1

Benefit-costanalysisofstormwatergreeninfrastructureforGrandRapids,MichiganErikE.Nordman,GrandValleyStateUniversity

ElaineIsely,WestMichiganEnvironmentalActionCouncil

PaulIsely,GrandValleyStateUniversity

RodDenning,GrandValleyStateUniversity

AbstractGrandRapids,Michigan,USAisamedium-sizedcitylocatedwithintheLakeMichiganwatershed,onethefiveNorthAmericanGreatLakes.Likemanycities,GrandRapidsspendsconsiderablemoneymanagingstormwater.Impervioussurfacescollectandconcentratevolumesofwaterandassociatedsedimentsandpollutants.Thiscreatesflooding,erosion,andpollutionproblemsespeciallyfordownstreamcommunities.However,stormwaterquantitycanbereducedandqualitycanbeimprovedby,forexample,mimickingnaturalhydrology,enhancingbiodiversity,linkingecologicalandeconomicsustainability,takinganintegratedapproachatmanageablescales,andviewingstormwaterasaresource.Evidenceismountingthatonsitestormwatermanagementsystemscanbecost-effective,butthedetailedbenefit-costanalysesarestilllacking.ThereforetheWestMichiganEnvironmentalActionCouncil,togetherwithresearchersfromGrandValleyStateUniversity,estimatedtheeconomicbenefitsandcostsofvarious“greeninfrastructure”(GI)practices.EachGIpracticewasstandardizedtotreat3,000ft3ofstormwaterper1.0-incheventplusthefirstinchofstormwaterfromlargerevents.Thisequatestoabout113,000ft3ofstormwaterperyear.Theeconomicanalysisusedabenefittransferapproachtoestimatethenetpresentvalue(NPV)ofcapital,operations,andmaintenancecostsaswellasthedirectandindirectbenefits.ThesuiteofbenefitsvariedforeachGIpracticeandincludedfloodriskreduction;reductionsinstormwatervolume,phosphorus,totalsuspendedsolids(TSS),andairpollution;scenicamenityvalue;andCO2storage.A3.5percentdiscountratewasappliedtoallcostsandbenefits,andeachpracticewasanalyzedover50years.Conservednaturalareashadthelargestnetpresentvalueat$3.10/ft3,followedbystreettreeplantersat$1.48/ft3,raingardensat$1.12/ft3,porousasphaltat$0.68/ft3,andinfiltrationbioretentionbasinsat$0.03/ft3.Greenroofshadanegativenetpresentvaluesof$-1.12/ft3suggestingtheirlifetimecostsexceedtheirbenefits,atleastinGrandRapidswhereground-levelopenspaceisplentiful.IfthegreenroofisusedtoattaincertificationsuchasLeadershipinEnergyandEnvironmentalDesign(LEED),whichhasahighamenityvalue,thenthenetbenefitsturnpositive($0.16/ft3).Rainbarrelsareanothersmall-scalegreeninfrastructurepracticethatcanbeusefulandcost-effectiveatthehouseholdscale($1.06/ft3).However,thereisalotofvariabilityinthecostsandbenefitsassociatedwitheachoftheseGIpractices,whichwillaffectthenetpresentvalue;weutilizedlikelyvaluesfortheregion.NooneGIpracticeisappropriateforallsituations.RatherthechoiceofGIpracticewillbedrivenbythesiteandbudget.Thisbenefit-costanalysisofGIpracticeshaspolicyimplicationsforGrandRapidsandothersmalltomid-sizeMidwesterncities.Withthearrayofoptionsavailabletomanagestormwateronsite,municipalitieslikeGrandRapidsarewell-positionedtoadopttheGIpracticesthataremostappropriate.

2

IntroductionLocalgovernmentunits,includingvillages,cities,andcounties,expendsignificantresourcestomanagestormwater.TheCityofGrandRapids,Michigan,USA,forexample,operatesstormwaterinfrastructurevaluedat$533million(CityofGrandRapids,2014).Thesegovernmententitieshaveastrongincentivetoreduceexpendituresbyreducingthevolumeofstormwatertheymanage.Reducingrunoffvolumesalsoreducesboththeriskoffloodsandtheamountofpollutionenteringthewatercourses.

Thedominantparadigminstormwatermanagementformostofthe20thcenturywastomovethewateroffsiteasquicklyaspossiblethroughditchesandpipes,so-called“grayinfrastructure”,andintotheneareststreamorriver.Whileeffectiveatpreventingponding,movinglargequantitiesofwater,withitssedimentsandpollution,intowaterbodiesresultedinflooding,erosion,andpollutionproblemsfordownstreamcommunities.Sincethe1990sanecologicalparadigmhasemergedthatplacesstormwaterquantityandqualitywithinthecontextofintegratedwatershedmanagementandlowimpactdevelopment.Stormwaterquantitycanbereducedandqualitycanbeimprovedby,forexample,mimickingnaturalhydrology,enhancingbiodiversity,linkingecologicalandeconomicsustainability,takinganintegratedapproachatmanageablescales,andviewingstormwaterasaresource(DeboandReese,2002).

Thegrayinfrastructureparadigmemphasizespublicinfrastructurebuilt,maintained,andoperatedbythemunicipality.Stormwaterinfrastructureisapurepublicgood;thatis,itisnon-rivalandnon-exclusive,whichmeanthatundernormalcircumstances,everyonecanbenefitfromitwithout“usingitup”andonceitisbuilt,themunicipalitycannotexcludeanyonefromenjoyingitsbenefits,respectively(WeimerandVining,2010).Thereislittleincentiveforprivatelandownerstoinvestinstormwatermanagementpracticesbecausethebenefitsoftheiractionswouldlargelyaccruetotheirneighbors.Theecologicalparadigmbasedononsitemanagementandlowimpactdevelopment,however,requiressignificantinvestmentsonprivatepropertysuchasraingardens,green(vegetated)roofs,andpermeablepavement.Themisalignmentofincentivesresultsinamarketfailure.Intheabsenceofpolicy,actorsinthemarketplacewillunderprovideonsitestormwatermanagementsystemsandpractices.Thiswillbethecaseevenifonsitemanagementislessexpensivethanthetraditionalsewerinfrastructure.It’snotjustaboutthecosts;it’saboutwhopaysthem.

Evidenceismountingthatonsitestormwatermanagementsystemscanbecost-effective.TheCenterforNeighborhoodTechnologyfoundthatamunicipallevelgreeninfrastructureplancouldhavesignificantnetbenefitsforthecommunitybyreducinggrayinfrastructurecapitalcostsby$120millionandprovidingmorethan$4millioninenergy,airquality,andclimatebenefitsannually(CenterforNeighborhoodTechnology,2014).Ifthenetbenefitsofgreeninfrastructurearepositive,thereisacompellingcasethatmunicipalitiescouldsavemoneyandprovidebetterenvironmentaloutcomesbyprovidingincentivesforprivateinvestmentinonsitestormwatermanagementthroughgreeninfrastructure.

ThispaperanalyzesthebenefitsandcostsofstormwatermanagementusinggreenandgrayinfrastructureintheCityofGrandRapids,Michigan,USA.Specifically,itaddressessevengreeninfrastructurepractices:porousasphalt;greenroofs;raingardens;bioretentioninfiltrationponds;conservationofnaturalareas;streettrees,whichincludetreeplantersandtreepits;andrainbarrels.Thisbenefit-costanalysisispartoftheRainwaterRewardsprojectwhichincludesaweb-basedstormwatervaluecalculatorthatestimatesthebaselinestormwaterrunoffquantity,thereducedrunoff

3

quantityaftertheadoptionofgreeninfrastructuresystems,andtheneteconomicbenefitofthosesystems(http://www.RainwaterRewards.com).TheRainwaterRewardscalculatorisanaccessibletoolforcitizens,landowners,andpolicymakerstocalculatethepublicbenefitsofgreeninfrastructureandcraftpolicyinstruments,suchasrefundsortaxcredits,toencourageprivateinvestmentingreeninfrastructure.

Theresearchteambeganworkingonvaluationofecosystemservicesassociatedwithdifferenttypesofgreeninfrastructuresince2005.TheINtegratedValuationofEcosystemServicesTool(INVEST)wasdevelopedtoeducatecommunityplannersandlandownersaboutthevalueofecosystemservicesassociatedwithnon-urbanlandusesinWestMichigan.However,itwasdifficulttotranslateregionalvaluesforuseinparcel-baseddecisionmaking(Iselyetal.,2010a)In2010,INVESTwasexpandedandappliedtoasingleparceltohelpresolvealandusedisputebetweenthepropertyownerandBlueLakeTownship.Acalculatortemplatewasputtogethertodemonstratetheecosystemservicesassociatedwiththatparcel(Iselyetal.,2012).

Theteam’sworkonquantifyingthecostsandbenefitsofstormwatermanagementpractices,specifically,beganwithanintegratedassessmentprojectintheSpringLakeWatershedin2007.Theteamcalculateddirect,indirect,andopportunitycostsandbenefits,andperformedcosteffectivenessandcost-benefitanalysesofbioretention/raingardens,vegetatedbio-swales,perviouspavement,constructedwetlands,andstormwaterretrofits(Iselyetal.,2010b).In2013,teammemberscompletedareviewofbestpracticesinincentivizingtheimplementationofstormwatergreeninfrastructure(Isely,2014).ThenewRainwaterRewardscalculatorhasupdatedcostandbenefitinformationforstormwatergreeninfrastructurepracticesmostlikelytobefoundinsmall-tomediumurbancentersintheGreatLakesbasin–GrandRapidsandMuskegon,Michigan.TheRainwaterRewardscalculatorwillbethecenterpieceofacommunityengagementcurriculumonstormwatermanagementthroughgreeninfrastructure.

Whatwecallgreeninfrastructureintheremainderofthispapergoesbymanynames:lowimpactdevelopment(LID),stormwaterbestmanagementpractices(BMPs),stormwatermanagementpractices(SMPs),andothers.Whiletheirdefinitionsmaydifferslightly,theyallrefertodecentralizedpracticesthatreducethequantityofstormwaterenteringwatercourses.Forthesakeofconsistency,wewillsimplyrefertoallofthesepracticesasgreeninfrastructure(GI).

LiteraturereviewThemostcomprehensiveandaccessibleresourcetodateistheGreenValuesStormwaterToolboxCalculatorfromtheCenterforNeighborhoodTechnology(CNT)(CenterforNeighborhoodTechnology,2007).TheCNTcalculatorusedarelativelysimplewebinterfacethatallowsuserstoenterlot-specificinformation.Itcalculatedthestormwaterrunoffvolumeundertypicalcircumstancesandestimatesthereductionthroughtheuseofgreeninfrastructure.Costsestimatesconsideredbothconstructionandoperationandmaintenancecosts.Thecalculatorestimatedthefollowingbenefits:reducedairpollutants,carbondioxide,compensatoryvalueoftrees,groundwaterreplenishment,reducedenergyuse,andreducedtreatmentbenefits.NoteveryGIpractice,however,deliverseachofthesebenefits.CNTcurrentlyoffersthreeversionsofthecalculator:theoriginal,oneforChicago,andanationalcalculator.

4

BeauchampandAdamowski(2012)usedtheCNTcalculatorandothervaluationtoolstoestimatethevalueofGIcomparedtoconventionalinfrastructure.GIdevelopmentincludedreducedpavementdesigns,separatepotableandnon-potablewatersystems,greywaterandblackwatersewagesystems,andstormwatermanagementusingbioswales,wetlands,greenroofs,andraingardens.TheplanneddevelopmentintheMontrealsuburbofVaudreuil-DorionbasedonGIwouldcost11-29percentmorethanaconventionaldesign.Housingvalues,however,areexpectedtoincreaseby15-27percentwhichwouldoffsettheinitialcostgap.

