This article was downloaded by [20991221152]On 28 August 2014 At 1100Publisher Taylor amp FrancisInforma Ltd Registered in England and Wales Registered Number 1072954 Registered office Mortimer House37-41 Mortimer Street London W1T 3JH UK

Urban Water JournalPublication details including instructions for authors and subscription informationhttpwwwtandfonlinecomloinurw20

Illicit discharge detection and elimination Low costoptions for source identification and trackdown instormwater systemsKim Irvine a Mary C Rossi b Stephen Vermette a Jessica Bakert a amp Kerry Kleinfelder aa Department of Geography and Planning Buffalo State State University of New York USAb Erie County Department of Environment and Planning Buffalo New York USAPublished online 15 Nov 2011

To cite this article Kim Irvine Mary C Rossi Stephen Vermette Jessica Bakert amp Kerry Kleinfelder (2011) Illicit dischargedetection and elimination Low cost options for source identification and trackdown in stormwater systems Urban WaterJournal 86 379-395

To link to this article httpdxdoiorg1010801573062X2011630095

PLEASE SCROLL DOWN FOR ARTICLE

Taylor amp Francis makes every effort to ensure the accuracy of all the information (the ldquoContentrdquo) containedin the publications on our platform However Taylor amp Francis our agents and our licensors make norepresentations or warranties whatsoever as to the accuracy completeness or suitability for any purpose of theContent Any opinions and views expressed in this publication are the opinions and views of the authors andare not the views of or endorsed by Taylor amp Francis The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information Taylor and Francis shall not be liable forany losses actions claims proceedings demands costs expenses damages and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with in relation to or arising out of the use ofthe Content

This article may be used for research teaching and private study purposes Any substantial or systematicreproduction redistribution reselling loan sub-licensing systematic supply or distribution in anyform to anyone is expressly forbidden Terms amp Conditions of access and use can be found at httpwwwtandfonlinecompageterms-and-conditions

RESEARCH ARTICLE

Illicit discharge detection and elimination Low cost options for source identification

and trackdown in stormwater systems

Kim Irvinea Mary C Rossib Stephen Vermettea Jessica Bakerta and Kerry Kleinfeldera

aDepartment of Geography and Planning Buffalo State State University of New York USA bErie County Department ofEnvironment and Planning Buffalo New York USA

(Received 11 March 2011 final version received 21 September 2011)

Permit regulations in the US for Municipal Separate Storm Sewer Systems (MS4) require the MS4 to develop aprogram that detects and eliminates illicit discharges (eg improper wastewater connections) into the storm sewersystem Municipalities are interested in cost-effective methods to meet the permit requirements of this federalmandate Our demonstration project with municipalities in Western New York State evaluated low cost options forillicit discharge trackdown First a visual reconnaissance was used to document flowing stormwater outfalls in dryweather Subsequently a sampling program was conducted in combination with decision-making tools to identifypossible sources of illicit discharges Colorimetric techniques were tested for a number of chemical parameters andthe Coliscan Easygel1 system was tested for E coli analysis Results from these various cost-effective analyticaltechniques were compared with analysis by standard methods The E coli test in particular had good precision andwas useful in trackdown

Keywords illicit discharge stormwater MS4 outfall E coli

1 Introduction

11 Background and regulatory need

Over the past 30 years municipal stormwater manage-ment has experienced an important paradigm shiftevolving from an urban flood control function to anenvironmental protection and regulatory function(Tucker and Harrison 2006) Municipal stormwatermanagement increasingly is thought of as an elementof a comprehensive integrated urban water resourcemanagement service that may have a watershed focus(Prince Georgersquos County 1999 US EPA 2000Clarkson 2003 Stark 2003 Tucker and Harrison2006) Due to the concerns about stormwater qualityand impacts on receiving waters the US EPApromulgated a Phase I stormwater program in 1990followed by Phase II in 1999 (US EPA 1999) Phase Irequired medium and large MS4 owners and operatorswith populations of 100000 or more to obtainNational Pollutant Discharge Elimination System(NPDES) permit coverage for their stormwater dis-charges Phase II extended the regulations to smallMS4s servicing populations under 100000 within USCensus-defined Urbanized Areas (41000 peoplemi2 41000 people259 km2)

Regulated MS4 owners and operators in New YorkState are required to obtain a General Permit for theirseparate storm sewer system discharges from the stateregulatory authority the Department of Environmen-tal Conservation Under the General Permit the MS4must design a stormwater management program toreduce the discharge of pollutants to the lsquolsquomaximumextent practicablersquorsquo and protect water quality Thestormwater management programs developed by theMS4s must include six minimum control measures(MCMs)

Public Education and Outreach on StormwaterImpacts

Public InvolvementParticipation

Illicit Discharge Detection and Elimination(IDDE)

Construction Site Runoff Control

Post-construction Stormwater Management

Pollution PreventionGood Housekeeping forMunicipal Operations

These complementary measures when successfullyimplemented are expected to reduce the volume of

Corresponding author Email irvineknbuffalostateedu

Urban Water Journal

Vol 8 No 6 December 2011 379ndash395

ISSN 1573-062X printISSN 1744-9006 online

2011 Taylor amp Francis

httpdxdoiorg1010801573062X2011630095

httpwwwtandfonlinecom

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

pollutants discharged into receiving bodies of waterHowever the measures also represent an un-fundedfederal mandate off-loaded to municipalities andimplementation of the measures particularly theIDDE efforts (and associated analytical costs) can bea financial burden for the municipalities The objectiveof this study therefore was to develop and demonstratea cost-effective and robust protocol for small MS4s todetect sample and track down the source of illicitdischarges to their storm sewer systems

12 Western New York Stormwater Coalition

In 2002 the regulated MS4s in two counties ofWestern New York State (Erie and Niagara) startedmeeting regularly and established the Western NewYork Stormwater Coalition (WNYSC) The primaryobjective of the WNYSC is to utilize intergovern-mental cooperation to collectively meet the Phase IIstormwater regulations The WNYSC membershipincludes 29 regulated MS4s in Erie County 10regulated MS4s in Niagara County two county agencyMS4s and two non-traditional MS4s Through parti-cipation in the WNYSC the MS4s have workedcooperatively to develop and implement a stormwatermanagement program that meets the Phase II storm-water regulations

The Erie County Department of Environment andPlanning (ECDEP) is lead agency for the WNYSC andin that capacity acts as administrator and coordinatorA detailed discussion of the WNYSC activities toaddress the six MCMs (including local ordinancedevelopment) is provided by Rossi et al (2009) TheIDDE requirement was the most costly and laborintensive measure of all the MCMs to implementPrioritization of the tasks to develop an IDDEprogram dictated that first the stormwater outfallsmust be mapped The WNYSC bid out the mappingproject and hired a single contractor to identifyinspect and geocode over 5000 outfalls The resultsof this effort were made available to the MS4s andgeneral public through web-based mapping (httpgis1eriegovENSSO) and the benefits of this colla-borative effort were two fold realization of significantsavings based on economies of scale and fulfilment ofthe first round of inspections required by the Phase IIregulations

With the outfall mapping completed theWNYSC turned its attention to developing anIDDE protocol Buffalo State State University ofNew York (Buffalo State) collaborated with theWNYSC on this task through a grant from the USEPA This collaborative project illustrates theapproach that the WNYSC has successfully utilizedto address Phase II regulations

2 Methods

21 The demonstration project

New York Statersquos IDDE regulations require the MS4to develop implement and enforce a program to detectand eliminate illicit discharges into the system managedby the MS4 Illicit discharges to storm sewer systemsinclude illegal dumping improper businessindustrialconnections failing septic systems recreational sewageand wastewater connections The regulations specifi-cally suggest Pittrsquos (2004) Illicit Discharge Detection andElimination A Guidance Manual for Program Develop-ment and Technical Assessment and its Outfall Recon-naissance Inventory (ORI) tool be used forclassification of outfalls This document was valuablein developing the basic framework for our projectalthough local conditions required some variances toimplementation It was the intention of our project totake those national standards apply them to a subset ofmunicipalities in the WNYSC and refine them asneeded to provide an efficient and cost-effectiveprogram to successfully meet the regulations

Of the 41 regulated MS4s in the WNYSC six wereselected to participate in the demonstration projectSelection was based on willingness to participate aswell as a desire to represent a range in socio-economicstructure urban development and technical capabilityEach of the six demonstration municipalities provideda list of at least 10 priority outfalls that would beincluded in an ORI as the first step in developing theillicit discharge trackdown protocol The WNYSCdeveloped guidelines to help the municipalities identifythe priority outfalls The guidance directed the MS4sto consider factors such as whether or not the outfallwas known to experience dry weather discharges thepotential impact of the surrounding land use for anoutfall outfalls located in environmentally sensitiveareas and the age of the outfall and contributing area

The study team met with representatives of eachmunicipality once the priority outfalls had been identi-fied to review general procedures for the assessmentreview site locations on the municipal maps develop ageneral schedule for the work and ensure coordinationof all efforts between the team and municipal staff

22 Visual inspection of priority outfalls ndash completingthe ORI field sheet

Field work for the project commenced with visualinspections for the ORI A 72 hour antecedent dryperiod was observed prior to the site visit to minimizethe possibility of sampling stormwater runoff ratherthan illicit connections and it also acted to standardizeconditions to facilitate between-site data comparisonsA 72 hour antecedent dry period also has been used todefine the start of independent wet weather events in

380 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

pollutants discharged into receiving bodies of waterHowever the measures also represent an un-fundedfederal mandate off-loaded to municipalities andimplementation of the measures particularly theIDDE efforts (and associated analytical costs) can bea financial burden for the municipalities The objectiveof this study therefore was to develop and demonstratea cost-effective and robust protocol for small MS4s todetect sample and track down the source of illicitdischarges to their storm sewer systems

12 Western New York Stormwater Coalition

In 2002 the regulated MS4s in two counties ofWestern New York State (Erie and Niagara) startedmeeting regularly and established the Western NewYork Stormwater Coalition (WNYSC) The primaryobjective of the WNYSC is to utilize intergovern-mental cooperation to collectively meet the Phase IIstormwater regulations The WNYSC membershipincludes 29 regulated MS4s in Erie County 10regulated MS4s in Niagara County two county agencyMS4s and two non-traditional MS4s Through parti-cipation in the WNYSC the MS4s have workedcooperatively to develop and implement a stormwatermanagement program that meets the Phase II storm-water regulations

The Erie County Department of Environment andPlanning (ECDEP) is lead agency for the WNYSC andin that capacity acts as administrator and coordinatorA detailed discussion of the WNYSC activities toaddress the six MCMs (including local ordinancedevelopment) is provided by Rossi et al (2009) TheIDDE requirement was the most costly and laborintensive measure of all the MCMs to implementPrioritization of the tasks to develop an IDDEprogram dictated that first the stormwater outfallsmust be mapped The WNYSC bid out the mappingproject and hired a single contractor to identifyinspect and geocode over 5000 outfalls The resultsof this effort were made available to the MS4s andgeneral public through web-based mapping (httpgis1eriegovENSSO) and the benefits of this colla-borative effort were two fold realization of significantsavings based on economies of scale and fulfilment ofthe first round of inspections required by the Phase IIregulations

With the outfall mapping completed theWNYSC turned its attention to developing anIDDE protocol Buffalo State State University ofNew York (Buffalo State) collaborated with theWNYSC on this task through a grant from the USEPA This collaborative project illustrates theapproach that the WNYSC has successfully utilizedto address Phase II regulations

2 Methods

21 The demonstration project

New York Statersquos IDDE regulations require the MS4to develop implement and enforce a program to detectand eliminate illicit discharges into the system managedby the MS4 Illicit discharges to storm sewer systemsinclude illegal dumping improper businessindustrialconnections failing septic systems recreational sewageand wastewater connections The regulations specifi-cally suggest Pittrsquos (2004) Illicit Discharge Detection andElimination A Guidance Manual for Program Develop-ment and Technical Assessment and its Outfall Recon-naissance Inventory (ORI) tool be used forclassification of outfalls This document was valuablein developing the basic framework for our projectalthough local conditions required some variances toimplementation It was the intention of our project totake those national standards apply them to a subset ofmunicipalities in the WNYSC and refine them asneeded to provide an efficient and cost-effectiveprogram to successfully meet the regulations

Of the 41 regulated MS4s in the WNYSC six wereselected to participate in the demonstration projectSelection was based on willingness to participate aswell as a desire to represent a range in socio-economicstructure urban development and technical capabilityEach of the six demonstration municipalities provideda list of at least 10 priority outfalls that would beincluded in an ORI as the first step in developing theillicit discharge trackdown protocol The WNYSCdeveloped guidelines to help the municipalities identifythe priority outfalls The guidance directed the MS4sto consider factors such as whether or not the outfallwas known to experience dry weather discharges thepotential impact of the surrounding land use for anoutfall outfalls located in environmentally sensitiveareas and the age of the outfall and contributing area

The study team met with representatives of eachmunicipality once the priority outfalls had been identi-fied to review general procedures for the assessmentreview site locations on the municipal maps develop ageneral schedule for the work and ensure coordinationof all efforts between the team and municipal staff

