0

ITÍ0616/EGK056ÞV05186

Guðjón Atli Auðunsson

Summary and evaluation of environmental impactstudies on the recipient of sewage from the STP at

Ánanaust, Reykjavík

Work for Orkuveita Reykjavíkur (Reykjavík Energy)

Final reportNovember 2006

1

SUMMARY AND EVALUATION OF ENVIRONMENTAL ASSESSMENT STUDIESON THE RECIPIENT OF SEWAGE FROM THE STP AT ÁNANAUST, REYKJAVÍK

Table of contents

1 Topographic and physical oceanographic conditions 21.1 Disposal sites, location 21.2 Seafloor topography 51.3 Nature of the seafloor at the disposal sites 61.4 Water exchange and the freshwater regime in Faxaflói 81.5 Currents in Faxaflói and in the coastal area north of Reykjavík, general remarks 9

2 Dilution of sewage at the disposal sites. General description and simple empirical models 132.1 General description of the behaviour and fate of effluent constituents 132.2 Forecasting dilutions and effects by simple empirical and semi-empirical models 19

3 Hydrodynamic research and dispersal of faecal bacteria 253.1 Introduction 253.2 Flow model 263.3 Transport model 273.4 Model validation 283.5 Summary and conclusions 29

4 Chemical composition of sewage in Reykjavík 304.1 Main characteristics of the sewage in Reykjavík 304.2 Main chemical constituents of the sewage 304.3 Trace elements 314.4 Organic micropollutants 33

5 Suspended solids and sediment transport: bedload and suspension load 345.1 Introduction 345.2 Preexaminations and experimental layout 415.3 Flux of suspended load studied with sediment traps 435.4 Benthic photography 46

6 Biological surveys of the benthos 50

7 Monitoring of contaminants with blue mussels 53

8 Conclusions 59

9 References 61

2

1. TOPOGRAPHIC ANDPHYSICAL OCEANOGRAPHICCONDITIONS

1.1 Disposal sites, locationThe area of sewage disposal from Reykjavikand neighbouring communities (Reykjavík,Seltjarnanes, Kópavogur and Garðabær) is in

Kollafjörður, SE-Faxaflói, see maps 1 and 2.Kollafjörður may be defined as the areabounded by the line between Grótta andKjalarnes with an area of about 83 km2. Thewhole of Faxaflói, i.e. within Garðskagi andMalarrif, Snæfellsnes, is about 90 km wideand about 50 km long with an area of about4940 km2.

Map 1. Faxaflói is the bay within Garðskagi and Malarrif.

3

Map 2. Kollafjörður is the bay within Grótta and Kjalarnes. Small red dots indicate pump stations while the twolarger dots show the STPs at Ánanaust and Klettagarðar.

4

All sewage from the above communities iscollected and treated by only two sewagetreatment plants (STP) located on the northernshore of Reykjavik, one is the STP atÁnanaust, Skolpa, and the second is atKlettagarðar, see map 2 and figure 1. Thedisposal of sewage from the STP at Ánanaustis along a 500 m diffuser 3.6 to 4.1 km NWoff the station but the disposal site of sewagefrom the STP at Klettagarðar is along a 1000m diffuser 4.45 to 5.5 km NW offKlettagarðar (3.3 km northeast of the disposalsite from the STP at Ánanaust). Only thesetwo outfalls are used for all sewage fromReykjavik and neighbouring communities butdischarge from the former was started in thebeginning of year 1998 while the second wasput into operation in the middle of 2002 andin full operation in the autumn of 2005 whenall sewage to be treated had been collectedand directed to the station. Future disposal isestimated to amount to 400,000 personequivalents (pe), 150,000 p.e. from Ánanaust-STP and 250,000 p.e. from the Klettagarðar-STP. At present, the disposal is much less,however, being approximately hundredthousand p.e. from each STP. One personequivalent is defined as 60 g of BOD5 per day(C.E.C. 1991). Because of the short distancebetween these two disposal sites in addition totheir environmental similarities as describedbelow, both sites should be considered as oneintegral recipient of the sewage.In addition to data collected since theoperation of the STP at Ánanaust started in1998, the following report is based on studiesprior to its operation, which were summarisedearlier as a basis for the classification of thetwo recipients (City of Reykjavík’sEnvironmental and Technical Sector 1997 and2000). The environmental impact studiesmade after the operation started in the STP ofÁnanaust and discussed below have appearedin the following reports (all in Icelandic):

Vatnaskil Consulting Engineers 2000.Dreifing mengunar frá útrás við Ánanaust.Unnið fyrir Gatnamálastjórann í Reykjavík.April 2000. Skýrsla 00.06. (Dispersal ofpollutants from the Ánanaust outfall).

Jón Ólafsson og Sólveig R. Ólafsdóttir 2001.Ástand sjávar á losunarsvæði skolps undanÁnanaustum í febrúar 2000. Unnið fyrirGatnamálastjórann í Rreykjavík. Nóvember2001. Hafrannsóknastofnunin. Fjölrit nr 81.(Marine conditions in the recipient of sewageoff Ánanaust in February 2000).

Guðjón Atli Auðunsson 2002. Hegðun ogsamsetning fráveituvatns í Skolpu 2000-2001.Unnið fyrir Orkuveitu Reykjavíkur. Mars2002. Verkefnaskýrsla Rf 06-02. (Behaviourand composition of wastewater in the sewagetreatment plant at Ánanaust 2000-2001).

Jörundur Svavarsson 2002. Lífríki botns viðskólpútrásarstað undan Ánanaustum,-staðaeftir opnun útrásar. Skýrsla til Gatna-málastjórans í Reykjavík. Líffræði-stofnunHáskólans.(Benthic biota in the recipient ofsewage from the outfall of Ánanaust,-statusafter the operation started).

Kjartan Thors 2003. Ánanaustræsi: Set-flutningur við hafsbotn kannaður meðljósmyndun. Jarðfræðistofa Kjartans Thorshf. Unnið fyrir Gatnamálastofu Reykjavíkur.Júní 2003. (The Ánanaust outfall: Sedimenttransport examined by bottom photography).

Guðjón Atli Auðunsson 2005. Setgildru-rannsóknir út af Ánanaustum ’00-’01: hafrænmeðferð skolps. Unnið fyrir OrkuveituReykjavíkur. Desember 2005. SkýrslaITÍ0605/EGK01. (Sediment trap studies offÁnanaust ’00-’01: Marine treatment ofsewage).

Guðjón Atli Auðunsson 2006. Kræklings-rannsóknir: Ánanaust 2000. Unnið fyrirOrkuveitu Reykjavíkur. Júní 2006. SkýrslaITÍ0606/EGK02. (Monitoring contaminantswith blue mussels: Ánanaust 2000).

The following Figure 1 shows the routes ofthe ocean outfalls in relation to the nearbycoastal areas.

5

Figure 1 Route of the ocean outfall pipelines.

1.2 Seafloor topographyThe volume of the 83 km2 area of theKollafjörður at spring tide ebb is about 1.4km3. Therefore, the bay is shallow with anaverage depth of 17 m and the area within the10 m isobath is about 44% of the total area ofKollafjörður. The area beween the 10 and 20m isobaths is 26% and the rest, 30%, is withinthe 20 m and 50 m isobaths. No sills arefound in the Kollafjörður area that mightprevent mixing with the water outside.

The inner part of Faxaflói is relatively shallowand the area within the 50 m depth contour is60% of the total area (37% within the 20 and50 m isobaths) but the area within thecontours between 50 and 100 m depths isabout 30%. Out from this deep between 50and 100m is a channel, Kambsleira, towardssoutheast of Faxaflói reaching towardsHvalfjörður and Kollafjörður (Svend-AageMalmberg, 1968). Near the mouth ofFaxaflói there is a small region, 10-11% ofthe area, with depths greater than 100 m. Thisis the innermost part of the Jökuldjúp, cuttinginto the continental shelf from southwest(Unnsteinn Stefánsson and Guðmundur

Guðmundsson 1978). The total volume ofFaxaflói at ebb on spring tide is 245 km3

while the volume of Kollafjörður is 1.4 km3.No sills are between the disposal sites and theopen ocean that might prevent mixing ofwaters in Faxaflói with ocean water.

In the summer of 1993 and again in May andJune 1994 the Marine Research Institutecarried out depth measurements byechosounding (with a resolution of ±1 m)along the planned route of the outfall fromÁnanaust. The nature of the bottom withregard to rocks and other possibleobstructions was described by a diver (KjartanThors et al. 1994). Similar studies on depthsand of the bottom for the Klettagarðar-pipeline were carried out on November 261998 and in February and March 1999(Kjartan Thors 1999). The depth profiles ofthe two pipelines are illustrated in Fig. 2.

The diffuser of the Ánanaust site starts at3600 m distance at a depth of 19.5 m whilethe depth is 32 m at the diffuser’s end at 4100m.

6

-40

-35

-30

-25

-20

-15

-10

-5

0

0 1000 2000 3000 4000 5000 6000

Distance from shore along the pipeline route, m

Dep

th,m

Ánanaust

KlettagarðarDiffuser Ánanaust

Diffuser Klettagarðar

Figure 2 Depths along the route of the pipelines.

The pipe from Klettagarðar crosses a line ofrocks which lies north of Laugarnes towardsViðey after which the depth increases fastdown to about 20 m at a 1000 m distancealong the pipeline. Thereafter the depthincreases more slowly and the depth is about32 m at the start of the diffuser and about 34m at its end.

1.3 Nature of the seafloor at the disposalsites.On May 20 1994, the bottom along the path ofthe Ánanaust pipeline was photographed withdeepwater camera (Kjartan Thors 1994). Theoutermost 500 meters or so (from 4200 m to3700 m) are characterised by coarse sediments(gravel and coarse sand) with frequent ripplesof 10-20 cm wave heights and 20-50 cmwavelengths. Sampling by two types ofbottom grabs (Shipek and Van Veen) during

biological surveys of the benthos between andaround 4000-3700 m turned out to beimpossible as the thin carpet ofunconsolidated sediments gave rise to verysmall samples (Jörundur Svavarsson 1996).At 3700 m or 24 m depth, a hard rockybottom appears and at about 3540 m the lowerlimit of the brown kelp Laminaria hyperboreais reached but the plants of L. hyperborea inthis area are small and scattered. The kelp-forest grows denser in direction to shore andthe plants become taller. Still closer to shorethe plants become scattered again with zonesof no plants and at 3200 m the kelpsdisappear. Within 3200 m off shore thebottom is sandy.Figure 3 is a photograph of a typical area ofthe bottom at the disposal site of the ÁnanaustSTP.

7

Figure 3 Picture of the bottom at the disposal site of Ánanaust site, May 1997. Coarse sand with large ripples. Ripple crests fromupper left to lower right

Corresponding photographing and descriptionof the bottom floor by a diver took place inFebruary and March 1999 along the plannedroute of the pipeline from the STP atKlettagarðar and simultaneously 10 samplesof sediments were taken for determination ofparticle size distribution (Kjartan Thors,1999). Next to shore and out to about 2300m the seafloor is muddy, after which itbecomes more sandy with gravel zones out to5500 m. The disposal site itself is very flatand smooth. During a biological survey at thebottom at the planned disposal site, sixsamples were also collected and their particlesize distribution determined (JörundurSvavarsson, 2000). Around the New Year1997-1998 unconsolidated sediments wereinvestigated in the area and their depth to rockwas examined, both with seismic profilingand coring, and their particle size distributionsdetermined (Jón Skúlason 1998). Theseinvestigations showed that a thin layer ofsediment lies upon a hard layer along the pathof the planned pipeline from the STP at

Klettagarðar to the line of rocks to the west ofthe STP, see figure 1. In the area from theline of rocks to a gravel pit at the eastern endof Engey, there is a 7 m thick layer of sandysilt with shells but only few meters down torock in the gravel pit itself. North of thegravel pit, the thickness of the sediment layeris only 1 to 2 meters in a 100 m wide region(sandy silt with shells) after which thesediment layer becomes thicker towardsnorth. This thick sediment layer covers thebottom to the end of the pipeline.