TheWaterEnvironmentResearchFoundation(WERF)developedasuiteofspreadsheet-basedbestmanagementpracticeandlowimpactdevelopmentwholelifecostmodels(MoellerandPomeroy,2009).Thecosttoolincludesninedifferentpractices,includingpermeablepavements,greenroofs,raingardens,andin-curbtreeplantervaults.Thecostmodelsallowpractitionerstoestimatethecapital,operations,andmaintenancecostsforeachpracticeandcomparethecost-effectivenessofeach.Thedefaultspreadsheetispopulatedwithstandardvaluesbutallowstheusertoinputlocally-appropriateinformationaboutprojectcosts,timelines,wages,anddiscountrates.

AlocallyimportantcostanalysiswasthatofClarketal.(2008)whoassessedthenetpresentvalueofgreenroofscomparedtothoseofconventionalroofs.ThestudysitewastheUniversityofMichigancampusinAnnArbor.Themeancostofaconventionalflatroofwas$167/m2in2008($17.14/ft2in2015).Themeancapitalcostofagreenroof(includingtheconventionalroofunderneath)was39percenthigherthantheconventionalroofalone.Theresearcherstalliedthebenefitsofgreenroofs,includingstormwaterfeereductionswherethecityimplementsastormwaterchargebasedonimpervioussurfaces),energysavings,airpollutionreduction,andalongerlifespanfortheconventionalroof.Theamenityvalueofgreenroofswasnotincluded,norweretheoperationandmaintenancecostsforgreenorconventionalroofs.Ratherthanusingastandardrealdiscountrateintheireconomicanalysis,Clarketal.multipliedtheannualbenefitsandcostsbyathreepercentinflationrateandthendividedbyanominalfivepercentinterestrate.Thisresultsinaneffectivediscountrateoflessthantwopercent.Thenetpresentvalueanalysisshowedthat,overthelifeoftheroof,greenroofscost25-40percentlessthanconventionalroofs.Energysavingsandpollutionreductionbenefitsweregreaterthantheavoidedstormwaterfees.Despitethehighercapitalcosts,thelifetimebenefitsoutweighedthegreenroof’shighercapitalcosts.Howeverthisfindingisverylikelyduetotheuseofaloweffectivediscountrateoflessthantwopercent.

BianchiniandHewage's(2012)alsoreportedapositivenetbenefitforgreenroofs.Theirprobabilisticassessmentofgreenroofcostsandbenefitsfoundthemostlikelyscenarioproducedanetbenefitof$37/ft2.LikeClarketal.(2008)theireconomicanalysisincludedbotha1-4percentinflationrateanda2-8percentdiscountrate.Theresultswerehighlysensitivetothechoiceofinflationanddiscountrates.Otherresearchershavefoundnegativenetbenefitsforgreenroofs.Forexample,CarterandKeeler(2008)foundthatthepresentvaluecostsofagreenroofinGeorgiawas10-14percenthigherthanthatofaconventionalroof.Theyusedafourpercentrealdiscountrateintheirbenefit-costanalysis.Likewise,Sprouletal.(2014)foundthatgreenroofshaveahighernetcostovertheirlifetime.Sprouletal.usedathreepercentrealdiscountrate.Neitherofthesestudies,however,includedamenityvaluesforgreenroofsintheiranalyses.Allofthesestudiessuggestthatagreenroof’seconomicefficiencyishighlysensitivetothechoiceofdiscountrate.LowdiscountratestendtoresultinpositiveNPVswhilehigherdiscountratesofthreepercentorhighertendtoresultinnegativeNPVs.

5

ResearchersattheUniversityofNewHampshire’sStormwaterCenterassessedthecostandperformanceofseverallowimpactdevelopmentpracticesincludingporousasphalt.Theyfoundthat,contrarytoconventionalwisdom,porousasphalthadthelowestmaintenanceburdenintermsofstaffhoursandthesecondlowestinannualcosts.Porousasphaltalsoperformedwellinremovingbothtotalsuspendedsolidsandphosphorus(Houleetal.,2013).

TheForestServiceanalyzedthecostsandbenefitsofstreettreesinMidwesterncities.Theyfoundthat,forpublicstreettrees,thebenefitsoutweighthecostsoveraforty-yearperiod.Forsmalltreeslikeacrabapplethenetbenefitwas$160(in2005),whileformediumandlargetreesthebenefitswere$640and$2,320,respectively.TheForestServiceanalysisdidnot,however,usediscountingwhenassessingthesebenefits.Streettreesprovideheatingandcoolingenergysavings,increasepropertyvalues,reducestormwatervolumesbyinterceptingrainfall,andreduceairpollution(McPhersonetal.,2006).

Economistsusepropertyvaluemodels,alsoknownashedonicmodels,toestimatetheeffectofhousingattributesonsalesprices.ThehousingattributescanincludeenvironmentalvariablesandseveralstudieshavefocusedonGI.Hellman(2011)usedahedonicmodeltostudytheeffectofstormwatervolumesonhousingpricesintheRochester,NewYorksuburbofBrighton.Hellmanfoundadditionalstormwaterquantitiesnegativelyaffectpropertyvaluesandthatthemarginalabatementcostsarelessthanthemarginaldamage.Holdingallotherattributesconstant,aonepercentincreaseinstormwatervolumeleadstoapproximatelyaonepercentdecreaseinahome’sassessedvalue.

AnotherhedonicmodelinvestigatedtheeffectofgreenroofsonapartmentrentsinNewYorkCity(IchiharaandCohen,2010).Thepresenceofagreenroofadded16percenttotherentalprice.Thoughthegreenroofvariablewasstatisticallysignificant,thenumberofobservations(44)wasrelativelysmallandthefindingsshouldbeviewedwithcaution.Thestudysitewasaheavilyurbanizedareawheregreenspaceisscarce.Inthecontextofhighwealthandscarceopenspace,residentsmaybewillingtopayahighpremiumforagreenroof.AhedonicanalysisfromTaiwan,however,foundtheopposite–thatgreenroofs(aswellasotherGIpracticeslikeporouspavementandabalconygarden)haveanegativeeffectonresidentialpropertyprices.Theauthorsassumedthiswasduetoperceptionsofhighermaintenancecosts(Chenetal.,2014).Asgreenroofsbecomemorecommonandstarttofeatureinthepropertymarketthereshouldbemoredefinitivestudiesontheirpropertyvalueeffects.

GreeninfrastructurepracticescanhelpabuildingearnacertificationsuchasEnergyStarorLeadershipinEnergyandEnvironmentalDesign(LEED).Oneanalysisofcertifiedcommercialbuildingsfoundthatsuchcertificationscommandrentpremiumsof3.1percentforEnergyStarand7.0percentforLEED.LEEDbuildingswerealsofoundtoreduceoperatingcostsbyabout5.4percentperyear.Nodecreaseinoperatingcosts,however,wereobservedforEnergyStarcertifiedbuildings(Reichardt,2013).

BarnhillandSmardon(2012)facilitatedafocusgrouparoundGIinSyracuse,NewYork,USA.Theyfoundthreemajorbarrierscurrentlylimitgreeninfrastructureimplementation.Firstisthehomeownerfinancialcost.Thecostsof,forexample,aresidentialraingardenarebornebythehomeownerwhilethestormwaterabatementbenefitsaccruetothecommunityatlarge,especiallydownstreampropertyowners–aclassicmarketfailure.ThesecondbarrierisalackofknowledgeaboutGIbenefits,maintenanceissuesincludingcosts,andtheuselocally-appropriatepractices.Thethirdbarrierisafailuretoproperlyframetheissue.FramingGIintermsofneighborhoodregenerationandsustainabilitycanleadtomoreeffectiveengagement.EngaginglocalstakeholdersindevelopingGIcanimprovesocialequity.

6

BenefittransferThedemandforenvironmentalvaluationinformationhasoutpacedresearchandfundingforvaluationprojects.Consequently,manyprojectsmakeupforthelackofdatabyusingbenefittransfer.Freeman(2003,p.453)definesbenefittransferas“thepracticeofapplyingnonmarketvaluesobtainedfromprimarystudiesofresourceorenvironmentalchangesundertakenelsewheretotheevaluationofaproposedorobservedchangethatisofinteresttotheanalyst.”Thelocationpresentlyunderinvestigationiscommonlycalledthe“policysite”andthelocationfromwhichthevaluesaredrawnisthe“studysite.”

Thepolicyandstudysitesmaydifferforavarietyofreasons,suchasdifferencesinincomeorpreferencesamongthepopulationsatthesides(demandsidefactors)orvariationintheenvironmentalattributesbeingvalued(supplysidefactors).Thebenefittransferprocessadjuststhestudysitevaluestoreflectthesedifferences.Benefittransferissimplerandmoreaccurateifthepolicyandstudysitesarerelativelysimilar(Freeman,2003).

Johnstonetal.(2015)reviewedthegenerallyacceptedmethodsofbenefittransfer.Theydescribedseveraltypesofbenefittransfertechniques:unitvaluetransferandbenefitfunctiontransfer,thelatterofwhichincludesstructuralbenefittransferandmeta-analysis.Unitvaluetransfer,thoughthesimplest,hasseveraldrawbackswhichmakeitlessdesirableforpolicyapplications.Unitvaluetransferappliesasingle,unadjustedwillingness-to-pay(WTP)valuefromthestudysitetothepolicysite.Thismaybeappropriateifthestudyandpolicysitesarenearlyidentical.Inmostcases,however,unitvaluetransfersresultinunacceptablyhigherrorsandareusuallynotrecommended.

RatherthansimplytransferringtheWTPnumberfromstudytopolicysites,benefitfunctiontransferappliesthemathematicalfunction,includingallorasubsetofvariables,tothepolicysite.Applyingthefunctionallowstheresearcherstoadjustfordifferencesbetweenthesitesandreduceerrors.Theadjustmentalsoallowsawiderrangeofcontextstoserveasstudysites.Thisisimportantwherefew,orno,studysitesaresufficientlysimilartothepolicysite.

Instructuralbenefittransfer,alsoknownaspreferencecalibration,theresearcherdefinestheutilityorpreferencefunctionthatdescribesanindividual’schoicesoverarangeofmarketandnon-marketgoods.Onestudysiteistypicallyusedasthesourceofthepreferencefunction.Thevariablesusedinthepreferencefunctionthatweredevelopedatthestudysitearemeasuredatthepolicysiteandanempiricalrelationshipisestablished.Thisapproachisconsistentwiththebudget-constrainedutility-maximizationfoundationsofstandardeconomictheory.Thedrawback,however,isitscomplexityandtheexpertiseinmathematicaleconomicsrequiredtoemploythetechnique(Johnstonetal.,2015).

Meta-analysisisanalternativeformofbenefitfunctiontransfer.Meta-analysisis“thequantitativesynthesisofevidenceonaparticularoutcome,withevidencegatheredfrompriorprimarystudies”(Johnstonetal.,2015,p.26).Thequantitativesynthesisismostoftenaccomplishedusingameta-regressionmodelinwhichthedependentvariableisthatoftheprimarystudies,e.g.housingprice,fecalcoliformcount.Theindependentvariablesareobservablefactorsthatinfluencethedependentvariableatthevariousstudysites.Thesecaninclude,forexample,economic,demographic,andresourcecharacteristicsofthestudyandpolicysites.Thoughmeta-analysis,andespeciallymeta-regressionmodels,canimprovetheaccuracyofbenefittransferscomparedtounitvaluetransfers,meta-analysisisnotasrigorousoraccurateasstructuralbenefittransfer/preferencecalibration.Meta-analysesaremost

7

appropriatewhenasubstantialamountofstudiesonthattopichavebeenpublished,thereisnosinglestudythatcloselymatchesthepolicysite,andthereisadesiretoestimatebenefitsunderdifferentpolicycontexts(Johnstonetal.,2015).

