ModelingPollutantLevelsinLakeOrono

EnvironmentalProblemSolving(ESCI310)CourseDescription:This course explores methods of solving real-world environmental problems. Theseproblems are interdisciplinary by nature and are rarely addressed in a substantivefashion in traditional science textbooks. In this course, students and faculty worktogether to developmodels of earth systems and learn how tomake calculations ofhuman-inducedchanges.Studentsworktogetheroninterdisciplinaryresearchprojectsusing modeling and systems analysis software to more fully understand specificenvironments and the quantitative methods of assessing challenges to thoseenvironments.

Project1:AnalyzingSpatialPatternsofWaterChemistryintheElkRiver AnthonyMandile&CassiClubbProject2:EffectsofRemovingPhosphorusInputsatLocalvs.UpstreamSites JennyWalz&EvanCrossProject3:EffectivenessofAlumTreatmentonWaterColumnPinLakeOrono Quinn&KatherineProject4:RoleofCurly-LeafPondweedinSeasonalPdynamicsinLakeOrono CourtneyPelissaro&EthanJustProject5:EffectivenessofSandFiltersinRemovingPfromLakeOronoTributaries EricUdermann,KatieDennis,&TylerSchmitt

3

Project1:LongitudinaltrendsofphosphorusintheElkRiver

AnthonyMandile&CassiClubbTheFederalCleanWaterAct(CWA)requirestheMinnesotaPollutionControlAgency(MPCA)toidentifywaterbodiesthatdonotmeetwaterqualitystandards.TheCityofElkRiverisaccountablefordecreasingPinputbasedontheTMDLbudgetcalculationsforLakeOrono.ThepurposeofthisanalysisistoanalyzelongitudinalandtemporalpatternsofwaterchemistryfortheElk,andtousethisanalysistomakeinferencesaboutpotentialpointsources/tributariesthatcontributesignificantamountsofP.ToassessthemovementofPhosphorusdownstreamfromElkLaketoLakeOronothroughElkRiver,wefirstlookedatwerethelakeandbiologicalmonitoringstationsthathadenoughrecentphosphorousdatatobeconsideredrelevant.Ninemonitoringstationswerechosenbasedonnumberofdatapointspresentandyeartheyweretaken.ThesepointswereaveragedandreportedastotalmgPperliter.DistanceofeachstationfromElkLakewasestimatedtoshowchangeinTPthefartheritmovedfromtheLake.Fig.1

4

Figure1describestherelationshipbetweendistancefromElkLaketoLakeOrono,throughthetributaryElkRiver.Fig.2

Figure2showstheseasonalchangesovertimeinphosphorusconcentrationatsiteS003-686in2009.Fig3.

Figure3showsseasonalchangeinphosphorusconcentrationatsiteS001-526in2009.

5

Fig.4.

Fig.5

Thegraphsconductivity(Fig.4)anddissolvedoxygen(Fig.5)showmeasurementstakenatfourgagestationsalongtheElkRiver-betweenElklakeandlakeOrono.SiteMO84ismostupstreamoflakeOrono,followedbysite’sMO14,MO38,andMO17,successively.

Conclusions-advice

SolvingtheLakeOronophosphorousproblemwillbeneithereasynorinexpensive.TheElkRiverisasignificantcontributorofphosphorusinflowintoLakeOrono.TheElkRIveristhemajoroutflowofElkLakeandemptiesintoLakeOrono.TherearewatershedsdirectingphosphorousandothernutrientleachateintotheElk

6

River.TheagriculturallydenselandareasadjacenttotheElkRiverarelargelyresponsibleformuchofthephosphoruspollutioninlakeOrono.

Itispossibletoarguefractionalpaymentfromtheresponsiblecountiesforthispollution.ThemapshowsthechangeinphosphorusconcentrationatdifferentpointsalongtheElkRiver.ThedataweusedwascollectedfromgagestationsalongtheElkRiver.Weusedmeanannualvaluesforphosphorusconcentrationfrom2009.Wesuggestyoucouldusethisdataalongsidemicro-watersheddatatodelineateexactlythedirectionsinwhichphosphorousisflowingfromthesurroundinglandintotheElkRiver.Ifyouknowthevolumeofwatermovingthroughaparticularsectionoftheriverbetweenthegagingstations(oranytwopoints)aswellastheconcentrationchangebetweenthosezones,youcanroughlycalculatethequantityofphosphorusenteringElkRiverbetweeneachstationandeachperpetratorsfractionalcontributiontolakeOronospollution. Wesuggestyouraiseawarenessanddirectattentiontotheresponsibleagriculturalbusinesseswithconviction.Showthemthedata,showthematopographicalmapdescribingthedirectionofwaterflowinthearea,showthemthechangeinphosphorusconcentrationintheElkRiveroneitherendoftheirproperties-frompointAtopointB.ExplainthenegativeeffectsthenutrientpollutioniscausingtoLakeOronoandtheElkrivercommunity-andthesteeppricetosavelakeOrono-whoseburdenislefttothetaxpayersofElkRiver.Thiswillundoubtedlyraiseawarenessofthesituationandhopefullyencouragebetterdecisionmakinginthefuture.

7

Project2:

EffectsofRemovingPhosphorusInputsatLocalvs.UpstreamSites

JennyWalz&EvanCross

ThepurposeofourmodelistohighlighttheSherburneCountyphosphoruscontributioninLakeOronorelativetoupstreamsources.Thegoalwasforourmodeltoportrayhowinfluentialphosphorusremovalfromlocalsourceswouldbecomparedtotheremoval,or reduction, of phosphorus from these upstream sources. Our model is useful topotentialusersbecauseitillustrateswherepossibleinterventionsshouldbemade,andforcestheviewertotakeastepbackandconsiderwhereitwouldbemostbeneficialtoimplementaphosphorusremovalplan,whetherthatbeatthelocalorwidespreadlevel.Phosphorus removalwithin theWatershedDistrictmayunderstandablyhelp reducePlevels in Lake Orono, however, if there are alternate sources of phosphorus beingcontinuously poured in upstream, the local solution may only provide a temporarybandaid.Toexplorehoweffectivephosphorusremovalplanswouldbeateither level,wecreatedthismodelthatdemonstratescurrentphosphoruslevelswithinLakeOrono,and factors thatmaybecontributing toan increaseordecreaseofphosphorus in thelake.

