologies with low energy consumption. A constructed wetland with horizontal sub-surface flow was designed at a pilot scale involving Thalia geniculate as vegetation, with a wastewater loading rate of 204±66 L/day, using gravel as inert medium with a porosity and density of n= 56.3 ± 3.5 and 1666.7 ± 119.3 kg/m3, respectively. The reactor allows the biological treatment of 0.85 ± 0.05 and 0.66 ± 0.05 m3 of wastewater, with 4.2 days as a hydraulic retention time, favoring the removal of 85% of the average values of BOD with a k of -0.43 days-1. The pollutant analysis showed a pH value of 7.5 ± 0.1 in the reactor. The temperature (30.44 to 28.32°C), the electrical conductivity (4010 to 2922 μS/cm), the turbidity (144 to 17 UTN) and the bacterial biomass (30000 to 2646 mg/kg) decreased substantially from inlet to oulet across the reactor. The efficiency of the wastewater treatment in the CW is notable, nevertheless, keeping the appropiate hydraulic retention time is important in order to fully comply with the maximum permissible limits of 30 mg/L established in the Mexican environmental legislation (NOM-001-SEMARNAT1996). Keywords: wastewater treatment, macrophytes, subsurface flow, pollutant removal efficiency.

1 INTRODUCTION Constructed wetlands (CW) is not a new technology around the world. They have been studied because of its efficiency to remove organic matter through microbial degradation and settling of colloidal particles, pathogen elimination in domestic water and alternatives in construction design of wastwaster treatment [1]–[7]. Nevertheless, in Mexico and Latin America the applications of CW technology has been incipient, despite the technology has demonstrated the versatility of applications in small and medium urban areas, easy installation, operation and maintance, with highly competitive costs [8]. The most common type of constructed wetlands are the free water surface constructed wetlands (FWS-CW) and the horizontal subsurface flow constructed wetlands (HF-CW). In this context, it is estimated that CW plants require two or three growing seasons to achieve the maxium removal effiency [9]. The vegetation in the CW is important because the pollutant removal through direct assimilation into their tissues provides an adequate medium for microbial activity through the transport of oxygen to the rhizosphere, stimulating the aerobic degradation of organic matter and nitrifying bacteria growth [10], [11]. It has also been shown that the efficiency removal of contaminants depends on the support material and the hydraulic retention time (HRT) [12]. The communities interested in adopting this technology must develop CW based on local parameters [13], since the design can be influenced by hydrometeorological factors, which must be taken into account for the operation of the system [14], [15]. In an experimental study with vertical flow constructed wetlands (VF-CW) a bacterial culture of fungi and actinomycetes was used [4]. It was concluded that microorganisms play a key factor in the decontamination process and can be an indicator in the removal of



chemical oxygen demand (COD), biochemical oxygen demand (BOD), total nitrogen (TN), total phosphorus (TP) and total suspended solids (TSS). A train of treatment was also evaluated, consisting in three CW connected in series involving a sub-superficial flow (HF CW), using vegetation such as reed (Scirpus americanus), enea (Typha latifolia) and water lilies (Eichhornia crassipes). The system was stabilized in 44 days, having an HRT of 15 days in each CW, reporting high removal efficiencies of COD (71%), calcium (91%), chloride (77%), nitrite (82%), ammonium (99.9%), phosphate (77%) except for nitrate ion (36%) and electrical conductivity (93%) [9]. The effect on the removal of ammoniacal nitrogen in contaminated water was also studied from the Erh-Ren River in southeastern Taiwan [16]. The evaluation was performed through an experimental control and coupled HF and FWS, CW systems connected in series. Wetland vegetation such as Chinese grass (Pennisetum alopecuroides L.) and Pacific Island grass (Miscanthus floridulus) were involved in a HF-CW and FWS-CW respectively. These species did not survive to the winter due to the low temperature and the high salinity of the water, due to the salt intrusion from the sea water into the river. As a consequence, common reed (Phragmites australis) in both CW systems was shown, with an initial density of 2 plants per m2 and growing around 100 plants per m2 after 3 months. The control system during the experiment worked without species, concluding that season time affected the CW performance, particularly for the removal of ammoniacal nitrogen. Mexico has 110 CW systems throughout the country, three of them in the state of Tabasco, and only one operates in the Municipality of Centro at 64% of its capacity, because it was designed to treat a flow of 0.125 m3/s and currently operates with a flow of 0.080 m3/s [17]. In Tabasco-Mexico, the coverage of water treatment is low since 60 out of 93 wastewater treatment plants are known for their defficient performance operation. There is a notable predominance of conventional technologies and primary wastewater system, whose treatments are inefficient, and the costs have not yielded the expected results. In this respect, it is necessary to look for economic treatments that are easy to operate and appropriate for the climatic conditions and natural resources of the region, presenting technical and economical advantages over chemical treatment methods [17]. The objective of the current research work was to evaluate the phytoremediation potential of wastewater with Thalia geniculata in a CW, analizing the basic pollutan distribution and degradation at a pilot scale. CWs built in Tabasco consider HF and FWS system operated mainly with introduced species, such as Thypa latifolia. In the case of Thalia geniculata, it is also a native specie from Tabasco, but not studied so far and present advantages from introduced species in CW systems. 2 MATERIALS AND METHODS 2.1 Location of the pilot-scale CW The experimental HW-CW was installed at the Division Academica de Ciencias Biologicas (DACBiol), which is a campus from the Universidad Juarez Autonoma de Tabasco. The vegetation was collected in swampy areas from the Municipality of Centro, Tabasco. 2.2 CW design characteristics The reactor is 2.5 m long x 1.2 m wide x 1 m high [18]. The preparation began with the cleaning of the corrosion areas using an anticorrosive primer (white enamel finish layer) was applied in the external part. In the internal part, waterproofing of acrylic paste based on resins


Water Pollution XIV

55

and mineral fillers (1 cm thick) was applied. Afterwards, an elastomeric waterproofing layer was applied with a textile fiber reinforcing the internal part to avoid possible filtrations (five layers were placed). Once the reactor was waterproofed, all accessories, 1-inch hydraulic polyvinylide (PVC) pipes and fittings (valves, elbows, T’s, connectors, etc.) were installed for the supply and distribution of the wastewater. For natural aeration, internal sampling points were placed. Finally, 50 cm of mixed gravel was placed in the reactor, and then proceed for the stabilization phase of vegetation. 2.3 Planting and stabilization of vegetation The vegetation was placed into the gravel support medium. The stem size on the surface was 10 cm long and roots were placed 15 cm below the surface [8]. The reactor was fed with clean water at the beginning, maintaining a level of 40 cm of water for stabilization of the vegetation [19]. Thereafter, wastewater from the carcamus of the DACBiol was added to the CW. The stabilization phase in CW lasted six months from February to July 2016. 2.4 Hydraulic retention time, removal efficiency and degradation rate In the reactor, a mixed gravel support medium (crushed rock from the Teapa River, southern region of Tabasco) was placed and the hydraulic retention time (HRT) was calculated with the operation flow of the wastewater [8] HRT = n d A / Q,

(1)

where n is the porosity, d is the heigh of the support medium, A is the cross section of the reactor and Q is the water flowrate.

ɳ = [(C1 – C2) / C1] x 100,

(2)

where represents the removal efficiency in %, C1 the wastewater influent concentrationand C2 the wastewater effluent concentration. The behavior of wastewater is a first order kinetic reaction, the degradation rate k was estimated with the following eqn (3). Ko = - ln (Cn / Co) / τ,

(3)

where τ = retention time for BOD removal, Cn = BOD influent concentration of the reactor “n” (mg/L), Co = influent concentration, Ko = degradation constant. 2.5 Wastewater characterization In order to study the variables of the wastewater, sampling points were established and sampled throughout the reactor (Table 1, Fig. 1). Three samples were taken by triplicate per day during 5 days. For the kinetic study of the BOD, the influent and effluent of the reactor was monitored, taking a simple daily sample for 7 days, and up to one year of operation in the reactor.



Table 1: Wastewater methods of analysis applied for control parameter determinations. Parameter

Environmental regulation in Mexico

Temperature Turbidity Electrical conductivity pH Biological oxygen demand Total volatile solids

NMX-AA-007-SCFI-2000 NMX-AA-038-SCFI-2001 NMX-AA-093-SCFI-2000 NMX-AA-008-SCFI-2000 NMX-AA-028-SCFI-2001 NMX-AA-034-SCFI-2001

Figure 1:

The red points from M1 to M9 indicate the sampling point in the Subsurface Horizontal Flow – Constructed Wetland (HF-CW).

2.6 Biomass on the support medium The biomass, refering to the quantity of microorganisms on the rocks was determined by gravimetry adapting the total volatile solids (SVT) method to a sample of the support medium at each sampling point (Fig. 1). Each sample considers the density and porosity of the system [21].


Water Pollution XIV

57

2.7 Modeling of the pollutants distribution To achieve the modeling, the daily average of each sampling point referring to variables such as temperature, turbidity, electric conductivity, pH and biomass were monitored and analyzed. The pollutants distribution inside the reactor was plotted using the software Surfer 8.0 [22], which allows the determination of the spatial distribution within a coordinate system based on a linear interpolation and a quadratic diagram (isoconcentration map). 3 RESULTS AND DISCUSSION 3.1 Hydraulic retention time (HRT) The HF-CW was designed to operate with 200 L/day. However, when performing the corresponding gaugings and the volumetry of the wastewater flowrate, the HRT was estimated in 4.2 days. This value fulfilled the design criteria for CW established by several authors [8], [13], [15], [19] Fig. 2 shows the HRT at different operating wastewater flowrates. The average operating flow in this period was 204 ± 66 L/day, with 108 L/day and 370 L/day, as minimum and maxium operation wastewater flow values, respectively. 3.2 Degradation rates and kinetic coefficient When the wastewater flowrate exceeded 200 L/d, the HRT was observed to decrease. Therefore, the wastewater did not have enough time to be in contact with the microorganisms and vegetation, resulting in a low degradation. In order to analyse the influence of design parameters (loading rate, flow and temperature) on the degradation of pollutants in the wastewater, the removal of BOD was measured. It is important to note that facultative microorganisms eliminate part of the BOD through biological and physic process, mainly. The pollution removal rate in wastewater is related to HRT and temperature. The temperature

Hydraulic retention time (days)

20 18 16 14 12 10 8 6 4 2 0 50

100

200

300

400

500

600

Wastewater flowrate (L/day)

Figure 2: HRT for different operating wastewater flowrates in the HF-CW.



during the evaluation was 27°C on average. The HRT was 4.5 days and k was calculated in -0.43 days-1. The Mexican water regulations established a daily discharge average of 75 mg/L for the HRT parameter. This value can be achieved between the fourth and fifth day of operation, as shown in Fig. 3 [23]. The HF-CW (HRT = 6 days) removed nearly 90% of the organic pollutants in the wastewater and comply with the most stringent criteria established by the Mexican regulations for the aquatic life protection. Similar results were reported elsewhere [24], concluding that 8 days of HRT is adequate for the removal of organic matter at temperatures above 25°C. The kinetic degradation behavior of the organic matter in the HF-CW is described as a first order kinetic (Table 3). 3.3 BOD removal efficiency The maximum BOD removal efficiency was 92.8%. The average removal efficiency was found to be between 80 and 85%. Unlike the operation of other experimental reactors in series [19], [25], the current reactor is setup with a primary and secondary treatment, satisfying the regulations for wastewater discharge to rivers for urban public use (75 mg/L). Furthermore, it complies occasionally to the aquatic life protection limit (30 mg/L) [23]. 3.4 Spatial distribution of pollutants in the HF-CW The basic parameters of pollutants monitored in wastewater for the spatial distribution analysis can be seen in Table 3. Regarding the pH values, some differences were observed in the spatial distribution, but the values remained in the neutral range of 7.2–7.9 (Fig. 5). This variation may be explained by the type of substrate and biofilm employed in this investigation [25]. The temperature was 29.2 ± 0.8°C on average, with variation from the input to the output of the reactor with 30.5 to 28.1°C, respectively (Fig. 6). This temperature behaviour favors the growth and the stabilization of mesophilic microorganisms [26]. The electrical conductivity values were measured around 3442.7 ± 408.0 μS/cm, in compliance to the regulation for agricultural 400

BOD concentration (mg/L)

350 300 250 200 150 100 50 0 0

1

2

3

4

5

6

Time (days)

Figure 3: Degradation of organic matter in the HF-CW with Thalia geniculata vegetation.


Water Pollution XIV

Table 2:

59

Estimation of the kinetic degradation constants evaluated in the HF-CW with Thalia geniculata vegetation.

Days

BOD influent (mg/L)

1 2 3 4 5 6 7 Average

375.50 369.40 403.20 437.10 391.70 407.70 254.20 376.97

BOD effluent (mg/L) 66.20 65.30 66.50 65.90 43.30 29.50 50.50 55.31

k (d-1)

η (%)

-0.39 -0.39 -0.40 -0.42 -0.49 -0.58 -0.36 -0.43

82.4 82.3 82.5 84.9 88.9 92.8 80.1 85.3

Table 3: Basic pollutant parameters measured in wastewater for the HF-CW. Days Parameters pH Temperature (°C) 1 Conductivity (µS/cm) Turbidity (NTU) pH Temperature (°C) 2 Conductivity (µS/cm) Turbidity (NTU) pH Temperature (°C) 3 Conductivity (µS/cm) Turbidity (NTU) pH Temperature (°C) 4 Conductivity (µS/cm) Turbidity (NTU) pH Temperature (°C) 5 Conductivity (µS/cm) Turbidity (NTU)

M1 7.9 32.3

M2 7.7 32.2

M3 7.5 31.2

M4 7.8 29.4

M5 7.6 30.9

M6 7.5 31.3

M7 7.8 29.6

M8 7.4 29.5

M9 7.4 29.5

4160

4010

3770

3450

3640

3280

2760

2870

2530

226 7.5 30.5

176 7.5 30.8

176 7.5 29.7

165 7.6 27.2

171 7.5 27.5

110 7.5 26.7

23.1 7.5 26.7

17.6 7.4 25.1

6.9 7.2 25.1

4010

3680

3890

3070

2370

2950

2300

2240

2520

195 7.4 30.0

196 7.4 29.3

178 7.4 29.6

103 7.6 29.7

95 7.6 29.1

90 7.6 29.1

89 7.6 28.8

72 7.6 28.3

68 7.7 29

3950

3920

3920

2710

4020

3840

4000

3740

3640

42.0 7.3 29.9

42.6 7.5 29.2

42.6 7.8 29.5

12.4 7.4 29.8

12.5 7.5 29.3

8.71 7.8 29.4

3.64 7.4 29.6

2.8 7.6 29.1

2.1 7.8 29.4

3930

3930

3940

3940

4080

4000

3910

4080

3650

38.7 7.3 29.5

35.8 7.3 29.2

39.7 7.3 30.2

10.7 7.7 28.7

10.7 7.7 28.9

11.5 7.7 29.1

2.95 7.7 28.8

3.49 7.6 28.8

3.67 7.6 28.6

4000

4020

3860

3190

3190

3040

2300

2350

2270

220

227

210

81

83

76

3.1

2.57

6.73

irrigation in Mexico. Salinity is considered low, with values from 4000 to 2900 μS/cm (Fig. 7). The HF-CW reduced the salinity of the wastewater; thus water is suitable to be used without restriction for crops irrigation of [27].