The net sedimentation rate the last 8-10thousand years decreases with distance fromshore as is generally the case with increaseddepth worldwide. It may be assumed that thenet sedimentation at the Ánanaust disposalsite the last 8-10 thousand years is somewherebetween 30 and 90 cm thick (Kjartan Thors,pers.comm.), i.e. 3-4 cm per 1000 years or0.03-0.1 mm per year. This is very slow rateso close to shore but it is not known whetherthis rate is equally distributed over these 10

8

thousand years and it could well be slowerstill at present. The reasons for this slowsedimentation rate are the strong currents inthe area, storm waves during winter, erodingthe seabed and keeping the sediment particlesin suspension whereupon the currentstransport them away from the area. Inaddition, a lack of sediment supply (silt, clay,and fine grained sand) may also be a reason aswell as bioturbation, which keeps the toplayers of the sediments loose (non-cohesive)and amenable to transport. Therefore, thelikelihood of fine grained particles or the typeof particles originating from the sewerssettling at the diposal site is very small, cf.further chapter 5 below. For comparison, thesedimentation rate in the deep oceans wherethe rate is slowest is of the order of 1 cm/ka(Degens and Mopper 1976), usually below 3cm/ka, i.e. of the same order of magnitude asat the disposal sites (ka: kiloannum =thousand years). In valleys and basins on thecontinental shelves and continental slopes, therate is 10-100 cm/ka while it can reach muchhigher rates as for example in the continentalshelf off Eastern USA, where it can reach1000-3000 cm/ka. In enclosed or semi-enclosed seas and in fjords it is 30-500 cm/kawhile in some estuaries and deltas where tidalwater exerts an influence, e.g. Fraser Riverdelta, British Columbia, the rate can be700,000 cm/ka.

1.4 Water exchange and the freshwaterregime in FaxaflóiFaxaflói contains mostly relatively warm andsaline Atlantic water but impacts offreshwater occur with seasonal variation. Theamount of freshwater in Faxaflói wasestimated to be about 3000x106 m3 on averagein the years 1966 and 1967 but high seasonalvariation occurs (Unnsteinn Stefánsson andGuðmundur Guðmundsson 1978). In theseyears, the total flow of rivers into Faxaflói,mainly from its eastern part, was about29×106 m3/day on average or 336 m3/s, theriver Hvítá in Borgarfjörður (glacier water)contributing most. Elliðaár, the river closestto the disposal sites, has an average flow of4.6 m3/s (average of 30 years (National

Energy Authority, Hydraulic Laboratory2006)) but the freshwater effects of Elliðaár inthe area have been evaluated (VatnaskilConsulting Engineers, 1984). There is alsoconsiderable inflow of freshwater into thesouthwestern part of Faxaflói, the more sowhen the winds are from the south or 20-24.5×106 m3 d-1 for each knot of southerlywinds, i.e. about 0.5 m s-1 which is unusuallylow for the area, even for southerly winds, cfchapter 1.5 below. This water originates fromrivers in Southern Iceland (mostly glacierwater) that enters the coastal current whichcarries this water around Reykjanes and intoFaxaflói. From direct flow of rivers intoFaxaflói, the mean residence or flushing timeof freshwater in the whole of the bay is about100 days, but if freshwater entering the bayfrom S-Iceland is also included, assumed tobe equal to direct river flow, the residencetime of freshwater is probably 50 days orshorter (Unnsteinn Stefánsson andGuðmundur Guðmundsson 1978). In otherwords, inflow of freshwater into Faxaflóifrom S-Iceland is of a similar magnitude orgreater than from all the rivers flowing intothe bay.Because of unhindered mixing of the bay-water with ocean water, freshwater effects aremainly seen within the 100 m depth contourof Faxaflói (Unnsteinn Stefánsson andGuðmundur Guðmundsson 1978; UnnsteinnStefánsson and Jón Ólafsson 1991).

The data presented by Unnsteinn Stefánssonand Guðmundur Guðmundsson (1978)together with total volume of Faxaflói may beused to give a very rough estimate of the flowinto Faxaflói if the average inflow offreshwater from the south is assumed to beequal to direct river inflow and that waterwithin the bay (fresh water and seawater) isassumed to be well mixed. Under theseconditions, conservation of volume andsalinity gives a total flow into the bay ofabout 5-6104 m3/s and a flushing time of 7-8weeks for Faxaflói, which, of course is thesame as for fresh water. It has to be borne inmind that these are rough estimates and mayfluctuate greatly during the year and from one

9

month to another. For example, if waterinflow from south is nil as above, the flushingtime is 100 days but if it is half of the directinflow or double, the residence times are 70days and 35 days, respectively.The total anticyclonic transport aroundIceland has been estimated to be about 106

m3/s (as cited in Unnsteinn Stefánsson andJón Ólafsson 1991), which means that lessthan 10% of the anticyclonic transport entersFaxaflói. If half of the anticyclonic transporttakes place within the 200 m isobath as hasbeen roughly estimated (Unnsteinn Stefánssonand Jón Ólafsson 1991), one may assume thatvery roughly 1/8 of the transport takes placewithin the 100m isobath. This means that 40-50% of the flow within the 100 m isobathenters Faxaflói. To give a further example,this flow into Faxaflói would equal the flowof a current of velocity 5 cm s-1 at a waterdepth 50 m (about the average depth north ofGarðskagi) and a width of 22 km (12 nauticalmiles).

These flushing or residence times are ofconsiderable importance since they suggestthat most of the discharge of nutrients duringOctober to February will be flushed out beforephytoplankton might start to utilise them. Thegrowth season usually starts in mid-Marchand lasts until mid-October (Þórunn

Þórðardóttir 1986). Additionally, not lessimportantly, these short residence times,especially during summer, mean that thenutrients released cannot sustainphytoplankton or macrofaunal biomass asevidenced by a number of studies in estuariesand bays of the North Atlantic Ocean (Nixonet al. 1996; Josefson and Rasmussen 2000).Short residence times result in net export ofthe nutrients from the system and therebydecreased risk of eutrophication. Using theempirical model of Nixon et al. (1996), thisresidence or flushing time of Faxaflói meansthat 60-80% of the total nitrogen load and 70-95% of the total phosphorus load will beexported from Faxaflói. Discussion onflushing times and water exchange ofKollafjörður is given in chapter 2.2 below.

1.5 Currents in Faxaflói and in the coastalarea north of Reykjavík, general remarksClockwise coastal currents predominate inmost parts of Faxaflói (Svend-AageMalmberg 1967 and 1968). The mainresidual currents in Faxaflói are depicted infigure 4 below. The figure shows surfacecurrents during summer. The bottom currentshave similar paths and are about half anautical mile per day (1 cm/s) while surfacecurrents are about 5 nautical miles per day (11cm/s) (Svend-Aage Malmberg 1968).

Figure 4 An outline of surface currents in Faxaflói during summer (reprintedfrom Svend-Aage Malmerg 1968 with permission).

10

Closer to shore at Reykjavik the residualcurrents have been measured to be 1-5 cm/s atthe disposal sites. The net current follows thecoastline with net currents towards northeastat the Ánanaust site but southeast at theKlettagarðar site.

The current system close to Reykjavik ishowever dominated by tidal currents. Thetidal range in Reykjavík at spring tide is 4m atflood (0.2 m height above zero level on ebb)on average while it is 3.0 m on average atneap tides (1.3 m height above zero level onebb) (Vatnaskil 1994). Average tidal range isabout 2.1 m. The tidal currents are thereforemuch larger than the residual current orbetween 5 and 25 cm/s. The maximumcurrents occur at high and low tides (actually1-2 hours after slack water (Svend-AageMalmberg 1968)), lowest at neap tides andhighest at spring tides. The direction of thetidal currents at the Ánanaust site are towardsnortheast on flood, thus reinforcing theresidual current and southwest on ebb.However, the vectoral currents during the

tidal cycle of 12 h and 25 minutes form anellipse (hodogram) with significant currents inall directions although the currents at high andlow tides take up most of the time. At theKlettagarðar area the tidal directions aretowards southeast on flood reinforcing the netcurrent and northwest on ebb. For the sake ofsimplicity, one cycle time (12 h and 25 m) atthe Ánanaust-area may be divided into fourintervals:

Slack water,low water

From 0.5 h prior to and 0.75h afterslack water, 1.25 hours in total;turbulent mixing

Flood 0.75 h after inflowing slack wateruntil 0.5 h after high tide, 5 hours intotal; currents dominantly towardsnortheast, about 10 cm/s on average(lunar month).

Slack water,high water

From 0.5 h prior to slack water until0.75 h, 1.25 hours in total; turbulentmixing

Ebb From 0.5 h after the outflowingslack water until 0.5 h after low tide,5 hours in total; currents dominantlytowards southwest, 10-15 cm/s onaverage (lunar month)

Therefore, the tidal currents only partiallyflush the sewage in each cycle since outgoingwater on ebb returns on flood and vice versa,the returning water has been estimated to beabout 70% in an area closer to shore(Isotopcentralen, 1971). However, in additionto affect net transport from the area, theturbulence of the water during the tidal cyclescauses enormous mixing of the seawater andthereby seawater with sewage water, i.e.contributes significantly to the initial dilution.The tidal currents are approximately at rightangles to the diffuser of the Ánanaust pipewhich have the openings on the sides of thepipe directing the jets horizontally into theseawater (or more exactly at 30° fromvertical, see below).

Wind induced currents and accompanyingwave motion are also of considerableimportance. Wind induced currents are due tothe friction the wind imparts to the surfacelayer by wind stress (proportional to the windspeed squared). This motion is transferred to

deeper layers involving drag forces beneaththe sea surface and at the seabed and therebyturbulence in the water masses in betweenthese boundaries by vertical transport of smallmasses of water (eddies) and their inherentmomentum. Turbulence is caused by allcurrents but the vertical turbulent diffusion isusually at maximum at mid-depths for windinduced currents (Quetin and De Rouville1986). Most of the energy transferred fromthe wind to the sea surface, however, ismanifested by waves that cause circularmotion of water particles at the surface. Theorbital velocity of this circular motion(proportional to wave height) at the surface istransferred vertically down the water columnbut decreases fast and exponentially withdepth where the orbits become smaller andgradually elliptical. At any particular place,the waves may have formed locally or havearrived from elsewhere, even originating quitefar from the observed waves. The verticaltransfer of this motion caused by winds maybe hampered by density-stratification, which,however, is neither of long duration nor very

11

steep in the areas north of Reykjavík, seebelow.Both of these types of motion caused bywinds are of considerable importance down tothe seabed as seen by the flux of sediments tothe overlying water above the seabed (GuðjónAtli Auðunsson 2005). It is known byexperiments that the most important processesfor sediment being transported are the socalled burst-sweep cycles where downwardsweeps of high velocity water penetrate lowervelocity layers of water near the seabedexerting instantaneously high shear stress atthe bed and displacing both sediment and theimmediately overlying lower velocity water asbursts upward into the main flow, whichcarries the suspension away. Additionally,wave motion is particularly effective inmoving sediment particles from the seabed byits circular or stirring to and fro motion at theseabed carrying sediments to the main flow.Figure 5 shows an example of wind speeds atReykjavík harbour (Miðbakki) grouped ondirectional sectors. The figure shows that thehighest wind speeds may be expected fromthe north, northeast and east. The averagewind speed for this particular year was 4.7m/s (median 3.9 m/s). Figure 6 shows thedistribution of wind directions in Reykjavikharbour indicating similar behaviour from oneyear to another and dominance of wind fromeast and southeast, i.e. parallel to the coast ofReykjavík. Winds parallel to the coast impartabout 1.5-2 times higher current velocitiesthan directions perpendicular to the coastline,a fact that was observed experimentallyduring sediment trap studies at both theÁnanaust site and Klettagarðar site (GuðjónAtli Auðunsson 2005). Finally, Figure 7shows monthly average of wind speeds in theyears 1961-2005 where lowest wind speedsmay be expected in July (median of all yearsbeing 4.6 m/s) and highest in January andFebruary (median of all years being 6.5 m/s).Surface currents are proportional to the windspeed and directed about 45° to the right ofthe wind direction. Therefore, one mightexpect that most of the time the wind-drivencurrent gives surface currents from shoretowards north. These currents may be

expected to be of the order of more than 13-24 cm/s more than 90% of the time on amonthly basis using a semi-empiricalrelationships for coastal areas (Quetin and DeRouville 1986). However, these currentsdecrease sharply with depth at the same timeas they turn more to the right and at a certaindepth they are in opposite direction to that onthe surface. Already at 2-3 m depth at theÁnanaust-discharge area, these currents areabout half the surface current when wind isparallel to the coast while they are only 15%of the surface current when wind directionsare perpendicular to the coast. At 1 m fromthe bottom at the Ánanaust site, the wind-induced currents are 25% of the surfacecurrent if winds are parallel to the coast whilethey are only 10% of and in opposite directionto the surface current if winds areperpendicular to the shore. An investigationof the sediment fluxes show that wind speedsgreater than 3 m/s from southeast do set thebottom fines (clay and silt) into suspension(Guðjón Atli Auðunsson 2001b and 2005)showing that the wind-generated waves andcurrents reach down to the bottom.According to the Beufort scale the significantwave heights produced are greater than about0.5 m if these winds prevail more than 2hours. Significant wave heights measured atGarðskagi, about 40 km WSW of Reykjavíkacross open sea, correlated well with sedimentfluxes at the Ánanaust discharge area whengreater than 0.9 m (periods always less than15s) (Guðjón Atli Auðunsson 2001b and2005). These wave heights have empiricallybeen shown to move particles of 0.2-0.3 mmdiameter (fine and medium sized sand) at adepth of 30 m. Winds have thus a verysubstantial influence on the currents andmixing in the area.