OurconceptualmodelportraysthedifferentwaysinwhichphosphorusisbroughtintoLake Orono, and how quickly that phosphorus leaves. Our currency throughout thesystemisphosphorus.Ourtwostocks inthesystemareconcentrationsofphosphorusrepresented in grams per liter. Flows of our system include upstream phosphoruscoming in from Elk River upstream, local phosphorus inputs coming from withinSherburneCounty,miscellaneoussourcesofphosphoruscoming fromotherupstreamwater sources like Tibbett’s Brook and Big Lake, and finally, phosphorus leaving thesystemeitherthroughrunofforsedimentation.

Anumberofcontrolswereimplementedtomakeourdatamoreaccurate.Weobtaineddata from the TMDL document to implement discharge rates of water andconcentrations of phosphorus coming from upstream sources like the upstream andmiscellaneous inflows.Datawas also collected from Sherburne county to incorporatedailyprecipitation intoour systemsince the rainfall collects localizedphosphorusandbrings it into LakeOrono.We did not include snowfall in our precipitation converterbecause only a small amount of phosphoruswould likely go into LakeOrono directlyfrom snow. Likewise, the data would have to be manipulated in order to correctly

8

identifyhowmuchwatercontained in thesnowmelts,andof thatmeltedsnow,howmuch phosphorus it bringswith it. Sedimentation andmineralization datawere usedfromapreviousComoLakemodel,and lakevolume,area, turnover rate,andstartingconcentrationsofphosphoruswerecollectedusingboththeTMDLandDNRdocuments.

Wedecidedtorunourmodelusingdataobtainedfromthesummer,sinceitcontainedthemost activity in terms of discharge rates, phosphorus levels, and turnover rates.Sincetherewerefewdiscrepanciesforeachsummerovertheyears,wedecidedtorunourmodelforonesummer(2003),atotalof183days.Resultsareshownbelow.

Figure 1:

9

Results:

ModeldataofLakeOronowascomparedtobothUpperandLowerLakedata,asshowninFigure2.ThemodelestimatesapproximatelythesameamountofTPinLakeOronoasisobserved.Therefore,themodelisanaccuratetoolthatcanbeusedtoestimatetheeffectsofvariouschangesonLakeOronoTP.

Figure2:

Figure2 shows the total phosphorus in LakeOronoas estimatedby themodel (blue)comparedtoLowerOronodata(red)andUpperOrono(green).ThemodelTPisaboutthesameasbothLowerandUpperOronoTP.

To test thedifferent effectsphosphorus removal from local orupstream inputs couldhave on Lake Orono TP, we removed 1,728 g/day. We wanted to see what would

10

happen to Lake Orono TP if we removed half of the local input. In order to remainconsistent,weremovedthesameamountofPfromlocalinputorupstreaminput.Thisnumberwasdeterminedusingthefollowingcalculations:

LocalPinput=1.3million

LocalPinput*0.5=671,000

TotalPblocked=LocalPinput-LocalPinput*0.5=629,000

Pblocked/day=TotalPblocked/365=1728g/day

We then removed1728 g/dayof phosphorus fromeither local or upstream inputs todeterminewherePremovalcouldhavethelargesteffect.Figure3showstheeffectsofthesechangesontheTPinLakeOrono.ChangingPinputonalocallevelseemstohaveagreatereffectonLakeTPthanupstreamlevels.

Figure3:

Figure3comparesthenormalTPlevelsinLakeOrono(blue)comparedtoadecreaseinlocalPinput(red)andupstreaminput(green).Thoughdecreasingupstreaminputdoes

11

seem to result in slightly lower TP levels, decreasing local input completely removesseasonalTPspikes.

Our model indicates that decreasing local P input could be more effective thandecreasing upstream input. Local P input is almost all runoff, and so therefore it isheavily affected by rainfall. On the other hand, in our model, upstream P inputmaintains a relatively constant and high level throughout the season, no matter therainfall levels. Therefore, ourmodel suggests that upstream input would need to bedecreasedbyamuchhigherlevelthanlocalinputtoachievesimilarLakeOronoTP.

Basedoffoftheseresults,webelievethatdecreasingPinputwithinElkRivercouldhaveadisproportionatelyhigheffectontheLakeOronoTP.SmallchangeswithinElkRivercouldresultinmuchhigherdecreasesofLakeOronoTP.Theupstreaminputaccountsforapproximately94%oftheLakeOronoTP,andsotoeffectivelycutmassiveamountsofPfromLakeOrono,largechangesneedtobemadeupstreaminadditiontolocalareas.

12

Project3:TheEffectivenessofAlumasaTreatmentforExcessPhosphorusinLakeOrono

QuinnWhitingandKatherineConnelly

Aluminum sulfate (Alum) has been used to reduce the total amount of

phosphorusinshallowlakes,whichalsoimprovestheclarityofthelakes.Itdoesthisbybindingwiththesolublephosphate,makingitinsolubleandthensettlestothebottomofthelake.ThistreatmenthasbeenusedbeforeintheMinneapolisChainofLakesandwaseffectiveatimprovinglakeclarityandreductionofinternalphosphorusreleaseby85% (Huser, et al.). LakeOrono is classified as a shallow lake, however, it has a verylargethroughput,whichmakesitdifficulttocomparetotheseotherlakes.

TohelpusseehowanAlumtreatmentwouldaffectthetotalphosphorusinLakeOrono, we constructed a model of the lake using the program STELLA. This modelincludes stocks, flows, and converters, which calculate the movement of totalphosphorus in pounds into and out of the lake. The purpose of our model is todetermine whether or not Alum is a good treatment option with dealing with theoverloadofPhosphorusinLakeOrono.