Johnstonetal.(2015)presentedaten-stepprocedureforconductingabenefittransfer.Firstistodefinethebenefittransfercontextsuchasthecircumstances,theenvironmentalresources,thecurrentandproposedpolicies,andtheusesforthevalueestimates.Secondistoestablishtheneedforbenefittransfer.Ifsufficientresourcesareavailable,primaryvaluationstudiesarepreferableoverbenefittransfers.Thirdistodefinethepolicy,environmentalgoods,andpopulationofthepolicysite.Fourthistodefineandquantifythepolicyoptionsandchangesintheenvironmentalgoods.Thisincludesdeterminingthebaselinelevelsandmarginalchangesintheprovisionofenvironmentalgoodsandservices.Fifthistogatherandevaluatevaluationdataandevidencethroughacomprehensivereviewofliterature,bothpeer-reviewedandso-called“grayliterature.”Theresearchersmustscreenthedocumentsforqualityandrelevancetothepolicysite.Thesixthstepistodeterminethemethodofbenefittransfer.Asdiscussedabove,benefitfunctiontransfers,suchasmeta-analysisandpreferencecalibration,aregenerallypreferredoverunitvaluetransfers.Seventhistodesignandimplementthebenefittransfer.Eightistoaggregatethevaluesoverpopulations,areasandtime.Thebenefittransferresultsinaper-unitvalue.Theper-unitvaluemustbeaggregatedacrossthepolicysite.Theninthstepistoconductasensitivityanalysisandtotestreliability.Thismayincludeconductingthetransferusingarangeofdiscountratesorchangingthefunctionalform.Cross-validationmethodscanbeusedtotesttheperformanceofmeta-regressionmodels.Thetenthandfinalstepistoreporttheresults.Thisten-stepprocesswasusedtoestimatethebenefitsandcostsofGIinGrandRapids,Michigan.

MethodsRunoffestimationTheNewYorkStateDepartmentofEnvironmentalQualitycreatedtheConstructionStormwaterToolboxtoassistownersandoperatorswithcompliancewithplanningrequirementsundertheNewYorkStatePollutantDischargeEliminationSystem(SPDES).TheToolboxincludesadesignmanualandasetofExcel-basedrunoffreductionworksheets(NYSDept.ofEnvironmentalConservation,2014).TheworksheetsarerigorousenoughforSPDEScompliance,yetflexibleenoughtobeadoptedinmanycircumstances.MuchofupstateNewYorklieswithintheGreatLakesbasinandhasaclimatesimilartothatofMichigan’sLowerPeninsula.Aftercarefulreview,theprojectteamdeemedtheNewYorkStaterunoffreductionworksheetssuitableforuseinMichigan.TherunoffreductionworksheetswereusedtoestablishbaselinerunoffvolumesandtocalculatetherunoffreducedbyimplementingparticularGIsystems.

Theproject’sunitofanalysiswasthe2010censusblock.Censusblockswerechosenbecausetheyarewell-established,publiclyavailable,andaresmallenoughforfinescaleanalysis.Individualparcelswerenotusedbecausetheprojectteamdidnotwanttogiveindividuallandownerstheideathattheywouldbecompensatedfortheestimatedmarketandnon-marketbenefitsofgreeninfrastructureontheirproperties.Thecensusblockprovidestheminimumlevelofaggregationnecessarywhileenablingfine-scaleanalysis.

TheToolbox,aswellasotherstudies(e.g.Houleetal.,2013),usethe90thpercentile24-hourraineventasthedesigncriterionforstormwatermanagement.InMichigan,the90thpercentilerangesfrom0.8

8

inchesto1.0inches(KuhnsandUlasir,2015).Wechosetousetheupperbound(1.0inches)asthedesigncriterion.WeassumedthattheGIpracticeswouldpreventallstormwaterrunoffforraineventsuptoandincluding1.0inches.Tenyears(2006-2015)ofrainfalldatafromtheGeraldR.FordAirportinGrandRapidswereanalyzed(WeatherUnderground2016).Theten-yearaverageannualrainfallinGrandRapidswas40.0inchesandrangedfrom32.4(2007)to48.8(2008).Thesumofrainfalleventsuptoandincluding1.0inchesaswellasthefirstinchofeventsgreaterthan1.0inchesaveraged37.75inchesperyear.

EconomicvaluationTheinstallation,maintenance,andopportunitycostsoftheGIpracticeswillbecomparedtothebenefitsofavoidedstormwaterrunoffcosts,pollutionreduction,andaestheticenhancement.Thesecostsandbenefitswillbeapportionedovertheexpectedlifeofthesystemandanalyzedusingnetpresentvalue:

𝐵"(1 + 𝑟)"

−𝐶"

(1 + 𝑟)"

*

"+,

Wheregreeninfrastructureiscomparedto“grayinfrastructure,”thenetcostofgreeninfrastructurewascalculatedby:

𝐶"(1 + 𝑟)"

=𝐶"./00*

(1 + 𝑟)"−

𝐶"./12

(1 + 𝑟)"

Thedirectbenefitsincludereducedmaintenancefromavoidedstormwaterandreducedenvironmentalandhealthcostsrelatedtowaterpollution.Greeninfrastructurealsohasindirectbenefits.Streettrees,raingardens,andgreenroofsenhanceaneighborhood’saestheticqualityandmaybemeasuredthroughhomeprices.Streettreesalsoremoveairpollutionandreduceenergycostsbyshadingbuildings.Agreenroofmayalsoextendthelifeoftheconventionalroofunderneathandprovidesenergy-savinginsulation.

Wherepossible,thevalueestimatesweretakenfromprojectsinGrandRapidsandadjustedforinflation.Inothercases,thevaluesreportedinpeer-reviewedandgrayliteraturefromotherlocationswereused.ThesevalueswereadjustedtothepresentGrandRapidscontextusingbenefittransfermethods.

ValueofavoidedrunoffTheprojectassessedthecostsandbenefitsofstormwatermanagementthroughgrayandgreeninfrastructure.Thebenefitstransferapproachwasusedtomodifycostandbenefitvaluesfromdifferenttimesandlocations.Costsforbothtypesofsystemswerecatalogedthroughliteraturereviewandconversationswithlocalgovernmentsandserviceproviders.

Thedirectcostofstormwatermanagementwasestimatedfromgovernmentdocuments.TheCityofGrandRapidscompletedasustainabilityplanwhichincludedaStormwaterAssetManagementandCapitalImprovementPlan(CityofGrandRapids,2014).Theprojectedannualcosttoprovidetheexisting“levelofservice”forstormwatermanagement,includingbothfixedandvariablecosts,was$3.60millionin2014.Stormwaterreductionpractices,however,onlyreducevariablecosts.Fixedcosts,suchassystemrenewalorend-of-lifereplacement,inspections,andregulatorycosts,werenotincluded.Thereforeonlytheannualvariablecostsofcorrectiveandpreventativemaintenancewereusedto

9

estimatethevalueofavoidedrunoff.Thecityrecognizedthattheexistinglevelofserviceisinadequate.Aftercarefulreviewofthreeadditional“levelofservice”scenarioswhichprovideincreasinglevelsofannualspendingrequirementsforbasicstormwatermanagementservices,thecityrecommendedpursuingLevelofServiceCwhichwouldincreasetheannualbudgetforstormwatermanagementto$10.38million.Thislevelofservicefocusesonmaintainingcriticalinfrastructureandhighpriorityareas.ThebudgetforinspectionsofcatchbasinsanddetentionbasinsinLevelofServiceCincludes$639,000and$6,500,respectively.Unliketheinspectiondefinitionforotherassets,inspectionofcatchanddetentionbasinsincludescleaning.Cleaningactivitieswerelistedundermaintenancefortheexistinglevelofservice.Thereforecatchbasinanddetentionbasininspectionswereincludedinthevariablecosts(Table1).Thecity’sreportused2014dollars.Afteradjustingforinflationto2015dollarsusingtheConsumerPriceIndex(CPI),thetotalannualmaintenancecostis$2,898,804.

Table1:AnnualmaintenancecostsforstormwatermanagementunderLevelofServiceC(lowerestimate).

Asset/Activity Annualmaintenancecost(2014dollars)

Gravitymains $946,000Manholes $40,000Laterals $73,000Catchbasins $677,000Culverts $43,000Openchannels $3,000Dischargepoints $67,200Detentionbasins $6,500Streetsweeping $1,020,000Total $2,875,700Totalin2015dollars $2,898,804

Thestormwatermanagementsystemprocessesrunofffromthecity’simpervioussurfaces.AfeatureextractionprocessusingLandsatimagerywithagroundsampledistanceof30mx30mfound12,671acresofimpervioussurfaceinthecity(Xianetal.,2011).Thatis44%oftheentirecityarea.Attheaverage40inchesofannualrainfall,eachacregenerates137,940ft3/yearofrunoff,or1,747,837,740ft3/yearforthewholecity.UnderLevelofServiceC,theannualmaintenancecostperunitofstormwatertreatedis$0.0017/ft3/year.

Table2showsthepresentvalueof50yearsofavoidedstormwateratadiscountrateof3.5%.

Table2:Valueofavoidedstormwater(2015$/ft3)

LevelofService Unitcostofavoidedstormwater

Presentvaluecostofavoidedstormwater(50years)

Current $0.00090/ft3/year $0.023/ft3C $0.00017/ft3/year $0.040/ft3A $0.00444/ft3/year $0.108/ft3

10

StandardizingthegreeninfrastructurepracticeEachGIpracticewasstandardizedbasedonawater-qualityvolume(WQv)reductionof3,000ft3fora1.0inchraineventusingtheNYSStormwaterConstructionToolbox.ThetoolcalculatesthesizeoftheGIpracticeneededbasedonthearea’srainfallregime,totalarea,andimperviousarea.Asphaltandbuildingroofsaretypically100%impervioussurface.Raingardensareassumedtobeplacedinresidentialareas.AnalysisofLandsatimageryshowedthatresidentialareasinGrandRapidshaveanaverageof41percentimpervioussurface.Thecorrespondingareasthatproduce3,000ft3WQv(actually3,002ft3)isatotalareaof2.0acresandanimperviousareaof0.8acres.Residentialareas,onaverage,have5.8residencesperacreorjustover11for2.0acres(0.2acresperresidence).Ifeachofthe11houseshadaraingarden,eachgardenwouldneedtohaveanareaof195ft2.Thebioretention-infiltrationbasinwassizedfora0.9acreparkinglot,whichisroughlyequivalenttotheporousasphaltparkinglot,plusanadditionalareaforthebasin.Conservingnaturalareaswouldreducethetotalareathatwouldgeneraterunoff.Weassumedthattheconservedareawouldbereplacedbyanimpervioussurface.Thestreettree(treeplanter/pit)calculatorrequiresamaximumof33%impervioussurface.Thecorrespondingacreageproducing3,000ft3WQvis2.4totalacresand0.8imperviousacres.Thebasecasestreettreewasassumedtobeamedium-sizeddeciduoustree,suchasaredoak(McPhersonetal.,2006)(Table3).

Rainbarrelswerecalculatedseparatelyandwerenotstandardizedto3,000ft3.Rainbarrelsareahouseholdscalepracticeandweassumedahousewouldinstalltwo55-gallonbarrels.Asnotedabove,residentiallotshaveanaverageimperviouscoverof41percent.TheaverageresidentiallotsizeinGrandRapidsis0.17ac.Theimpervioussurfaceinaresidentiallot,mostofwhichisassumedthebetheroof,is0.07ac.Tworainbarrelswithacombinedstorageof110gallons(14.67ft3)cancapturetherunofffromatypicalresidentiallotuptoa0.055inevent.Weassumethateventslargerthanthatwillbestoreduptothemaximumcapacityandtheremainderwilloverflow.Tworainbarrelscanavoid1,625ft3ofrunoffperyear.TheresultsforrainbarrelsshouldnotbedirectlycomparedtotheotherGIpracticesbecauseofthedifferentmethodology.

Table3:Amountofgreeninfrastructurerequiredtoreduce3000ft3ofrunoffper1”rainevent.

