issn 2317-126x • us$: 15.00 a j h , water e s · 2019-02-04 · american journal of hydropower,...

American Journal of Hydropower, Water and Environment Systems, july 2016 1

Volume 1 • August, 2014ISSN 2317-126X • US$: 15.00

AmericAn JournAl of Hydropower, wAter And

environment SyStemS

Technical Papers

06 SETTING ENVIRONMENTAL FLOWS IN A MEDITERRANEAN RIVER USING 2D HABITAT MODELINGIsabel Boavida, José Maria Santos, Rui Cortes, Teresa Ferreira, António Pinheiro

12 METHODOLOGY FOR THE USE OF RADIOACTIVE TRACERS IN HYDROSEDIMENTOLOGICAL STUDIESJefferson Vianna Bandeira, Lécio Hannas Salim, Cláudio José Chagas, Duarte Augusto Costa, Namir Souza Vieira, Vinícius Verna Magalhães Ferreira

17 COMPUTATIONAL MODELING OF FLUID FLOW ON ENCAPSULATED PIPELINE TRANSPORT Gabriel de Carvalho Nascimento, Carlos Alexandre Bastos de Vasconcellos, Marcelo de Miranda Reis, José Carlos Cesar Amorim

20 BRAZILIAN HYDROELECTRIC REHABILITATION POTENTIAL AND VIABILITYElisa de Podestá Gomes, Sérgio Valdir Bajay

25 VALIDATION OF A COMPUTATIONAL FLUID DYNAMICS MATHEMATICAL SIMULATION WITH A PHYSICAL MODEL OF A PUMPING STATIONSergio Liscia, Ezequiel Lacava, Milagros Loguercio, Cecilia Lucino

29 CFD OPTIMIZATION OF LOW HEAD TURBINES INTAKE USING FISHER-FRANKE GUIDELINESMauricio Angulo, Sergio Liscia

34 GUIDE VANE INFLUENCE OVER PRESSURE FLUCTUATION AT THE DISCHARGE RING IN A KAPLAN TUR-BINE: EXPERIMENTAL ASSESSMENTArturo Rivetti, Cecilia Lucino, Sergio Liscia

38 HYDRAULIC TRANSITORY STUDY IN THE SMALL HYDROPOWER BY CHARACTERISTICS METHOD IN ORDER TO SURGE TANK DIMENSIONINGRegina Mambeli Barros, Geraldo Lúcio Tiago Filho, Ivan Felipe Silva dos Santos, Fernando das Graças Braga da Silva

Technical Notes

48 AN APPROACH TO THE MAIN PROBLEMS FACED WHEN DEVELOPING ENVIRONMENTAL STUDIES IN BRAZILMaria Rita Raimundo e Almeida, Maria Inês Nogueira Alvarenga

Published with the support of Hydraulic Machinery and Systems

International Association

WORKING GROUPlatinamerican

Editorial

During “25th IAHR Symposium on Hydraulic Machinery and Systems”, worldwide event of IAHR that took place on September 2010 in ROMENIA, a proposal was born for creating a Latin American group with the purpose of increasing discussions on hydraulic machines and systems area under the coordination of International Association Hydraulic Research. In that instance, Prof. Geraldo Lúcio Tiago Filho was appointed as responsible for coordination of structuring the group that now counts with 18 participating companies and institutions, ALSTOM, ANDRITZ, IMPSA, KSB, VOITH, Brazilian universities: IME, UFMG, UFMT, UFRJ, UnB, UNICAMP, UNIFEI, USP and Argentinian universities: UNAM, UNCOMA, UNCU, UNLP and UTN-Regional Mendoza.

The LAWG-IAHR group was created on September 2011 to fit demands of researchers and specialists in Latin America, in disclosure of knowledge on hydraulic machines, associated components and systems area.

IAHR, founded on 1935, is a worldwide and independent organization of engineers and specialists acting on areas related to hydraulics, environment and its applications. Activities range from basic hydraulics applied to rivers and tides to water resources development through the use of computer hydraulics tools and regular training. IAHR stimulates and promotes research and its application, and thus contributes for sustainable development, optimization of water resources management and industrial processes of worldwide effluents. IAHR reaches its purposes through a range of activities including groups of work, research performing, events, short term trainings, journals publication, monographs and papers. These activities count with engagement of international programs as UNESCO, WMO, IDNDR, GWP, ICSU and cooperation organizations related to water subject worldwide.

IAHR publishes two international scientific magazines from its headquarters in Madrid, Spain, in partnership with Taylor and Francis - Journal of Hydraulic Research and Journal of River Basin Management. The International Journal of Hydro-Environment Research (JHER) is published by Asia Division of IAHR in partnership with Water Resources Korean Association and Elsevier. IAHR publishes semimonthly “Hydrolink” Magazine and monthly “NewFlash”, newsletter targeted to international hydraulic community.

For disclosure of Latin American group scientific production, we have launched with this edition the American Journal of Hydropower, Water and Environment Systems, with quadrimestral periodicity and our goal being to fulfill existing gap for publications in Latin America. Thus, we hope to count on a much more expressive participation of researchers with the purpose to strengthen scientific initiation of American countries. Special thanks for our supporting members and Prof. Augusto N.C. Viana for their contribution for feasibility of this publication.

Geraldo Lúcio Tiago FilhoEditor in Chief

IAHR DIVISION I: HYDRAULICSTECHNICAL COMMITTEE: HYDRAULIC

MACHINERY AND SYSTEMS

Editors in Chief Prof. Geraldo Lucio Tiago Fº - UNIFEIProf. Eduard Egusquiza - UPC

Executive Editors Prof. Carlos Martinez - UFMGEng. Humberto Gissoni - VOITH

Technical EditorsProf. Regina Mambeli Barros - UNIFEIProf. Cecilia Lucino - UNLP

Journalist in chargeAdriana Barbosa MTb - MG 05984JP

Graphic Projec/ DiagrammingLidiane Silva

Circulation1,000 copies

Federal University of Itajubá - UNIFEIAv. BPS, 1303 - Bairro Pinheirinho

Itajubá - MG - Brasil - CEP: 37500-903

ISSN 2317-126X


environment SyStemS

LAWG-IAHR Executive Secretariat Lucia Garrido Rios

[email protected]

2 American Journal of Hydropower, Water and Environment Systems, aug 2014

American Journal of Hydropower, Water and Environment Systems

A publication of Latin American Working Group of the International Association for Hydro-Environment Engineering and Research-IAHR

All papers must be submitted in English. In case the author wants to translate the article through the journal all costs for the translation will be charged on the account of the author.

1. Formatting articles

1.1. Article structure

1.1.1 Subdivision - numbered sections

Divide your article into clearly defined and numbered sections. Subsections should be numbered 1.1 (then 1.1.1, 1.1.2, ...), 1.2, etc. (the abstract is not included in section numbering). Use this numbering also for internal cross-referencing: do not just refer to ‘the text’. Any subsection may be given a brief heading. Each heading should appear on its own separate line.

1.1.2 Format

All text of the manuscript must be located within a 170 mm by 252 mm rectangle of a white A4 page or within 170 mm by 240 mm for the letter format. The margins are given in Table 1. An example of the page format is given in Fig. 1

[Table 1]: Page margin for manuscripts.

Margin Position Top Bottom Left Right

Margin size (cm) 2.0 2.5 2.0 2.0

All text should be single spaced, black and in 12-point type. “Times News Roman” or a similar proportional font should be used. Total length 15 pages in Word.

The terminology given in the IEC Technical Report for the Nomenclature of Hydraulic Machinery is recommended.

Introduction

State the objectives of the work and provide an adequate background, avoiding a detailed literature survey or a summary of the results.

Material and methods

Provide sufficient details to allow the work to be reproduced. Methods already published should be indicated by a reference: only relevant modifications should be described.

Theory/calculation

A Theory section should extend, not repeat, the background to the article already dealt with in the Introduction and lay the foundation for further work. In contrast, a Calculation section represents a practical development from a theoretical basis.

Results

Results should be clear and concise.

Discussion

This should explore the significance of the results of the work, not repeat them. A combined Results and Discussion section is often appropriate. Avoid extensive citations and discussion of published literature.

Conclusions

The main conclusions of the study may be presented in a short Conclusions section, which may stand alone or form a subsection of a Discussion or Results and Discussion section.

INSTRUCTIONS FOR AUTHORS

References

Within the text, references should be cited in numerical order according to their order of appearance. The numbered reference citation within text should be enclosed in brackets.

After the second edition all papers must have at least one reference of the American Journal of Hydropower, Water and Environment Systems.

Example: It was shown by Prusa [1] that the width of the plume decreases under these conditions.

In the case of two citations, the numbers should be separated by a comma [1,2]. In the case of more than two references, the numbers should be separated by a dash [5-7].

List of References. References to original sources for cited material should be listed together at the end of the paper; footnotes should not be used for this purpose. References should be arranged in numerical order according to the sequence of citations within the text. Each reference should include the last name of each author followed by his initials.

(1) Reference to journal articles and papers in serial publications should include:

• last name of each author followed by their initials• year of publication• abbreviated title of publication in which it appears• full title of the cited article in quotes, title capitalization• volume number (if any) (Do not include the abbreviation,

“Vol.”)• issue number (if any) in parentheses (Do not include the

abbreviation, “No.”)• inclusive page numbers of the cited article (include “pp.”)

(2) Reference to textbooks and monographs should include:

• last name of each author followed by their initials• year of publication• titles in examples may be in italic• publisher• city of publication• inclusive page numbers of the work being cited (include “pp.”)• chapter number (if any) at the end of the citation following

the abbreviation, “Chap.”

(3) Reference to individual conference papers, papers in compiled conference proceedings, or any other collection of works by numerous authors should include:

• last name of each author followed by their initials• year of publication• full title of the cited paper in quotes, title capitalization• individual paper number (if any)• full title of the publication• initials followed by last name of editors (if any), followed

by the abbreviation, “eds.”• publisher• city of publication• volume number (if any) in boldface if a single number,

include, “Vol.” if part of larger identifier (e.g., “PVP-Vol. 254”)• inclusive page numbers of the work being cited (include “pp.”)

(4) Reference to theses and technical reports should include:

• last name of each author followed by their initials• year of publication• full title in quotes, title capitalization• report number (if any)• publisher or institution name, city

Sample References

[1] Ning, X., and Lovell, M. R., 2002, “On the Sliding Friction Characteristics of Unidirectional Continuous FRP Composites,”

American Journal of Hydropower, Water and Environment Systems, aug 2014 3


ASME J. Tribol., 124(1), pp. 5-13.[2] Barnes, M., 2001, “Stresses in Solenoids,” J. Appl. Phys.,

48(5), pp. 2000–2008.[3] Jones, J., 2000, Contact Mechanics, Cambridge University

Press, Cambridge, UK, Chap. 6.[4] Lee, Y., Korpela, S. A., and Horne, R. N., 1982, “Structure of

Multi-Cellular Natural Convection in a Tall Vertical Annulus,” Proc. 7th International Heat Transfer Conference, U. Grigul et al., eds., Hemisphere, Washington, DC, 2, pp. 221–226.

[5] Hashish, M., 2000, “600 MPa Waterjet Technology Development,” High Pressure Technology, PVP-Vol. 406, pp. 135-140.

[6] Watson, D. W., 1997, “Thermodynamic Analysis,” ASME Paper No. 97-GT-288.

[7] Tung, C. Y., 1982, “Evaporative Heat Transfer in the Contact Line of a Mixture,” Ph.D. thesis, Rensselaer Polytechnic Institute, Troy, NY.

[8] Kwon, O. K., and Pletcher, R. H., 1981, “Prediction of the Incompressible Flow Over A Rearward-Facing Step,” Technical Report No. HTL-26, CFD-4, Iowa State Univ., Ames, IA.

[9] Smith, R., 2002, “Conformal Lubricated Contact of Cylindrical Surfaces Involved in a Non-Steady Motion,” Ph.D. thesis, http://www.cas.phys.unm.edu/rsmith/homepage.html

1.1.2 Essential title page information

• Title. Concise and informative. Titles are often used in information-retrieval systems. Avoid abbreviations and formulae where possible.

• Author names and affiliations. Where the family name may be ambiguous (e.g., a double name), please indicate this clearly. Indicate all affiliations with a number immediately after the author’s name and in front of the appropriate address. Provide the full postal address of each affiliation, including the country name and, if available, the e-mail address of each author.

• Author résumé. The author must inform the graduation degree, post graduation, affiliation and email address. The résumé must not exceed 150 characters.

• Corresponding author. Clearly indicate who will handle correspondence at all stages of refereeing and publication, also post-publication. Ensure that e-mail address and the complete postal address are provided. Contact details must be kept up to date by the corresponding author.

• Present/permanent address. If an author has moved since the work described in the article was done, or was visiting at the time, a ‘Present address’ (or ‘Permanent address’) may be indicated as a footnote to that author’s name. The address at which the author actually did the work must be retained as the main, affiliation address. Superscript Arabic numerals are used for such footnotes.

Abstract

A concise and factual abstract is required. The abstract should state briefly the purpose of the research, the principal results and major conclusions. An abstract is often presented separately from the article, so it must be able to stand alone. For this reason, References should be avoided, but if essential, then cite the author(s) and year(s). Also, non-standard or uncommon abbreviations should be avoided, but if essential they must be defined at their first mention in the abstract itself.

Keywords

Immediately after the abstract, provide a maximum of 6 keywords, using American spelling and avoiding general and plural terms and multiple concepts (avoid, for example, ‘and’, ‘of’). Be sparing with abbreviations: only abbreviations firmly established in the field may be eligible. These keywords will be

used for indexing purposes.

Abbreviations

Define abbreviations that are not standard in this field in a footnote to be placed on the first page of the article. Such abbreviations that are unavoidable in the abstract must be defined at their first mention there, as well as in the footnote. Ensure consistency of abbreviations throughout the article.

Acknowledgements

Collate acknowledgements in a separate section at the end of the article before the references and do not, therefore, include them on the title page, as a footnote to the title or otherwise. List here those individuals who provided help during the research (e.g., providing language help, writing assistance or proof reading the article, etc.).

Nomenclature and units

Follow internationally accepted rules and conventions: use the international system of units (SI). If other quantities are mentioned, give their equivalent in SI.

Math formulae

Present simple formulae in the line of normal text where possible and use the solidus (/) instead of a horizontal line for small fractional terms, e.g., X/Y. In principle, variables are to be presented in italics. Powers of e are often more conveniently denoted by exp. Number consecutively any equations that have to be displayed separately from the text (if referred to explicitly in the text).

Footnotes

Footnotes should be used sparingly. Number them consecutively throughout the article, using superscript Arabic numbers. Many wordprocessors build footnotes into the text, and this feature may be used. Should this not be the case, indicate the position of footnotes in the text and present the footnotes themselves separately at the end of the article. Do not include footnotes in the Reference list.

Table footnotes

Indicate each footnote in a table with a superscript lowercase letter.

Artwork

Electronic artwork General points • Make sure you use uniform lettering and sizing of your

original artwork. • Save text in illustrations as ‘graphics’ or enclose the font. • Only use the following fonts in your illustrations: Arial,

Courier, Times, Symbol. • Number the illustrations according to their sequence in

the text. • Use a logical naming convention for your artwork files. • Provide captions to illustrations separately. • Produce images near to the desired size of the printed

version. • Submit each figure as a separate file.• Pictures, graphics and images must be submitted in a JPG

or GIF format with 300 dpi.

2 Conducting the Review

2.1 Originality

You might wish to do a quick literature search using tools such as Scopus to see if there are any reviews of the area. If the research has been covered previously, pass on references of those works to the editor.


2.2 Structure

Consider each element in turn: Title; Abstract; Introduction (It should describe the experiment, the hypothesis(es) and the general experimental design or method); Method; Results; Conclusion/Discussion; Language: you do not need to correct the English. You should bring this to the attention of the editor, however.

2.3 Previous Research

If the article builds upon previous research does it reference that work appropriately? Are there any important works that have been omitted? Are the references accurate?

2.4 Ethical Issues

Plagiarism: If you suspect that an article is a substantial copy of another work, please let the editor know, citing the previous work in as much detail as possible

Fraud: It is very difficult to detect the determined fraudster, but if you suspect the results in an article to be untrue, discuss it with the editor

AUTHORIZATION FOR PUBLICATION OF PAPERS

LICENSE FOR USE OF INTELLECTUAL WORK (Author)

For this private instrument the AUTHOR, below signed authorizes the IAHR Latin American Working Group, to publish its work authorship, without any obligation and in exclusiveness character for the period of six months starting from the publication in the AMERICAN JOURNAL OF HYDROPOWER, WATER AND ENVIRONMENT SYSTEMS, or in another official publication of IAHR.

In case of joint authorship, the first author signs as AUTHOR, assuming before IAHR the commitment of informing the other authors of the granted license.

AUTHOR (full name in form letter):

Title of the Paper:

JOINT AUTHORS [full name in form letter]:

ADDRESS:

.

Email:


American Journal of Hydropower, Water and Environment Systems, aug 2014 5

Technical Papers

06 SETTING ENVIRONMENTAL FLOWS IN A MEDITERRANEAN RIVER USING 2D HABITAT MODELINGIsabel Boavida, José Maria Santos, Rui Cortes, Teresa Ferreira, António Pinheiro

12 METHODOLOGY FOR THE USE OF RADIOACTIVE TRACERS IN HYDROSEDIMENTO-LOGICAL STUDIESJefferson Vianna Bandeira, Lécio Hannas Salim, Cláudio José Chagas, Duarte Augusto Costa, Namir Souza Vieira, Vinícius Verna Magalhães Ferreira

17 COMPUTATIONAL MODELING OF FLUID FLOW ON ENCAPSULATED PIPELINE TRANSPORT Gabriel de Carvalho Nascimento, Carlos Alexandre Bastos de Vasconcellos, Marcelo de Miranda Reis, José Carlos Cesar Amorim

20 BRAZILIAN HYDROELECTRIC REHABILITATION POTENTIAL AND VIABILITYElisa de Podestá Gomes, Sérgio Valdir Bajay

25 VALIDATION OF A COMPUTATIONAL FLUID DYNAMICS MATHEMATICAL SIMULATION WITH A PHYSICAL MODEL OF A PUMPING STATIONSergio Liscia, Ezequiel Lacava, Milagros Loguercio, Cecilia Lucino

29 CFD OPTIMIZATION OF LOW HEAD TURBINES INTAKE USING FISHER-FRANKE GUIDELINESMauricio Angulo, Sergio Liscia

34 GUIDE VANE INFLUENCE OVER PRESSURE FLUCTUATION AT THE DISCHARGE RING IN A KAPLAN TURBINE: EXPERIMENTAL ASSESSMENTArturo Rivetti , Cecilia Lucino , Sergio Liscia

38 HYDRAULIC TRANSITORY STUDY IN THE SMALL HYDROPOWER BY CHARACTERISTICS METHOD IN ORDER TO SURGE TANK DIMENSIONINGRegina Mambeli Barros, Geraldo Lúcio Tiago Filho, Ivan Felipe Silva dos Santos, Fernando das Graças Braga da Silva

Technical Notes

48 AN APPROACH TO THE MAIN PROBLEMS FACED WHEN DEVELOPING ENVIRONMENTAL STUDIES IN BRAZILMaria Rita Raimundo e Almeida, Maria Inês Nogueira Alvarenga

IAHR DIVISION I: HYDRAULICSTECHNICAL COMMITTEE: HYDRAULIC

MACHINERY AND SYSTEMS

Scientific Committee

Alessandro Perroti Martins

Alexandre Ferretti

Alexandre Kepler

André Mesquita

Antonio Brasil Júnior

Augusto Nelson Viana

Benedito Márcio de Oliveira

Carlos Barreira Martinez

Cecilia Lucino

Daniel Fernández

Daniel Rodriguez

Edmundo Koelle

Facundo González

Fernando Zárate

Gabriel Tarnowski

Geraldo Lucio Tiago Filho

Humberto de Camargo Gissoni

José Carlos Amorim

José Geraldo P. de Andrade

Juan Carlos Cacciavillani

Lubienska Cristina Lucas Jaquie Ribeiro

Luciano dos Santos

Miguel Tornell

Orlando Anibal Audisio

Pablo Martin Magistocchi

Rafael Acedo Lopes

Regina Mambeli Barros

Ricardo Vasconcellos

Roque Zanata

Segen Farid Estefen

Sergio Liscia

Zulcy de Souza

Number 1 August 2014


environment SyStemS

6 American Journal of Hydropower, Water and Environment Systems, aug 2014, page 6-11

Setting environmental flowS in a mediterranean river uSing 2d habitat modeling

1Isabel Boavida, 2José Maria Santos, 3Rui Cortes, 2Teresa Ferreira, 1António Pinheiro

1CEHIDRO, Instituto Superior Técnico, Technical University of Lisbon, Av. Rovisco Pais, 1049-001 Lisboa, Portugal, E-mail address of the corresponding author: [email protected] de Estudos Florestais, Technical University of Lisbon, Tapada da Ajuda, 1349-017 Lisboa, Portugal 3Centro de Investigação e de Tecnologias Agro-Ambientais e Biológicas, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5001-801Vila Real, Portugal

ABSTRACT

Over the years, the need to sustain and maintain the aquatic ecosystems has been widely recognized and embraced by river managers and environmental policies and legislation. In particular, minimum flow regimes have been determined and implemented all over the world. The Instream Flow Incremental Methodology (IFIM), one of the most used methodologies to evaluate the effects in biota of changing the flow regime, has been criticized for several reasons. These problems arise from a limited knowledge of many species, the lack of rigorous description of habitat requirements and the fact that habitat suitability curves have been developed for a very limited range of conditions. To overcome these issues and thus determine the environmental flows in a regulated Mediterranean-type stream, the present study applies the IFIM using the River2D in the physical habitat simulation. The IFIM procedures and limitations were taken into consideration and a few guidelines were proposed. The findings of this study point out to the need to select target fish species and account for their different life-history stages, and, furthermore, consider shifts in habitat preferences among the different flow seasons therefore, providing a sound basis for future environmental flow studies in other Mediterranean-type streams.

KEYWORDS: Environmental flows, River2D, Mediterranean-rivers, IFIM

1. INTRODUCTION

Human impairment with freshwater ecosystems has increased in the last few decades severely affecting their natural physical characteristics and biological dynamics, consequently undermining their productivity and resilience. Nowadays, almost no stream or river worldwide has been left in its natural state, especially in Europe and North America. Dams, among other things, are widely known to alter rivers physical, chemical, and biological processes. Bank erosion, channel degradation, width adjustment, flattening slope, and bed armouring with coarse material are the typical downstream morphological alterations, which result in a general loss in morphological diversity [1]. Furthermore, seasonal reductions in flows, wetted perimeter and water quality during late summer when releases are minimized are due to occur. This tends to be exacerbated in Mediterranean regions due to their high hydrologic variability, also aggravated by high water demands.

The solution to overcome these issues involves identifying the quantity, timing, and variability of flow required to maintain desired levels of population biomass and biotic diversity [2], currently named of environmental flows. There are many methods available to assess the environmental flows [3]. Most common are hydrological methods that simply allocate water based on fixed percentages of historical natural flow regime [4]. Moreover, there are hydraulic methods based on the wetted perimeter of the river channel. Other frequently used methods, named habitat models, link the physical habitat preferences of target species or life-stages, established empirically or by expert opinion, with detailed hydrodynamic models that simulate the habitat availability as it varies across discharge [5, 6]. These methods often involve an assessment of changes in habitat quantity with stepwise changes in discharge under the assumption that habitat quantity is a limiting factor for aquatic species [7, 8]. Such methods attempt to preserve the ecological processes necessary for population viability by maintaining a baseline level of suitable habitat.

The Instream Flow Incremental Methodology (IFIM) is an interdisciplinary framework that can be used in a holistic way to determine an appropriate flow regime by considering the effects of flow changes on physical habitat, river morphology, water temperature, water quality, and sediment processes [9]. The use of IFIM requires a high degree of knowledge about seasonal and life-stage requirements of species. To determine the physical habitat changes due to flow, the IFIM applies one or two-dimensional hydrodynamic models. However, for studies examining habitat alterations, 1D models have limitations. Specifically, they are unable to predict the spatial variation of velocities [10, 11]. Growing evidence suggests that these spatial flow patterns play a vital role in determining the types and quality of habitat available within a stream. According to Crowder and Diplas [12] 2D approaches are required for realistic simulation of flow patterns and to better represent the diversity of physical habitats. These hydrodynamic models are usually coupled with a biological model of habitat selection, the Habitat Suitability Criteria (HSC) [13]. The classic approach of the HSC consists in converting structural and hydraulic characteristics of streams into indices of habitat quality, the Habitat Suitability Index (HSI). The HSI is an analytical tool used to represent preferences of different aquatic species for various instream habitat variables, provided that they are related to the hydraulic or the physical characteristics of the stream (i.e. depth, velocity, substrate and/or cover) [14]. Fish habitat use markedly changes among species and life-stages [15]. Knowledge that large fishes (i.e. adults) prefer deep-sheltered habitats than smaller con-specifics is well-known [e.g. 16]. Other factors such as competition, predation and seasonal time scale also affect the habitat use [17, 18]. Therefore, while defining the environmental flows one has to account to species life-stages and the seasonal time scale, once they are easily quantified in habitat use.