The system into which the sewage isdischarged is apparently very well mixed mostof the time. During the summer months,notably July and August, the system iscalmest due to comparatively low windvelocities. Only during this time, smalldensity gradients are formed, which decreasesomewhat the buoyancy of the sewage plume

12

as well as inhibiting vertical transfer of wind-induced motion from surface down. This is

therefore the period of the year in which mostof the studies below have been concentrated.

0

2

4

6

8

10

12

14

0 50 100 150 200 250 300 350 400

Direction (clockwise from north)

Win

dsp

eed,

m/s

90% percentileMedian10% percentile

Figure 5 Median wind speeds in Reykjavík harbour in the period 01/05/200-30/04/2001, i.e. awhole year. Measurementscarried out every hour.

0

5

10

15

20

25

0 50 100 150 200 250 300 350 400

Direction (clockwise from north)

Fre

que

ncy

,%

01/05/00-30/04/01

01/11/95-31/12/9602/07/98-07/03/99

Figure 6 Distribution of wind direction in Reykjavík harbour in the years indicated, whole yearsexcept ’98-’99.

13

0

1

2

3

4

5

6

7

8

9

0 2 4 6 8 10 12

Month

Win

dsp

eed

,m/s

Median10% percentile90% percentile

Figure 7 Median monthly wind speeds in Reykjavík in the years 1961-2005.

2 Dilution of sewage at the disposalsites. General description andsimple empirical modelsThe general picture of the currents at theÁnanaust and Klettagarðar disposal sites maybe used to estimate the dilution and flushingof water in the areas. The issue of sewagedilution has been the subject of innumerableworldwide studies in the laboratory, in thefield, and with mathematical models. Simplecalculations based on the general currentspresented above may give an idea of how thesystem works but results of more elaboratemathematical modelling are presented inchapter 3. In the discussion below, a generalpicture of the dilution and dispersionprocesses is given, after which simplecalculations are presented with comparisonswith actual measurements.

2.1 General description of the behaviourand fate of effluent constituentsThe diffuser at the Ánanaust site is 500 mlong with circular 0.075 m diameter orificeson each side at about 60° from horizontal line

through the centrum of the pipe. The 60°angle for the nozzles has prevailed in outfalldesign ever since Zeitoun et al. (1972, cited ine.g. Roberts and Toms 1987) found this angleto give the longest trajectory and therefore thegreatest dilution. The orifices are about 14 mapart on each side, i.e. 7 m between holesalternating between sides. Therefore, theplumes from two adjacent diffuser holes willnot overlap on their way up to the surface (thediameter on the surface containing 90% of theliquid discharge will be 10-12 m in stagnanthomogenous media). The jet dischargevelocity from each orifice will be about 3 m/sat an average effluent flow from the ÁnanaustSTP of 1 m3/s. This is the first stage ofdilution since this orifice velocity will impactthe surrounding media with turbulence causedby the transfer of high momentum flow.Empirically, it has been found that thedilution in a stagnant medium is roughlyinversely proportional to the orifice diameter(and directly proportional to the water depth)(Fischer et al. 1979) while a more detailedexperimental and theoretical analysis for thejet in cross flow results in buoyancy and

14

orifice velocity entering the equation (Chuand Goldberg 1974; Roberts and Toms1988).

Dilution may be considered to occur in fourphases, where the first is the dilutionimmediately after release of sewage from thediffuser by the jet from the orifices. Thesecond stage is the subsequent immediatemixing that is effected by the buoyancy of thedischarge. Thses two first stages arecollectively the initial dilution in the nearfield area. The third stage is the dispersion ofthe sewage on the surface, the far fielddilution, and the fourth is the long run effectson a larger scale.

Initial dilution or near fieldSince wastewater has a density close to freshwater, it is lighter than the surroundingseawater (which is colder and more saline)and the resulting buoyancy deflects the jetsfrom each hole of the diffuser upward,forming a plume which is swept downstreamby the current. The plumes entrain seawateras they rise causing dilution and increaseddensity and finally, if the surroundingseawater is not stratified, will reach thesurface. However, if the seawater has a largeenough density gradient, the plume maybecome trapped under the surface. This initaldilution (jet momentum and buoyancy) occurswith vigourous mixing and causes rapiddilution well within 10 minutes after therelease. This is the so called “near field” or“initial mixing region”. Within the initialmixing region, the fluctuations inconcentrations may be large compared to theiraverage value. At the end of the inital mixingregion the fluctuations are small, however.When the plume hits the surface, the averagedilution at the surface is about two times theminimum dilution (maximum concentration)at the centerline of the plume (Roberts 1991).With the dominating tidal currents at both thedisposal sites, Ánanaust and Klettagarðar, forwhich the current vectors form an ellipticalpath, the plume at the Ánanaust site forexample will appear as an ellipse on thesurface with the long axis perpendicular to the

diffuser. This elliptical form is also due to thenature of the dispersion when the sewageplume enters the surface, see below.Subsequently, the established wastefield driftswith the surface currents to a region called the“far field” with further dilution by dispersion(diffusion and convection by turbulence).This farfield dilution is the third stage ofdilution and is acting with totally differentmechanisms from those in the nearfielddilution.

If there is any residual density difference leftwhen the plume hits the surface, i.e. still someupward or positive buoyancy, the plume mayspread or float like an oil slick on the densernatural seawater and will be pushed by thecontinous flow without any vertical mixingwith the underlying water. For this to happen,however, the surface has to be relativelysmooth (e.g. wind velocities less than 3 m/s)and the density difference must amount to aninital dilution of less than about 125 (Quetinand De Rouville 1986). Neither of theseconditions are likely to ever existsimultaneously in the coastal waters north ofReykjavík.

Another aspect of this initial dilution appliesif the current speeds are large or in the range8-80 cm/s for the Ánanaust site, when a socalled forced entrainment regime may takeplace, where the discharge is rapidly carriedout to sea. In this case the plume cannotentrain all of the oncoming flow and the riseheight and thickness decreases with increasingcurrent speed (Roberts 1991). This situationmay occur in the area of discharge, especiallyduring winter.

The initial mixing zone ends when there is nofurther dilution due to turbulent mixingcaused by the jet momentum flux, buoyancyand accompanying friction. During the firsttwo phases, the thickness of the wastefieldincreases. In the end of the initial dilutionregion, the turbulence collapses, i.e. when theinital kinetic energy of the jet and the initialpotential energy due density difference(gravitational forces) have been dissipated.

15

For the Ánanaust site at 10 cm s-1 averagecurrent, this will take place within 50 metersfrom the diffuser or within 10 minutes.Further downstream, the wastefield widthincreases and its thickness decreases due togravitational collapse. In the transitionbetween inital dilution and far field, theplume flows at a rate of the discharge flowtimes the inital dilution factor which is alsoapproximately equivalent to a plume movingat the ambient current speed with a widthroughly equal to the diffuser length(Medcalf&Eddy 1991). With theseassumptions and using a depth averagecurrent of 10 cm/s for the Ánanaust site, anaverage dilution of 1:1000, see below, and adischarge flow of 1 m3/s (Guðjón AtliAuðunsson 2000 and 2002), and assuminghomogenous medium like in winter, thethickness of the plume would take up about80% of the water column. Alternatively, adepth average current of 5-8 cm s-1 wouldmake the transition plume extend over thewhole water column at Ánanaust under thesame assumptions. Therefore, it seems likelythat in the transition between near field andfar field, the plume probably extends overmost of the water column most of the time.

Far fieldIn the far field, the mixing is caused byambient oceanic turbulence and proceedsmuch slower than in the near field. Generally,an additional dilution of 5-20 is expectedwithin a zone of 1 km (Quetin and DeRouville 1986). The plume broadens on itsway downstream by lateral spreading and withincreasing rate since larger and largerturbulent eddies participate in the dispersion(i.e. the turbulent diffusion coefficientsincrease with time and/or distance from thesource, see e.g. Stacey et al. 2000, Okubo1971, and Okubo 1974). In a tidalenvironment, the reduction in contaminant onthe plume centerline may be relatively smallfor the first hours of plume travel. Atmoderate distances the levels are intermittent,consisting of long periods of no detectableconcentrations or of concentrations close tothe low background levels intermixed with

much higher concentrations when the plumeis present. These higher concentrations wouldoccur in a region about the size of the tidalexcursions, i.e. within about 5 km east andwest of the diffuser at the Ánanaust site. Theturbulent diffusion is mainly caused by shearstress at the seabed and wind stress at thesurface and are far from being equal in alldirections, the horizontal dispersioncoefficient in direction of the current beinghighest although of little importance inrelation to the current speed with which theplume moves while the one perpendicular tothe current increases with time (distance)usually following the 4/3-law, i.e.proportional to the length of plume (diffuserlength at the start) to the power of 4/3 (Staceyet al. 2000). Therfore, the first hours it maybe 0.1-1 m2/s, then dispersion that is due to acombination of turbulent diffusion and shearstress becomes increasingly importantresulting in a dispersion coefficient around 25m2/s. For longer times, horizontal sheardominates, and the dispersion may grow up to100-1000 m2/s (Roberts 1999a; Zimmerman1986). This large ultimate dispersioncoefficient combined with flushing, seebelow, results in very low backgroundcontaminant concentrations. The verticaldispersion coefficient is usually about onethird of the transverse coefficient (Quetin andDe Rouville 1986) but it decreases withincreasing stratification (inverselyproportional to the density gradient) and isaffected by wave characteristics, i.e. increaseswith wave height (proportionally) and shorterwave periods (inversely proportional to thewave period squared) (Koh and Brooks 1975).The dispersion coefficients may beapproximated fairly well but measurements ofthese at the discharge site by for examplerelease of dyes or radiaoctive tracers areusually necessary since the coefficients mayvary in space and time with the flowcharacteristics, i.e. they are not physicalconstants. If the coefficients have beenapproximated or found by measurements, theycan be used to estimate downstream dilutionwith simple calculations (Metcalf & Eddy1991; Quetin and De Rouville 1986) where

16

the simple solution of the diffusion equationby Brooks is most commonly applied (Brooks1960). The Brooks solution assumes a steadycurrent and negligible vertical andlongitudinal dispersion. These calculationsmay give quite accurate avereage values forsimple systems, i.e. for uniform currents asopposed to varying current speeds anddirection across the water column and timeand when density gradients are linear. Thismay also give good insights into theprocesses, e.g. at the worst conditions.Otherwise, the use of more elaboratemathematical models is required (Roberts1991; Medcalf&Eddy 1991; Roberts 1999a),see chapter 3, but the same general principlesas described above are, however, always used.

The dilution factor in the far field multipliedby the initial dilution factor gives the totaldilution at any distance from the plume centerwithin the mixing zone.

