Development and applcaton of a rural water supply assessment tool n Brazl W. T. P. Slva 1,2, A. A. Santos 3 and M. A. A. Souza 1 1 Post-graduate Program on Envronmental Tecnology and Water Resources, Department of Cvl and Envronmental Engneerng, Faculty of Tecnology, Unversty of Brasla, Brasla, Brazl. E-mal: wttatom@terra.com.br (W. T. P. Slva), maasouza@unb.br (M. A. A. Souza) 2 Department of Santary and Envronmental Engneerng, Faculty of Arctecture, Engneerng and Tecnology, Federal Unversty of Mato Grosso, Cuabá, Brazl 3 Envronmental Engneerng, nsttute of Educaton, Agrculture and Envronmental, Federal Unversty of Amazonas, Humatá, Brazl. E-mal: aldecy_allmeda@yaoo.com.br (A. A. Santos) Abstract Te objectve of ts paper s to develop and to apply a rural water supply systems (WSS) assessment tool to typcal cases n Brazl. Te metodology developed comprses te followng steps: (1) a bref descrpton of te tools prevously developed; (2) combnaton and adaptaton of prevously developed tools employng a multcrtera approac; (3) data gaterng and smulaton of real cases; (4) applcaton of te developed tool to typcal cases n Brazl. Adaptatons were done and tey conssted of some ndexes modfcatons (safety of drnkng water and rural WSS structural condton), ncluson of te TOPSS Metod for selecton of te adaptve soluton, and consderaton of local expertse nformaton for defnton of wegts. Some of modfed ndexes were tose for condtons of water reservors and water dstrbuton network. t can be concluded tat te Brazlan rural WSS sowed operatonal and mantenance problems. Te developed tool was consdered as an useful decson support to te plannng of te Brazlan rural WSS. Keywords: rural ouseolds; tecncal assessment; water problem ntroducton Te process operaton s one of te man ponts of an operator of santaton, snce t s drectly ted to ts core competency. ts management s crtcal to success and to ensure contnued market presence. Te vast majorty of carrers of santaton servces operate ter systems wtout a metodology tat enables te trackng and montorng of ter processes, as well as te process of decson makng. Te lack of an operaton metodology brngs some consequences suc as te low qualty of products and servces provded by te servce provder of santaton servces, resultng n te unavalablty of water to te user and terefore publc ealt problems. Te ncdence of problems n water supply systems or te unavalablty of drnkng water can lead to rsky beavor suc as poor water condtonng and lookng for water supples nadequated for consumpton. Ts s especally true for rural areas, were te populaton, n most cases, does not ave any soluton to te system of rural water supply (rural WSS). Ts paper descrbes a toolset for assessment and applcaton to rural WSS n remote areas. n te followng sectons of te paper, te secton 1 presents an ntroducton. Secton 2 presents a bref revew of some relevant lterature. Secton 3 descrbes te metodology. Secton 4 descrbes te results and secton 5 descrbes and sows te conclusons of paper. Revew of related researc n te current secton of te paper, a bref descrpton of some paper related to te subject evaluaton rural WSS are present. Sauer (2003) development a researc s focused on te effcency of water supplyng nfrastructure n rural areas. A cross-sectonal data set was collected wt respect to water companes n rural areas of East and West Germany. After formulatng partal performance ndexes, te relatve cost effcency was estmated by a stocastc fronter analyss. Due to ts analyss, frst (prelmnary) results sow tat te tecncal effcency n te (rural) water sector can be mproved by up to 45% wt te same tecnology and te same ydrologcal and supplyng caracterstcs. o sgnfcant effect on supplers neffcency by regonal locaton was found. Economes of scale are verfed for te sample suggestng te realzaton of effcency gans by te expanson of te average supplyng area. Assumed explct negatve effects on frms effcency by te exstng legal framework of te water sector are ence emprcally confrmed. Te analyss of te determnants of publc fundng reveals tat tere s no sgnfcant nfluence of operatonal effcency on te decson about fundng by publc nsttutons. Fnally t s sown tat frms n East Germany enjoy an easer access to publc funds, regardless ow effcent ter operatons are.