The present study aimed to determine the habitat availability considering the fish preferences of the endangered species

American Journal of Hydropower, Water and Environment Systems, aug 2014, page 6-11 7

south-western arched-mouth nase Iberochondrostoma almacai (hereafter nase) and Arade chub Squalius aradensis (hereafter chub), accounting for their life-stage and shift preferences in the dry and flowing season of a Mediterranean river, in order to establish the environmental flows downstream the Odelouca dam.

2. MATERIAL AND METHODS

2.1 Study Area

The study for setting the environmental flows was conducted in the Odelouca River, the largest tributary of the Arade basin (987 km2) in southwest Portugal (Figure 1). This Mediterranean river is medium-sized low-gradient with 92 km long. River topography varies from narrow steep sided valley walls to restricted meander valleys and small floodplains in the lower reaches and is dominated by schistose rocks, which are covered with alluvial deposits in the lower reaches.

The climate is typically Mediterranean, with large intra-annual variability. An annual rainfall following a seasonal pattern (wet season from October to March, dry season from June to August), resulting in a relatively slow running river, subject to high discharge peaks during the winter and leaving temporarily unconnected pools in the river bed during summer.

Nase (Iberochondrostoma almacai) and chub (Squalius aradensis) used to be the dominant species on the whole catchment [19], but are now almost absent from the lower part as a result the of human-induced habitat modification. The Odelouca dam is located just 8 km upstream the confluence with Arade River. The dam was constructed in order to respond to the water demand for human services and agriculture.

[Figure 1: Map of the study area, showing the location of the Odelouca River and the study site.]

2.2 Data collection

Downstream the Odelouca dam, a 250 m long reach representative of the river segment was selected to be modeled. Catchment area at the study site is 466 km2 with a mean annual flow of 4.05 m3/s. River width range from 7 to 30 m wide, with a 0.0035 m/m mean slope. The river bed topographic was surveyed with a combined association of a Nikon DTM310 total station with a Global Positioning System (GPS) (Ashtechy, model Pro Mark2). Additionally, bed elevation and substrate composition, measured according to a modified Wentworth scale [20] [(1) organic cover; (2) silt (1-2 mm); (3) sand (2-5 mm); (4) gravel (5-25 mm); (5) pebble (25-50 mm); (6) cobble (50-150 mm); (7) boulder (>150 mm) and (8) bedrock], were collected. Altogether, 4129 spots (x, y and z) were surveyed. To calibrate the model, a series of points were located along cross-sections where significant alteration in the physical habitat was due to occur. At each cross-section, depths were measured with a ruler and water velocities

were measured with a water flow probe (model FP101, Global Water Instrumentation, USA) at 60% of the distance from the water surface to the riverbed [21]. The boundary conditions were established by measuring discharge and water surface elevations at the downstream cross-section. Discharge was assumed to be constant as no tributaries joined the study reach. According to observations of bed material and bedform size, different bed roughnesses heights were estimated. The final values were obtained by calibration of the model results with the measured water surface elevations and the water mean velocity in vertical profiles along different cross-sections.

Habitat Suitability Curves (HSC) of depth, velocity and substrate were previously developed for specific fish size-classes based on reported differences in length and age structure [22], corresponding to fish life-stages juveniles (1+) and adults (>1+), respectively. Sampling took place at undisturbed or minimally disturbed sites of the Odelouca basin, so that habitat associations reflected the optimal species habitat rather than an externally imposed displacement towards sub-optimal situations [23]. Fish sampling was performed during the flowing season, which typically runs from mid-October to early July, when there is full connectivity between habitats (riffles, runs, pools) and fishes are therefore not restricted to isolated pools. Samplings were conducted by electrofishing throughout the year, encompassing both the dry (i.e. Jul-Sep) and wet seasons (i.e. Nov-Mar). Further details about site locations, sampling procedures and microhabitat measurements are given in Santos and Ferreira [22].

2.3 Data Analysis

The Instream Analysis Flow Incremental Methodology (IFIM) was selected for setting the environmental flows coupled with a two-dimensional model. These models have been shown to give a better representation of flow compared to one-dimensional models and are also able to simulate complex flow patterns, such as recirculation and transverse water surface slopes [24]. Specifically, the River2D model [25], a 2D hydrodynamic and fish habitat model, was used. This finite element model simulates hydrodynamic conditions and uses the habitat suitability index curves containing known biological preference data, to calculate the potential habitat for specific species life-history stages by the Weighted Usable Area (WUA).

The WUA, i.e. the surface (m2/km) that can be potentially used by a given fish life-stage, was then computed as the product of depth, velocity and substrate suitability indexes. Simulations were carried out with River2D for discharges up to 12 m3/s to ensure coverage of common flow discharges. The natural flow regime was computed using the daily flow data from a nearby gauging station (Monte dos Pachecos). The station is situated only 5 km upstream from the degraded site, and daily flow data, which were available for a period of 29 years (between 1962-2000 with a few years gap), were considered to be representative of the local conditions. The daily flow data for the studied reach was determined by multiplying the daily flow data available for the gauging station by an adjusting factor corresponding to the quotient between the study reach catchment area and the gauging station catchment area. The natural flow regime was then calculated considering the mean daily flow.

The environmental flows were calculated considering the WUA versus discharge curves and the natural flow regime. Three different scenarios of environmental flows were considered. These scenarios were designed considering (i) natural flow regime variability; (ii) maximum habitat availability in the different seasons (dry and wet season); and (iii) discharges equal or lower than the one expected in natural flow conditions, this is especially


important in the summer season. The mean monthly discharge of scenario A, B and C correspond to 30%, 5% and 10% of the mean monthly discharge of the natural flow regime. The WUA of the different scenarios was than compared with the WUA of the natural flow regime through XY - scatter plots

3. RESULTS

Discharged measured at the study reach was 2.05 m3/s. Figure 2 shows variations in WUA according to flow discharge for both species nase and chub and their life-stages, for the wet season (Figure 2a) and the dry season (Figure 2b) considered in HSC development. Differences were found to occur among seasons. In the wet season the species tend to have a habitat increase within the range of 0 to 1.6 m3/s, with a maximum at 0.8 to 1.0 m3/s for nase, and around 1.6 m3/s for chub. In the dry season differences among species and life-stages are more noticeable. Accordingly, for adult chub the habitat tends to increase within the entire studied period. For juveniles and nase, the trend is the same registered in the wet season. Maximum habitat was registered for juvenile chub and nase around 0.6 m3/s.

[Figure 2: WUA for the adult and juvenile life-stages of nase and chub at different flow discharges considering the (a) wet season and (b) dry season.]

The mean monthly discharges corresponding to the natural flow regime for the study reach is shown in Figure 3. To this flow regime corresponds a mean annual discharge of 4.05 m3/s. Table

1 shows the three different scenarios of environmental flows and the natural flow regime. Mean annual discharges of scenarios A, B and C are 2.0, 1.1, and 1.3 m3/s, respectively. In the dry season, the environmental flows are very low, corresponding to <0.02 m3/s (July to September).

[Table 1]: Mean monthly discharge (m3/s) at the natural flow regime and at the different flow scenarios.

Oct Nov Dec Jan Feb Mar Apr

Natural Flow Regime

1.6 8.0 12.1 11.1 11.6 6.2 3.0

Scenario A 0.7 3.4 5.1 4.8 4.9 2.6 1.3

Scenario B 0.7 1.9 2.8 2.6 2.6 1.4 0.8

Scenario C 0.5 2.5 3.8 3.5 2.6 1.4 0.8

May Jun Jul Aug Sep MMD(1)

Natural Flow Regime

0.9 0.2 0.0 0.0 0.0 4.1

Scenario A 0.4 0.2 0.0 0.0 0.0 2.0

Scenario B 0.7 0.2 0.0 0.0 0.0 1.1

Scenario C 0.5 0.2 0.0 0.0 0.0 1.3

(1) Mean annual discharge

[Figure 3: Mean monthly discharges corresponding to the natural flow regime.]

The habitat availability from these scenarios is shown in Figure 4. All the scenarios show higher habitat availability than the natural flow regime, except for May, where scenario A exhibits a smaller value of WUA compared to the natural flow regime. Overall, scenario B has the highest habitat availability, especially, in the wet season (i.e. from October to May). In the dry season, WUA tends to be the same among the different scenarios and the natural flow regime due to the lowest discharge that as to be equal to the natural flow regime.

Scatter plots of natural flow regime versus WUA along the year for the different scenarios are shown in Figure 5, considering the habitat availability for nase species (Figure 5a) and chub species (Figure 5b), and the habitat availability for juveniles (Figure 5c) and adults (Figure 5d). Overall, the results reveal a larger degree of scatter for the scenario B when compared with the others. The best adjustment was registered for chub (scenario A: R2=0.98; scenario B: R2=0.96; scenario C: R2=0.97). Among life-stages, the WUA for the juveniles (scenario A: R2=0.95; scenario B: R2=0.91; scenario C: R2=0.91) adjusts better to the WUA of the natural flow regime than the WUA for the adults (scenario A: R2=0.88; scenario B: R2=0.73; scenario C: R2=0.75).


[Figure 4: WUA (m2/km) at the natural flow regime and at the environmental flow scenarios for the nase and chub species.]

[Figure 5: Scatter plots of WUA (m2/km) at the natural flow regime versus WUA at the different environmental flow scenarios considering: (a) the nase species, (b) the chub species, (c) the juveniles from both species, and (d) the adults from both species.]

4. DISCUSSION

A two-dimensional modeling approach was applied to predict the habitat availability for endangered cyprinid species considering three different environmental flow scenarios. The results from this study highlight the importance of modeling the habitat when choosing the environmental flows. Modeling enables us to identify the environmental flow regime that better adjusts to the river reach and to the target species.

Modeling the habitat availability for the different environmental flow scenarios yielded distinct results. Chub was found to have the higher habitat availability across the range of discharges. This species was found to be a eurytopic species (habitat generalist) that inhabits a broad range of habitats [22]. Therefore, chub

would not restrict themselves to inshore shallow bays but would rather move alternatively between the bays and the free-flowing main river channel, therefore adapting better to different flow conditions. Moreover, significant size-related differences in the species habitat availability where found to occur. Juveniles tend to have higher habitat availability except for the adult chub in the dry season. Our findings support the hypothesis that during this season fish become increasingly confined to a reduced space, therefore, increasing competition for limited resources such as food and deep-sheltered areas. Depth has long been recognized to be an important factor that overrides fish assemblage structure in Mediterranean rivers [26]. The use by larger individuals of deeper areas in the dry season may reflect this shortage of water surface.

As aforementioned and considering other studies [22, 26], species do act differently in the dry and wet season. In the dry season shifts in habitat use are expected. These are closely related with habitat availability during this season, which tend to be scarce, thereby resulting in an increase in competition for space and resources [27]. In Mediterranean-type streams, the river can be reduced to a series of pools forcing fish to adjust to the new conditions [28]. Habitat studies should reflect seasonal-related shifts to habitat use in cases where this is notorious. Furthermore, habitat associations addressed in those conditions should not be considered to reflect the optimal habitat [23], since they represent an externally imposed displacement towards sub-optimal habitat conditions. However it should be noted that differences in habitat use may also be associated to other factors such as time of the day, temperature, food availability, presence of competitors and river discharge regime [29].

In natural, non-regulated streams, the biota varies in abundance and composition over time and space. The best way to judge the different flow scenarios is to maintain or recreate the hydrologic, hydraulic and morphologic conditions under which fish communities had existed prior to disturbance in order to meet ecological needs [30]. The habitat availability for the target species in the natural, non-regulated stream will act as the reference condition (guiding image) for comparing the degree to which a flow scenario deviates from the natural flow regime. The closer the environmental flow scenario is from the reference condition the “healthier” that flow scenario is judged to be. Conversely, the further from the reference condition, the less healthy. A logical consequence of this is to assume that under natural conditions habitat utilization is optimum [31]. Thus, the hydrologic circumstances that match the habitat availability for the native fish community found for the natural flow regime can be considered a rough estimate of the desired flow scenario.

The river flows in a Mediterranean climate vary greatly both seasonally and inter-annually. The conditions in summer impose particularly restrictive constraints on the survival of fishes [28], which normally occupy zones of higher depths [32]. Often, only part of a stream will go dry and affected reaches will be repopulated rapidly from remaining pools or upstream tributaries, thus supporting a unique biological community. In comparison with perennial streams (i.e. flow year-round), Mediterranean ones have a lower species richness. As a result, native species from the Iberian Peninsula have evolved and adapted to harsh hydrologic conditions, dictated by this high inter annual variability, over a long period of pre-development history. Therefore, we can assume that native fish fauna are more adapted to spatially and temporally variable habitats and such environmental variability is important to the long-term survival of fish stocks. Considering this there is a trade-mark to improve fish habitat and the population success. Goals should be


set to improve the fish habitat, bearing in mind the replication of such harsh conditions, even if this implies periods of almost no habitat for native fish species. Scenario B was found to reproduce the highest habitat availability along the year but on the other hand it was the scenario that reveals the greater discrepancy between the habitat availability of the flow scenario and the habitat availability under the natural flow regime.

Moreover, when a particular habitat alteration is imposed, one should predict what the outcome will be. Furthermore, hydrodynamic modeling is useful in extending the range of field observation either by incorporating different flows to evaluate different scenarios. Indeed quantitative predictions provide an important reference point for detection and interpretation of unexpected outcomes. This, in turn, provides outputs against which scenarios might be assessed after implementation. Accounting to this is easy to accept that the use of hydrodynamic modeling is a useful tool to support the ecological basis for defining the environmental flows.

This study is essential not only for an improved understanding of habitat requirements for fish fauna, but also for setting environmental flow recommendations elsewhere, in other typical Mediterranean-type streams, where knowledge of the specific habitat requirements of other cyprinid species still remains largely understudied. Nevertheless, scientists, managers and water users should be aware that species’ ability to respond positively to river restoration schemes will also depends on whether water quality in the new scenario is adequate to support them.

5. ACKNOWLEDGEMENTS

The authors would like to thank José Lourenço for assistance with the fieldwork and Paulo Pinheiro for helping to draw the study map. Isabel Boavida was supported by a grant (SFRH/BD/35801/2007) from FCT (Science and Technology Foundation).

6. REFERENCES

[1] Grant, G.E., Schmidt, J.C., Lewis, S.L., 2003, A geological framework for interpreting downstream effects of dams on rivers, in A Unique River: pp. 209-225.

[2] Poff, N.L., Allan, J.D., Bain, M.B., et al., 1997, “The natural flow regime” BioScience, 47, pp. 769–784.

[3] Tharme, R.E., 2003, “A global perspective on environmental flow assessment: emerging trends in the development and application of environmental flow methodologies for rivers”, River Research and Applications, 19, pp. 397–441.

[4] Gordon, N.D., McMahon, T.A., Finlayson, B.L., 1992, Stream hydrology : an introduction for ecologists, John Wiley & Sons, Chichester, 526 pp.

[5] Hardy, T.B., 1998, “The future of habitat modeling and instream flow assessment techniques”, Regulated River, 14, pp. 405–20.

[6] Rosenfeld, J., 2003, “Assessing the habitat requirements of stream fishes: an overview and evaluation of different approaches”, Transactions of the American Fisheries Society, 132, pp. 953–968.

[7] Rosenfeld, J.S., Hatfield, T., 2006, “Information needs for assessing critical habitat of freshwater fish”, Canadian Journal of Fisheries and Aquatic Sciences, 63, pp. 683-698.

[8] Vehanen, T., Sutela, T., Korhonen, H., 2010, “Environmental assessment of boreal rivers using fish data – a contribution to the Water Framework Directive”, Fisheries Management and Ecology, 17, pp. 165–175.

[9] Stalnaker, C., Lamb, B.L., Henriksen, J., Bovee, K.D., Bartholow, J.M., 1995, The instream flow incremental

methodology: a primer for IFIM, Biological Report 29, National Biological Service, US Department of the Interior, 45 pp.

[10] Leclerc, M., Boudreault, A., Bechara, J.A., Corfa, G., 1995, “Two-dimensional hydrodynamic modeling: a neglected tool in the instream flow incremental methodology”, Transactions of the American Fisheries Society, 124, pp. 645–662.

[11] Pasternack, G.B., Wang, C.L., Merz, J.E., 2004, “Application of a 2D hydrodynamic model to design of reach-scale spawning gravel replenishment on the Mokelumne River, California”, River Research and Application, 20, pp. 205–225.

[12] Crowder, D.W., Diplas, P., 2000, “Using two-dimensional hydrodynamic models at scales of ecological importance”, Journal of Hydrology, 230, pp. 172–191.

[13] Bovee, K.D., 1982, A guide to stream habitat analysis using the instream flow incremental methodology, Instream Flow Information Paper 12, USDI Fish and Wildlife Service. Washington, 248 pp.

[14] Ahmadi-Nedushan, B., St-Hilaire, A., Bérubé, M., Robichaud, É., Thiémonge, N., Bobée, B., 2006, “A review of statistical methods for the evaluation of aquatic habitat suitability for instream flow assessment”, River Research and Applications, 22, pp. 503–523.

[15] Copp, G.H., 1992, “Comparative microhabitat use of cyprinid larvae and juveniles in a lotic floodplain channel”, Environmental Biology of Fishes, 33, pp. 181–193

[16] Lamouroux, N., Olivier, J.M., Persat, H., Pouilly, M., Souchon, Y., Statzner, B., 1999, “Predicting community characteristics from habitat conditions: fluvial fish and hydraulics”, Freshwater Biology, 42, pp. 275–299.

[17] Power, M.E., 1984, “Depth distributions of armored catfish: predator-induced resource avoidance?” Ecology, 65, pp. 523–528.

[18] Orth, D.J., 1995, “Food web influences on fish population responses to instream flow”, Bulletin Français de Pêche et de Pisciculture 337/338/339: pp. 317–328.

[19] Pires, A.M., Costa, L.M., Alves M.J., Coelho M.M., 2004, “Fish assemblage structure across the Arade basin”, Cybium, 28, pp. 357–365.

[20] Bovee, K.D., 1986, Development and Evaluation of Habitat Suitability Criteria for Use in the Instream Flow Incremental Methodology, U.S. Fish and Wildlife Service, Washington, D.C., Instream Flow Information Paper No. 21., Biological Report 86(7), 235 pp.

[21] Bovee, K.D., Milhous, R.T., 1978, Hydraulic simulation in instream flow studies: theory and technique, Instream Flow Paper 5, FWS/OBS-78/33, US Fish and Wildlife Service, Fort Collins: CO.

[22] Santos, J.M., Ferreira, M.T., 2008, “Microhabitat use by endangered Iberian cyprinids nase Iberochondrostoma almacai and chub Squalius aradensis”, Aquatic Science, 70, pp. 272-281.

[23] Gorman, O.T., Karr, J.R., 1978, “Habitat structure and stream fish communities”, Ecology, 59, pp. 507–515.

[24] Ghanem, A., Steffler, P., Hicks, F., Katopodis, C., 1996, “Two dimensional hydraulic simulation of physical habitat conditions in flowing streams”. Regulated Rivers: Resources and Management, 12, pp. 185–200.

[25] Steffler, P., 2000, Software River2D. Two Dimensional Depth Averaged Finite Element Hydrodynamic Model, University of Alberta, Canada.

[26] Godinho, F.N., Ferreira, M.T., Cortes, R.V., 1997, “Composition and spatial organisation of fish assemblages in the lower Guadiana basin, southern Iberia”, Ecology of Freshwater Fish, 6, pp. 134–143.


[27] Schlosser, I.J., 1990, “Environmental variation, life history attributes, and community structure in stream fishes: implications for environmental management and assessment”, Environmental Management, 14, pp. 621–628.

[28] Moyle, P.B., 1995, “Conservation of native freshwater fishes in the Mediterranean-type climate of California, USA: a review”, Biological Conservation, 72, pp. 271—279.

[29] Copp, G.H., Vilizzi, L., 2004, “Spatial and ontogenetic variability in the microhabitat use of stream-dwelling spined loach (Cobitis taenia) and stone loach (Barbatula barbatula)”,

Journal of Applied Ichthyology, 20, pp. 440-451. [30] Karr, J.R., 1981, “Assessment of biotic integrity using fish

communities”, Fisheries, 6, pp. 21–27. [31] Parasiewicz, P., 2007, “Using mesohabsim to develop

reference habitat template and ecological management scenarios”, River Research and Application, 23, pp. 924–932.

[32] Facey, D.E., Grossman, G.D., 1992, “The relationship between water velocity, energy costs and microhabitat utilization on four North American stream fishes”. Hydrobiologia, 239, pp. 1-6.


methodology for the uSe of radioactive tracerS in hydroSedimentological StudieS

1Jefferson Vianna Bandeira, 2Lécio Hannas Salim, 3Cláudio José Chagas, 4Duarte Augusto Costa, 5Namir Souza Vieira, 6Vinícius Verna Magalhães Ferreira

1B.Sc. in Mechanical Engineering - UFMG (Federal University of Minas Gerais -1968), M.Sc. in Nuclear Sciences and Techniques - UFMG (1972), M.Sc. in Applied Hydrology by the UFRGS (Federal University of Rio Grande do Sul - 1974), Diploma in Hydraulic Engineering: Branch Costal & Offshore Engineering by The International Institute For Hydraulic and Environmental Engineering (IHE) – Delft, The Netherlands (1986) and D.Sc. in Sanitation, Environment & Water Resources - UFMG (2004). Presently is a CDTN Senior Researcher – [email protected]; Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.2B.Sc. in Electrical Engineering - UFMG (1977), M.Sc. in Hydro Resources and Environmental Sanitation - UFRGS (1998). Presently is a CDTN Senior Researcher – [email protected]; Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.3B.Sc. in Geography - UFMG (2011), presently working at the CDTN – [email protected]; Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.4Chemical technician, presently working at the CDTN – [email protected]; Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.5Electrical and mechanical technician, presently working at the CDTN – [email protected]; Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.6B.Sc. in Electrical Engineering - UFMG (1994), M.Sc. in Nuclear Sciences and Techniques - UFMG (1998), D.Sc. in Sanitation, Environment & Water Resources - UFMG (2004). Presently is a CDTN Senior Technologist – [email protected] (corresponding author). Centre for the Development of Nuclear Technology (CDTN) – Av. Antônio Carlos 6627 – Campus UFMG – 31270-901, Belo Horizonte-MG – Brazil. Tel. +55-31-30693127, Fax +55-31-30693174.

ABSTRACT

The use of radioactive tracers in hydrological and sedimentological studies is a technique applied in a restricted way in Brazil, due to the need for specialized labor and the legal limitations regarding the use of radioactive material. However, some studies have been done using technetium 99 as a tracer, and satisfactory results were obtained in surveys conducted in Brazil and abroad. The objective of this paper is to present the methodology for use of this radioisotope tracer in hydrosedimentological studies, in particular of one study done that aimed to verify the dynamic of the water and of the fine sediment. The results of several works done in the past demonstrate the feasibility of using this technique, which allowed its use in other researches in progress.

KEYWORDS: hydrology, sedimentology, tracers, technetium.

1. INTRODUCTION

Issues related to the transport of sediments are included in a large area of environmental studies. In Brazil, a country that possesses vast water resources, this line of research is particularly much relevant.

In places of high declivity, the rivers run relatively fast and generally contain a significant load of suspended sediments. However, when the velocity of the water decreases, the particles of the sediments tend to deposit more easily. Consequently, when the flow reaches a reservoir, the suspended load is deposited, and rarely is resuspended. Thin sediment layers provide habitat suitable only for certain types of invertebrates, particularly in the deeper and darker areas, where the sunlight does not penetrate, precluding the growth of algae and aquatic plants. Additionally, the slow decomposition of organic sediment and other wastes by bacterias, which use the available oxygen in the water, creates anaerobic conditions near the bottom [1].

Over the years, the river beds in places where the water speed is low, slowly begins to fill with sediments. As an example, it was estimated that Lake Powell, located in the U.S. State of Utah, which is among the largest artificial lakes in the world, will be filled with sediments in about 500 years [2]. This obviously has long-term implications with regard to power generation and water supply, among many others. Also, this creates questions about how to solve this problem, and what to do with the sediments that accumulate in the bottom of the reservoirs.