Turbidity shows a significant decrease from influent to effluent in the HF-CW. This behavior is associated with Thalia geniculata vegetation, which is very effective at removing up to 87% of the sediments (Fig. 8). Finally, the biomass adhered to the support medium presented the highest concentration of microorganisms at the reactor inlet and decreasing significantly at the reactor outlet. The biomass concentration was found to be higher than 33,000 mg/kg (mg of biomass on kg of rock “medium support”) in the influent and reaching values of 3000 mg/kg at the end of the reactor. The support medium has a diameter of 2.8 ± 0.8 cm, porosity n = 56.3 ± 3.5 and density of 1666.7 ± 119.3 kg/m3. These characteristics allow a water volume of 0.85 ± 0.05 m3 and 0.66 ± 0.05 m3 of gravel within the HF-CW. A HF-CW experimental study reported BOD removals greater than 90% with Typha and Phragmites vegetation [28]; while a removal efficiency of 79% was observed with Typha latifolia for combined HF reactors [29]. Similar removal efficiencies for BOD (80%) were reported with Typha and Phragmites in the second year of evaluation of an HF-CW [30]. In this study, Thalia genicualata vegetation showed a high removal efficiency of BOD (85%). The BOD removal in a CW is generally high because the organic components are degraded aerobically and anaerobically by the bacteria adhered to the roots and rhizomes of the plants with the porous medium [31]. The water treatment in a HF-CW with a HRT of 6 days and temperatures around 28°C is sufficient to obtain removals greater than 90%. Similar results were concluded for 8 days of HRT for the removal of organic matter at temperatures above 15°C [24]. Thalia geniculata is not frecuently cited in the literature, however in this study it is demonstratred the great potential for wastewater treatment from this native vegetation in the tropical region of Mexico. The most important effects of emerging macrophytes in wastewater treatment are plant tissue, wind speed reduction that supports the sedimentation of suspended solids, filtering effect or adherence of microorganisms to the roots plants that can be a significant route for the elimination of nutrients, especially under low loading rates [32]. Therefore, Thalia geniculata is highly efficient for the removal of basic pollutants in the treatment of domestic wastewater.

2

2

UpH 1.8

Temperature (°C) 1.8

30.5 30.4 30.3 30.2 30.1 30 29.9 29.8 29.7 29.6 29.5 29.4 29.3 29.2 29.1 29 28.9 28.8 28.7 28.6 28.5 28.4 28.3 28.2 28.1

7.615 7.605 1.6

7.595

1.6

7.585 7.575

1.4

7.565

1.4

7.555 7.545

1.2

1.2

7.535 7.525 1

7.515

1

7.505 7.495

0.8

7.485

0.8

7.475 0.6

0.6

0.2

0.4

0.6

0.8

1

Figure 4: pH spatial distribution.

0.2

0.4

0.6

0.8

1

Figure 5: Temperature spatial distribution.


Water Pollution XIV

2

2

EC (µS/cm) 1.8 4050 4000 3950 3900 3850 3800 3750 3700 3650 3600 3550 3500 3450 3400 3350 3300 3250 3200 3150 3100 3050 3000 2950 2900

1.6

1.4

1.2

1

0.8

61

Turbidity (UTN) 1.8

145 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15

1.6

1.4

1.2

1

0.8

0.6

0.6

0.2

0.4

0.6

0.8

0.2

1

Figure 6: Electrical conductivity spatial distribution.

0.4

0.6

0.8

1

Figure 7: Turbidity spatial distribution.

2

Biomass (mg/kg) 1.8 34000 32000 30000

1.6

28000 26000 24000

1.4

22000 20000 18000

1.2

16000 14000 12000

1

10000 8000 6000

0.8

4000 2000

0.6

0.2

0.4

0.6

0.8

1

Figure 8: Biomass spatial distribution.

4 CONCLUSIONS Thalia geniculata vegetation was found to be quite efficient for wastewater treatment in a subsurface flow constructed wetland (HF-CW) with 85% of BOD removal. The daily average temperature (28°C) showed a k of 0.43 days-1, promoting the pollutant removal, similarly as a secondary water treatment. The support medium influenced significantly the microorganism’s fixation, meaning that the mixed gravel employed was appropriate for the treatment.



From the operation standpoint, the experimental reactor attained the highest removal efficiencies of basic contaminants when the HRT ranged from 4.2 to 6 days. The experimental design of the HF-CW proposed in this research complied with the environmental regulation for water standards in Mexico (75 mg/L). [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

[15] [16]

REFERENCES Reed, S.C., Crites, R.W. & Middlebrooks, E.J., Natural Systems for Waste Management and Treatment. 2nd ed., McGraw Hill Co.: New York, NY, 1995. García, J., Ruiz, A. & Junqueras, X., Depuración de aguas residuales urbanas mediante humedales construidos. Tecnología del Agua, 165, pp. 58–65, 1997. García, M., Bécares, E., Soto, F. & de Luis, E., Macrófitos en la depuración de aguas residuales, su función en la eliminación de bacterias. Tecnología del Agua, 185, pp. 64–67, 1999. Liang, W., Zhen-bin, W., Shui-Ping, C., Qiao-Hong, Z. & Hong-Ying, H., Roles of substrate microorganisms and urease activities in wastewater purification in a constructed wetland system. Ecological Engineering, 21, pp. 191–195, 2003. Pigem, J.O., Marzo, R., de la Peña, J.L. & Llagostera, R., Infiltración/Percolación y humedales como tratamientos blandos en la depuración de aguas residuales. Tecnología del Agua, 186, pp. 48–53, 1999. Soto, F., Bécares, E., García, M. & de Luis, E., Macrófitos en la depuración de aguas residuales. Su función en la eliminación de nutrientes. Tecnología del Agua, 185, pp. 68–72, 1999. Llagas, C.W.A. & Guadalupe, G.E., Diseño de humedales artificiales para el tratamiento de aguas residuales en la UNMSM. Revista del Instituto de Investigación FIGMMG, 15(17), pp. 85–96, 2006. Crites, R.W., Tchobanoglous. G., Small and Decentralized Wastewater Management Systems, McGraw Hill Co.: New York, NY, 2000. Ramos, E.M.G., Rodríguez, S.L.M. & Martínez, C.P., Uso de Macrófitos acuáticas en el tratamiento de aguas para el cultivo de maíz y sorgo. Hidrobiología, 17, pp. 7–15, 2007. Moshiri, G.A., Constructed Wetlands for Water Quality Improvement, Lewis Publishers: USA, p. 615, 1993. Brix, H., Functions of macrophytes in constructed wetlands. Wat. Sci. Tech., 29(4), pp. 45–53, 1994. Padrón-López, R.M., Depuración de aguas residuales domésticas a través de humedales artificiales de flujo vertical en zonas Trópico-Húmedas. Tesis Maestría en Ciencias Ambientales UJAT, 2005. Kivaisi, A.K., The Potential for constructed wetlands for wastewater treatment and reuse in developing countries: a review. Ecological Engineering, 16, pp. 545–560, 2000. Folch, M., Huertas, E. & Salgot, M., Zonas Húmedas Artificiales como tratamiento de aguas residuales en pequeños núcleos urbanos: El caso de ELS Hostalets de Pierola (Barcelona). Manual de agua potable para comunidades rurales, reuso y tratamiento avanzados de aguas residuales domésticas, capitulo, 17, pp. 199–205, 2000. USEPA. Office of Research and Development Cincinnati, Ohio, EPA/625/R-99/010, Design Manual: Constructed wetlands treatment of municipal wastewater. Sept. 2000. Jing, S.R. & Lin, Y.F., Seasonal effect on ammonia nitrogen removal by constructed wetlands treating polluted river water in southern Taiwan. Environmental Pollution, 127, pp. 291–301, 2004.


Water Pollution XIV

63

[17] Comisión Nacional del Agua (CONAGUA). Inventario Nacional de Plantas Municipales de Potabilización y de Tratamiento de Aguas Residuales en Operación. Subdirección General de Agua Potable, Drenaje y Saneamiento. Diciembre 2015. www.gob.mx/conagua. [18] López-Ocaña G. et al., Diseño de sistemas experimentales de humedales artificiales de flujo libre y subsuperficial. Perspectiva Científica desde la UJAT. ISBN: 978607-606172-5, pp. 133–146, 2014. [19] CONAGUA, Manual de plantas de tratamientods de aguas residuales. Lechos de Hidrófitas, capítulo 5. Subdirección General de Agua Potable, Drenaje y Saneamiento. Diciembre 2007. [20] Chung, A.K.C., Wu, Y., Tam, N.F.Y. & Wong, M.H., Nitrogen and phosphate mass balance in a sub-surface flow constructed wetland for treating municipal wastewater. Ecological Engineering, 32, pp. 81–89, 2008. http://dx.doi.org/10.1016/ j.ecoleng.2007.09.007. [21] McCabe, W.L., Smith, J.C. & Harriot, P., Operaciones Unitarias en Ingeniería Química. Mc Graw Hill, 1112. Cuarta Edición, 1991. [22] Surfer 8.0. Powerful Contouring, Gridding, and 3D Surface Mapping Software for Scientists and Engineers. Surfer® Software. [23] NOM-001-SEMARNAT-1996, Que establece los límites máximos permisibles de contaminantes en las descargas de aguas residuales en aguas y bienes nacionales. Secretaría de Medio Ambiente, Recursos Naturales y Pesca. Diario Oficial de la Federación. 23 de Abril de 2003. [24] Akratos, S.C. & Tsihrintzis, A.V., Effect of temperature, HRT, vegetation and porous media on removal efficiency of pilot-scale horizontal subsurface flow constructed wetlands. Ecological Engineering, 29, pp. 173–191, 2007. http://dx.doi.org/ 10.1016/j.ecoleng.2006.06.013. [25] Jiménez-López, E.C., Wastewater treatment by constructed wetlands with thalia geniculata and paspalum paniculatum in a tropical system of Mexico. Int. J. Sus. Dev. Plann., 12(1), pp. 42–50, 2017. DOI: 10.2495/SDP-V12-N1-42-50. [26] Kadlec, R.H. & Knight, R., Treatment Wetlands, Lewis Publishers: Boca Raton, FL, 1996. [27] Silva, J., Torres, P. & Madera, C., Reuso de aguas residuales domésticas en agricultura: Una revisión. Agro Col., 2, pp. 347–359, 2008. [28] Morari, F., Giardini, L., Municipal wastewater treatment with vertical flow constructed wetlands for irrigation reuse. Ecological Engineering, 35, pp. 643–653, 2009. http://dx.doi.org/10.1016/j.ecoleng.2008.10.014. [29] Karathanasis, A.D., Potter, C.L. & Coyne, M.S., Vegetation effects on fecal bacteria, BOD, and suspended solid removal in constructed wetlands treating domestic wastewater. Ecological Engineering, 20, pp. 157–169, 2003. http://dx.doi.org/ 10.1016/S0925-8574(03)00011-9 [30] Solano, M.L., Soriano, P. & Ciria, M.P., Constructed wetlands as a sustainable solution for wastewater treatment in small villages. Biosystem Engineering, 87(1), pp. 109– 118, 2004. [31] Abidi, S., Kallali, H., Jedidi, N., Bouzaiane, O. & Hassen, A., Comparative pilot study of the performances of two constructed wetland wastewater treatment hybrid systems. Desalination, 246, pp. 370–377, 2009. http://dx.doi.org/10.1016/j.desal.2008.03.061. [32] Vymazal, J., Emergent plants used in free water surface constructed wetlands: a review. Ecological Engineering, 61, pp. 582–592, 2013. http://dx.doi.org/10.1016/ j.ecoleng.2013.06.023.



Water Pollution XIV

65

GLYPHOSATE IN RUNOFF FROM URBAN, MIXED-USE AND AGRICULTURAL WATERSHEDS IN HAWAII, USA STEVEN R. SPENGLER, MARVIN D. HESKETT & SAMUEL C. SPENGLER Element Environmental, Hawaii, USA

ABSTRACT High-frequency sampling and analysis was conducted for the broad-spectrum herbicide glyphosate (Roundup) in ephemeral and perennial streams receiving storm water runoff generated within agricultural, urban and mixed-used watersheds on the island of Oahu, Hawaii. Glyphosate was selected for analysis since it is the most widely used herbicide in the world, and as a result, tends to be ubiquitous in the environment and our food supply. Samples were collected under both baseflow and storm conditions from five streams. The pervasiveness and maximum concentration levels of glyphosate detected in these streams are greater than any other pesticide currently present in Hawaiian streams. Glyphosate was detected in 96% and 65% of the stream samples collected during storm events (53 samples) and under baseflow conditions (34 samples), respectively (detection limit = 0.05 µg/L). The mean glyphosate concentrations measured in stream samples collected under storm conditions were between five to fifty times higher than mean glyphosate levels measured in the same stream under groundwater dominant baseflow conditions. The highest glyphosate concentrations were measured during a small runoff event in Manoa stream which flows through residential communities in urban Honolulu. The mass of dissolved phase glyphosate measured in stream water during the individual storm events monitored ranged from 0.5 to 18 grams. Between 11% and 23% of the total glyphosate load was present in suspended sediment during three sampled storm events in Honouliuli, Waimanalo and Kawa streams. The estimated total mass of dissolved phase glyphosate that discharged into Kaneohe Bay from Kawa Stream over a four-month monitoring period from December 2017 to March 2018 was 987 grams, with 92% of the pesticide load entering under storm conditions. Keywords: glyphosate, streams, suspended sediment, Hawaii, Oahu.

1 INTRODUCTION Pesticides have been applied to agricultural lands on the island of Oahu, Hawaii since the early 1900s. The annual amounts of herbicides historically applied on sugarcane and pineapple crops in Hawaii may have been as much as five times the amounts applied in major temperate region field crops elsewhere in the United States [1]. The intensive historical use of herbicides in Hawaii, year-round cultivation practices, and proximity of agricultural lands to streams enhance the possibility of pesticide transport to Oahu streams [2]. Some of the highest concentration levels of pesticides measured in fish during national water quality studies conducted by the United States Geological Survey (USGS) in the 1990s were detected in fish caught from urban Honolulu streams [3]. There is increasing public concern on the impact to human and ecological health from exposure to pesticides currently used in Hawaii and bills within the state legislature have recently moved ahead to regulate the use of glyphosate and chlorpyrifos. The USGS and Hawaii Department of Agriculture initiated a comprehensive pesticide-monitoring program of surface water in Hawaii in 2016 for 225 current-use pesticide compounds, but glyphosate is not among the compounds being monitored [4]. This study was conducted to provide data on the concentration levels of glyphosate in streams that flow through agricultural, residential and mixed-use areas on the island of Oahu under baseflow and storm conditions.