It is important to keep in mind that thedescriptions above relate to time-averagedsituations since, during short travel times afterrelease from the diffuser, the plume meandersaround and breaks up into successive “puffs”like a smoke from a chimney (Quetin and DeRouville 1986; Roberts 1999a). This causesepisodic, or intermittent, “fumigation” of agiven fixed point by the contaminants itcarries, i.e. the local contaminantconcentrations there are negligible most of thetime but is relatively high during fumigationepisodes. For this reason, the picture of boththe near field and far field are patchy at anygiven time which many studies have revealed(e.g. Carvalho et al. 2002; Petrenko et al.1998). For far field modelling the horizontaltrajectories of these patches have beenexamined with progressive vector diagramsand resulting visitation frequency diagrams(Robets 1999; Washburn et al. 1999) and withparticle tracking, computationally notpractical, however, for very long simulationsextending over many months (Roberts 1999a).In short, the deterministic prediction oftransport in far field may not be possiblebecause the trajectory of the plume in an

unsteady tidal environment is a chaosphenomenon (Zimmerman 1986), resulting inindividual plume trajectories that areunpredictable, random, and not repeatable.However, the best present day modelsdescribe the gross features of the plumesreasonably well but there remains aspectswhich cannot be modelled yet, particularly thepatchy nature of the wastefield. Furthercomplications are for example due to decayprocesses, e.g. the decay rate of bacteria isfastest during daylight hours and slowest atnight, and decreases with increased depthbelow the water surface. This variable decayrate of bacteria therefore also results invarying die-off times with season in Iceland,see chapter 3.

Long run effectsFinally, exchange rates on a larger scale, evenat continental shelf scales, and biological andchemical decay processes determine thepossible build-up of contaminants in the longrun. This is the fourth stage of dilution. Thisphase is often described by simple tidal prismmodels (e.g. Isotopcentralen 1971a; Strain,Wildish and Yeats 1995) which work well ifsills are not present or/and if there are nostrong/stable stratifications (Strain and Yeats1999).

In the UK, the different mixing zones havebeen the subject of comprehensive studies andmodelling. The European Union UrbanWaste Water Treatment Directive (CEC1991) sets out the general conditions forwhich primary treatment, rather thansecondary, may be sufficient for a marinedischarge. The Urban Waste Water TreatmentDirective (UWWTD) sets secondarytreatment of urban wastewater as standard(article 4), which may though be simplified toprimary treatment only if discharging fromagglomerations of between 10,000 and 150,000 p.e. to coastal waters defined as “lesssensitive” recipients and if comprehensivestudies indicate that such discharges will notadversely affect the environment (article 6).In exceptional cases, primary treatment is alsosufficient for agglomerations over 150,000

17

p.e. when “it can be demonstrated that moreadvanced treatment will not produce anyenvironmental benefits” (article 8). Enhancedor tertiary treatment is required if discharginginto “more sensitive” waters (article 5). Thepresence or absence of eutrophic conditions isone of the indicators of “sensitive” or “lesssensitive” waters. Article 6 requires“comprehensive studies” and so the in the UKthe “Comprehensive Studies Task Team”(CSTT) was set up to provide guidelines onthe implementation of the directive. Theseguidelines cover four main aspects of theproblem: initial dilution; oxygen demand;nutrient loadings, particularly in respect ofeutrophication; and suspended solids (allthese parameters will be addressednumerically in 2.2 below except forsuspended solids which will be dealt with inchapter 5). CSTT introduced a distinctionbetween “zones” or spatial scales (CSTT1997; Sherwin 2000):

o Zone A: that around and close to theend of a pipeline or a fish farm;concentrations will be similar to theinitial dilution in the boil.

o Zone B: that of a sea-loch basin orother water area in which the residencetime of pollutants is a few days; this isthe near field region according toCSTT, denoted to be at least the sizeof the tidal excursion and wherephytoplankton growth might takeplace in favourable circumstances.

o Zone C: a larger region, e.g. anoffshore water providing the"boundary conditions" for zone B.This is the far field region where theresidence time is of the order of weeksto months.

These guidelines have been used in the UKand are likely to be applied in other Europeancountries and the latest version of the CSTT-

model appears in Tett et al. (2003) where itwas applied to a large variety of systems(Kongsfjorden, Spitzbergen; Gullmaren, W-Sweden; Himmerfjärden, E-Sweden; Firth ofClyde, Scotland; Golfe de Fos, S-France; RiaFormosa, S-Portugal).

The UWWTD defines “eutrophication” as theenrichment of water by nutrients, especiallycompounds of nitrogen and/or phosphorus,causing an accelerated growth of algae andhigher forms of plant life to produce anundesirable disturbance to the balance oforganisms present in the water and to thequality of the water concerned. The definitionof OSPAR is essentially that of UWWTD butwith an emphasis on ”undesirable changes(“Eutrophication” means the enrichment ofwater by nutrients causing an acceleratedgrowth of algae and higher forms of plant lifeto produce an undesirable disturbance to thebalance of organisms present in the water andto the quality of the water concerned, andtherefore refers to the undesirable effectsresulting from anthropogenic enrichment bynutrients as described in the CommonProcedure”) (OSPAR 2001). Therefore,eutrophication is a process of acceleratedgrowth or a shift from lower to higher trophicstatus, see also Nixon (1995). The teamdeveloped the CSTT model (CSTT 1997), andproposed the “Water Quality Standards” (nowEnvironmental QS) in which winter values ofdissolved available inorganic nitrogen(DAIN) should not exceed 12 µM andsummer chlorophyll a should not exceed 10mg m-3 (DAIN is the sum of nitrite, nitrateand ammonia). The value of cholorophyll awas chosen 10 mg m-3 since at that level thephytoplankton will be seen by the bare eye.However, these are not applicableeverywhere, since the trophic status maydiffer from one location to another andnaturally eutrophic areas will exceed thesevalues. According to Nixon (1995) marinewaters are classified according to thefollowing scheme:

18

Status, trophic state Primary productivity, g C m-2 y-1

Oligotrophic <100Mesotrophic 100-300Eutrophic 300-500Hypertrophic >500

For example, the trophic status of the coastalwaters north of Reykjavík are naturallyeutrophic (Þórunn Þórðardóttir 1994) sincethe annual productivity is more than 300 g Cm-2. Thus, using the relationship betweenprimary productivity and chlorophyll a (Chla) in Icelandic waters (Kristinn Guðmundssonog Kristín J. Valsdóttir 2004) and an averageeuphotic zone down to a depth of 20m in therecipient (light attenuation evaluated bySecchi depth (kD=1.4/Ds) and relationshipbetween Secchi depth, Ds, and Chl a (KristinnGuðmundsson, Þórunn Þórðardóttir ogGunnar Pétursson 2004)), the average Chl a inthe waters north of Reykjavík during thegrowing season will be more than 15 mg m-3,a result also obtained if a yield of 1.1mg Chl aper mmol nitrate is used (Tett et al. 2003) andassuming most of the winter nitrate inFaxaflói, 13.7±0.3 µM (n=25 on a transectwest of Akranes out to Jökuldjúp) (Sólveig R.Ólafsdóttir 2006), to be used for primaryproduction. Since this is an average value ofChl a, considerably higher values of Chl aconcentration may occur during autumn butespecially spring blooms. The nitrate qualitystandard value of 12 µM in the CSTT modeldoes therefore not apply for Icelandic waterssince natural winter values of nitrate arehigher. Therefore, the approach proposed byOSPAR (2001) as a measure of significantshift from oligotrophic to eutrophic withoutproviding an absolute standard foroligotrophic or eutrophic, is more appropriatein general and involves the following criteria:

a 50% increase in winter DAIN (dissolvedavailable inorganic nitrogen) and DIP(dissolved inorganic phosphate)concentrations, and in maximum and meangrowth season phytoplankton chlorophyllconcentrations;

increase is relative to “historical region-specific background concentrations” or“spatial (offshore) backgroundconcentrations”;

in waters of varying salinity,concentrations should be corrected tostandard salinity.

The eutrophication proceeds as follows:

o nutrient enrichment;

o increased phytoplankton growth, biomass andprimary production;

o undesirable disturbances to water quality and“balance of organisms”

Finally, the undesirable disturbance used inthe definition of eutrophication of both theUWWTD and OSPAR may be described bythe following:

1. shifts in relative abundance of diatoms anddinoflagellates or flagellates (in favour of thelatter);

2. increases in species considered to indicateeutrophic conditions and which themselvesmight be nuisance or harmful species;

3. increase in harmful algal blooms (HABs)and/or “red tides”;

4. decrease in water transparency with adverseeffects especially on benthic sea-grasses;

5. increased sedimentation of organic matter,resulting in increased rates of oxygenconsumption in near-bottom water or seabed,and hence to potential deoxygenation in theseregions;

6. (increased) kills of plankton, benthos or fishas a result of above effects;

7. changes in food webs as a result of aboveeffects and also because change inphytoplankton might cause change in grazingzooplankton.

19

The simple CSTT model deals with theeutrophication potential by the followingcriteria (Tett 2000) after a zone has beendefined and water exchange in a single andwell mixed box has been evaluated from dataon currents:1. How much will the receiving water

concentration of nutrients increase as aresult of the discharge?

2. If the extra nutrients were used byphytoplankton, would the resultingconcentration of Chl a exceed theenvironmental criteria, 10 mg m-3 inCSTT but 50% increase by OSPAR,assuming total conversion of nutrientsinto Chl a?

3. Can phytoplankton in fact use thesenutrients, i.e. is the receiving water tooturbid to allow plant growth, or dilutedtoo fast to sustain an increase in theabundance of plankton, providedinformation on photosyntheticallyavailable radiation (PAR) is monitored?

4. Would the answers to the questions abovebe any different if the discharge receivedprimary rather than secondary or morestringent treatment?

Both the OSPAR and the CSTT models usechlorophyll a as a measure of microplanktonbiomass.

2.2 Forecasting dilutions and effects bysimple empirical and semi-empiricalmodelsIn the simplest of cases where no densitygradients are present, maximum dilution inthe vicinity of the diffuser may be found bycomparing the outfall discharge with theprobable seawater flow through a sea cross-section. Assuming an average current speedover the whole water column of 10 cm/sperpendicular to diffuser length, aconservative estimate for the Ánanaust area,the maximum initial dilution at the averagedepth of 25 m at the Ánanaust site (neap tidelevel), is between 1000 and 1500. Here, theaverage sewage flow of 1 m3/s from the

Ánanaust STP is used (Guðjón AtliAuðunsson 2000 and 2002).

Initial dilution or near fieldThis is equivalent to zone A in the CSTT-model (Sherwin 2000; CSTT 1997).Oceanographic parameters were monitoredclose to the discharge area of Ánanaust STPin the years 1966 and 1967, and showedstratification during June, July and August1966 with a maximum aproximately lineardensity gradient of 1.2 g/L over 30 m watercolumn for a short while at the end of Juneand the beginning of July (UnnsteinnStefánsson et al. 1987). The pycnoclines are,however, often disrupted by winds in this area(Þórunn Þórðardóttir 1986) althoughoccurring when wind speeds are expected tobe lowest as discussed above. A densitygradient, nonlinear however, was alsoobserved much closer to shore in the summerof 1977, i.e. a maximum of 1.3 g/L differenceacross 30 m water column between Viðey andEngey 24/07/1970 (Isotopcentralen 1971a).Therefore, this gradient seems to reflect themost adverse conditions as regards dilutionrate in the discharge area. Taking this intoaccount and using a simple but wellestablished semi-empirical model for theinitial dilution from a diffuser (Roberts 1991;Roberts, Snyder and Baumgartner 1989a,1989b, 1989c), a minimum dilution of 300 isfound or an average initial dilution of 600(assuming an average current of 10 cm/s and adischarge rate of 1 m3/s). However, apartfrom a short period in mid-summer, the wateris not stratified in this area (UnnsteinnStefánsson et al. 1987), and simple empiricalrelationships give a minimum dilution of 750and average dilution of 1400 (25 m averagedepth to diffuser, 10 cm/s currentperpendicular to diffuser, and a discharge rateof 1 m3/s). This empirical relationship isactually 60% of the simple box-model valueabove. These calculations show that anaverage initial dilution of about 1000 may beassumed most of the time. Additionally,under this stratification conditions and adepth-average current of 10 cm/s, the plumewill probably not reach the surface.