Amaral et al. (2003) assessed te santary qualty of drnkng water n rural farms troug counts of mcrobologcal ndcators. A total of 180 drnkng water samples from sources, reservors and water consumpton stes were collected n 30 rural farms located n te norteast regon of te state of São Paulo, Brazl. Te most probable number of total colforms, Escerca col and mesoplc mcroorgansms were determned. Also, te presence of protecton measures for water supples was verfed. Te study results sowed tat 90.0% of drnkng water samples from sources, 90.0% from reservors, and 96.7% from water consumpton stes, collected durng te rany season, and 83.3%, 96.7% and 90.0% of samples collected n dry season were below te qualty control standards for drnkng water. Te conclusons ndcated tat drnkng water n rural farms was consdered a potental uman ealt treat. Preventve measures for protectng water sources and water treatment would be necessary to sgnfcantly reduce te occurrence of waterborne dseases. Kesavarz et al. (2006) nvestgated te relatonsps between water consumpton and rural ouseold actvtes are determned by comparng a snapsot of water consumpton wt rural ouseold beavor of low, medum and g water consumers. n addton, te factors affectng water consumpton n rural ouseolds are also determned. Te data for ts study were collected from a survey of 653 rural ouseolds n 33 vllages of Ramjerd area, Fars Provnce, n soutern ran, usng a smple random samplng tecnque. Te daly water consumpton data for a 5-year perod (1999 2004) were used. Results of te study revealed tat te daly average water consumpton for te area was found to be 121.7 L per person per capta per day (Lpcd). Water consumpton was also found to be sgnfcantly correlated wt explanatory varables suc as ouseold sze and age of ouseold s ead. Fnally, te results of dscrmnant functon analyss sowed tat n rural ouseolds, garden sze, greenouse sze, and garden waterng tmes per mont wt tap treated water are assocated wt water consumpton. Azevedo (2006) sowed te experence related wt te water supply system mplantaton n te ndgenous vllages of ova Esperança, Vla ova and ossa Senora de azaré of te etnc group Sateré-Mawé, located n te regon of te Marau rver, dstrct of Maués, Amazon State, Brazl. Tese water supply systems began operatng n 1998 and were modfed to aggregate oter elements lnked to te cosmology of tat people and tat ad not been observed n te ntal project. Te study demonstrated te mportance of understandng te ndgenous culture wc allows te effectve partcpaton n te dscussons of te several pases of te project n order to facltate te acceptance and te correct utlzaton of te adopted soluton. Results of te study conducted by Somala Water and Land nformaton Management (SWALM) n rural area Somala, are present by Mutus et al. (2007). SWALM conducted a desk study assessment to determne te water supply stuaton n rural Somala. Te assessment was done n tree regons: Puntland, Somalland and Sout-Central Somala. Te assessment nvolved revew and analyss of avalable data and lterature on exstng water supples. ntervews were conducted n arob, Puntland and Somalland wt key players n te Somal water sector. Feld vsts to some water ponts were also made by two consultants, one n Puntland and anoter n Somalland, to see te condton of te water sources and to ntervew operators and water users. Te fndngs of te assessment were smlar n many aspects for te tree regons, wt only slgt dfferences between tem. n Puntland, te assessment concluded tat: (1) sortages n water supply are not only a functon of lmted water sources, but also as a result of frequent breakdown of te sources; (2) rural communtes n Puntland try to cope wt water sortages by mnmzng te amount of water usage per day and n many cases ts compromses ygene standards, creatng some ealt rsks; (3) te Puntland autorty s unable to support te muc-preferred communty-based water supply projects, makng people more relant on external donors. Te conclusons for Somalland were as follows: (1) sallow well aqufers n rural Somalland are not well developed and many of te sallow wells dry up n te Jlaal and Hagaa seasons; (2) tere are many sources wt water unsutable for uman consumpton, but locals use tem eter out of gnorance or because tey do not ave an alternatve; (3) ygene facltes n rural Somalland are lmted, te people lve n poor santary condtons and water avalablty s explaned n terms of quantty, not qualty. Conclusons for Sout Central Somala nclude: (1) Sout-Central Somala s muc advantaged n terms of surface water resources compared to oter regons of te country; (2) use of water resources for economc development s neffcent due to lack of a water polcy; (3) te Somal communty, due to g scarcty quantty of water avalable s more mportant tan qualty and ts as contrbuted to a large extent to frequent outbreaks of colera and oter waterborne dseases n te regon. Metodology Te metodology developed comprsed te followng steps: (1) a bref descrpton of te tools prevously developed; (2) combnaton and adaptaton of te tools prevously developed employng a multcrtera approac; (3) data gaterng and smulaton of real cases; (4) applcaton of te developed tool to te typcal cases n Brazl.