2 GLYPHOSATE Glyphosate is a non-selective herbicide that is widely used in agricultural, residential and commercial applications. Its most common mode of application is by direct spraying onto weeds. Its herbicidal activity is a result of glyphosate’s ability to block a metabolic pathway involved in the synthesis of three aromatic amino acids (tyrosine, tryptophan, phenylalanine) which are essential for a plant’s growth [5]. Glyphosate binds strongly to soils but is also very water soluble (more than 10,000 mg/l at 25°C). Glyphosate degrades in the environment, primarily by microbial processes, to aminomethylphosphonic acid (AMPA). The half-life of glyphosate in water and soil is highly variable and longer than previously recognized. In field studies, the half-life of glyphosate in soil ranged between a few days to up to two years, depending on soil composition [6], [7]. Factors that control the amount of pesticides that may be transported to streams in runoff include the amount and intensity of rainfall, antecedent rainfall or irrigation, topography, soil type and condition, crop type and practices, pesticide application rates and timing, and pesticide properties. Pesticides such as glyphosate that adhere to soil particles may become liberated by rain events and either dissolve into or flow with surface water as suspended sediment, eventually ending up in stream bed sediments or coastal waters. Glyphosate was first registered for use in the United States in 1974 and is now the world’s most heavily applied herbicide. The use of glyphosate increased dramatically in the early 2000s with the introduction of genetically-modified crops that are resistant to this herbicide. Annual farm-sector glyphosate usage in the United States increased to approximately 110 million kilograms by 2014, based on average annual usage data reported by the United States Department of Agriculture [8]. In the agricultural sector, glyphosate is typically applied “post-emergence” after crops and weeds have emerged from the soil. Due to this mode of application, glyphosate is mainly stored in topsoil layers in agricultural soils. Hawaii’s yearround growing season makes it an ideal location to develop new varieties of corn, soy and other commodity crops, since it often takes a dozen or more generations to develop a new hybrid plant that is ready for commercial distribution. As a result, pesticides including glyphosate are used throughout the year in Hawaii with the main growing season for seed crops extending from November to June. Glyphosate is also commonly used by homeowners and for other nonagricultural purposes, the annual use of which quadrupled between 1993 and 2011 [9]. In urban settings glyphosate is widely approved for use by public works departments for weed control along streets, and within parks and public spaces. These urban applications are frequently on or near impervious surfaces and can result in substantial pesticide inputs to urban drainage systems [10]. For instance, Connor et al. [8] found that runoff from small-urbanized tributaries contribute as much or more to the pesticide loads to San Francisco Bay than runoff from the agricultural Central Valley, even though 90% of the freshwater flow to the bay comes from the Central Valley via the Sacramento and San Joaquin rivers. 3 GLYPHOSATE LEVELS IN THE ENVIRONMENT The USGS has conducted the most comprehensive assessment of the environmental occurrence of glyphosate and its degradation product AMPA in the continental United States to date [12]. The USGS collected and analyzed a total of 3,732 water and sediment and 1,018 quality assurance samples from 38 states between 2001 and 2010 (no samples were collected from the State of Hawaii). This comprehensive study commonly detected either glyphosate or AMPA in surface water (59% detection frequency from 470 sampled sites) and infrequently in groundwater or soil water samples (8.4% detection frequency from 820 sampled sites). The European Glyphosate Environmental Information Sources (EGEIS)


Water Pollution XIV

67

summarized surface water (50,000 samples) and groundwater (36,000 samples) monitoring data collected from thirteen European countries between 1993 and 2009 [13]. Glyphosate and AMPA were detected in 29% and 50% of the surface water samples analyzed, respectively. Glyphosate was detected in 1.3% of the groundwater samples collected from 8,900 localities throughout Europe. Glyphosate was also detected in 60% and 100% of air and rain samples collected from agricultural areas in Mississippi and Iowa [14]. Measured glyphosate concentrations ranged from 4.0 > 4.0 0% 0.94 0.86 8.9% 0.75 0.97 25.6% 0.90 1.02 12.5% < 0.05 < 0.05 0% < 0.05 < 0.05 0% 0.12 0.16 28.6% < 0.05 < 0.05 0% < 0.05 < 0.05 0%

Shook sample glyphosate concentrations (µg/L) Primary sample result 0.65 1.84 < 0.05 0.16 0.09 < 0.05 1.20

Re-suspended sample result

Primary sample turbidity (NTU)

0.54 1.59 < 0.05 0.14 0.06 < 0.05 1.80

76 389 25 388 2,000 160 121



The replicate samples were vigorously shaken after the decanted primary sample had been pipetted and a duplicate analysis was conducted on the resulting turbid sample. Table 3 shows that four of five of these shook, re-suspended samples yielded slightly lower glyphosate concentrations than the primary sample. This result suggests that the ELISA method does not detect glyphosate that may be sorbed to the entrained sediment present in the analyzed samples. 6 FINDINGS Multiple stream samples were collected from fives streams during discrete rainfall events from February to April 2018. A single set of stream runoff samples were collected from Manoa, Waimanalo, Kahawai and Honouliuli streams while three discrete runoff events of varying size were sampled at Kawa Stream. The storm event hydrographs (in cubic meters per second (cms)) and associated glyphosate concentrations (µg/L) measured in Manoa, Honouliuli, Kahawai, and Waimanalo streams are depicted in Fig. 2. The Manoa Stream monitoring location receives rainfall runoff from the residential communities located in both Manoa and Palolo Valleys. Due to urbanization, the surface water flow to Manoa Stream is significantly altered from the “natural” state by the presence of impermeable surfaces, buildings, and on-going construction projects. The sampled storm for Manoa Stream on 3/31/2018 approached Oahu from the west just after noon. Due to the storm direction, the majority of the runoff that entered the stream during this rainfall event originated from rainfall runoff that fell on the residential communities located within Manoa and Palolo valleys. This likely contributed to the elevated glyphosate concentration levels detected in these samples. By comparison, the HDOH [15] and Surfrider [16] studies measured lower glyphosate concentrations in Manoa Stream ranging from Kahawai > Waimanalo > Kawa > Manoa. Despite this estimate, Manoa Stream contained the highest concentration levels of glyphosate. The high concentrations measured in Manoa Stream are likely an artifact of the nature of the particular storm event sampled. The majority of the runoff generated during this “Kona” storm event came from the residential portions of Manoa and Palolo Valleys and were not “diluted” by runoff from the surrounding forested mountain watersheds which likely contain little or no glyphosate. The paved surfaces in the residential communities within these two valleys also likely more efficiently transport glyphosate containing top soils from residential yards and street medians to the stream during moderate intensity rainfall events such as the one sampled. Fig. 4 plots the glyphosate concentrations versus the turbidity measured in the stream samples analyzed during this study. Fig. 4 shows the generally lower glyphosate concentrations measured in low turbidity samples (< 40 NTU) typically associated with baseflow conditions and the generally higher glyphosate concentrations measured in more turbid samples (> 40 NTU) typically associated with storm conditions. This finding supports the hypothesis that the source of some of the glyphosate measured in the stream samples during the storm events sampled originated from



Figure 4: Relationship between glyphosate concentration and stream turbidity. release of sorbed glyphosate present in contaminated sediments as they become resuspended from the stream bed during moderately intense storm runoff events. The mass of glyphosate present in stream water during the monitored runoff events was estimated as the product of the glyphosate concentrations measured in the stream (micrograms per liter) and the stream flow rates measured at the associated USGS gaging station. The gaging stations typically recorded stream flow at either a five or fifteen-minute time intervals. The glyphosate concentration in the stream was estimated for the same five or fifteen-minute time intervals based upon extrapolation of the individual glyphosate concentrations measured in the stream throughout the duration of each monitored storm event (Figs 2 and 3). Table 4 summarizes the duration of the monitored storm events, the storm event rainfall totals, the total volume of streamflow during the monitored event, and the calculated mass of glyphosate present in stream water during the monitored event. The concentration of glyphosate and AMPA in suspended samples collected from Waimanalo, Honouliuli and Kawa Streams were determined by Pacific Agricultural Laboratory in Sherwood, Oregon. These composite suspended sediment samples were collected by repeatedly filling up three five-gallon buckets with stream water collected over a thirty-minute time period during the rainfall runoff event. The Waimanalo Stream suspended sediment sample was collected between 17:15 and 17:45 on 3/22/18; the Table 4: Runoff event duration, rainfall, runoff volume and glyphosate mass. Stream

Sampled runoff event duration

Manoa 3/31/18: 11:45 to 14:30 Waimanalo 3/22/18: 15:45 to 18:30 Kahawai 3/22/18: 15:45 to 18:30 Honouliuli 3/24/18: 2:55 to 13:30 Kawa 2/5/18: 6:40 to 17:00 Kawa 3/13/18: 5:00 to 8:30 Kawa 4/2/18: 7:00 to 10:15

Rainfall event total (mm) 6.4 5.1 5.1 51.3 29.2 6.4 12.7

Stream volume during sampling (L) 5,333,094 3,427,271 1,131,000 6,063,241 125,987,860 1,152,020 4,330,925


Glyphosate mass during event (gm) 14.4 11.9 3.4 5.4 18.0 0.5 4.8

Water Pollution XIV

75

Honouliuli Stream suspended sediment sample was collected between 12:15 and 12:45 on 3/22/18; and the Kawa Stream suspended sediment sample was collected between 9:00 and 9:30 on 4/2/18. The total mass of glyphosate present in the suspended sediment during this thirty-minute period of time was estimated based upon the turbidity levels measured in the stream samples collected during this portion of the storm event. The turbidity levels were converted to a total suspended solids concentration using relationships between turbidity and TSS developed during this study, historic USGS water quality measurements, and from measurements made during previous environmental studies [21]. The glyphosate and AMPA concentrations measured in the suspended sediment samples collected in Honouliuli, Waimanalo and Kawa streams were 0.25/0.39, 0.75/0.63 and 0.82/1.0 mg/kg, respectively. The corresponding mass of glyphosate in suspended sediments present in these three streams during the thirty-minute time interval sampled were 0.05, 0.17 and 0.15 grams, respectively. By comparison, the mass of glyphosate dissolved in the stream water during this same thirtyminute time interval of the runoff event was 0.22, 1.55 and 0.82 grams, respectively. Based on this limited suspended sediment data, the glyphosate in suspended sediment represents between 11%–23% of the total glyphosate load in the stream. Because of the duration of monitoring (four months) and number of sampled storm events (3) at Kawa Stream, the mean and median glyphosate concentrations measured under baseflow and storm conditions from Kawa Stream are the most representative data for evaluating the magnitude of glyphosate transport to the offshore environment via streams in Hawaii. The total mass of dissolved phase glyphosate transported over this four-month period in Kawa Stream was estimated by multiplying the stream flow measured under baseflow conditions (stream flow < 0.10 cms) during this period by the mean baseflow glyphosate concentration measured over this four-month period (0.14 µg/L) and the storm related stream flow (> 0.10 cms) by the storm event mean glyphosate concentration (0.67 µg/L). The resultant total dissolved phase glyphosate mass input into Kaneohe Bay (the adjacent coastal body) is 987 grams for this four-month period, with about 92% of the load entering the bay during storm runoff events. 8 CONCLUSIONS High-frequency sampling for glyphosate in five streams on the island of Oahu revealed the ubiquitous nature of this herbicide in both agricultural and urban settings as reflected by the high rates of detection measured during both storm and baseflow conditions (96% and 65%, respectively). Measured stream glyphosate concentrations were generally significantly higher under storm conditions than under groundwater dominant baseflow conditions. The source of some (or perhaps a majority) of the glyphosate measured in the stream samples is likely from release of sorbed glyphosate present in sediments as they become re-suspended from the stream bed during storm runoff events. During the limited sampling conducted in this study, glyphosate concentrations measured in streams receiving runoff from an urbanized watershed (Manoa) was higher than concentrations measured in streams that flow through agricultural areas on Oahu. The mass of glyphosate measured discharging to the coastal environment during the monitored storm events ranged from 0.5 to 18 grams while the total mass of glyphosate discharging from one of the monitored streams (Kawa) over a four-month period was estimated to be around 1 kilogram. [1]

REFERENCES Green, R.E., Goswami, K.P., Mukhtar, M. & Young, H.Y., Herbicides from cropped watersheds in stream and estuarine sediments in Hawaii. Journal of Environmental Quality, 6(2), pp. 145–154, 1977.



[2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

[12] [13] [14]

[15] [16] [17]

Oki, D.S. & Brasher, A.M.D., environmental setting and the effects of natural and human-related factors on water quality and aquatic biota, Oahu, Hawaii. United States Geological Survey, Water-Resources Investigations Report 03-4156, 2003. Brasher, A.M. & Anthony, S.S., Occurrence of organochlorine pesticides in stream bed sediment and fish from selected streams on the island of Oahu, Hawaii, 1998: U.S. Geological Survey Fact Sheet 140-00, p. 6, 2000. Steinrücken, H.C. & Amrhein, N., The herbicide glyphosate is a potent inhibitor of 5enolpyruvyl-shikimic acid-3-phosphate synthase. Biochemical and Biophysical Research Communications, 94(4), pp. 1207–1212, 1980. Benbrook, C., Trends in the use of glyphosate herbicide in the U.S. and globally. Environmental Sciences Europe, 28(3), 2015. Grube, A.D.D., Kiely, T., & Wu, L., Pesticides Industry Sales and Usage, 2006 and 2007 Market Estimates, United States Environmental Protection Agency, EPA 733-R11-001, 34 p. 2011. http://1.usa.gov/1VN2QDo, 2011. Wittmer, I.K., Scheidegger, R., Bader, H., Singer, H. & Stamm, C., Loss rates of urban biocides can exceed those of agricultural pesticides. Science of the Total Environment 409, pp. 920–932, 2011. Connor, M.S. et al., The slow recovery of San Francisco Bay from the legacy of organochlorine pesticides. Environ. Res., 105, pp. 87–100. 2007. Szekacs, A. & Darvas, B., Forty years with glyphosate. Herbicides – Properties, Synthesis and Control of Weeds, ed. N. Hasaneen, M. InTech, 2012. doi: 10.5772/32491. Bento, C.P.M. et al., Persistence of glyphosate and aminomethylphosphonic acid in loess soil under different combinations of temperature, soil moisture and light/darkness. Sci. Total Environ., 572, pp. 301–311, 2016. Battaglin, W.A., Meyer, M.T., Kuivila, K.M. & Dietze, J.E., Glyphosate and Its Degradation Product AMPA Occur Frequently and Widely in U.S. Soils, Surface Water, groundwater, and precipitation. Journal of the American Water Resources Association (JAWRA), 50(2), pp. 275–290, 2014. doi: 10.1111/jawr.12159. Horth H., European Glyphosate Environmental Information Source (EGEIS), Monitoring results for surface and groundwater. www.egeis.org/documents/11%20 Detection%20in%20SW%20and%20GW%20draft%20v3.pdf, 2010. Chang, F, Simcik, M.F. & Capel, P.D., Occurrence and fate of the herbicide glyphosate and its degradant aminomethylphosphonic acid in the atmosphere. Environ. Toxicol. Chem., 30, pp. 548–555, 2011. Hawaii Department of Health (HDOH) Hazard Evaluation and Emergency Response Office. 2013–2014 State Wide Pesticide Sampling Pilot Project Water Quality Findings. A joint investigation by the Hawaii State Departments of Health and Agriculture, May 2014. Surfrider Foundation, Water Quality Study of Nearshore Environments, Oahu, Kauai and Molokai, Hawaii. Final Project Report, Version 1.1, Aug. 2017. Rosa, S.N., Measuring surface-water loss in Honouliuli Stream near the ‘Ewa Shaft, Oahu, Hawai‘i: U.S. Geological Survey Scientific Investigations Report 2017–5042, p. 14, 2017. https://doi.org/10.3133/sir20175042. Johnson, A.G. & Kennedy, J.J., Summary of dissolved pesticide concentrations in discrete surface-water samples collected on the islands of Kauai and Oahu, Hawaii, Nov. 2016–Apr. 2017, U.S. Geological Survey data release, 2018. https://doi.org/ 10.5066/F7BG2N79.