22 Visual inspection of priority outfalls ndash completingthe ORI field sheet

Field work for the project commenced with visualinspections for the ORI A 72 hour antecedent dryperiod was observed prior to the site visit to minimizethe possibility of sampling stormwater runoff ratherthan illicit connections and it also acted to standardizeconditions to facilitate between-site data comparisonsA 72 hour antecedent dry period also has been used todefine the start of independent wet weather events in

380 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

sampling done under several Long Term Control Plan(LTCP) studies for combined sewer overflow abate-ment in the city of Buffalo (eg Irvine et al 2005) The72 hour antecedent dry period criterion to initiatevisual inspection and sampling therefore is consistentwith other sewer studies done in the region

At each site the study team used a Garmin eTrexhandheld Global Positioning System (GPS) unit toconfirm the location of the outfall originally inventor-ied under the web-based mapping effort (httpgis1eriegovENSSO) The study team then conductedthe visual inspection of the site including completion ofthe ORI Field Sheet The ORI Field Sheet piloted inthis study was based on those developed by Pitt (2004)At the end of the two year field program and based onstudy team experience the ORI Field Sheet wasmodified somewhat to reflect local conditions

If flow at the outfall was observed flow rate wasmeasured using one of two methods For smaller flowsa bucket was used to collect water for a timed periodand the volume of water was measured in a graduated1 L beaker For larger flows the velocity was measuredusing either a Marsh McBirney Model 2000 meter or aGlobal Water FP 101 meter with the width and depthof flow being determined using a tape measure

The ORI served two purposes First it provided aquality assurance check on the original mapping doneunder contract for the WNYSC Second it allowed arapid determination of potential need for additionalinvestigation of illicit connection If flow was observedduring the ORI survey the site was categorized ashaving high potential for illicit connection If otherindicators of sanitary waste were observed during theORI (eg toilet paper oily sheen odour) the site wascategorized as having very high potential for illicitconnection If flow or the other indicators wereobserved the outfall was scheduled for sample collec-tion and potential trackdown

23 Sample collection at flowing outfalls

For sample collection at flowing outfalls the 72 hourantecedent dry period was again observed Using acertified clean 1 L amber glass bottle a grab samplewas collected directly from the flow and retained forlaboratory analysis In a separate clean 1 L amberglass bottle a sample was collected for on-site analysisof pH temperature conductivity and total dissolvedsolids using a Hanna Instruments HI9811 field meterBecause of the variety of outfall configurations andflow rates it sometimes was not possible to sampledirectly into the bottles A clean long-handled dipperwas used if the flow could not be directly collected in the1 L amber bottles The dipper was conditioned with theflow (ie rinsed three times) prior to collecting a

sample In other cases it was necessary to lsquolsquochannelrsquorsquo thediffuse shallow flow into the sample bottle using plasticroof gutter Flow rate was measured and recordedagain per the methods described in Section 22

From the lsquolsquolaboratory analysisrsquorsquo bottle 1 mL ofsample was extracted using a disposable sterile plasticpipette and dispensed into a Coliscan Easygel1

(Micrology Laboratories Goshen IN) growth mediascrew-top plastic vial for the E coli analysis The 1 Lamber glass sample bottle for the laboratory analysisand the Coliscan Easygel1 growth media (containingthe 1 mL water sample) were placed on ice forpreservation in the field All samples were processedwithin four hours of collection

24 Sample collection for trackdown

Based on the results of the outfall sampling and thedecision making tools (discussed in section 33)sampling was conducted progressively up-pipe fortrackdown purposes The within-sewer sample loca-tions were considerably refined by first driving andwalking the contributing area The contributing areawas defined based on discussions with municipal staffand guided by sewer drawings Areas with no observedflow in the sewer pipes were eliminated from furtherconsideration for the trackdown

Typically the trackdown sampling required thelifting of manhole covers However no confined spaceentry was done during the sampling all sampling wasdone from the street surface The sample collection me-thod depended on the depth of flow and the size depthand configuration of the manhole access Sampling wasdone using either a dipper telescoping sampler or abailer that could be laid horizontally into shallow flow

The sample handling and on-site parameter analysisat the trackdown sites was identical to that describedfor the outfalls (discussed in Section 23) Generallytwo to three sites upstream of a particular outfall wereselected for sampling and once the sample results werereviewed either further sampling was done or adecision regarding the source was determined

25 Receiving water body sampling

Water samples were collected in five different receivingwater bodies Cayuga Creek (Niagara County) Scaja-quada Creek Niagara River Two Mile Creek and atributary leaving Green Lake (Figure 1) The samplesites generally were located as an upstream anddownstream pair The Cayuga Creek site had a totalof three permitted stormwater discharges between theupstream and downstream location Niagara Riverhad four permitted stormwater discharges between theupstream and downstream location Two Mile Creek

Urban Water Journal 381

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

had five permitted stormwater discharges between theupstream and downstream location and the tributaryleaving Green Lake had one permitted stormwaterdischarge between the upstream and downstreamlocation All of the receiving water sites were sampledduring dry weather flow Scajaquada Creek receivesnot only stormwater discharges but also combinedsewer overflow (CSO) discharges from the City ofBuffalo Samples were collected at between two andfour sites (Figure 2) on Scajaquada Creek for threestorms that occurred on 24 July 2008 1 October 2008and 16 October 2008 in addition to dry weather

samples These sites were selected for sampling inassociation with the Buffalo Sewer Authorityrsquos CSOLong Term Control Plan Phase II Study

Samples collected on Scajaquada Creek werecollected either with a telescoping sampler or bywading into the creek and sampling directly into abottle depending on flow conditions Samples forCayuga Creek Two Mile Creek and the tributaryleaving Green Lake were collected by wading into thecreek and sampling directly into a bottle while samplesfor the Niagara River were collected from the shorelinewith a telescoping sampler The sample handling and

Figure 1 General location of receiving water samples

382 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

on-site parameter analysis at the receiving water bodysites was identical to that described for the outfalls(discussed in Section 23)

26 Municipal tap water and local groundwaterprofiles

To assist with the trackdown interpretation samples oftap water were collected at two different locations inthree of the participating MS4 communities The

sample handling and on-site parameter analysis forthe tap water was identical to that described for theoutfalls (discussed in Section 23) Supplementing themunicipal tap water profile data were the AnnualDrinking Water Quality Reports for the participatingMS4 communities The reports identify levels ofvarious contaminants detected as well as fluorideconcentrations

Local groundwater data also were compiled toassist with the trackdown interpretation The United

Figure 2 Scajaquada Creek sample sites The creek flows through an underground tunnel between sites 1 and 2

Urban Water Journal 383

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

States Geological Surveyrsquos Groundwater Quality inWestern New York 2006 (Eckhardt et al 2008) wasused to develop a geographical profile to characterizelocal groundwater quality for eight sites in WesternNew Yorkrsquos Urbanized Area

27 Analytical methods at Buffalo State

Routine analysis of all samples was done in the WaterQuality Laboratory Department of GeographyPlan-ning Buffalo State The sample parameters andanalytical methods were selected through a review ofPitt (2004) Pomeroy et al (1996) and through ourown experience in evaluating water quality withcommunity groups (eg httpwwwbuffalostateeduorgsaqua Wills and Irvine 1996 Irvine et al 2004)The guiding principle for our sampling program wasthat we wanted meaningful parameters that could beanalyzed easily and cost-effectively (particularly im-portant for municipalities with limited resources) withreasonable precision

Parameters sampled for the study were nitrateammonia total phosphorous potassium fluoride totalchlorine chromium VI copper dissolved oxygenBOD5 detergents (anionic surfactants) phenols gen-eral and carbonate hardness total suspended solidsturbidity and E coli Analyses for nitrate ammoniatotal phosphorus fluoride total chlorine chromiumVI and copper were done using a colorimetricapproach with a Hanna Instruments (wwwhannainstcom) C 200 multiparameter bench photometer Sam-ples were not filtered prior to analysis In considerationof space the analytical details are not provided herebut are presented in Irvine et al (2009) Briefly allanalyses carefully followed the Standard OperatingProcedures outlined in the Hanna Instruments Instruc-tion Manual C 99 amp C 200 HI 83000 Series Multi-parameter Bench Photometers The nitrate analysis usedan adaptation of the cadmium reduction method andresults were expressed as nitrate-nitrogen which weremultiplied by a factor of 443 to convert to nitrate(NO3

-) The ammonia analysis used an adaptation ofthe Nessler method total phosphorus analysis was anadaptation of the amino acid method fluoride analysiswas an adaptation of the SPADNS method totalchlorine analysis was an adaptation of the US EPADPDmethod chromiumVI analysis was an adaptationof the diphenylcarbohydrazide method and copperanalysis used the bicinchoninate reagent approachPotassium was analyzed using an adaptation of theturbidimetric tetraphenylborate method using a HannaInstruments HI 93750 single parameter photometer

Dissolved oxygen was measured using a CHE-Metrics Inc (wwwchemetricscom) visual test kit thatemployed the indigo carmine method (ASTM D

888ndash87) in which the reduced form of indigo carminereacts in a sample with dissolved oxygen forming ablue product The kit comes with a graduated colourcomparator with divisions of 1 2 3 4 5 6 8 10 and12 ppm To determine BOD5 the dissolved oxygen levelmeasured in the field was compared with the dissolvedoxygen measured in a BOD bottle held at roomtemperature 5 days after sample collection

Detergents (anionic surfactants) were measuredusing a CHEMetrics Inc (wwwchemetricscom)visual test kit that employed the methylene blue activesubstances method in which anionic detergents reactwith methylene blue to form a blue-coloured complexthat is extracted into an immiscible organic solventResults are expressed in ppm as linear alkylbenzenesulfonate (LAS) equivalent weight 325 The intensityof the colour reaction is determined using a graduatedcolour comparator having divisions of 0 025 050075 10 15 20 and 30 ppm

The Coliscan Easygel1 system from MicrologyLabs (httpwwwmicrologylabscomHome) GoshenIN was used to determine E coli levels The mediuminoculums mix collected in the field was plated imme-diately upon return to the laboratory at Buffalo StateThe poured samples were incubated at room tempera-ture for 48 hours and after this period all purple colo-nies were counted as E coli Pink colonies are coliformand teal green colonies are other types of bacteria thatmay include Salmonella spp or Shigella spp The pinkand teal green colonies were not counted in this study

Phenols were measured using a CHEMetrics Inc(wwwchemetricscom) visual test kit that employed the4-aminoantipyrine (4-AAP) method in which phenoliccompounds react with 4-AAP in alkaline solution in thepresence of ferricyanide to produce a red reactionproduct The intensity of the colour reaction is deter-mined using a graduated colour comparator havingdivisions of 0 5 75 10 15 20 25 and 30 ppm Generaland carbonate hardness were determined using atitration approach with reagents in the GH amp KHGeneral amp Carbonate Hardness Test Kit available fromAquarium Pharmaceuticals (httpaquariumpharmcom) Turbidity (NTU) was determined in the labora-tory using an Oakton T-100 Portable Turbidity Meter

28 Quality assurance checks on routine analyticalmethodologies

To check the calibration of the Hanna C 200 and HI93750 photometers certified standard solutions ofpotassium nitrate fluoride chromium VI and copperwere obtained from Ricca Chemical Co Arlington TXand certified standard solutions of phosphorus andammonia were obtained from Pointe Scientific IncCanton MI In addition to the standard solutions for

384 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

on-site parameter analysis at the receiving water bodysites was identical to that described for the outfalls(discussed in Section 23)

26 Municipal tap water and local groundwaterprofiles

To assist with the trackdown interpretation samples oftap water were collected at two different locations inthree of the participating MS4 communities The

sample handling and on-site parameter analysis forthe tap water was identical to that described for theoutfalls (discussed in Section 23) Supplementing themunicipal tap water profile data were the AnnualDrinking Water Quality Reports for the participatingMS4 communities The reports identify levels ofvarious contaminants detected as well as fluorideconcentrations

Local groundwater data also were compiled toassist with the trackdown interpretation The United

Figure 2 Scajaquada Creek sample sites The creek flows through an underground tunnel between sites 1 and 2

Urban Water Journal 383

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

States Geological Surveyrsquos Groundwater Quality inWestern New York 2006 (Eckhardt et al 2008) wasused to develop a geographical profile to characterizelocal groundwater quality for eight sites in WesternNew Yorkrsquos Urbanized Area

27 Analytical methods at Buffalo State

Routine analysis of all samples was done in the WaterQuality Laboratory Department of GeographyPlan-ning Buffalo State The sample parameters andanalytical methods were selected through a review ofPitt (2004) Pomeroy et al (1996) and through ourown experience in evaluating water quality withcommunity groups (eg httpwwwbuffalostateeduorgsaqua Wills and Irvine 1996 Irvine et al 2004)The guiding principle for our sampling program wasthat we wanted meaningful parameters that could beanalyzed easily and cost-effectively (particularly im-portant for municipalities with limited resources) withreasonable precision

Parameters sampled for the study were nitrateammonia total phosphorous potassium fluoride totalchlorine chromium VI copper dissolved oxygenBOD5 detergents (anionic surfactants) phenols gen-eral and carbonate hardness total suspended solidsturbidity and E coli Analyses for nitrate ammoniatotal phosphorus fluoride total chlorine chromiumVI and copper were done using a colorimetricapproach with a Hanna Instruments (wwwhannainstcom) C 200 multiparameter bench photometer Sam-ples were not filtered prior to analysis In considerationof space the analytical details are not provided herebut are presented in Irvine et al (2009) Briefly allanalyses carefully followed the Standard OperatingProcedures outlined in the Hanna Instruments Instruc-tion Manual C 99 amp C 200 HI 83000 Series Multi-parameter Bench Photometers The nitrate analysis usedan adaptation of the cadmium reduction method andresults were expressed as nitrate-nitrogen which weremultiplied by a factor of 443 to convert to nitrate(NO3