20

These simple calculations of inital dilution arein line with experimental results for severalmicrobial parameters in caged musselsshowing lowest dilution factor of 1000 duringJuly and August 2000, this minimum dilutionoccurring at the start of the diffuser (GuðjónAtli Auðunsson 2006). A somewhat highermedian dilution factor was obtained for faecalcoliforms when results for surface watersamples collected on 02/02/2000 (i.e. longdie-off times, sampling close to spring tideduring ebb, and the wind speed was 2.6 to 3.6m s-1 from SSW) in an area along and close tothe diffuser were compared with estimatedlevels in the discharge (using experimentalper capita value of 1.3x1010 MPN/p.e./d, seechapter 4). However, about two and a halfmonths later or on 13/04/00 (also close tospring tide where winds were 1.5-3.1 m s -1

from the north), hardly any sample gaveresults of faecal coliforms above 2MPN/100mL (Vatnaskil 2000), one valuebeing high, the station at the beginning of thediffuser, the same station giving highestresponse in the mussel samples.Investigations on the level of faecalenterococci in 60 beaches in S-Californiashowed highest values during ebb at springtides (Boehm and Weisberg 2005) suggestingthat these sampling dates, especially on02/02/2000, may have been at the mostadverse conditions. Additionally, a minimumdilution of 600 was observed in one sampleon 02/02/2000 above the middle of thediffuser when salinity-silicate relationshipsobtained prior to and after discharge from theSTP at Ánanaust were applied (Jón Ólafssonand Sólveig R. Ólafsdóttir 2000). However,the average dilution in 24 surface samplestaken in a small area close to the diffusershowed an average dilution of 1100 orbetween 600 and 1600 (Jón Ólafsson andSólveig Ólafsdóttir 2000). It is more difficultto use either nitrogen or phosphate to evaluatedilution due to their natural presence inseawater. However, application of the percapita values below together with dischargevolumes that particular day, and using thestation of highest salinity in the study of JónÓlafsson and Sólveig Ólafsdóttir (2000) as the

background or natural level of the nutrients, atotal average dilution factor can be estimated.Ammonia (ammonia being 33% of totalnitrogen in the effluent (Guðjón AtliAuðunsson 1992)), total nitrogen, and totalphosphorus in the close vicinity of and abovethe diffuser show dilution factors that are log-normally distributed as found in other studies(Roberts 1999b). The dilution factors forthese three nutrients did not differsignificantly and are on average 1000 asfound above but with 10 and 90% percentilesranging from 400 to 2000.

Therefore, all results above taken together, theaverage initial dilution is about 1000 bothinferred from experiments and calculations bysemi-empirical models, where the averagedilution is 1.5-3 times the minimum dilution.

Applying this dilution to nutrients, the relativeincrease in nitrogen and phosphorus from 150thousand person equivalents in the surfacewater above the diffuser is less than 15% and10% on average of the winter values of nitrateand phosphate in Faxaflói in 2002 (13.7 µMand 0.92 µM, respectively (Sólveig R.Ólafsdóttir 2006)). 150,000 personequivalents are the maximum amountexpected but the present value is only about100,000. During dry and warm summermonths when the dilution of the sewage itselfis minimal or only twofold (Guðjón AtliAuðunsson 2000 and 2002), the increase inconcentrations at the surface above thediffuser (in an area extending 20-30 m oneither side of the diffuser) amounts to lessthan 30% and 18% for nitrogen andphosphorus, respectively, from 150 thousandperson equivalents (using the measured percapita volume of 270 L/p.e./d, see chapter 4(Guðjón Atli Auðunsson 2000 and 2002).Therefore, even in the near field right abovethe diffuser the average elevation of thesenutrients is well below the limit set byOSPAR (2001) of 50% increase as a measureof significant shift up a trophic level.Applying this average dilution to chemicaloxygen demand, the near field depletion ofoxygen would be less than 2.5% resulting

21

from a discharge of 150,000 p.e., a decreasethat would require very extensive samplingand analysis to reveal. The sampling on02/02/2000 from 90,000 p.e. did not revealany significant decrease in oxygen, i.e. thedecrease was significantly less than 0.25%(95% CI).

Extended mixing and source region: farfieldThis is the region defined as zone B in theCSTT-model. The Brooks-solution (Brooks1960) to the diffusion equation with estimatedtransverse turbulent diffusion by semi-empirical means (i.e. the “4/3-law”) given ine.g. Roberts (1999a), Stacey et al. (2000),and Koh and Brooks (1975) may be used todefine this area. The solution of Brooksassumes negligible vertical diffusion orlongitudinal dispersion from a finite linesource in a steady current and is therefore aconservative estimate. Using a time periodof one tidal cycle to define the far field regionin an average current of 10 cm s-1 gives anellipse with a short axis of 1.7 km and a longaxis of 4.5 km, i.e. an area of about 24 km2.At these distances, the Brooks equationpredicts an additional dilution of a factor ofabout five at the edges of the ellipse andthereby the nutrients nitrogen and phosphorusare on average at levels less than 3 and 2%,respectively, of the winter values from150,000 p.e. or less than 6 and 4% during dryand warmest summer months.

Another way to look at the far field is toassume the water and sewage underneath theellipse to be well mixed as is done on theCSTT-model, and assuming further theresidual current to be 3 cm s-1 (neglectingturbulent mixing and water exchange bywinds, wave motion and tidal currents to havea very conservative estimate), then the waterexchange is 0.012 h-1 and the averageadditional or build-up concentrations ofnitrogen and phosphorus in this volume isabout 1 µM and 0.04 µM, respectively, from150 thousand person equivalents or less than10 and 5% of the winter values, well belowthe criteria of OSPAR (2001). Under these

assumptions, the average decrease in oxygenin this volume from 150,000 p.e. is atmaximum 0.12 mg/L or less than about 1.5%of the equilibrium level of oxygen neglectingdecrease by nitrification and increase byuptake of oxygen from the atmosphere andproduction during photosynthesis. Thus, eventhis conservative estimate of oxygen decreasewould be very difficult to detectexperimentally.

The question remains to be answered to whichdegree the nutrients in the far field may beconverted to phytoplankton biomass duringphotosynthesis. During the winter months(middle of October to middle of March),however, there is hardly any primaryproduction in Icelandic waters (ÞórunnÞórðardóttir and Unnsteinn Stefánsson 1977).Macoralgae may utilise nutrients for growthduring winter, however, where the growth iscontrolled by length of day, i.e. starts to growwhen daylight duration is less than a certainlimit, approximately 12 hours in late winter,and ceases to grow in the spring whendaylight duration is more than 12 hours. Thekelp tangle (Laminaria saccharina) starts togrow already in January while other speciesstart to grow in March (Sjötun andGunnarsson 1995; Karl Gunnarsson andAgnar Ingólfsson 1995). The maximumgrowth rate is usually in May. During thesummer months, when nutrients are at lowconcentrations, the kelp uses photosynthesisto produce polysaccharides, mainly mannitoland laminarin, i.e. compounds deficient innitrogen and phosphorus. Thus, the kelpclosest to the diffuser might utilise possibleadditional supply of nutrients during winterbut this addition is very limited as discussedabove.

Assuming total conversion and yield factorsfrom Tett et al. (2003), i.e. 1.1 mg Chl a permmol nitrogen and 30 mg Chl a per mmolphosphorus, the additional chlorophyll a inthe far field is at most 1-1.5 mg Chl a/m3.This amounts to less than 10-15% increase inthe average Cla a in the far field and muchless during the spring blooms, far below the

22

50% increase proposed by OSPAR (2001).Although these yield factors have been usedacross Europe (Tett et al. 2003), they have notbeen validated for Icelandic waters.For answering the question whether thephytoplankton is able to utilise the nutrientsand give maximum yield in Chl a, an estimateof the effective photosynthetic efficiency isneeded as well as information on radiation.Figure 8 shows the measured radiation inFaxaflói (station at Húsafell) showing June tohave the maximum PAR (photosyntheticallyavailable radiation) of about 21 µE m-2 s-1 onaverage. This maximum radiation is reflectedby 6% by the water surface (albedo) andfurther attenuated by Chl a and turbidity inthe water column, resulting in a total of 80%average decrease in PAR at the surface andthe euphotic zone of 20 m (using a radiationattenuation factor by k=1.7/Ds, Ds beingSecchi depth, and data on relationshipbetween Secchi depths and Chl a in Icelandicwaters from Kristinn Guðmundsson andKristín J. Valsdóttir (2004) and KristinnGuðmundsson et al. (2004). Furtherattenuation occurs as the light penetrates thesurface, somtimes called the hyperexponentialdecrease, due to different penetration

efficiency of some wavelengths. Thissubsurface decrease is commonly 35-40%(Tett et al. 2003) but is neglected here for aconservative estimate. Using the effectivephotosynthetic efficiency in Tett et al. (2003)of 0.006 d-1 (µE m-2 s-1)-1, the photosyntheticgrowth rate is 0.001 h-1, far less than theexchange rate in the area by the residualcurrent alone, 0.012 h-1. Therefore, thephytoplankton is only able to use a smallfraction of the additional nutrients dischargedby the STP at Ánanaust to build up biomass,provided the present form of the CSTT-modelis valid and that the effective photosyntheticrate holds true for Icelandic waters in which ithas not yet been validated.

From these results it may be concluded thatsecondary treatment of the sewage, decreasingnitrogen levels by 10% at best, provides littleextra benefit and the present primarytreatment alone will not result in a measurableincrease in primary production nor a decreasein oxygen saturation in the far field.Therefore, the present primary treatment issuitable for the STP at Ánanaust under Article6 of the UWWTD (C.E.C. 1991).

0

5

10

15

20

25

30

0 2 4 6 8 10 12 14

Month

PA

R,µ

Em-2

s-1

90% percentile

Median10% percentile

Figure 8. Median photosynthetically available radiation (PAR) in Faxaflói during 1990-2002 with 10 and 90% percentiles for eachmonth. PAR is calculated from data on total radiation measured by the Meteorological Office of Iceland using a conversion factorof 0.455 for the latitude of Faxaflói (http://www.atmos.umd.edu/~srb/par/04status.htm) and 1W m-2 PAR = 4.15 µE m-2 s-1 (valid for500 nm).

23

The distribution of faecal coliforms isdiscussed with in chapter 3 and suspendedsolids are described in chapter 5.

Long range effectsOne may use a simple tidal prism model toevaluate the dilution and flushing times (orturnover and half life) (e.g. Luketina 1998) ina larger area, here the Kollafjörður.Kollafjörður was defined above as the areainside the line from Grótta to Kjalarnes withan area of about 83 km2, volume at spring tideebb about 1.4 km3 and an average depth of 17m.Total freshwater inflow into Kollafjörður isabout 10 m3/s, mainly from Elliðaár (4.6 m3/s,average of 48 years), Leirvogsá (about 3.3m3/s) and Úlfarsá (1.5 m3/s) (National EnergyAuthority, Hydraulic Laboratory 2006).Assuming the concentrations of nutrients inall these rivers to be the same as in Elliðaár(winter time) (Sigurður Reynir Gíslason et al.1998), freshwater contributes 20 kg/d and 6kg/d of dissolved nitrate-N and dissolvedphosphate-P, respectively, to Kollafjörður.The total amounts of nitrogen and phosphorusare somewhat higher but these discharges are,however, small compared with the dischargesof the STPs at Ánanaust and Klettagarðar. In2004, the discharge from both STPs was fromabout 200,000 p.e. amounting to 3,400 kg/d ofnitrogen and 320 kg/d of phosphorus. In thefuture, the total discharge from these STPs isexpected to be twice as much or 400,000 p.e.in total.