Water Pollution XIV

77

[18] Yang, X.M. et al., Short-term transport of glyphosate with erosion in Chinese loess soil – a flume experiment. Sci. Total Environ., 512, pp. 406–414, 2015. [19] Todorovic-Rampazzo, G., Rampazzo, N., Mentler, A., Blum, W.E.H., Eder, A. & Strauss, P., Influence of soil tillage and erosion on the dispersion of glyphosate and aminomethylphosphonic acid in agricultural soils. Int. Agrophysics, 28, pp. 93–100. 2014. [20] Coupe, R.H., Kalkhoff, S.J., Capel, P.D. & Gregoire, C., Fate and transport of glyphosate and aminomethylphosphonic acid in surface waters of agricultural basins. Pest Management Science, 68, pp. 16–30. 2011. [21] Tomlinson, M. & DeCarlo, E., Final Report – Investigations of Waimanalo and Kaneohe Streams, p. 30, 2001.



Water Pollution XIV

79

PATH MODELLING ANALYSIS OF POLLUTION SOURCES AND ENVIRONMENTAL CONSEQUENCES IN RIVER BASINS ANTÓNIO FERNANDES1, ANA FERREIRA1, LUÍS SANCHES FERNANDES1, RUI CORTES1 & FERNANDO PACHECO2 1 Centre for the Research and Technology of Agro-Environment and Biological Sciences (CITAB), Universidade de Trás-os-Montes e Alto Douro, Portugal 2 Vila Real Chemistry Research Centre (CQVR), Universidade de Trás-os-Montes e Alto Douro, Portugal

ABSTRACT The Portuguese Index of Macroinvertebrates is used as a pollution index in the study of surface water in Portugal. From an environmental perspective, it is necessary to determine the pollution sources affecting surface water. Not only direct discharges of industrial and urban effluents are the cause; possibly the combination of other pollutions sources could result in ecological loss. To comprehend all the cause and effect relationships between pollution sources, water contamination and ecological integrity, it is necessary to apply complex models and possibly apply the use of thorough statistical tools. Structural Equation Modelling (SEM) has been used in the social sciences for a long time. Due to the present environmental concern and awareness of the phenomenon’s complexity, SEM was used for environmental studies. In this paper are present SEM-PLS models that are applied to collected data from two different river basins, that of the Ave and the Sabor, in order to understand which are their main pollution sources and which contaminants are restraining biodiversity. The applied models from each basin reproduced different realities, as expected, since the river Ave has has the notorious impact of industry, while the Sabor basin has a higher level of water quality. Keywords: macroinvertebrates, modelling, pollution modelling, Portuguese Index of Macroinvertebrates, river pollution, Structural Equation Modelling.

1 INTRODUCTION To study water quality of river basins, it is crucial to understand that there is a multitude of phenomena that affect biodiversity, such as several pollution sources, different contaminants, several reactions, morphological characteristics and even climate data. To comprehend the relationships between these variables, it is necessary to use advanced statistical tools such as Structural Equation Modelling (SEM). Generically, there are two types of SEM [1]: CB-SEM (covariance-based SEM), and PLS-SEM (Partial Least Squares SEM) which is also called PLS-PM (PLS path modelling). In the first case, the estimation procedure is based on a maximum likelihood estimation, while PLS-SEM is based on ordinary least squares regression [2]. The benefits of each type of SEM differ from case study to case study, so the opinion on which type is the most appropriate method is still divergent [1], [3]. Initially, SEM was used in the social sciences; nowadays it has been increasingly used in non-social sciences, such as environmental and biological issues. In an environmental study, PLS-PM was used to understand ground-level ozone concentrations, considering meteorology, chemical reactions and the presence of primary pollutants as main causes [4]. Already in 1994 SEM was applied to study surface water quality: natural and anthropogenic effects were declared as impossible to quantify [5]. SEM was used to understand the relationship between several concentrations of substances with the total dissolved solids in ground water [6]. The comparison of 3 types of pollution (organic, sediment and eutrophication in surface waters) was applied to Feitsui Reservoir Watershed [7], with the conclusion that sediment pollution was the main cause. From a social standpoint, SEM was



applied to comprehend community awareness about local water quality, through the usage of surveys [8]. For the Surabaya river [9], the citizens’ awareness was accounted for in a model that included variables such as oxygen demands and total solids, to understand their relationship with water pollution levels using two water quality indexes, and concluding that the application of more variables would strengthen the model. The comparison of all studies about water quality using SEM led to the understanding that there are a multitude of variables to apply. The purpose of this study is to understand the relationship between pollution sources (named/given as pressures), the concentration of contaminants in surface waters and ecological integrity, as measured by the Portuguese Index of Macroinvertebrates [10]. SEMPLS was applied in formative models, establishing a cause and effect model for two distinct river basins, the Ave and the Sabor, to act as industrialized and rural river basins, respectively. 2 METHODOLOGY To create the structural equation models, a seven-step methodology was followed (Fig. 1). For Step 1, the river basins and sub-basins were delineated using ArcHydro [11]. The collected data for models was divided into three groups: water contamination, pressures and ecological integrity, according to the conceptual model (Fig. 2). It is known that the biodiversity in surface waters is dependent on the presence of contaminants, while the pollutions sources are the pressures in surface waters which can cause contamination. In any case, it was established that a connection between pressures and ecological integrity was established, although the effect is indirect. For Ecological Integrity (Fig. 2), we used only one variable, IPtIN, which is a numerical indicator that represents the biodiversity of macroinvertebrates in the north zone of Portugal: high values are indicative of water quality [10]. For water contamination, we chose 11 concentrations of contaminants, As, Cr, Cu, Fe, Pb, Zn, the oxygen demands BOD5 and COD, the concentration of nutrients NO3 and PO4, and also the total suspended solids (TSS). For the pollution sources in pressures (Fig. 2), we collected a vast number of variables such as the industrial and urban discharges in surface waters and underground waters of phosphorous, nitrogen, the oxygen demands (CBO5 and COD), the percentage each sub-

Figure 1: Seven-step methodology.

Figure 2: Conceptual model.


Water Pollution XIV

81

basin is covered with artificial and agricultural land uses, the percentage of land covered with land use conflicts [12], the population density, soil loss, wildfire risk, nitrogen and phosphorous discharges from agriculture and forest, and from livestock production. After collecting all the data and assigning a value for each sub-basin, a data matrix was built for each sub basin (step 4 of Fig. 1), where the columns are variables while each row is a subbasin. Variables were discarded from the study in order to have a Pearson correlation coefficient between each variable of the same group below 0.8, and a variance inflation factor (VIF) below 5, because for values higher than this one, it means that a model with this data has multicollinearity [13]. In SEM-PLS there are two types of variables used: measured and latent. Latent variables are defined by the operator, while measured variables are the ones collected for the model. For this study, we chose to use 3 latent variables, according to the conceptual model: pressures, contamination and ecological integrity. We built a total of four models, two for each basin. In the first models (Ave 1 and Sabor 1) the latent variable pressures are connected to the ecological integrity, while in the second models (Ave 2 and Sabor 2) there is no direct connection between these two latent variables. SmartPLS [14] was used to build the models, which can be be either reflective or formative. In the first case, the measured variables were composed by the latent variables; while in the formative case, the latent variables are composed by the measured values. We chose to use formative models, due to the nature of the data. The aspect of a SEM is presented in Fig. 3. Each yellow square represents a measured variable (MV), while blue circles are latent variables (LV). The algorithm attributes values for weights (w) for MVs and path coefficients for LVs, through and iterative process, in order to achieve the highest determination coefficient (R2) for each latent variable. The model calculates a measured score for latent variables, based on the measured variables that compose it: eqn (1). For latent variables that are composed by other ones, such as B and C (Fig. 3), the model calculates a predicted score, as is demonstrated in eqn (2), for LVC from Fig. 3. Measured score: LVi = ∑ni=1 (MVi ×Wi )

(1)

Predicted score: LVc =LVa ×pcac +LVb ×pcbc

(2)

The determination coefficient is calculated for latent variables that have and measured and predicted scores. Because the present study is a comparison between models with a

Figure 3: Structural equation model (SEM). WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)


different number of data points in each sub-basin and even a different number of measured variables, we chose to use an adjusted R-Squared [3], instead of R-Squared, eqn (3). Radj 2 =1–(1–R2 )×

(n–1) (n–k–1)

(3)

For this adjusted R, n is the total sample size, while k is the number of predictors. Normally, as more predictors are used in a model, the tendency of the R-squared value is to increase, so the adjustment of R-squared is useful for the kind of models that may include a high number of predictors. 3 RESULTS AND DISCUSSION The occupied area of the Ave river basin is 1321 km2 while for the Sabor, it is 3513 km2. Since the Sabor is an international river, we only considered the zone within the Portuguese territory, thus reducing the study area to 2969 km2. The number of delineated sub-basins was 92 for the Ave and 100 for the Sabor. Fig. 4 presents both river basins, their sub-basins and interpolated values of IPtIN. As shown in Fig. 4, the IPtIN in the Ave river basin is much lower than in the Sabor basin. Biodiversity lowers from upstream to downstream, not only due to an accumulation effect, but due to the fact that important pressures such as industrial discharges and higher population density are closer to the basin outlet. For the Sabor basin, the locations with low biodiversity are the sub-basins 11, 12, 13 and 14 (Fig. 4). Around these sub-basins, there is urban discharge that is possibly the cause. Since all the variables are normalized (thus, mean of each one becomes equal to 0 and standard deviation is equal to 1), before being applied to the model, it is possible to do a comparison of path coefficients and weights. For each model, measured variables where chosen in order to eliminate multicollinearity issues, such as with inflated weights. We built a total of 4 models (Fig. 5), 2 for the river Ave (models A1 and A2) and two for the river Sabor (models S1 and S2). Initially, only models A1 and S1 were built, according to the conceptual model, and Pressures (P) were connected to Contamination (C) and Ecological Integrity (EI), with C connected to EI. We noticed that for A1 the results were reliable, had

Figure 4: Map of Portugal (A) with the Ave (B) and Sabor (C) river basins, IPtIN values.


Water Pollution XIV

83

Figure 5: Structural equation models. a highly adjusted R-squared, 0,861 for C and 0,921 for EI; and the path coefficients according to theory, from Pressures to Contamination the path coefficient is positive (0.929), which means that pressures are the cause of the high concentrations of contaminants. Negative path coefficients from P to EI (0.316), and from C to EI (0.660), mean that identified pressures and contaminants are the ones that decrease EI, even more by the effect WIT Transactions on Ecology and the Environment, Vol 228, © 2018 WIT Press www.witpress.com, ISSN 1743-3541 (on-line)


of C in EI is the higher P in EI. On the other hand, model S1 achieved lower values of adjusted R-squared: 0.716 for C and 0.618 for EI. Moreover, the path coefficients from P to C, and C to EI are unexpected, since it was foreseen that they would have the same signal as Model A1. Possibly the contaminants that lower ecological integrity in the Sabor river basin are not included in this study and that is why the path coefficients connected to contamination are unexpected; however, the chosen pressures have impact in ecological integrity, because the path coefficient between P and EI is negative. Besides, model A1 is representative of the reality, the path coefficients might be inflated, the latent variables (C and P) that compose EI appear to be collinear, as their VIF is higher than 5 (Table 1), while for model S1 the VIF values are below 5. In order to create a more reliable model, we removed the connection between P to EI, so the A1 model would have no multicollinearity, creating model A2. The same was done for model S1, in order to understand if the sign of the path coefficients would change, creating model S2. From Model A1 to Model A2, and from Model S1 to Model S2, we noticed a small change for the adjusted R-Squared value; for the latent variable C it increases, while for EI decreases. C adjusted R-Square increases due to the fact that for latent variable P the weights that compose it are calculated in order to increase the determination of the coefficient for C (in Model 2) while in the models A1 and S1, the weights of MVs that compose P were calculated in order to increase both determination coefficients (for C and EI); that is why in the models, the adjusted R-squared lowered from Models 1 to 2. In Model A2, the path coefficients remain according to theory, in the path coefficient between P and C there are no significant changes, but the path coefficient between C and EI increases in the module, due to the fact that it absorbs the effect of P in EI. For Model S2, the path coefficients are totally discordant, because the pressures can’t decrease C, and C cannot increase EI. For the river Sabor, the best model is certainly Model S1, because for this river basin, there is at least one reliable path coefficient from P to EI (Fig. 5). We analysed the product of total effect and the weight of each variable (Table 2). For example, in Model A1 the direct effect of P in EI is –0.316, but since P is connected to C, there is and indirect effect of P in EI that is equal to –0.613 (–0.660×0.929), so the total effect of P in EI is equal to the sum of direct and indirect effect, which results in –0.928. As an example, the product of total effect by Population Density is –0.160. Besides, if the weight of some measured variables is positive, it does not mean that in reality they increase biological diversity, but that the model attributes positive values for these variables, since in order to maximize the R-Squared value some of the weights must have a positive sign, or because they have a positive correlation and do not explain the variation. Besides the unreliable path coefficients in models S1 and S2, it is still possible to identify which are the measured variables that have a negative effect in ecological integrity. Through Table 2 it is possible to identify variables that lower EI (for Models S1 and S2) are Agriculture and Forest P, Urban P, Urban P (s), Wildfire Risk and PO4. Since Conflicts have a positive effect in S1 and a negative effect in S2, it is uncertain if the real effect of this variable is positive or negative. For the Ave river basin, the variables that do not lower EI are Cr, Agriculture and Forest P, Conflicts and Industry COD. For both river basins, the models show that effluent discharges of nutrients are a cause to lower EI, but not oxygen demands. In water treatment stations, one of the major concerns is to lower oxygen demands: in order to do this some biological treatments require the addition of nutrients, which results in low values of oxygen demand, but a high concentration of nutrients. For the Ave river basin, all contaminants have a negative impact in EI; only Cr does lower EI, probably due to its low concentration in surface waters, while PO4 is the only contaminant that restrains biodiversity for the Sabor river basin.