Since the sediments settle at the bottom of lakes, they cannot reach the areas downstream the dams. Ibanez et al. [3] report that after the construction of a dam on the River Ebro, sediment transport decreased by 99%, leading to a higher rate of erosion downstream because the river did not carry sediments

that compensate the erosion caused by the action of water. The U.S. Department of the Interior acknowledged that the erosion of estuarine areas in Washington State is caused due to accumulated sediments in lakes Aldwell and Mills, both located in the Elwha River [4].

The dams intercept the movement of sediment downstream the riverbed, however, this movement continues to occur after the dams. Without the refueling of the upstream sources, there is an increase in the average size of the materials in riverbeds [5]. This fact has important implications for the fish that usually spawn in the main river basin, and for those forced to spawn below the dam, because the substrates can become very rough for successive spawnings. As example, an increase in the size of surface sediments downstream the dam of the River Tyne, UK was observed [5]. These findings may explain that the downstream siltation is possibly caused by the alteration of the normal speed of the rivers. It is suspected, also, that roughness of the substrates and the sedimentation rate growth may explain the decreased activity of spawning below the dam [6].

The deposition of sediments in lakes becomes more serious question when it generates the accumulation of toxins such as heavy metals, polychlorinated biphenols, pesticides or insecticides, especially adsorbed to the fine sediments. This fact has important implications for future uses that may be given to the sediments, as they become poisonous and potentially dangerous for some applications. The contaminants in sediments may persist for decades or centuries, and bring ecological consequences for the entire food chain [7]. High levels of toxic compounds in the tissues of animals bring serious problems to their health and reproduction. The bioaccumulation of toxins through the food chain generally causes more harm to the animals that are on the


top of the chain, such as whales, dolphins, swordfish or raptors. An example of this fact was caused by the insecticide DDT (Dichloro Diphenyl Trichloroethane) which made the eggshells of some birds of prey very fragile, precluding the survival of the cubs. The birds were affected because DDT was present in tissues from animals killed by them [1].

2. OBJECTIVE

The objective of this paper is to present a methodology that allows the use of radioactive tracers in hydrosedimentological studies. Some work successfully done in the last years by technicians of CDTN/CNEN (Center for Development of Nuclear Technology/ National Commission of Nuclear Energy) in Brazil and abroad prove the effectiveness of these techniques.

3. CURRENT AND PAST WORKS

As example of the several hydrological and sedimentological studies already conducted, and other ones that are being executed by the CDTN team using radioactive tracers, since the late 1990s, it is valid to mention, among others, the following researches:

• study of the behavior of fine sediment suspended in the basin of Montevideo - Uruguay, under the scope of a project supported by the IAEA - International Atomic Energy Agency. The goal of this technical cooperation was the acquisition of parameters to be used in the calibration of a mathematical model [8];

• environmental studies aiming to dump the dredging of fine sediment that comes from the Pampulha dam (Belo Horizonte - Brazil), in waterways downstream, due to the serious sedimentation issues related to this dam [9];

• study of the dynamics of fine sediment in suspension and of the water that transports it (figure 1), to evaluate the influence of the dead zones in a sector of the Serra Azul creek, in the spread of mud in suspension (NEEBH project, supported by FAPEMIG - Foundation that Patronage Research Projects in Minas Gerais State). It is valid to observe that the region of this study is used as a hydrological laboratory by the CDTN technicians, and in the last 20 years many researches were done in the basin with the help of tracers. As example, in previous studies, some hydrological parameters were determined through the use of tritium as a tracer, like time of concentration, instantaneous unit hydrograph, infiltration rate, water balance and evapotranspiration values. Also, tritium was used to separate the surface and groundwater parcels of flood hydrographs, among other studies [10].

[Figure 1: Preparation of mud marked with radioactive tracer.]

• evaluation of the bottom and suspended sediments in the Orinoco River, Venezuela, using radioactive tracers, under the scope of a IAEA Technical Cooperation Project, related to dredging in the areas of Barrancas and Puerto Ordaz [11,12];

• assessment of the environmental impact (advection, dispersion and decantation rate of the dumped material) in watercourses downstream, of material removed through bottom discharge and / or dredging of the reservoir of the Paciência small hydro power plant, in the Paraibuna River (Figure 2), located in the Paraíba do Sul Basin (project GT 198, in execution and supported by CEMIG – Minas Gerais Energetic Company, and ANEEL - Brazilian Agency of Electrical Energy). It is valid to mention here that the implementation of this project is under the responsibility of the CDTN and CETEC (Technological Centre of Minas Gerais) technicians [13].

[Figure 2: Field work in the Paraibuna River.]

3.1 Radioactive tracer

The only tracer that represents the fine sediment, in a proper way to its detection in situ, is the same sediment to which a radioactive substance is added through the chemical process of adsorption. The fine sediment is generally electronegative and, if placed under the presence of cations or electro-colloids, the adsorption process occurs (the radioisotope is fixed to the sediment). If these cations or colloids are from radioactive elements with half-life and energy of the radiation appropriate emitted, a tracer of the fine sediment will be obtained. It is worth mentioning here that two conditions must be satisfied: the material adsorbed by the fine sediment cannot be released by it as long as the detection process goes on, and hydrodynamic behavior of marked and unmarked sediment has to be the same.

To study the dynamics of fine sediments in suspension, its marking was historically made with gold, in form of a radioactive isotope - 198Au [14]. However, this tracer is obtained only by direct irradiation, in a nuclear reactor, of its stable isotope 197Au, being necessary one irradiation for each experiment. Furthermore, due to the energy of the emitted radiation (410keV), it requires a relatively heavy lead shielding. Also, the chemical manipulation of the irradiated gold leaves (hot dissolution by acid attack) before the marking process requires much attention to be properly performed. These aspects limit, by far, the application of this tracer in many studies. Sometimes is necessary the air transport of irradiated material for its use in locations far away from nuclear reactors that produce it, what is also a complicating factor. Thus, the use of metastable technetium 99 becomes a very interesting option in hydrological studies.


3.2 Properties of the metastable technetium-99 and its use in nuclear medicine

The metastable technetium 99 (99mTc) is a radioactive tracer with short half life, that emits low energy gamma radiation. Its half life (time required for a quantity to fall to half its value as measured at the beginning of the time period) is 6.02 hours and its maximum energy is around 140 keV. 99mTc is derived from 99Mo (molybdenum) through a transition metastable – 99Mo→99mTc→99Tc [15]. 99mTc decays to technetium 99 (99Tc), a beta radiation emitter, which has little penetrating power (half life = 2.1 x 105 years; β = 290 keV).

The technetium generators (figure 3) are supplied to nuclear medicine laboratories, and due to the longer half life of the 99Mo, they can be used for up to two weeks in the extraction of 99mTc by a simple elution process, when the 99mTc is fully extracted from plate containing 99Mo and 99mTc. This elution consists in passing by the mentioned plate, a volume of 6 ml of a parenteral solution of NaCl (9 mg / ml), which is collected in a bottle which was initially had only vacuum. This electrolyte containing the radioisotope is the one injected into the vein of the patients. The aspect of portability, even in sedimentological applications, is a huge advantage over the use of 198Au, which needs to be irradiated each time a new experiment happens. Also, this aspect makes the technique accessible in countries that not possess nuclear reactors or that are not close to those ones that have.

[Figure 3: Technetium generator.]

After each elution, the generator continues to produce 99mTc, and in 24 hours, due to growth factor of the 99mTc, taking into account the radioactive decay of 99Mo, it is possible to extract an activity of about 88.5 % of that previously extracted. Thus, the generators are easily used by the laboratories of nuclear medicine, not requiring the irradiation of the material in a nuclear reactor for each exam.

99mTc is used to obtain mappings of various organs. As main applications in the medical field, it is valid to mention:

• scintilography of kidneys, cerebral, liver, lungs, placenta and bones;

• diagnosis of acute myocardial infarction;• circulatory system studies.

4. 4. METHODS

A field work done in November 2011 will be used as an example of the application method of 99mTc as a tracer in hydrosedimentological studies. This work aimed to study the dynamics of water and fine sediment using tracers in the Serra Azul creek. In the opportunity, the technetium 99 was injected as TcO4

- after its elution in a parenteral solution of NaCl (0.9%), to mark the water. After its reduction with stannous chloride

dihydrate (SnCl2.2H2O), to mark the mud, the TcO4- was eluted

from the Mo/Tc generator (volume = 6 ml). The procedure adopted follows explained below:

Step 1: first it is necessary to mark the mud, formerly removed from the same watershed where the studies will be conducted, in order to not introduce an external element to the existing natural scenario. The 99mTc must be reduced, which is obtained by the addition of stannous chloride dihydrate (SnCl2.2H2O) to the pertechnetate (TcO4

-) eluted from the generator. As result a colloidal electro-positive compound will be obtained, able to mark the mud, which is electro-negative (equation 1).

TcO2+ + 2H2O = TcO(OH)2 + 2H+ (equation 1).

Previous results (figure 4) showed that the stannous chloride, among others reducers studied, showed the best adsorption efficiency [16].

[Figure 4: Effect of the nature of the reducing agent on the adsorption of 99mTc in the fine sediment.]

Step 2: The mud was injected in the river cour se and its passage in the sector shown in Figure 5 was registered through the use of scintillometers. Another elution was perfo rmed in a second generator, and the TcO4

- was injected directly mixed with water, which being electro-negative, is not adsorbed by the muddy material that may exist in the creek, that in this step works as a tracer of water.

[Figure 5: Studied sector: Serra Azul creek Minas Gerais State – Brazil.]

Thus, based on the responses of the detectors and the analysis of the results, it will be possible to see the differences in the behavior of the two phases: water and mud in suspension, in the presence of dead zones, which was the scope of this particular study. The launch of the tracer, marking the water and / or the fine sediment, was performed in a drum surrounded by a blanket of lead with a thickness of 3.2 mm, enough to protect the operator. From the drum, the tracer was drained by gravity towards the creek. The total volume of the mixture injected


was 42 liters. The launch was done in the central section of the Serra Azul creek indicated as PC2 (Figure 5). The detection was performed by scintillometers immersed in water, to measure the passage of the radioactive cloud in some sections downstream the injection point. The results of this experiment are still under study, and will be published shortly.

It is emphasized here that in environmental studies, the required concentrations to develop hydrosedimentological stu-dies (Bq / ml of water) are much smaller, on the order of 10-7, than those ones used in medical applications (Bq / ml of blood). This occurs due to the geometry of counting in the environment, where the detector is immersed in water. However, in medical applications, the detector stays external to the patient [9]. Among other advantages, this reduces the involved costs and facilitates the public acceptance of the activity.

It is also necessary to emphasize the obligation to obey the Brazilian standard for road transport of radioactive sources [17], and the basic guidelines for radiological protection [18] for the execution of the work. A radiological protection technician should be included in the group that will execute the activities.

5. DISCUSSION

As stated before, the methodology explained above has been applied in several studies in the last decades. Small changes in the methods are needed due to the characteristics of the region and the goal of the study. As example, part of the study performed in the Pampulha dam was done to evaluate the viability of an environmental and perennial solution to dredge the fine sediment that accretes (400,000 m3/year) the reservoir. Previous researches showed that due to this reason, the reservoir can lose in the next years, two of the main purposes for which it was built: flood damping and leisure region. Experiments done in the dam, with simultaneous and instantaneous injections of sediment and water, labeled respectively with 99mTc and Rhodamine WT, were performed to measure the hydro transport capability downstream the dam, in a 25 km section. The experiment allowed comparing the different hydrodynamic behaviors of the mud in suspension and the water transporting it. One of the main conclusions of the studies done in the dam showed that there is no impediment for the dumping of the dredged fine sediments in the watercourses downstream.

In 2006, under the scope of a project named “Management of Sediments throughout the Navigation Channel of the Orinoco River”, 99mTc was used to evaluate bottom and suspended sediment transport in this watercourse. The study was able to determinate the following characteristics of the behavior of the fine sediment in suspension: advection, velocity, dispersion coefficient, sedimentation rate and dilution, considering that the fine sediment is the main carrier of heavy metals and other pollutants in the water environment. The sedimentation rate was determinate from the decrease of the maximum count rate with time. The maximum count rates obtained for each cloud that crossed the section where the probes were placed, were plotted in a logarithmic scale as a function of time. In this study, two sub superficial injections of mud labeled with 99mTC were performed. The initial activities used during the injections were respectively 2.1 and 1.6 Curies. The detection was performed by a boat positioned by GPS with scintillators detectors placed at 1.5 m (probe 1) and 0.5 m (probe 2) below the water surface. The measured count rates were corrected for the background and radioisotope decay. The results obtained for the behavior of the natural sediment in suspension, at the end of the low water season of the Orinoco River (April), could be used for preliminary

designs of outfalls for industrial effluents which will discharge particulate material with a density similar to the fine sediment or for pollutant material that can be adsorbed by the fine sediment, since the Orinoco River Basin region is undergoing a fast industrial development, with many industries being installed in the river margins.

6. CONCLUSIONS

Despite the great public rejection that exists in the society against the activities of the nuclear industry, this sector is present in many areas of human activity, including agriculture, medicine, engineering and hydrology, among others.

The use of radioactive tracers in several hydrological and sedimentological studies is a very useful tool, although not much explored, due to some reasons as the need of a very specialized manpower. Its use enables the realization of non-destructive researches, contributing to a better understanding of many questions associated to hydrology and sedimentology. This type of research is very relevant in Brazil, since many times the country’s researchers use in their works international reference values due to lack of researches done in the national territory that can provide the needed parameters.

7. REFERENCES

[1] Ferreira, V. V. M., 2004, “Evaluation of externalities in the hydro electrical sector in Minas Gerais State”, Ph. D. Thesis in Sanitation, Environment and Hydro Resources, UFMG Engineering School, Belo Horizonte, Minas Gerais, Brazil.

[2] Friends of Lake Powell, http://www.lakepowell.org, site accessed in 01/14/2011.

[3] Ibanez, C., Prat, N. & Canicio A., 1998, “Changes in the hydrology and sediment transport produced by large dams on the lower Ebro River and its estuary”, Regulated Rivers: Research & Management 12, pp 51-62.

[4] DOI – Department of the Interior, 1996, “Elwha River ecosystem restoration implementation. Final environmental impact statement”, Rep. No. NPS D-271A, USA.

[5] Sear, D.A., 1995, “Morphological and sedimentological changes in a gravel-bed river following 12 years of flow regulation for hydropower”, Regulated Rivers: Research & Management 10, pp 247-264.

[6] Dauble, D. D., Johnson, R. L. & Garcia, A. P., 1999, “Fall Chinook Salmon Spawning in the Tailraces of Lower Snake River Hydroelectric Projects”, Transactions of the American Fisheries Society 128, pp 672-679.

[7] Reinhold, J.O., Hendriks, A.J., Slager, L. K. & Ohm, M., 1999, “Transfer of Microcontaminants from Sediment to Chironomids, and the Risk for the Pond Bat Myotis Dasycneme (Chiroptera) Preying on them”, Aquatic Ecology 33, pp 363-376.

[8] Pinto, G.G. & Bandeira, J.V., 1998, “Estudios sedimentológicos en la Bahía de Montevideo, con el empleo de técnicas nucleares y trazadores radiactivos y fluorescentes”, Proc. III National Meeting of Sediments Engineering - ENES, Belo Horizonte – MG, Brazil.

[9] Bandeira, J.V., 2004, “Desenvolvimento de Técnicas Nucleares e Correlatas para Estudos em Hidrologia Urbana – Aplicações na Bacia Hidrográfica da Pampulha e no Rio das Velhas”, Ph. D. Thesis in Sanitation, Environment and Hydro Resources, UFMG Engineering School, Belo Horizonte, Minas Gerais, Brazil.

[10] Drumond, M.M., Rodrigues, P.C.H., Camargos, C.C. & Minardi, P.S.P., 2008, “Tracer techniques as a contribution for studying the hydrological behavior of a São Francisco River Sub


Basin” – Parts I, II and III, Proc. 5th International Conference on Tracers and Tracing Methods - Tracer 5. Tiradentes, MG – Brazil, pp 101-120.

[11] Bandeira, J.V., Salim, L.H. & Brisset, P., 2006, “Management of Sediments throughout the Navigation Channel of the Orinoco River – Project VEN/8/019-02-01”, End-of-mission Report to International Atomic Energy Agency - IAEA, Vienna, Austria.

[12] Machado, M.L., Bandeira, J.V., Salim, L.H. & Moreira, R.M, 2008, “Orinoco River suspended sediment studies using 99mTc - Venezuela”, Proc. 5th International Conference on Tracers and Tracing Methods - Tracer 5. Tiradentes, MG – Brazil, pp 83-91.

[13] Bandeira, J.V., Salim, L.H., Ferreira, V.V.M., Junqueira, M.V., Barbosa, G.H.S.P.C., Carvalho, M.D. & Mota, H.R., 2012, “Assoreamento de reservatórios, descargas de fundo e avaliação de impactos ambientais: caso estudo da PCH Paciência, Rio Paraibuna, MG”, Proc.10th National Meeting of Sediments Engineering (X ENES), Foz do Iguaçu – PR, Brazil – to be published (paper already approved).

[14] Bougault, H., 1970, “Étude de la sorption de quelques radioéléments artificiels par les sédiments pélitiques en vue de son application au marquage radioactif de ces matériaux”,Ph. D. thesis, Faculté des Sciences , l´Université de Paris, France.

[15] Sandler, M.P., Coleman, R. E., Patton, J.A., Wackers, F.J.T. & Gottschalk, A., 1996, Diagnostic Nuclear Medicine, Williams & Wilkins, 3 ed., Baltimore, USA.

[16] Bandeira, J.V., Salim, L.H., Sabino, C.V.S., Agudo, E.G., Aun, P.E. & Mendes, V.L., 2008, “99mTc Mud labelling and its application in hydrodynamic studies of fine suspended sediment in Brazil”, Proc. 5th International Conference on Tracers and Tracing Methods - Tracer 5, Tiradentes, MG – Brazil, pp 75-82.

[17] CNEN – National Commission of Nuclear Energy, 2011, Rule NN 3.01 – “Basic Guidelines of Radiological Protection”, http://www.cnen.gov.br/seguranca/normas/m ostra-norma.asp?op=301.

[18] CNEN – National Commission of Nuclear Energy, 1998, Rule NE 5.01 – “Radioactive Materials Transport”, http://www.cnen.gov.br/seguranca/normas/mostra-norma.asp?op=501.


computational modeling of fluid flow on encapSulated pipeline tranSport

1Gabriel de Carvalho Nascimento, 2Carlos Alexandre Bastos de Vasconcellos, 2Marcelo de Miranda Reis, 2José Carlos Cesar Amorim

1Federal Fluminense University - UFF, rua Passos da Pátria 156/464-D - São Domingos - Niterói/RJ - Brasil, contato: [email protected] Institute of Engineering - IME, Praça General Tibúrcio s/n - Urca - Rio de Janeiro/RJ - Brasil. The author graduated in Civil Engineering in 2007 and post graduated in Transport Engineering in 2013, both at Military Institute of Engineering (IME). Currently, he is professor of Fluid Mechanics at Federal Fluminense University (UFF) and works as consultant for computational modeling at several offshore oil and gas projects.

ABSTRACT

The overload of the transport system that increases even more due to the growing economic development in Brazil, in addition to the search for more economic and environmental friendly solutions demands research for enhanced technologies at this subject. One of the possible alternatives to meet the growing need for transport capacity is the encapsulated pipeline. As the main objective of this work it was done a computational modeling of the flow that occurs inside the pipe and its consequent interaction with the capsule in order to sustain the methodology that could points to the technical and economic viability. Such modeling aims to calculate the velocities produced by the fluid and possibly by the capsule at several different flows and design parameters. The results achieved from computer modeling are very close to those found in experimental and theoretical literature.

KEYWORDS: CFD, encapsulated pipeline, HCP

1. INTRODUCTION

In many ways the pipelines have the best rates among all transportation modes, including reliability, frequency, and low en-vironmental impact. The incident of annual causalities in pipeline operations is singular making it the safest transportation. The independence regarding weather and the nature of its behavior, demanding few interruptions and allowing it to be used 24 hours per day and 7 days per week, results a frequency of virtually 100% of the time. Finally, in the rating of an aspect which is gaining more and more importance for the transportation analysis, it is highlighted the low environmental impact of pipelines due to the negligible physical change of the environment during the installa-tion and an operation based on electrical pumps, resulting in the best energetic efficiency among all other modes of transportation.

Those positive factors create motivation to research methods that can allow the increase of transportable loads variability. That would make this mode of transportation very attractive even to the solution of the transportation infrastructure overload. Within this context, it is emphasized the encapsulated pipeline transport, that consist basically in propelling capsules by the flow of a fluid through the pipe. Such capsules can be the load itself if it’s possible to cast it in a shape that fits inside the pipe (e.g.: cylindrical or spherical) and the constitutive material of the capsule resists to the interaction with fluid and eventual impacts with internal surface. If isn’t, the capsules will be containers transporting the load, generalizing the type of load that can be transported. According to the type of fluid used for propelling the pipes will be classified as HCP (Hydraulic Capsule Pipeline) if it’s liquid (water) or PCP (Pneumatic Capsule Pipeline) if gas (air).

The first suggestion of HCP was in the 2nd World War by Jeffrey Pyke to transport military materials from China to Burma (Lumpe [3]), however due to the lack of developed technology it was not used at that time. The idea came back on Canada in 1959 and between 70’s and 80’s researches was initiated about this matter in USA, Japan and others countries (ASCE [1]).

The most know application of HCP is the CLP (Coal Log Pipeline). CLP is a capsule in cylindrical shape formed by

compressed coal, which is resistant to the water so needs no container capsule. Therefore, it’s not necessary to retrieve empty capsules to the source of the system, what virtually duplicates the transport capacity.

[Figure 1: CLP’s (left.) and the machine of compression (dir.).]

The CLP’s was developed and widely researched by CPRC (Capsule Pipeline Research Center) at the University of Missouri-Columbia. It can include a sealer layer of asphalt emulsion in addition to other techniques that decrease the loss of material and also the addition of polyethylene oxide in order to reduce the friction that causes loss of energy. Studies show that the application of CLP is feasible for distance between 50 and 2.000 km (ASCE [1]).

2. COMPUTATIONAL MODELING

2.1. Theory basis and methodology

In order to calculate the pressure distribution around the capsule it’s necessary to predict the behavior of the fluid (water) inside the pipe, which flow is classified as turbulent. The basis of theory applied was the Reynolds Averaged Navier-Stokes equations (RANS):

(1)

Source: Liu [4]


and

(2)

where u–i is the component of mean velocity on the xi, p– is the

mean pressure and “ν” the cinematic viscosity. The referred fluid can be considered incompressible so the density r0 is constant.

In order to calculate u'iu'j, referred as Reynolds tensor, it was adopted the SST (Shear Stress Transport) turbulence model, which accounts for the shear stress transport and considers an proportional rate between the Reynolds shear stress and turbulent kinetic energy.

The RANS equations doesn’t have an analytical solution for the referred physical problem, therefore a numerical solution is evaluated by the Finite Volume Method (FVM). This method consists in discretizing the domain into finite control volumes, creating a mesh, and to apply the transport equation by numerical integration for each one of them, according to the equation:

(3)

where φ is the physical quantity that will be integrated, and may be mass, momentum and energy. Δnj is the vector normal to the surface, Γφ is the diffusion term and S–φ is the mean value of the source along the control volume. The summations are done for each integration point (ip).

The application of the Equation (3), complemented by the Equations (1) and (2) in all elements of the mesh results in a linear system of equation which provides an approximated solution for the related flow.

The software used in this work was the Ansys CFX v14, which is able to apply all the aforementioned theory in order to derive a CFD solution. However, many other known programs are developed based on the same methodology and hence the choice of this software doesn’t restrict the results obtained.

3. FLUID-STRUCTURE INTERACTION

In order to be able to predict the capsule motion, the model must consider both capsule solid mechanics and fluid flow. At the solid domain (capsule), it’s necessary to include in calculations the inertia, gravity force as well as contact with the inner surface of the pipe. For those purposes, the Lagrangian approach is more appropriate, referring the coordinate system to the matter and having displacements as unknowns. However, at the fluid domain (water), the equations must represent the turbulent flow of the particles over the entire pipe interior, making it more recommended the Eulerian approach, which is referred to the space instead of matter and velocities are the unknowns.

Fluid-structure interaction is a relatively new area in numerical methods, therefore better techniques still under development in order to overcome the challenge of modeling both fluid and solid domains “simultaneously” in a more efficient way. Actually, most of the current methods basically consists of not solve both domains simultaneously.

Between the solid and fluid domains, there is the forces applied by the fluid and the displacement resulted by the capsule. To allow the interaction of such different domains, they are solved separately from each other. However, pressure and shear stresses are transmitted from fluid to solid and displacements are transmitted back. This procedure is repeated until a desired convergence is achieved.