Tidal prism models are very simplified pictureof an estuary (here the sewage effluent is the“river”) where the main simplification is theassumption of very well mixed watersvertically and horizontally during the wholetidal cycle. The models provide a firstestimate of possible effects, and as such haveproven a useful management tool. They arefrequently used when there is a lack ofcomprehensive data on currents andoceanographic parameters and due to theirsimplicity which enables the user to get anunderstanding of the major physical processesat work. If a return factor of 0.5 is assumed asproposed by USEPA in absence of any data

(USEPA 1985 as cited in Luketina 1998), theflushing time of Kollafjörður is about 9-10days on average. Additional flushing byresidual currents is estimated to be roughly ofthe same magnitude and the total flushingtime of Kollafjörður therefore about 5 days.Under these conditions, the increase innitrogen in the Kollafjörður would be about0.9 µM and phosphorus about 0.04 µM from200 thousand p.e. but 1.7 and 0.075 µM,respectively, from the future 400 thousandperson equivalents. These latter valuesrepresent 10-15% increase in winter nitratenitrogen and less than 10% increase in winterphosphorus values of Faxaflói, both wellbelow the 50% increase proposed by OSPAR(2001) to warrant caution for increase introphic status of the area. Nitrogen is thelimiting nutrient in Faxaflói and otherIcelandic waters (Sólveig R. Ólafsdóttir 2006;Unnsteinn Stefánsson and Jón Ólafsson 1991)and therefore the sewage nutrient that mightaffect the trophic status of the waters. Themaximum increase in Chl a is about 1 mg/m3

if phytoplankton could utilise this increase innitrogen level. However, the estimated waterexchange rate in Kollafjörður, about 0.01 h-1,is greater than the growth rate ofphytoplankton under the CSTT-assumptionsabove, about 0.001 h-1, and therefore thephytoplankton is only able to utilise a fractionof these nutrients for growth withinKollafjörður.With the estimated exchange rate ofKollafjörður, and neglecting nitrification aswell as oxygen uptake from air and oxygenproduction during photosynthesis, the oxygendecrease will only be about 0.2 mg/L onaverage when sewage from 400 thousandperson equivalent are discharged intoKollafjörður Bay or roughly 2% of the naturaloxygen levels in the area. This is well belowthe criterium set by CSTT (1997), 0.5 mg/Lwith oxygen above 7 mg/L.As regards nutrients and oxygen levels,secondary treatment of sewage where thenitrogen decrease is less than 10% will notresult in any environmental benefits greaterthan those already offered by the presentprimary treatment.

24

Source TN* (tons/day) TP* (tons/day) CODCr (tons/day)All rivers into Faxaflói 1.6 0.12 78Hvítá in Borgarfjörður 0.9 0.07 52Elliðaár, Reykjavík 0.02 0.006 0.8400,000 p.e.** 6.8 0.64 54*In case of rivers these values pertain to nutrients in dissolved state which is total amount used for sewage. **Untreated sewage.

Finally, the whole of Faxaflói Bay may beconsidered. The following table givesestimates of the total discharge into the bayassuming the number of person equivalentsfrom the Ánanaust STP will be 150,000 p.e.and 250,000 p.e. from the Klettagarðar STP orabout 400,000 p.e. in total. The contributionfrom sewage is estimated from the per capitavalues given in chapter 4 while contributionfrom all rivers is estimated from thecomposition of Hvítá in Borgarfjörður in theyears 1973-1974 (Sigurjón Rist 1986) andtotal direct flow of freshwater into Faxaflói inthe years 1966 to 1967 from UnnsteinnStefánsson and Guðmundur Guðmundsson(1978). The contribution from Elliðaár in TNand TP are estimated from its winter values(Sigurður Reynir Gíslason et al. 1998). CODin Hvítá is derived from the permanganatenumber by asuming CODMn ≈0.4CODCr butobviously the organic matter of rivers andsewage is of different nature.As discussed above (1.5), the inflow offreshwater by rivers of Southern Iceland intoFaxaflói is of the same order as direct inflowby rivers into the bay, which thereby roughlydoubles the contribution of freshwater in theinput of nutrients by rivers into Faxaflói.From the table above, it is seen that the inputof the nutrients nitrogen and phosphorusreleased by the expected 400,000 personequivalents from Ánanaust and KlettagarðarSTPs, is about 4 to 5 times the discharge fromall rivers that direcly enter the Faxaflói.However, the seawater entering the baycontains nutrients which account for its highprimary productivity. Using the estimatedflow of seawater into the bay in 1.4 and thewinter values of nutrients from Sólveig R.Ólafsdóttir (2006), the amount of nitrate-nitrogen inflow is 800-900 tons/d while theinput of phosphate by seawater is 120-150tons/d, exceeding by far the direct load by

rivers and sewage. This input of nitrate is alsopredicted by the empirical relationshipbetween primary productivity and nitrateinput found by Nixon et al. (1996) whereaverage primary productivity of Faxaflói isabout 300 g C m-2 y-1 (Þórunn Þórðardóttir1994). The effects of the input rates of thenutrients nitrate, phosphate and silicate fromrivers on their levels in coastal waters ofIceland within the 200 m depth contour havebeen assessed (Unnsteinn Stefánsson and JónÓlafsson 1991). According to theirassessment, the contribution from rivers isnegligible in the case of nitrate and phosphatebut the effects of dissolved silicate from riversare considered to be substantial.The longest estimate of residence time offreshwater in Faxaflói is about 100 days(Unnsteinn Stefánsson and GuðmundurGuðmundsson 1978). As the volume of theFaxaflói is 245 km3, the increase in nitrogenand phosphorus caused by sewage dischargefrom 400,000 p.e. is estimated to be 0.2 µMand 0.08 µM on average, respectively.However, the increase will be smaller sincethe flushing time is shorter than 100 daysmost of the time. This increase in nutrients,about 1%, will be impossible to detectexperimentally when seawater is analysed.However, the phytoplankton may utilise thisaddition resulting in increased biomass sincethe residence time of water in Faxaflói islonger than the growth rate of phytoplanktonunder the conditions of the CSTT-model (Tettet al. 2003). The relative increase in biomasswill be extremely small, however.

The residence time of sewage from 50,000p.e. that was released 500 m from the shore atthe STP at Ánanaust on 28/11/1995, wasestimated to be 15-22 hours and this release ofsewage had no effect on the oxygen-saturationof the seawater (Jón Ólafsson et al. 1996).

25

Additionally, on a transect along the piplene'sposition from Ánanaust 8/11/1996, when therelease of sewage was about one fourth of thaton 28/11/1995, the undersaturation of oxygennext to shore was only about 3% as comparedwith a station at 6000 m NW of the STP atÁnanaust (Jón Ólafsson et al. 1995). Thisinvestigation showed that the affected area ordistribution of sewage was best estimated bysalinity and silicate but organic matter, eitheras BOD5 or COD, was not a good tracer dueto fast mixing times and low sensitivity of themethods for BOD5 and COD (Guðjón AtliAuðunsson 1996).

Finally, it may be noted that in February ofboth 1991 and 1992, nutrients were analysedon a transect west of Akranes (Jón Ólafsson etal. 1994). In short, the results showed that theconcentration of nitrate was lowest at smalldepths and at similar level as seawater at thesame salinity south of the Reykjanespeninsula. It was therefore concluded that theanthropogenic influence was very small onthis transect or that dilution due to mixing ofsewage made that source undetectable.

From the discussion above, it is clear thateven in the near field, the risk ofeutrophication (increase in microplanktongrowth) or reduced oxygen levels (due to totalorganic matter, BOD or COD) is not at handin the recipient of the present primary treatedsewage from Reykjavík and neighbouringcommunities. This holds true for the futuredischarge of up to 400,000 personequivalents. Secondary treatment reducesnitrogen only to a small degree, the limitingnutrient in the receiving waters. Therefore,secondary treatment will only result in verylimited, if any, environmental benefit asregards eutrophication potential in therecipient.

The fate of suspended solids will be describedin chapter 5 while the possible effects ofcontaminants in the sewage on sediments andbiota are discussed in chapters 5, 6 and 7.

3. HYDRODYNAMIC RESEARCHAND DISPERSAL OF FAECALBACTERIA.

3.1 Introduction.According to the water quality standards inIceland, the following two conditions need tobe fulfilled regarding faecal coliforms atsewage disposal sites.

1. The number of faecal coliforms orenterococci outside the dilution areashall during at least 90% of the timebe under 1000 per 100mL for aminimum of 10 samples.

2. Where recreational areas are located atthe coastline or food processing plantsnear the shore, the number of faecalcoliforms or enterococci, shall during90% of the time be under 100 per100mL ouside the dilution areas for aminimum of 10 samples.

These requirements are additional to the onesfound in the Council directive 91/271/EEC(C.E.C. 1991) concerning urban waste-watertreatment which is also in force in Icelandwith some minor amendments.

In order to calculate the concentration offaecal coliforms and other pollutants fromsewage outlets in the sea north of Reykjavík,Iceland, a numerical model of surface waterflow and transport was constructed. Theoptimal length of the sewage outlets wasdetermined so that calculated concentrationsof pollutants fulfilled water quality standards.Calculated results were compared withmeasured data in order to calibrate the model.The calibration involved determining thefollowing parameters:

1. Current speed and direction in a largearea surrounding the proposed outlets.

2. Dispersion coefficients of pollutants inthe sea.

3. Sewage flow from the outlets.4. Concentration of faecal coliforms in

the sewage.

26

5. Concentration of other pollutants inthe sewage.

6. Lifetime of faecal coliforms in the sea.

In order to determine the above-mentionedfactors, extensive measurements wererequired, and are discussed in detail in thefollowing sections.

3.2 Flow modelThe flow model covers all of Faxaflói andtakes into account tidal and wind-inducedcurrents in the sea. The coastline fromSandgerði on the Reykjanes peninsula toMalarrif on the Snæfellsnes peninsuladelineates the inner model boundary. Theouter boundary is defined by a straight lineacross Faxaflói from Malarrif to Sandgerði.The boundary conditions at the outerboundary are estimated by extrapolatingmeasured tidal elevation in Reykjavíkharbour. The tidal amplitude and phase wereadjusted so that measured and calculated tidalelevation were in agreement. The results ofthese adjustments were that the amplitude onthe boundary was 93% of the measuredamplitude in Reykjavík harbour and the phaseshift across Faxaflóí was approximately 15minutes.

The flow model was calibrated against currentmeasurements taken in Faxaflói between 1967and 2000. The first current measurementswere made in Fossvogur (creek betweenReykjavík and Kópavogur, see map 2) andSkerjafjörður by the National EnergyAuthority’s Hydraulic Laboratory for the Cityof Reykjavík’s Environmental and TechnicalSector (National Energy Authority 1967). In1970, the Danish company Isotopcentralencompiled and carried out extensivemeasurements for the City of Reykjavík’sEnvironmental and Technical Sector inSkerjafjörður and in the sea north ofReykjavík. Among these were currentmeasurements at five locations made by theMarine Research Institute (Isotopcentralen,1971). Current measurements were made attwo locations in 1989 for the City ofReykjavík’s Environmental and Technical

Sector (Vatnaskil Consulting Engineers1989). In addition, current measurementswere performed at high tide under Gullinbrúbridge (across Grafarvogur) as well as off theisthmus of Geldinganes. The most extensivecurrent measurements in the area weresupervised by Vatnaskil Consulting Engineersfor the City of Reykjavík’s Environmentaland Technical Sector. The actualmeasurements were performed by the MarineResearch Institute and the National EnergyAuthority during the summer of 1994(Vatnaskil Consulting Engineers 1994) andincluded:

1. Measurements of tidal elevation inReykjavík harbour.

2. Continuous current measurements at theend of the proposed Ánanaust sewageoutlet, 4 kilometers from the coast.Currents were recorded for approximatelytwo months.

3. Current measurements at different depthsalong a line from Bygggarðsboði toGunnunes, made up of segments fromBygggarðsboði to Akurey, Akurey toEngey, Engey to Viðey, Viðey toGeldinganes and finally Geldinganes toGunnunes. Measurements were recordedover a period of 12 hours at low and hightides.

4. Current measurements at different depthsalong the length of the proposedÁnanaust outlet pipe. Measurementswere recorded over a period of 12 hoursat low and high tides.5. Current measurements atdifferent depths at the end of the proposedÁnanaust sewage outlet. Thesemeasurements were taken for comparisonwith the continuous measurementsdescribed in number 2 above.Measurements were taken over a periodof six hours.