Water Pollution XIV

85

Table 1: Inner model VIF for model Ave 1 and Sabor 1. Model Ave 2

Model Ave 1 C

EI

C P

1.000

C

7.263

C

7.263

P

3.556 1.000

Model Sabor 1 C C

3.556

Model Sabor 2

EI

C

1.000

P

EI

1.000

EI

C P

1.000 1.000

Table 2: Product total effect and the weight for each measured variable. A1

A2

S1

S2

COD

-

-

As

+

+

Cr

+

+

BOD5

+

+

Fe

-

-

Cu

+

+

NO3

-

-

Fe

+

+

PO4

-

-

NO3

+

+

Agricultural areas

-

-

PO4

-

-

Agriculture and forest P

+

+

TSS

+

+

Conflicts

+

+

Agricultural areas

+

+

Industry COD

+

+

Agriculture and forest P

-

-

Industry N

-

-

Artificial surfaces

+

+

Livestock N

-

-

Conflicts

+

-

Population density

-

-

Livestock N

+

+

Urban N

-

-

Soil loss

+

+

Wildfire risk

-

-

Urban COD (s)

+

+

Urban P

-

-

Urban P (s)

-

-

Wildfire risk

-

-

Through the analysis of the models, it is possible to establish an order of pressures which require environmental intervention. By decreasing the order of severity, the pressures that require intervention in the Ave River basin are industrial discharges of nutrients, livestock production discharges in underground water and urban discharges. Besides wildfire risk and agricultural areas having a negative impact, they should not be considered as dangerous, since



the weight of these variables in the models A1 and A2 are considerably low. For the Sabor River Basin, the pressures of most concern are urban discharges, phosphorous in soil, forest and agriculture-based flows of nutrients, and urban discharges into surface waters. 4 CONCLUSIONS The application of SEM-PLS models to the Sabor River and the Ave River datasets, which describe multiple pressures, surface water quality and biodiversity loss, proved to be efficient in discriminating factors that explain biodiversity loss. The chosen dataset for the Ave basin proved to be descriptive of the reality, while for the Sabor river basin, it appears that other phenomenona and variables should be considered in order to improve the models’ reliability, such as morphological data. Possibly the fact that the Ave river basin is polluted resulted in an explanatory and reliable model; while for a clean basin such as the Sabor, it is harder to explain biodiversity loss. For the application of these models in other river basins, we advise to always collect as much data as possible and then proceed to the analysis of the correlation matrix, as some variables can explain each other. For further studies, we recommend the use of the transformation of variables and even the use of reflective models, in order to compare their results. ACKNOWLEDGEMENTS This research was funded by the INTERACT project: “Integrated Research Environment, Agro-Chain and Technology”, no. NORTE-01-0145-FEDER-000017, in its line of research entitled BEST, which was co-financed by the European Regional Development Fund (ERDF) through NORTE 2020 (North Regional Operational Program 2014/2020). For the authors integrated in the CITAB research centre, the research was further financed by the FEDER/COMPETE/POCI Operational Competitiveness and Internationalization Programme, under Project POCI-01-0145-FEDER-006958, and by National Funds from FCT, the Portuguese Foundation for Science and Technology, under the project number UID/AGR/04033/2013. For the author who is integrated into the CQVR, the research was additionally supported by National Funds from the FCT, under the project number UID/QUI/00616/2013. [1] [2] [3] [4] [5] [6]

REFERENCES Binz, A.C., Patelb, V. & Wanzenried, G., A comparative study of CB-SEM and PLSSEM for theory development in family firm research. J. Fam. Bus. Strat., 5(1), pp. 116–128, 2014. Hair, J.F., Hult, G.T.M., Ringle, C. & Sarstedt, M., A Primer on Partial Least Squares Structural Equation Modeling, Sage Publications Inc., 2014. https://doi.org/10.1016/ j.lrp.2013.01.002. Garson, G.D., Partial Least Squares: Regression and Structural Equation Models, 2016. www.statisticalassociates.com/pls-sem.htm. Kumar, G., Tuluri, A.F. & Tchounwou, P.B., Development of PLS: path model for understanding the role of precursors on ground level ozone concentration in Gulfport, Mississippi, USA. Atmos. Pollut. Res., 6(3), pp. 389–397, 2015. Zou, S. & Yu, Y.-S., A general structural equation model for river water quality data. J. Hydrol., 162(1), pp. 197–209, 1994. Chenini, I. & Khemiri, S., Evaluation of ground water quality using multiple linear regression and structural equation modeling. Int. J. Environ. Sci. Technol., 6(3), pp. 509–519, 2009


Water Pollution XIV

[7] [8] [9]

[10]

[11] [12] [13] [14]

87

Wu, E.M.-Y., Tsai, C.C., Cheng, J.F., Kuo, S.L. & Lu W.T., The application of water quality monitoring data in a reservoir watershed using AMOS Confirmatory Factor Analyses, 2014. Levêque, J.G. & Burns, R.C., A structural equation modeling approach to water quality perceptions. J. Environ. Manag., 197, pp. 440–447, 2017. Nugroho, A., Structural Equation Modelling as an instrument for water pollution factor analysis: Study case on the Surabaya River. The Fourth International Conference on Advances in Applied Science and Environmental Technology (ASET), Institute of Research Engineers and Doctors: USA, pp. 28–32, 2016. INAG, Critérios para a classificação do estado dasmassas de água superficiais - rios e albufeiras. Technical Report. [Criteria for classifying the condition of the volumes of surface waters in rivers] Ministério do Ambiente, do Ordenamento do Território e do Desenvolvimento Regional [Ministry of the environment, care of the territory and regional development], Instituto da Água [Water Institute], IP in Portuguese, 2009. www.apambiente.pt. ESRI, ArcHydro Tools for ArcGIS 10 – Tutorial, 2012. Pacheco, F.A.L., Varandas, S.G.P., Sanches Fernandes, L.F. & Valle Junior, R.F., Soil losses in rural watersheds with environmental land use conflicts. Sci. Total Environ, pp. 110–120, pp. 485–486, 2014. Monecke, A. & Leisch, F., SemPLS: Structural Equation Modeling using partial least squares. J. Stat. Softw., 48(3), pp. 1–32, 2012. Ring, C.M., Wende, S. & Will, A., Smart PLS, 2005. www.smartpls.de.



Water Pollution XIV

89

RE-EVALUATING HYDROCHEMICAL DATA FROM AQUIFERS OCCURRING IN THE RIO CLARO CITY REGION, SÃO PAULO STATE, BRAZIL RAQUEL CURTOLO QUIRINO & DANIEL MARCUS BONOTTO Departamento de Petrologia e Metalogenia, Instituto de Geociências e Ciências Exatas – UNESP, Rio Claro, Brazil

ABSTRACT Groundwater use increases every year in different sectors of the society, therefore, hydrochemical studies of aquifers and their relations with the anthropic environment are crucial. With an area of approximately 1,880 km², comprising Rio Claro city and near districts, the study assembles a series of groundwater data from previous assessments and aims using available tools to reach a new approach regarding the data in the literature. The database comprised a total of 41 groundwater samples from regional wells exploiting the aquifer systems Rio Claro, Serra Geral, Guarani, and Tubarão, as well the Passa Dois aquiclude, all them situated in the Paraná basin geological context. Locally, the study area is associated with a sedimentary deposit of sandstones, shales, limestones and unconsolidated sediments. The data of the chemical analyses of major ions and main physical parameters of the water samples were re-evaluated through The Geochemist’s Workbench 11.0.5 software and compared with three water potability standards: São Paulo State Environmental Agency, Brazil Health Ministry, and World Health Organization. Moreover, a series of hydrochemical diagrams and graphs (Piper, Schoeller, Stiff and Durov) were generated in order to classify the samples and evaluate their local behavior in comparison with the data of other aquifers in São Paulo State. Keywords: aquifers, hydrochemical, water standards.

1 INTRODUCTION This study was held at Rio Claro city and adjacent municipalities, São Paulo State, southwest Brazil (Fig. 1). The region occupies an area of ~1,880 km² and its population is circa 520 thousand inhabitants. The largest ceramic pole in Latin America is located in the area; moreover, agriculture, livestock, sugar-alcohol industries and other industrial activities represent the local economic base. According to CETESB [1], in 2016, about 80% of the municipalities used groundwater in their water-supply systems, including two cities (Itirapina e Ipeúna) focused in this study. A detailed study of the groundwater composition and relationship with the biotic and anthropic environment is increasingly necessary due to its enhanced use coupled to the economic activities developed in the surrounding districts. In addition, detailing the behavior of the groundwater contributes to improve the water management system, providing resources to prevent crises like that affecting São Paulo State in 2014 and 2015. This research also contributes for identifying areas of potential groundwater pollution by comparison with different potability standards (Brazil Health Ministry, CETESB – São Paulo State Environmental Agency, and WHO – World Health Organization). Previous investigators [2]–[4] realized hydrogeological and hydrochemical studies in the area. Annual reports about the groundwater quality in São Paulo State from DAEE (Water and Electricity Department) and CETESB were used in this study, as well the hydrochemical database reported by [4].



Figure 1: Location of studied wells in the selected area of this paper. 2 GENERAL FEATURES OF THE STUDIED AREA The study area is inserted in the geological context of Paraná Sedimentary Basin, which covers about 1.1 million km² in Brazil. Distinct aquifer systems (Rio Claro, Serra Geral, Guarani, Tubarão) and Passa Dois aquiclude occur there (Fig. 2). Rio Claro aquifer (RCA) has restricted local occurrence, where unconsolidated sandstones from Rio Claro Formation mark its lithological framework. Its flow ranges between 5 and 25 m³/h [3] and the water is used for supplying small local communities. Serra Geral Aquifer System (SGA) is a fractured aquifer occurring in the western part of São Paulo State. The groundwater is stored in fractures of basaltic lava flows setting from Serra Geral Formation. Its flow ranges between 7 and 100 m³/h (average = 23 m³/h) [5]. Guarani Aquifer System (GAS) is considered one of the largest aquifers in the world (total area = 1.2 million km²). It occupies 76% of the territory of São Paulo State and is confined in the majority of its occurrence [5]. The host rocks of this aquifer are well-selected sandstones with a high amount of interconnected pores from Pirambóia and Botucatu formations, which grant a high storage capacity. Reports indicate flows over 500 m³/h, although the average flow from exploited wells is 360 m³/h [6]. The siltstones, mudstones, shales, diamictites and sandstones that compose the Itararé Group form the framework of the Tubarão Aquifer System (TAS). Its productivity is low, with an average flow of 10 m³/h, being exploited in the outcropped portion [5]. Passa Dois aquiclude (PDA) is a hydrogeological unit separating GAS and TAS. The characteristics of the host rocks (shales, siltstones, mudstones and limestones) make difficult the vertical water movement. Thus, in a regional scale, Passa Dois Group is considered an aquiclude. Two geomorphological compartments occur in the studied area: Peripheral Depression (altitudes between 500 e 650 m), marked by broad, tabular and convex hills tops; Basaltic “cuestas” (up to 850 m high), characterized by hills occurring around the area [7]. Rio Claro city is located in the hydrographic basin of Corumbataí River, whose main tributaries are Passa Cinco River and Claro creek.


Water Pollution XIV

91

Figure 2: Hydrogeological map of the studied area. According to [8]. According to the Köppen-Geiger classification, the studied area present two climatic variations, i.e. tropical climate (dry winter season) and temperate (dry winter and hot summer). The annual average temperature is 22°C, whereas the annual rainfall varies between 1,325 and 1,500 mm. The vegetation covering the region is mostly characterized by semideciduous broadleaved forest with traces of savanna and trees formation. 3 MATERIALS AND METHODS This study started from bibliographic review of researches developed at Rio Claro area, and the database reported by [4] was chosen due to its wide coverage area, including the occurrence of different aquifers and a large number of analyzed parameters. The database is composed of information about 62 wells with hydro chemical analysis performed in 41. The Geochemist’s Workbench Student Edition 11.0.5 software was used to generate Piper and Schoeller diagrams for the data re-evaluation. The statistical analysis was held with Microsoft Excel 2013 in order to build box plot graphs for evaluating the dispersion of variables involved in the potability standards. The flowchart showed in Fig. 3 outlines the steps performed in this study. Hydro chemical analysis characterized the groundwaters in terms of pre-established targeted parameters aimed for the objectives of this study. The selected major ions practically comprise the whole composition of the groundwater, allowing its classification. They are: calcium (Ca²⁺), magnesium (Mg²⁺), sodium (Na⁺), potassium (K⁺), bicarbonate (HCO3¯),



Figure 3: Flowchart of the steps involved in this study. sulphate (SO4²ˉ) and chloride (Clˉ). Concentrations of nitrate (NO3ˉ) and fluoride (Fˉ) were also selected due to some high levels found in area. The pH and water temperature values were also used for checking the quality standards. The groundwater sampling of the database was done between November 1998 and November 1999 [4]. Five litters of water was collected from the nearest exit well and waters were kept at 4°C until analysis. The use of filters and chlorination were interrupted for sampling. In order to check the groundwater potability, three guidelines were used at global, national and state (local) levels. The WHO publication “Guidelines for Drinking-water Quality” was used at a global level [9]. In Brazil, at a national level, it was adopted the Ordinance No. 2914 from Health Ministry that regulates the water standards for human consumption. In São Paulo State, CETESB acts as the body regulating the water quality standards in the state (local level). The values proposed by CETESB are based in biennial reports about the groundwater quality in the state, differing in some parameters from those established by the Brazilian Health Ministry and WHO because specific characteristics from each aquifer are taken into account. 4 RESULTS 4.1 Serra Geral Aquifer (SGA) The SGA samples show a pH range between 6.5 and 7.0 with average temperature of 25.7°C. According to the Piper diagram (Fig. 4), most of the waters are of sodium chloride-sulfate type. The Schoeller diagram (Fig. 4) shows, in logarithmic scale, the following ions concentration variation: Na⁺>K+≥ Mg²⁺and Ca²⁺ and HCO3¯>Cl¯>NO3->SO4²->F. Previous researches classify these waters as calcium-magnesium bicarbonate and sodium bicarbonate type. Groundwater from basaltic aquifers are typically characterized as calcium bicarbonate as a consequence of the dissolution of pyroxene (Ca²⁺) and plagioclase (Na⁺ and Mg²⁺). 4.2 Guarani Aquifer System (GAS) The GAS samples present a pH range between 5.6 and 6.2, exhibiting acid characteristics. The average temperature is 25.2°C. The Piper diagram (Fig. 5) classifies the water as sodium chloride-sulfate type. The Schoeller diagram (Fig. 5) clearly shows the sodium dominance in comparison to other cations and a higher content of chloride in relation to other anions.