SMP Totalarea(acres)

Imperviousarea(acres)

Amountrequiredtoreduce3,000ft3WQvper1”event

Annualrunoffavoided(allevents<1.0”plus1.0”fromlargerevents)

Porousasphalt 0.87 0.87 0.87ac 113,257Raingarden 1.96 0.81 0.04ac 113,326Greenroof 0.87 0.87 0.85ac 113,257Infiltrationbioretention 1.00 0.86 0.07ac 113,248Conservationofnaturalareas*

0.87 0.00 0.87 113,257

Streettree(treepit)** 2.40 0.79 342trees 113,257Rainbarrel*** 0.17 0.07 N/A 1,625*reducedtotalareaby0.87ac,notactualstormwatervolume**reducedimpervioussurfaceareaby0.79ac,notactualstormwatervolume***Storesraineventsupto0.05”,notcomparableinscaletoothergreeninfrastructurepractices

11

Oncethesizeofthegreeninfrastructurepracticewasdetermined,thecostforeachwasestimatedusingtheLowImpactDevelopment(LID)CostToolsfromtheWaterEnvironmentResearchFoundation(WERF).TheLIDCostToolsareExcel-basedspreadsheetsthathavedefaultparametersbutcanbemodifiedforparticularsituations.ThedefaultcasewasmodifiedtofittheGrandRapidsstudyarea,includinglocalandcurrentwagesand,wherepossible,costestimatesfromlocalserviceproviders.EachGIpractices’sizewasadjustedbasedonthedesired3,000ft3WQvreductionper1.0inchevent.Thedefaultcostswereadjustedforinflationto2015from2005usingtheConsumerPriceIndexfromtheUSBureauofLaborStatistics.Maintenancecostswerescaledtotheprojectsizewhereappropriate.TheCityofGrandRapidsexpectsthatitsdetentionbasinsandraingardenstohavea50yearlifespan.Thecityplanstoreplaceporouspavementafter25years(CityofGrandRapids,2014).Weusedtheselifespanestimatesinourmodel.

Thecapital(installation)andperiodicoperationsandmaintenance(O&M)costswerecombinedintoapresentvaluecalculation.A3.5percentdiscountratewasusedforallpresentvaluecalculations.Thisrateisappropriateforenvironmentalprojectswithalifespanof30-75years(AlmansaandMartínez-Paz,2011).TheCityofGrandRapidsusesa50-yearinfrastructureplanninghorizonwhichisreplicatedinthisanalysis.

PollutionandfloodriskreductionInadditiontoreducingstormwatervolumes,GIpracticesreducepollutionenteringwaterways.Theannualpollutionloadfromaparticularsitecanbeestimatedusingthefollowingformula(Landphairetal.,2000):

𝐿𝑜𝑎𝑑 𝑙𝑏𝑠 = 0.2266 ∗ 𝐴𝑟𝑒𝑎 𝑎𝑐 ∗ 𝑅𝑎𝑖𝑛𝑓𝑎𝑙𝑙 𝑖𝑛 ∗ 𝑅G ∗ 𝐶(𝑚𝑔

𝐿)

WhereRvistherunofftorainfallratioandCisthepollutioncoefficient.Therainfallamountwastheannualtotalforevents<1.0incheswhichtotaled27.95inches.RvwascalculatedforeachGIpracticeusingtheNYSStormwaterToolbox.Weissetal.(2007)reviewedseveralsourcesandfoundthatcontaminantloadswerefairlyconsistent.ReportedvaluesforCaveraged131+/-77mg/L(ppm)witha67%confidenceintervalfortotalsuspendedsolids(TSS)and0.55+/-0.41mg/L(ppm)witha67%confidenceintervalfortotalphosphorus.Theseaveragevalueswereusedinthecalculations.AlloftheGIpractices,exceptrainbarrels,weresizedtotreat3,000ft3WQvfora1.0inchevent.TheMinnesotaStormwaterManualreportedthepollutionreductionefficiencyforvariousGIpractices(Table4).Notethatgreenroofsdonotremovephosphorusfromstormwater(MinnesotaPollutionControlAgency,2015).

Table4:PollutionreductionfromgreeninfrastructureSMPs.

PollutionreductionefficiencySMP Totalsuspendedsolids PhosphorusPorousasphalt 74% 45%Greenroof 85% 0%Raingarden 85% 100%Bioretentioninfiltration 85% 100%

12

Treepit 85% 80%

TheeconomicvalueofremovingTSSandphosphoruswasestimatedfromthetreatmentcostfromawastewatertreatmentplant.A2007reportestimatedthatTSSremovalcosts$5.70/lbandphosphorusremovalcosts$220/lb(WSB&Associates,2008).Adjustedforinflationto2015usingtheCPI,thesecostsare$5.93/lband$251.25/lb,respectively.Thesevalueswerebasedonthetreatmentplant’s30-yearlifecyclecostincludingcapitalandO&Mcosts.Multiplyingthepollutantreductionamount(lb)bytheunitcost($/lb)resultedinthevalueofstormwaterremovalforeachft3ofWQvavoidedperyear(Table5).Weassumedthatrainbarrelswouldhavethesamepollutionreductioneffectasraingardens.

Table5:Unitpriceofpollutionreduction.

GIpractice Annualpollutionreduction(lbs/ft3WQv/year)

Economicvalueofavoidedpollution($/ft3WQv/year)

Porouspavement TSS0.00818P0.00003

TSS$0.036P$0.004

Greenroof TSS0.00818P0.00000

TSS$0.041P$0.000

Raingarden TSS0.00692P0.00003

TSS$0.041P$0.009

Bioretentioninfiltration TSS0.00692P0.00003

TSS$0.041P$0.009

Treeplanter/treepit TSS0.00686P0.00003

TSS$0.041P$0.007

Rainbarrel TSS0.00692P0.00003

TSS$0.041P$0.009

Reducingthevolumeofstormwaterenteringarealakesandriversalsoreducestheriskoffloodingindownstreamlocations.In2013,theGrandRiver,whichflowsthroughdowntownGrandRapids,flooded,causinganestimated$450millionindamages.Thetotalwatervolumeoverthe18-footfloodstagefortheGrandRiverfortheentiretwo-weekfloodperiodinApril2013was3.9billionft3(USGS2016).Thiscomesoutto$0.11/ft3offloodwater.Thoughthiswasalargeflood,itwasnotrecordsetting.Afloodofthismagnitudehasa10to25-yearrecurrencetime.Thatis,agivenyearhasa4-10percentchanceofafloodofthissize.Assuminga25-yearrecurrencetimetheannualexpecteddamagewouldbe$0.11/ft3*0.04=$0.005/ft3.Reducingonecubicfootofstormwatervolumeisexpectedtoavoid$0.005indamageseachyear.Thisistheconservativeestimate.Usingtheten-yearrecurrencetime(tenpercentannualchance)wouldresultinhigherdamageestimates.

OtherbenefitsfromspecificGIpracticesGreenroofsResearchersattheUniversityofMichigandocumentedthebenefitsofgreenroofsoncampusbuildings(Clarketal.,2008).UsingtheUSEnvironmentalProtectionAgency’s(EPA)EnergyPlus2.0simulator,theyfoundthattheinsulatingpropertiesofgreenroofssaved$0.36/m2(2006dollars)or$0.04/ft2in2015dollars.Thisequatesto$0.013/ft3WQv/year.Theanalystsalsofoundthatthegreenroof’sgrowing

13

plantstookuptheairpollutantNO2atarateof0.27kg/m2/year(0.06lb/ft2/year).Theeconomicbenefitsfromthepollutionreductionwereestimatedat$1,680-6,380/Mg($0.76-2.89/lb)in2006($1,982-$7,526/Mg($0.90-3.41/lb)in2015).Applyingtheseratestothe37,200ft2ofgreenroofundertheGIpracticescenarioyieldsabenefitof$0.016-0.062/ft3WQv/year.Themoreconservative,lower-boundestimateof$0.016/ft3/yearwasusedinouranalysis.

Greenroofsalsoprovideascenicamenityvaluewhentheyarevisiblefromupperfloorsoradjacentbuildings.IchiharaandCohen(2010)usedahedonicmodeltoestimatetheamenityvalueofgreenroofsinNewYorkCity.Theyfoundthat,allelsebeingequal,thepresenceofagreenroofaddedabout16percenttoaresidence’ssalesprice.Thecoefficientforthegreenroofindicatorvariablewas0.1496+/-0.0729whichyieldsalowerboundofeightpercent.Thestudy’sNewYorkCitylocationisnotacloseanalogforGrandRapidswherehousingvaluesaremuchlowerandaccesstoground-levelgreenspaceisplentiful.TheIchiharaandCohenpaperis,todate,theonlyhedonicmodelthatconsiderstheamenityvalueofgreenroofs.Sanderetal.(2010),forcomparison,foundthatatenpercentincreaseinstreettreecanopyraisedpropertyvaluesinMinnesotaby0.29-0.48percent.Thiseffectheldupto40percenttreecover.Agreenroofsubstitutesvegetatedcoverforanentirelyimpervioussurface.Assumingthattheeffectofagreenroofissimilartostreettrees,thenagreenroofsamenityvaluecouldbeintherangefora40percentincreaseintreecanopy(1.16-1.92percent).Giventhelackofsolidregionaldatafortheamenityvalueofgreenroofs,ourbestjudgmentisitliesbetweenzeroforroofslackingvisibilityandaccessto2.0percentforhighlyvisibleroofswitheasyaccess.

MostgreenroofsinGrandRapidsareonofficebuildings,arehighlyvisible,andareaccessible(Greenroofs.com,2015).Thereforeweusedthehigh-endestimateofa1.9percentpropertyvalueamenity.TheaverageaskingpriceofofficepropertyinGrandRapidsis$71.26/ft2(LoopNet,Inc.,2015)whichyieldsacapitalizedamenityvalueforgreenroofsinGrandRapidsof$1.35/ft2.Theannualizedamenityvalueofgreenroofsata3.5percentdiscountrateis$0.06/ft2/yearor$0.019/ft3WQv/yearofWQvreduced.

Greenroofsprotecttheconventionalroofsunderneaththem.Analystsreportthatgreenroofscandoublethelifetheconventionalroofandeliminatetheneedforafullroofreplacementaftertwenty-fiveyears.Sinceanewroofcostsabout$10/ft2(K.Menard,personalcommunication),thisisasubstantialbenefit.

RaingardensandinfiltrationbioretentionbasinsRaingardensalsoprovideascenicamenity.Polyakovetal.(2015)studiedtheamenityvalueofraingardensplacedatstreetintersectionsinSydney,Australia.Raingardenswerefoundtoincreasethemedianpropertyvaluebysixpercentforthosewithin50m(164ft)andfourpercentbetween50mand100m(164-328ft)fromtheraingarden.ApplyingthesixpercentratetoGrandRapidsmediansalespriceof$129,900resultsinaneffectof$7,794.Ata3.5percentdiscountrate,theannualizedvalueofaraingardenis$332or$0.032/ft3WQv/yearofrunoffreduced.Stormwaterinfiltrationbioretentionbasinsalsocan,ifcarefullydesigned,haveanamenityvalue.LeeandLi(2009)foundthatinTexas,ordinary(singleuse)detentionbasinshadnoinfluenceonresidentialhousingprices.Multi-usedetentionbasins,ontheotherhand,includerecreationamenitiesintheirdesign.Homesclosertothemulti-usebasinssoldforhigherpricesthanthosefurtheraway,allelsebeingequal.Forouranalysis,weconservativelyassumedthattheinfiltration-bioretentionpracticewassimilartotheordinary,single-usedetentionbasinandprovidesnoamenityvalue.

14

StreettreesandconservedopenspaceUrbantreesprovidemanyecosystemservicesbeyondstormwatermitigation.TheMidwestCommunityTreeGuidedocumentedandquantifiedthebenefitsprovidedbyurbantrees(McPhersonetal.,2006).Theguideliststhebenefitsofstreettreesbyunit(kWhofelectricitysaved,poundsofairpollutantsavoided,etc.)aswellastheprice($/unit)ofeach.Themagnitudeofthebenefitschangesasthetreegrowsinsizeandmaturity.Forthisanalysis,weusedtheguide’sunitsandupdatedthemwithcurrentandlocallyappropriateprices.InadditiontothebenefitsdescribedbyMcPhersonetal.,wealsoincludethereducedfloodingrisk,reductionintotalsuspendedsolids,andreductioninphosphorus.Table6belowshowstheunits,prices,andsourcesforeachbenefit.TheavoidedrunoffvolumeestimatesreportedbyMcPhersonetal.werehigherthanthoseresultingfromtheNYSStormwatertoolbox.ThevolumereductionestimatedbyMcPhersonetal.wasbasedmostlyonrainfallinterception.ThevolumereductionestimatedbytheNYSStormwatertoolboxfocusedmostlyoncapturingrunoffintheperviousareaunderthetree.Aftersomedeliberation,theteamdecidedtousetheMcPhersonrunoffreductionestimatesinthebenefitcalculation.TheNYSStormwatertoolboxwasused,however,todeterminethesizeoftheGIpractice(numberoftrees)tobeconsistentwiththeotherGIpractices.