We want our model to act as a calculator that allows you to put in variousamountsofalumandtobeabletomakeagraphtoseewhenthealumrunsout,aswellas the effectiveness of the treatment. By using a price per unit converter, it candetermine the cost of the treatment,whichwould helpwith determiningwhether ornotthisisarealisticoptioneconomically.

Ourmodelisrobust,bywhichitcanbeappliedtootherlakesandwatersystemsandnotjustLakeOrono.Onecansimplychangetheflowsofwatercominginandout,thevolumeofthelake,andtheinitialamountofphosphorusinthelakeaswellastheinflowandoutflowrateofphosphorus.Itcouldalsobehelpfulintheplanningoffuturetreatmentaswellascalculatingthecostofthetreatments.

Below is a description of ourmodel in whichwe go through and define eachstock,flow,andconverteraswellasjustifytheequationsandinitialvaluesweused.Wealsoincludemultiplescenariossuchasdecreasingthethroughput,addingtheadditionofvariousamountsofalum,aswellaschangingthewashawayrateofthealuminordertoseewhattheeffectswouldbe.

13

Figure1.ImageofourSTELLAmodelwhichrepresentsLakeOronoandcanbeusedtocalculatethetotalamountofPhosphorusintheLakebeforeandafterAlumtreatment.DescriptionofModel:Currency-(Whatthestocksmeasure)LakeTPStock:Thetotalunreactedphosphorusinthelakeinpounds(lbs)AlumStock:TheamountofAluminpounds(lbs)Stocks:(aquantityofmaterialorinformationthathasbuiltupovertime)Totallakephosphorous(LakeTP):initiallyat437lbs(198.2kg),thiswascalculatedfromthe TMDL document, which found the concentration to be 115 ug/L, and using unitconversions to lbs/m^3andmultiplyingby the lakevolume inm^3,yields the initialPamountinthelake.Amount of alum (ALUM): initially at 190000 lbs (86182.5 kg), thiswas used becauseusing thisamountyields50mg/L,which is themaximumconcentration thatdoesnotnegativelyaffectfish.Flows(flowsofphosphorus/alumgoingintooroutofastock)Pin:theamountofPcomingintothelakeeveryday,sameasRunoffTP,(lbs/day).P-out: the amount of P flowing out of the lake every day. This is a function of theoutflow of water and the concentration of total P, (outflow*(lbs/m^3)) which yields(lbs/day)unitsAlum TP: this is the amount of P that the Alum “soaks up” and is a function of theamount of unreacted Alum and the amount of P in the lake. The equation is

14

Lake_TP*(Kmax*(ALUM/(ALUM+Kss))), this makes it so all of the alum doesn’t reactinstantaneously.Washaway: this is theamountof theunreactedalumthatwill getwashedoutof thelakeduetothehighthroughputofthelake.Rightnowwearesayingthat2.5%ofthetotalalumisgettingwashedawaypereverythroughputofthelake.Theequationusedis(.025*ALUM)/ThruPutAlumuptake:thisistheamountofAlumthatisbeingreactedwiththePandisfoundtobeALUM_TP/STOIC,becauseforevery1molofalumitreactswith2molsofP.Converters:(affectstheflowsandstocksofthesystem)RunoffTP:This isthetotalphosphorusamountfromtheTMDLpaperwhich included:inflowfromdrainageareas,pointsourcedischarges,fallingsepticsystems,inflowfromupstreamlakes,atmosphere,groundwater,andinternal.Thevalueinthepaperwasinpoundsperyearwhichwethendividedby365togetpoundsperday.Thisisaconstantvalue.Lake Volume: the total volume of lake Orono. Found value in TMDL paper andconvertedtocubicmeters.Volume:1,732,788cubicmeters.Thisisaconstantvalue.LakeTPlbsm^3:Thisistheconcentrationoftotalphosphorousinthelakebasedonthelakevolumeinpoundspercubicmeters.Outflow:TheoutflowvaluewasfoundintheTMDLpaperandlabeledwiththevariableQ.Weconvertedthatvaluetocubicmetersperdayandgot985479cubicmeters/day.Thisisaconstantvalueinourmodel.Turnover Rate: This value was found in the TMDL paper labeled with the variable Twhichwasthelakevolumedividedbythelakeoutflow.Theturnoverrateisafunctionoflakevolumeandoutflowwhichisequalto3.65days.Kmax:This isthemaximumrateofuptakeofphosphorusbythealum.Thisvaluewasobtainedfromanonlinesource(LennTech)whichsaidthatalumtakes80%-90%ofthetotalphosphorussoweused80%forourmodel.Ks:Athalf themaximumrate, this is thevalueof theremainingamountofunreactedalum,wesetthisat76000lbs.Alumconcentration:Theconcentrationoflbsofaluminthelakebasedonthevolumeincubicmetersofthelake.STOIC:Thisistheratioofmolesofalumtomolesofphosphorous.Foreveryonemoleofalum,twomolesofphosphorusaretakenup.

Al2(SO4)3·14H2O+2H3PO4↔2AlPO4+3H2SO4+18H2O

ResultsandConclusion:

15

Figure2:ShowsthetotalamountofAluminlbs(Blue)inthelakeovertime.TheamountofPthattheAlumreactswith(ALUMTP,RED).AndtheamountofPinthelake(Pink).Theadditionof the alum to the lakedecreased the total P to209 lbsor 54ug/L andslowlyincreasedastheAlumwentawayto475lbsor124ug/L.ThemaxconcentrationofPinashallowlakeis60ug/L.Thisisachievedinthefirstdayofthemodel,howeveritgrows to past that concentration very quickly and eventually leveled off at arounddouble the maximum desired concentration. Total cost of this treatment would beabout$256,500.

16

Figure3:Thisscenarioshowswhat theLakeTPwouldbe ifnoneof thealumwashedawayandstayedinthelake.Thismaybethecasebecausedamsactassedimentbarriersandpreventsedimentfromflowingdownstream.Thetotalcostofthistreatmentwouldbe$256,500.