Water Pollution XIV

Figure 4:

93

(a) Piper diagram; and (b) Schoeller diagram for SGA groundwaters, showing the relation of major ions.

Figure 5: (a) Piper diagram; and (b) Schoeller diagram for GAS groundwaters. According to CETESB [1], the GAS waters are mainly characterized as calcium bicarbonate and sodium bicarbonate type. 4.3 Passa Dois Aquiclude (PDA) Only three samples were collected from PDA, since it is a restricted unit with low productivity. Two samples are of sodium chloride type, following the trend in the area, whereas the remaining sample is of chloride-sulfate magnesium type (Fig. 6). The relation of the main ions is: Na⁺>K⁺>Mg²⁺>Ca²⁺ and Clˉ>SO4²ˉ>NO3ˉ≥HCO3ˉ and Fˉ. PDA is the



most unknown unit due to its low productivity and lack of detailed hydrochemical characterization. 4.4 Tubarão Aquifer System (TAS) The TAS samples present a pH range of 8.5 to 9.5, showing alkaline conditions for these waters. The average water temperature is 26°C. Most of the sampled waters are of sodium sulfate type (Fig. 7). The Schoeller diagram shows the following relation: Na⁺>Mg²⁺>K⁺>Ca²⁺ and SO4²ˉ>Fˉ>Clˉ>HCO3ˉ> NO3ˉ. According to previous studies [10], the TAS groundwater may present two hydrochemical facies: sodium bicarbonate type and calcium bicarbonate type.

Figure 6: (a) Piper diagram; and (b) Schoeller diagram for PDA groundwaters.

Figure 7: (a) Piper diagram; and (b) Schoeller diagram for TAS groundwaters.


Water Pollution XIV

Figure 8:

95

Box plot graphs for (a) pH and (b) chloride showing the spatial distribution of values from studied aquifers compared to the reference values established by CETESB.

5 DISCUSSION The values reported in the database were compared with the guideline reference values established by CETESB, Brazilian Health Ministry and WHO. Usually, the pH does not present a direct health impact on the consumers. However, it is one important parameter as exerts control on the reactions occurring in subsurface. The WHO [9] suggests that the pH values must range between 6.5 and 9.5. The box plot graph (Fig. 8) shows that part of all sampled waters are outside that range (the majority of the GAS samples are under 6.5). Almost 50% of the TAS waters exhibit pH higher than the guideline reference value established by CETESB.



Sodium and its salts are commonly found in food and water, despite, in potable drinking water, its concentration usually does not exceed 20 mg/L. According to WHO [9], values above 200 mg/L may influence the water taste and should not be used for human consumption. Sodium is the cation with the highest concentration in relation to others in the studied area. Six samples from GAS and TAS exhibit values of 200 mg/L, which could restrict these waters for human consumption. WHO does not establish a limiting value for potassium as this ion rarely appears in concentrations that are of concern for human consumption. The selected database shows that some samples exhibit values above of the guideline reference established by CETESB (Table 1). Sulfate occurs naturally in several minerals, whose dissolution in the aquifer host rocks may cause its introduction into groundwater [9]. The WHO does not suggest a limit value for this anion in drinking water, despite levels above 200 mg/L may result in laxative effects and give a characteristic flavour to the waters. A few TAS samples exhibit values above it (Table 1). Despite chloride occurs naturally in groundwater due to water/rock-soil interactions, it may also originates from sewage, industrial effluents, and saline intrusions. The WHO [9] guideline reference value is 250 mg/L, which was not reached by any sample in the database (Table 1). Four TAS samples present values between 60 and 140 mg/L (Table 1). For nitrate concentration, WHO establishes a guideline reference value of 50 mg/L, whereas the Brazilian Health Ministry suggests 10 mg/L. The nitrate concentration in the database does not reach the WHO guideline value, however, it is sometimes above the value orientated by CETESB (Fig. 9). High nitrate concentrations could indicate the presence of contamination from garbage deposits, sewage, septic tanks, and cemeteries among others. In the study area it could reflect the unplanned urban expansion. Fluoride is an ion that in low levels (up to 1.5 mg/L, according to WHO) is beneficial for human health, although can cause dental fluorosis and bones deformation if consumed for long periods in waters containing higher levels. The TAS samples showed the highest levels, reaching values close to 140 mg/L (Table 1), possibly related to the presence of phyllosilicate clays and mica that exhibit fluorine in their structures. 6 CONCLUSION This study enabled the use of different tools for the re-evaluation of the hydrochemical database from previous work. The hydrochemical diagrams are essential to understand the relations between the main water ions and the environment in which water percolates. The statistical analysis allowed the data examination in the sample space, allowing the identification of discrepant levels and the comparison with data available in the literature comprising the studied aquifers. Based on the analysis of the results we may conclude that the studied area, in general, present sodium chloride-sulfate waters. This may occur due several factors, being the majority related to water mixing of the sampled wells. Most of the analyzed wells were built without adequate casing, which can lead to water mixing. In addition, the aquifers in Rio Claro region are in unconfined areas, therefore, more subjected to pollution. The large aquifer systems Tubarão, Guarani and Serra Geral extend over wide areas in São Paulo State, are heterogeneous and may present different hydrochemical characteristics when analyzed locally. Thus, it may be concluded that most of the groundwaters selected in the database are in accordance with the quality standards and values orientated by WHO which guarantees the safety of their use for human consumption. However, an accentuated disparity was identified between the database values and those suggested by CETESB, which may be related to the heterogeneous nature and variable characteristics of the aquifer systems throughout São Paulo State.


Water Pollution XIV

Table 1: Concentration of the major ions from database utilized in this study [4].


97


Figure 9:

[1] [2] [3]

Box plot graphs for (a) fluoride; and (b) nitrate showing the spatial distribution of values from studied aquifers compared to the reference values established by CETESB.

REFERENCES CETESB (Companhia Ambiental do Estado de São Paulo). Qualidade das águas subterrâneas do estado de São Paulo: 2013–2015, CETESB: São Paulo, p. 308, 2016. Bonotto, D.M. & Mancini, L.H., Estudo Hidroquímico dos Aquíferos de Rio Claro (SP). Geochimica Brasiliensis, 6(2), pp 153–167, 1992. Oliva, A., Estudo hidrofaciológico do aquífero Rio Claro no município de Rio Claro – SP. PhD thesis, Rio Claro (SP): São Paulo State University, p. 196, 2006.


Water Pollution XIV

[4]

99

Tonetto, E.M., Hidroquímica em aquíferos de Rio Claro (SP) e adjacências. PhD thesis, Rio Claro (SP): São Paulo State University, p. 108, 2001. [5] Departamento de Águas e Energia Elétrica, Instituto Geológico, Instituto de Pesquisas Tecnológicas do Estado de São Paulo, Serviço Geológico do Brasil, Rocha G. Mapa de Águas Subterrâneas do Estado de São Paulo, São Paulo, 2005. [6] Iritani, M.A. & Ezaki, S., As Águas Subterrâneas do Estado de São Paulo, 2nd ed., Secretaria de Estado do Meio Ambiente: São Paulo, p. 104, 2009. [7] Zaine, J.E. & Penteado-Orellana, M.M., APA Piracicaba no município de Rio Claro, SP: proposta de mudança com base em critérios geomorfológicos e políticos. Presented at XXXVIII Congresso Brasileiro de Geologia, Balneário Camboriú, SC, Oct. 1994. [8] Departamento de Águas Subterrâneas e Energia Elétrica, Instituto de Geociências e Ciências Exatas: Laboratório de Estudo de Bacias. Águas Subterrâneas no Estado de São Paulo: diretrizes de utilização e proteção. DAEE/LEBAC: São Paulo, p. 44, 2013. [9] World Health Organization, Guidelines for Drinking-Water Quality, 4th ed. Gutenberg: Malta, p. 564, 2011. [10] Diogo, A., Bertachini, A.C., Campos, H.C.N.S. & Rosa, R.B.G.S., Estudo preliminar das características hidráulicas e hidroquímicas do Grupo Tubarão no estado de São Paulo. Presented at III Simpósio Regional de Geologia, Curitiba, PR, 1987.



Water Pollution XIV

101

ARSENIC ADSORPTION INTO THE FIXED BED COLUMN FROM DRINKING GROUNDWATER VASILE MINZATU1, ADINA NEGREA1, CORNELIU MIRCEA DAVIDESCU1, CORINA SEIMAN DUDA2, MIHAELA CIOPEC1, NARCIS DUŢEANU1, PETRU NEGREA1, DANIEL DUDA SEIMAN3 & BOGDAN IOAN PASCU4 1 University Politehnica Timisoara, Faculty of Industrial Chemistry and Environmental Engineering, Timisoara, Romania 2 West University of Timisoara, Faculty of Chemistry, Biology and Geography, Timisoara, Romania 3 Victor Babes University of Medicine and Pharmacy Timisoara, Timisoara, Romania 4 Research Institute for Renewable Energy of the Politehnica University of Timişoara, Timisoara, Romania

ABSTRACT One of the main goals of the World Health Organization (WHO) and its Member States is that “all people, regardless of their stage of development and their social-economic condition, have the right to have access to adequate drinking water”. The problem of remediation of waters contaminated with toxic substances, using natural resources, is a special concern at a high level. Researchers’ efforts have sought to find effective and cheap ways to reduce the impact on the environment. As(III) gets into the environment from a variety of natural and anthropogenic sources. As As(III) is commonly found in rocks, soil or sediments, these sources are particularly important determinants of the As(III) zonal level in groundwater and surface water. As(III) contamination is currently one of the major problems with serious consequences on the biosphere. The main objective of this paper is the establishment of a method for the removal of As(III) from real waters by adsorption in a fixed bed column, using ecological materials, friendly to the environment. For this reason, the content of As(III) in groundwaters in the Romania–Hungary border area was determined, knowing the fact that the Pannonian Basin was given special attention, being an area heavily affected by As(III) contamination of natural water sources. The paper studies the adsorption of As(III) from a natural water taken from the monitored area. At the same time, it has followed the behaviour of microorganisms naturally existing in the soil of Timisoara city and in the Bega river in the presence of different amounts of As(III) in order to establish the relative toxicity of As(III) in them. Keywords: As(III), groundwater, adsorption, fixed bed column, toxicity.

1 INTRODUCTION One of the main goals of the World Health Organization (WHO) is: “all people, regardless of their developmental stage and their social economic condition, have right to access adequate drinking water” [1]. Last decade’s research proved that the groundwater sources used as drinking water sources are contaminated with As(III), which represent a real problem and become a challenge for scientists. As(III) is a metalloid which can be found in nature as trivalent or penta-valent ions. Generally, the trivalent form was founded in ground waters and the penta-valent one into the surface waters [2], [3]. As(III) toxicity is a complex phenomenon and in general is classified as acute and sub-acute, his toxicity being dependent on the form in which it is found into the contaminated water. Inorganic form of As(III), typically found in the drinking water resource is more toxic then As(III) organic compounds found in the seawater. From all As(III) compounds those with the highest toxicity are the compounds containing trivalent As(III) ions [4], [5]. Acute poisoning with As(III) occurs mostly by ingestion of contaminated food or water and requires immediate medical attention.



After consumption of As(III) contaminated water, the As(III) ions are adsorbed and transported into the body through the bloodstream. The human body is able to eliminate As(III) mainly through urine, but small amounts are eliminated through skin pores, hair, nails and sweat. Short term exposure (days, weeks) at a higher As(III) concentration present in drinking water can lead to abdominal pain, vomiting, diarrheal disease, muscular cramp, skin weakness and redness, rash, numbness, tingling or burning sensation in the hands and feet, skin thickening at level of palms or feet soles, and can culminate with the loss of movement and sensorial responses [6], [7]. Long-term exposure to As(III) contaminated water can mainly cause dysfunctions of the kidneys and after that affects the internal organs such as: lungs, liver, gall bladder, and can also cause skin lesions [6]–[8]. Regardless of the sources that generate As(III) (natural or anthropic), it reaches the underground or surface waters where it suffers a series of physical, chemical and biological processes, such as: oxidation – reduction reactions, ligands exchanges and/or biotransformation. Physical–chemical processes that can affect the evolution and the transport of As(III) ions into the ground and surface waters are similar and are influenced by pH, oxidation–reduction potential, temperature, salinity, concentration of iron and manganese ions, content of other ions found in water, and by the composition of Earth’s crust [10], [11]. Most knowledge regarding natural microorganism activity was deduced from their study under laboratory conditions. Subsequently, studies were carried out in microorganism natural environment using progressively improved analytical methods [5], [7]. In order to understand the activity of natural microorganisms we need to study the activity of a single class of microorganisms in order to see how they are affecting the environment and in order to see how the environment is affecting the microorganism population. Simultaneously it is important to study the microorganism populations, and complex microorganism systems [7]. Presence and development of microorganism into the aquatic environment are influenced by light, temperature, hydrostatic pressure, turbidity, dissolved gases, flow rate, depth, mineral and organic content [6]. Microorganisms take from surrounding environment nutrients and energy used in their fundamental activities; such nutrients are essential elements for cell life and are called biogenic elements [9]. As(III) removal from water represents an important concern and is realised using a series of techniques, such as: precipitation-filtration [12], coagulation-precipitation [13], separation by combined processes such as: photo-catalysis combined with complexation and filtration, electrocoagulation, electro-dialysis, nano-filtration, inverse osmosis, ion exchange technology, adsorption on zeolites and activated carbons [14], [15]. Cheap adsorbent such as waste and agriculture secondary products [16], oxides (manganese dioxide, activated alumina, titanium dioxide, iron oxide), hydroxides (lanthanum hydroxides), oxo-hydroxides, macroporous polymers, ion exchange resins, chelating resins, biopolymers (cellulose, human hair) are used today for As(III) removal [17]–[22]. Adsorbent material used in the present study is a composite material based on carbon and iron oxide, synthesized from cheap precursors, starch and ferric chloride. The main purpose of this study was to establish a mechanism for As(III) removal from water by adsorption in a fixed bed column, using eco-friendly materials. For this purpose, the As(III) content was determined in groundwaters from Romanian–Hungarian border, part of the Pannonian Basin. Waters from this basin receive a special attention because they are contaminated with As(III). Also, was studied the influence of different other ions dissolved into the water onto the