Te bref descrpton of te tools prevously developed (1) comprsed te descrpton of te tool developed by Retveld et al. (2009). For step (2) te tools prevously developed were combned and adapted for te tradtonal multcrtera approac, by revewng patterns of local water qualty, ncluson of te TOPSS Metod (for selecton of te most adaptve soluton) and also local expertse (for defnton of wegts). Te tool of assessment of rural WSS was developed n Mcrosoft Excel workseet. For step (3) data gaterng and smulaton, te necessary data were gatered from te local operators of rural WSS and local communty. Te ard data were smulated from lterature revew about local communty. n step (4), te developed tool was appled to typcal cases n Brazl, usng data gatered and smulatng nput to te developed tool, and te results were analyzed. Results Te tool for tecncal assessment of rural WSS developed by Retveld et al. (2009) s based on four crtera, namely avalablty, capacty, contnuty and condton. Avalablty pertans to te adequacy of te source n terms of quantty and qualty of te water tat can be obtaned from t. Te avalablty fals wen quantty and qualty are below te standard for drnkng water supply. Te capacty s te adequacy of te storage, transport and dstrbuton to supply water to te communty. Unacceptable capacty wll typcally arse wen a vllage grows n sze, per capta demand and populaton to te pont were a prevously good system wll begn to fal n ts nadequacy. Contnuty s te consstency wereby water s conveyed from te source to te water pont. Water supply can be nterrupted wen a pump s broken, ppes are damaged or taps cannot be opened, among oters. Te condton of a system would reveal more about te expected lfetme tan durablty because t takes nto account te external factors wc play a major role n rural and poor communtes. Te condton s te status of te system n terms of ts servceablty. Te crtera are quantfed by ndexes, wc allow comparson of performance of dfferent rural WSS. Some of te varables tat relate to accounts for te water losses n te rural WSS, turbdty n water, total colforms n water, requred water flow dstrbuton, among oters. Addtonal nformaton about te tool, equatons and varables nvolved, can be found n te artcle. Te adaptaton of te tool developed by Retveld et al. (2009) bascally conssted of modfcatons to te crtera of water avalablty and condton of te rural WSS crtera for ncluson of a socal nature. Wereas te rural WSS systems, ts classfcaton as alternatve water supply, and tat te Brazlan standards of water qualty (Decree 518/2004 of te Mnstry of Healt) for suc a stuaton, tere s te necessty of ncludng some water qualty varables (e.g., resdual clorne free and color n water). For te crtera condton tere s need to nsert some components (e.g., condtons of te resdental reservor). A crteron of socal caracter, dsposton of te populaton to cooperate wt actons to mprove te rural WSS, was nserted nto te tool. A metod of multcrtera decson support, te TOPSS metod, was also avalable at tool for evaluatng te rural WSS n order to elp select te best alternatve workaround for te rural WSS. Te equatons consdered n eac of te crtera are presented n sequence. Te avalablty ndex ( av ) s defned by Equatons 1, 2, 3, 4, 5, 6, 7 and 8. av Vmont 1000, quant 1 (1) 30 D D D sp yc c..... 1 (2) av, qual 1 ph 2 TU 3 TC 4 CL 5 CO ph e e TU e TC e. 1 ph7. 2 7pH 3. TU 4. TC CL. CL TC e 5 6 7. CO se ph 7, 0 se ph 7, 0 av 1. av, quant 2. av, qual Were: s te number of people wt access to a standppe or and pump (sp), yard connecton (yc) or ouse connecton (c), D s te mnmum demand related to one te mentoned (n Lppd), V mont s te montly producton (n m 3 /mont), accounts for te water losses n te rural WSS, are constants dependng on te allowable values n te drnkng; and, are wegng factors, were = = 1; ph s te water ydrogenc potental; TU s water turbdty; TC s water total colforms; CL s concentraton free clorne n te water; CO s color n water. (3) (4) (5) (6) (7) (8)