17

Figure4:TopgraphshowstheLakeTP(InPink)ofLakeOronoifyouadded260,000lbsofAlumondayoneand70 lbsofalumeverydayafter that, thismakes it so that theconcentrationofphosphorous iswithintheparameters forashallow lake.Thecostofthiswouldbe$(351,000+94.5/day).Thiswouldalsomakeitsothattheamountofalumisabove the recommendedconcentration.Thebottomgraphshows the flowofAlumintothelake.

18

Figure5:This showswhathappenswhenwedecrease the flowrate through the lake to10000m3/day.ThetotalPlowersto4.8lbswhichisfarlowerthanthe60ug/Lmaximum.Soifthis lake had a slower inflow and outflow, then alum may be a good treatment,however,this isnotthecaseandsoalumisnotagoodtreatmentforlakeOrono.Thetotalcostofthistreatmentwouldbe$256,500.Conclusion: Given the model outputs we realize that one of the limiting factors in alumtreatment of Lake Orono is the throughput of water and phosphorous. This eithermakes it that thealum isbeingwasheddownstreambefore it reactsor that the totalamountreactsquickerduetotheconstantinputoffreshphosphorous.Webelievethattreatmentwithalumwouldnotbecostbeneficialduetothefactthatthealumeithercompletely reactsorwashesawaywithin a shortamountof timewhencompared toothershallowlakeswithlessthroughput.Summarizedpoints:● Based on the various scenarios we explored using our model, we would not

suggestusingthealumtreatmentmethodinlakeOrono.● Togetunderthe60ug/Lmarkwithalum,youwouldneedtoaddmorethanthe

environmentallyrecommendedamount.● Evenifthealumdidnotwashawayandstayedinthelake,thephosphoruslevels

wouldstillbeabovethesafelevelsforashallowlake.● 40 x 50lbs pallet for 2700$, (40*50)= 2000lbs, 190000/2000=95 pallets,

95*2700= $256500 for one 190000 lbs treatment. (from Factory DirectChemicals)

19

Project4:

RoleofCurly-LeafPondweedinSeasonalPdynamicsinLakeOrono

CourtneyPelissaro&EthanJustIntroduction

Curly-leaf pondweed is an invasive specie that has spread its way throughMinnesota.Itproducesathicklayerofmassatthesurfaceofthewaterinterferingwithrecreational lake use. The invasive plant overpowers and kills many native plantsultimatelytakingoverlakes.Theninmidsummertheplantdiesoffandwashestotheshore. Theseplants have causedmanyproblems to the lakes inMinnesota.After themid-summer die-off the plant reduces water clarity and often creates algae blooms,creating unpleasant swimming conditions. Curly-leaf pondweed has been a majorconcernfor lakesthatbenefitfromhavingcleanwater.ForourprojectwefocusedonLakeOrono,ashallow,highphosphorousyieldinglakeinMinnesotathathasaproblemwithCurly-leafpondweed.Lakedrawdownsareknownforhavingasignificanteffectonloweringplantlifeinlakes,butthereisalackofknowledgeonhowdrawdownsimpacttothetotalamountofphosphorusinalake.Inthismodelwedesignedittorepresentthe effects Curly-leaf pondweed has on the phosphorous levels in Lake Orono and ifkillingoffCurly-leafpondweedcouldbeasolutiontothelakeshighphosphorouslevels.Oncewecreatedaworkingmodelof the lake,wedesignedawaytorepresenta lakedrawdown in the fall that would kill some of the pondweed in hopes to lowerphosphorouslevelsinthelake.Wewentthisroutebecauseweknowadrawdownhasbeentalkedaboutandthelakehasbeendrawndown4.5feetinthepast.

Lake drawdowns can be a successful way to manage aquatic plant problems.SpecificallyforLakeOrono,wewanttosimulatealakedrawdownduringthewinter,atimewheretheseedsandrootsoftheCurly-leafpondweedcanbeexposedtothecoldwinterweather.Thiswillkilltheminhopethattheywouldnotcomebackasstrongthenextyear.WereadaboutLakeOronodoingasuccessful lakedrawdownof4.5feet in2003 andnot done since.Wedidn't receive anydataon the2003drawdownbutweassumeasuccessfuldrawdownwouldresult inasignificantimpactinaquaticplantsinthelake.

Parameters

We found that the total amount of Phosphorus in LakeOronowas 211,596 g.NextwefoundthedailyinflowofwaterfromElkRiveroverafouryearperiodfromthe

20

USGSwebsiteandplugged thesepoints intoourmodel. Wealso found the inflowofwaterfromothersourcesbytakingthetotalamountofwaterthatentersintothelakeperyearandsubtractingtheamountofwaterfromElkRiverenteringthelakeperyear.To find the yearly amount of inflow fromElk River,we added a year’sworthof dailyinflow. The information was found athttps://www.pca.state.mn.us/sites/default/files/wq-iw8-46e.pdf;p.106.

Next, we calculated how much phosphorus was entering the lake each day.Usingthesamewebsitesource,wefoundthatthenetdischargeofwateris3.59x1011liters/yearand theamountofphosphorous is 44,690,642grams/year.We then foundthe average amount of phosphorus entering the lake per liter to be .0001244g ofphosphorusperliter.WemultipliedthewaterinflowbythephosphorusconcentrationtofindthePinflowintothelake.

Weassumedthatthewater inflowandwateroutflowrateswerethesame.Tofindhowmuchphosphoruswasleavingthelake,wemultipliedtheconcentrationofthelakeby thewateroutflow.For thecurly-leafpondweed,weestimated that therewas200,000 pounds of wet mass or 20,000 pounds dry mass pondweed in Lake Oronoduringthesummer.Wealsoestimatedthat.25%ofthedrymassisphosphorus(Snow&Mason, 1979). This is equivalent to 22,680 grams of phosphorus in the pondweedduringthesummer.WeusedthisvalueasthemaximumgramsofPthatthepondweedholdcouldreachandadjustedthetakeupflowtoaccountforthismaximum.