Table6:Benefitsofstreettrees(basedonMcPhersonetal.2006).

Benefit Price($/unit) SourceAvoidedrunoff $0.0002/gallon CityofGrandRapidsdata($0.0017/ft3/yearWQv)Electricitysavings $0.126/kWh 2014EIAEastNorthCentralresidential,adjusted

forinflationHeatingsavings $0.009/kBtu 2014EIAMichiganaverageresidentialnaturalgas

priceCO2sequestered $0.018/lbCO2 EPAsocialcostofcarbonfor2015,$40/tonCO2Airpollutionavoided

Various McPhersonetal.valuesadjustedforinflationusingCPI

Aestheticvalue 0.81%ofresidentialhousingprice

McPhersonetal.percentageappliedtoGrandRapidsaveragehousingsalesprice,$129,900

Floodriskreduction

$0.005/ft3WQv/year Currentanalysis

Totalsuspendedsolids

$0.041/ft3/year Currentanalysis

Phosphorus $0.007/ft3/year Currentanalysis

Conservednaturalareashavebeenshowntoincreasethepropertyvaluesoftheadjacentlots.Thorsnes(2002)usedahedonicmodeloftheGrandRapids,Michiganareaandfoundthatforestpreservesadd19-35%tothesellingpriceoflotsadjacenttothepreserve.WithaverageGrandRapidshomessellingfor$129,900in2015,the19%premiumis$24,681.Annualizedover50yearsata3.5percentdiscountrate,thevalueis$1,052perhomeperyear.Wealsoassumethatthepreservednaturalareawouldbeadjacentto12lots.Thetotalamenityvalueistherefore$12,627.Ouranalysisassumesthattheconservednaturalareawouldotherwisebeconvertedto0.9acresof100percentimpervioussurface,whichwouldgenerate3,000ft3WQvforaone-inchrainevent.Theresultingamenityvalueis$0.111/ft3WQv/year.Manyoftheservicesprovidedbymature(25+yearold)streettreeswereadaptedfortheconservednaturalareagreeninfrastructurepractice,includingthefollowing:carbondioxidestorageat$0.029/ft3/year;reducedairpollutionat$0.004*342trees=$0.013/ft3/year;avoidedstormwaterat

15

$0.0017/ft3/year;floodriskreductionat$0.005/ft3/year;reducedtotalsuspendedsolidsat$0.041/ft3/year;andreducedphosphorusat$0.007/ft3/year.Thetotalannualbenefitfromconservednaturalareaswas$0.208/ft3/year.

ThebenefitsfromallGIpracticesaresummarizedinTable7andFigure1.

Table7:SummaryofbenefitsfromGIpractices.

Unitpriceofstormwatermitigationbenefits($/ft3WQv/year)

GIpractice Porousasphalt

Greenroof

Raingarden

Streettree(medium)*

Infiltrationbioretention

Conservenaturalarea

Rainbarrel

Avoidedvolume 0.0017 0.0017 0.0017 0.0017 0.0017 0.0017 0.0017

Floodriskreduction 0.0051 0.0051 0.0051 0.0051 0.0051 0.0051 0.0051

TSSpollution 0.0359 0.0412 0.0410 0.0407 0.0410 0.0407 0.0410Phosphoruspollution 0.0039 0.000 0.0086 0.0068 0.0086 0.0068 0.0086

Amenityvalue 0 0.0190 0.0323 0.0116 0 0.1115 0Energysavings 0 0.0132 0 0.0243 0 0 0Airpollutionreduction 0 0.0163 0 0.0034 0 0.0126 0

CO2storage 0 0 0 0.0066

0 0.0292 0

Totalannualbenefits 0.0466 0.0965 0.0887 0.10-0.25 0.0564 0.2077 0.0564

*Benefitsduringfirstfiveyearsandincreasesthereafter.

Figure1:BenefitsofvariousGIpractices.*Streettreebenefitsforthefirstfiveyears,benefitsincreasewithtreesize.

$- $0.0500 $0.1000 $0.1500 $0.2000 $0.2500

Porousasphalt

Raingarden

Infiltrationbioretention

Rainbarrels

$/ft3/year

Benefitsofgreeninfrastructurepractices($/ft3/year)

Avoidedvolume Floodriskreduction TSSpollution

Phosphoruspollution Amenityvalue Energysavings

Airpollutionreduction CO2storage

16

GIcostsGreeninfrastructurecostswereestimatedusingtheWERFLIDspreadsheettoolsasastartingpointandadjustedforinflation,technologicaladvancements,andlocation-specificdata.

PorousasphaltCenturyWestEngineeringcompared,side-by-side,thecapitalcostsofconventionalandporousasphaltfora3,200ft2parkinglot.Theconventionalasphaltlotcost$23,680toconstruct($7.40/ft2)whiletheporousasphaltcost$25,960($8.11/ft2).ThecomparisonwasconductedinPortland,Oregonin2013.AccordingtotheBureauofLaborStatistics,Portland’smeanconstructionwageis$25.64/hourandGrandRapids’is$21.04.Thisratio(0.8)wasusedtoadjustthePortlandasphaltconstructioncoststoGrandRapidsandthecostswereadjustedforinflationto2015usingtheCPI.Theadjustedcapitalcostswere$6.20/ft2forconventionalasphaltand$6.79/ft2forporousasphalt.

ThefigureswereenteredintotheWERFLIDcostmodelwhichincludes10%forengineeringandplanningand20%forcontingency.Thetotalcapitalcosts,includingconstructionanddevelopmentcosts,were$8.19/ft2($2.74/ft3/yearWQv)forconventionalasphaltand$8.96/ft2($3.00/ft3/yearWQv)forporousasphalt.

Bothconventionalandporousasphalthavemaintenancecosts.Itwasassumedthatporousasphaltwouldhaveallthemaintenanceofconventionalasphaltplusitsownspecializedmaintenance.TheWhitestoneFacilitiesMaintenanceandRepairReferenceandtheOperationsReferencelistrecommendedmaintenancehoursforspecifictasksrelatedtofacilitiesmanagementincludingparkinglots(Abateetal.,2009).Thescheduleforeachtaskislistedinthetablebelow.TheBureauofLaborStatisticsreportsthattheaverageconstructionwageinGrandRapidswas$19.86in2015.Thiswagewasusedtocalculatethecostforeachtask(Table7).

Table8:Operationsandmaintenancecostsforconventionalasphalt.

O&Mtask Frequency Hours/ft2 Materialscost$/ft2

Totalcost$/ft2

Totalcost$/ft3/yearWQv

Patchandseal 5years 0.0006 0.07 0.09 0.03Resurface 15years 0.0127 0.46 0.76 0.26Repair 25years 0.0073 0.80 1.04 0.35Snowplowandsweep(weekly)

Annual - 0.03 0.03 0.01

Houleetal.(2013)estimatedthemaintenancecostsforporousasphaltinNewHampshire.NewHampshire’scoldclimateissimilartothatofGrandRapidsandthusisanappropriatecomparison.Houleetal.estimatedthatthepersonnelcostswouldbe$939/ha/year($380/acre/year)andsubcontractorcostswouldbe$1730/ha/year($700/acre/year).ThesevalueswereadjustedtofittheGrandRapidsareabyscalingthembytheaverageconstructionandextractionoccupationwagesfromthetwocitiesasreportedintheUSBureauofLaborStatistics’OccupationalEmploymentStatistics.Thefigureswerealsoadjustedforinflationto2015usingtheCPI.Theadditionalmaintenancecostsforporouspavementwereestimatedat$0.02/ft2($0.008/ft3WQv).Thepresentvaluecostoverfiftyyearscomesoutto$3.54/ft3WQvforconventionalasphaltand$3.99/ft3WQvforporousasphalt(Table9).

17

GreenandconventionalroofsAMichigan-basedroofingcontractorestimatedthecostofaconventionalcommercialroof(2,000-10,000ft2)witha25yearlifespanatroughly$10/ft2(K.Menard,BloomRoofing,personalcommunication).Mr.Menardalsoestimatedmaintenancecostsatabout$0.05/ft2/year.TheWhitestoneFacilitiesMaintenanceandRepairReferencelistedaspecificscheduleforroofmaintenanceoveritslifetime.Annualmaintenancetaskswere$0.06/ft2($0.07in2015dollars)withmoresubstantialperiodictasksatacostof$0.11/ft2everyfiveyearsand$3.15/ft2inyear15(Abateetal.,2009).Theentireroofwouldbereplacedattheinitialcostinyear25.Forouranalysis,wesubstitutedMenard’slocalestimateforroutinemaintenance($0.05/ft2)intheWhitestonereferencemanual,keepingtheperiodicmaintenancethesame.StandardizedbyWQv,aconventionalroofcosts$3.31/ft3WQvinthefirstyearwithmaintenancecosts$0.02/ft3WQv/year.

Agreenroofrequirestheinstallationofaconventionalroofunderneathit.Thereforethecostofagreenroofisadditionaltotheconventionalroof.Localrefinementsinthegreenroofestimateswereprovidedbyalocalgreenroofcompany(J.Aleck,LiveRoof,personalcommunication).IntheGrandRapidsarea,adeliveredgreenroofsystemscosts$9-20/ft2plusinstallationcostsofabout$3/ft2.Forthisproject,weusedanestimatedinstalledcostof$15/ft2.Greenroofsrequiresomemaintenancewhichvariesbythelocalenvironment,soilsystem,andtypeofplantsused.Astandardgreenroofwithroutinemaintenanceperformedbyin-housestaffcostsabout$0.13/ft2/yearandaboutdoublethatifthemaintenanceisoutsourcedtothegreenroofinstaller.Ouranalysisassumedthatthemaintenancewouldbedonein-house.Agreenroofmayextendthelifeoftheconventionalroofunderneathit(Clarkeetal.2008).Weassumedthatthepresenceofthegreenroofwouldeliminatetheneedtoreplacetheconventionalroofinyear25.StandardizedbyWQv,agreenroofhasaninstallationcost,includingthecostofaconventionalroofof$8.24/ft3andamaintenancecostof$0.04/ft3/year.Thepresentvaluecostfortheconventionalroofwasestimatedat$5.77/ft3WQvand$9.23/ft3WQvforthegreenroof(Table9).

RaingardenThe2,145ft2ofraingardensneededtomitigate3,000ft3ofstormwaterwasassumedtobespreadover11residentialhomes(195ft2perraingarden).TheWashingtonStateDepartmentofEcologyestimatedtheinstallationandmaintenancecostsofraingardensunderthecategoryofbioretentionbasin(HerreraEnvironmentalConsultants,2012).Themeanofthereportedlowandaveragecostswere$17.95/ft2and$0.73/ft2/yearforcapitalandmaintenancecosts,respectively.TheconstructionwagedifferentialbetweenWashingtonandMichiganwasusedtoadjustforlocationandthecostswereadjustedforinflationusingtheCPI.TheGrandRapidscostsfor2015were$14.00/ft2forcapitaland$0.57/ft2/yearformaintenance.TheWERFLIDcosttoolincludesothercosts,suchaslandscapedesign,first-yearestablishment,andperiodicmaintenancecosts.Forthe2,145ft2ofraingardenGIpractice,thetotalcapitalcostis$32,788plusregularmaintenancecostsof$1,223/year.Periodicmaintenancetasksincludereplacingmulcheverythreeyears($4,605)andtillingthesoileveryfiveyears($3,105).Itwasassumedthattheraingardenswouldbeinstalledprofessionally.Opportunitycostsoflandwereincludedasdescribedabove.Capitalandmaintenancecostscouldbesubstantiallylowerifhomeownersorvolunteersdidtheworkthemselves.Standardizedonaperft3WQvbasis,thetotalfirst-yearcostis$0.30/ft3WQv/yearwithatotalpresentvaluecostof$1.03/ft3WQv(Table9).