Te capacty ndex s composed by four ndexes: (1) pumpng capacty from te source ( pump ); (2) storage volume to overcome dfferences between demand and pumpng capacty and calamtes ( res ); dstance of water ponts from consumers ( wp,dens ); and, number of water ponts wt too rates ( wp,funct ). Te capacty ndex ( cap ) s defned by Equatons 9, 10, 11, 12 and 13. Qpump pump 1 (9) Q res Q wpdens, demand V demand res T res 1 wp 1 wp,mn wpgood, wpfunct, 1 wptotal, cap 1. pump 2. res 3. wpdens, 4. wpfunct, (13) Were: Q pump s te actual pumpng capacty (n m 3 /); Q demand s te water flow for dstrbuton, ncludng losses n te system (n m 3. -1 ), V res s te nstalled reservor (n m 3 ), T res s te requred mnmum storage tme (n ), wp s te actual water pont (wp) densty (n wp/km 2 ); wp,mn s te mnmum requred water densty (n wp/km 2 ), wp,good s water pont wt a suffcent flow rate, wp,total s te total number of water ponts and are wegng factors, were = 1. Two ndexes caracterzed te crteron of contnuty: te number of ours per day of unplanned nterrupton of water supply to te ouseolds ( ours,day ) and te number of days per mont wtout unplanned water supply ( day,mont ). Te contnuty ndex ( cont ) s defned by Equatons 14, 15 and 16. (10) (11) (12) day e, mont T T d d, max 1 sup ours, day 1 sup, mn cont 1. day, mont 2. ours, day (14) (15) (16) Were: T sup s te supply per day (n ); T sup,mn s te mnmum, predetermned supply tme per day (n ); d s te number of days per mont wtout water, wle water was avalable; d,max s te maxmum allowable number of days per mont wtout water, and are wegng factors, were = 1; s factor depends on local crcumstances and nsgts. Te elements consdered to ndex of condton are: damage and/or leakage of te tap ( tap ); damage and/or leakage of te resdental reservor ( ouse,res ); and, damage and/or leakage of te water net ( water,net ). Te condton ndex ( cond ) s defned by Equatons 17, 18, 19 e 20. tap tap, good (17) ouse, res water wptotal, ouse, res, good ouse, res, total 0,1 f condtonsare bad, net 0,5 f condtonsare reasonable (19) 1,0 f condtonsare good. (20) cond 1. tap 2 ouseres, 3. water, net (18) Were: wp,total s te total number of water ponts n te communty; ta,good s te number of taps tat are s good state; ouse,res,good s te number of resdental reservor n good state n te communty; ouse,res,total s te total number of resdental reservor state n te communty; and, are wegng factors, were =1. Two ndexes caracterzed te crteron of socal caracter. Te frst ndex s te percent of famly tat sows good acceptance for te practce of water conservaton ( acceptance ), and te second s te percent of famly
tat perform exstng best-practce for water conservaton ( good,practce ). Te socal ndex ( socal ) s defned by Equatons 21, 22 and 23. acceptance famly,acceptance famly, total famly,practce good, practce famly, total socal 1. acceptan ce 2. good,practce (23) (21) (22) Were: famly,acceptance s te number of famly tat ave good acceptance te practce of water conservaton n te communty; famly,total s te total number of famly n te communty; famly,practce s te number of famly tat current exstng best-practce; s of wegng factors, were =1. To te judge