We ignoredthesedimentbecauseourmodel isconcernedaboutthedie-offofpondweedinthelatesummerandwhateffectthathasonthelakestotalphosphorus.Weassumedthesedimentationwouldbeatasteadystatewiththewatercolumnandthenetfluxwouldbeequalto0.

21

ModelDesignandOutput

Firstwesetupourmodelwithalakephosphorusstockindicatingtheamountofphosphorus in the lake.Weaddedasimple inflowandoutflowtothestockwhichwecould then add the concentration of P coming into and flowing out of the lake. TorepresentthepondweedgrowingandtheirPhosphoroustakeupduringtheirgrowthwemadeastocknamedPondweedPhold.InthisstockphosphorouswillbetakenfromtheLakePstockandstoredinthePondweedPholdduringthespringandsummer.Whenthe pondweed dies during late summer the phosphorous hold is released and thephosphorousflowsoutofthePholdandintothelake.

During the spring and summer, the pondweed can take up P, but after thesummerdieoffthepondweedcannottakeupanyPuntilspring.Tomodelthistakeup,we added a converter that made the takeup inflow either multiplied by 1 or 0 torepresent spring/summer and fall/winter, respectively. In other words, the takeupequaled0duringthefall/winter.

Sincewehadwater inflowdatafor4yearswewereabletorunourmodel forthatamountoftime.StartingourmodelJanuary1st2013,weranituntilDecember10,2016.Inthesummer,wehadtomodelthesummerdieoff.Duringthistime,allofthepondweedPisreleasedfromtheholdandbecomespartofthetotallakeP.Duringtherestoftheyearthereisnorelease.Tosimulatethisweaddedaconverterthatmadethe

22

releasebe0atalltimeswhenitwasnotthesummerdieoffperiod.Forthesummerdieoffperiod,wehadthereleasehappenover20days. Thefirstday5%oftheholdwasreleased,theneverydaythereleaseratewasincreasedby5%untilday20when100%ofthePwasreleased.Figure1showsthepondweedPoverthe4yearperiodwithnoadditionalmanagementplanbesidestheregularsummerpondweeddie-offcycle,whichis technically amanagementplan for the lake. The suddendrops in thepondweedPrepresentsthesummerdieoff.Figure2showsthetotal lakePoverthe4yearperiodwithnoadditionalmanagementplan.Figure1

Figure1.PondweedPafternoadditionalmanagementplanFigure2

Figure2.LakePwithnoadditionalmanagementplan

Forourmanagementplan,wewantedtomodelawinterdrawdownthatwouldresultinkillingoffalargeamountofthepondweed.Tosimulatetheeffectsofawinterdrawdown,wedecidedtodecreasetheamountofPtakeupsincewewouldbekilling

23

offthepondweedandtherefore,theabilityforpondweedtotakeupP.Tomodelthis,weaddedanother converter thatwas similar to thenormal takeup function. For thefirst year, we had the pondweed take up the normal amount of P during thespring/summer. Thenext spring/summer, thepondweedwasonly allowed to takeup50%ofthePtosimulatethat50%ofthepondweedwaskilledoffthepreviouswinterduringthedrawdown.Thenextspring,wereducedthetakeupto25%.Thenthenextspring,wereducedthetakeupto12.5%.Eachyearwewantedtosimulatethattherewas less and less pondweed. Figure 3 illustrates the pondweed P during thismanagement plan and Figure 4 illustrates the lake P during thismanagement plan. Ifyou compare Figure 1 and Figure 3, it can be observed the there is a reduction inpondweed P. On the other hand, if you compare Figure 2 and Figure 4, it can beobserved that there is almost no difference between the levels of total lake P. Theimplicationhereisthatthemanagementplanwassuccessfulatreducingpondweed,buthadlittleimpactonthetotallakeP.Figure3

Figure3:PondweedPafter50%killoffmanagementplanFigure4

Figure4:LakePafter50%killoffmanagementplan

24

Wealsowantedtomodelwhathappenedif75%ofthepondweedwaskilledoff

ofthesecondwinter(insteadofonly50%).Soweonlyletthepondweedtakeup25%ofP that year, then12.5% thenext year, and6.25% the last year. Figure5andFigure6shows the pondweed P and total lake P, respectively, after this management plan.Figure5illustratesalargereductioninpondweedP,butagainthereis littleimpactontotallakeP.Further,Figure7graphsthelakePatallmanagementstrategies.Itcanbeobserved that the lines are completely overlapping, meaning there is almost nodifferencebetweenthestrategies.ThisimpliesthatnoadditionalmanagementisjustassuccessfulasthekilloffmanagementplansinrelationtototalLakePlevels.WedonotseeadifferencebecausetheamountofPthatisbeingreleasedbythepondweedissominimalcomparedtotheamountofPflowinginandoutofthelakeeveryday.Figure5

Figure5.PondweedPafter75%killoffmanagementplanFigure6

25

Figure6.LakePafter75%killoffmanagementplanFigure7

Figure7.ComparisonsoftheLakePwithnoadditionalmanagementandthetwolevelsofkilloffafterwinterdrawdown.Note thatall threemanagementplansaregraphed,butcannotbeseenduetocompleteoverlap.

LakeOrono is not a standard lakebecause it has sucha fast turnover rate. Amore normal lake will have lower rates of water entering and leaving the lake. WewantedtomodelwhatwouldhappenifLakeOronohadamorestandardturnoverrate.

26

Tosimulatethis,wemultipliedthewater inflowandwater flowby10%to illustratealakethathasonly10%oftheturnoverratethat.Figure8 illustratesthedifferences inlakePforthenormalLakeOronoturnoverratecomparedtoLakeOronoturnoverrateifitwas10%ofwhatitisnow.EssentiallyweseethattheLakePisreducedwithalowerturnover rate, which makes sense since there is less P entering the lake. Next, wewantedtoseeifkillingoffthepondweedwouldmakemoreofadifferenceinastandardlake.Figure9showstheLakePshowsthepondweedPwithnoadditionalmanagementplanwhen at the reducing turnover rate and Figure 10 shows the respective Lake P.Thenwecomparedthesefindingstoifwedidthe50%killoffmanagementplan.Figure11andFigure12showthepondweedPandlakeP,respectively, forthismanagementsituation.Whencompared,thereisalmostnodifferenceintheLakePwithorwithoutthe additionalmanagement plan, evenwhen the turnover rate is reduced. Figure 13comparestheLakePforallmanagementscenarioswhentheturnoverrateisreducedto10%.Figure8

Figure8.ComparisonoflakePwithnormalandreducedturnoverrate.