Fluid

Eulerianapproach:

Velocities

Fluid- Structure

Interaction

Forces

Displacemen

Solid

Lagrangian approach:

Displacements

[Figure 2: Fluid-structure interaction scheme.]

The displacements calculated at the solid domain must be applied at the fluid domain. Since the Eulerian approach doesn’t account for displacements, an additional equation is considered to distribute the displacement from solid-fluid interface over the fluid mesh:

(Гdisp) = 0 (4)

This diffusion equation is evaluated based on Γdisp term, which is defined by larger values where should be smaller deformations of the mesh.

2.3. Model details and boundary conditions

The computational model was developed aiming to represent the same conditions of the experiment performed by Richards [5]. In this experiment, capsules with different dimensions (cases 1 to 7) were placed inside a pipe and the water flow was increased until the capsule start to move. Then the measured mean water velocity is called incipient velocity.

The pipe was made of Plexiglas and had an inner diameter of 54.0 mm. Two different materials were tested for the capsules: acrylic and coal. The geometry was based on capsule length (Lc) and diameter (Dc). The values for the friction were measured by the tilt angle applied to the pipe for which the capsule starts to move due to self-weight.

[Table 1]: Experimental data.

Case Capsule Material

Capsule Geometry Friction factor

Material a = Lc/Dc k = Dc/D MeanStandard deviation

1acrylic

d=1,176

2.667 0.884 0.404 0.0302

2 1.300 0.884 0.421 0.0523

3 4.073 0.695 0.453 0.0520

4 2.053 0.695 0.476 0.0378

5 coald=1,174

3.792 0.858 0.569 0.0262

6 1.861 0.858 0.513 0.0339

7 1.295 0.858 0.530 0.0317

Source: Richards [5].

As boundary conditions, a uniform velocity was applied at the inlet with 5% of turbulence and constant static pressure at the outlet. The inner surface of pipe was considered smooth as well as the capsule surface. Only one half of the pipe was modeled with a vertical symmetry plane containing the axis.


It was adopted hexagonal elements at the regions where the flow has a predominant velocity on axial direction and tetrahedral elements on the rest of the domain. To capture a satisfactory representation of the boundary layer, the mesh was refined at the proximities of the surrounding surfaces. The generated mesh has nearly 400.000 cells and can by observed on Figure 3.

[Figure 3: Mean monthly discharges corresponding to the natural flow regime.]

Because it wasn’t expected large stresses for the capsule and pipe a simple mesh could be applied to these bodies. The main concern about the solid domain is the inertia of the capsule and the contact (normal and friction force) with the pipe.

[Figure 4: Applied mesh for the pipe and capsule (solid domain).]

4. RESULTS AND DISCUSSION

In the Figure 5 are represented streamlines calculated in the case 4, whi ch are very similar in the other cases. There are two recirculation zones: one on the top of upstream end of the capsule and the other at the downstream. The first one is the most important because it causes a contraction of flow section further then capsule obstruction and a consequent head loss. This pressure reduction results in a pressure difference between upstream and downstream that is the main responsible for the capsule movement.

[Figure 5: Streamlines and recirculation zones (case 4).]

The simulation was performed increasing the inlet velocity until the capsule starts to move. The results obtained for each case are listed at Table 1 and compared with the experimental

data measured by Richards [5] and empirical results derived by Gao [2].

[Table 2]: Results and comparisons.

Case Incipient Velocity (m/s)

Measured 1 Empirical 2 Numeric Error (%)

1 0,066 0,058 0,053 -19,7

2 0,047 0,046 0,039 -17,0

3 0,224 0,182 0,200 -10,7

4 0,164 0,139 0,153 -6,7

5 0,092 0,097 0,095 3,3

6 0,066 0,071 0,069 4,5

7 0,053 0,061 0,059 11,3

Source: 1 Richards [5]. 2 Gao [2].

Most of the numeric results are very close to the empirical ones, however, it was found larger differences (up to nearly 20%) compared to the experimental values.

The larger differences in comparison to the experiment are concentrated on cases 1 to 4, where the capsule was made of acrylic. It can be observed on Table 1 that the friction values for acrylic capsules are significantly different despite of been expected the same value for the same materials. The imprecision of measured friction is confirmed by the relatively large standard deviations. In fact, bigger differences between numerical and experimental are noticed for the smaller values measured for the friction of acrylic capsules.

For the coal capsules (cases 5, 6 and 7), the mean difference between numeric and experimental values of incipient velocity was around 6% and the numerical results are even closer to the analytical.

According to the aforementioned results, the CFD analysis can be a very useful tool to predict the incipient velocity and many other operational information necessary to the design of HCP’s. The computational modeling can be extended to PCP application as well, when the fluid that is propelling the capsule is compressible.

Finally, it is concluded that the fluid structure interaction (FSI) applied at this work is a reliable method to be used in future researches of encapsulated pipeline transport, making it possible to analyze both solid and fluid domain simultaneously, according to the requirement of this study.

5. REFERENCES

[1] American Society of Civil Engineers (ASCE), 1998, “Fright pipelines: current status and anticipated future use”, Journal of transportation engineering, 124(4), pp. 3-8.

[2] Gao, X. and Liu, H., 2000, “Predicting Incipient Velocity of Capsules in Pipe”. Journal of Hydraulic Engineering, 126(6), pp. 1-3.

[3] Lump, D., 1959, “Pyke: The unknown genius”. Evans Brothers Ltd., London, England.

[4] Liu, H., 2006, Transportating Freight Containers by Pneumatic Capsule Pipeline (PCP): Port Security and Other Issues, U.S. Transportation Research Board (TRB) Meeting, California/EUA.

[5] Richards, J. L., 1992, Behavior of Coal Log Trains in Hydraulic Transport Through Pipe, University of Missouri-Columbia, Missouri-Columbia.


brazilian hydroelectric rehabilitation potential and viability

1Elisa de Podestá Gomes, 2Sérgio Valdir Bajay

1 Elisa de Podestá Gomes is a research engineer at the Linz hydraulic laboratory of Andritz Hydro GmbH, Austria. She graduated in mechanical engineering at the State University of Campinas (Unicamp) in Dec. 2003, in Brazil, where she also obtained a master degree on planning of energy systems (2013). Research engineer at Andritz Hydro GmbH, Austria. Lunzerstrasse 78, 4031, Linz, Austria. [email protected]. Phone: +43 732 6986 3061. 2 Sergio Valdir Bajay is a mechanical engineer (1973) with a MSc degree from Unicamp (1976) and a PhD degree from the University of Newcastle upon Tyne, England (1981). He is professor and senior research fellow at the State University of Campinas (Unicamp), Brazil. He is also a member of the Sao Paulo State Council for Energy Policy (CEPE) and president of the Brazilian Society for Energy Planning (SBPE). Professor and senior research fellow at the State University of Campinas (Unicamp), Brazil. NIPE/Unicamp, Cidade Universitária “Zeferino Vaz”, CEP: 13083-896, Campinas, SP, Brasil. [email protected]. Phone: +55 19 3521 3281.

ABSTRACT

The average annual growth of the Brazilian economy from 2003 to 2010 was 4.6%. In the same period, electricity consumption rose 36.5%, i.e. 5.2% per year on average, going from 306,987 GWh in 2003 to 419,016 GWh in 2010. Thus, to ensure the national supply of electricity, Brazil needs to increase substantially the installed capacity of the country’s power plants to meet this growing demand. This paper estimates the potential to increase the capacity of hydroelectric power plants in the country through the rehabilitation of such type of plant.

Hydro power plant rehabilitation is the process of increasing the plant’s installed capacity, the efficiency of the turbine-generator set, or the plant’s capacity factor. Adding new generation units in available spaces left over in the power house of operating power plants for this purpose is also a type of repowering and is considered in this paper.

KEYWORDS: Hydraulic power plant, rehabilitation, upgrade, Brazilian potential

1. INTRODUCTION

Hydroelectric power plants provide most of Brazil’s electrical energy supply. Many of these power plants were built during the seventies and eighties and have been running for more than 30 years. The country’s economy has been growing at an average rate of 4.6% from 2003 to 2010 [1], requiring, therefore, much more electricity to sustain such a growth; besides, there will be large international sporting events in Brazil in the coming years, requiring extra power.

Countries that have old hydraulic power plants have been carrying out rehabilitations projects regularly. Developing countries, such as Brazil, where several hydropower plants are now aging could also undertake this kind of project. The current available electrical power generating capacity in Brazil is 132,064.36 MW and 65% of this capacity, i.e. almost 80 GW, is installed in hydropower plants [2]. The share of the hydro power production is even higher; in 2011 it represented 89% [3].

Very common in North America and Europe, but still a novelty in Brazil, rehabilitation projects are a good opportunity for the country. They can increase the power plants’ life and reliability with no extra environmental problems. Furthermore, increasing the power stations’ capacity and/or efficiency is usually aimed at in this kind of project, providing more electrical energy to the country.

This paper presents a list of hydro power plants in Brazil that could go through a rehabilitation process. The current regulatory barriers to undertake this kind of project in the country are also discussed here, together with some proposals to overcome them.

2. REHABILITATION OF HYDRO POWER PLANTS

2.1 Main features of a hydro power station

A hydro power station uses river flows as its “fuel”. In this type of power plant, the potential energy provided by the river flows and the difference between the headwater levels and the tailwater levels (power plant head), minus the head losses in the

plant’s intake, is converted into mechanical energy by the plant’s turbines, which are connected to electric generators, where the mechanical energy is transformed into electrical energy that is delivered to the grid through transmission lines. Figure 1 depicts the scheme of a typical hydro power station with Francis medium head turbines.

Generically speaking, the river inflows go through the gate to the penstock and, from there, to the turbine’s spiral casing, stay vanes, guide vanes and runner, in this sequence. After producing mechanical power in the turbine runner, the stream flow reaches the station’s tailwater level through the turbine’s draft tube. The mechanical energy that rotates the shaft between the turbine and the electric generator is transformed into electricity in the latter equipment. The electric power flow from the generator is injected into the grid through transmission lines, after having the voltage level upgraded in a power transformer.

[Figure 1: Typical scheme of a hydro power station with a Francis medium head turbine.]

Some advantages of hydro power stations are: (i) they use a renewable energy source; (ii) they are reliable; (iii) their operation is flexible; and (iv) they are efficient, as they can convert about 90% of the hydraulic energy input into electricity, while fossil-fuelled power plants have their efficiencies in the range of 35% and 60% [4].


The power (P) produced by a hydraulic turbine can be calculated as:

P = ρ × g × H × Q × η (1)

whereρ = water density (kg/m³);g = gravity acceleration (m/s²);H = net head (m);Q = turbine discharge (m³/s); andη = turbine-generator efficiency.

So, according to equation 1, in order to produce more power in a hydro unit it is necessary to increase the head, the discharge or the efficiency. In the case of an operating unit, the easiest parameter to be changed is the efficiency. This is relatively easy to carry out with current technology and engineering knowledge, for power generating units that have been in operation for more than 30 years.

The electrical energy generated in a power plant does not depend only on the installed generating capacity, but also on the plant’s average capacity factor, which is the ratio between the energy produced in a specified time interval and the energy that would be generated by the plant’s installed capacity fully operated all the time. Thus, the electricity generated by the power plant during the period of a year, EG, in MWh, can be calculated by:

EG = P × FC × T (2)

whereP = power plant’s installed capacity;FC = plant’s average capacity factor during the year; andT = number of hours in the year.

2.2 Reasons to rehabilitate

A rehabilitation project usually contemplates: (i) increase in the plant’s reliability and availability; (ii) life extension and performance restoration; (iii) performance improvements; and (iv) reduction on operation and maintenance (O&M) costs.

According to the IEC 62256 [5], rehabilitation is defined as the combination of:

• Restoration of equipment capacity and/or equipment efficiency to near “as-new” levels;

• Extension of equipment life by re-establishing mechanical integrity;

• Increase capacity or efficiency beyond those of the original machine.

This process can be done in all kinds of power plants – hydroelectric, thermoelectric, nuclear, wind power stations, etc. However, in Brazil, as referred to before in this paper, hydro power is the dominant electricity source, with several plants over 30 years old [2]. The second largest source is natural gas, with an installed capacity of gas-fuelled power stations 6 times smaller than the total capacity of hydro power stations; most of the gas-fired power plants are relatively new.

Degradation of a power plant is a natural phenomenon after many years of operation. The same IEC code estimates an average life between 25 and 50 years old, depending on the project, construction and operation conditions [5]. Along the years, the generation units no longer perform as originally designed for and the O&M costs tend to rise substantially. At this point, a rehabilitation project can bring the performance of the power plant back to the original conditions, or even achieve better ones.

2.3 Rehabilitation types

When a power plant reaches an age that it no longer can operate as efficiently as before, maintenance and operating costs become too high and many unexpected outages occur due to some

failure. So, the owner has to take one of the following decisions: to close down the plant, to reconstruct it, or to rehabilitate it.

When the rehabilitation option is chosen, the main objective can be to extend the power plant’s life time. Besides that, to decrease the O&M costs and to increase the efficiency and/or the plant’s capacity are also very common goals.

Increasing the capacity of existing hydro power plants in Brazil can be achieved in two ways: rehabilitating old hydro power stations or adding new generating units in available spaces left over in their power houses.

The technical literature classifies rehabilitation (rehab) projects in three categories: minimum, light and strong rehabilitations. The minimum rehab is defined as when the turbine and generator are repaired to similar conditions as new ones and the performance is back to original values. The average capacity gain in this case is 2.5%. Light rehabilitation occurs when there is a wider check-up of the main components, with repairs and changes of some of them, in order to improve the power plant performance above the original values. A typical capacity gain with light rehab is 10%. Strong rehabilitation happens when the latest hydrological studies provide a new physical condition to the generation unit, and so its energy production is substantially increased. It is usual to change the turbine runner and other components in this case, as well as the main generator parts. The capacity gain can be between 20% and 30%; the Brazilian regulatory agency for the electric power sector – ANEEL - considers 23.3% as a reference gain.

3. CANDIDATE HYDRO PLANTS FOR REHABILITATION IN BRAZIL

Two criteria were used to select candidate hydro power stations for rehabilitation in Brazil: (i) plants older than 30 years and with generating units above 15 MW each; and (ii) operating plants with room available in their power houses for the installation of extra generating units.

3.1 Rehabilitation of old hydro power stations

Applying the first criterion referred to above, 43 power plants with 193 generating units were selected as candidates for rehabilitation. Table 1 lists some features of these power plants.

The total installed capacity of the generating units of Table 1 is 25,872.37 MW, which represents 19.60% of the total installed capacity of the Brazilian power system.

If all of the generating units listed in Table 1 would undertake a minimum rehab, 646.81 MW would be added to the country’s installed capacity. A light rehab of these units would add 2,587.24 MW, while a strong rehab would contribute with 6,028.26 MW for the Brazilian power generating capacity.

Regardless of the real value of new capacity that could be tapped from these potential rehab projects, it could be added to the country´s grid much faster than the construction of new power plants, and with no environmental damage.

3.2 Addition of extra generating units in the power house of operating hydro power plants

Some hydro power plants in Brazil were built leaving room, in the form of “empty wells”, in their power houses for future additions of extra generating units. Such cases are consequences of the Brazilian power sector practice, when most of the power plants were state-owned, of meeting peak demand with the power generated by extra units installed in low-capacity-cost hydro power plants. In the plants listed in Table 2, the planned extra generating units were not installed due to political or economic reasons, but spaces in their power houses were made available for this purpose.


[Table 2]: Brazilian hydro power plants that could be rehabilitated.

The total generating capacity of the 27 new units, distributed in 12 operating hydro power plants, in Table 2 is 5,096 MW.This new capacity, together with the capacity additions accruing from rehab projects could provide substantial amounts of cheap

and reliable electric power, substituting expensive thermal power dispatching.

4. BRAZILIAN POWER SECTOR REGULATION AFFECTING REHABILITATION PROJECTS

During President Lula’s government (2003-2010), a second reform of the Brazilian electric power system took place. Act no. 10848/2004, which enforced some changes brought about by the reform, had the following guidelines:

• Increasing supply security;• Fostering affordable tariffs, through the promotion of energy auctions in the electricity wholesale market; and• Enlarging social inclusion, through the program “Light for everyone”.


The reform kept the electricity supply market divided into two parts, one comprising the free consumers and the other made up by the captive consumers. Free consumers are those with power demands greater or equal to 3 MW; they can choose their suppliers among independent power producers or traders and their energy requirements are met through freely negotiated bilateral contracts. The rest of the consumers are captive ones; they are served by distribution utilities that transact in a pool managed by a new entity, the Electric Power Trading Chamber (CCEE) [6].

A new state-owned company, the Energy Research Company (EPE), was created to perform 10- and 20-year expansion planning for the Brazilian Ministry of Mines and Energy (MME).

To meet the forecast demand of the pool consumers, the 10-year plan sets the commissioning schedule for hydro power plant projects, thermal power plants and wind power farms, identify the regional constraints, and name the transmissions lines to be built. Procurement auctions are organized for all these categories of plants/energy sources. MME has the power to define which energy sources can participate in each auction. The proposals requiring the least revenue during the concession period are the winning bids.

Two types of electrical energy auctions were defined by this reform: one for the generation of existing plants and the other for new plants. The winning bids in the former tend to be lower than those of the latter, thus contributing to lower the average prices. The forecast demand of the distribution utilities for the next 5 years should be fully assured through long-term power purchase contracts signed with the winning generators of these auctions.

The current institutional model of the Brazilian electric power industry tries to integrate competition with the traditional forward planning of this industry, carried out by EPE and MME. The predominantly hydroelectric power system in Brazil is centrally dispatched by the National Operator of the Electrical System (ONS).

So, the amount of electricity that each generator can produce and sell in the Brazilian Interconnected Power System (SIN) depends on the expansion and operation plans elaborated by EPE/MME and ONS, respectively. These entities, consequently, calculate for them their guaranteed generation, the so-called „physical guarantee“, which is the amount they can sell for both distribution utilities, representing the captive consumers, and free consumers, eventually represented by traders.

A rehabilitation project can increase, or not, the physical guarantee of a hydro power plant, but it usually increases the available generation at peak power periods and the reserve margins of the generation system.

Only in 2010 Ordinance no. 861 of the Ministry of Mines and Energy set a clear procedure allowing the sale of the extra power reaped by rehabilitated hydro power plants that achieved increases in their physical guarantee. The owners of such plants should require a review of their physical guarantee to ANEEL [7].

In Brazil, as in several other countries, the users of the transmission and distribution grids should pay a regulated tariff for the power transported through them. When the increase in the physical guarantee of a hydro power plant that underwent a rehabilitation process is not large enough, the increase in the cost of using the transmission grid may eventually cut harshly the extra revenue obtained with the sale of the extra physical guarantee, as actually happened recently in some cases in the country.

The benefits for the interconnected system, in terms of larger, cheaper and more reliable generation at peak load periods and enhanced reserve margins, brought about by rehabilitated hydro power plants are not recognized yet by the current power sector regulation in Brazil.

5. STRATEGIES TO PROMOTE HYDRO POWER PLANT REHABILITATION PROJECTS IN BRAZIL

As shown in the previous section, the current regulation of the Brazilian power sector does not provide enough financial incentives to materialize a substantial part of the large potential, in terms of extra power generation, that could be exploited with the rehabilitation of hydro power plants in the country.

In the Brazilian electric power sector ancillary services have always been treated as obligations charged to the generators, and the two reforms that the sector underwent in the last 18 years have not changed this rule. Auctions for ancillary services, as practiced in Norway, or, simply, separate payments for such services in the long-term contracts required by the current legislation in Brazil could encourage rehabilitation projects.

Another possibility to encourage such projects would be to organize specific auctions for the rehabilitation of hydro power plants. This alternative would profit from the wide experience gathered by EPE, MME, CCEE and ANEEL in organizing auctions for new power plants in the last 8 years.

Act no. 12.783, enacted in 2013, sets new rules for the payment of services provided by the operators of hydro power stations whose concession period run out. Unfortunately, no payments for rehabilitation services are provided for in the articles of this act. The inclusion of such payments within the scope of this act would be a third alternative to foster rehabilitation projects in the country.

6. CONCLUSIONS

Rehabilitations demand much less time than building a new power station; besides, they do not generate additional environmental impacts and have lower costs.

Brazilian power stations are aging and need to go through modernizations processes in order to keep (or increase) their generation.

Rehabilitated hydro power plants in Brazil can provide more power generation near to the main load centers, compared to most of the new hydro power projects, which are located in the Amazon region, far from such load centers and facing fierce opposition from environmentalist groups within and outside the country.

The increase in the capacity of operating hydroelectric power plants in the country through rehabilitation and installation of extra generating units varies, according to the assumptions and calculations shown in this paper, from 5,743 MW if only minimum rehabs would be undertaken, to 11,124 MW if all the potential rehabs would the strong ones. These figures are substantially higher than those calculated by EPE a few years ago [8].

To materialize such potentials, however, some regulatory barriers should be overcome. This paper presents some alternatives ways to achieve this.

Acknowledgements

Special acknowledgements are directed to Andritz Hydro colleagues in Austria, Brazil and Canada, which provided information and data for the study reported in this paper.

7. REFERENCES

[1] Empresa de Pesquisa Energética (EPE), 2011, Consumo nacional de energia elétrica na rede por classe, Report, http://epe.gov.br/mercado/Paginas/Consumonacionaldeenergiael%C3%A9tricaporclasse%E2%80%931995-2009.aspx

[2] ANEEL, 2013, “Banco de Informações de Geração” (BIG),


Report, http://www.aneel.gov.br/aplicacoes/capacidadebrasil/capacidadebrasil.asp

[3] LEITE, N. F., 2013, A geração térmica e seus efeitos na tarifa de energia. Valor Econômico.

[4] US Army Corps of Engineers, 2009, “Hydropower – Value to the Nation”, brochure, www.CorpsResults.us

[5] International Electrotechnical Commission, 2008, IEC62256: Hydraulic turbines, storages pumps and pump-turbines – Rehabilitation and performance improvement, Switerland.

[6] Bajay, S. V., 2006, “Integrating competition and planning: a mixed institutional model of the Brazilian electric power sector”, Energy, 31 (6-7): 865-76.

[7] MME, 2010, “Portaria Nº 861“, Ministry of Mines and Energy (MME).

[8] EPE, 2008, Nota técnica DEN 03/08: “Considerações sobre repotenciação e modernização de usinas hidrelétricas”, Empresa de Pesquisa Energética (EPE), Série recursos energéticos, Rio de Janeiro.


validation of a computational fluid dynamicS mathematical Simulation with a phySical model of a pumping Station

1Sergio Liscia, 2Ezequiel Lacava, 1Milagros Loguercio, 1Cecilia Lucino

1UIDET Hidromecánica, Facultad de Ingeniería, Universidad Nacional de La Plata – Calle 47 No 200, La Plata, Buenos Aires, Argentina. (+54)0221-4236684 – E-m ail: [email protected] Hidromecánica, Facultad de Ingeniería, Universidad Nacional de La Plata – Calle 47 No 200, La Plata, Buenos Aires, Argentina. (+54)0221-4236684 – E-mail: [email protected]. The author is graduated in Hydraulic Engineer at the Faculty of Engineer of the National University of La Plata (UNLP), Argentina. He started to work at the Laboratory of Hydromechanics in 1987 and he has been directing it since 2004 . He is also Professor at the Area “Use of Water Resources and Hydraulic Machines” and Career Director of Hydraulic Engineer.

ABSTRACT

The present work is aimed at verifying the capability of Computational Fluid Dynamics (CFD) to predict the hydraulic behavior of a wastewater pumping station.

The project is under the framework of the Strategic Tunnel Enhancement Program (STEP) that is being developed in Abu Dhabi, UAE. For several months now both the physical and the mathematical modeling are being carried out at the Laboratory of Hydromechanics of the National University of La Plata, Argentina.

The comparison of results was made as a function of the evolution of hydraulic variables at certain points of the domain under study at which uniform behavior was expected, also allowing for the measurement in the physical model with the instrumental available.

Variables chosen for the calibration of the CFD model were the levels of the free surface, velocities and trajectories of water particles. Measurements on the physical model were made with a Doppler acoustic velocimeter, a conventional Pitot tube and a moving scale.

The comparison between the results of both models indicates that errors made are acceptable according to several international standards.

KEYWORDS: mathematical modeling, physical modeling, pump station design

1. INTRODUCTION

The present work is aimed at validating the mathematical model built with the help of the computer program FLOW3D® of the hydraulic behavior of a wastewater pumping station to be built in Abu Dhabi, UAE.

The calibration was made in terms of the velocities measured and calculated at different points of the domain and the elevation of the free surface at the inlet of the pumping station. Modeling results were compared with those obtained experimentally in the physical model. Tests were carried out in the facilities of the Laboratory of Hydromechanics of the National University of La Plata, as a part of the project STEP (Strategic Tunnel Enhancement Program).