-) The ammonia analysis used an adaptation ofthe Nessler method total phosphorus analysis was anadaptation of the amino acid method fluoride analysiswas an adaptation of the SPADNS method totalchlorine analysis was an adaptation of the US EPADPDmethod chromiumVI analysis was an adaptationof the diphenylcarbohydrazide method and copperanalysis used the bicinchoninate reagent approachPotassium was analyzed using an adaptation of theturbidimetric tetraphenylborate method using a HannaInstruments HI 93750 single parameter photometer

Dissolved oxygen was measured using a CHE-Metrics Inc (wwwchemetricscom) visual test kit thatemployed the indigo carmine method (ASTM D

888ndash87) in which the reduced form of indigo carminereacts in a sample with dissolved oxygen forming ablue product The kit comes with a graduated colourcomparator with divisions of 1 2 3 4 5 6 8 10 and12 ppm To determine BOD5 the dissolved oxygen levelmeasured in the field was compared with the dissolvedoxygen measured in a BOD bottle held at roomtemperature 5 days after sample collection

Detergents (anionic surfactants) were measuredusing a CHEMetrics Inc (wwwchemetricscom)visual test kit that employed the methylene blue activesubstances method in which anionic detergents reactwith methylene blue to form a blue-coloured complexthat is extracted into an immiscible organic solventResults are expressed in ppm as linear alkylbenzenesulfonate (LAS) equivalent weight 325 The intensityof the colour reaction is determined using a graduatedcolour comparator having divisions of 0 025 050075 10 15 20 and 30 ppm

The Coliscan Easygel1 system from MicrologyLabs (httpwwwmicrologylabscomHome) GoshenIN was used to determine E coli levels The mediuminoculums mix collected in the field was plated imme-diately upon return to the laboratory at Buffalo StateThe poured samples were incubated at room tempera-ture for 48 hours and after this period all purple colo-nies were counted as E coli Pink colonies are coliformand teal green colonies are other types of bacteria thatmay include Salmonella spp or Shigella spp The pinkand teal green colonies were not counted in this study

Phenols were measured using a CHEMetrics Inc(wwwchemetricscom) visual test kit that employed the4-aminoantipyrine (4-AAP) method in which phenoliccompounds react with 4-AAP in alkaline solution in thepresence of ferricyanide to produce a red reactionproduct The intensity of the colour reaction is deter-mined using a graduated colour comparator havingdivisions of 0 5 75 10 15 20 25 and 30 ppm Generaland carbonate hardness were determined using atitration approach with reagents in the GH amp KHGeneral amp Carbonate Hardness Test Kit available fromAquarium Pharmaceuticals (httpaquariumpharmcom) Turbidity (NTU) was determined in the labora-tory using an Oakton T-100 Portable Turbidity Meter

28 Quality assurance checks on routine analyticalmethodologies

To check the calibration of the Hanna C 200 and HI93750 photometers certified standard solutions ofpotassium nitrate fluoride chromium VI and copperwere obtained from Ricca Chemical Co Arlington TXand certified standard solutions of phosphorus andammonia were obtained from Pointe Scientific IncCanton MI In addition to the standard solutions for

384 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

States Geological Surveyrsquos Groundwater Quality inWestern New York 2006 (Eckhardt et al 2008) wasused to develop a geographical profile to characterizelocal groundwater quality for eight sites in WesternNew Yorkrsquos Urbanized Area

27 Analytical methods at Buffalo State

Routine analysis of all samples was done in the WaterQuality Laboratory Department of GeographyPlan-ning Buffalo State The sample parameters andanalytical methods were selected through a review ofPitt (2004) Pomeroy et al (1996) and through ourown experience in evaluating water quality withcommunity groups (eg httpwwwbuffalostateeduorgsaqua Wills and Irvine 1996 Irvine et al 2004)The guiding principle for our sampling program wasthat we wanted meaningful parameters that could beanalyzed easily and cost-effectively (particularly im-portant for municipalities with limited resources) withreasonable precision

Parameters sampled for the study were nitrateammonia total phosphorous potassium fluoride totalchlorine chromium VI copper dissolved oxygenBOD5 detergents (anionic surfactants) phenols gen-eral and carbonate hardness total suspended solidsturbidity and E coli Analyses for nitrate ammoniatotal phosphorus fluoride total chlorine chromiumVI and copper were done using a colorimetricapproach with a Hanna Instruments (wwwhannainstcom) C 200 multiparameter bench photometer Sam-ples were not filtered prior to analysis In considerationof space the analytical details are not provided herebut are presented in Irvine et al (2009) Briefly allanalyses carefully followed the Standard OperatingProcedures outlined in the Hanna Instruments Instruc-tion Manual C 99 amp C 200 HI 83000 Series Multi-parameter Bench Photometers The nitrate analysis usedan adaptation of the cadmium reduction method andresults were expressed as nitrate-nitrogen which weremultiplied by a factor of 443 to convert to nitrate(NO3

-) The ammonia analysis used an adaptation ofthe Nessler method total phosphorus analysis was anadaptation of the amino acid method fluoride analysiswas an adaptation of the SPADNS method totalchlorine analysis was an adaptation of the US EPADPDmethod chromiumVI analysis was an adaptationof the diphenylcarbohydrazide method and copperanalysis used the bicinchoninate reagent approachPotassium was analyzed using an adaptation of theturbidimetric tetraphenylborate method using a HannaInstruments HI 93750 single parameter photometer

Dissolved oxygen was measured using a CHE-Metrics Inc (wwwchemetricscom) visual test kit thatemployed the indigo carmine method (ASTM D

888ndash87) in which the reduced form of indigo carminereacts in a sample with dissolved oxygen forming ablue product The kit comes with a graduated colourcomparator with divisions of 1 2 3 4 5 6 8 10 and12 ppm To determine BOD5 the dissolved oxygen levelmeasured in the field was compared with the dissolvedoxygen measured in a BOD bottle held at roomtemperature 5 days after sample collection

Detergents (anionic surfactants) were measuredusing a CHEMetrics Inc (wwwchemetricscom)visual test kit that employed the methylene blue activesubstances method in which anionic detergents reactwith methylene blue to form a blue-coloured complexthat is extracted into an immiscible organic solventResults are expressed in ppm as linear alkylbenzenesulfonate (LAS) equivalent weight 325 The intensityof the colour reaction is determined using a graduatedcolour comparator having divisions of 0 025 050075 10 15 20 and 30 ppm

The Coliscan Easygel1 system from MicrologyLabs (httpwwwmicrologylabscomHome) GoshenIN was used to determine E coli levels The mediuminoculums mix collected in the field was plated imme-diately upon return to the laboratory at Buffalo StateThe poured samples were incubated at room tempera-ture for 48 hours and after this period all purple colo-nies were counted as E coli Pink colonies are coliformand teal green colonies are other types of bacteria thatmay include Salmonella spp or Shigella spp The pinkand teal green colonies were not counted in this study

Phenols were measured using a CHEMetrics Inc(wwwchemetricscom) visual test kit that employed the4-aminoantipyrine (4-AAP) method in which phenoliccompounds react with 4-AAP in alkaline solution in thepresence of ferricyanide to produce a red reactionproduct The intensity of the colour reaction is deter-mined using a graduated colour comparator havingdivisions of 0 5 75 10 15 20 25 and 30 ppm Generaland carbonate hardness were determined using atitration approach with reagents in the GH amp KHGeneral amp Carbonate Hardness Test Kit available fromAquarium Pharmaceuticals (httpaquariumpharmcom) Turbidity (NTU) was determined in the labora-tory using an Oakton T-100 Portable Turbidity Meter

28 Quality assurance checks on routine analyticalmethodologies

To check the calibration of the Hanna C 200 and HI93750 photometers certified standard solutions ofpotassium nitrate fluoride chromium VI and copperwere obtained from Ricca Chemical Co Arlington TXand certified standard solutions of phosphorus andammonia were obtained from Pointe Scientific IncCanton MI In addition to the standard solutions for

384 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

QAQC of the Hanna Instruments photometers severalother QAQC measures were taken to assess thereliability of the routine analytical methods employedfor this study First a total of 11 receiving water samplesand 12 samples from storm sewers were collected induplicate and analyzed at Buffalo State and at a NewYork State Health Department certified laboratory Theparameters analyzed at Buffalo State and the certifiedlaboratory were chromium VI copper potassiumammonia BOD5 nitrate total phosphorus and totalrecoverable phenolics All samples were refrigerated at48C until analysis was started At the certified labchromium VI copper and potassium were analyzedusing US EPA method 2007 ammonia was analyzedusing US EPA Method 3501 nitrate was analyzedusing US EPA method 3537 and total recoverablephenolics were analyzed using US EPA method 4204(US EPA 1983 1991 1992 1993) Biochemical oxygendemand was analyzed using Standard Method 5210Band total phosphorus was analyzed using StandardMethod 4500-P E (APHA 1998) Detection limits werechromium ndash 0004 mgL copper 001 mgL potassium ndash05 mgL ammonia ndash 002 mgL BOD5 ndash 20 mgLnitrate ndash 005 mgL total phosphorus ndash 001 mgL totalrecoverable phenolics ndash 001 mgL QAQC measures(except for BOD5) included matrix spikes methodblanks and blank spikes for each batch of samples

A total of 21 samples (10 receiving water and 11storm sewer) were collected in duplicate and analyzed atBuffalo State and at the Erie County Public HealthLaboratory for E coli The Erie County Public HealthLaboratory analyzed E coli using membrane filtrationaccording toUS EPAMethod 1603 Finally a YSI 6920datasonde was used to compare dissolved oxygenmeasurements with the CHEMetrics visual test kit meas-urements as well as measurements of pH and conductiv-ity with the Hanna Instruments HI9811 field meter

3 Results

31 MS4 participants

Among the six MS4s that participated in the projectthere were five traditional land use control MS4s (ie

municipally owned) and one non-traditional non-landuse control MS4 (college campus) The demographiccharacteristics of the MS4s (Table 1) offered diversitywith the storm sewer system configurations receivingwater features and population

32 Outfall sampling

A total of 64 outfalls were investigated for the projectapproximately ten per MS4 The first round offieldwork for the project consisted of conducting anORI for the 64 outfalls selected for the study The ORIserved to verify existing outfall data confirm dryweather flow conditions measure flow rates andevaluate potential for an illicit discharge

Once the ORIs were completed fieldwork transi-tioned to outfall sampling and field measurementsfollowed by lab analysis and interpretation For eachparticipating MS4 the ORI results were used tostreamline the outfall sampling by prioritizing field-work toward flowing outfalls or those with compellingevidence of prior illicit discharges (ie intermittentthough not necessarily flowing at initial inspection)

To determine whether an outfall flowing during dryconditions was illicit it was necessary to collect andanalyze a sample Samples were collected according tothe methods detailed in Section 23 under 72 hour dryantecedent conditions

33 Illicit discharge trackdown

331 Determination of need

Lab results and field measurements determinedwhether or not a trackdown investigation was neededThe primary decision-making tools were a flow chart(Figure 3) and industrial benchmarks developed by Pitt(2004) and the New York State Department ofEnvironmental Conservationrsquos (NYSDEC) Part 703Surface Water Quality Standards (NYSDEC 1999) Inthe absence of effluent limitation guidelines theNYSDEC Standards provided a baseline value todetermine tolerable levels for contaminants detected

Table 1 Demographics of participating MS4s

MS4Population

(2000 Census)Area(km2)

Length ofPipe (m)

Number ofOutfalls

Length ofStreams (m) Land Use

1 12575 047 17320 6 1049 Academic institution (urban area)2 18621 666 64370 331 120500 Low density residential3 22458 815 167400 175 168800 Medium density residential4 24343 898 261300 218 140300 Medium density residential5 61729 431 458700 62 45690 Light industrial Commercial High density residential6 45920 466 159000 278 96830 Light industrial Medium density residential

represents number of students and full time faculty fall 2008 predominant land use in the contributing area of the sampled outfalls in eachMS4

Urban Water Journal 385

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

The Pitt (2004) flow chart values and the NYSDECStandards as summarized in Table 2 allowed for arapid determination as to whether or not a trackdownwas necessary Simply put if a given parameterexceeded the Pitt values or the NYSDEC Standardsfurther investigation was warranted The Pitt (2004)

flowchartindustrial benchmarks were used to char-acterize the (possible) source of flow as sanitarywastewater washwater irrigation tap water ground-water or industrialcommercial Municipal tap waterand ground water profiles (Eckhardt et al 2008) alsowere used to determine a possible source in instances

Table 2 Summary of IDDE Guidance Manual Benchmarks (Pitt 2004) and NYSDEC Part 703 Surface Water amp GroundwaterQuality Standards (NYSDEC 1999)