27

Figure9

Figure 9. Lake P after no additional management plan when there is 10% of theinflow/outflow

Figure10Figure10.Total lakePafternoadditionalmanagementplanwhenthere is10%oftheinflow/outflowFigure11

28

Figure11.PondweedPafter50%killoffmanagementplanwhenthere is10% of theinflow/outflowFigure12

Figure12.Total lakePafter50%kill offmanagementplanwhen there is10% of theinflow/outflowFigure13

29

Figure13.ComparisonoftotallakePwithdifferentmanagementstrategieswhenthereis10%oftheinflow/outflow

Lastly,wewantedtoillustrateifthelakenotonlyhad10%theturnoverrate,butalso 10% the P coming in. Figure 14 and 15 show the pondweed P and lake P,respectively,forthissituationwhenthe50%killoffmanagementplanisapplied.Figure14 shows that there is a heightened level of pondweed P when compared to thepreviouspondweedPfigures.Forsimplicity’ssakewekeptthebaselinetotallakePthesame,sothis iswhythere isabigdrop in thebeginningofFigure15. Figure15doesshowageneraltrendofdecreasingtotallakeP,butitisunclearifitisafunctionofthepondweedremoval.

30

Figure14

Figure14.PondweedPafter50%kill offmanagementplanwhen there is 10%of theinflow/outflowand10%oftheconcentrationofPflowingin.Figure15

Figure 15. Lake P after 50% kill off management plan when there is 10% of theinflow/outflowand10%oftheconcentrationofPflowingin.Conclusions

WhenusingourmodelwefoundthattherewasnotenoughphosphorousbeingheldintheCurly-leafpondweedtomakeasignificantimpactonphosphorouslevelsinLakeOrono.Thisisduelargelytothefactthatthereissuchashortturnoverrateinthelake.When testingourmodelwith a lakewith an Inflowandoutflowof 10%of Lake

31

OronoandwithaPconcentration inflowof10%ofLakeOrono(Figure14and15)wefound that killing off pondweed does lower the amount of phosphorous in the lake.Althoughkillingoffpondweedmightnothaveasignificanteffect inLakeOrono, itwillstillslowthegrowthandspreadoftheweed.Doingalakedrawdownwillcleanupthelakebyreducingtherepopulationoftheweedafterbeingexposedtothewinterair.Thephosphorous levels in Lake Orono are due to upstream problems so the removal ofcurly-leafpondweedwillnotchangethephosphorouslevelsinthelake.TheCityofElkRiver is interestedin loweringphosphorus levels inanattempttocleanupthelakesothatLakeOronowillbemoreappealing for recreationaluseandresidentialuse.Afterrunning our model we have concluded that a lake drawdown will not lowerphosphorous levels. However, if a drawdown does have a significant impact on theamountofpondweedinthelakethenTheCityofElkRivermightbeinterestinginusingadrawdowntocleanthelake.

References

Snow,P.D.,&Mason,R.P.(1979).Monitoringofhydraulicdredgingforlakerestoration Lakerestoration:Proceedingsofanationalconference(pp.195-204). Washington,D.C.:OfficeofWaterPlanningandStandards.

32

Project5:

EffectivenessofSandFiltersinRemovingPfromLakeOronoTributaries

EricUdermannm,KatieDennis,TylerSchmitt As a result of runoff from fields using fertilizer containing Phosphorus (P), theconcentrationofPinsurroundingriversandlakesishigh.HighPconcentrationleadstoalgaeblooms,whichhavenegativeeffectsonlakehabitatsaswellasrecreationalareas.Anareaofsignificanceforourproject,theElkRiver,eventuallyrunsintotheMississippiRiver, whichmeans that P is also entering the river that stretches across the UnitedStates.A largepercentageof theP in theElkRiverandLakeOrono is fromupstream,outsideofSherburneCounty.IflittlecanbedonetoreducethePinputupstream,thentheproblemmustbeaddressedinSherburneCounty. Onemanagementmethod is theuseofan iron-enrichedsand filter. The filterconsists of aggregate sands and Iron filings. The sand catches suspended solids andparticulateP,whiletheIronchemicallyreactswiththedissolvedPandremovesitfromthewater. Onaverage,afilter lastsabout30yearsbeforetheIronmustbereplaced.The effective working time of the filter depends on water flow rate and Pconcentrations. OneofLakeOrono’stributariesistheOronoinlet.Theinletdrains3,410acresofland into LakeOrono. TheOrono inlet is the location chosen to implement the iron-enriched sand filter for a number of reasons. First, the inlet runs into a stormwaterpondbeforeitentersLakeOrono.ThestormwaterpondallowssomeoftheparticulatePtosettlebeforeitentersthelake.TheOronoinletalsohasasignificantlyhigherwaterflowratethantheElkRiver.Itwouldbeeasiertoputafilterintoanareawithalowerflowratethanintoarushingriver.Also,afilterbigenoughtospantheElkRiverwouldcosta substantialamountofmoneyand thenwouldneed tobe replacedbeforea30yearperiodbecauseofthehighwaterflowrate. Thepurposeofourmodelsistoshowtheeffectofdifferentsizediron-enrichedsand filters on Lake Orono by implementing them on the Lake Orono inlet.Model 1showstheeffectivenessofanIron-enrichedsandfilteranditscapacitytoabsorbPovera30yeartimeperiod.Model2usesefficiencyvaluesfromModel1andshowstheLakeOronoPoverasinglesummer.Together,theresultsofthesetwomodelsallowsforacomparison of costs/benefits of different sized Iron-enriched sand filters. ThecomparisonswillbeusedbytheCityofElkrivertodetermineiftheIron-enrichedsandfilterisapracticalmanagementtechniqueforhighconcentrationsofP.Model#1