18

InfiltrationbioretentionThecostofbioretentioninfiltrationbasinswasbasedonthecostestimatesfromthePugetSoundgreeninfrastructurecostdatabase(HerreraEnvironmentalConsultants,2012).Theaveragepublishedestimate,from2014,was$31.61/ft2and$1.27/ft2inconstructionandmaintenancecosts,respectively.TheaverageconstructionwageasreportedbytheBLSEmploymentSurveywasusedtoadjustthecostsfromtheSeattleareatoGrandRapids.ConstructionwagesinGrandRapidsareabout75.4percentofthoseinSeattle.TheCPIwasusedtoadjustthecoststo2015.Thecostforbioretentionconstruction,adjustedforlocationandtime,was$24.02/ft2andtheadjustedmaintenancecostwas$0.97/ft2.Thebioretentioninfiltrationbasinwassizedat3,049ft2.Thetotalpresentvaluecostforabioretention-infiltrationbasinwas$1.34/ft3WQvoverfiftyyears(Table9).

ConservationofnaturalareasConservingnaturalareascomeswithahighopportunitycost–thelandwillnevercontainincome-producingstructures.ThisopportunitycostofopenspacewasestimatedusingThorsnes'(2002)hedonicanalysisofopenspacepreservationintheGrandRapids,Michiganarea.Themodelincludedavariableforlotsize.ThorsnesanalyzedthreedevelopmentsaroundGrandRapids.WechosetobaseourcalculationsonthemodelforthedevelopmentclosesttothecityinadjacentPlainfieldTownship.Thehedonicmodelyieldedanelasticityof0.0031forlotsize;thatis,aonepercentchangeinlotsize(ft2)resultsina0.3percentchangeinhousingsalesprice.Theaverageresidentialsalespricein2015was$129,900andtheaveragelotsizeforthenortheasternportionofGrandRapidswas9,148ft2.Thevalueofanadditionalsquarefootoflotsize,therefore,was$4.40/ft2.Theannualizedvalueata3.5percentdiscountratewas$0.24/ft2.Thereforeconserving0.9acresofnaturalareawouldhaveanopportunitycostof$0.08/ft3WQv/yearandapresentvaluecostoverfiftyyearsof$1.94/ft3WQv(Table9).

StreettreesThecostsofstreettrees,plantedinstormwater-retainingtreepits,wastakendirectlyfromtheMidwestCommunityTreeGuidewhichliststhecostsforplantingandmaintainingatreefor40yearsinfiveyearincrements(McPhersonetal.,2006).Thecostswereadjustedforinflationto2015dollars.Theguidepresentssmall,medium,andlargetreesizeoptions;wechosemedium,suchasredoak,whichiscommoninthearea.Thecostofplantingatreewas$200($244in2015)or$0.74/ft3WQv.Totalmaintenancecosts,whichincludepruning,removalanddisposal,treatingpestsanddisease,infrastructurerepair,irrigation,cleanup,liabilityandlegalcosts,andadministrativecosts,were$55.21-33.00peryear($18.57-40.26peryearin2015dollars).StandardizedtoWQv,themaintenancecostsrangefrom$0.06/ft3to$0.12/ft3.ThoughtheWERFLIDcosttoolincludesaconcretetreevaultinthedefaultsettingforstreettrees,theForestServiceanalysisdidnotincludeatreevault.Treevaults,whichmaycostmorethan$1,000,werenotincludedinouranalysis.Thetotalpresentvaluecostofthestreettreesoverfiftyyearswas$2.92/ft3WQv(Table9).

RainbarrelsRainbarrelsarecommerciallyavailableathardwarestores.Anationalchainwassellingbasic-style55-gallonrainbarrelsfor$90perbarrel(TheHomeDepot,2015).Therainbarrelswereassumedtobereplacedatthesamecosteverytenyears.Thereisnooperationandmaintenancecostforrainbarrels.Thetotalpresentvaluecostfortworainbarrelswas$0.11/ft3WQv(Table9).

19

OpportunitycostoflandGreenroofsandporousasphaltparkinglotsareco-locatedwithexistinginfrastructure.Raingardensandstreettrees,however,replaceothervaluableresourcessuchaslawnspaceorsidewalks.Theopportunitycostneedstobeaccountedfor.TheopportunitycostwascalculatedusingthevalueofasquarefootofresidentiallotsizeintheGrandRapidsmetropolitanarea.Theopportunitycostoflandforraingardens,bioretentionponds,andstreettreeswascalculatedusingthesamemethodasthatforconservationofnaturalareas.Forthe2,145ft2ofraingardentheopportunitycostequatesto$0.0036/ft3/yearofWQv.Thissameopportunitycostwasappliedtothe342streettreesandthebioretention-infiltrationsystems.Rainbarrelshavenoopportunitycostforland.

Table9:Costforgreeninfrastructurepractices.

Infrastructure/GItype GIpracticesize(for3,000ft3WQvreductionperone-inchevent)

PVcost PVcost/ft3WQv

PVcost/unitofGIpractice

Conventionalasphalt 37,897ft2 $400,395 $3.54 $10.57/ft2Porousasphalt 37,897ft2 $451,397 $3.99 $11.91/ft2Conventionalroof 37,500ft2 $653,062 $5.77 $17.41/ft2Greenroof 37,200ft2 $1,045,565 $9.23 $28.11/ft2Raingarden 2,145ft2 $116,816 $1.03 $54.46/ft2Infiltrationbioretention 3,049ft2 $151,447 $1.34 $49.67/ft2Conservenaturalareas 37,897ft2 $220,017 $1.94 $5.81/ft2Streettree(medium) 342trees $359,665 $3.18 $1,051.65/treeRainbarrels 2-notstandardizedto3000ft3 $507.66 $0.31 $253.85/barrel

ResultsTheNPVanalysisshowsthatfiveofthesevengreeninfrastructurepracticeshavepositivenetpresentvaluesunderthebasecaseassumptions(Table10,Figure2).Conservingnaturalareashadthehighestnetbenefits($3.10/ft3)followedbystreettrees($1.48/ft3)andraingardens($1.12/ft3)and.Porousasphaltalsohadapositivenetpresentvalue($0.68/ft3)asdidinfiltrationbioretention($0.03/ft3).RainbarrelshadapositiveNPV($1.06/ft3)butthesewerenotanalyzedonthesamescaleastheotherpracticesandshouldnotbedirectlycompared.Greenroofs,however,hadanegativenetpresentvalueof$-1.12underthebasecaseassumptions.Greenroofsprovidedthehighestbenefitsbutalsohadthehighestcosts.

Porouspavementreplacestheconventionalpavement“grayinfrastructure.”Thegreenroofiscomparedtotheconventionalroofitwouldreplace.Inallothercases,thegreeninfrastructureisadditionalto,anddoesnotreplace,grayinfrastructure.ThebenefitsofgreeninfrastructureinthisstudycomeprimarilyfromavoidedstormwatervolumeswhichareassociatedwithreducedO&Mcosts,flooding,andpollutionaswellas,insomecases,enhancedscenicamenities.Newdevelopmentsinwhichgreeninfrastructurepracticesareimplementedexplicitlytomanagestormwateron-sitemayreducethecapitalcostsofgrayinfrastructure.HoweverintheCityofGrandRapids,asinmosturbanareas,theexistinggrayinfrastructurewillnotberemovedorsignificantlyreduced.

20

Table10:NetpresentvalueforGIpractices.

Infrastructure/GItype

GIsize(for3,000ft3WQvper1”event)

PVbenefits($/ft3WQv)

PVcostGI($/ft3WQv)

PVcostofgray($/ft3WQv)

NetPresentValue($/ft3WQv)

Porousasphalt 37,897ft2 $1.13 $3.99 $3.54 $0.68Greenroof 37,200ft2 $2.34 $9.23 $5.77 $-1.12Raingarden 2,145ft2 $2.15 $1.03 - $1.12Bioretentioninfiltration

3,049ft2 $1.37 $1.34 - $0.03

Conservenaturalarea

37,897ft2 $5.04 $1.94 $3.10

Streettree(treepit)

342trees $4.66 $3.18 - $1.48

Rainbarrel 2barrels* $1.37 $0.31 $1.06*Notstandardizedto3,000ft3

Figure2:Benefits,costs,andNPVsofGIpractices.

DiscussionTheGIpracticesshowedahighdegreeofvariabilityamongtheirnetpresentvalues.ConservationofnaturalareasowesitshighNPVprimarilytotheamenityvalueitbringstoaneighborhood.Thescenicamenityvalueaccountsformorethanhalfofthetotalannualbenefit($0.11/ft3outof$0.21/ft3).Thecostofconservingnaturalareascomesfromtheopportunitycostofdevelopment.Weassumedthese

$(2.00)$- $2.00

$4.00

$6.00

$8.00

$10.00

NPV

$/ft3WQv

Benefitsandcostsofgreeninfrastructurepractices

TotalPVbenefits(includingavoidedgrayinfrastructure)

PVcostofgreeninfrastructure

NetPresentValue

21

areaswouldbekeptinarelativelynaturalstatewithoutmaintenancecosts.Whiletherecouldbesomeadditionalcostsassociatedwiththis,suchasdeerandotherwildlifeeatingresidentialgardenplants,theseweredifficulttoquantifyandwerenotincludedintheanalysis.Thepremiumpaidonlotsadjacenttotheconservednaturalarea,especiallywhencombinedwiththesuiteofotherecosystemservices,outweighstheopportunitycost.Thissuggeststhatlow-impactdevelopmentpatternsthatconcentratedevelopmentinoneareawhileleavingnaturalareasintactcanbeahighlycost-effectivepractice.Itischeapertoavoidgeneratingstormwaterrunoffratherthantreatingitlateron.Thisrequires,however,considerableplanningandlong-termcommitment.NaturalareasareoftenscarceincitieslikeGrandRapidssothispracticemayhavelimitedpotentialoutsideofgreenfielddevelopmentsites.

Streettreesweresecondintermsofnetpresentvalueat$1.48/ft3WQv.Streettrees,whenplantedinstormwaterretainingtreepits,providesubstantialbenefitsovertheirlifetimes.Trees,however,taketimetomatureandthefullbenefitofstreettreestakesdecadestoberealized.Since2006,costsforelectricityandheatinghaveincreasedfasterthanthegeneralrateofinflation.UpdatingtheMcPhersonstudywithcurrentcosts,aswellaswithadditionalwaterpollutionbenefits,showsthatstreettreesareevenmorevaluablethanoncethought.Thepresentvaluecostsarerelativelylowcomparedtoporousasphaltandgreenroofs.Maturetreesprovideahighlevelofbenefitbutittakesdecadesforthetreestogrow.Evenwithareasonablediscountrate,thebenefitsofstreettreesstillexceedthecosts.Thisallsuggeststhatstreettreeplantersarecosteffectiveunderawiderangeofassumptions.

BecauseofthelowcapitalandO&Mcosts(PVcost=$2.15/ft3WQv),raingardensareanattractiveGIpracticesforhomeownersandsmallcommercialpropertyowners.Thesehadthethird-highestNPVofthegreeninfrastructurepracticesevaluated.Ouranalysisassumedthattheraingardenswouldbeprofessionallyinstalled.Thenetbenefitscouldbeevenhigherifthepropertyownersinstalltheraingardenthemselvesorwithvolunteerhelp.Raingardensarealsohighlyscalableandcanbeusedonlargeorsmallcitylots.

Inouranalysis,thepresentvaluecostofporousasphaltisabouttenpercenthigherthanthatofconventionalasphalt.Porousasphalthaspositivenetbenefitof$0.68/ft3WQv.StudiesfromtheUniversityofNewHampshire’sstormwatercentershowedthatporousasphaltcanbeacost-effectivesolutionevenincoldclimatessimilartothatofGrandRapids(Houleetal.,2013).Thoughporousasphaltiseffectiveatreducingstormwatervolumesandtreatingwaterpollution,itdoesnotprovideanyamenitybenefitsliketheothergreeninfrastructurepracticesconsideredhere.Parkinglotsareubiquitousand,accordingtoourresults,managingstormwaterfromparkinglotsusingporousasphaltresultsingreateroverallnetbenefitsthanusingbioretentioninfiltrationsystems.