Contaminant NYS Standard Pitt Benchmark Possible Source

Ammonia mgL 2 mgL Fingerprint Value SewageDetergents ppm 4025 mgL Sewage washwater industrial

commercialE coli per 100 mL 412000 CFU100 ml SewageConductivity mScm Fingerprint Value IndustrialFluoride mgL 15 mgL Fingerprint Value Tapirrigation waterNitrate mgL 10 mgL Sewage fertilizerpH 65ndash85 IndustrialPotassium ppm Fingerprint Value Sewage industrialTurbidity NTU 5 NTU IndustrialAmmonia Potassium Ratio 410 Sewage

Other Indicators

Phosphorous - Total mgL Sewage fertilizerChlorine - Total mgL Tapirrigation water

See Tables 48 and 49 of Pitt (2004)

Figure 3 Flow chart for trackdown (after Pitt 2004)

386 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

where no pollutants were detected and no indicatorswere present for a flowing outfall

332 Trackdown process

Once an outfall sample indicated the possibility of anillicit discharge to a storm sewer system the trackdownprocess was initiated Field work first entailed a physicaltrackdown in the contributing area to determine pointsin the system where flow was present In doing so thevarious portions of the system where flow was absentcould be eliminated from further considerationWith thecontributing area defined the source of the illicitdischarge was identified more efficiently

There are two possible approaches for tracking downthe source of an illicit discharge collect samplesprogressively up-pipe at manholes and other accesspoints or use a sewer camera For the purpose of thisproject source trackdown did not utilize a sewer camerabecause the majority of the MS4 project partners do notown or have access to a camera However the cameraapproach was recommended in some instances toquickly and specifically identify a discharge point

Of the 64 outfalls evaluated through the ORI atotal of 24 were flowing at the time of inspection andwere sampled Fifteen of the outfalls required atrackdown investigation to identify a source ofcontamination Sixty seven samples were collectedand analyzed as part of the trackdown and recom-mendations for specific MS4 inspection to correct illicitconnections were made for a total of seven outfallsThe inspections and subsequent corrective actions arefacilitated by the local ordinance that the MS4sadopted as part of the WNYSC work (see Rossiet al 2009) It was beyond the scope of this project tofollow up with each MS4 to determine if correctiveactions were completed For the purpose of discussionone outfall known as Outfall 7 has been selected todescribe the process follow up reporting andrecommendations

333 Case study ndash outfall 7

Previous inspections at the outfall indicated flow waspresent in 2005 (GIS mapping survey httpgis1er-iegovENSSO) and 2007 (ORI for this study) Inspring 2008 Outfall 7 was revisited to collect asample Elevated levels of E coli (21600 CFU100 ml)ammonia (39 mgL) and fluoride (16 mgL) weredetected in the sample Storm sewer and land use mapswere obtained from the MS4 project partner for reviewand planning for additional trackdown samplingFollow up sampling at the outfall confirmed elevatedlevels of E coli (14000 CFU100 ml) ammonia(45 mgL) and nitrate (64 mgL) At that time a field

assessment was conducted to define the contributingarea for Outfall 7 confirm the accuracy of the sewermaps and plan for trackdown sampling

A trackdown was initiated on 8 July 2008 Onceagain the outfall was sampled and E coli levelsremained high (28000 CFU100 ml) as did theammonia (37 mgL) The first upstream site sampled(2 manhole 68 m upstream of Outfall 7) exhibiteda sharp increase in E coli (49000 CFU100 ml)detergent (10 mgL) and total phosphorus (25 mgL)Nitrate (48 mgL) and ammonia (38 mgL) levels wereelevated as well At the 3 manhole site (148 mupstream of Outfall 7) E coli levels (44000 CFU100 ml) and detergent (10 mgL) remained high whilenitrate and total phosphorus decreased substantiallyAmmonia remained consistent with previous levels (40mgL) Still farther upstream at the 4 manhole(238 m upstream of Outfall 7) the E coli (43000CFU100 ml) and ammonia (38 mgL) levels remainedhigh Detergent levels although slightly elevateddecreased substantially (125 mgL) One additionalupstream point was sampled on 6 August 2008 At thissite the 5 manhole (329 m upstream of Outfall 7)E coli levels declined (12500 CFU100 ml) to nearbenchmark limits (12000 CFU100 ml as defined byPitt 2004)

At the levels detected E coli contamination fromOutfall 7 became a flag of concern The elevatedlevels of E coli alone were indicative of raw sewagecontamination The high ammonia concentrationswere consistent with raw sewage and further reinforcedit as the source of contamination Similarly in thepresence of high ammonia concentrations the elevateddetergent levels provided credence to the suspectedsource The trackdown investigation defined thegeographic boundaries of the source of contaminationas being downstream from the 5 manhole

Further investigation by the MS4 to identify thesource of E coli contamination and eliminate it (usingthe newly established local stormwater ordinance) wasrecommended It also was recommended the MS4 re-inspect this site one year after eliminating the E colisource

34 Receiving water body results

341 E coli in receiving waters

High levels of pathogenic indicators including E coliare a concern in many water bodies throughout theUnited States and this contamination frequently is theprimary focus of watershed Total Maximum DailyLoad (TMDL) studies (eg US EPA 2001 He et al2007 Kay et al 2007 US EPA 2009) Water bodies inErie and Niagara Counties are no exception to this

Urban Water Journal 387

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

The Pitt (2004) flow chart values and the NYSDECStandards as summarized in Table 2 allowed for arapid determination as to whether or not a trackdownwas necessary Simply put if a given parameterexceeded the Pitt values or the NYSDEC Standardsfurther investigation was warranted The Pitt (2004)

flowchartindustrial benchmarks were used to char-acterize the (possible) source of flow as sanitarywastewater washwater irrigation tap water ground-water or industrialcommercial Municipal tap waterand ground water profiles (Eckhardt et al 2008) alsowere used to determine a possible source in instances

Table 2 Summary of IDDE Guidance Manual Benchmarks (Pitt 2004) and NYSDEC Part 703 Surface Water amp GroundwaterQuality Standards (NYSDEC 1999)

Contaminant NYS Standard Pitt Benchmark Possible Source

Ammonia mgL 2 mgL Fingerprint Value SewageDetergents ppm 4025 mgL Sewage washwater industrial

commercialE coli per 100 mL 412000 CFU100 ml SewageConductivity mScm Fingerprint Value IndustrialFluoride mgL 15 mgL Fingerprint Value Tapirrigation waterNitrate mgL 10 mgL Sewage fertilizerpH 65ndash85 IndustrialPotassium ppm Fingerprint Value Sewage industrialTurbidity NTU 5 NTU IndustrialAmmonia Potassium Ratio 410 Sewage

Other Indicators

Phosphorous - Total mgL Sewage fertilizerChlorine - Total mgL Tapirrigation water

See Tables 48 and 49 of Pitt (2004)

Figure 3 Flow chart for trackdown (after Pitt 2004)

386 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

where no pollutants were detected and no indicatorswere present for a flowing outfall

332 Trackdown process

Once an outfall sample indicated the possibility of anillicit discharge to a storm sewer system the trackdownprocess was initiated Field work first entailed a physicaltrackdown in the contributing area to determine pointsin the system where flow was present In doing so thevarious portions of the system where flow was absentcould be eliminated from further considerationWith thecontributing area defined the source of the illicitdischarge was identified more efficiently

There are two possible approaches for tracking downthe source of an illicit discharge collect samplesprogressively up-pipe at manholes and other accesspoints or use a sewer camera For the purpose of thisproject source trackdown did not utilize a sewer camerabecause the majority of the MS4 project partners do notown or have access to a camera However the cameraapproach was recommended in some instances toquickly and specifically identify a discharge point

Of the 64 outfalls evaluated through the ORI atotal of 24 were flowing at the time of inspection andwere sampled Fifteen of the outfalls required atrackdown investigation to identify a source ofcontamination Sixty seven samples were collectedand analyzed as part of the trackdown and recom-mendations for specific MS4 inspection to correct illicitconnections were made for a total of seven outfallsThe inspections and subsequent corrective actions arefacilitated by the local ordinance that the MS4sadopted as part of the WNYSC work (see Rossiet al 2009) It was beyond the scope of this project tofollow up with each MS4 to determine if correctiveactions were completed For the purpose of discussionone outfall known as Outfall 7 has been selected todescribe the process follow up reporting andrecommendations

333 Case study ndash outfall 7

Previous inspections at the outfall indicated flow waspresent in 2005 (GIS mapping survey httpgis1er-iegovENSSO) and 2007 (ORI for this study) Inspring 2008 Outfall 7 was revisited to collect asample Elevated levels of E coli (21600 CFU100 ml)ammonia (39 mgL) and fluoride (16 mgL) weredetected in the sample Storm sewer and land use mapswere obtained from the MS4 project partner for reviewand planning for additional trackdown samplingFollow up sampling at the outfall confirmed elevatedlevels of E coli (14000 CFU100 ml) ammonia(45 mgL) and nitrate (64 mgL) At that time a field

assessment was conducted to define the contributingarea for Outfall 7 confirm the accuracy of the sewermaps and plan for trackdown sampling

A trackdown was initiated on 8 July 2008 Onceagain the outfall was sampled and E coli levelsremained high (28000 CFU100 ml) as did theammonia (37 mgL) The first upstream site sampled(2 manhole 68 m upstream of Outfall 7) exhibiteda sharp increase in E coli (49000 CFU100 ml)detergent (10 mgL) and total phosphorus (25 mgL)Nitrate (48 mgL) and ammonia (38 mgL) levels wereelevated as well At the 3 manhole site (148 mupstream of Outfall 7) E coli levels (44000 CFU100 ml) and detergent (10 mgL) remained high whilenitrate and total phosphorus decreased substantiallyAmmonia remained consistent with previous levels (40mgL) Still farther upstream at the 4 manhole(238 m upstream of Outfall 7) the E coli (43000CFU100 ml) and ammonia (38 mgL) levels remainedhigh Detergent levels although slightly elevateddecreased substantially (125 mgL) One additionalupstream point was sampled on 6 August 2008 At thissite the 5 manhole (329 m upstream of Outfall 7)E coli levels declined (12500 CFU100 ml) to nearbenchmark limits (12000 CFU100 ml as defined byPitt 2004)

At the levels detected E coli contamination fromOutfall 7 became a flag of concern The elevatedlevels of E coli alone were indicative of raw sewagecontamination The high ammonia concentrationswere consistent with raw sewage and further reinforcedit as the source of contamination Similarly in thepresence of high ammonia concentrations the elevateddetergent levels provided credence to the suspectedsource The trackdown investigation defined thegeographic boundaries of the source of contaminationas being downstream from the 5 manhole

Further investigation by the MS4 to identify thesource of E coli contamination and eliminate it (usingthe newly established local stormwater ordinance) wasrecommended It also was recommended the MS4 re-inspect this site one year after eliminating the E colisource

34 Receiving water body results

341 E coli in receiving waters

High levels of pathogenic indicators including E coliare a concern in many water bodies throughout theUnited States and this contamination frequently is theprimary focus of watershed Total Maximum DailyLoad (TMDL) studies (eg US EPA 2001 He et al2007 Kay et al 2007 US EPA 2009) Water bodies inErie and Niagara Counties are no exception to this

Urban Water Journal 387

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

where no pollutants were detected and no indicatorswere present for a flowing outfall

332 Trackdown process

Once an outfall sample indicated the possibility of anillicit discharge to a storm sewer system the trackdownprocess was initiated Field work first entailed a physicaltrackdown in the contributing area to determine pointsin the system where flow was present In doing so thevarious portions of the system where flow was absentcould be eliminated from further considerationWith thecontributing area defined the source of the illicitdischarge was identified more efficiently

There are two possible approaches for tracking downthe source of an illicit discharge collect samplesprogressively up-pipe at manholes and other accesspoints or use a sewer camera For the purpose of thisproject source trackdown did not utilize a sewer camerabecause the majority of the MS4 project partners do notown or have access to a camera However the cameraapproach was recommended in some instances toquickly and specifically identify a discharge point

Of the 64 outfalls evaluated through the ORI atotal of 24 were flowing at the time of inspection andwere sampled Fifteen of the outfalls required atrackdown investigation to identify a source ofcontamination Sixty seven samples were collectedand analyzed as part of the trackdown and recom-mendations for specific MS4 inspection to correct illicitconnections were made for a total of seven outfallsThe inspections and subsequent corrective actions arefacilitated by the local ordinance that the MS4sadopted as part of the WNYSC work (see Rossiet al 2009) It was beyond the scope of this project tofollow up with each MS4 to determine if correctiveactions were completed For the purpose of discussionone outfall known as Outfall 7 has been selected todescribe the process follow up reporting andrecommendations

333 Case study ndash outfall 7

Previous inspections at the outfall indicated flow waspresent in 2005 (GIS mapping survey httpgis1er-iegovENSSO) and 2007 (ORI for this study) Inspring 2008 Outfall 7 was revisited to collect asample Elevated levels of E coli (21600 CFU100 ml)ammonia (39 mgL) and fluoride (16 mgL) weredetected in the sample Storm sewer and land use mapswere obtained from the MS4 project partner for reviewand planning for additional trackdown samplingFollow up sampling at the outfall confirmed elevatedlevels of E coli (14000 CFU100 ml) ammonia(45 mgL) and nitrate (64 mgL) At that time a field

assessment was conducted to define the contributingarea for Outfall 7 confirm the accuracy of the sewermaps and plan for trackdown sampling