Formodel#1wesimulatedusingan iron-enrichedsand filterand lookedat itsefficiency(thecapacitytoabsorbP)overa30yearperiodintheLakeOronoinlet.Weused this information to model seasonal total phosphorus (TP) and P concentration

33

scenarios forLakeOrono (model#2).Thestock inmodel#1 is shown inFigure1.Thestock is the iron filter’s capturedphosphorous ingrams.Thecontrolson this flowarethe rate of discharge and concentrationof P coming into the filter. At the topof themodelwehaveacontrolforthesizeoffilterdesired,whichthencalculatesthecapacityin the control called takeout. We calculated the correlation between filter size andtakeout capacitywith the data found in the LakeOrono Stormwater Retrofit Analysisdocument.EachsizefilterinthedocumentgaveacorrespondingaveragetakeoutvalueinpoundsofPperyear.Onaveragethefilterslastfor30years,sowemultipliedeachofthe average takeouts by 30 to show the overall capacity. Each size filter had acorresponding capacity, so we graphed capacity with filter size as the dependentvariable.

OurinflowstothestockaredissolvedPandsuspendedPingrams/L.Wesplitupthe inflowof phosphorusbetweendissolvedP andparticle P because the filter has acapacityofdissolvedPitcanhold,yettheparticlePcanstillbecaughtbythesand.Ouroutflowisthecombinedamountofdissolvedandparticulatephosphorouscomingoutof the filter. We set this upwith an outflow (dissolved P * Dissolved P efficiency) +(ParticleP*ParticlePefficiency). Controlsonbothinflows(dissolvedandparticulate)aretheconcentrationofphosphoruscominginingrams/LandtherateofdischargeinL/year. Asthestockofourfilterfillsup,wecancalculatethesaturationinthefilterusingtheequation(filter(g)/takeout(g)).Thisiscalculatedinthecontrolcalledpercentfiltersaturation.As the filter fillsup the saturation reaches100%and theamountofPnotbeing caught in the filter increases.We demonstrate this by the Dissolved efficiencycontrol increasingexponentiallyup to100%as thesaturationapproaches1.Once thedissolvedefficiencyreaches100%theoutflowofdissolvedPisthesameastheinflowofdissolvedP.Essentiallywhatgoesiniscomingstraightoutafterthesaturationreaches1.However,oncethesaturationreaches1,someoftheparticulatePstillgetsthrough.WesetthisupwithacontrolcalledParticlePefficiency.AsthesaturationgetshighertheefficiencyofParticlePgetshigheruntiltaperingoffat90%.Thisrepresents10%oftheparticulatePthatstillgetsthroughthefilterafteritistechnically“filled.”Sincetheefficienciesareontheoutflows,thehigherthe“efficiency”thelessefficientthefilteris.Sowhen saturation is high and the efficiencies are high the filter is not collecting asmuchP.Thischangeswiththesizeofeachfilter.ThedatathatwillbeusedfortheLakeOronomodel(model#2)andcomesfromtheoverallefficiencyconverter.Wefindthisbydividingthemodel’soutflowbyitsinflow,givenapercentageoftotalPcomingoutofthefilter.

34

Figure1:Iron-enrichedSandFilterStellaModelResultsofModel#1

Graph 1 below illustrates the total P in the filter over a 30 year period. Thebeginningyearof the filterare themostefficient inabsorbingPas seenby thesteepslope.OncetheIroncannotholdanymoreP,thesandcontinuestofilterthewaterandcollecttheparticleP.ThetotalPtakenoutfromthefilter(model#1)isthedatathatisusedinmodel#2toexaminetheeffectsonaseason.ResultsofthetotalPintakeofthefilter alignwith estimatesmade in the lakeOrono Stormwater Retrofit Analysis. TheestimatesgivetotalP intakeperyear,somultiplyingthatvalueby30 iswhattheendtotalPisforthefilterafter30years.Fora980ft2filter,thetotalestimatedPintakeisaround 100,000 g. Graph 1 shows the agreement between Model #1 results andestimatedvalues.

35

Graph1:TotalPcapturedbyiron-enrichedfilterintheOronoinletModel#2Model#2showstheeffectsoftheIronfilteronLakeOronoduringoneseason.ItisoriginallybasedoffofamodelforComoLake,butthevaluesarechangedtosimulateLakeOronoTPinthewatercolumnforthesummermonths.ControlsthataffectTPinthewatercolumnincludemineralization,sedimentation,temperature,dailyrainfall,andturnover data for Lake Orono. The sedimentation rate for the lake feeds out of thewater columnand into a stock for sediment P,which ties intopermanent burial. ThemineralizationrateofthesedimentPflowsoppositeofsedimentationandintoTPinthewater column. Further, the temperatureof the lake affects the rateofmineralizationandtheareaofthelakeaffectssedimentationcomingfromthewatercolumn.ThereisalsoanintroductionofPintowatercolumnthatwascalculatedusingflowratesandPvaluesfromotherbodiesofwaterthatenterOrono,asdescribedbelow.ValuesforTPin thewater columnand forP comingoutof the systemweredeterminedusingpre-existingOronodata.Pcomingoutofthesystemisaffectedbydailyrainfalldata,whichplaysintothelake’sturnoverrate.Thisrateisalsoaffectedbythelake’soverallvolume,therateofdischargecomingfromElkRiver,andtheamountofPinElkRiver.