Water Pollution XIV

103

As(III) removal process. Subsequently the behaviour of some natural microorganisms found in Timisoara soil and Bega river was studied in the presence of different amounts of As(III) in order to determine its toxicity. 2 MATERIALS AND METHODS In order to establish the adsorption mechanism of As(III) ions from natural ground water in dynamic regime through fixed bed column, using as adsorbent a composite material containing carbon and iron oxide, water samples were taken from public fountains located along Romanian – Hungarian border. For all samples some parameters were determined: As(III) concentration, pH, nitrite, nitrate, ammonium, phosphates, and bicarbonates quantities using standardized analytic methods, specific for each parameter. For material synthesis as environmental friendly precursor starch was used for the carbon support (starch soluble extra pure, pH 6.0–7.5, 20 g/L-1 H2O, 298 K), and as precursor for iron oxide was used ferric chloride bought from Merck. 20 g of soluble starch was dissolved in 25 mL of DI water and heated at 331 K, over which was added under continuous stirring a 33% ferric chloride. These amounts of precursors were chosen in order to get a final ratio Fe:C equal to 1:10. The obtained paste was dried at 325 K for minimum 20 hours, grinded and subjected to thermal treatment at minimum 850 K in inert atmosphere. The obtained material was characterized using Quanta FEG 250 X-ray dispersion coupled electron microscope (SEM-EDX).The adsorption of As(III) was performed in dynamic mode using a glass column with inner diameter of 2.5 cm and length of 30 cm, loaded with 10 g. of prepared adsorbent material. In order to prevent clogging of the column, the composite material was mixed with quartz sand with grain between 0.6–1.85 mm using different ratios of sand: adsorbent material: 9:1; 1:1 and 3:7 Real water containing 191 μg/L-1 As(III) with pH 7.5 was introduced into the column using a HEIDOLF peristaltic pump with a flow-rate of 10 mL min-1. As(III) concentrations were determined by using inductively coupled plasma mass spectrometry ICP-MS BRUKER AuroraM90. As(III) toxicity was determined by observing the behaviour of microorganisms naturally existing in Timisoara soil and in the Bega river water in presence of different As(III) concentrations (10, 50, 100, 200, 500, 800, 1000 µg As(III)/L and 10, 100, 500 mg As(III)/L). Microorganisms culture was obtained by using two different methods: by flooding of the environment with the enriched culture and by embedding the microorganism into the environment, for both water and soil. The filter papers soaked into the As(III) solution with different concentrations were placed over the microorganism culture. Filter papers were placed in ascending orders of As(III) concentrations, clockwise. As(III) was added into the culture environment in form of As(III) ions and the microorganism growth was realized at 310 K. Culture plates where then analysed in order to quantify the number of microorganisms growth in each environment, depending on the As(III) concentration. Similar environment but in liquid form was used for microbial biomass production in order to test it on the As(III) contaminated soils. 3 RESULTS AND DISCUSIONS 3.1 Determination of As(III) content in ground waters in Romania–Hungary border area The content of As(III) in ground waters (public water wells and fountains) in the Romanian– Hungarian cross-border area is shown in Table 1.



Table 1: Cross-border groundwater monitoring. Index number

Water samples

As(III) content, (μg/L-1)

1

Pecica, Arad county, drilling 1

191

2


0.31

3


34.9

4


8.50

5

Timișoara, drilling 1

4.73

6


8.38

7


3.44

8

Arad, drilling

8.45

9

Algyő, Hungary, drilling

7.20

10

Újszentiván, Hungary, family drilling

17.5

11

Ásotthalom, Hungary, family fountain

90.8

12

Szatymaz. Hungary, family fountain

14.3

13

Szeged, Hungarya, drilling

57.7

14

Sandorfalva, Hungary, drilling

7.26

Maximum allowed concentration (MTC)

10 μg/L-1 As(III)

Obtained experimental data depicted in Table 1 confirms that the fountains located in the Hungarian part of the border contain higher amounts of As(III). In almost all of the cases this concentration exceeds the maximum allowed concentration. However, it is observed that there is no exceeding of maximum allowed concentration in case of Timisoara wells, but one exceeding was observed in Pecica, Arad county. Due to this fact, the further step into the experimental study is the elimination of As(III) for all the samples by adsorption on carbon based composite materials, using a dynamic regime. 3.2 Characterization of adsorbent material 3.2.1 Material characterization by electronic scanning microscopy SEM Images obtained from SEM are depicted in Fig. 1. Analysing the data presented in Fig. 1 we can observe the presence of iron oxide particles onto the surface of carbon material, and a relative homogenous distribution of these particles over de material surface Fig. 1(a). All iron oxide particles present the crystallite size between 630 nm and 1060 nm Fig. 1(b). 3.2.2 Material characterization using X-ray dispersion (EDX) To study the interaction between carbon and iron particles, the synthesized material was subjected to elemental EDX analysis (X-ray dispersion spectral analysis) presented in Fig. 2. From the obtained EDX spectrum the presence of peaks characteristic of iron and oxygen atoms is observed, which confirms the modification of the graphite surface.


Water Pollution XIV

(a)

105

(b) Figure 1: SEM images.

Figure 2: EDX elemental analysis. 3.3 Elimination of As(III) from deep water by dynamic adsorption. Influence of foreign ions. 3.3.1 As(III) adsorption on the column During experiments was observed the clogging of the adsorbent columns, which can be explained if we are taking into account the small dimensions of the synthesized composite materials. In order to avoid the clogging of the adsorbent column the adsorbent material was mixed in different ratio with quarts sand. As(III) ions adsorption on the column was studied following the dependence of residual concentration and the adsorption capacity of used adsorbent material, as function of the mixing ration between adsorbent material and quartz sand, dependencies depicted in Figs 3 and 4. Analyzing experimental data depicted in Fig. 3 it can be observed that by increasing the quantity of the composite adsorbent material into the mixture, the volume of water contaminated with As which is passed through column is increased until the breakdown of the column.



Figure 3:

Dependence of residual concentrations on the volume of solution passed through the column, for different (w/w) composite material:sand ratios.

Figure 4:

Dependence of the adsorption capacity of As(III) on the adsorbent material for different composite materials and ratios.

It can also be observed that for the sand: adsorbent material ratio of 9:1, the water volume passed through the column until its breakdown was 600 mL, similar for a sand:adsorbent material ratio of 1:1 the volume of solution was 2950 mL and for the 3:7 ratio, the volume of solution was 4150 mL. Based on data depicted in Fig. 4 it can be observed that the maximum adsorption capacity of the produced adsorbent material is 110 μg per gram of composite material, corresponding to the maximum efficiency. At this value the residual concentration of As ions is minimum, and further usage of the adsorbent material leads to decrease of the adsorption capacity until the residual concentration reaches the initial concentration of the As ions, representing the moment when the column is breakdown. 3.3.2 Influence of other ions present in water on the adsorption process of As(III) Experimental data presenting the influence of other ions present in water on the process of As(III) elimination using carbon based composite material are presented in Table 2.


Water Pollution XIV

107

Table 2: Influence of ions present in water on the adsorption process of As(III). Parameter

pH

H2O initial H2O treated

7.5 6.5

As(III), μg/L 191 230 m). In order to establish the adsorption performance of the synthesized adsorbent material in dynamic regime the following were studied: the variation of As(III) residual concentration, the adsorption time, the capacity and the efficiency of the adsorption process as a function of volume of water passed through column filled with adsorbent material, that depends on the quartz sand to adsorbent material ratio. From experimental data it can be observed that with an increase in the adsorbent material to quartz sand ratio, the volume of water passed through the column increases until the breakdown of the column at a ratio of 3:7; over this ratio the clogging of the adsorbent column is observed. Thus, when a quantity of 10 grams of mixture was introduced into the column at a sand to adsorbent material ratio of 1:9, the water volume which passed through the column was 600 mL in one hour until the column was broke down. For the 1:1 ratio the volume of water was 2950 mL which passed through the column in 5 hours, and for the 7:3 ratio the volume of water which passed through column was 4150 mL in 7 hours.


Water Pollution XIV

109

Based on experimental data it can be concluded that the optimal sand to adsorbent material ratio is 7:3. The maximum adsorption capacity is 112 µg As(III) per gram of adsorbent material. Also, it was found that the produced adsorbent material exhibits adsorbent properties for the other ions dissolved into the groundwater. As(III) toxicity studies were carried out using microorganism obtained from Timisoara soil and from Bega river. From the obtained experimental data, a natural resistance of the microorganism to the As(III) concentrations was observed until 10 g As(III) L-1. When the As(III) concentration reach 25 g/L-1, solution becomes toxic for microorganisms obtained from Timisoara soil and Bega river. ACKNOWLEDGEMENTS This work was partially supported by research grants PCD-TC-2017. Authors gratefully acknowledge use of the services and facilities of the Research Institute for Renewable Energy of the Politehnica University of Timişoara, founded by National Authority for Scientific Research and Innovation, through Priority Axis 2: Competitivity through research, technological development and innovation, Domain 2.2: Investments in the researchdevelopment-innovation infrastructures; Operation 2.2.1: Development of existing RD infrastructures and creation of new RD infrastructures. REFERENCES World Health Organization, Preventing Disease through Healthy Environments Exposure to Lead: A Major Public Health Concern, WHO: Geneva, 2010. [2] Zhang, L., Qin, X., Tang, J., Liu, W. & Yang, H., Review of arsenic geochemical characteristics and its significance on As(III) pollution studies in karst groundwater, Southwest China. Applied Geochemistry, 77, pp. 1–9, 2016. [3] Barringer, J.L. & Reilly, P.A., Arsenic in Groundwater: A Summary of Sources and the Biogeochemical and Hydrogeological Factors Affecting Arsenic Occurrence and Mobility, INTECH Open Access Publisher, pp. 83–116, 2013. [4] Alexander, M., Biodegradation and Bioremediation, Academic Press: San Diego, New York, Boston, London, Tokyo and Toronto, 1994. [5] Huang, Q., Xi, G., Alamdar, A., Zhang, J. & Shen, H., Comparative proteomic analysis reveals heart toxicity induced by chronic arsenic exposure in rats. Environmental Pollution, 229, pp. 210–218, 2017. [6] Kesici, G.G., Arsenic ototoxicity. Journal of Otology, 11, pp. 13–17, 2016. [7] Beard, S.J., Hashim, R., Hernandez, J., Hughes, M. & Poole, R.K., Zinc (II) tolerance in Escherichia coli K-12: Evidence that the zntA gene (o732) encodes a cation transport ATPase. Molecular Microbiology, 25(5), pp. 883–891, 1997. [8] Halina, B.R., Kalavati, C., Bukola, G.O. & Odland, J.O., Evaluation of in utero exposure to arsenic in South Africa. Science of the Total Environment, 575, pp. 338– 346, 2017. [9] Alexander, M., Biodegradation and Bioremediation, Academic Press: San Diego, New York, Boston, London, Tokyo and Toronto, 1994. [10] Arpan, S. & Biswajit, P., The global menace of arsenic and its conventional remediation – A critical review. Chemosphere, 158, pp. 37–49, 2016. [11] Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G. & Ahmed, K.M., Mechanism of arsenic release to groundwater, Bangladesh and West Bengal. Applied Geochemistry, 15(4), pp. 403–413, 2000. [1]



[12] Dinesh, M., & Pittman, C.U. Jr., Arsenic removal from water/wastewater using adsorbents—a critical review. Journal of Hazardous Materials, 142(1)–(2), pp. 1–53, 2007. [13] Song, S., Lopez-Valdivieso, A., Hernandez-Campos, D.J., Peng, C., MonroyFernandez, M.G. & Razo-Soto, L., As(III) removal from high-arsenic water by enhanced coagulation with ferric ions and coarse calcite. Water Research, 40, pp. 364– 372, 2006. [14] Bora, A.J., Gogoi, S., Gautam, B. & Robin, K., Dutta, utilization of co-existing iron in arsenic removal from groundwater by oxidation-coagulation at optimizated pH. Journal of Environmental Chemical Engineering, 4, pp. 2683–2691, 2016. [15] Rajesh, M. et al., Removal of arsenic (III) from water by magnetic binary oxide particles (MBOP): Experimental studies on fixed bed column. Journal of Hazardous Materials, Part B, pp. 469–478, 2017. [16] Samfira, I. et al., Remediation of rare earth element pollutants by sorption process using organic natural sorbents. International Journal of Environmental Research and Public Health, 12, pp. 11278–11287, ISSN-1660-4601, 2015. [17] Dandan, Z., Yang, Y., Chenghong, W. & Chen, J.P., Zirconium/PVA modified flatsheet PVDF membrane as a cost-effective adsorptive and filtration material: A case study on decontamination of organic arsenic in aqueous solutions. Journal of Colloid and Interface Science, 477, pp. 191–200, 2016. [18] Salazar, H., Poly(vinylidene fluoride-hexafluoropropylene)/bayerite composite membranes for efficient arsenic removal from water. Materials Chemistry and Physics, 183, pp. 430–434, 2016. [19] Chang-Gu, L., Arsenic(V) removal using an amine-doped acrylic ion exchange fiber: Kinetic, equilibrium, and regeneration studies. Journal of Hazardous Materials, 325, pp. 223–229, 2017. [20] Iberhan, L. & Wisniewski, M., Removal of arsenic(III) and arsenic(V) from sulfuric acid solution by liquid–liquid extraction. Journal of Chemical Technology & Biotechnology, 78, pp. 659–665, 2003. [21] Borah, D., Satokawa, S., Kato, S. & Kojima, T., Surface-modified carbon black for As(V) removal. Journal of Colloid and Interface Science, 319(1), p. 53, 2008. [22] Mohan, D. & Pittman, C.U. Jr., Arsenic removal from water/wastewater using adsorbents—a critical review. Journal of Hazardous Materials, 142(1)–(2), pp. 1–53, 2007.


Water Pollution XIV

111

A NEW ADSORBENT FOR ARSENIC REMOVAL FROM WATER MIHAELA CIOPEC1, IOSIF HULKA2, NARCIS DUŢEANU1, ADINA NEGREA1, OANA GRAD2, PETRU NEGREA1, VASILE MINZATU1 & CRISTINA ARDEAN1 1 Faculty of Industrial Chemistry and Environmental Engineering, Polytechnic University Timisoara, Romania 2 Research Institute of Renewable Energy, Polytechnic University Timisoara, Romania

ABSTRACT Water represents an essential resource for earth life and for all-natural processes. It is well known that in all developing countries the underground water resource represents the main source of drinking water and its contamination with arsenic presents a real problem. Thus, we have developed a new adsorbent based on cellulose doped with crown ethers (dibenzo-18-crown-6) functionalized with iron ions and used for the removal of arsenic from water. Usage of such extractants involves only a small amount of crown ether indicating higher efficiency of produced material, and, in order to improve the adsorbent properties and selectivity for arsenic removal, the modified cellulose was functionalized with iron ions. The new obtained adsorbent material was characterized by using energy dispersive X-ray analysis, scanning electron microscopy and Fourier Transform Infrared Spectroscopy. In order to investigate the adsorbent properties for arsenic removal equilibrium, kinetic and thermodynamic studies were performed. Arsenic adsorption from water onto a new adsorbent was studied under different experimental conditions such as reaction times, initial arsenic concentration and temperature. Obtained results show that the new produce adsorbent has a higher efficiency for arsenic removal, leading to lower residual concentration (under 10 µg As L-1 – value accepted by WHO). Keywords: arsenic, crown ethers, iron ions, water, adsorption.