te dfferent rural WSS n te selected, te followng scema was used: bad 0.00 < < 0.50; far 0.50 < < 0.75; suffcent 0.75 < < 0.90; and, good 0.90 < < 1.00. Havng te knowledge of te caracterstcs of rural WSS, alternatves for tool mprovement can be proposed. Wen lookng for ts, multcrtera decson support metods are useful. n ts way, te TOPSS metod (Tecnque for Order Preference by Smlarty to deal Soluton), developed by Hwang and Yoon (1981), was coupled to te same envronment as te assessment tool of rural WSS. Te use of TOPSS as appeals as smplcty (ease of applcaton) and te mode ow t faces a decson problem by comparng two ypotetcal stuatons: deal and undesrable. Ts approac s based on te tecnque of order preference by smlarty to deal soluton. Te deal soluton s defned as tat ave a smaller dstance (or gest smlarty) compared to deal soluton (A+) and also olds te greatest dstance (lower smlarty) n relaton to undesrable soluton (A ). Ts metod s ratonal and compreensble and evaluate te decson matrx n 6 steps: (1) normalzed matrx representng te relatve performance of alternatves, (2) create standardzed and wegted matrx, (3) provdes an deal soluton wt te aggregaton process, (4) computes a n- dmensonal measure of separaton, by usng te Eucldean dstance for all alternatves, (5) calculates te relatve closeness to te deal soluton, and (6) orders te alternatves accordng to decreasng values. Te Equatons 24, 25 and 26 present te dstances equatons. d A d A n 1 n 1 w w * * f f f f x * f x ** f f ** f ** da A da 1 1 d A, A (26) d Were: d(a+) s te dstance between te deal soluton vector and te vector of te studed case; w s te wegng of te consdered crtera, w = 1; f * s te value of te deal soluton n te crtera ; f (x) s te value s te performance of te studed alternatve n te crtera ; f ** s te value of te undesrable soluton n te crtera ; s te parameter of sensblty verfcaton, te value s 1, 2 and ; s dentfcaton of te crtera n te study; d(a ) s te dstance between te undesrable soluton vector and te vector of te study case; d(a+,a ) s te dstance among te deal soluton vector and te undesrable soluton vector, and te vector te studed case. n ts researc, suggested crtera nclude: cost of nvestment ( = 1); reducng water consumpton ( = 2); reducton of water losses ( = 3); tecnologcal level ( = 4); and acceptablty of te populaton ( = 5). Te alternatves desgned to mprove te rural WSS are closely related to caracterstcs of eac case. However, wen consderng te man desgn elements of rural WSS (water extracton n deep wells, dsnfecton by clorne soluton, elevated water storage reservor, dstrbuton water net and ouseold connectons), presented by Azevedo (2006), and te ypotess tat tere are problems of mantenance and operaton of rural WSS, some alternatves suggested for mprovement comprses te followng: (1) mplementaton of a government program for qualfcaton of people for operaton and mantenance of rural WSS, (2) replacement of te clornaton treatment system for a clornaton system at ome n order to smplfy system operaton, and (3) replacng te rural WSS for a system composed by surface water use and slow fltraton. As (24) (25)