2. PROJECT DESCRIPTION

[Figure 1: View of the physical model]

[Figure 2: View of the physical model]

The pumping station that is the subject of this study is aimed at bridging the 90-meter vertical gap between the discharge of the tunnel (5.5 m diameter and 42 km long), that carries the wastewater from the city, and the treatment plant. The project is located at the end of the tunnel and is therefore some tens of meters underground. This factor is of paramount importance, as it has constrained the space available and forced the geometry to be extremely compact as the whole work should fit into a 50-m-diameter pit.

Figure 1 shows a plan and a front view of the pumping station. Two main sections can be distinguished, namely, the screen chamber and the pumping manifold chamber.


3. MATERIAL AND METHODS

3.1. CFD

The domain was discretized so as to allow the simulation of flow under a finite-volume scheme [1]. The mesh density was emerged from the compromise between, above, the reliability of results, on the one hand, and the computational requirements that such density imposes, on the other hand. To that aim, the evolution of the relative error made on a series of simulations performed over increasingly refined meshes. The mesh eventually adopted is the one beyond which any further refining is meaningless.

According to the criterion, the mesh cells are 1 cm , grouped into three blocks corresponding to the charge chamber; the inlet channel and screen chamber; and the pumping chamber (Figure 3).

[Figure 3: Perspective view of mathematical domain]

The number of active cells (i.e., those limited by the boundaries of the domain into which flow takes place) is ten million. Boundary conditions are of the Dirichlet type at the inlet section of the charge chamber, imposing the level of the physical model; and of the Neumann type at the outlet section, at which the discharge corresponds to the scenario associated to the operating conditions adopted. Fluid was considered to be one-phase and incompressible, at a temperature of 20 °C. Gravity was assumed to be 9.81 m/s2 and the turbulence model was the so-called Large Eddy Simulation (LES). The selection of this model lies on the fact that it is the most suitable to identify the unstable behavior of submerged and surface vortices that are frequent in the neighborhood of the suction section of pumps. It is also able to represent with acceptable accuracy the dominant frequencies of unsteady phenomena if an appropriate cell size is chosen [2-4].

3.2 Physical Model

[Figure 4: Front view of the Physical Model]

The physical model was designed so as to create a watertight container built in 3-mm-thick stainless steel with structural reinforce-ments of the same material. The flat bottom was placed at a lower level than that of the project, in order to allow for further changes in the geometry. The inner configuration was built in different materials in order to allow for the visualization of flow at certain key points and the implementation of required correcting measures.

Figure 4 shows a front view of the physical model, all of whose parts can be seen as well as their general geometric distribution.

The geometric scale of the physical model, 1:10, was chosen so as to preserve the similitude of inertial forces (i.e., the Froude number corresponding to both the model and the prototype are the same).

The design of the pumps and their adduction allow for the easy replacement of the elbow upstream the section of the pumps and, in order to ensure the discharge, a siphon was used with reaches of stainless steel, acrylic and PVC, that flows into a restitution chamber into which the level of energy can be controlled.

Measurements were made by means of a conventional Pitot tube (Figure 6) associated to a electronic differential piezometer; a moving scale; and an accoustic Doppler velocimeter (ADV Vectrino).

Water velocities were measured in the section shown in Figure 5 with the help of the ADV. Measurements were made with the Pitot tube at pump 5 (Figure 6) and the free surface was surveyed at every point of the charge chamber (Figure 7).

[Figure 5: Measurement section and instrument used (ADV)]

[Figure 6: Measurement section and instrument used (Pitot)]


[Figure 7: Free surface elevation measured]

Different scenarios were tested, corresponding to operating conditions determined by the number of pumps that are turned on simultaneously. The present work is focused on the results of the model of the scenario in which pumps 3, 4, 5, 6, 7 and 8 (Figure 1) are turned on, releasing a total discharge of 30m3/s in prototype, standing for 94,9 l/s in the model.

4. 4. RESULTS

For the sake of presentation, this section will be divided on the basis of the locations at which comparisons were made.

4.1 Inlet section

The section at which measurements were made corresponds to a plane located at 45 cm upstream the pumping station, in coincidence to the end of the run of the moving scale. Three vertical axes were surveyed with the ADV, as shown in Figure 5.

Figure 8 shows the comparison of results obtained by means of both CFD simulations and experimental measurements in terms of the velocity in the direction of flow observed in this section.

[Figure 8: Velocities at inlet section.]

4.2 Free surface

Three longitudinal axes were analyzed in coincidence with the dividing piles. Figure 9 shows the results corresponding to the middle and right piles, taking advantage of the symmetry of flow.

[Figure 9: Free surface comparison.]

4.3 Section of the pumps

The section studied in this case corresponds to the level of installation of the pumps, where two orthogonal axes are lying in line with the mesh of the mathematical model.

For the sake of simplicity, two mouths were set so as to introduce the measuring tool in the three pumps of the left-hand side chamber (pump 5) that correspond to the scenario under analysis.

Figures 10 and 11 show the ratio between the absolute and the mean velocities measured for each of these sections.

[Figure 10: Velocities at Pump 5. Section AC.]

[Figure 11: Velocities at Pump 5. Section BD.]

4.4 Left-hand side chamber

In order to compare the results, three vertical axes were considered, lying on a coincident with the section defined in Figure 5. Six points of each axis were surveyed: P1, P2, P3, P4, P5 and P6 at distances from the bottom of 30, 25, 20, 15, 10 and 7.5 cm, respectively.

Results obtained with the mathematical model correspond to a 20-second reading, while 3-minute readings were made with the ADV for each point.

Figure 12 show the horizontal components of the velocity vectors as a function of the depth at which they were measured, in order to compare quantitatively the direction of flow.

[Figure 12: Vertical component of the velocity vectors. The color scale denote the depth at which measurements were made.]


Figure 13 shows the results of the measurements on the physical model and those obtained by means of CFD simulations. The mean absolute velocity was determined at every point and referred to the mean velocity of the section.

[Figure 13: Results obtained at the section of the manifold.]

5. DISCUSSION

Comparisons made seem to suggest that the greatest difference obtained between the measurements on the physical model and the results provided by the mathematical model:

Inlet section: 5,6 % Free surface: 12% Section of the pump:15% Left-hand side chamber: 33%

Nevertheless, mean errors that were calculated were considerably less than 10 %, as shown below:

Inlet section: 3,6 % Free surface: 1,2% Section of the pump: 3% Left-hand side chamber: 0%

The 0 % error calculated at the section of the manifold is associated to the fact that experimental results are as similarly distributed with respect to the mean value as the numerical results, in spite of their different velocity profiles.

The source of divergences can be related to the experimental errors of the measuring tools, their actual placement with respect to that in the simulated domain, and reading errors.

Test results show that all recommendations by the American National Standard for Pump Intake Design were taken [5].

6. CONCLUSIONS

The capability of the numerical model to predict the hydraulic behavior was of great help in the predesign of the pumping station and results shown in this work suggest that it is accurate enough for the evaluation of a specific geometry.

A proper turbulence model was chosen and the finite-volume mesh was refined o as to meet the compromise between accuracy of the results and acceptable computing times.

A good agreement was found between the results provided by CFD simulations with respect to the measurements made on the physical model, amounting to errors of less than 4 % at any of the sections studied. This is in the order of magnitude of errors made because of the theoretical approach and the measurements made on physical models.

Thanks to the computer-assisted predesign, a geometric layout was defined that ensured a minimum circulation at the suction section of the pumps, corroborated by tests made on the physical model, conformed to values recommended by international standards.

7. REFERENCES

[1] Matsui J., K. , Okamura T. (2006). “CFD Benchmark and a Model Experiment on the Flow in a Pump Sump”. Proceedings of 23rd IAHR XXIX Symposium, Yokohama, Japan.

[2] Okamura T., Kyoji K. and Matsui J. (2007). “CFD Prediction and Model Experiment on Suction Vortices in Pump Sump”. Proceedings of the 9th Asian International Conference on Fluid Machinery. Jeju, Korea.

[3] Tokyay, T. and Constantinescu, S.G. (2005), “Large Eddy Simulation model to simulate flow in pump intakes of realistic geometry. II: Investigation of dynamics of coherent structures”, Journal of Hydraulic Engineering, ASCE.

[4] Iwano, R., Shibata T. (2002). “Numerical prediction method of submerged vortex and its application to the flow in pump sumps and without a baffle plate”. Proceedings of the 9th International Symposium on Transport Phenomena and Dynamics of Rotating Machinery. Honolulu, Hawai.

[5] Hydraulics Institute Standards (1998) American National Standard for Pump Intake Design - Section 9.8


cfd optimization of low head turbineS intake uSing fiSher-franke guidelineS

1Mauricio Angulo, 1Sergio Liscia

11UIDET Hidromecánica, Facultad de Ingeniería, Universidad Nacional de La Plata – Calle 47 Nº 200, La Plata, Buenos Aires, Argentina. (+54)0221-4236684 –E-mail: [email protected], [email protected]. The author is graduated in Hydraulic and Civil engineering at the Faculty of engineering of UNLP. He works at the Laboratory of Hydromechanics since his graduation in 2005 and is assistant professor at the cathedra of hydraulic machines.

Acknowledgments: This work is ongoing in the context of the counseling work that the National Universities of La Plata (UNLP) and Misiones (UNaM) carry out for the Binational Entity of Yacyretá (EBY) of the Argentina and Paraguay Republics. The Laboratory of Hydromechanics from the Faculty of engineering (UNLP) and the Center of studies for Energy and Development (CEED) are the executing units of the agreement about the expansion project of the Yacyretá Hydropower complex. So the authors wish to thank to EBY for its support and collaboration to do this investigation.

ABSTRACT

Poor design of low head turbine intakes can lead to a non-uniform flow incoming to the machine inlet. This can affect performance and operational features along its useful life.

In most of the cases, turbine acceptance test are performed with ideal intake conditions without considering the actual upstream flow condition. This situation may result in different behavior between the model and the prototype, so it is advisable to model the flow approaching the inlet section. Such modeling is usually studied by physical modeling which can be costly and may require excessive construction and test time, especially when an optimized design is required.

This paper aims to propose the use of a methodology that allows the optimization of low head turbine intakes. The methodology takes as its starting point the Fisher-Franke´s [1] authors criteria regarding the conditions to be achieved by flow at the turbine inlet. These criteria are applied to the numerical modeling results for each one of the intake geometries proposed and comparing it with respect to a reference situation where flow conditions are optimal or “ideal”.

Numerical modeling provides more information in a section than the measurements that could be surveyed on physical model test. Due to this reason the comparison between different designs is sensitive enough to notice the effects that an introduced geometry change provokes within the flow.

In particular this work shows the design optimization of a bulb intake, where the flow pattern upstream the inlet section it is not usual, so the main concern was to guarantee a high quality flow at the turbine inlet section. It has been studied four different designs and it has been also modeled the “ideal” flow condition. The velocity field at a reference section was captured. The magnitude, direction and distribution of incoming velocity components were analyzed and compared.

It was concluded that the design and optimization process by means of numerical modeling of an intake of a low head power station beside the design evaluation with the Fisher-Franke criteria, constitute a very useful tool because it allows the designer to test their proposal quickly, objectively and with a degree of sensitivity enough to note changes on each proposed design.

KEYWORDS: Intake, bulb turbine, Kaplan turbine, CFD

1. INTRODUCTION

Poor design of low head turbine intakes can lead to a non-uniform flow incoming to the machine inlet. This can affect performance and operational features along its useful life.

In most of the cases, turbine acceptance test are performed with ideal intake conditions without considering the actual upstream flow condition. This situation may result in different behavior between the model and the prototype, so it is advisable to model the flow approaching the inlet section. Such modeling is usually studied by physical modeling which can be costly and may require excessive construction and test time, especially when an optimized design is required, due to this fact is preferred CFD modeling.

This work aims to propose the use of a methodology that allows the optimization of low head turbine intakes. The methodology takes as its starting point the Fisher-Franke´s [1] authors guidelines.

In particular this paper shows the design optimization of a bulb intake, where the flow pattern upstream the inlet section it is not usual, so the main concern was to guarantee a high quality flow at the turbine inlet section.

2. MATERIALS AND METHODS

In order to optimize an intake design it is proposed a methodology that has its basis on Fisher-Franke´s guidelines. The proposed methodology has some differences with its precedent mainly in two aspects:

• It is applied on CFD results and not on physical model survey.• The criteria that indicate the wright direction of proposed

designs it is not a static criteria but a criteria that depends on the particular case named as “ideal” flow condition.

According to F&F guidelines, there are some requirements that a good low head turbine intake design should satisfy. These requirements are summarized below:Requirement 1: The normalized axial flow velocity at the

intake section should fit within a certain spatial distribution named as Fisher-Franke limits (red dash lines, Figure 7).

Requirement 2: At trash rack section, no cross flow velocity should exceed +/- 5 % of mean axial velocity.

Requirement 3: Flow free from air entraining vortices.Requirement 4: Deviation of the mean axial velocity for each

quadrant should not vary more than 10 % from mean axial velocity.


Requirement 5: A deviation of the local flow velocity should not exceed 5º. This value could be exceeded only if it not results in a rotating flow.

All of these requirements are evaluated at a unique section close to the turbine inlet, named a reference section. The recommended section should be behind the trash racks, near to the turbine components.

The new methodology reformulates the 1st Requirement. Instead of requesting that the axial velocity be closer to a fixed and specific velocity spatial distribution, it is proposed to get closer to a spatial distribution named as “Ideal”.

To know the velocity distribution at the reference cross section, it is proposed to model an intake by means of CFD where the inlet flow distribution is perfectly uniform. To achieve this condition on a bulb turbine type, a tubular intake was modeled (Figure 4).

On bulb type turbines the reference section can be placed between the bulb and the gate slot. Probably the velocity profile is influenced by the bulb, but this is not a problem because the 1st requirement is assessed in comparison to the “Ideal” flow condition where the same reference cross section was computed.

On Kaplan type turbines it is recommended to take the reference section at the spiral case inlet, behind the trash racks and also behind the gates or stoplogs slots.

The 2nd Requirement it is proposed to be evaluated behind the trash racks. The main reason to this change is that the methodology takes a unique reference cross section to simplify the simulations results postprocessing.

For the 3th requirement, the observation of air entraining vortices on physical model tests is simple, but it is not so easy to detect them by CFD. Nevertheless some efforts in this field have been done to detect swirling flows on pump stations by means of CFD [2].

Originally, the 4th Requirement evaluates the mean velocities for each quadrant. This requirement is modified expressing the same in terms of flow for each quadrant. Although the ratio between the quadrant mean velocity and the section mean velocity is mathematically equal to the ratio between the quadrant flow and the total section flow, it is preferred the last one, because this parameter has a more clear physical sense for hydraulic designers.

The 5th requirement stays the same as it was stated by F&F´s guidelines.

One additional requirement is established, 6th Requirement. Here it is proposed the use of α Coriolis coefficient which is the kinetic energy correction factor. This parameter gives an idea of how far is the actual velocity profile from the uniform velocity profile. Our experience indicates that for an acceptable intake design, α coefficient should not exceed 5% from the “Ideal” flow condition.

Therefore, the requirements of the proposed methodology can be formulated as follows:Requirement 1: The normalized axial flow velocity at the reference

section should be closer as possible to the spatial distribution named as “Ideal” flow condition.

Requirement 2: At reference cross section, no cross flow velocity should exceed +/- 5 % of mean axial velocity.

Requirement 3: Flow free from air entraining vortices.Requirement 4: Deviation of flow for each quadrant should not

vary more than 10 %. Requirement 5: The deviation angle of flow velocity from axial

direction should not exceed 5º. This value could be exceeded only if it not results in a rotating flow.

Requirement 6: The α Coriolis coefficient at the reference section should not exceed 5% from calculated for the “Ideal” flow condition.

All these statements must be assessed at a unique cross section that must be exactly the same for each design and for the “ideal” flow condition. Besides, it is also preferred to be placed at a section where geometry changes occurs upstream of it, in order to not distort a comparative analysis.

2.1. STUDY CASE

Applying this new methodology, it has been studied the optimization of a low head power station with bulb type turbines. The powerhouse has the particularity that allows flood passage over it. This is possible operating two radial gates, one located at the spillway crest and the other at the entrance of the spillway channel (Figure 1). The flood passage function it is not secondary, such is the case that the ratio between the flooding flow and the turbine flow is approximately 7. So the powerhouse design must accomplish both functions.

[Figure 1: Longitudinal section of bulb unit.]

The CFD modeling was done for one unit vane using the specialized CFD code FLOW-3D®. This code solves the Navier-Stokes and continuity equations by finite differences approximation.

In particular this software has the ability to model transient free surface flows with high accuracy because cells can be occupied partially by fluid.

2.2. CFD MODEL SETUP

Geometry construction: The optimization process requires the intake geometry to have a flexible format that allows hydraulic designers to introduce geometry changes in a simple and quick way. For this purpose it has been used a CAD tool which is widely extended in civil works design. The exchange format between CAD and CFD was stereo-lithography (.STL). Four intake designs and the “Ideal” intake were modeled.

Geometry Meshing: For each design it has been used 3 orthog-onal meshes composed by cubic cells. To reduce the CPU total time for each design, CFD calculations have been divided in two stages:

1st stage: model warm-up with coarse meshes.2nd stage: model simulation with fine meshes In all of the cases and for all the geometry components,

the absolute roughness was 1 mm following the experience on previous simulations.

A baffle was added at the plane which contains the blade centerline. This baffle have porosity properties to produce a local energy loss in a such way that, given the reservoir and tailwater levels, the flow thought the turbine is equal to the nominal flow. The porosity properties of it are: porosity and linear loss coefficient are zero, and the quadratic loss coefficient is 1,75. To find this value it is necessary to iterate until the flow thought the turbine is the nominal one within a certain tolerance.


[Table 1]: Geometry meshing and simulation stages.

Mesh 1st Stage 2nd Stage

Size (m) Cells Size (m) Cells

1 Reservoir) 0,50 241.000 0,50 241.000

2 (Intake) 0,50 612.000 0,25 4.450.000

3 (Draft Tube) 0,50 27.000 0,50 27.000

Number of cellsTotal 880.000 Total 4.718.000

Active 489.000 Active 1.850.000

Simulation time (sec) 200 100

CPU time (hs) 7,5 – 14,5 hs 52 – 60 hs

[Figure 2: Geometry meshing.]

Boundary conditions: For the mesh number 1 (upstream) it was imposed the reservoir water level that corresponds with generating condition (NRE). For the mesh number 3 (downstream) it was fixed the same arbitrary constant level for each design (NTE). Summarizing, for the upstream and downstream faces for the whole domain, the boundary condition is specified pressure/fluid height and for the rest of the mesh faces a symmetry boundary condition was set.

Although it is possible to impose flow condition at the upstream face, the set boundary conditions have been always resulted on stable and convergent simulations.

Turbulence Model: The turbulence model chosen was k-ε RNG. A constant maximum turbulent mixed length of 1,19 m was set. This value corresponds with the bibliography recommendations, taking 7 % of the depth at the spillway crest, as the reference length.

Initial Conditions: The initial condition for the first stage was the domain filled with fluid up to the normal reservoir elevation (NRE). The 1st stage run for 200 sec to stabilize at a flow close to the turbine nominal flow. The initial condition for the following stage was the results for the last tim=e interval of the first stage (restart condition). These linked simulations reduce hugely CPU time when high spatial resolution is required.

Simulation: For each intake modeled it is important to know the steady state of flow. To achieve this state it is required an unknown simulation time that is defined according to the first simulation experience. The simulation time and CPU time consuming is summarized on Table 1.

Simulations were computed on a CPU with these main features: Windows 7 64-bits, Intel® Xeon® E5645 (6 cores at 2.4 GHz) and 32 GB of RAM memory.

The flow crossing the baffle at the runner section is monitored during simulation. The criteria to stop it, establishes that flow must be between +/- 3% of nominal flow and must last, at least 50 second; that is the half simulation time for the 2nd stage.

Postprocessing: A reference cross section was defined as the methodology indicates. For each intake design modeled, the results for the last time interval of the 2nd stage are saved. For the reference

section, the values of velocities for the 3 axis directions (Vx, Vy, Vz), the fraction of fluid, the volume fraction, and the position of each cell, are extracted. Calculations are needed to evaluate the requirements; this is explained and detailed in the next section.

3. THEORY AND CALCULATION

To assess each of the requirements it should proceeded as follows:

Requirement 1:The normalized velocity for the main flow direction (axial)

must be calculated for all the reference section cells. Vyn [-] = Vy/Uy

Vyn values must be in descending order. Then, it can be plotted two curves, one for Vyn > 1 and the other for Vyn < 1. The abscissa indicates the cumulated area that equalizes or exceeds a particular value of Vyn (Figure 7).

Requirement 2:For the reference section it is calculated a histogram for

each of the normalized velocities Vxn and Vzn, expressed as the percentage of mean axial velocity.

Vxn [%] = |Vx*100/Uy|

Vzn [%] = |Vz*100/Uy|

The histogram contains on abscissas the normalized velocity value (Vxn or Vzn) and on the ordinate the relative frequency. Figure 7 shows the Vzn histogram.

Requirement 3: This condition was not assessed with CFD results. If the

different intake designs do not modify the submergence of the bulb turbine, this requirement can be excluded of the comparison analysis. It will only be necessary to check it for the final intake design selected as the optimum one.

Requirement 4: To compute flow for each quadrant, the reference section was

divided in 4 quadrants as indicates the next Figure.

[Figure 3: four quadrant division.]

Firstly the flow for each quadrant is computed: Q1, Q2, Q3, Q4. Then the flow deviations are computed for each quadrant as follows:

dQq [%] = (Qq/Qt) * (Aq/At)

The subscript “q” indicates the quadrant order and “t” indicates the total reference section.

The results are plotted on a graph with the quadrant order on abscissas and the flow deviation on the vertical axis (Figure 7).

Requirement 5: For the reference cross section the histograms of deviation

angles α (XY plane) and β (ZY plane) are computed.

α [°] = arctan |Vx/Vy|

β [°] = arctan |Vz/Vy|

The histograms have the α angle on abscissas and the relative frequency on the ordinate. The same can be done for β angle (Figure 7).


Requirement 6: The correction factor for the kinetic energy named as α

Coriolis, was computed for the reference section by means of the following expression:

αc [-] = Σ [(Vyi/Uy)3 . Ai] / ΣAi

The subscript “i” corresponds to the cell order that forms the reference section.

If the CFD modeling is computed with FLOW-3D® code, as it is our case, it is necessary to consider the Fraction of Fluid (FF) and the Volume Fraction (VF) of each cell in order to know the actual fraction of the cells filled with fluid. This area is computed with this expression:

Ai = Acell . FF . VF

The results are summarized on the Figure 7.

4. RESULTS

In our study case it has been modeled the intake named as “Ideal” flow condition (Figure 4) and four different intake designs (Figure 6).

Changes introduced in the hydraulics designs modified one or more of these elements (Figure 1):

• The spillway channel nose• The spillway step• The connection ramp between the spillway crest and the

bulb inlet.• The strut / deflector (position and number)Figure 6 shows the longitudinal sections for each intake.

There, it could be observed geometrical changes introduced from the original design (intake 1) to the final design (intake 4).

Figure 5 summarized all the reference cross sections studied. Maybe only looking at them, designers can take a decision about which is the best or which are the best designs, but this is not enough to take a decision based on quantitative and objective analysis.

Due to this reason the new methodology has been applied and the results are shown on Figures 7 and 8. With only these five graphics we have enough elements to indicate which the alternative that gets closer to the “Ideal” intake is. Also, as we are making changes we can know whether the optimization process is going in the wright way or not.

If we observe these Figures it is clear that intake number 4 is the design that fits better to the six requirements stated previously.

In respect to the 1st requirement, the normalized velocity distribution is very close to the ideal one (Figure 7 – upper left). Looking to the 2nd requirement, the intake 4, has a maximum deviation of 30 %, with a peak value at 15 %. The requirement of reaching 100 % of cross vectors under a deviation of 5 % could be excessive and maybe impossible to achieve, taking account that for the ideal case, it cannot fulfill this condition due to the 22 % of the total cross vectors has a deviation of 10 % (Figure 7 – Lower right).

4th requirement, about quadrant flow distribution, is widely achieved by intake 4. It presents a distribution under +/- 5 % deviation, when the requirement is 10 %. It can also be observed that the upper quadrants take the mayor part or total flow as it is observed on the other designs (Figure 7 – Upper right).