Weassumedthattheentireimperviousareawouldbepavedwithporousasphalt.Thatmaynotbenecessary,however,asstrategicallyplacedareasofporousasphaltcaneffectivelytreatimperviousareasthatdraintoit.Thiswouldreducetheneededareaofporousasphaltandthusreducetheprojectcost.TheCityofGrandRapidsisalreadyexperimentingwithstripsofporousasphaltintheparkinglanesofsomecitystreets.

Thebioretention-infiltrationbasinpracticehadabarelypositivenetpresentvalue($0.03/ft3WQv).Bioretention-infiltrationbasinsactaslargeraingardens.Unlikeraingardens,thebasinsareusuallynotplantedwithwildflowersandarenotviewedasscenicamenities(LeeandLi,2009).Incaseswheredetentionpondsweredesignedasmulti-usecommunityresources,includingrecreationfacilities,LeeandLididfindanamenityvalue.Buildingsuchmulti-usestructuresrequiresadditionalcoststoachieve

22

thosebenefitsandthosearenotdirectlytiedtothefunctioningofthebasinitself.Thenetbenefitsoftheinfiltrationbioretentionpracticecouldbeimprovedifcost-effectivescenicandrecreationalamenitiesareincludedinthedesign.LeeandLifoundthat,allelsebeingequal,decreasingthedistancetothemulti-usedetentionbasinincreasedhomesalepricesatarateofabout$16/foot.Thecostofbuildingandmaintainingabioretention-infiltrationbasinwasalsohigherthanthatofaraingardenbecauseofthecommunity-levelscaleofmostprojects.

Agreenroofhasthehighestpresentvaluecost($9.23/ft3WQv)ofallthepracticessurveyedandapremiumof$4.83/ft3WQvoverastandardroof.Howeveragreenroofalsohassubstantialpresentvaluebenefits($2.34/ft3WQv)whichclimbevenhigherwhentheavoidedcostofroofreplacementisaccountedfor($8.11).Thenetbenefits,however,arenegative($-1.12/ft3WQv)usingthemid-rangeinstallationcostof$15/ft2.Thegreenroof’sPVcost(includingtheconventionalroofbelow)is60percenthigherthanaconventionalroofalone.Thisisconsiderablyhigherthanthegreenroofcapitalcostpremium(39percent)foundbyClarketal.(2008)butconsistentwithotherestimates(CarterandKeeler,2008,Sprouletal.,2014).Thegreenroofinstallerprovidedarangeofcapitalcostsfrom$9/ft2to$20/ft2.Acapitalcostof$11.50/ft2isthebreak-evenpoint.Belowthiscost,thegreenroof’snetpresentvalue,allelsebeingequal,wouldbecomepositive.Notethat$11.50/ft2isstillwithinthequotedcapitalcostrange.Undercertaincircumstancesthatenablealowcostinstallation,thegreenroofcouldbecost-effective.Alternatively,apositiveNPV($0.25/ft3WQv)canbeachievediftheamenityvalueisequaltoorgreaterthansevenpercentofthepropertyprice.ThislevelwouldbesimilartothelowerboundofIchiharaandCohen’sanalysisofgreenroofsinNewYorkCity.Manysmalltomid-sizeMidwesterncitieshaveadequateaccesstoground-levelpublicgreenspacecomparedtohighlyurbanNewYorkCity.TheCityofGrandRapidshaseighteenbuildingswithgreenroofs(Greenroofs.com,2015).Buildingownersevidentlyarewillingtopayforgreenroofs,sotheiramenityvaluesmaybegreaterthanthe1.9percentofsalespricehigherthanweestimatedhere.

ManyofthegreenroofsinGrandRapidsareinstalledtoachieveLEEDcertification(J.Aleck,LiveRoof,personalcommunication).StudieshaveshownthatofficespaceinLEEDcertifiedbuildingsrentsatapremiumof4-7percentascomparedtosimilarbuildingswithoutsuchcertification(FuerstandMcAllister,2011;Reichardt,2013).OfficespaceinGrandRapidsrentsforonaverage$13.25/ft2/year.TherentpremiumforaLEEDcertifiedbuilding,therefore,wouldbeabout$0.53-0.93/ft2/year.Ifweassumethatthecommercialbuildingis37,200ft2(LoopNet,Inc.,2015),whichistheareaofthegreenroofinourscenario,andonestory,theLEEDcertificationpremiumwouldbe$19,716-34,596/year.Thispremiumcouldoffsetthecostofsomeofthemoreexpensivegreeninfrastructurepractices,suchasgreenroofs.Agreenroofcancontributeupto5to23pointstowardthe40pointsneededforbasicLEEDcertification.AssumingallLEEDpointsarevaluedequally,agreenroofthatcontributed8.5-15pointstowardcertification(21-38percent)wouldhaveaLEEDamenityvalueofabout$0.11-$0.20/ft2or$0.04-$0.07/ft3WQv.AmodestLEEDamenityvalueof$0.15/ft2($0.05/ft3WQv)wouldbeenoughtoflipthegreenrooftoapositiveNPV($0.16/ft3WQv).ThisLEEDamenityvalueofgreeninfrastructurewasnotincludedintheanalysisbecausenotallgreeninfrastructurepracticesareimplementedtoachieveLEED,EnergyStar,orothersustainabilityratings.

Thisbenefit-costanalysiscomprehensivelydocumentedthevaluesassociatedwithGIpractices.Somevalues,however,aremorecertainthanothers.Theamenityvaluesforraingardensandgreenroofsinparticularareunderstudied.Ourliteraturereviewfoundonestudyofraingardenamenityvalues(Polyakovetal.,2015)andoneforgreenroofs(IchiharaandCohen,2010).Raingardenshavegrownin

23

popularitysoitshouldbepossibletoseewhethertheirpresenceaffectshousingvalues.Greenroofsarestillrelativelyrarebutbecomingmorecommon.GrandRapidsitselfishometoaboutonepercentofallknowngreenroofs(Greenroofs.com,2015).Greenroofsareamajorinvestmentforacommercialbuildingsowemaynotexpectbuildingownerstosellthemsoon.Intime,however,commercialbuildingswithgreenroofsshouldcomeonthemarketandtheiramenityvaluecouldbeassessed.

Rainbarrelsdonotfunctionatthesamescaleastheothergreeninfrastructurepracticesassessedhere.Forindividualhouseholds,however,raingardensareacosteffectivechoiceformanagingstormwater.Rainbarrelsarealow-costoptionwithapositivenetpresentvalue($1.06/ft3WQv).Thewaterfromrainbarrelsisusuallyusedtoirrigateflowerbeds.Thisanalysisdidnotincludetheeconomicvalueofthewaterusedforthispurposewhichwouldraisethenetpresentvaluefurther.Tworainbarrelscaptureasmallfractionofproperty’stotalrunoff.Rainbarrelscanbeusedtocomplementother,morecomprehensivegreeninfrastructurepractices.

ValidationTheestimatesusedinthisbenefit-costanalysiswerevalidatedagainstanactualgreeninfrastructureprojectimplementedintheCityofGrandRapids.GrandRapidsbuiltthePlainfieldIslandsbioretentionstructuresin2015.Itisasystemofsevenbioretention“islands”alongalargeurbanroad.Thoughitisreferredtoasabioretentionstructure,theislandsareplantedwithtreesandflowersthataddasubstantialscenicamenity.Inthisuniquecase,itseemsappropriatetousethecostsfortheinfiltrationbioretentionpractice,reflectingthehighly-engineeredstructure,andthebenefitsfortheraingarden,whichincludethescenicamenity.Thetotaldrainageareais2.2acresandthebioretentionareais0.1acres.BasedontheNYSStormwaterToolbox,a2.2acareawouldgenerate286,398ft3WQvperyearinrainfalleventsuptooneinchplusthefirstinchoflargerevents.Weestimatedthetotalpresentvaluecostofbioretention-infiltrationpracticesat$1.34/ft3WQv.Thereforeweestimatedthecapitalcostat$185,013($31/ft2)andthetotalpresentvaluecostofthePlainfieldIslandsbioretentionsystemat$518,380($87/ft2).Theactualconstructioncost,notincludingO&Mcosts,was$328,000or$55.12./ft2.Ourcapitalcostestimateislowerthantheactualcost.ThePlainfieldIslandsstructurewassignificantlylargerandmorecomplicatedthanourbasecasescenarioofasimpleinfiltrationbioretentionbasin.Whileimperfect,thisvalidationexercisesuggeststhatourcostestimatesarereasonable.

PolicyThisbenefit-costanalysisofgreeninfrastructurepracticeshaspolicyimplicationsforGrandRapidsandothersmalltomid-sizeMidwesterncities.First,greeninfrastructurepracticesprovideasuiteofbenefitsincludingstormwatervolumereduction,airandwaterpollutionreduction,andscenicamenities.Thetraditionalgrayinfrastructureprovidesafarmorelimitedsuiteofbenefits,oftenmanagingjustforstormwatervolumes.Thereforeinvestmentsingreeninfrastructurepracticesprovidefarmorevalueforeachdollarinvested.Second,thenetbenefitsfromgreeninfrastructurepractices,thoughvariable,aremostlypositive.Third,thebenefitsarelargelyexternaltothepropertyowners.Thatis,thecostsofgreeninfrastructurepracticesarebornebythelandownerbutthebenefitsaccruetothepublic.Thisresultsinamarketfailure–fewergreeninfrastructurepracticeswillbeimplementedthanaresociallyoptimal.Thereisastrongargumentforpublicpolicytoprovideincentives–financial,knowledge,orotherwise–formoreprivateinvestmentingreeninfrastructurepractices.

Astormwaterutilityfeemaybethemosteconomicallyefficientpolicytoincentivegreeninfrastructurepractices.ForexampletheCityofAnnArbor,Michigan,USA,implementsastormwaterutilityfee(City

24

ofAnnArbor,2015).Theadvantageofastormwaterutilityisthatitputsapriceonrunofffromimpervioussurfaces,thusinternalizingtheexternality.Landownershavetheflexibilitytoeitherpaythefee,orreducetheirrunoffbyinvestinginthemostcost-effectivegreeninfrastructurepractices.Pricingstormwaterrunoffalsofostersinnovationbyrewardingentrepreneurswhocaninventnext-generationgreeninfrastructurepractices.

Policiescanalsoovercomeknowledgeandinstitutionalbarriers.BarnhillandSmardon(2012)identifiedthreemajorbarriersthatlimitinvestmentingreeninfrastructure:the“publicgood”marketfailure;alackofknowledgeaboutthetruecostsandbenefits;andchallengesinframingtheissue.Themarketfailurecouldbeaddressedbyastormwaterutility.Acomprehensiveoutreachandeducationprogramcouldprovidemoreaccurate,relevant,andtimelyinformationtoresidentsandlandownersaboutgreeninfrastructure.BarnhillandSmardon’sworksuggesttheoutreachwillbemoreeffectiveifitisframedintermsofneighborhoodregeneration,sustainability,andsocialequity.Thisbenefit-costanalysiswasusedtocreateapubliclyavailable,webandmobile-basedGIcalculator,andtoinformoutreachandeducationcampaignsintheCitiesofGrandRapidsandMuskegon,Michigan.Theresearchteamiscollaboratingwithregionalpartnersandlocalgovernmentunitstoencouragelandownerstoadoptgreeninfrastructurepractices.