mentoned earler, tese alternatves must consder te specfctes of eac case and are presented merely as possble suggestons. Te Fgures 1a, 1b, 1c and 1d sows te water collecton and water supplyng nfrastructure n rural WSS. (a) (b) (c) (d) Fgure 1. Communty Cnturão Colna Verde, n Cuabá cty, Mato Grosso state: water collecton n ouse (a); reservor of te communty (b); ppes of water dstrbuton (c); water net n contact wt wastewater (d) n Table 1 are presented te results for some real cases n Brazl. Table 2 summarze te ndexes results for te case studes. Table 1. Some ndexes of fve rural WSS n Brazl Some data nput A B C D E umber of ouseolds* 426 125 90 131 115 Volume reservor (n m 3 )* 30 5 2 5 5 Total number of water ponts* 420 25 22 26 23 Condtons of water net* bad bad bad bad bad Total number famly* 107 36 30 43 38 ph 7,50 5,30 5,00 5,00 5,50 TU 1,0 3,0 2,0 0,5 4,0 TC (n MP.100mL -1 ) 56 0 0 0 0 CL (n mg.l -1 ) 0,0 0,0 0,0 0,0 0,0 CO (n mgpt.l -1 ) 5 2 2 1 2 Water supply per day (n ) 8 3 2 3 3 Pumpng capacty (n m 3. -1 ) 8 2 1 3 3 *Data obtaned by a questonnare n te study stes, by smulaton and lterature revew A s te communty Cnturão Colna Verde, n Cuabá cty, Mato Grosso state; B s te communty pxuna, n Humatá cty, Amazona state; C s te communty Parasno, Humatá cty, Amazona state; D s te communty Parasno Grande, Humatá cty, Amazona state; and, E s te communty São Mguel, Humatá cty, Amazona state
Table 2. Results of te tecncal assessment of fve rural WSS n Brazl Crtera A B C D E Avalablty ndex ( av ) 0,6507 0,8617 0,6656 0,7708 0,9875 Capacty ndex ( cap ) 0,9181 0,9187 0,9427 0,9481 0,9324 Contnuty ndex ( cont ) 0,1990 0,0500 0,0334 0,0500 0,0500 Condton ndex ( cond ) 0,3936 0,1479 0,2333 0,5500 0,4167 Socal ndex ( coop ) 0,3150 0,1583 0,5167 0,8209 0,4079 A s te communty Cnturão Colna Verde, n Cuabá cty, Mato Grosso state; B s te communty pxuna, n Humatá cty, Amazona state; C s te communty Parasno, Humatá cty, Amazona state; D s te communty Parasno Grande, Humatá cty, Amazona state; and, E s te communty São Mguel, Humatá cty, Amazona state We also made te prortzaton of te performance crtera accordng to respectve ndexes and wegt of eac crteron. Te assumpton made for te prorty was tat te lower te performance crteron te ger te prorty actons to mprove rural WSS and te greater te wegt te ger te prorty for mprovement n te crteron, accordng to Equaton 27. Table 3 sows te values of te crtera wegts obtaned by consultng te publc and experts. Table 3. Crtera wegng (average value) Crtera Weg Avalablty ndex ( av ) 0,55 Capacty ndex ( cap ) 0,20 Contnuty ndex ( cont ) 0,01 Condton ndex ( cond ) 0,20 Socal ndex ( coop ) 0,04 Fgures 1a, 1b, 1c, 1d and 1e sow te prortzaton crtera to mprove te condtons of rural WSS for eac of te cases studed. P av cap cont cond coop 1 2 3 4 5 (27) Were: 1 s wegng avalablty ndex; 2 s wegng capacty ndex; 3 s wegng contnuty ndex; 4 s wegng condton ndex; and, 5 s wegng socal ndex. Te sum of te wegts sould be equal =1. Prorty 1>> avalablty Prorty 1>> avalablty Prorty 1>> avalablty Prorty 1>> avalablty Prorty 1>> avalablty Prorty 2>> capacty Prorty 2>> capacty Prorty 2>> capacty Prorty 2>> capacty Prorty 2>> capacty Prorty 3>> condton Prorty 3>> condton Prorty 3>> condton Prorty 3>> condton Prorty 3>> condton Prorty 4>> socal Prorty 4>> socal Prorty 4>> socal Prorty 4>> socal Prorty 4>> socal Prorty 5>> contnuty Prorty 5>> contnuty Prorty 5>> contnuty Prorty 5>> contnuty Prorty 5>> contnuty (A) (B) (C) (D) (E) A s te communty Cnturão Colna Verde, n Cuabá cty, Mato Grosso state; B s te communty pxuna, n Humatá cty, Amazona state; C s te communty Parasno, Humatá cty, Amazona state; D s te communty Parasno Grande, Humatá cty, Amazona state; and, E s te communty São Mguel, Humatá cty, Amazona state Fgure 2. Prortzaton crtera Accordng to Fgure 2, te model adopted for prortzng crtera sowed slgt senstvty. Tere are stll smlartes among te cases, as sown n te Table 2, and te man problems are related to te contnued supply of water.