1 INTRODUCTION In accordance with principle of sustainable development a new and efficient material for arsenic removal from underground waters is presented in this paper. Major objective of present paper is to produce and test a new material with efficient properties for arsenic removal from underground waters. In order to achieve this, new chemically modified materials have been obtained and tested by functionalization with crown ethers and iron ions. Water represents an essential element for entire earth life and for all-natural processes. Our existence and our economic activities are very dependent on this precious resource. Especially into the developing countries main resource of drinkable water is represented by underground water so, their contamination with arsenic represent a major problem which must be solved [1]–[3]. Arsenic represent an element which can reach and contaminate water resources from a variety of anthropogenic and natural sources. Chronic exposure of humans at inorganic arsenic increases the risk of cancer. Studies showed that the arsenic inhalation can be associated with the increases of number of patients presenting lung cancer. Moreover, arsenic ingestion was associated with skin, bladder, liver and lung cancer. Because of its toxicity, the arsenic content in drinking water over maximum admitted level by the World Health Organization (10μg As L-1) has a negative impact on the human health [1]–[8]. Valence of arsenic inorganic species depends on the redox conditions and pH of the underground waters. In aqueous solutions at pH between 6 and 9 As(V) can be found in form of four species: H3AsO4, H2AsO4–, HAsO42–, AsO43–, and As(III) species appear at pH= 9 as H3AsO3, H2AsO3–, HAsO32–, AsO3 [4]–[6], [9], [10]. Therefore, to reduce the negative impact of arsenic towards human health is necessary to develop new and efficient technologies for its removal [3]. Conventional methods used for arsenic removal from aqueous solutions are



precipitation, coagulation and filtration, inverse osmosis, ion exchange and adsorption. From all these methods, adsorption represent the most eloquent one, both form the efficiency of arsenic removal point of view and for the economical point of view [1]–[12]. In past years, it has been urgently necessary to obtain advanced composite materials with applications into the field of selective recovery and separation of metals and non-metals ions. Adsorption process efficiency can be improved by developing new methods to produce composite materials by chemical modification of inorganic and organic solid supports, through functionalization with different extractants. Currently, the used method for advanced adsorbent material production are: wet method (diluted extractant dissolved in different solvents is placed in contact with the solid support, being absorbed by the support), dry method (extractant is diluted in proper solvent is placed in contact with the support followed by slow evaporation of the solvent under vacuum), modifier addition (represent a hybrid between dry and wet methods), and the column dynamic method which present the advantage of the short time of functionalization correlated with increased efficiency of the adsorption process [13]–[24]. In order to apply these methods, the extractants should be liquid or be kept in liquid form by adding a proper solvent; extractant and solvent must have a minimum solubility, the support must be prepared for impregnation and the functionalization method does not have to change the extractant or support properties [13], [18], [25]. In last decade as support were used macroporous polymeric resins (Amberlite XAD resins) with a right three-dimensional structure appropriate to incorporate higher quantities of extractants due to proper specific surface area. In addition, they present great mechanical resistance and low solvent swelling time during functionalization process [25]. According to the specialty literature for the arsenic removal using Amberlite class polymers it is necessary to functionalize the used polymeric support through chemical synthesis, which leads at increase of the price of obtained material [26]. Other types of adsorbent materials used in arsenic removal process through adsorption are activated alumina, active carbon, copper–zinc granules, iron hydroxide granules, iron, aluminium, manganese oxides, etc. All these materials can be used as adsorption media as they are, or they can be impregnated on various supports such as silicates, ceramic materials, cellulose, etc. [27]–[29]. Such modified materials point out great adsorption capacities (>100 mg As per each gram of material), but such materials cannot be used for removal of trace arsenic from underground waters (under 100μg per litter, case of waters form south-west of Romania due to the natural fond) [30]. In order to solve this stringent item, it is really important to develop new materials with efficient adsorbent properties. Adsorbent properties of solid supports can be improved by chemical modification using functionalization method with different extractants. The main extractants used in the specialty literature are: organophosphoric acid extractants (di (2-etylhexyl) phosphoric acid – DEHPA, di (2-etylhexyl) ditiophosphoric acid – DEHTPA, 2-etylhexyl phosphoric acid mono – 2-etylhexil ester (HEHEHP), di (2, 4, 4-trimetyl-pentyl) phosphoric acid – DTMPPA), neutral organophosphoric extractants (three – n-butyl phosphate – TBP, three – n-octyl phosphine oxide – TOPO), mixture of organophosphoric acid extractants, bifunctional organophosphoric basic extractants, etc. [13], [15], [17], [18], [25]. Based on data presented onto the specialty literature is well known that the crown ethers can be successfully used for metal ions removal because they allowed the complexation and embedding of metallic ions inside or at their surface. For the trace arsenic removal through adsorption is important that the adsorbent materials present good adsorbent properties, which can be obtained through chemical modification [31]. Based on that, the main objective was to obtain new materials with improved adsorbent


Water Pollution XIV

113

properties by functionalization of solid supports using crown ethers as extractants, followed by loading with iron. The original aspect of the paper consists in the fact that literature is not presenting studies regarding the chemical modification of cellulose through functionalization with crown ethers and loaded with iron ions in the view of arsenic removal from waters. Even if the crown ethers are expensive reagents, the proposed method for their usage imply only small quantities, this way combining the advantage of crown ethers usage with the properties of solid supports. Obtained materials presents therefore high adsorbent performance at affordable price, moreover obtained materials are addressed to some niche applications such as arsenic removal from underground waters to achieve the WHO requirements. 2 MATERIALS AND METHODS In order to obtain new adsorbent materials by chemical modification of solid support were used: as support cellulose (AVICEL 101, microcrystalline, powder, Sigma-Aldrich, Merck) with a particle size of ~50μm, as extractant Di-benzo 18 crown 6 (DB18C6) (purity, 98%, Sigma-Aldrich, Merck), and iron chloride purchased from Sigma-Aldrich. 2.1 Functionalization of polymers To obtain the modified adsorbent material, 0.1 g of Di-benzo 18 crown 6 (DB18C6) were accurately weighed and dissolved in 25mL acetone (99.9% SC ECO-MOLD Invest SRL, Romania). The obtained extractant solution is added over 5 g of support-cellulose with a particle size of ~50μm. The SIR dry method is used to functionalization the support. Thus, the extractant and the support remain in contact for 24 hours, after which it is filtered and dried in oven at 50°C for 24 hours. To load the material with iron ions 25mL of FeCl3 solution with concentration of 100mgL-1 were added to the material, left in contact for 24 hours, then filtered and dried in the oven for 24 hours. 2.2 Characterization of the functionalized polymers The materials obtained were characterized by X-ray dispersion (EDX) using a FEI Quanta FEG 250 instrument and Fourier Transformed Infrared Spectroscopy (FTIR) using a Bruker Platinum ATR-QL Diamond instrument in the range of 4000–400cm-1. 2.3 Sorption studies In order to determine optimum conditions for the arsenic adsorption process were carried out experiments regarding contact time, initial concentration of As(V) and temperature. The experiments were carried out in a Julabo SW23 thermostatic and shaking water bath with a stirring speed of 200rpm. To establish the influence of contact time and temperature on the adsorption capacity of the functionalized material 0.1g of material was mixed with 25mL As(V) solution with concentration of 50µgL-1. Samples were shaken for different time periods (0.5, 1, 2, 3, 4, 5, 6, 7, 8 hours) in a thermostatic bath at different temperatures (298K, 308K and 318K) with a stirring speed of 200rpm. After that all samples were filtered and the filtrate was analysed to evaluate the arsenic residual concentration using an Inductively coupled plasma mass spectrometer-ICP-MS Bruker Aurora M90 type. In order to establish the effect of the initial concentration of As(V) on the adsorption capacity of materials, different concentrations of As(V) solutions (10, 25, 50, 75, 100, 150, 175 and 200μgL-1) were prepared. Adsorption processes were carried out at pH~7, for a contact time of 4 hours at 298 K.



3 RESULTS AND DISCUSSION 3.1 Characterization of the doped materials 3.1.1 X-ray energy dispersive spectroscopy Cellulose functionalized with DB18C6 and loaded with iron ions was characterized by using X-ray dispersion (EDX) technique. Obtained data are represented in Fig. 1. From the EDX spectra, can observe the presence of specific peaks of the elements from chemical structure of cellulose, crown ether namely C and O and iron. 3.1.2 Fourier transform infrared spectroscopy The use of FTIR spectroscopy is used to identify chemical groups and for quantitative analysis of different samples. The FT-IR spectra for the studied material – cellulose functionalized with DB18C6 and iron are shown in Fig. 2.

Figure 1:

EDX spectra of materials obtained by functionalized cellulose with crown etherDB18C6 and iron doped.

Figure 2:

IR spectra of materials obtained by functionalized cellulose with crown etherDB18C6 and iron doped.


Water Pollution XIV

115

From FTIR spectra (Fig. 2) obtained for cellulose functionalized with DB18C6 and loaded with iron ions were identified the bands specific for cellulose groups, in the range of 3200– 3400cm-1 associated with the stretching vibrations of O-H bonds [32]. Simultaneously was identified a band located around 1600cm-1 which are characteristic for stretching vibrations water specific O-H bonds. The bands located at 1720 and 1600cm-1 are assigned to aromatic C=C bond stretching vibrations [33]. Bands characteristic of DB18C6 extractant appear in the range 1550–500cm-1, the most intense are located at 1000cm-1 and 550cm-1 [34]. Thus, dibenzo-18-crown-6-crown ether specific vibrations bands can be observed around 1000cm-1 and 1100cm-1 that can be attributed to Calphatic-O-Caromatic respectively Calphatic-O-Calphatic bonds [31]. At the same time, iron surface loading is evidenced by the presence of 1037cm-1 of some peaks characteristic of the Fe-OH bonds [30]. 3.2 Sorption studies results 3.2.1 The effect of contact time, temperature and sorption kinetics studies The effect of contact time on the adsorption of As(V) on the obtained functionalized material was studied as was previously described. Obtained experimental data are showed in Fig. 3. For this study, contact time was varied in the range between 0.5–8 hours, and temperature in the range 298–318K. From data depicted in Fig. 3 can observe that the adsorption capacity increases with the increase of contact time, and the equilibrium is reached after about 4 hours for all used temperatures. At the same time, it is observed that with the temperature increase, the adsorption capacity of the material is not significantly affected. In order to study the As(V) adsorption mechanism the experimental data were modelled using pseudo-first-order and pseudo-second order kinetic models. Fig. 4 depicts the pseudo-first-order and pseudo-second order kinetic model’s plots obtained at temperatures used during adsorption experiments (298, 308 and 318K).

Figure 3: Effect of contact time on the adsorption capacity of the studied material.



(a) Figure 4:

(b)

Pseudo-first-order and pseudo-second-order plots for As(V) adsorption onto the material. (a) Pseudo-first-order kinetic model; (b) Pseudo-second-order kinetic model.

Kinetic parameters for As(V) sorption onto studied material at three different temperatures obtained from used kinetic models are presented in Fig. 4. The value of constant k1 for the pseudo-first-order model was calculated using the slope of the linear graphical representation of ln (qe-qt) function of time. Similarly, the constant k was determined for pseudo-second-order model, from the slope of the linear graphical representation of the t/qt as function of time. If the correlation coefficient R2 is closer to 1, the adsorption process presents a better linearization for one of the two kinetic models presented. Thus, from the experimental data presented in the tables from Fig. 4, it is observed that pseudo-second-order model is better describing the arsenic adsorption onto the studied material, because the obtained correlation coefficient R2 is in the range 0.9970–0.9983 depending on the temperature and is higher than that found for the pseudo-first-order model (R2 = 0.8438–0.8924). Due to the fact that adsorption kinetics was better described by the pseudo-second-order kinetic model compared to the pseudo-firstorder model suggesting that the As(V) removal process correspond to a chemisorption mechanism. In order to establish the As(V) behaviour on the surface of the adsorbent material during the adsorption process, the experimental data obtained were modelled according to Langmuir, Freundlich and Sips isotherms (depicted in Fig. 5) generally used to describe the adsorption processes. The correlation coefficient R2 was calculated to establish which adsorption isotherm describes better the adsorption process of As(V). Inset of figures are presented the parameters obtained for each used isotherm. The highest value of the correlation coefficient R2 obtained when the experimental data were modelled using Langmuir isotherm (R2=0.91685) versus the Freundlich isotherm (R2=0.79081) allows us to consider that the Langmuir isotherm describes better the As(V) adsorption process onto the chemically modified material. But because the value of the correlation coefficient R2 for Sips isotherm is much closer to 1 (R2=0.99726), can conclude that the Sips model best describes the adsorption process. It is also observed that the value of maximum adsorption capacity evaluated based on Sips isotherm model have a value of 12.24μg As g-1, much closer to the experimental value of


Water Pollution XIV

117

Figure 5: Adsorption isotherm of As(V) onto material. 12.079 μg As g-1. Based on that can conclude that the adsorption process of As(V) on the obtained material is mono-layer adsorption on to the heterogeneous surface. The adsorption mechanism is controlled by chemisorption processes as a result of strong chelation between As(V) and OH- groups or Fe(III) ions present on the functionalized material surface. The value of the coefficient ns

Water Pollution XIV

Water Pollution XIV

Suggest Documents

Water pollution

Let's Talk About Pollution . . .Water Pollution Water - City of Gary

water pollution

Water, Air, & Soil Pollution

Water and Soil Pollution

Pesticides and Water Pollution

Modelling indicators of water security, water pollution

Technologies for Water Pollution Control

Water Pollution in India - IDFC

Thermal Pollution in Water - eolss

Water, Air, & Soil Pollution

Water Pollution - Singapore Environment Council

Water Pollution in India - IDFC

Water, Air, & Soil Pollution

Water Pollution - India Environment Portal

Water Pollution - Singapore Environment Council

Water, Air, & Soil Pollution - Springer

Water Pollution in India

Lesson 2. Pollution and Water Quality Pollution Sources

Chapter 2 Water Pollution and Riverbank Filtration for Water ... - Springer

Water Pollution and Water Quality Assessment of Major ...

Chapter 2 Water Pollution and Riverbank Filtration for Water Supply ...

NSW Cold Water Pollution Strategy - NSW Office of Water

Chapter 2 Water Pollution and Riverbank Filtration for Water ... - Springer