ConclusionsThebenefit-costanalysisforthevariousgreeninfrastructureGIpracticeshowsthatporouspavement,raingardens,infiltrationbioretention,conservingnaturalareas,andstreettreesarecost-effectiveoptions.Thelife-cyclecostsofgreenroofsontheirownexceedtheirbenefitsbuttheycanbecost-effectiveaspartofaLEEDcertifiedbuilding.NooneGIpracticeisappropriateforallsituations.RatherthechoiceofGIpracticewillbedrivenbythesiteandbudget.PorousasphaltisanattractiveGIpracticegiventhatparkinglotsarenecessaryandtheadditionalcapitalandO&Mcostsoverconventionalasphaltaremodest.Raingardensarelow-costandattractiveoptionsforsmallsiteslikehomesandstreetcorners.Infiltrationbioretentionbasinscanbeeffectivefortreatinglargerareasofimpervioussurface.Ifscenicandrecreationalamenitiesareincorporatedintothedesigntheymaybeevenmorecost-effective.Conservingnaturalareasrequiressubstantialup-frontplanningandawillingnesstoforgoimmediateincome.Overthefifty-yearprojectlifecycle,thebenefitsoftheconservedareasmorethanmakeupfortheopportunitycostofdevelopment.Streettreestaketimetofullyprovidethesuiteofstormwatermitigationandotherecosystemservices,buttheirbenefitsarestillgreaterthanthelifetimecosts.Rainbarrelsareacost-effective,small-scaleoptionforresidences.

Withthearrayofoptionsavailabletomanagestormwateronsite,municipalitieslikeGrandRapidsarewell-positionedtoadopttheGIpracticesthataremostappropriate.

AcknowledgementsTheRainwaterRewardsprojectwasfundedbytheUnitedStateForestServicewithfundsfromtheGreatLakesRestorationInitiative.Theauthorswouldliketothankthemembersoftheprojectteamwhoparticipatedinthisprojectfortheirexpertise,time,andresources.TheyalsothankthecommunitypartnersfromtheCityofGrandRapidsandtheWestMichiganShorelineRegionalDevelopmentCommissionwhoprovideddataandreal-worldapplicationstobuildandtestthecalculator.

25

LiteratureCitedAbate,D.,Towers,M.,Dotz,R.,Romani,L.,2009.WhitestoneFacilityMaintenanceandRepairCost

Reference.

Almansa,C.,Martínez-Paz,J.M.,2011.Whatweightshouldbeassignedtofutureenvironmentalimpacts?Aprobabilisticcostbenefitanalysisusingrecentadvancesondiscounting.Sci.TotalEnviron.409,1305–1314.doi:10.1016/j.scitotenv.2010.12.004

Barnhill,K.,Smardon,R.,2012.GainingGround:GreenInfrastructureAttitudesandPerceptionsfromStakeholdersinSyracuse,NewYork.Environ.Pract.14,6–16.doi:http://dx.doi.org.ezproxy.gvsu.edu/10.1017/S1466U611000470

Beauchamp,P.,Adamowski,J.,2012.DifferentMethodstoAssessGreenInfrastructureCostsandBenefitsinHousingDevelopmentProjects.J.Sustain.Dev.5,p2.doi:10.5539/jsd.v5n4p2

Bianchini,F.,Hewage,K.,2012.Probabilisticsocialcost-benefitanalysisforgreenroofs:Alifecycleapproach.Build.Environ.58,152–162.doi:10.1016/j.buildenv.2012.07.005

Carter,T.,Keeler,A.,2008.Life-cyclecost–benefitanalysisofextensivevegetatedroofsystems.J.Environ.Manage.87,350–363.doi:10.1016/j.jenvman.2007.01.024

CenterforNeighborhoodTechnology,2014.TheEconomicBenefitsofGreenInfrastructure:ACaseStudyofLancaster,PA.

CenterforNeighborhoodTechnology,2007.GreenValuesStormwaterToolbox[WWWDocument].Cent.NeighborhoodTechnol.URLhttp://www.cnt.org/tools/green-values-stormwater-toolbox(accessed7.31.15).

Chen,F.-Y.,Peng,I.-W.,Liang,J.-H.,Liang,Y.-Y.,2014.Greenpremiumingreencondobuildings?EvidenceinTaiwan,in:GreenBuilding,MaterialsandCivilEngineering.CRCPress,Leiden,TheNetherlands.

CityofAnnArbor,2015.StormwaterRatesandCredits[WWWDocument].URLhttp://www.a2gov.org/departments/systems-planning/water-resources/Stormwater/Pages/StormWaterRates.aspx(accessed7.1.15).

CityofGrandRapids,2014.StormwaterAssetManagementPlan[WWWDocument].URLhttp://grcity.us/enterprise-services/Environment-Services/Pages/Stormwater-Asset-Management-Plan.aspx(accessed6.12.15).

Clark,C.,Adriaens,P.,Talbot,F.B.,2008.GreenRoofValuation:AProbabilisticEconomicAnalysisofEnvironmentalBenefits.Environ.Sci.Technol.42,2155–2161.doi:10.1021/es0706652

Debo,T.N.,Reese,A.,2002.MunicipalStormwaterManagement,SecondEdition,2nded.CRCPress,Hoboken.

Freeman,A.M.,2003.TheMeasurementofEnvironmentalandResourceValues:TheoryandMethods.ResourcesfortheFuture.

Fuerst,F.,McAllister,P.,2011.Eco-labelingincommercialofficemarkets:DoLEEDandEnergyStarofficesobtainmultiplepremiums?Ecol.Econ.70,1220–1230.doi:10.1016/j.ecolecon.2011.01.026

26

Greenroofs.com,2015.TheGreenroof&GreenwallProjectsDatabase[WWWDocument].URLhttp://www.greenroofs.com/projects/(accessed7.7.15).

Hellman,K.,2011.AnEconomicapproachtotheecologicalissuesofurbanstormwaterrunoff:ACasestudyoftheAllenCreekWatershedinMonroeCounty,NewYork(Thesis).

HerreraEnvironmentalConsultants,2012.PugetSoundStormwaterBMPCostDatabase.WashingtonStateDepartmentofEcology,EnvironmentalAssessmentProgram,Olympia,Washington.

Houle,J.,Roseen,R.,Ballestero,T.,Puls,T.,Sherrard,J.,Jr.,2013.ComparisonofMaintenanceCost,LaborDemands,andSystemPerformanceforLIDandConventionalStormwaterManagement.J.Environ.Eng.139,932–938.doi:10.1061/(ASCE)EE.1943-7870.0000698

Ichihara,K.,Cohen,J.P.,2010.NewYorkCitypropertyvalues:whatistheimpactofgreenroofsonrentalpricing?Lett.Spat.Resour.Sci.4,21–30.doi:10.1007/s12076-010-0046-4

Isely,E.S.,2014.SustainingstormwaterinvestmentsinGrandRapids.WMEAC.

Isely,E.S.,Denning,R.,Isely,P.,2012.ApplyingtheINtegratedValuationofEcosystemServicesTool(INVEST)attheparcellevel:theOwasippeScoutReservationdemonstrationproject.FinalProjectReportfortheWestMichiganStrategicAllianceandTheNatureConservancy.WestMichiganStrategicAlliance,GrandRapids,Michigan.

Isely,E.S.,Isely,P.,Seedang,S.,Mulder,K.,Thompson,K.,Steinman,A.D.,2010a.AddressingtheinformationgapsassociatedwithvaluinggreeninfrastructureinWestMichigan:INtegratedValuationofEcosystemServicesTool(INVEST).J.Gt.LakesRes.36,448–457.doi:10.1016/j.jglr.2010.04.003

Isely,E.S.,Steinman,A.D.,Thompson,K.,VanderMolen,J.,Koches,J.,Sinha,S.,Huntington,L.,Isely,P.,Penning,T.,2010b.ReinintheRunoff:StormwaterIntegratedAssessmentinSpringLake[WWWDocument].URLhttp://www.gvsu.edu/wri/director/rein-in-the-runoff-stormwater-integrated-assessment-in-22.htm(accessed4.29.16).

Johnston,R.J.,Rolfe,J.,Rosenberger,R.S.,Brouwer,R.,2015.IntroductiontoBenefitTransferMethods,in:Johnston,R.J.,Rolfe,J.,Rosenberger,R.S.,Brouwer,R.(Eds.),BenefitTransferofEnvironmentalandResourceValues,TheEconomicsofNon-MarketGoodsandResources.SpringerNetherlands,pp.19–59.

Kuhns,T.,Ulasir,M.,2015.FirstFlushDesignforStormwater.

Landphair,H.C.,McFalls,J.A.,Thompson,D.,2000.Designmethods,selection,andcosteffectivenessofstormwaterqualitystructures(No.FHWA/TX-01/1837-1).TexasTransportationInstitute.

Lee,J.S.,Li,M.-H.,2009.Theimpactofdetentionbasindesignonresidentialpropertyvalue:CasestudiesusingGISinthehedonicpricemodeling.Landsc.UrbanPlan.89,7–16.doi:10.1016/j.landurbplan.2008.09.002

LoopNet,Inc.,2015.GrandRapids,MIMarketTrends[WWWDocument].URLhttp://www.loopnet.com/Grand-Rapids_Michigan_Market-Trends(accessed7.2.15).

McPherson,E.G.,Simpson,J.R.,Peper,P.J.,Maco,S.E.,Gardner,S.L.,Cozad,S.K.,Xiao,Q.,2006.Midwestcommunitytreeguide:benefits,costs,andstrategicplanting.

27

MinnesotaPollutionControlAgency,2015.MinnesotaStormwaterManual[WWWDocument].URLhttp://stormwater.pca.state.mn.us/index.php/Main_Page(accessed6.29.15).

Moeller,J.,Pomeroy,C.A.,2009.BMPandLIDWholeLifeCostModels:Version2.0[WWWDocument].URLhttp://www.werf.org/i/a/Ka/Search/ResearchProfile.aspx?ReportId=SW2R08(accessed6.22.15).

NYSDept.ofEnvironmentalConservation,2014.ConstructionStormwaterToolbox[WWWDocument].URLhttp://www.dec.ny.gov/chemical/8694.html(accessed6.3.15).

Polyakov,M.,Iftekhar,S.,Zhang,F.,Fogarty,J.,2015.Theamenityvalueofwatersensitiveurbaninfrastructures:Acasestudyonraingardens.

Reichardt,A.,2013.OperatingExpensesandtheRentPremiumofEnergyStarandLEEDCertifiedBuildingsintheCentralandEasternU.S.J.RealEstateFinanceEcon.49,413–433.doi:10.1007/s11146-013-9442-z

Sander,H.,Polasky,S.,Haight,R.G.,2010.Thevalueofurbantreecover:AhedonicpropertypricemodelinRamseyandDakotaCounties,Minnesota,USA.Ecol.Econ.69,1646–1656.doi:10.1016/j.ecolecon.2010.03.011

Sproul,J.,Wan,M.P.,Mandel,B.H.,Rosenfeld,A.H.,2014.Economiccomparisonofwhite,green,andblackflatroofsintheUnitedStates.EnergyBuild.71,20–27.doi:10.1016/j.enbuild.2013.11.058

TheHomeDepot,2015.RainBarrels-Watering&Irrigation[WWWDocument].URLhttp://www.homedepot.com/b/Outdoors-Garden-Center-Watering-Irrigation-Rain-Barrels/N-5yc1vZc7pb(accessed11.9.15).

Thorsnes,P.,2002.TheValueofaSuburbanForestPreserve:EstimatesfromSalesofVacantResidentialBuildingLots.LandEcon.78,426–441.doi:10.3368/le.78.3.426

WeatherHistory&DataArchive|WeatherUnderground[WWWDocument],n.d.URLhttp://www.wunderground.com/history/airport//2016/01/23/DailyHistory.html?req_city=Grand%20Rapids&req_state=MI&reqdb.zip=49512&reqdb.magic=5&reqdb.wmo=99999(accessed1.23.16).

Weimer,D.L.,Vining,A.,2010.PolicyAnalysis:ConceptsandPractice,5edition.ed.Routledge,Boston.

WSB&Associates,2008.APublicWorksPerspectiveontheCostvs.BenefitofVariousStormwaterManagementPractices.

Xian,G.,Homer,C.,Dewitz,J.,Fry,J.,Hossain,N.,Wickham,J.,2011.Thechangeofimpervioussurfaceareabetween2001and2006intheconterminousUnitedStates.Photogramm.Eng.RemoteSens.77,758–762.