Te tool was developed n an envronment Excel (spreadseet program from Mcrosoft) n order to make t accessble, easy to use. Fgures 3a and 3b sow data screens of te tool developed n Excel envronment. Te developed tool allows to obtan te values of crtera, after data entry, and also elps te decsonmakng, usng te TOPSS metod. Te tool s capable do receve te nput of fve alternatves and fve crtera of selecton. However, a larger number of alternatves and crtera can be easly added. (a) (b) Fgure 3. Assessment tool: data entry screen for evaluaton (a); entry screen data for decson makng (b) Concluson A tool for assessment system of rural water supply was developed and appled to some rural communtes n Brazl, more specfcally n communtes of Mato Grosso and Amazonas states. Te tool as some mprovements over te tool prevously developed by Retveld et al. (2009), ncludng te nserton of te TOPSS metod to mprove te selecton of actons to enance communty water systems, mplementaton of models n Excel spreadseet, and te proposton n order to prortze crtera of rural WSS. Tere s stll need for mprovement of te developed tool, especally wt regard to te prortzaton of te crtera for rural water supply systems. Te more accurately collecton of data ponts to better results. Acknowledgement Te autors would lke to express ter grattude for te fnancal support from te Brazlan agences CPq (project o. 556084/2009-8) and CAPES. References Amaral, L.A.; ader Flo, A.; Ross Junor, O.D.; Ferrera, F.L.A. (2003). Água de consumo umano como fator de rsco à saúde em propredades ruras. Revsta de Saúde Públca, 37(4), 510 514. Azevedo, R.P. (2006), Sstemas de abastecmento de água: a experênca nas aldeas Sateré-Mawé do Ro Marau no muncípo de Maués - Amazonas, n proceedngs of te 3º Semnáro nternaconal de Engenara de Saúde Públca, n Fortaleza, BRA.
Hwang, C.L.; Yoon, K. (1981). Multple attrbute decson makng. Metods and applcatons. A State of te Art Survey. ew York: Sprnger-Verlag. Kesavarz, A.R.; Sarfzade, M.; Hagg, A.A.K.; Amn, K.; Kestkar, S.; Bamdad, A. (2006). Rural domestc water consumpton beavor: A case study n Ramjerd area, Fars provnce,.r. ran. Journal Water Researc, 40, 1173 1178. Mnstéro da Saúde. (2004), Portara MS nº. 518/2004, avalable at: <ttp://portal.saude.gov.br/portal/arquvos/pdf/portara_518.pdf>. [Accessed 29 Aprl 2010]. Mutus, F.M.; Maamud, G.; Abdalle, A.; Gadan, H.M. (2007). Rural Water Supply Assessment, Tecncal Report o-08, FAO-SWALM, arob, Kenya. [on lne]. <avalable at: <ttp://www.faoswalm.org/ftp/water_reports/cleared/w- 08%20Rural%20Water%20Supply%20Assessment.pdf>. [Accessed 29 Aprl 2010]. Retveld, L.C.; Haaroff, J.; Jagals, P. (2009). A tool for tecncal assessment of rural water supply systems n Sout Afrca. Pyscs and Cemstry of te Eart, 34, 43 49. Sauer, J. (2003). Te Effcency of Rural nfrastructure Water Supply n Rural Areas of Transton. [on lne]. avalable at: <ttp://deas.repec.org/p/ww/wwrsa/ersa03p463.tml>. [Accessed 04 July 2010].