A trackdown was initiated on 8 July 2008 Onceagain the outfall was sampled and E coli levelsremained high (28000 CFU100 ml) as did theammonia (37 mgL) The first upstream site sampled(2 manhole 68 m upstream of Outfall 7) exhibiteda sharp increase in E coli (49000 CFU100 ml)detergent (10 mgL) and total phosphorus (25 mgL)Nitrate (48 mgL) and ammonia (38 mgL) levels wereelevated as well At the 3 manhole site (148 mupstream of Outfall 7) E coli levels (44000 CFU100 ml) and detergent (10 mgL) remained high whilenitrate and total phosphorus decreased substantiallyAmmonia remained consistent with previous levels (40mgL) Still farther upstream at the 4 manhole(238 m upstream of Outfall 7) the E coli (43000CFU100 ml) and ammonia (38 mgL) levels remainedhigh Detergent levels although slightly elevateddecreased substantially (125 mgL) One additionalupstream point was sampled on 6 August 2008 At thissite the 5 manhole (329 m upstream of Outfall 7)E coli levels declined (12500 CFU100 ml) to nearbenchmark limits (12000 CFU100 ml as defined byPitt 2004)

At the levels detected E coli contamination fromOutfall 7 became a flag of concern The elevatedlevels of E coli alone were indicative of raw sewagecontamination The high ammonia concentrationswere consistent with raw sewage and further reinforcedit as the source of contamination Similarly in thepresence of high ammonia concentrations the elevateddetergent levels provided credence to the suspectedsource The trackdown investigation defined thegeographic boundaries of the source of contaminationas being downstream from the 5 manhole

Further investigation by the MS4 to identify thesource of E coli contamination and eliminate it (usingthe newly established local stormwater ordinance) wasrecommended It also was recommended the MS4 re-inspect this site one year after eliminating the E colisource

34 Receiving water body results

341 E coli in receiving waters

High levels of pathogenic indicators including E coliare a concern in many water bodies throughout theUnited States and this contamination frequently is theprimary focus of watershed Total Maximum DailyLoad (TMDL) studies (eg US EPA 2001 He et al2007 Kay et al 2007 US EPA 2009) Water bodies inErie and Niagara Counties are no exception to this

Urban Water Journal 387

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

concern and bacterial contamination along the Buffalowaterfront has been and continues to be a focus in theBuffalo Sewer Authority CSO Long Term ControlPlan (NYSDEC 2005 Irvine et al 2005 2006a) Assuch this section separately discusses the issue of Ecoli levels in receiving waters

342 E coli characteristics Cayuga CreekNiagara County

A total of 62 samples (21 dry weather days 41 wetweather days) were collected at the upstream anddownstream sites on Cayuga Creek between 27 July2007 and 25 July 2008 for E coli analysis and theresults are summarized in Figure 4 The maximummeasured E coli level for a storm event at theupstream site was 18900 CFU100 mL and themaximum measured E coli level for the downstreamsite was 11500 CFU100 mL However as shown inFigure 4 there was a general trend towards increasingE coli levels in the downstream direction for both dryweather and storm event conditions In general theupstream site exhibited relatively low E coli levelsMean storm event levels were higher than dry weatherflow at both sites

Although the municipalities in the Cayuga Creekwatershed were not participants in this demonstrationproject there are three permitted stormwater outfalls(NI6 NI5 NI4) between the upstream and down-stream sites (Figure 5) Samples were not collected atthese outfalls during dry weather flow Howeversamples were collected at most of the stormwateroutfalls shown in Figure 5 in association with thestorm event of 21 October 2008 (04 inches (101 mm)of rain) The geometric mean E coli level at theoutfalls for this event was 711 CFU100 mL which istaken as the lsquolsquoexpectedrsquorsquo level (Figure 6) Site NI4(Figures 5 and 6) in particular could be an

important source of E coli during storm eventsThere also may be other sources of E coli to CayugaCreek between the upstream and downstream sitesincluding failing septic systems and natural wildlifeand these various sources should be investigatedfurther

343 E coli levels at Niagara River Two Mile Creekand Green Lake Tributary sites

Although two of four outfalls examined between thetwo Niagara River receiving water sites had E colilevels in the range of 21600ndash115900 CFU100 mLthere was no observable impact on the Niagara Riverdownstream site The two samples collected at theNiagara River upstream site had E coli levels of 200CFU100 mL while the two samples collected at theNiagara River downstream site had E coli levels of300 and 800 CFU100 mL on 22 July 2008 and 13August 2008 The high flow volume in the NiagaraRiver was quite sufficient to dilute the small volumestormwater discharges The E coli levels at theupstream (0 CFU100 mL) and downstream (600CFU100 mL) sites on the tributary from GreenLake also did not reflect significant impact from thesingle stormwater discharge point between them (150CFU100 mL)

E coli levels at the upstream and downstreamlocations on Two Mile Creek may reflect inputs fromthe stormwater discharge points For example on 22July 2008 E coli levels at the upstream site were 2200100 mL and the downstream site were 4100 CFU100 mL One of the sampled outfalls between the twosites had an E coli level of 9700 CFU100 mL

344 E coli levels in Scajaquada Creek ndash storm eventconditions

Sampling was conducted during storm events atvarious combinations of the four sample sites onScajaquada Creek depending on team availability andstorm duration These sites were selected as beingcommon to the Buffalo Sewer Authorityrsquos CSO LongTerm Control Plan Phase II Study Samples werecollected for the storm events of 24 July 2008 1October 2008 and 16 October 2008 and the results aresummarized in Table 3 The levels of E coli observedin Scajaquada Creek for storm events generally werehigher than those observed for Cayuga Creek Severalcity of Buffalo CSOs discharge to Scajaquada Creekbetween sites 1 and 2 and CSO points also are locatedin the vicinity of sites 3 and 4

Because of the likely contributions of CSOs toScajaquada Creek during storm events it is difficult toseparate contributions from stormwater dischargesFigure 4 Geometric mean E coli levels in Cayuga Creek

388 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

and CSOs For the 24 July 2008 event three of fourBuffalo State stormwater discharge points locatedbetween Scajaquada Creek sample sites 3 and 4 had

E coli levels of 2500ndash4200 CFU100 mL which werecomparable to the levels observed in ScajaquadaCreek It also seems likely that the CSO impact was

Figure 5 Permitted stormwater discharge locations Cayuga Creek Niagara County NI6 NI5 and NI4 are between theupstream and downstream sample sites on Cayuga Creek

Urban Water Journal 389

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

observed at site 2 for the sample collected at 1320(35800 CFU100 mL) but the CSO discharge was notreflected in the samples collected at sites 3 and 4 earlierin the day Further sampling and possible modelapplication needs to be done to more adequately assessCSO vs stormwater inputs

For the storm event of 16 October 2008 a decreasein E coli levels was observed through the event inassociation with a decreasing flow Irvine et al (2002)showed that fecal coliform levels were significantlycorrelated with suspended solids concentrations inlocal rivers and we expect E coli would exhibit asimilar relationship As such the E coli and suspendedsolids from the various sources (including stormwaterdischarges CSOs and bed and bank sedimentresuspension) will have flushed through the systemearlier in the 16 October 2008 event and towards theend of the event also will start to settle out of the watercolumn with the lower flow

Table 3 E coli levels for storm events on Scajaquada Creek

Date Time SiteE Coli

CFU100 ML Comments

24 July 2008 1030 3 4400 Rain during and immediately prior to sampling totaled 86ndash102 mm11201320

42

650035800

(2 gauges)

1 October 20081 October 2008

820835

34

2320029500

Rain around 2300 the night before of 48-102 mm (multiple gaugesthroughout the city sewage odor at site 4 flow was turbid and visually

16 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 2008

110016409051130171092511501730940121317451000

111222333444

33000182006900

47600202003820027600178009400

17400128004400

higher than baseflow)Rain the previous night of 201 mm

Figure 6 E coli results for stormwater discharges to Cayuga Creek event of 102108 Site mapping ID codes for locationsbetween the upstream and downstream sites in Cayuga Creek are NI6 frac14 16508 NI5 frac14 16507 and NI4 frac14 16506

35 QAQC results

351 E coli analysis Coliscan Easygel1 vsmembrane filtration

The results of the Coliscan Easygel1 and membranefiltration analysis (Erie County Public Health Depart-ment) for receiving water (n frac14 11) and stormwater sewersamples (n frac14 12) combined (total n frac14 21) are shown inFigure 7 The samples represented a range of E colilevels and the regression indicates there was goodcorrespondence between the Coliscan Easygel1 resultsand the standard membrane filtration The slope of theregression line between the Coliscan Easygel1 resultsand membrane filtration was close to 1 and was signifi-cantly different from 0 (a frac14 005) while the interceptwas not significantly different from 0 (a frac14 005)

Citizenrsquos groups throughout the US successfullyhave used the Coliscan Easygel1 system in monitoringprograms (eg Alabama Water Watch (1999) VirginiaCitizen Water Quality Monitoring Program (2003)

390 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Texas Watch (httpwwwtexaswatchgeotxstateeduNewsletters98ndash04pdf ) Hoosier Riverwatch (httpwwwingovdnrriverwatchpdfmanualChap4pdf)Alliance for the Cheasapeake Bay Citizens Monitor(httpwwwacb-onlineorgpubsprojectsdeliverables-87ndash1ndash2003pdf) University of Vermont (2003)) and wehave used the system for research projects in Cambodiaand Thailand (eg Krueger et al 2004 Gugino et al2006 Irvine et al 2006b Visoth et al 2010) The VirginiaDepartment of Environmental Quality approved theColiscan Easygel1 method for screening purposes andindependent testing (eg Alabama Water Watch 1999Deutsch and Busby 2000) has shown that ColiscanEasygel1 results are comparable to standard methodsOur results also suggest that the Coliscan Easygel1

system can be used to reliably and cost-effectively trackdown illicit connections and problem discharges

352 Certified standard analysis with HannaInstruments photometers

Certified standards for potassium nitrate fluoridechromium VI copper total phosphorus and ammoniawere analyzed in triplicate All parameters exceptpotassium were measured using the Hanna Instru-ments C 200 photometer Potassium was measuredusing the Hanna Instruments HI 93750 photometerResults of the mean of triplicate samples for potas-sium nitrate chromium VI copper and fluoride areshown in Figures 8ndash12 The Hanna Instruments C 200photometer recorded a mean triplicate result of 103mgL for the total phosphorus standard of 10 mgLThe Hanna Instruments C 200 photometer alsorecorded a mean triplicate result of 041 and 103mgL for the ammonia standard of 05 and 10 mgLrespectively Results of the certified standards analysisshow that the Hanna Instruments photometers werecapable of producing reliable results suitable for atrackdown study

353 Comparison of results between Hanna Instru-ments photometers and certified laboratory

Results of the 11 receiving water samples and 12 stormsewer samples analyzed using the Hanna Instrumentsphotometers and the New York State Health Depart-ment certified laboratory are summarized in Table 4The mean level for all parameters was higher for the

Figure 7 Comparison of E coli levels determined usingColiscan Easygel1 and membrane filtration Erie CountyPublic Health Lab (ECPHL)

Figure 8 Potassium standards analysis with Hanna HI93750 photometer Each point is the mean of a triplicateanalysis

Figure 9 Nitrate as NO3- standards analysis with the

Hanna C 200 photometer Each point is the mean of atriplicate analysis

Figure 10 Chromium VI standards analysis with theHanna C 200 photometer Each point is the mean of atriplicate analysis

Urban Water Journal 391

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

and CSOs For the 24 July 2008 event three of fourBuffalo State stormwater discharge points locatedbetween Scajaquada Creek sample sites 3 and 4 had

E coli levels of 2500ndash4200 CFU100 mL which werecomparable to the levels observed in ScajaquadaCreek It also seems likely that the CSO impact was

Figure 5 Permitted stormwater discharge locations Cayuga Creek Niagara County NI6 NI5 and NI4 are between theupstream and downstream sample sites on Cayuga Creek

Urban Water Journal 389

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

observed at site 2 for the sample collected at 1320(35800 CFU100 mL) but the CSO discharge was notreflected in the samples collected at sites 3 and 4 earlierin the day Further sampling and possible modelapplication needs to be done to more adequately assessCSO vs stormwater inputs

For the storm event of 16 October 2008 a decreasein E coli levels was observed through the event inassociation with a decreasing flow Irvine et al (2002)showed that fecal coliform levels were significantlycorrelated with suspended solids concentrations inlocal rivers and we expect E coli would exhibit asimilar relationship As such the E coli and suspendedsolids from the various sources (including stormwaterdischarges CSOs and bed and bank sedimentresuspension) will have flushed through the systemearlier in the 16 October 2008 event and towards theend of the event also will start to settle out of the watercolumn with the lower flow

Table 3 E coli levels for storm events on Scajaquada Creek

Date Time SiteE Coli

CFU100 ML Comments

24 July 2008 1030 3 4400 Rain during and immediately prior to sampling totaled 86ndash102 mm11201320

42

650035800

(2 gauges)

1 October 20081 October 2008

820835

34

2320029500

Rain around 2300 the night before of 48-102 mm (multiple gaugesthroughout the city sewage odor at site 4 flow was turbid and visually