For the design number 4, the 5th requirement shows an important improvement in the deviation angle. It has been only plotted the deviation angle β because it is the more significant direction. The α angle it is not affected by design changes so it was excluded from the comparative analysis. It is evident that depending on the intake geometry this angle could be relevant

and must be considered in other powerhouse studies. For the intake 4, none of the vector exceeds a deviation of 20°, moreover, most vectors have a deviation of 10°. As it was observed for the 2nd requirement, it is also difficult to achieve the condition of no vectors over 5° of deviation, because it cannot be fulfilled by the ideal intake. Therefore the design was accepted because it accomplished a minimum deviation and a rotating flow cannot be observed (Figure 7 – Lower left).

Finally the comparison between the α Coriolis coefficient is convincing when we are talking about being closer to the ideal flow condition. For the ideal intake de αc value is 1,01 and 1,03 for the intake number 4 (Figure 8).

[Figure 4: Ideal Intake - Absolute velocity distribution on the vane´s middle plane (left). Vy velocity profile at ref. section (right).]

[Figure 5: Intake´s reference sections colored by Vy velocity.]

[Figure 6: Four intake designs – absolute velocity profile for the longitudinal cross sections.]

Bellow this Figures have been repeated the more significant graphs where it can be observed a clear improvement on the hydraulic designs, starting with the intake 1 and finalizing with 4. For the 2nd requirement the histogram has its peak centered at 40 %, with a minimum of 25 % and a maximum of 50 % of deviation for the intake 1; changing to a histogram centered on 15 %, with a minimum of 5 % and a maximum of 30 % for the best design (Figure 9 - center).

The same could be said for β deviation angle, changing from an histogram centered on 15°, with a minimum of 10° and a maximum of 35° for the intake 1, to a histogram centered on 10°, with a minimum of 5° and a maximum of 20° for the intake 4 (Figure 9 - right).

Finally de α Coriolis coefficient for intake 1 is 1,24 and 1,03 for intake 4.


[Figure 7: Req. 1: Spatial distribution of Vyn (Upper left). Req. 4: Flow deviation for each quadrant (Upper right). Req. 5: Histogram of β deviation angle (Lower left). Req. 2: Histogram of Vz deviation from axial mean velocity (Lower right).]

[Figure 8: Req. 6: α Coriolis coefficient for each intake design.]

[Figure 9: Intake 1 and 4 comparisons. Req. 1: Spatial distribution of Vyn.(left).Req. 2: Histogram of Vz deviation from axial mean velocity (center). Req. 5: Histogram of β deviation angle (right).]

5. DISCUSSION AND CONCLUSIONS

The incorporation of the 6th requirement, suggesting the use of de Coriolis coefficient, is really important due to it is sensitive enough to detect improvements during the optimization process.

Finally it was concluded that the design and optimization process by means of numerical modeling of an intake of a low head power station beside the design evaluation with an upgraded methodology based on Fisher-Franke´s guidelines, constitute a very useful tool because it allows the designer to test their proposal quickly, objectively and with a degree of sensitivity enough to note changes on each proposed design.

At present we are working on the implementation on physical model tests to survey velocity distribution at the reference section with ADV technics. We expect to verify the design selected as optimum and we also expect to validate predictions done by CFD modeling, probing at the same time the ability of this methodology to reach the optimum design.

It is also of our interest to extend the use of this methodology in the study of the whole low head power stations. The goal will be evaluate different scenarios of power production and flood passage on isolated power station or contiguous to spillway works.

6. NOMENCLATURE

α : Deviation angle, XY plane. αC : Coriolis coefficient of kinetic energy correction.β : Deviation angle, YZ plane.ADV: Acoustic Doppler VelocimetryAcell: Cell área.Ai: Effective cell area.Aqi: Effective cell area for the quadrant “i”.At: Effective section areaCFD: Computational Fluid DynamicsFF: Fraction of Fluid.NRE: Normal Reservoir Elevation.NTE: Normal Tailwater Elevation.Vx: Axial velocity.Vy: Vertical velocity.Vz: Cross Velocity.Vxn, Vyn, Vzn: Normalized velocities for three directions.Uy: Mean axial velocity.VF: Volume Fraction.

7. REFERENCES

[1] Fisher, F. K. Jr., and Franke, G. F., 1987, “The Impact of Inlet Flow Characteristics on Low Head Hydro Projects. International confererence on hydropower , Porland, Oregon.

[2] Lucino, C., Liscia, S. and Duró, G., 2010, “Detección de vórtices en dársenas de bombeo mediante modelación matemática”. XXIV Congreso latinoamericano de Hidráulica, Punta del Este, Uruguay.


guide vane influence over preSSure fluctuation at the diScharge ring in a kaplan turbine: experimental aSSeSSment

1Arturo Rivetti , 2Cecilia Lucino , 2Sergio Liscia

1UIDET Hidromecánica, Facultad de Ingeniería, Universidad Nacional de La Plata – Calle 47 Nº200, La Plata, Buenos Aires, Argentina, Zip Code: 1900. (+54)0221 3562155 – E-mail: [email protected]. The author is graduated in Hydraulic Engineer at the Faculty of Engineer of the National University of La Plata (UNLP), Argentina. Since 2009 he is working at the Laboratory of Hydromechanics and he is doing a doctoral degree in the field of CFD simulation in Kaplan turbines. He also is teacher assistant in the are of “Use of Water Resources and Hydraulic Machines”.2UIDET Hidromecánica, Facultad de Ingeniería, Universidad Nacional de La Plata – Calle 47 Nº 200, La Plata, Buenos Aires, Argentina. (+54)0221-4236684 – E-mail: [email protected]

ABSTRACT

In this work, experimental research of a large Kaplan turbine at prototype scale for the minimum guaranteed head and high discharge is carried out with the goal of studying the influence of the guide vane opening in the flow pattern over the discharge ring. Measurements of wall pressure were made at different locations in the discharge ring wall between the guide vanes outlet to the draft tube inlet.

The test was performed at constant head, while varying the opening of the guide vanes in discrete steps. The spectral analysis of these signals shows that main frequencies are the blade passage and their harmonics. One accelerometer was located at the man door entrance and, after a demodulated analysis of the signal, a good correlation was observed against pressure pulsation spectra. This shows that the vibrations of the turbine structure are modulated by its hydraulic components. Furthermore, a frequency corresponding to the first rotor-stator diametral mode appeared in the acceleration signal, showing the interaction between these two components. This frequency component appeared for guide vanes openings greater than 84%, reaching its peak at 100%.

KEYWORDS: Pressure Fluctuation; Discharge Ring; Kaplan Turbine

1. INTRODUCTION

Kaplan turbines are used in low-head hydroelectric power plants, which correspond to alluvial rivers carrying high flow rates. In the last decades, the number of such developments has increased, pushed by the ever-growing electrical demand. With the goal of reducing costs, and on account of the high flow rates involved, larger diameters have been required in order to reduce the number of units. At the time when these machines were designed, techniques were not sufficiently developed to study certain hydrodynamic features, giving rise, on occasions, to damage related to cavitation erosion on the blades surface and the discharge ring. One specific pattern of erosion occurs in the form of discrete spots on the discharge ring corresponding to the number of guide vanes [1]. This situation was observed in the Kaplan turbine that is studied in this work, with 9.5 m of runner diameter where the guide vanes present an overhang for an opening greater than 82%. For this reason, prototype measurements were performed to study the influence of the guide vane opening in the flow pattern on the discharge ring.

In the present work, firstly the case study of a large Kaplan turbine at prototype scale operating at the minimum guaranteed head is introduced. Then, a theoretical model of the interaction between guide vanes and runner blades (RSI) is presented. Measurements are analyzed in the time and frequency domains. A demodulated analysis of the accelerometer signal is done showing the presence of the hydraulic components. Finally, a qualitative analysis of the guide vane aperture influence in the dynamic behavior of the machine is done.

2. 2. CASE STUDY AND EXPERIMENTAL SETUP

A Kaplan turbine at prototype scale was studied in this work. The runner has 5 blades and a diameter of 9.5 m. The stator is composed of 24 guide and stay vanes. The draft tube and the semi-spiral casing are divided into three channels by two vanes.

The actual unit, featuring a rated power of 157 MW, is located in the power plant of Yacyretá on the Paraná River, on the boundary between Argentina and Paraguay.

The test was carried out at constant head, while varying the opening of the vanes in discrete steps from 0% up to 100% (Fig. 1 and Fig. 2). At each step, the guide vane opening is kept constant for 350 runner revolutions, long enough time to stabilize the flow and record the signals by means of sensors. The location of pressure sensors and the accelerometer is showed in Fig. 3. The sampling rate is 1000 fn for the pressure and 42000 fn for the accelerometer. The guide vane and runner blade angle, the output power, and the net head were monitored during the whole test with a sampling rate of 150 fn. All the signals were connected to an acquisition system that saves the recordings into a computer hard disk.

In the moment of the test, the machine was operating at the minimum guaranteed head as shown in the efficiency chart. All the operation points measured correspond to on cam combination.

[Figure 1: Steps measured over the discharge hill.]


[Figure 2: Time history showing every measured step.]

[Figure 3: Pressure transducer and accelerometer locations.]

3. RSI INTERACTION IN A KAPLAN TURBINE

Hydraulic turbines are subjected to pressure pulsations due to the interaction between rotating and stationary parts. The guide vanes produce a non-uniform flow that interacts with the blade runner passage producing pressure fluctuations. This effect, known as Rotor Stator Interaction (RSI), is more significant in Francis and Pump turbines, where the gap between guide vanes and runner blades is small. However, this longer distance does not avoid the guide vane influence on the flow pattern at the runner section. The cavitation frosting in Kaplan turbines present on the discharge ring occurs at different locations corresponding to the number of guide vanes, evidencing that the influence of the guide vane is not negligible. An analytical model proposed by Ruchonnet [2] to predict the frequencies of the RSI phenomenon for pump turbines was applied in this work.

Diametral rotating pressure modes were obtained composing the pressure fields from both rotor and stator. Considering a simplified model, Eq. (1) and Eq.(2) were used to describe the pressure fields in stationary and rotating domains as Fourier series. Combining this two equations, and considering that the runner angle coordinate is related to the stationary system of reference θr = θs - ωt, the pressure field in the gap is expressed by Eq. (3). This equation describes the RSI interaction, which is function of time and space having two diametral pressure modes, k1 and k2, rotating in the same direction as the runner when k is

positive, and in the opposite direction when k is negative. In this case, with Zo = 24 and Zb = 5, the higher amplitude expected is for m = 1 and n = 5. Eq. (4) gives k1 = 1 which represent one diametral mode rotating with in same direction of the runner with the corresponding frequency f/fn = mZb = 25 in the stationary frame of reference. k2 is usually not relevant because of the high harmonic number. This frequency of 25 fn indicates the presence of RSI interaction, and the amplitude of this component, the intensity.

4. RESULTS AND DISCUSSION

4.1. Pressure Signal analysis

The pressure signals are plotted as a function of the dimensionless coefficient Cp (Eq. 6) considering only the dynamic behavior removing the mean value of the pressure. In this analysis, a guide vane opening of 86% was chosen to show the results.

The dominant frequency component is the blade passage (5 fn) with the maximum amplitude for the sensor SPB as can be seen in Fig. 4. This sensor is the closest to the blade where the difference of pressure between the suction and pressure side is noticeable. The runner passage frequency (fn) appears in the sensors SPA and SPD induced by the non-uniformity in the pressure profile due to the presence of the semi-spiral casing and the draft tube. For the other steps, the frequency components observed are the same but with different amplitude values.

In the spectrum of the sensor SPD the frequency 25 fn related with the RSI that was analyzed in the previous section appears, but is not clear if the origin of this component is the RSI or a harmonic component. Further research is needed to clarify this issue.

[Figure 4: a) Pressure temporary signals. b) Pressure spectra.]


4.2. Acceleration signal analysis

The acceleration signal is expressed as a fraction of the gravity acceleration. The instrumentation of this sensor could be done in a fast and easy way since access is guaranteed during the operation period of the machine. Therefore, it is important to perform a correlation analysis against pressure values, in order to have an external parameter of the internal hydraulic behavior, since the instrumentation of the pressure sensors is quite challenging.

It is not possible to identify any main frequencies from the FFT of the acceleration (Fig. 5a). In the spectrum the energy is located in a range of frequency (100 to 10000 fn) that is not related with the hydraulic components. This spectrum shows the structural response to hydraulic forces and is associated to the vibration modes of the man door structure.

In order to obtain information of the hydraulic behavior by means of the accelerometer signal, a demodulated analysis was required. First, a low-pass filter at 5000 fn was applied and a Hilbert transform was used to obtain the envelope of the signal. Doing the FFT of the envelope, all the hydraulic components are present, as can be observed in Fig. 5a and Fig. 5b. There is good correlation with the pressure signal for sensor SPC, since it is the closest to the location of the accelerometer. Moreover, performing the same analysis for all the experimental steps, it is shown that the frequency of 25 fn, that represents the RSI interaction, becomes more important for great openings of the guide vane. Beyond an opening of 90%, the amplitude increases exponentially. This can be explained by the fact that for great openings, the influence on the flow perturbation due the overhanging increases.

[Figure 5: a) Comparison between the modulation of the acceleration and the pressure pulsation in the FFT analysis for opening 86%. b) Comparison between the modulation of the acceleration and pressure pulsation in the FFT analysis for opening 100%.]

4.3 Guide vane opening influence

To study the guide vane influence on the hydraulic behavior in the surroundings of the machine, pressure and acceleration signals were analyzed in terms of different parameters, namely:

The standard deviation: For the acceleration signal, the magnitude of the standard deviation (Eq. 7) represents the level of vibration of the structural components that is the response of the hydraulic forces.

The blade passage amplitude: It is the amplitude of the frequency component of 5 fn taken from the spectrum analysis. In the case of the acceleration signal is obtained from the spectrum of the envelope of acceleration.

The RSI frequency amplitude: It is the amplitude of the frequency component of 25 fn taken from the spectrum of the acceleration envelope. It represents the magnitude of the interaction between the guide vanes and the runner blades. The evolution of these three parameters as a function of the guide vane opening is shown in Fig. 6.

[Figure 5: a) Standard deviation for accelerometer signals. b) Amplitude of the envelope of acceleration for the components 5 fn and 25 fn. c) Amplitude of the blade passage for pressure sensors.]

As is shown in Fig 6a, the level of vibration exhibits an exponential increase for guide vanes opening greater than 82%. This is in agreement with the increase of the blade passage amplitude since this is the main hydraulic frequency that induces the structure vibration.

The SPB pressure is only plot up to a guide vane opening of 84% because it was out of service during the measurements for greater openings. As was explained in the previous section, the frequencies of 5 fn and 25 fn are not present in the accelerometer signal.

For this reason, these amplitudes were analyzed in the demodulated spectrum (Fig 6b). The blade passage amplitude shows a behavior analogous to that of the level of vibration (Fig 6a) and the pressure amplitudes (Fig 6c). For the frequency of 25 fn that represents the RSI interaction, the amplitude is zero until a guide vane opening of 84%. Then it increases to reach the peak for an opening of 100%.

5. CONCLUSIONS

Experimental research of a large Kaplan turbine at prototype scale operating at the minimum guaranteed head has been presented. Pressure transducers were located in the prototype machine over the discharge ring from guide vanes outlet to draft tube inlet. The pressure fluctuation is dominated by the blade passing frequency.

One accelerometer was placed at the man door entrance during experiments. The spectral analysis shows range of frequencies components that are not related with the hydraulic ones that appears in pressure spectra. These components correspond to vibration modes of the mechanical structure. Applying a filter over the accelerometer signal and doing a demodulated analysis using the Hilbert Transform, it was found that all hydraulic components are present. This means that the structure vibration is modulated by the hydraulic pressure pulsation. The frequency component of 25 fn starts growing for guide vane aperture greater than 82%. This frequency is the result of the runner-stator interaction that generates a rotating diametral mode.


The level of vibration represented by the standard deviation of the acceleration increases exponentially for openings of guide vanes greater than 82%.

6. NOMENCLATURE

ƒn [Hz] Passage runner frequency.g [m∙s-2] Gravity acceleration. ρ [kg∙m-3] Water density. Cp [-] Pressure coefficient.pabs [Pa] Time average absolute pressure. pabs [Pa] Absolute pressure. E [j∙kg-1] Specific energy. Ap [%] Guide vane opening.Zo [-] Number of guide vanes, Zo=24. Zb [-] Numberof runner blades, Zb=5. n [-] Harmonic Order. m [-] Harmonic Order. k1 [-] Diametrical pressure mode. k2 [-] Diametrical pressure mode.θs [-] Angle in stationary system.θr [-] Angle in rotating system.øn [-] Phase for the nth Harmonic.øm[-] Phase for he mth Harmonic.Bn [-] Amplitude for the nth Harmonic.Bm [-] Amplitude for the mth Harmonic.

7. ACKNOWLEDGEMENTS

The study introduced here was carried out under the framework of a project focused on the dynamic behavior of Kaplan turbines that combines prototype measurements, model tests and CFD simulations, and is supported by the Yacyretá Binational Entity (EBY), the National University of La Plata, Argentina (UNLP) and the National University of Misiones, Argentina (UNAM). Special thanks to Oscar Héctor Capezio and the technical team of the Yacyretá power station for their contribution in the prototype measurements.

8. REFERENCES

[1] Nennemann, B. and Vu, T. C., 2007 “Kaplan turbine blade and discharge ring cavitation prediction using unsteady CFD”. 2nd IAHR international meeting of the workgroup on cavitation and dynamic problems in hydraulic machinery and systems. Timisoara, Romania.

[2] Ruchonnet, N. Nicolet, C. Avellan, F., 2006 “Hydroacoustic Modeling of Rotor Stator Interaction inFrancis Pump-Turbine”. IAHR Int. Meeting of WG on Cavitation and Dynamic Problems in Hydraulic Machinery and Systems. Barcelona, Spain.

[3] Dörfler, P., Sick, M. Coutu, A., 2013, “Flow-Induced Pulsation and Vibration in Hydroelectric Machinery”, Engineer’s Guidebook for Planning, Design and Troubleshooting. Springer, pp pp.84 – 86, pp. 132 – 134.


hydraulic tranSitory Study in the Small hydropower by characteriSticS method in order to Surge tank dimenSioning

1Regina Mambeli Barros, 2Geraldo Lúcio Tiago Filho, 3Ivan Felipe Silva dos Santos, 4Fernando das Graças Braga da Silva

1Civil Engineer, Phd. and Masters from PPG-SHS/EESC/USP, Phd. Professor - IRN/ UNIFEI, Researcher from National Reference Center in SHP-CERPCH; Natural Resource Center – IRN; Federal University of Itajubá, UNIFEI, Av.BPS, 1303, Itajubá-MG, CEP: 37500-903 Av.BPS, 1303, Itajubá-MG, CEP: 37500-903, phone number +55 35 36291224, [email protected] 2Mechanical Engineer, Phd. in Hydraulic Systems from USP and Masters in Mechanical Engineering in Flow Machines from UNIFEI, Director and Phd. Professor - IRN/ UNIFEI, Researcher from National Reference Center in SHP-CERPCH; Natural Resource Center – IRN; Federal University of Itajubá, UNIFEI, Av.BPS, 1303, Itajubá-MG, CEP: 37500-903, phone number +55 35 36291156, [email protected] Engineering and Masters’ student in Engineering of Energy - UNIFEI, Av, BPS, 1302, Itajubá- MG, Zip Code: 37500-903, Researcher from National Reference Center in SHP-CERPCH; Natural Resource Center – IRN; Federal University of Itajubá, UNIFEI, Av.BPS, 1303, Itajubá-MG, CEP: 37500-903, Brazil, phone number +55 35 36216935, [email protected] Engineer, Phd. and Masters from PPG-SHS/EESC/USP, Phd. Professor - IRN/ UNIFEI, Researcher from National Reference Center in SHP-CERPCH; Natural Resource Center – IRN; Federal University of Itajubá, UNIFEI, Av.BPS, 1303, Itajubá-MG, CEP: 37500-903, phone number +55 35 3629 1485 [email protected]

ABSTRACT

This study is intended to make a study about hydraulic transient regarding to a design of fictional Small Hydropower (SHP). For that, it was conducted the characteristics method, which purposes to transform two partials differentials equations (Saint Venant) into ordinary equations, that have more suitable properties for numerical calculations, while may allow explicit solutions. The case study was an application for a simple pipe, in the case of valve-closing in the downstream boundary, with a reservoir in upstream boundary. The main data consisted of design head of the project, of 185.56 meters, 2.80 m3/s for turbine discharge for a typical wet month (February), diameter and length pipe (penstock), of respectively, 1.84 meters and 920 meters. The obtained results showed the simulated values for valve pressure with variation for turning valve between 4 and 12 seconds results in maximums values of pressures that oscillated 205.49mca and 204.01 mca (4s) as well as 192.12mca and 191.19mca (12s). The surge tank dimensioning resulted with total height of 19.16m, and a diameter of 1.60m. Another distance between valve and surge tank less than 920.00 meters, as well a greater diameter could be considered in this dimensioning, in order to become, respectively, the size of this surge tank lower, and to prevent a so slender concrete structure.

KEYWORDS: Hydraulic Transitory, Small Hydropower, Water Hammer, Surge Tank.

1. INTRODUCTION

According to Afshar et al. (2010), hydraulic transient events are caused during a change in state, from one steady or equilibrium condition to another, for example, sudden valve opening or closure, starting or stopping of pumps or turbines, mechanical failure of an item, rapid changes in demand condition, etc. The authors (op. cit.) remembered that the main components of this disturbance are pressure and flow changes that bring about propagation of pressure waves throughout the system, as well the velocity of this wave may exceed 1000 m/s, which may lead to severe damages. Design and operation of any pipeline system require that the distribution of head and flow in the system is predicted at different operating conditions, which justify the modeling of these phenomena, as have been purposed in this paper.

Various numerical approaches have been introduced for calculation of the pipeline transients, namely: method of characteristics (MOC), finite volume method (FVM), finite element method (FEM), wave characteristics method (WCM) and finite difference method (FDM). Among these methods, MOC is the commonly used method by of its simplicity as well superior performance in comparison with other methods (Afshar et al., 2010).

The water hammer effects caused by closure of spherical valves against the discharge have been studied by Karadžić et al. (2010), in which Perućica high-head hydropower plant (HPP), Montenegro, safety spherical valves (inlet turbine valves) have been refurbished on the first two Pelton turbine unit. According to the authors (op. cit.), the spherical valve boundary condition was incorporated into the MOC algorithm. As a result, it has been found that flow conditions do not have a significant impact on the spherical valve closure time for the cases investigated by Karadžić et al. (2010), as well as the developed numerical models shown reasonable agreement with measured results.

Bortoni (2008) has presented the basis of Gibson’s method (Figure 1) for flow measurement in a SHP plants, the recent developments related to the employed instrumentation, as well as the methods for overcoming the leakage flow measurement, and some actual applications with pressure–time records for flow calculation. For the case study presented by this author (op. cit.), the flow measurements were conducted in five loading conditions, namely: 25%, 50%, 75%, 85% and 100% of full load. The resulting differential overpressure records are presented in Figure 2 (with the differential overpressure presented in meters of water column).

[Figure 1: Geometric, kinematic and dynamic characteristics of Gibson’s method. Source: Bortoni, Souza e Santos (1997 apud Bortoni, 2008).]


[Figure 2: Pressure–time records for several load conditions.Source: Bortoni (2008).]

The MOC algorithm has been very useful for several engineering purposes, it also can be studied for de pumping purposes. For example, the method of characteristic line (MOC) was adopted by Tian et al. (2008), to evaluate the valve-induced water hammer phenomena in a parallel pumps feedwater system (PPFS) during the alternate startup process of parallel pumps. The MOC was used by the authors (op. cit.) to compute the transient phenomena including the local flow velocity, pressure wave vibration, and slamming of the check valve disc, etc. Some interesting results were obtained and it was shown by Tian et al. (2008) that severe slamming between the valve disc and valve seat occurred during the alternate startup of parallel pumps. According Tian et al. (2008), the induced maximum pressure vibration amplitude was up to 5.0 MPa which has occured under high–high speed startup condition. For the purpose of mitigating the slamming times and pressure pulse amplitude that follows the water hammer, an optimum damping torque was adopted by Tian et al. (2008) to slow the valve disc closing speed, and it has been numerically shown to be an effective approach.

As the same purpose of the present paper, in spite of the lack of experiments for quantitative validation, the present computational results are expected to be instructive for the optimum design of the SHPs to purposes to mitigate the potential damage caused by valve-induced closing-time water hammer.

2. METHODOLOGY

The methodology proposed by Chaudry (2001) concerning the development of hydrodynamic models has been used, in which runoff is regarded as a phenomenon using the laws of physics, namely conservation of mass (assuming space), conservation of momentum etc. Figure 3 shows an elementary volume analysis aimed at conservation of mass.