Forthismodel,themainfocuswastouseafiltertoreducetheamountofPinthewatercolumnintroducedfromoutsidesources.UsingthedatafromtheMississippiRiverWatershed TMDL document (pg. 106), we created 4 controls on P coming intoLake Orono. The four inflows come from Elk River, Direct, Orono Inlet, and TibbitsBrook.Ourfocusforthismodelwasputtingafilterontheditchedinlet.Alongwithflowdatawealsohadconcentrationsforeachofthe4streams.ThePinflowintoLakeOronowascontrolledbymultiplyingeachofthedischargesbyconcentration.Weranmodel#1forthreedifferentbenchareasoffilters(980ft²,4262ft²,9,365ft²),andtookdataoftheefficiencyofthefilteratdifferentyears(1yr.,10yrs.,20yrs.,30yrs.).Usingthisdata,wemultipliedtheditchedinletportionoftheinflowbytheefficienciesatdifferentyearswithdifferentsizedfiltersandgraphedtheLakeOronoTPtoseehowmuchofaneffect

36

eachfiltermade.Essentiallywhatthisshowed,washoweffectiveafilterwouldbeovera30yearperiodonlylookingatsummerdata.

Figure2LakeOronoTPModelApplyingFilterEfficiencytoModel#2 Figure3comparesanormalseasonforLakeOrono(regardingPconcentrationinthewatercolumn)withaseasonwhereafilterisappliedthathasa0.5efficiencyvalue.Althoughthefilterallowsfora lowerconcentrationvalue,thedifferencebetweenthetwo concentrations is significant. Using a filter at 0.5 efficiency only allows for adifferenceofapproximately2ppbbetweentheregularwatercolumnPconcentration,andwhatitwouldbewithafilter(Table1).Althoughduringcertaintimesoftheseasonthefilterisremovingalmost4,000gramsofPfromOrono,thisdoesn’thaveabigeffectbecauseofthelake’svolume.Duetothisfinding,afilterstillmaybeafeasibleoptionifit is paired with other mitigation efforts, or if a better efficiency can be obtained,thereforetakingmoreTPoutofthewatercolumn.

37

Lookingatthedifferentfilteroptionsinouriron-enrichedsandfiltermodel,forOronotoexperiencethegreatestdecreaseinPconcentrationforthelongestamountoftime,thelargestfilterneedstobeused(9,365ft²).Thisfilterrunsatanefficiencyof.5for17years,butit isalsothemostexpensiveoption(Figure3).Further,thisefficiencyonly allows for a small concentrationof P to be takenout of thewater column, so abetterefficiencywouldneedtobeobtainedtomakethepriceofthefilterworthit.

Figure3:AComparisonoftheConcentrationofPinOronoduringaSingleSeasonwithandWithoutaFilterOperatingat0.5Efficiency

Average WaterColumnTP(Days0-40)

Average Concentrationof P in the WaterColumn(Days0-40)

Average WaterColumnTP(Days100-120)

AverageConcentration ofP in the WaterColumn(Days100-120)

WithoutFilter

211448 0.000114284 235395 0.000127227

WithFilter 207940 0.000112388 231529 0.000125137

38

Table1-AverageLow(Days0-40)andHigh (Days100-120)WaterColumnTPValuesandConcentrationsofPintheWaterColumnforLakeOronowithandWithoutaFilter

Bench size(ft²)

Year1eff Year10eff Year15eff Year20eff Year30eff

980 .5 .9 .9 .9 .9

4,262 .5 .5 .6 .7 .9

9,365 .5 .5 .5 .6 .8

Table2-FilterEfficiencyValuesandCorrespondingBenchSizes

Figure3

Oneoftheconcernspresentedinanymanagementtechniqueiscost.Thecostofthe filter is related to its size. The cost estimateswere formulated in the LakeOronoStormwaterRetrofitAnalysis. The initial costof any filter ishigh,butas the filter sizeincreases,theincreaseinpriceisrelativelylow.Figure3showstheestimates.

Additionally, the largestbench size allows for a greaterefficiencyover the30-year period, but as Model 2 demonstrates, this efficiency only allows for a 2 ppbdifferenceregardingPconcentrationinLakeOrono.ThiswouldmeanthatimplementingalowerbenchsizewouldhaveevenlessofaneffectonLakeOrono’sPconcentration,since itwould runat thebestefficiency foranevenshorterperiodof time.For thesereasons, either a larger bench area must be used, or a better filter efficiency beobtainedusingthebenchsizesdepictedinFigure3.

39

Conclusion

Inmodel#1wedemonstrated theeffectivenessofdifferent sized filters in theOrono inlet over time. As expected, the bigger filters lasted longer and were moreeffective,andasdischargeincreases,theeffectivenessdecreases.Inmodel#2weusedthe efficiencies generated fromModel #1 and simulatedusing the filter in theOronoinlet foroneseason.Thoughthefiltercouldremoveapproximately4,000gramsofTPfromthewatercolumn,thisonlyresultedinremovingaPconcentrationaround2ppb.Further,we tested the9365 ft2 filteron theElkRiver, and found thatbecauseof therelativelyhigherwaterflowrate,thefilterfillsupwithPquickly(withinafewyears)andisnolongerremovingdissolvedP.

ThegoalofthesandfilteristoremovePfromLakeOrono.UsingthesandfilterineithertheOronoinletortheElkRiverwouldaccomplishthisgoal,however,itismorepractical to implement the filter in the Orono inlet. The Orono inlet flows into astormwater pond before entering Lake Orono. This gives some sediment that issuspended in thewaterachancetosettle to thebottomof thepondbeforegoing toLakeOrono.ThisisbeneficialbecauseinsteadofsuspendedPgettingcaughtinthefilteritsettlesinthestormwaterpondandaddstothelongevityofthefilter.TheOronoinletis the most advantageous spot for the Iron-enriched sand filter, but a greater filterefficiencymustbereachedbeforeimplementingthisremediationtechnique.

Webelievethattheamountofmoneyneededtoimplementthesandfilterisnotworththereturn.InordertomakeasignificantimpactonthelevelofPinLakeOrono,thesandfilterwouldneedtobeusedwithothermanagementtechniques.ItwouldbeconvenientifwecouldfindonemanagementtechniquetosolvethePprobleminLakeOrono, but in reality if thiswas feasible, itwould have been done already. Workingthrough severalmanagement techniqueswill decreaseP concentration to thedesiredlevel.