16 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 2008

110016409051130171092511501730940121317451000

111222333444

33000182006900

47600202003820027600178009400

17400128004400

higher than baseflow)Rain the previous night of 201 mm

Figure 6 E coli results for stormwater discharges to Cayuga Creek event of 102108 Site mapping ID codes for locationsbetween the upstream and downstream sites in Cayuga Creek are NI6 frac14 16508 NI5 frac14 16507 and NI4 frac14 16506

35 QAQC results

351 E coli analysis Coliscan Easygel1 vsmembrane filtration

The results of the Coliscan Easygel1 and membranefiltration analysis (Erie County Public Health Depart-ment) for receiving water (n frac14 11) and stormwater sewersamples (n frac14 12) combined (total n frac14 21) are shown inFigure 7 The samples represented a range of E colilevels and the regression indicates there was goodcorrespondence between the Coliscan Easygel1 resultsand the standard membrane filtration The slope of theregression line between the Coliscan Easygel1 resultsand membrane filtration was close to 1 and was signifi-cantly different from 0 (a frac14 005) while the interceptwas not significantly different from 0 (a frac14 005)

Citizenrsquos groups throughout the US successfullyhave used the Coliscan Easygel1 system in monitoringprograms (eg Alabama Water Watch (1999) VirginiaCitizen Water Quality Monitoring Program (2003)

390 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Texas Watch (httpwwwtexaswatchgeotxstateeduNewsletters98ndash04pdf ) Hoosier Riverwatch (httpwwwingovdnrriverwatchpdfmanualChap4pdf)Alliance for the Cheasapeake Bay Citizens Monitor(httpwwwacb-onlineorgpubsprojectsdeliverables-87ndash1ndash2003pdf) University of Vermont (2003)) and wehave used the system for research projects in Cambodiaand Thailand (eg Krueger et al 2004 Gugino et al2006 Irvine et al 2006b Visoth et al 2010) The VirginiaDepartment of Environmental Quality approved theColiscan Easygel1 method for screening purposes andindependent testing (eg Alabama Water Watch 1999Deutsch and Busby 2000) has shown that ColiscanEasygel1 results are comparable to standard methodsOur results also suggest that the Coliscan Easygel1

system can be used to reliably and cost-effectively trackdown illicit connections and problem discharges

352 Certified standard analysis with HannaInstruments photometers

Certified standards for potassium nitrate fluoridechromium VI copper total phosphorus and ammoniawere analyzed in triplicate All parameters exceptpotassium were measured using the Hanna Instru-ments C 200 photometer Potassium was measuredusing the Hanna Instruments HI 93750 photometerResults of the mean of triplicate samples for potas-sium nitrate chromium VI copper and fluoride areshown in Figures 8ndash12 The Hanna Instruments C 200photometer recorded a mean triplicate result of 103mgL for the total phosphorus standard of 10 mgLThe Hanna Instruments C 200 photometer alsorecorded a mean triplicate result of 041 and 103mgL for the ammonia standard of 05 and 10 mgLrespectively Results of the certified standards analysisshow that the Hanna Instruments photometers werecapable of producing reliable results suitable for atrackdown study

353 Comparison of results between Hanna Instru-ments photometers and certified laboratory

Results of the 11 receiving water samples and 12 stormsewer samples analyzed using the Hanna Instrumentsphotometers and the New York State Health Depart-ment certified laboratory are summarized in Table 4The mean level for all parameters was higher for the

Figure 7 Comparison of E coli levels determined usingColiscan Easygel1 and membrane filtration Erie CountyPublic Health Lab (ECPHL)

Figure 8 Potassium standards analysis with Hanna HI93750 photometer Each point is the mean of a triplicateanalysis

Figure 9 Nitrate as NO3- standards analysis with the

Hanna C 200 photometer Each point is the mean of atriplicate analysis

Figure 10 Chromium VI standards analysis with theHanna C 200 photometer Each point is the mean of atriplicate analysis

Urban Water Journal 391

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

observed at site 2 for the sample collected at 1320(35800 CFU100 mL) but the CSO discharge was notreflected in the samples collected at sites 3 and 4 earlierin the day Further sampling and possible modelapplication needs to be done to more adequately assessCSO vs stormwater inputs

For the storm event of 16 October 2008 a decreasein E coli levels was observed through the event inassociation with a decreasing flow Irvine et al (2002)showed that fecal coliform levels were significantlycorrelated with suspended solids concentrations inlocal rivers and we expect E coli would exhibit asimilar relationship As such the E coli and suspendedsolids from the various sources (including stormwaterdischarges CSOs and bed and bank sedimentresuspension) will have flushed through the systemearlier in the 16 October 2008 event and towards theend of the event also will start to settle out of the watercolumn with the lower flow

Table 3 E coli levels for storm events on Scajaquada Creek

Date Time SiteE Coli

CFU100 ML Comments

24 July 2008 1030 3 4400 Rain during and immediately prior to sampling totaled 86ndash102 mm11201320

42

650035800

(2 gauges)

1 October 20081 October 2008

820835

34

2320029500

Rain around 2300 the night before of 48-102 mm (multiple gaugesthroughout the city sewage odor at site 4 flow was turbid and visually

16 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 200816 October 200816 October 200817 October 2008

110016409051130171092511501730940121317451000

111222333444

33000182006900

47600202003820027600178009400

17400128004400

higher than baseflow)Rain the previous night of 201 mm

Figure 6 E coli results for stormwater discharges to Cayuga Creek event of 102108 Site mapping ID codes for locationsbetween the upstream and downstream sites in Cayuga Creek are NI6 frac14 16508 NI5 frac14 16507 and NI4 frac14 16506

35 QAQC results

351 E coli analysis Coliscan Easygel1 vsmembrane filtration

The results of the Coliscan Easygel1 and membranefiltration analysis (Erie County Public Health Depart-ment) for receiving water (n frac14 11) and stormwater sewersamples (n frac14 12) combined (total n frac14 21) are shown inFigure 7 The samples represented a range of E colilevels and the regression indicates there was goodcorrespondence between the Coliscan Easygel1 resultsand the standard membrane filtration The slope of theregression line between the Coliscan Easygel1 resultsand membrane filtration was close to 1 and was signifi-cantly different from 0 (a frac14 005) while the interceptwas not significantly different from 0 (a frac14 005)

Citizenrsquos groups throughout the US successfullyhave used the Coliscan Easygel1 system in monitoringprograms (eg Alabama Water Watch (1999) VirginiaCitizen Water Quality Monitoring Program (2003)

390 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Texas Watch (httpwwwtexaswatchgeotxstateeduNewsletters98ndash04pdf ) Hoosier Riverwatch (httpwwwingovdnrriverwatchpdfmanualChap4pdf)Alliance for the Cheasapeake Bay Citizens Monitor(httpwwwacb-onlineorgpubsprojectsdeliverables-87ndash1ndash2003pdf) University of Vermont (2003)) and wehave used the system for research projects in Cambodiaand Thailand (eg Krueger et al 2004 Gugino et al2006 Irvine et al 2006b Visoth et al 2010) The VirginiaDepartment of Environmental Quality approved theColiscan Easygel1 method for screening purposes andindependent testing (eg Alabama Water Watch 1999Deutsch and Busby 2000) has shown that ColiscanEasygel1 results are comparable to standard methodsOur results also suggest that the Coliscan Easygel1

system can be used to reliably and cost-effectively trackdown illicit connections and problem discharges

352 Certified standard analysis with HannaInstruments photometers

Certified standards for potassium nitrate fluoridechromium VI copper total phosphorus and ammoniawere analyzed in triplicate All parameters exceptpotassium were measured using the Hanna Instru-ments C 200 photometer Potassium was measuredusing the Hanna Instruments HI 93750 photometerResults of the mean of triplicate samples for potas-sium nitrate chromium VI copper and fluoride areshown in Figures 8ndash12 The Hanna Instruments C 200photometer recorded a mean triplicate result of 103mgL for the total phosphorus standard of 10 mgLThe Hanna Instruments C 200 photometer alsorecorded a mean triplicate result of 041 and 103mgL for the ammonia standard of 05 and 10 mgLrespectively Results of the certified standards analysisshow that the Hanna Instruments photometers werecapable of producing reliable results suitable for atrackdown study

353 Comparison of results between Hanna Instru-ments photometers and certified laboratory

Results of the 11 receiving water samples and 12 stormsewer samples analyzed using the Hanna Instrumentsphotometers and the New York State Health Depart-ment certified laboratory are summarized in Table 4The mean level for all parameters was higher for the

Figure 7 Comparison of E coli levels determined usingColiscan Easygel1 and membrane filtration Erie CountyPublic Health Lab (ECPHL)

Figure 8 Potassium standards analysis with Hanna HI93750 photometer Each point is the mean of a triplicateanalysis

Figure 9 Nitrate as NO3- standards analysis with the

Hanna C 200 photometer Each point is the mean of atriplicate analysis

Figure 10 Chromium VI standards analysis with theHanna C 200 photometer Each point is the mean of atriplicate analysis

Urban Water Journal 391

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Texas Watch (httpwwwtexaswatchgeotxstateeduNewsletters98ndash04pdf ) Hoosier Riverwatch (httpwwwingovdnrriverwatchpdfmanualChap4pdf)Alliance for the Cheasapeake Bay Citizens Monitor(httpwwwacb-onlineorgpubsprojectsdeliverables-87ndash1ndash2003pdf) University of Vermont (2003)) and wehave used the system for research projects in Cambodiaand Thailand (eg Krueger et al 2004 Gugino et al2006 Irvine et al 2006b Visoth et al 2010) The VirginiaDepartment of Environmental Quality approved theColiscan Easygel1 method for screening purposes andindependent testing (eg Alabama Water Watch 1999Deutsch and Busby 2000) has shown that ColiscanEasygel1 results are comparable to standard methodsOur results also suggest that the Coliscan Easygel1

system can be used to reliably and cost-effectively trackdown illicit connections and problem discharges

352 Certified standard analysis with HannaInstruments photometers

Certified standards for potassium nitrate fluoridechromium VI copper total phosphorus and ammoniawere analyzed in triplicate All parameters exceptpotassium were measured using the Hanna Instru-ments C 200 photometer Potassium was measuredusing the Hanna Instruments HI 93750 photometerResults of the mean of triplicate samples for potas-sium nitrate chromium VI copper and fluoride areshown in Figures 8ndash12 The Hanna Instruments C 200photometer recorded a mean triplicate result of 103mgL for the total phosphorus standard of 10 mgLThe Hanna Instruments C 200 photometer alsorecorded a mean triplicate result of 041 and 103mgL for the ammonia standard of 05 and 10 mgLrespectively Results of the certified standards analysisshow that the Hanna Instruments photometers werecapable of producing reliable results suitable for atrackdown study

353 Comparison of results between Hanna Instru-ments photometers and certified laboratory

Results of the 11 receiving water samples and 12 stormsewer samples analyzed using the Hanna Instrumentsphotometers and the New York State Health Depart-ment certified laboratory are summarized in Table 4The mean level for all parameters was higher for the

Figure 7 Comparison of E coli levels determined usingColiscan Easygel1 and membrane filtration Erie CountyPublic Health Lab (ECPHL)

Figure 8 Potassium standards analysis with Hanna HI93750 photometer Each point is the mean of a triplicateanalysis

Figure 9 Nitrate as NO3- standards analysis with the

Hanna C 200 photometer Each point is the mean of atriplicate analysis

Figure 10 Chromium VI standards analysis with theHanna C 200 photometer Each point is the mean of atriplicate analysis

Urban Water Journal 391

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Hanna Instruments photometers than the resultsobtained from the certified laboratory In the case ofchromium and copper frequently the levels were nearor below the detection limit for the certified laboratoryThe standard deviations for the Hanna Instrumentsresults were considerably higher than those for certifiedlaboratory Considering the good results of the HannaInstruments photometers with the certified standardsthe comparisons with certified laboratory were some-what disappointing and should be investigated furtherA review of the Hanna Instruments data indicated thatindividual sample results could be high In the futureanomalous values should be flagged and re-done in

triplicate Furthermore the samples were not filteredprior to analysis with the Hanna Instruments photo-meters Higher suspended solids levels can result ininterference with the photometric analysis and testingof both filtered and unfiltered environmental samplesshould be carried out The samples for the certifiedlaboratory were not filtered prior to analysis so thisrules out analytical differences as a result of differentsample preparation

354 Comparisons for dissolved oxygen and conduc-tivity tests with YSI 6920 results

The regression between the CHEMetrics visual test fordissolved oxygen and measurements made by a YSI6920 datasonde is shown in Figure 13 The relationshipis strong and some of the scatter occurred due to thecoarser resolution of the visual test (05 mgL) Giventhe ease of the visual test and its low relative cost theseresults were quite encouraging

The Hanna Instruments HI9811 field meter mea-surements of conductivity (project meter) were com-pared to the YSI 6920 datasonde measurements(Figure 14) The Hanna Instruments meter is arelatively low-cost unit but the results comparedfavourably with those provided by the YSI 6920

Table 4 Mean values (standard deviation in brackets) ofcertified laboratory and Hanna photometer results