[Figure 3: Elemental volume for mass conservation equation. Source: elaborated by the authors as based on Chaudhry (2001).]

The one-dimensional mass conservation equation (Equation 4) is developed as follows in Equations (1) to (3). Considering

that is the change in discharge rate (Q) and that

q’.T = q(m3 / s / m) is the unitary side entrance (for meters), as developed by Chaudhry (2001):

(1)

Assuming that ρ is incompressible and therefore it is constant, as well that surface area.

(2)

(3)

Ρ and Δx were cutting off.

(4)

Where:T: width of the surface [L]; A, cross-sectional area [L2]; and

Q: discharge rate [L3/T].The volume element for analysis of the momentum

conservation equation, qm (mass.velocity) is presented in Figure 4. Equation (6) presents the momentum conservation equation, after some arrangements of algebraic equation (5), as presented by Chaudhry (2001).

[Figure 4: Elemental volume for the momentum conservation equation. Source: elaborated by the authors as based on Chaudhry (2001).]

(5)

Instead the using of mass, the value of mass per time (in seconds) is used, as ρ.Q.v.

By the substituting of the forces of flow and inflow of momentum, the following is obtained.

(6)

Where:ν: speed; S0: slope of the water line (slope of the bottom),

and Sf: slope of the energy line.Equations (4) and (6) - mass and momentum conservation

equations – have been called Saint-Venant: partial differential equations with few explicit solutions, as recommended by Chaudhry (2001).


By assuming one-dimensional flow as well based on the continuity and momentum equations describing the general behavior of fluids in a closed duct in terms of two variables, namely, y, piezometric head, and ν, fluid velocity the analyses of most hydraulic transients in pressurized systems are carried out. Wave propagation velocity or celerity, c, friction f, and pipe diameter D, are pipe parameters which can be considered constant with time, despite of they be spatial functions (Izquierdo and Iglesias, 2002).

The alternatives for solving such equations for both the discharge and depth of water variations along both the flow (x) and over time (t) are (Chaudhry, 2001):

I) To simplify the equations.II) To use numerical methods (by replacement of derived by

differences).III) Make changes.

2.1 Methods of characteristics (MOC)

This method aims to transform the two partial differential equations (there are two independent ∂v/∂x and ∂v/∂t) into ordinary equations which have more convenient properties for numerical calculation at the same time that explicit solutions can be allowed (Chaudhry, 2001).

Referring to Figure 5 which allows the schematic representation of the cross-sectional area of the river trough, the riverbed increasing (y) increases the area (A) as can be seen (Chaudhry, 2001).

[Figure 4: Elemental volume for mass conservation equation. Source: elaborated by the authors as based on Chaudhry (2001).]

[Figure 5: Schematic representation of the cross-sectional area of the river trough. Source: elaborated by the authors as based on Chaudhry (2001).]

By substituting of T = dA/dy into the mass conservation equation (Equation 4), after some algebraic arrangements, Equation (7) can be obtained (Chaudhry, 2001).

(7)

As proposed by Chaudhry (2001), in the analyzing it must be remembered that the momentum equation (Equation 6), and calling it by L1, as well L2 and L1 being = 0, a linear combination of L1 and L2 will also be zero.

L1 + λL2 = 0 To be replaced L1 e L2

Chaudhry (2001) has described the arrangements for the above algebraic linear combination as follows.

Assembling the partial derivatives into ν (Chaudhry, 2001):

Compared with the chain rule, it has been concluded that (Chaudhry, 2001):

There are two expressions for dx/dt. So that they are identical, since they describe the same thing, then these two expressions must be equals. There is a particular value of λ that makes possible to do this substitution (Chaudhry, 2001).

The particle trajectories within the wave can be observed as follows Chaudhry (2001):

I: (8)

and

II: (9)

Rewriting to λ (Chaudhry, 2001):

III: (10)

and

IV: (11)


By observing Equations (8) to (10), the two partial differential equations have become into four ordinary differential equations simplifying the work of resolution as can be seen (Chaudhry, 2001) in the Figure 6.

[Figure 6: Characteristics equations. Source: elaborated by the authors as based on Chaudhry (2001).]

If c+ and c- intersect in P point, the (9) and (11) equations can be solved simultaneously (P as intersection of c+ and c-).

2.1.1 Solving strategy of the four equationsWhen the numerical solution of differential equation is gotten,

it is possible to replace the derivative by these approaches. Thus, the equations (8) and (10) approach as:

(8)

(10)

The approximation of an implicit differential equation is stable (sometimes unconditionally stable) while the explicit is unstable, unless Δt very small and consistent with Δx has been chosen. (E), and according to the Courant condition, as recommended by Tucci (1998).

According Izquierdo and Iglesias (2002), such stability of this calculation scheme is assured by the so-called Courant-Friedrics-Lewy (CFL) condition, as well, in the case of linear systems, the stability is studied through the Fourier expansion of the error. As well as remembered by Chaudhry (2001), the CFL condition is given by (Izquierdo and Iglesias, 2002):

The simultaneous resolution of these equations provides coordinates of P as shown in Figure 4. The solution to the explicit scheme is straightforward (Chaudhry, 2001). Considering, as described by Tucci (1998) that the speed of the wave is given by:

Equations (8) to (11) are written as follows in Equations (12) to (15).

I: (12)

II: (13)

III: (14)

IV: (15)

The solution for a problem of wave propagation goes from a given initial condition at t = 0, in terms of flow or height (or both), etc. Therefore, the length L of Figure (6th) must be subdivided in sub-sections. If the calculation of points P proceeds to other points afterward, the grid (x, t) will be completely irregular and difficult to manage (Figure 7a). Chaudhry (2001; Streeter and Wylie, 1982) have been proposed a regular grid as well as an evaluating of the information at the point A and B, etc., which must be obtained by interpolation at each time interval (Figure 7b), as also recommended by Chaudhry (2001).

[Figure 7: Solving strategy of the characteristic equations a) irregular grid; e b) with a regular grid. Source: elaborated by the author as based on Chaudhry (2001).]

Since P is known, the equations I and IV (equations 12 and 15) must be searched, with the coordinates xA and xB, the equations II and IV (equations 13 and 15) are numerically solved to obtain vp and yp into the all grid points, t + t (Chaudhry, 2001). The contour points have no negative feature (c+ and c-), so that the channel have been ended. This calculation is applied only to “interior points”, but not for points where the contour has only one characteristic curve. At x=0 there is only the negative curve (c-) and x=L there is only positive curve (c+) (Chaudhry, 2001).

In these contours, the characteristic equation for v and y (II and IV, i.e., 13 and 15 equations) must be supplemented by another equation from the boundary condition. Generally, at x=0, the hydrograph is used as input boundary condition and the far downstream (x=L), a relationship between flow and elevation (curve-key) as boundary condition (Chaudhry, 2001).

The sequence of calculations is described in Streeter and Wylie (1982) and was recommended by Chaudhry (2001):

(i) To located in any point P inside (Figure 8).(ii) To obtain coordinates R (if, c+) and S (if, c-, observing if

or v > gT/A or v < gT/A).(iii) To obtain v and y in the R and S points, from the known

values in A, B and C by interpolation.(iv) To obtain values of v and y in P.


[Figure 8: Boundary conditions in the characteristic equations solving strategy with a regular grid. Source: elaborated by the author based on Chaudhry (2001).]

By observing of Figure 7, Chaudhry (2001) have been recommended that in P1 c- can be cast and achieved S and therefore, yR and vR. Another equation in yP1 and VP1 (flow value) must be complemented with. An equation in the form VP1

A(yP1)

of flow is the amount which, together with the above equation for S, yp, vP can be obtained. In the Point P2 into the downstream boundary, yR and vR, yP2 and vP2 can be located. For this purpose, Chaudhry (2001) has mentioned the need for complementation with the condition downstream, i.e., the boundary condition: key-curve.

2.2 Dimensioning of the surge tank

The surge tank is a reservoir of vertical axis, normally positioned at the end of the low-pressure adduction pipe, and the upstream of the penstock, for the following purposes (Eletrobras, 2000):

• To dampen the pressure oscillations, which propagate through the penstock, water hammer, resulting from the fast closing of the turbine, and

• Storing water to supply the initial penstock flow caused by the new opening of the turbine until the system has been established continuously.

When necessary, the surge tank should be installed as close to the power house in order to reduce the length of the penstock and lessen the effects of water hammer (Eletrobras, 2000).

In order to assess the need of installing a surge tank in a hy-droelectric exploitation, the following characteristics were adopted as shown in the Table 1, for the case study developed in this paper.

[Table 1]: Characteristics adopted for the SHP case study.

Penstock

Order number

Description Symbol Value Units

1 Nominal diameter Dn 1.84 [m]

2Internal diameter D 1.80 [m]

Internal area A 2.54 [m2]

3 Wall thickness e 20.10-3 (20mm) [m]

4 Total length Lcf 920 [m]

5 Gross head Hb 185.56 [m]

6 Discharge of the design Q 2.80 [m3/s]

The calculation of the velocity of the water inside the pipe, v [m/s], was made according to Eletrobras (2000) as shown by the Equation (16). Then the verification of the necessity for a surge tank was made in accordance with the limits established by the in Inequation (17) and Inequation (18).

(16)

(17)

(18)

Where:Lcf: is the length of the penstock [m];Hb: Gross head [m];th: is the acceleration time of flow in the penstock [s];vcf: is the flow velocity in the penstock [m/s];g: is the gravitational acceleration, equal to 9.81 [m/s2].

For th u than 3.0 s, there is no need for an installation of the surge tank. Values for th among 3 and 6 the installation of the surge tank is desirable, but it is not mandatory. For values for th upper to 6.0 s, it is mandatory to install the surge tank. It is emphasized that the constant of acceleration of the flow into the penstock keeps a relationship with the constant of acceleration of the turbine-generator, which must satisfy equally, to the criteria for maximum allowable pressure, as studied in the warmer hammer, and to the maximum permissible overspeed, the last one, in case of load rejection (Eletrobras, 2000).

In order to establishes the dimensioning of penstock, it should be emphasized the economic diameter, De [cm]. Therefore, the economic diameter is the boundary diameter for which an increasing of its size, which would reduce hydraulic losses and consequently would be obtained higher power installed, also would promote increase of the energetic benefit without this outweighs the extra cost associated. According to Eletrobras (2000), once is considered the difficulties of obtaining a formula that considers the exact parameters mentioned above is adopted, the diameter calculated by the Bondshu formula as considered as economic (Equation 19).

(19)

Where:De: is the economic diameter [cm];Q: is the design discharge [m3/s];Ht is the total hydraulic load over the conduit [m], equal to

the sum of the gross head (Hb) plus the pressure due to water hammer [hs].

According to Eletrobras (2000), for SHP designs, it can be assumed that: hs = 0.2.Hb. Therefore, it has been obtained by this assumption that: Ht = 1.2.Hb. By replacing it in the Equation (19) as proceeding, Equation (20) can be obtained. After the calculation of the economic diameter, it must be verified if the maximum allowable speed for each type of pipe, listed in the Table 2, is reached.

(20)

[Table 2]: Maximum allowable speed for each type of pipe. Source: Eletrobras (2000).

Material vmax allowable (m/s)

Steel 5.0

Concrete 3.0


2.2.1 Dimensioning of the simple surge tank with a constant section

In order to ensure the stability of the oscillations of the water level inside the surge tank, this structure must have a cross section with a minimum internal area, calculated by the Thoma formula, as shown by Equation (21), as preconized by Eletrobras (2000).

(21)

Where:Ac: is the minimum internal area of the cross section of the

surge tank [m2];v: flow velocity in the adduction pipe [m/s];g: gravitational acceleration, equal to 9.81 [m/s2];Lta: length of the adduction pipe [m];Ata: internal cross-sectional area of the adduction pipe [m2];Hmin: minimum head [m];hta: head loss in the adduction system between the water

outlet and the surge tank [m].

The height of the standpipe (Hc) is determined according to Eletrobras (2000) by the oscillation level water inside it, by disregarding the losses in the adduction system or by considering the losses in the adduction system. In the first case, it can be calculated from the elevation (Ye) of the maximum static water level and the depletion (Yd) of the minimum static water level by Equation (22); and in the second case it can be calculated by using the Equations (23) to (25), according to Eletrobras (2000).

(22)

Ye = ze.Ye, where: (23)

(24)

(25)

Where:k: is the relative load loss;hta: is the head loss in the adduction system between the

water intake and the surge tank [m], with the load loss due to friction in the pipeline (ha) calculated for smooth walls: ka equal to 0.32 (Scobey) or ka equal to 100 (Strickler), as shown in Eletrobras (2000).

- Calculation of Yd:

For the calculation of depletion, it is necessary to determine which of the two cases would be the most unfavorable from the following situations:

(1) Depletion consecutive at the maximum lift, due to the closing total (100%) of turbine; or

(2) Depletion resulting from partial opening of 50% to 100% of the turbine.

For the first (1) verification, the procedure is as follows:

Calculation of Yd = zd.Yd.

The coefficient is obtained from the graph of the Figure 9, and Table 3, based on graphs from M.M. Calame and Gaden (apud Eletrobras, 2000), by entering with the parameter that is called as k’ (Equation 26). For the second verification the table 4 must be used in order to obtain k’.

(26)

Where:h’ta: is the head loss in the adduction system between the

water intake and the surge tank [m], as a load loss due to friction in the pipeline (h’a) calculated for the rough walls: ka equal to 0.40 (Scobey), or ka equal to 80 (Strickler), as shown in Eletrobras (2000).

[Figure 9: Curve zd=f(k’) as shown in the Eletrobras (2000).]

[Table 3]: Characteristics adopted for the SHP case study.

k’ 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0.00 1.000 0.982 0.964 0.946 0.928 0.910 0.895 0.881 0.866 0.8520.10 0.837 0.823 0.809 0.794 0.780 0.766 0.755 0.744 0.734 0.7230.20 0.712 0.702 0.692 0.683 0.673 0.663 0.654 0.645 0.637 0.6280.30 0.619 0.611 0.603 0.594 0.586 0.578 0.570 0.562 0.555 0.5470.40 0.539 0.532 0.526 0.519 0.513 0.506 0.500 0.494 0.487 0.4810.50 0.475 0.469 0.464 0.458 0.453 0.447 0.442 0.437 0.432 0.4270.60 0.422 0.417 0.412 0.408 0.403 0.398 0.394 0.390 0.386 0.3820.70 0.378 0.374 0.371 0.367 0.364 0.360 0.357 0.353 0.350 0.3760.80 0.343 0.340 0.337 0.334 0.331 0.328 0.325 0.322 0.319 0.3160.90 0.313 0.310 0.308 0.305 0.303 0.300 0.298 0.296 0.293 0.2911.00 0.289 - - - - - - - - -

Note:The zd values listed in the table are negative.

[Table 4]: Depletion resulting from partial opening of 50% to 100% of the turbine. Determination of the coefficient z’d as a function of k’. Source: Eletrobrás (2000).

k’ 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0.00 0.500 0.504 0.507 0.511 0.514 0.518 0.522 0.525 0.529 0.5320.10 0.536 0.540 0.544 0.548 0.552 0.556 0.560 0.564 0.569 0.5730.20 0.577 0.581 0.585 0.589 0.593 0.597 0.601 0.605 0.610 0.6140.30 0.618 0.622 0.627 0.631 0.636 0.640 0.644 0.649 0.653 0.6580.40 0.662 0.666 0.671 0.675 0.680 0.684 0.689 0.693 0.698 0.7020.50 0.707 0.711 0.716 0.720 0.725 0.729 0.734 0.739 0.744 0.7490.60 0.754 0.759 0.764 0.770 0.775 0.780 0.791 0.791 0.797 0.8020.70 0.808 0.814 0.819 0.825 0.830 0.836 0.848 0.848 0.854 0.8600.80 0.866 0.872 0.878 0.885 0.891 0.897 0.910 0.910 0.917 0.923

0.90 0.930 0.937 0.944 0.952 0.959 0.966 0.980 0.980 0.986 0.993

1.00 1.000 - - - - - - - - -

Note:The zd values listed in the table are negative.


The height of the surge tank (Figures 10 and 11) is then determined by using the following Equation (25), according to Eletrobras (2000).

Hc = Ye + ye + (YD or YD) + yD + YR (25)

Where;ye and yd ≈ 1.0 [m]: is the increasing in the height of elevation

and depletion for security; andYR: is the maximum depletion of the water level of the

reservoir [m]. In this case study, as considered was 0.0 m.

[Figure 10: Surge tank for run-in-the-river SHP as shown in the Eletrobras (2000).]

[Figure 11: Surge tank for SHP with small daily regulation (Yr depletion) as shown in the Eletrobras (2000).]

3. DATA OF A CASE STUDY

A spreadsheet in Microsoft® Excel® for modeling water hammer, as proposed in Chaudhry (2001) and presented by Streeter and Wylie (1982) has been developed to conduct a simple case study which has been aimed at valve-closing at the end of downstream, with a reservoir an extreme upstream without level variation (Figure 12). The valve-closing equation was specified by CdAv/(CdAv)0 = (1-t / tc)m where tc was the closing-time, whose value ranged from 4.0 to 12.0 s; m=3.2; L=920 meters; D=1.80 meters; f=0.019, ν0= 1.10 m/s and H0=185.56 meters. For these calculations equal to 301.80 meters and s were used.

Equations (4) and (6) have been solved by the MOC using numerical grid (Streeter and Wylie, 1982; Chaudhry, 2001). At the upstream and downstream boundaries (reservoir and valve), a device-specific equation were used (considering a constant level of reservoir) and instead one of the MOC water hammer compatibility equations. This has been proposed by Chaudhry (2001) and has been studied for several authors (Karadžić et al., 2010; Izquierdo and Iglesias, 2002; Afshar et al., 2010; Tian et al., 2008).

[Figure 12: Small Hidro Power Plant system of the case study.]

4. RESULTS

4.1 Water hammer results

The pressures over the valve (mca) for the various closing times (between 4s and 12s) have been presented in the graph in Figure 13. Table 5 shows the pressure over values for the first two peak pressures observed in Figure 14, referring to the times of 1.65s and 3.15s.

[Figure 13: Variation of pressure over the valve.]

The calculation of pressure and depression for t = 4s.Overpressure: (+hs) resulted in a pressure with value of pi

equal to 217.26 mca (or 21.73 kgf/cm2);Depression: (-hs) resulted in a pressure with value of pi equal

to 158.37 mca.Then, the calculation of the thickness of the penstock, by using

the Equations (27) and (28), according to Eletrobras (2000), as following shown.

(27)

(28)

Where:σƒ: 1,400 kgf/cm2;kƒ: is equal to 0.80; andes: is equal to 1.00 mm


In function of values found for emin, it was concluded that the minimum allowable thickness of 5.77 mm, would not be properly adopted to the wall thickness of the penstock. But, as can be seen in the Table 1, the value that was adopted as 20mm could be considered as proper, by considering that the minimum thickness of the penstock should be 18.46 mm.

[Table 5]: Pressure over a valve as a function of the valve closing time.

Valve closing time (s)

4 5 6 7 8 9 10 11 12

Elapsed time (s)

1.65 217.26 211.71 207.32 203.82 201.01 198.70 196.79 195.17 193.79

3.15 209.02 206.61 203.43 200.16 197.08 194.30 191.81 189.59 187.63

Reduce the pressure for tc = 4.0 s

1.65 2.55% 4.54% 6.18% 7.48% 8.54% 9.42% 10.17% 10.80%

3.15 1.15% 2.68% 4.24% 5.71% 7.04% 8.24% 9.29% 10.24%

[Figure 14: Pressure over valve for elapsed times of 1.65 s and 3.15 s, depending on the closing times between 4.0 s and 12.0 s.]

The measurement of peak pressure over the valve is seen to be reduced for the valve closing-time 4s to 12s, as increases of the time value, can be observed from the graphs of Figures 12 and 13, as well as in Table 5, especially for two first peaks as observed in the 1.65 s and 3.15 s elapsed times after valve-closing. Respectively, for such values of time after valve-closing, i.e., 1.65 s and 3.15 s, for a closing time of 4s, values such as 217.26 mca and 209.02 mca were been obtained. These values for a valve closing-time of 12s would be 193.79 mca and 187.63 mca, which would represent a decrease, in relation to the closing-time of 4s, of respectively 10.80% and 10.24% of the pressure over valve (in mca). The behavior of discharge values is much milder for longer periods of valve-closing as can be seen from the graphs of Figures 15 to 17.

Negative values of discharge for smaller valve closing-times, for example, for the closing-time of 4s (Figure 14) it have been presented negative values of up to -3.53 m3/s (4.35s elapsed after valve-closing), while for the closing time of 12s, the smaller value that has been obtained would be 1.50 m3/s (also, 4.35s elapsed after valve-closing). In the middle of the tube (Figure 15) for closing time of 4s, the minimum flow rate has been -4.67 m3/s (3.75s elapsed after valve-closing) and for the closing time of 12s, the value would be -0.13 m3/s (also, 7.05s elapsed after valve-closing).

[Figure 15: Pressure over valve for elapsed times of 1.65 s and 3.15 s, depending on the closing times between 4.0 s and 12.0 s.]

[Figure 16: Discharge in the middle of pipe, in terms of closing times between 4.0 s and 12.0 s.]

[Figure 17: Discharge at the end of pipe, in terms of closing times between 4.0 s and 12.0 s.]

4.2 Surge tank dimensioning

– Calculation of speed of the water inside the pipeConsidering, according to values that have been presented

in Table 1:Internal diameter, D = 1.80 mThe internal area, A, resulted in a value of 0.2545 m2. By using

of the Equation (16), it can be obtained the value for speed, v [m/s].


– Verification of the need for a surge tank by using the Inequations (17) and (18). By considering, according to Eletrobras (2000) the head losses as a maximum as 3% of the gross head, the value for the head, H, which can be obtained is 180 meters. But, in the case of Inequation (17) the value to be considered is Hb, i.e., the gross head.

It can be verified that the value for the th has been satisfied, but the relationship between Lcf and Hb could reach and super this one as described by Eletrobras (2000). Then, the decision for dimensioning the surge tank was taken, in order to assurance the safety, in the cases of load rejection.

As above mentioned, the total head losses in the pipeline system as considered as a maximum allowable value (3% of Hb) by Eletrobras (2000), i.e., resulting in a value for these losses, ht, as 5.56 to which can be obtained is 180 meters.

– Determination of minimum internal area of the cross sectionAssuming the “run-in-the-river” exploitation, the water level

in the reservoir does not change, the by using of Equation (21), it can be obtained.

Then, it can result in a diameter, Dc, with the value of 0.43m.In order to reduce the height of the surge tank, also for

economic reasons or to make its construction easier, it is possible to increase its cross-sectional area, subsequently by adding to the internal diameter. Then, for purposes of better construction, it is suggested a diameter, Dc, with 1.60 meters, resulting in an area of 2.00 m2.

For this diameter, the height of the surge tank will be obtained according to the following procedure (Equations 22 to 25).

– Determination of water oscillation inside the surge tank. Maximum elevation of water, for a 100% closing, by

considering the load losses:

Ye = ze . Ye, where:

– Consecutive depletion of water after closing for a maximum elevation of 100%, by considering the losses (Equation 26).

By entering with k’ equal to 0.4628 in Table 3 or Figure 8, it can be obtained zd equal to 0.386 (Figure 18).

(b) Depletion resulting from partial opening of 50% to 100% of the closing device.

By entering with k’ equal to 0.4628 in Table 4, it can be obtained z’d equal to 0.712 by using the interpolation for partial closing of the device as 50%, among the values of 0.684 and 0.729.

[Figure 18: zd obtained as a function of (k’) for the case study for a consecutive depletion of water after closing for a maximum elevation of 100% according to procedures preconized by Eletrobras (2000).]

Yd = zd.Yd = 0.50x12.02 = 6.01 m (a)

Y’d = z’d.Yd = 0.712x12.02 = 8.56 m (b)

This identifies the depletion Y’d, with partial opening of 50% to 100%, as more favorable than the depletion Yd, immediately after the elevation with 100% closing.

– Calculation of height of the surge tank, according to Equation (25), which results are show in Figure 19.

Hc = Ye + ye + (YD or Y’D) + yD + YR

Hc = 8.60 + 1.00 + 8.56 + 1.00 + 0.00 = 19.16 m

[Figure 19: Dimensions of the surge tank obtained for the case study.]

This makes possible to suggest that the surge tank should be located a distance less than 900 meters, in order to reduce the thickness of the pipeline forced, aiming to become cheaper to the acquisition of it. Also, a larger diameter could still be considered in order to prevent a so slender concrete structure. Maybe, the dimensions of the surge tank would be reduced, which would lead also to cheapen the SHP total cost.