6. Wind speed and direction measurementstaken from a two-meter high mast inAkurey during the same time period asthe continuous current measurements.

In 1998 during the design of the Klettagarðarsewage outlet, it was decided to performadditional current measurements in the seanorth of Reykjavík (Vatnaskil Consulting

27

Engineers 1999). Current flow was measuredalong three cross sections (Table 3.1) andcurrent speed was measured at differentdepths at the end of the proposed outlet pipe.

Tavle 3.1 Current flow measurements

Cross section Time period TidesSkarfaklettur-Viðey 28-29 Oct ‘98 LowViðey-Engey 28-29 Oct ‘98 LowEngey-Laugarnes 28-29 Oct ‘98 Low

Skarfaklettur-Viðey 3-4 Dec ‘98 HighViðey-Engey 3-4 Dec ‘98 HighEngey-Laugarnes 3-4 Dec ‘98 High

Similar current measurements have beenconducted in Hafnarfjörður in associationwith the design of sewage outlets. TheMarine Research Institute performed currentmeasurements at four locations inHafnarfjörður: off Garðar on Álftanes,Langeyri, Hvaleyrarhöfði and near Helgasker(west of Hvaleyrarhöfði) (VatnaskilConsulting Engineers 1990). In 1994 theNational Energy Authority performed currentmeasurements for the city of Hafnarfjörður.Current measurements were taken at differentdepths along a line from Hvaleyrarhöfði toHelgasker and then back to the mainland atÁlftanes. Measurements were made over a12-hour period during low and high tides(Vatnaskil Consulting Engineers 1995).Finally, current measurements were made bythe Marine Research Institute at four locationsnear Keflavík and Njarðvík in connectionwith sewage disposal studies in the area(Vatnaskil Consulting Engineers 1993).

All of the above-mentioned currentmeasurements have in one way or anotherbeen used to calibrate the flow model. Inorder to calibrate current speed and direction,both the bottom friction coefficient (C) andthe wind shear stress coefficient (CD) wereadjusted. The following values gave the bestresults:

C D20 1 6/ where D is the average depthCD 8 10 4

Both values are within published limits. Ascan be seen, the flow model has been

calibrated against extensive currentmeasurements in the sea north of Reykjavík.

3.3 Transport model.As mentioned previously, in order to calibratethe transport model it was necessary todetermine the dispersion coefficients ofpollutants in the sea, volume of sewage flow,concentration of faecal coliforms and otherpollutants in the sewage and lifetime of faecalcoliforms in the sea.

Dispersion coefficients of pollutants weredetermined by fitting measured and calculatedpollutant concentrations in the sea. The firstmeasurements of pollutant concentrationswere made by the National EnergyAuthority’s Hydraulic Laboratory in 1967. Acolored dye was released into the sea inSkerjafjörður and its dispersion measured. In1970 Isotopcentralen conducted similarexperiments (Isotopcentralen 1971) usingradioactive tracer. In 1995 the MarineResearch Institute conducted extensivemeasurements of nutrients in the sea nearÁnanaust (Jón Ólafsson et al. 1996). Thesemeasurements were primarily used for thecalibration of the dispersion coefficients.Measured and calculated concentrations fitbest with dispersion coefficients of 10 m2/s inthe current direction and 5 m2/s perpendicularto the current direction.

In order to determine the amount of sewageflow expected from the outlet, the City ofReykjavík’s Environmental and TechnicalSector performed flow measurements at thefollowing four stations:

1. Pumping station at Laugalækur2. Pumping station at Ingólfsgarður3. Pumping station at Ánanaust4. Manhole at Neðra-Breiðholt

The results of the measurements showed thatthe sewage flow, not including rainwater orheating water, was approximately 275liters/person/day (Vatnaskil ConsultingEngineers 1991, 1992, and 1996).

28

The concentration of faecal coliforms in thesewage was measured at the four above-mentioned stations during the same time asthe flow measurements. The average dailyconcentration was approximately nine millionfaecal coliforms per 100 mL sample.Concentrations of other pollutants were alsomeasured at the four locations.

The lifetime of faecal coliforms in the sea wasdetermined by fitting measured and calculatedconcentrations. The first measurements offaecal coliforms were performed byIsotopcentralen (Isotopcentralen 1971).Measurements were later performed by theIcelandic Fisheries Laboratory (1991) in thesea north of Reykjavík. Severalmeasurements were taken in Grafarvogur in1984 (Vatnaskil Consulting Engineers 1984).The most extensive measurements of faecalcoliforms concentration were taken in the seanear Laugarnes, Ingólfsgarður and Ánanaustin conjunction with the above-mentionedmeasurements of flow at the pumping stations(Vatnaskil Consulting Engineers 1991, 1992,and 1996). The results showed that thelifetime of faecal coliforms is longer duringwinter months when there is less sunlight thanduring summer months. The lifetime offaecal coliforms, T90, is the time it takes for90% of the faecal coliforms population to die.Table 3.2 shows the measured T90 values on amonthly basis.

Table 3.2 T90 values

Month T90, hoursJanuary 9February 8*March 5April 4*May 3*June 1*July 2August 3*September 5*October 8November 9December 10*Estimated

3.4 Model validationIn 1999, Vatnaskil Consulting Engineerspublished a report for the City of Reykjavík’sEnvironmental and Technical Sectordescribing the above-mentioned flow andtransport model. The report includedcalculations of dispersion of pollutants fromthe current and proposed outlets in Ánanaustand Klettagarðar, respectively. Theconcentration of faecal coliforms,biochemical oxygen demand (BOD), chemicaloxygen demand (COD), nitrogen,phosphorous and suspended particles havebeen calculated. Figure 9 shows thecalculated faecal coliforms concentrationfrom 130,000 p.e. from each outlet.

A monitoring programme was developed forthe Ánanaust outlet where concentrations offaecal coliforms and other pollutants would bemeasured at determined intervals in order tocompare with concentrations predicted by thetransport model, and therefore test its validity.The first set of measurements were taken inFebruary 2000 at 31 locations around the endof the Ánanaust outlet. A similar monitoringschedule was developed for the Klettagarðaroutlet after its construction.

Additional measurements of concentration offaecal coliforms were performed on April 13,2000 at 35 locations covering a moreextensive area around the end of the Ánanaustoutlet (Vatnaskil Consulting Engineers 2000).

The results show that the calculatedconcentrations of faecal coliforms comparewell with measured concentrations, takinginto account uncertainties in both themeasurements and calculations.

3.5 Summary and conclusionsThe flow and transport models have beencalibrated against extensive measurementsspanning a 30-year time period. The resultsshow that the Ánanaust and Klettagarðaroutlets shown in Figure 9 fulfill all of thedemands of the water quality standards. Insummary, the measurements and calculationsshow the following:

29

1. The measured net current is 1-5cm/sec and flows in an easterlydirection at the proposed outlet sites.The direction changes farther out intoFaxaflói as depth increases.

2. Calculated current speed and directioncompare well with measurements.The current at the proposed outlet sitesis roughly 5 to 30 cm/sec in an east-west direction.

3. The proposed outlets satisfy thedemands of the water quality standardsfor concentration of faecal coliforms.

4. With respect to nutrients and organicmatter, the dilution zone is defined asan area confined by 50 meters distancein all directions around the diffuser.Within this dilution zone, the excessconcentration resulting from thedischarge has reached a smallpercentage increase in the backgroundconcentration in the sea. Considerablewater exchange takes place at thelocation of the outlets.

5. The main conclusion of these studiesis that the recipient north of Reykjavíkcan be classified as less sensitive asdefined by the water quality standards.

Scale: 1:100,000

Calculated concentration of fecal-coli

Calculated concentration withinthe contour lines is higher thanthe given value 10% of the timeor more

Fecal-coli per 100 mL

1000

100

0 2000 4000m

Aku re y

Eng eyGelding anesV

iðe y

Gu nnu n es

Se ltj arna rnes

R eykjaví k

Ánanaust

Klettagarðar

Figure 9 Calculated dispersal of faecal coliforms at the disposal sites for 130,000 p.e. from each outlet.

30

4. CHEMICAL COMPOSITIONOF THE SEWAGE INREYKJAVIKThe composition and behaviour of sewage inReykjavik have been studied extensivelywhereupon a good baseline has beenestablished for later monitoring as well as abase for decisison-making regarding sewagetreatment and for the impact assessment of thesewage disposal into the recipient (GuðjónAtli Auðunsson 1992, 1994a, 1996, 2000, and2002).

4.1 Main characteristics of the sewage inReykjavíkThe main characteristic of sewage in thepumping stations of Reykjavik is its greatdilution. This dilution is both diurnal due todaily activities of humans and stronglyseasonal which mainly relates to hydrothermalwater used for space heating. Most of thiswater ends up in the sewage, either directly incomposite systems or after its heat has beenused up in dual systems (Guðjón AtliAuðunsson 2000). At the Ánanaust STP, thedilution is about two times in summer while itis up to five times in winter (Guðjón AtliAuðunsson 2000 and 2002).

This dilution results in relatively lowconcentrations of the various chemicalconstituents but the composition of thediluting water may also affect the compositionof especially ash content and the content oftrace elements in the sewage. The dilutionalso results in short residence times in thesewage system.

Other peculiarities of the sewage is that theinput of hydrothermal water adds to thesilicate content and results in relatively highpH and temperature.

Studies on the treatment of sewage in thecatchment area of the Ánanaust STP and theplant itself, showed that the cleanup oforganic matter (CODCr) was 20% and 15%with respect to suspended matter (Guðjón AtliAuðunsson 2000). The same treatment isapplied in the Klettagarðar STP.

4.2 Main chemical constituents of thesewage

Per capita values of the main nutrients ofuntreated sewage are shown in the tablebelow.

Constituent Per capita valueg/p.e./ d

Source of information

Biological oxygen demand, BOD5 60 Definition of p.e. (used to calculate per capita values)Chemical oxygen demand, CODCr 135 Evaluated from the definition of p.e. and direct

experiments in Skolpa (2.25xBOD5)Total nitrogen, TN 17 Faxaskjól, Nov. 1995, 50,000pe.

[March, July and Oct. 1991: Breiðholt (3.8 thous. pe): 35;Laugarnes (62 thous. pe): 25; Ingólfsgarður (10.9 thous.pe): 40. Gelgjutangi Sept. 1993 (8.5 thous. pe): 97 g/pe d]

Total phosphorus, TP 1.6 Faxaskjól Nov. 1995, 50 thous. pe.[Breiðholt (3.8 thous. pe) : 2.8; Laugarnes (62 thous. pe) :1.4; Ingólfsgarður (10.9 thous. pe) : 2.2. Gelgjutangi Sept.1993 (8.5 thous. pe): 1.9 g/p.e./d]

Total volatile nitrogen, TVN 6.3 Faxaskjól Nov. 1995: 50,000 p.e.[Breiðholt March, July and Oct. 1991 (3.8 thous. p.e.) andGelgjutangi Sept. 1993 (8.5 thous. p.e.): 13g/p.e./d]

Organic matter (TS-ash) 125 Breiðholt March, July and Oct. 1991, Gelgjutangi Sept1993.

Suspended solids, SS (<1.6µm) 65 Breiðholt March, July and Oct. 1991, Gelgjutangi Sept1993.

31

The per capita volume in the STP atÁnanaust has been found to be 270 L/p.e./d, avalue in agreement with estimated usage ofcold (150 L/p.e./d) and warm water (120L/p.e./d) for household purposes inReykjavík, i.e. excluding warm water forspace heating (Guðjón Atli Auðunsson 2000and 2002).

Extensive analyses have been carried out forfaecal coliforms in sewage (Vatnaskil 1991and 1996). In most of these samples CODand/or BOD were analysed at the same time(Guðjón Atli Auðunsson 1992 and 1996).From these measurements, altogether 119pairs, a good per capita value of faecalcoliforms could be estimated and found to be1.31010 MPN/p.e./d with 10 and 90%percentiles being 0.51010 and 3.81010

MPN/p.e./d, respectively. This value agreesvery well with a per capita value calculatedfrom data for the STP of Hveragerði 2004(averages of monthly sampling)(Hveragerðisbær og Háskólasetrið íHveragerði 2004).