Parameter Certified LabHanna

Photometer

chromium (mgL) 00024 (00013) 00127 (00195)copper (mgL) 0008 (0007) 0288 (0310)potassium (mgL) 39 (23) 193 (163)ammonia (mgL) 0421 (0977) 213 (295)nitrate (mgL) 0574 (0418) 58 (79)total phosphorus (mgL) 0135 (0198) 143 (112)

Figure 11 Copper standards analysis with Hanna C 200photometer Each point is the mean of a triplicate analysis

Figure 12 Fluoride standards analysis with Hanna C 200photometer Each point is the mean of a triplicate analysis

Figure 13 Regression between CHEMetrics and YSIdissolved oxygen measurement

Figure 14 Hanna HI9811 (Project) conductivitymeasurements compared with the YSI 6920 measurements

392 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

One pair of measurements exhibited some deviationfrom the regression line where the HI9811 meterrecorded a value of 1000m Scm but the YSI 6920recorded a value of 512 mScm For this particular setof measurements the flow from the outfall pipe wasquite shallow and it was difficult to get the YSI 6920sensor submersed in the flow The HI9811 measure-ment was taken from a sample bottle and it is likelythat the HI9811 reading more accurately reflected thetrue conductivity of the outfall

4 Technology transfer

The primary objective of this project was to demon-strate a protocol for small MS4s to detect sample andtrack down the source of illicit discharges to theirstorm sewer systems The MS4s of the WNYSCreceived a series of training workshops (samplecollection and analysis ORI interpretation of track-down results) and opportunities to provide feedbackthroughout the course of the project

In March 2009 a final training workshop was heldfor all MS4s in the WNYSC Based on the feedbackand concerns voiced at the previous training work-shops the sampling protocol and trackdown proce-dures were streamlined to be an efficient cost-effectiveprocess that is compliant with New York Statersquos PhaseII Stormwater regulations

5 Cost comparisons

One of the objectives of this study was to identify cost-effective robust and accurate analytical methods forthe trackdown studies Two comparisons are offeredhere

51 E coli

Erie County Public Health Lab frac14 $1607sample(note that this is a low price for a New YorkState Laboratory)

Coliscan Easy Gel1 - $160 (materials) thorn $333(labour) frac14 $493sample

52 Suite of total phosphorus nitrate potassiumchromium and copper

New York State Health Department certifiedlaboratory frac14 $105sample

Kit reagents ($331) thorn $1333 (labour) thorn $177 (Capital equipment costs) frac14 $1841sam-ple(Capital costs here assume new photo-meters every 500 samples which is veryconservative)

Clearly the analytical approaches examined in thisproject offer a cost-effective alternative to contractlaboratory costs

6 Conclusion

The outcome of this project was two-fold a simpleprocess to detect and eliminate illicit discharges tostorm sewer systems and a reliable cost-effectiveanalytical method to perform the sampling andchemical analysis required to properly identify anillicit discharge

The demonstration component of the projectentailed sampling and analysis for a broad spectrumof parameters to determine a source for the contam-ination andor flow Initially there were twentyanalytical parameters measured for each sampleEach parameter was examined as an indicator tocharacterize the flow as illicit and to identify the sourceof flow (ie sanitary washwater industrial ground-water etc) As the various flow identification andtrackdown analyses were conducted it soon becameevident that a number of the parameters analyzedsimply did not provide a hard indicator for sourceidentification In some instances such as copperdetections at a wide range of concentrations wereprevalent in all samples collected tap water andgroundwater included In other cases the parameterjust did not come into play for source identificationUltimately twelve parameters were selected for theIDDE guidance document temperature pH conduc-tivity E coli ammonia nitrate fluoride totalchlorine potassium detergents phosphorous andturbidity

To achieve the goal of determining cost-effectivechemical analyses procedures the project entailed useof simple kits and colorimetric approaches Theanalytical methods were inexpensive and easy to use(detailed in Section 27) To confirm the reliability andaccuracy of the analytical methods a variety of QAQCassessments were applied and these indicated themethods provided a level of accuracy that wasconducive to their application for illicit dischargedetection and trackdown sampling purposes Theresults were particularly encouraging for the E colianalysis Care should be taken with the photometricanalyses and if extreme values are measured the sampleshould be re-analyzed in triplicate

Following the success of the demonstration projectthe WNYSC utilized grant funds to purchase equip-ment and supplies for outfall sampling and in-housechemical analyses for its entire membership Theseoutfall sampling kits are comprehensive in that every-thing required for an illicit discharge trackdowninvestigation is included

Urban Water Journal 393

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Through its partnership with the WNYSC mem-bership the Erie County Department of Environmentand Planning coordinated the purchase assembly anddelivery of the outfall kits In addition field trainingfor MS4 staff on sampling procedures laboratorytraining for sample analysis results interpretationguidance and trackdown and elimination strategiesare currently underway Erie County Department ofEnvironment and Planning will continue to be aresource for the MS4 project partners for assistancewith interpretation of analytical results trackdownprocedures and source identification

While the two primary objectives of the study wereachieved in reality it is often difficult to progress to thenext level that being effective elimination of the sourceof illicit connection The foremost reason for this is alack of financial resources for many of the MS4sSocio-economic variability among and within the MS4communities is considerable At the municipal levelthis translates into challenges to dedicate manpowerandor equipment at a minimum More costlyconstraints would be the need for large scale infra-structure improvements or re-construction projects toeliminate illicit discharges On a smaller scale similarfinancial limitations may exist for residential propertyowners when corrective action is the responsibility of aproperty owner Simply put medium-high densityresidential areas are common among MS4 commu-nities and often have a large proportion of economic-ally disadvantaged sections within them Stormwaterutilities have been successfully established in somestates to help address financial concerns (eg Brisman2001) but a recent study completed for the WNYSCfound a general resistance to this approach at the MS4and public levels in Western New York (WendelDuchscherer et al 2010) Specifically public consulta-tion suggested that a stormwater utility was perceivedas another layer of government with increased fees andless local control

Finally although this study was site-specific toWestern New York in fact many of the trackdowntechniques discussed herein could be used at locationsthroughout North America and even globally sincefrequently there are common water quality concernswith respect to contaminant type and source Thelaboratory methods used in this study were cost-effective and robust making them particularly attrac-tive for developing countries and we have applied thesemethods successfully in Cambodia Thailand andMalaysia The trackdown approach also is flexibleenough to accommodate local knowledge aboutspecific contaminants of concern and these could beadded as needed to help identify sources of illicitconnections

Acknowledgements

1 Thanks to Mary Perrelli Joseph Drakes RatchadawanNgoen-klan Brent LaSpada Jason Pantano John Karczand Gordon Irvine for their assistance in the field andlaboratory2 Funding for this project was provided by the US EPARegion 2 Project Number CP-97280200ndash0 Mention ofproduct names in this report does not imply endorsementby the US EPA nor does this paper necessarily reflect theviews of the US EPA

References

Alabama Water Watch 1999 Quality Assurance Plan forBacteriological Monitoring Report for the US Envir-onmental Protection Agency Region 4

APHA (American Public Health Association) 1998 Stan-dard Methods for the Examination of Water and Waste-water 20th ed Washington DC

Brisman A 2001 Considerations in establishing a storm-water utility Southern Illinois University Law Journal 26505ndash528

Clarkson HE 2003 Creating a stormwater utility withnear-zero opposition Stormwater The Journal of Sur-face Water Quality Professionals JulyAugust 13pp

Deutsch WG and Busby AL 2000 A Field andLaboratory Manual for a Community-based WaterQuality Monitoring Program in the Philippines Reportfor the Heifer Project International

Eckhardt DAV Reddy JE and Tamulonis KL 2008Ground-water Quality in Western New York 2006 USGeological Survey Open-File Report 2008ndash1140 36 pavailable only online at httppubsusgsgovofr20081140

Gugino H Irvine K Ngern-klun R Sukontason KSukontason K Prangkio C and Thunyapar P 2006Impacts of the urban environment on area source waterthe Klong Mae Kha - Chiang Mai Thailand Proceedingsof the Fourth International Synposium on Southeast AsianWater Environment Asian Institute of Technology Bang-kok pp 85ndash88

He L-M Lu J and ShiW 2007Variability of fecal indicatorbacteria in flowing and ponded waters in southernCalifornia Implications for bacterial TMDL developmentand implementationWater Research 41 3132ndash3140

Irvine KN Somogye EL and Pettibone GW 2002Turbidity suspended solids and bacteria relationships inthe Buffalo River watershed Middle States Geographer35 42ndash51

Irvine KN Vermette SJ Tang T Sampson M andMurphy TP 2004 Partnering to provide water qualityand GIS training in Cambodia Proceedings 25th AsianConference on Remote Sensing Asian Association onRemote Sensing Bangkok

Irvine KN Perrelli MF McCorkhill G and Caruso J2005 Sampling and modeling approaches to assess waterquality impacts of combined sewer overflows ndash theimportance of a watershed perspective Journal of GreatLakes Research 31 105ndash115

Irvine KN Perrelli M Brown N and Rossi MC 2006aSeptic system pollution prevention BMPs Developmentof public outreach approaches and decision making toolsfor local government Clearwaters 36 (3) 51ndash55

394 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

One pair of measurements exhibited some deviationfrom the regression line where the HI9811 meterrecorded a value of 1000m Scm but the YSI 6920recorded a value of 512 mScm For this particular setof measurements the flow from the outfall pipe wasquite shallow and it was difficult to get the YSI 6920sensor submersed in the flow The HI9811 measure-ment was taken from a sample bottle and it is likelythat the HI9811 reading more accurately reflected thetrue conductivity of the outfall

4 Technology transfer

The primary objective of this project was to demon-strate a protocol for small MS4s to detect sample andtrack down the source of illicit discharges to theirstorm sewer systems The MS4s of the WNYSCreceived a series of training workshops (samplecollection and analysis ORI interpretation of track-down results) and opportunities to provide feedbackthroughout the course of the project

In March 2009 a final training workshop was heldfor all MS4s in the WNYSC Based on the feedbackand concerns voiced at the previous training work-shops the sampling protocol and trackdown proce-dures were streamlined to be an efficient cost-effectiveprocess that is compliant with New York Statersquos PhaseII Stormwater regulations

5 Cost comparisons

One of the objectives of this study was to identify cost-effective robust and accurate analytical methods forthe trackdown studies Two comparisons are offeredhere

51 E coli

Erie County Public Health Lab frac14 $1607sample(note that this is a low price for a New YorkState Laboratory)

Coliscan Easy Gel1 - $160 (materials) thorn $333(labour) frac14 $493sample

52 Suite of total phosphorus nitrate potassiumchromium and copper

New York State Health Department certifiedlaboratory frac14 $105sample

Kit reagents ($331) thorn $1333 (labour) thorn $177 (Capital equipment costs) frac14 $1841sam-ple(Capital costs here assume new photo-meters every 500 samples which is veryconservative)

Clearly the analytical approaches examined in thisproject offer a cost-effective alternative to contractlaboratory costs

6 Conclusion

The outcome of this project was two-fold a simpleprocess to detect and eliminate illicit discharges tostorm sewer systems and a reliable cost-effectiveanalytical method to perform the sampling andchemical analysis required to properly identify anillicit discharge

The demonstration component of the projectentailed sampling and analysis for a broad spectrumof parameters to determine a source for the contam-ination andor flow Initially there were twentyanalytical parameters measured for each sampleEach parameter was examined as an indicator tocharacterize the flow as illicit and to identify the sourceof flow (ie sanitary washwater industrial ground-water etc) As the various flow identification andtrackdown analyses were conducted it soon becameevident that a number of the parameters analyzedsimply did not provide a hard indicator for sourceidentification In some instances such as copperdetections at a wide range of concentrations wereprevalent in all samples collected tap water andgroundwater included In other cases the parameterjust did not come into play for source identificationUltimately twelve parameters were selected for theIDDE guidance document temperature pH conduc-tivity E coli ammonia nitrate fluoride totalchlorine potassium detergents phosphorous andturbidity

To achieve the goal of determining cost-effectivechemical analyses procedures the project entailed useof simple kits and colorimetric approaches Theanalytical methods were inexpensive and easy to use(detailed in Section 27) To confirm the reliability andaccuracy of the analytical methods a variety of QAQCassessments were applied and these indicated themethods provided a level of accuracy that wasconducive to their application for illicit dischargedetection and trackdown sampling purposes Theresults were particularly encouraging for the E colianalysis Care should be taken with the photometricanalyses and if extreme values are measured the sampleshould be re-analyzed in triplicate

Following the success of the demonstration projectthe WNYSC utilized grant funds to purchase equip-ment and supplies for outfall sampling and in-housechemical analyses for its entire membership Theseoutfall sampling kits are comprehensive in that every-thing required for an illicit discharge trackdowninvestigation is included

Urban Water Journal 393

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Through its partnership with the WNYSC mem-bership the Erie County Department of Environmentand Planning coordinated the purchase assembly anddelivery of the outfall kits In addition field trainingfor MS4 staff on sampling procedures laboratorytraining for sample analysis results interpretationguidance and trackdown and elimination strategiesare currently underway Erie County Department ofEnvironment and Planning will continue to be aresource for the MS4 project partners for assistancewith interpretation of analytical results trackdownprocedures and source identification