5. CONCLUSIONS

A method of characteristics (MOC) has been used in this paper to simulate the response of a pipe system upstream of power plants in the case of valve closing. The case study was a fictional Small Hidro Power (SHP) Plant which has been aimed at valve-closing at the end of downstream, with a reservoir an extreme upstream without level variation. The valve-closing value ranged from 4.0 to 12.0. As results, the measurement of peak pressure over the valve is seen to be reduced for the valve closing-time 4s to 12s, as increases of the time value. Values of time after valve-closing, i.e., 1.65 s and 3.15 s, for a closing time of 4s, values such as 205.49 mca and 204.01 mca were been obtained while for a valve closing-time of 12s these values would be 192.12 mca and 191.19 mca. Behavior of discharge values is much milder for longer periods of valve-closing also have been observed through these simulations.

The benefits that are obtained on reducing the peak pressure and the minimum (negative) discharge are greatly justified, since both could reach a value of zero flow at the exit of the tube with lower possible of damages on the pipe by pressure values.

The surge tank dimensioning resulted with a diameter of 1.60m and total height of 19.16m. Maybe another distance valve-surge tank less than 920.00 meters, as well a greater diameter could be considered in this dimensioning, in order to become, respectively, the size of this surge tank lower, and to prevent a so slender concrete structure.

As recommendation, the authors suggest a validation with systems both, actual and laboratory scales, which will help to produce more realistic results. Also, it is recommended that the assessment of the surge tank be made by considering the lower distance from this surge tank to the valve than 920 meters, in order to become the size of this last one lower, and, subsequently, cheaper the total cost of the SHP.

6. AKNOWLEDGMENTS

We are enormously grateful to Prof. Dr. Fazal Hussain Chaudhry for their valuable contributions and to the National Center for Small Hydropower Plant (CERPCH, in Portuguese) by providing any information when necessary.

7. REFERENCES

[1] AFSHAR, M.H.; ROHANI, M.; TAHERI, R. (2010). Simulation of transient flow in pipeline systems due to load rejection and load acceptance by hydroelectric power plants. International Journal of Mechanical Sciences, v. 52, n. 1, p 103-115. Disponível em:< http://www.sciencedirect.com/science/article/pii/S0020740309002161>. Acesso em 30 jun. 2011.

[2] BARROS, R.M.; TIAGO FILHO, G. L.; SILVA, F. G. B. (2011) Hydraulic transitory study in the small hydropower by characteristics method. IN: 1st. HYDROVIOSION BRAZIL. Rio de Janeiro, Brazil. Proceddings... Penwell, Rio de Janeiro, 2011.

[3] BORTONI, E. C. (2008). New developments in Gibson’s method for flow measurement in hydro power plants. Flow Measurement and Instrumentation, Vol. 19, n. 6, p. 385–390. http://dx.doi.org/10.1016/j.flowmeasinst.2008.06.003

[4] CHAUDHRY, F. (2001). Hydrology: quantitative aspects. Class notes. São Carlos.

[5] ELETROBRAS. The Brazilian Electric Power. (2000). Guidelines for Studies and Projects of SHPs. 458 p.

[6] IZQUIERDO, J.; IGLESIAS, P. L. Mathematical Modelling of Hydraulic Transients in Simple Systems. Mathematical and Computer Modelling, v. 35, p. 801-812

[7] KARADŽIĆ, U.; BERGANT, A.; P VUKOSLAVČEVIĆ, P. (2010). Water hammer caused by closure of turbine safety spherical valves. IN: 25th IAHR Symposium on Hydraulic Machinery and Systems IOP Publishing; 2010. IOP Conf. Series: Earth and Environmental Science 12 (2010) 012096 doi:10.1088/1755-1315/12/1/012096.

[8] STREETER, V. L.; WYLIE, E. B. (1982). Fluid mechanics. 7. São Paulo: McGraw-Hill do Brasil. 585 p.

[9] TIAN, W.; SU, G.H.; WANG, G.; QIU, S.; XIAO, Z. (2008). Numerical simulation and optimization on valve-induced water hammer characteristics for parallel pump feedwater system. Annals of Nuclear Energy, v.35, n.12, p. 2280-2287.

[10] TUCCI, C. E. M. (1998). Modelos Hidrológicos. Porto Alegre: Ed. Universidade / UFRGS / Associação Brasileira de Recursos Hídricos.


an approach to the main problemS faced when developing environmental StudieS in brazil

1Maria Rita Raimundo e Almeida, 2Maria Inês Nogueira Alvarenga

1Environmental Engineer, PhD, adjunct Professor at the Agricultural Sciences Institute at Universidade Federal de Uberlândia, Uberlândia, Minas Gerais, Brazil. Universidade Federal de Uberlândia, Av. João Naves de Ávila, 2121, Uberlândia, Minas Gerais, Brazil, 38408-100, [email protected] Engineer, PhD, associeted Professor at the Natural Resources Institute at Universidade Federal de Itajubá, Itajubá, Minas Gerais, Brazil. Universidade Federal de Itajubá, Av. BPS, 1303, Itajubá, Minas Gerais, Brazil, 37500-903, [email protected].

ABSTRACT

Environmental Impact Assessment (EIA) is a set of procedures which identifies future consequences of a proposed project. The EIA was introduced in Brazil by the 1981 Federal Law n° 6938, which considered it as one of the tools of the National Environmental Policies; however, the technical criteria and the general guidelines were established by CONAMA Resolutions n° 1 of 1986 and n° 237 of 1997. The EIA is a tool recently put into use, which underwent an evolution process along the years, but still needs enhancements. In this context, the current work aims at carrying out a bibliographic survey of the main problems found in studies that make up the Environmental Impact Assessment and the Environmental Licensing Process in Brazil. As a result, it was observed that the deficiencies found in the early EIA years still exist consisting in obstacles which must be overcome so that EIA be effectively implemented.

KEYWORDS: Environmental Impact Assessment, Environmental Study, deficiency.

1. INTRODUCTION

Since its appearance on Earth, mankind has caused changes in the environment in order to ensure survival. With population growth and Industrial Revolution, these changes have intensified, becoming increasingly significant and began to receive special attention. Consequently, understanding the changes that will be caused by an action, searching for more environmentally viable alternatives and considering these issues in the decision-making process have been the subject of several studies in order to minimize the degrading action of man over nature and decrease the impact.

Environmental impact is the change in an environmental component at a given time and place, which results from an activity, compared with the situation had that activity not begun [1]. Analysis and measurement of these impacts are made by Environmental Impact Assessment (EIA), an instrument of environmental policy which includes a set of procedures that enable a systematic examination of the environmental impacts of a proposed action and its alternatives [2].

Among the EIA procedures is the preparation of environmental studies, considered by Araújo [3] as the core step. The kind of environmental impact statement (EIS) used in Brazil for environmental licensing of enterprises which cause significant impacts is called “Estudo de Impacto Ambiental” and should include: project background and overview; baseline description of physical, biotic and anthropic components; delimitation of the influence areas; environmental effects assessment - impacts identification, prediction and determination of relevance; proposition of environmental mitigation measure and monitoring program; and follow-up of the measures proposed [4]. A reliable EIS presents to users outcomes and conclusions that cover all the tasks of assessment appropriately, using pertinent methods for collecting, analysing and reporting information [5].

Several researches have focused on the evaluation of effectiveness of the environmental impact statement as an indicator of the performance of the EIA process [6], and the poor quality of environmental impact statement pointed out as one of the major problems of applying this instrument [7]. However, it is verified a shortage of work in this field in Brazil [8]. Thus, this

work aims at conducting a literature survey of the main problems found in the studies that make up the process of environmental impact assessment and environmental licensing in Brazil.

2. HISTORY OF THE ENVIRONMENTAL IMPACT ASSESSMENT (EIA)

EIA terminology was introduced in 1969 by the United States by means of the National Environmental Impact Assessment Act (NEPA), which required the preparation of a study on the environmental impact caused by government actions. The second country to introduce the EIA process was Canada [9], with a Council Resolution in 1973, where all the projects proposed by federal agencies and government sponsored should be subjected to the Environmental Assessment and Review Process (EARP).

In Europe, EIA introduction occurred in 1985 with the publication of Directive 85/337/EEC, where certain public and private projects had to be assessed concerning their environmental effects.

As has occurred in other developing countries, the first EIA requirements in Latin America have been designed by international financial agencies [10], such as Inter-American Development Bank and International Bank for Reconstruction and Development, a World Bank institution.

The first environmental studies in Brazil were prepared for big hydropower projects during the 1970s. EIA was carried out for Sobradinho dam and hydroelectric plant (state of Bahia), Tucuruí hydroelectric plant (state of Pará) and Ponta da Madeira harbour (state of Maranhão). Since there were no Brazilian environmental standards, the studies were conducted according to the norms of international agencies.

In 1977, the state of Rio de Janeiro created the pioneering regulation for the use of EIA in Brazil. The act of regulating the Licensing System of Polluting Activities established provisions for the State Commission for Environmental Control to require, when deemed necessary, the preparation and submission of the Environmental Impact Report to get any license.

At federal level, we highlight: Decree Law 1413/1975, which introduced the legal zoning of the critical pollution areas [9]; and Law 6803/1980 - Law of Industrial Pollution Zoning in critical


areas, which established the need for the presentation of ‘special studies of alternatives and impact assessments’ to areas of strictly industrial use.

However, EIA was officially introduced by Federal Law 6938/1981 which saw it as an instrument of the National Environmental Policy. The technical criteria and general guidelines for the preparation of the EIS, however, were established only by CONAMA Resolution No. 001/1986, when the EIA was adopted by all Brazilian states. CONAMA Resolution No. 237/1997 is also important in the EIA regulation in Brazil, providing for environmental licensing; Union, States and Municipalities competences; and list of activities subject to environmental licensing.

3. ENVIRONMENTAL IMPACT STATEMENT PROBLEMS

EIA practice is far from what the law requires, although the EIA benefits are undeniable [11]. Therefore, we show a literature review of EIS problems in chronological order. The choice of using the chronological order is to emphasize the progressive aspect in improving studies quality and effectiveness, because it is expected that quality improves over time, as both teams (the one that prepares it and the one that analyses it) gain experience [12].

At the beginning of EIA implementation, the EIS was constituted by extensive and overly descriptive documents, intended to ratify a decision already taken. This did not allow recognizing important details and recommendations that can contribute to minimize the negative environmental impacts [13].

EIS prepared during the first five years of the CONAMA Resolution No. 001/1986 neglected the project alternatives, presented superficial monitoring plans, generic measures and no technical procedures for identification and prediction of impacts [14]. Encyclopaedic character, disregard of cumulative effects and tendency for a broader description of environmental elements to be affected by the proposed project instead of identification and assessment of impact, were found as EIS problems [15].

Among several problems encountered in relation to EIS preparation, the lack of appropriate analysis of design alternatives, environmental risks, cost/benefit relation should be highlighted; and methods of prediction and evaluation of environmental impacts are adapted from other countries and not suitable to the Brazilian reality [16].

Zanzini [17], analyzing biotical and natural ecosystems parts in EIS approved in Minas Gerais State, between the years 1986/1999, concluded that these neither satisfactorily met the legal requirements nor the basic technical recommendation issued for conducting studies. On the legal aspect, the main flaws were related to baseline description and to the proposition of mitigation measures of negative impacts and monitoring programs. On the technical aspect, among major flaws were ignoring the season, species vulnerability analysis and application of ecological indexes and multivariate analyses.

For Dias [18], serious problems of organization and language and lack of precision in the formulation of measures occur in EIS preparation, which requires subjective assessments by government agents, subject to controversy and contestation. In the case of mining, in general, the EIS does not establish a direct link between identified environmental impacts and proposed actions for its minimization [8].

In 2004, Federal Public Ministry carried out a survey of the main weaknesses identified in each component of the studies involved in EIA process in Brazil which are shown in Table 1.

Glasson, Therivel and Chadwick [20] point out as failure of EIS: inadequate monitoring of positive impacts that have not happened in practice, but counted in the decision-making

of project environmental feasibility, and that have created expectations in the affected population.

Silveira [21], assessing studies about biotical environmental in EIS, found that among the variables that showed flaws are: habitat description, description and quality of the methods used and quantity of species identified.

In his studies, Caldas [22] pointed out that analysis of alternatives was deficient, since it did not present details of other possible alternatives; baseline description was insufficient to characterize the local reality of the enterprise, because of doubts about effectiveness of the methodology used; environmental impacts have been addressed in a generic way, without taking into account the actions and specific characteristics of the project and of the area, and there was no assessment of cumulative and synergistic effects; mitigation measures based on the results obtained have not been prepared. They were not correlated with the proposed environmental programs and there was no definition of time and cost.

[Table 1]: Environmental Impact Statement weaknesses. Source: based on MPF [19].

EIS part Weaknesses

Project objectives Adoption of the set of interdependent work objectives as a justification for approval of just one of the sections or one of the projects.

Analyses of alternatives Absence of proposition of alternatives/Prevalence of economics over environmental aspects in choosing alternatives.

Delimitation of Influence Area (IA)

Disregards the watershed/IA elimination not supported by characteristics and vulnerability of regional natural and social environments.

Baseline description Deadlines insufficient for field research/Environmental characterization based on secondary data/Poor information about the methodology used/Lack of integration of data from specific studies.

Baseline description of physical and biotic aspects

Submission of inaccurate or contradictory information/Weaknesses in sampling/Incomplete characterization/Insufficient data about flora and certain groups of organisms/Absence of diagnostic on animal breeding and feeding sites.

Baseline description of social-economic aspects

Methodologically ineffective searches/Poor knowledge of native communities’ lifestyles /Generic socio-economic characterizations.

Impacts identification, prediction and analysis

Absence or partial identification of impacts /Under-utilization or disregard of the diagnostic data/Trend to minimize negative impacts and to overestimate positive impacts/No assessment of cumulative and synergistic impacts.

Compensatory and mitigating measures

Proposition of measures that do not address the impact/Little detailing/Indication of technical and legal obligations as mitigating measures/No assessment of efficiency/Absence of detailed information about proposed financial resources.

Environmental programs Insufficient monitoring efforts/Stipulation of monitoring periods incompatible with periods of occurrence of impacts.

Non technical summary Inappropriate language/EIS results distortion in order to decrease negative impacts.

For Aguilar [23], the analysis of location alternatives is treated precariously and its importance is underestimated by the entrepreneurs. Moreover, the economic criteria prevail in the choice of location alternatives over the environmental criteria.


Xingó hydroelectric dam EIS had several flaws, including inadequate demarcation of the influence area, which did not consider the watershed, and resulted in negative impacts not foreseen [24].

The quality of Preliminary Environmental Reports (a kind of EIS used in São Paulo state) of pipelines Lins/Marília and Bauru/Agudos/Pederneiras were regarded as unsatisfactory because they did not meet legislation requirements, the percentages of failure were above 80% and the studies were inconsistent with best practices, except for some specific issues [25].

In EIS for landfills in the state of São Paulo it was observed that the main formulated measures were vague and imprecise and should be implemented soon, independently from EIA process, since they are intrinsic to landfills technology. In baseline description, despite contextualizing the influence areas, integration and contextualization were omitted [26].

For Sandoval and Cerri [27], the main problems are the lack of impacts identification and omissions regarding methodology and criteria. They also highlight technicians’ difficulties in written communication.

Gomes et al. [28] analyzed Mucuri and Santo Antonio do Porto small hydropower EIS, and concluded that these studies did not present projects environmental feasibility. They found that neither of them includes the impacts of the decommissioning phase of project and that baseline description was made by an exhausting approach, which does not guarantee that all information collected is relevant.

To facilitate the visualization of problems reported, Table 2 presents a summary of deficiencies found and respective authors. It is noticed that problems are found in every part of environmental studies. In addition, several authors have pointed out that, in many cases, the preparation of environmental studies is part of a process in which projects are already defined and studies have only a documentary effect.

According to Table 2, the deficiencies most found in this review are:

1. no proposition of alternatives, making it impossible to select the most environmentally viable alternative;

2. baseline description is superficial and incomplete, without characterizing comprehensively and objectively the project influence area;

3. proposition of measures: usually, the best solution to impacts is not indicated and (usually) the description of measure is incomplete.

Interestingly, the deficiencies that were found in early EIA processes still persist, constituting obstacles to be overcome and that deserve special attention in order the purpose of implementing EIA can be guaranteed.

4. CONCLUSION

Environmental Impact Assessment is an important tool to put the environmental issue in the decision-making process. Due to its recent implementation, EIA process and environmental studies development still need improvements to fully achieve their objectives.

Researches to evaluate environmental studies quality are extremely important, since by pointing out the weaknesses, they make the formulation of recommendations and changes that improve the EIA process possible.

According to the review carried out among main shortcomings of the environmental impact statement are: absence of proposition of alternatives, superficial and incomplete baseline description and inadequate proposition of measures.

[Table 2]: Mean monthly discharge (m3/s) at the natural flow regime and at the different flow scenarios.

Weaknesses Authors

Technological and location alternatives

Absence of proposition of technological and location alternatives

Agra Filho [14]; Salvador [16]; MPF [19]; Caldas [22]; Aguilar [23]; Gomes et al. [28]

Influence area

Inadequate influence area delimitation

MPF [19]; Santos [24]

Baseline description

Superficial and incomplete baseline description

Zanzini [17]; Silveira [21]; MPF [19]; Caldas [22]; Santos [26]; Gomes et al. [28]

Absence of information about used methodology

MPF [19]; Caldas [22]

Insufficient deadline for such studies Zanzini [17]; MPF [19]

Environmental impacts

Superficial analysis of impacts Ronza [29]; MPF [19]; Caldas [22]

Inadequate identification and assessment of impacts

Salvador [16]; Sandoval and Cerri [27]; Gomes et al. [28]

Inadequate prediction of impacts Agra Filho [14]; Salvador [16]; Santos [24]

Omission of methodology and of criteria adopted

Sandoval and Cerri [27]

Insufficient treatment of cumulative impacts

Bursztyn [15]; MPF [19]; Caldas [22]

Inadequate risk assessment Salvador [16]

Failure monitoring impacts Glasson, Therivel and Chadwick [20]

Mitigating, maximizing and compensatory measures

Proposition of measures (measures that are not a solution to the impact, incomplete measures)

Agra Filho [14]; Zanzini [17]; Dias [18]; Prado Filho and Souza [8]; MPF [19]; Caldas [22]; Santos [26]

Mitigation measures implementation Santos [26]

Monitoring programs

Proposition of monitoring programs (incomplete or superficial, without defining parameters to be monitored)

Agra Filho [14]; Zanzini [17]; MPF [19]

Other aspects

Long and overly descriptive documents

Pádua [13]; Bursztyn [15]

Organizational problems and texts’ language

Dias [18]; Sandoval and Cerri [27]

5. REFERENCES

[1] Wathern, P., 1988, “An introductory guide to EIA”. In: Wathern, P., (Org) “Environmental impact assessment: theory and practice,” London, Unwin Hyman, pp.3-30.

[2] Moreira, I .V. D., 1992, “Vocabulário básico de meio ambiente,” Feema, Rio de Janeiro.

[3] Araújo, M. G., 2004, “Políticas Públicas de Meio Ambiente,” Especialização em Gestão Ambiental, UNIVIX.

[4] CONAMA – Conselho Nacional do Meio Ambiente (Brasil), 1986, “Resolução n° 001, de 23 de janeiro de 1986,” Diário Oficial da União, Brasília.

[5] Lee, N., 2000, “Integrating appraisals and decision-making”. In: Lee, N., and George, C., “Environmental assessment in developing and transitional countries,” John Wiley & Sons, Chichester, pp. 161-175.


[6] Badr, E. A., Zahran, A. A., and Cashmore, M, 2011, “Benchmarking performance: Environmental impact statements in Egypt,” Environmental Impact Assessment Review, 31, pp. 279–285.

[7] Mendes, D., and Feitosa, A, 2007, “ IBAMA reduzirá em mais de 50% prazo para concessão de licença ambiental,” Brasília, www.mma.gov.br/ascom/ultimas/index.cfm?id=4241.

[8] Prado Filho, J. F., and Souza, M. P., 2004, “O Licenciamento Ambiental da mineração no Quadrilátero Ferrífero de Minas Gerais – uma análise da implementação de medidas de controle ambiental formuladas em EIAS/RIMAS,” Engenharia Sanitária e Ambiental, 9(4), pp. 343-349.

[9] Antunes, P. B., 2000, “Direito Ambiental,” 4º ed, Lumen Júris, Rio de Janeiro.

[10] Moreira, I. V. D., 1989, “Avaliação de impacto ambiental – instrumento de gestão,” Cadernos FUNDAP, 9(16), pp.54-63.

[11] Milaré, E., and Benjamin, A. H. V., 1993, “Estudo prévio de Impacto Ambiental,” Revista dos tribunais, São Paulo.

[12] Sánchez, L. E., 2008, “Avaliação de impacto ambiental: conceitos e métodos,” Oficina de Textos, São Paulo.

[13] Pádua, M. T. J., 1990, “Estudos e relatórios de impacto ambiental como instrumentos de conservação da natureza,” 1th Seminário sobre Avaliação e Relatório de Impacto Ambiental, Curitiba, pp.9-17.

[14] Agra Filho, S. S., 1993, “Situação atual e perspectivas da avaliação de impacto ambiental no Brasil”. In: Sánchez, L. E., “Avaliação de impacto ambiental: situação atual e perspectivas,” Epusp, São Paulo, pp.153-156.

[15] Bursztyn, M. A. A., 1994, “Gestão Ambiental: Instrumentos e Práticas,” IBAMA, Brasília.

[16] Salvador, N. N. B., 2001, “Análise crítica das práticas de avaliação de impactos ambientais no Brasil,” 21th Congresso brasileiro de engenharia sanitária e ambiental, ABES, João Pessoa.

[17] Zanzini, A. C. S., 2001, “Avaliação comparativa da abordagem do meio biótico em Estudos de Impacto Ambiental no Estado de Minas Gerias,” Ph.D. thesis, Universidade de São Paulo, São Carlos.

[18] Dias, E. G. C. S., 2001, “Avaliação de impacto ambiental de projetos de mineração no Estado de São Paulo: a etapa de acompanhamento,” Ph.D thesis, Universidade de São Paulo, São Paulo.

[19] MPF – Ministério Público Federal, 2004, “Deficiências em Estudos de Impacto Ambiental: síntese de uma experiência,” 4th Câmara de Coordenação e Revisão, Brasília.

[20] Glasson, J., Therivel, R., and Chadwick, A., 2005, “Introducion to Environmental Impact Assessment,” Ed. Routledge, Inglaterra.

[21] Silveira, R. L., 2006, “Avaliação dos métodos de levantamento do meio biológico terrestre em estudos de impacto ambiental para a construção de usinas hidrelétricas na região do Cerrado,” Msc. thesis, Escola Superior de Agricultura Luiz de Queiroz, Piracicaba.

[22] Caldas, F. V., 2006, “Estudos de Impacto Ambiental em empreendimentos dutoviários: análise da elaboração, acompanhamento e monitoramento durante a fase de construção,” Msc. thesis, Universidade Federal Fluminense, Niterói.

[23] Aguilar, G. T., 2008, ‘Análise do tempo de tramitação de processo de licenciamento ambiental: estudo de casos de termelétricas no Estado de São Paulo,” Msc, thesis, Universidade de São Paulo, São Carlos.

[24] Santos, R. G., 2008, “Impactos sócio-ambientais à margem do rio São Francisco: um estudo de caso,” Msc. thesis, Universidade de São Paulo, São Paulo.

[25] Lopes, A. L., 2008, “O relatório Ambiental Preliminar como Instrumento da Avaliação da Viabilidade Ambiental de Sistemas de Distribuição de Gás,” Msc. thesis, Universidade de São Paulo, São Carlos.

[26] Santos, C. N., 2008, “Avaliação das medidas mitigadoras relacionadas ao meio físico, propostas em Estudos de Impactos Ambientais e Relatórios de Impactos Ambientais (EIAs/Rimas) de aterros sanitários no Estado de São Paulo,” Msc. thesis, Universidade Estadual de Campinas, Campinas.

[27] Sandoval, M. S., and Cerri, L. E. S., 2009, “Proposta de padronização em avaliação de impactos ambientais,” Engenharia Ambiental, 6(2), pp.100-113.

[28] Gomes, C. S. et al., 2009, “Avaliação da qualidade de estudos de impacto ambiental de Pequenas Centrais Hidrelétricas,” http://www.ambiente-augm.ufscar.br/uploads/A2-052.pdf.

[29] Ronza, C., 1998, “A política de meio ambiente e as contradições do Estado, a avaliação de impacto ambiental em São Paulo,” Msc. thesis, Universidade Estadual de Campinas, Campinas.

1 American Journal of Hydropower, Water and Environment Systems, july 2016

published by ACTA Editora

issn 2317-126x • us$: 15.00 a j h , water e s · 2019-02-04 · american journal of hydropower,...

Documents