While the two primary objectives of the study wereachieved in reality it is often difficult to progress to thenext level that being effective elimination of the sourceof illicit connection The foremost reason for this is alack of financial resources for many of the MS4sSocio-economic variability among and within the MS4communities is considerable At the municipal levelthis translates into challenges to dedicate manpowerandor equipment at a minimum More costlyconstraints would be the need for large scale infra-structure improvements or re-construction projects toeliminate illicit discharges On a smaller scale similarfinancial limitations may exist for residential propertyowners when corrective action is the responsibility of aproperty owner Simply put medium-high densityresidential areas are common among MS4 commu-nities and often have a large proportion of economic-ally disadvantaged sections within them Stormwaterutilities have been successfully established in somestates to help address financial concerns (eg Brisman2001) but a recent study completed for the WNYSCfound a general resistance to this approach at the MS4and public levels in Western New York (WendelDuchscherer et al 2010) Specifically public consulta-tion suggested that a stormwater utility was perceivedas another layer of government with increased fees andless local control

Finally although this study was site-specific toWestern New York in fact many of the trackdowntechniques discussed herein could be used at locationsthroughout North America and even globally sincefrequently there are common water quality concernswith respect to contaminant type and source Thelaboratory methods used in this study were cost-effective and robust making them particularly attrac-tive for developing countries and we have applied thesemethods successfully in Cambodia Thailand andMalaysia The trackdown approach also is flexibleenough to accommodate local knowledge aboutspecific contaminants of concern and these could beadded as needed to help identify sources of illicitconnections

Acknowledgements

1 Thanks to Mary Perrelli Joseph Drakes RatchadawanNgoen-klan Brent LaSpada Jason Pantano John Karczand Gordon Irvine for their assistance in the field andlaboratory2 Funding for this project was provided by the US EPARegion 2 Project Number CP-97280200ndash0 Mention ofproduct names in this report does not imply endorsementby the US EPA nor does this paper necessarily reflect theviews of the US EPA

References

Alabama Water Watch 1999 Quality Assurance Plan forBacteriological Monitoring Report for the US Envir-onmental Protection Agency Region 4

APHA (American Public Health Association) 1998 Stan-dard Methods for the Examination of Water and Waste-water 20th ed Washington DC

Brisman A 2001 Considerations in establishing a storm-water utility Southern Illinois University Law Journal 26505ndash528

Clarkson HE 2003 Creating a stormwater utility withnear-zero opposition Stormwater The Journal of Sur-face Water Quality Professionals JulyAugust 13pp

Deutsch WG and Busby AL 2000 A Field andLaboratory Manual for a Community-based WaterQuality Monitoring Program in the Philippines Reportfor the Heifer Project International

Eckhardt DAV Reddy JE and Tamulonis KL 2008Ground-water Quality in Western New York 2006 USGeological Survey Open-File Report 2008ndash1140 36 pavailable only online at httppubsusgsgovofr20081140

Gugino H Irvine K Ngern-klun R Sukontason KSukontason K Prangkio C and Thunyapar P 2006Impacts of the urban environment on area source waterthe Klong Mae Kha - Chiang Mai Thailand Proceedingsof the Fourth International Synposium on Southeast AsianWater Environment Asian Institute of Technology Bang-kok pp 85ndash88

He L-M Lu J and ShiW 2007Variability of fecal indicatorbacteria in flowing and ponded waters in southernCalifornia Implications for bacterial TMDL developmentand implementationWater Research 41 3132ndash3140

Irvine KN Somogye EL and Pettibone GW 2002Turbidity suspended solids and bacteria relationships inthe Buffalo River watershed Middle States Geographer35 42ndash51

Irvine KN Vermette SJ Tang T Sampson M andMurphy TP 2004 Partnering to provide water qualityand GIS training in Cambodia Proceedings 25th AsianConference on Remote Sensing Asian Association onRemote Sensing Bangkok

Irvine KN Perrelli MF McCorkhill G and Caruso J2005 Sampling and modeling approaches to assess waterquality impacts of combined sewer overflows ndash theimportance of a watershed perspective Journal of GreatLakes Research 31 105ndash115

Irvine KN Perrelli M Brown N and Rossi MC 2006aSeptic system pollution prevention BMPs Developmentof public outreach approaches and decision making toolsfor local government Clearwaters 36 (3) 51ndash55

394 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Through its partnership with the WNYSC mem-bership the Erie County Department of Environmentand Planning coordinated the purchase assembly anddelivery of the outfall kits In addition field trainingfor MS4 staff on sampling procedures laboratorytraining for sample analysis results interpretationguidance and trackdown and elimination strategiesare currently underway Erie County Department ofEnvironment and Planning will continue to be aresource for the MS4 project partners for assistancewith interpretation of analytical results trackdownprocedures and source identification

While the two primary objectives of the study wereachieved in reality it is often difficult to progress to thenext level that being effective elimination of the sourceof illicit connection The foremost reason for this is alack of financial resources for many of the MS4sSocio-economic variability among and within the MS4communities is considerable At the municipal levelthis translates into challenges to dedicate manpowerandor equipment at a minimum More costlyconstraints would be the need for large scale infra-structure improvements or re-construction projects toeliminate illicit discharges On a smaller scale similarfinancial limitations may exist for residential propertyowners when corrective action is the responsibility of aproperty owner Simply put medium-high densityresidential areas are common among MS4 commu-nities and often have a large proportion of economic-ally disadvantaged sections within them Stormwaterutilities have been successfully established in somestates to help address financial concerns (eg Brisman2001) but a recent study completed for the WNYSCfound a general resistance to this approach at the MS4and public levels in Western New York (WendelDuchscherer et al 2010) Specifically public consulta-tion suggested that a stormwater utility was perceivedas another layer of government with increased fees andless local control

Finally although this study was site-specific toWestern New York in fact many of the trackdowntechniques discussed herein could be used at locationsthroughout North America and even globally sincefrequently there are common water quality concernswith respect to contaminant type and source Thelaboratory methods used in this study were cost-effective and robust making them particularly attrac-tive for developing countries and we have applied thesemethods successfully in Cambodia Thailand andMalaysia The trackdown approach also is flexibleenough to accommodate local knowledge aboutspecific contaminants of concern and these could beadded as needed to help identify sources of illicitconnections

Acknowledgements

1 Thanks to Mary Perrelli Joseph Drakes RatchadawanNgoen-klan Brent LaSpada Jason Pantano John Karczand Gordon Irvine for their assistance in the field andlaboratory2 Funding for this project was provided by the US EPARegion 2 Project Number CP-97280200ndash0 Mention ofproduct names in this report does not imply endorsementby the US EPA nor does this paper necessarily reflect theviews of the US EPA

References

Alabama Water Watch 1999 Quality Assurance Plan forBacteriological Monitoring Report for the US Envir-onmental Protection Agency Region 4

APHA (American Public Health Association) 1998 Stan-dard Methods for the Examination of Water and Waste-water 20th ed Washington DC

Brisman A 2001 Considerations in establishing a storm-water utility Southern Illinois University Law Journal 26505ndash528

Clarkson HE 2003 Creating a stormwater utility withnear-zero opposition Stormwater The Journal of Sur-face Water Quality Professionals JulyAugust 13pp

Deutsch WG and Busby AL 2000 A Field andLaboratory Manual for a Community-based WaterQuality Monitoring Program in the Philippines Reportfor the Heifer Project International

Eckhardt DAV Reddy JE and Tamulonis KL 2008Ground-water Quality in Western New York 2006 USGeological Survey Open-File Report 2008ndash1140 36 pavailable only online at httppubsusgsgovofr20081140

Gugino H Irvine K Ngern-klun R Sukontason KSukontason K Prangkio C and Thunyapar P 2006Impacts of the urban environment on area source waterthe Klong Mae Kha - Chiang Mai Thailand Proceedingsof the Fourth International Synposium on Southeast AsianWater Environment Asian Institute of Technology Bang-kok pp 85ndash88

He L-M Lu J and ShiW 2007Variability of fecal indicatorbacteria in flowing and ponded waters in southernCalifornia Implications for bacterial TMDL developmentand implementationWater Research 41 3132ndash3140

Irvine KN Somogye EL and Pettibone GW 2002Turbidity suspended solids and bacteria relationships inthe Buffalo River watershed Middle States Geographer35 42ndash51

Irvine KN Vermette SJ Tang T Sampson M andMurphy TP 2004 Partnering to provide water qualityand GIS training in Cambodia Proceedings 25th AsianConference on Remote Sensing Asian Association onRemote Sensing Bangkok

Irvine KN Perrelli MF McCorkhill G and Caruso J2005 Sampling and modeling approaches to assess waterquality impacts of combined sewer overflows ndash theimportance of a watershed perspective Journal of GreatLakes Research 31 105ndash115

Irvine KN Perrelli M Brown N and Rossi MC 2006aSeptic system pollution prevention BMPs Developmentof public outreach approaches and decision making toolsfor local government Clearwaters 36 (3) 51ndash55

394 K Irvine et al

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014

Irvine KN Murphy T Sampson M Dany V VermetteSJ and Tang T 2006b An overview of water qualityissues in Cambodia In Intelligent Modeling of UrbanWater Systems Monograph 14 Computational Hydrau-lics International Guelph Ch 2

Irvine KN Rossi MC and Vermette SJ 2009Demonstration of Illicit Connection Trackdown andReceiving Water Impact Evaluation for MS4s Report toUS EPA Region 2 Project Number CPndash97280200ndash0

Kay D Edwards AC Derrier RC Francis C Kay CRushby L Watkins J McDonald AT Wyer MCrowther J and Wilkinson J 2007 Catchmentmicrobial dynamics The emergence of a research agendaProgress in Physical Geography 31 (1) 59ndash76

Krueger AM Irvine KN Prangkio C Chaokasad KSukontason K Sukontason KL Ngern-klun R 2004Visualizing water quality trends in Chiang Mai ricepaddies Possible links between environment and healthrisks Middle States Geographer 37 1ndash8

New York State Department of Environmental Conserva-tion 1999 Part 703 Surface Water and GroundwaterQuality Standards and Groundwater Effluent Limita-tions 6 NYCRR Chapter X Article 2 Part 703

New York State Department of Environmental Conserva-tion 2005 The 2002 Niagara RiverLake Erie BasinWaterbody Inventory and Priority Waterbodies ListNYSDEC Division of Water Report

Pitt R 2004 Illicit Discharge Detection and Elimination AGuidance Manual for Program Development and Techni-cal Assessments Center for Watershed Protection MD

Pitt R and Lalor M 1993 A Userrsquos Guide for theAssessment of Non-Stormwater Dischargers Into SeparateStorm Drainage Systems EPA600-Rndash92ndash238

Pomeroy C Cave K and Tuomari D 1996 TechnicalMemorandum Summary of Illicit Connection DetectionPrograms in Michigan Rouge River National WetWeather Demonstration Project

Prince Georgersquos County Maryland 1999 Low-Impact Deve-lopment Design Strategies An Integrated Design Ap-proach Department of Environmental Resources Report

Rossi MC Irvine KN Hersey TR Vermette SJ 2009Western New York Stormwater Coalition Compliancethrough Collaboration In Intelligent Modeling of UrbanWater Systems Monograph 17 Computational Hydrau-lics International Guelph Ch 7

Stark H 2003 Cuyahoga County Board of Health Storm-water Program Stormwater The Journal for SurfaceWater Quality Professionals SeptemberOctober 12 pp

Tucker S and Harrison D 2006 Background and Introduc-tion In Guidance for Municipal Stormwater FundingReport to the National Association of Flood and Storm-water Management Agencies for the US EPA ch 1

University of Vermont 2003 Watershed Alliance WatershedEducatorrsquos Handbook

US EPA 1983Methods for Chemical Analysis of Water andWastes EPA6004ndash79ndash020

US EPA 1991 Methods for Chemical Analysis of Waterand Wastes Update EPA6004ndash91ndash010

US EPA 1992 Methods for Chemical Analysis of Waterand Wastes Update EPA600Rndash92ndash129

US EPA 1993Methods for Chemical Analysis of Water andWastes Update EPA600R-93ndash100

US EPA 1999 Report to Congress on the Phase II StormWater Regulations US EPA Report EPA 833-R-99ndash001

US EPA 2000 Low Impact Development (LID) ALiterature Review EPA Report EPA-841-B-00ndash005

US EPA 2001 Protocol for Developing Pathogen TMDLsUS EPA Report EPA 841-R-00ndash002

US EPA 2009 EPArsquos BEACH Report 2008 SwimmingSeason US EPA Report EPA 823-F-09ndash005

Virginia Citizen Water Quality Monitoring Program 2003Methods Manual Report for Virginia Department ofEnvironmental Quality

Visoth T Yim M Vathna S Irvine K and KoottatepT 2010 Efficiency of Phnom Penhrsquos natural wetlands intreating wastewater discharges Asian Journal of WaterEnvironment and Pollution 7 (3) 39ndash48

Wendel Duchscherer Malcolm Pirnie Advanced DesignGroup Bond Schoeneck and King 2010 Feasibility of aRegional Stormwater Utility District in Erie and NiagaraCounties Report to the Western New York StormwaterCoalition

Wills M and Irvine KN 1996 Application of the NationalSanitation Foundation Water Quality Index to theCazenovia Creek pilot watershed management studyMiddle States Geographer 29 105ndash113

Urban Water Journal 395

Dow

nloa

ded

by [

209

912

211

52]

at 1

100

28

Aug

ust 2

014