4.3 Trace elementsThe total concentrations of trace elementswere determined in the sewage of the pumpstations of Ingólfsgarður and Laugarnes inaddition to a manhole in Breiðholt in March,July and October 1991 (at two-hour intervalsfor 24 hours each time), but organic

micropollutants were also analysed in thesamples from Laugarnes (Guðjón AtliAuðunsson 1992). Analyses of the total traceelement concentrations were also made on 24hour composite samples of sewage from thepump station at Gelgjutangi in September1993 (Guðjón Atli Auðunsson 1994a) and afew samples from the STP of Ánanaust 2000(Guðjón Atli Auðunsson 2005 and 2006).

Per capita valuesAll the measurements of trace elements in thesewage of Reykjavík are summarised as percapita estimates in the table below.

It was noteworthy that in the 1991investigation that the per capita values of thetrace elements (copper, lead, cadmium, zinc,arsenic, and mercury) in all three stationswere proportional to the per capita values ofwater volumes. This, in addition to the factthat there was not a significant difference inconcentrations in these three stations at thethree different seasons, renders the hypothesisreasonable that the source of the traceelements is not solely from households butrather the water itself after transport by thedistribution system, weathering of soils anddrainage from roofs and streets. In thiscontext, it may be mentioned thathydrothermal water may be quite corrosive formetal pipes.

Trace element Per capita value, mg/p.e./dIron (Fe) 2200Selenium (Se) 0.2Chromium (Cr) 5Silver (Ag) 1.3Arsenic (As) <2Zinc (Zn) 85Nickel (Ni) 1-1.5Copper (Cu) 12Lead (Pb) 8Mercury (Hg) 0.15Manganese (Mn) 60Cadmium 0.2

32

In the examination of sewage water in thepump station at Gelgjutangi in 1993 (copper,lead, cadmium, zinc, arsenic, mercury,chromium, nickel, silver, iron, andmanganese) all per capita values were foundto be lower than the lowest in the 1991investigation (Breiðholt). Only copper wasfound in comparable concentrations.Noticeable were much lower per capitavalues of cadmium and especially lead.The per capita values for inorganic traceelements in Reykjavik are high whencompared with sewage water in Europe,particularly in the case of zinc. Even thoughthe per capita value of zinc was lower in theexamination at Gelgjutangi 1991, it isnonetheless higher in Reykjavik than in mostother places. The reason for this high percapita value for zinc may be due to extensiveusage of galvanised and corrugated roof andgutter materials in Reykjavik in addition togalvanised pipes for water, lamp posts, carparts etc. The high flow of water and thepotential relationship between per capitavolumes and trace elements is a possibleexplanation of these relatively high per capitavalues.

ConcentrationsThe levels of trace elements in the sewagewater are very low and one has to comparethem to maximum permissible levels indrinking water to find similar concentrations.The lead level in sewage water is only about5% of the permissble level in drinking waterwhile arsenic, chromium cadmium, silver, andcopper are about 10% of their maximumlevels or guidance concentrations in drinkingwater. Mercury, however, is at 20% of itsmaximum level in drinking water and zinc at50% of its guidance value for drinking water.Only iron and manganese are above theirmaximum levels in drinking water, ten andthree times higher, respectively. The reasonfor these high levels in sewage water isprobably dissolution of the pipes in the waterdistribution system since for example castiron contains manganese. No maximumlimits or guidance values do exist for thesetwo metals in sewage water.

A comparison with undiluted sewage water inStockholm 1989 (Stockholm Vatten AB,1990), where the same trace elements wereanalysed as at Gelgjutangi (except for arsenic,iron and manganese), reveals lower or equallevels in Reykjavik. Levels are similar (Cd),lower (Cr, Ag, Zn, Ni, Pb, and Hg) or muchlower (Cu) than in other Nordic countries(Nordisk Ministerråd 1993). Recentinvestigations of the sewage in Ánanaust STPand drainage water in Reykjavik (Guðjón AtliAuðunsson, 2005 and 2006) confirm the lowlevels found in the sewage water inGelgjutangi. It is worth noting that the levelof cadmium in the Gelgjutangi examination isequal to the level of cadmium found in tapwater in central Reykjavik in 1986 (JónÓlafsson 1987) and the level of zinc in thesame tap water sample was only one fifth ofthe zinc concentration found in the sewagewater in the pump station at Gelgjutangi.These levels in the tap water stem mostprobably from the dissolution of pipelines inthe water distribution systems. At theirsources (both cold and warm water), thelevels are, however, 10-1000 times lower thanin the sewage water except for mercury whichmay be at the same level to ten times lower inwarm water than in the sewage water. Inseawater the levels of these trace elements are1000 times lower than in sewage wateralthough copper is only at 10-500 times lowerlevel and cadmium 100 times lower.

Finally, the initial 1000-fold dilution ofsewage in the near field from the STP atÁnanust, takes all trace elements except silverto lower or much lower levels than found inNorth-Atlantic seawater.

4.4 Organic micropollutantsPCBs, AOX (Adsorbable organic boundhalogens), EOX (Extractable organic boundhalogens), total phenols, anionic surfactants(methylene blue active substances, CTAS),and nonionic surfactants (cobaltothiocyanateactive substances) were analysed in sewagewater from the pump station at Gelgjutangi in1993 (Guðjón Atli Auðunsson 1994a).

33

PCBs, AOX, anionic surfactants, andnonionic surfactants were analysed in thesewage water of the pump station atLaugarnes in March, July, and October 1991(Guðjón Atli Auðunsson 1992).

PCBs All results for PCBs in sewage waterof the Gelgjutangi pump station 1993 werebelow the detection limits or less than0.05µg/L, which is half the permissible levelin drinking water. This means that the percapita value is less than 0.06 mg/p.e./d to becompared with 2.1 mg/p.e./d in the pumpstation at Laugarnes 1991 and about 1.2mg/p.e./d in Stockholm 1989 (StockholmVatten AB, 1990). In the sewage of theLaugarnes pump station, the levels of PCBswere found to be from being less than 0.05µg/L up to 6.6 µg/L. From these results itseems that the sources of PCBs were relatedto the older districts of Reykjavík rather thanindustrial areas of Reykjavik.Research on the possible sources of PCBs wascarried out in 2000. This showed very lowlevels in the sewage water of the ÁnanaustSTP on a dry summer day (below or a slightlyabove the detection limit, which now was tentimes lower than in 1991-1993 or 0.005µg/L). The level of PCBs had fallendramatically in the sewage water of the pumpstation at Laugarnes (Guðjón Atli Auðunsson2001a, 2005 and 2006). The per capita valueof PCB153, the congener usually highest ofthe PCBs in nature, was less than 5 µg/p.e./d(Guðjón Atli Auðunsson 2006).

AOX and EOX AOX-substances in the pumpstation at Gelgjutangi 1993 were on averageat a level of 51±23 µg/L (from 18 to 90 µg/L).This is a similar concentration as may befound in ground water around Stockholm(Stockholm Vatten AB, 1990), but twice thelevel found in sewage water from the pumpstation at Laugarnes in 1991, which was in theinterval from <10 µg/L to 59 µg/L. However,the per capita value of AOX at Gelgjutangi islittle less than 60 mg/p.e./d which is twice theper capita value in Laugarnes 1991, 27mg/p.e./d, and Stockholm 1989, 25 mg/p.e./d.The level of AOX in sewage water in

Stockholm is twice the ground water level inthat area and its source accounted for bybleaching of clothes with chlorine-solutions.The EOX-analysis gave a weekly average of22±15 µg/L (from <10 µg/L to 47 µg/L) atGelgjutangi, about half of the AOX, which isa relatively high ratio since EOX is usually inthe range of 2-20% of AOX. The twofoldlevels of per capita values at Gelgjutangicompared with Laugarnes indicate that thesource of these substances at Gelgjutangi maybe for example organic solvents fromindustry. No further studies have been carriedout to elucidate the nature of the compoundsincluded in the EOX but there is reason tobelieve that EOX-compounds may indeedaccumulate in organisms. However, the useof AOX in environmental science has beenquestioned in a recent review on this summaryparameter (Müller 2003).

Anionic surfactants Anionic surfactants inthe pump station at Gelgjutangi ranged from0.5 to 1.0 mg/L, which is ten times the levelsfound in Laugarnes in 1991, 0.09-0.4 mg/L.The per capita value, 800 mg/p.e./d, isfurthermore nearly ten times the per capitavalue in Laugarnes, 75-100 mg/p.e./d,indicating industrial sources. The per capitavalue in Stockholm 1989 was, however,higher than the per capita value atGelgjutangi or 1200-2300 mg/p.e./d. EU(80/778/EC) has a maximum level of 0.2mg/L in drinking water.

Nonionic surfactants Nonionic surfactants inthe sewage water of the pump station atGelgjutangi ranged from 0.2 to 0.8 mg/L,resulting in a per capita value of 0.5 g/p.e./d.Both levels and per capita values are aboutsix times higher than those detected in sewagewater in the pump station at Laugarnes in1991. Similar levels for nonionic and anionicsurfactants in the sewage waters are in linewith their production in Iceland and becauseof their use in car washing for example itcomes as no surprise that their levels arehigher in Gelgjutangi than Laugarnes.

34

Phenols The total amount of phenols in thesewage of the Gelgjutangi pump station wasin the range <5 µg/L to 24 µg/L, which is 25times the maximum EU level for drinkingwater. Individual chlorophenols in domesticsewage in Stockholm 1989 were in the range0.01 to 0.45 µg/L and therefore much highertotal concentrations. The EU has much higherpermissible maximum levels for somespecific industries (Commission Directive86/280/EC of 12 July 1986). Thus. themonthly average of pentachlorophenol may beup to 1000 µg/L in discharge water from theproduction of sodium pentachlorophenol anda daily average of 2000 µg/L. Forpentachlorophenol, the EU has a quality targetaccording to this Directive, which for surfacewaters, estuarine waters, and coastal waters isless than 2 µg/L. A high ratio of EOX toAOX calls attention to the nature of thephenols that comprise the total phenol contenthere analysed, where it is especially importantto examine whether pentachlorophenols arepresent.

5. SUSPENDED SOLIDS ANDSEDIMENT TRANSPORT:BEDLOAD AND SUSPENSIONLOAD

5.1 IntroductionSuspended matter behaves differently inseawater than dissolved constituents of thesewage. However, models for predicting theirfate are less well established than those formicrobes and dissolved compunds althoughmodels for suspended solids are obtainingmore attention lately, especially due to theemerging problems associated with the heavyload of organic solids from aquaculture(Cromey et al. 1998 and 2002). Predictionsof the fate of suspended solids involve all theproblems of dissolved pollutant modelling butwith the added problems of variable particlesizes, density of particles, colloidal andflocculant properties, and response of bottomsediment and benthic infauna to settlingparticles (nutrients and pollutants).

Additionally, the fate of the suspended matteris also dependent on the shear stress on theseabed by currents slightly above the bottom(0.5-1m), the bed form and type, the extent ofconsolidation with natural sediments and theability of water currents to resuspend anderode the sediments at high tidal velocitiesand storm events. Modelling of the fate ofsuspended solids from ocean outfalls havebeen approached by widely differenttechniques (Mearns and Word 1982; Koh1982; Farley 1990; Neves et al. 1995 and2002; CSTT 1997; Cromey et al. 1998) but itmay be concluded that they are still only intheir infancy and modelling has to beaccompanied by monitoring and verifyingstudies.

Chemical nature of suspended solidsStudies on the suspended solids at ÁnanaustSTP showed that they are about 90% organicmatter by weight and their share in the totalorganic matter in the sewage is about 40%(Guðjón Atli Auðunsson 2000). According toAustralian data, the composition of suspendedsolids is 35% organic carbon, 5% nitrogen,1% phosphorus, and 5% silicate (Bickford1996). Assuming the same composition ofsuspended solids in Reykjavík and using theper capita values in 4.2, approximately 20%of the total nitrogen and 40% of the totalphosphorus are found in the suspended solids.Thus, primary treatment, which should reducesuspended solids by at least 50% and BOD5by at least 20% according to UWWTD (CEC1991), will reduce nitrogen by only about10% and phosphorus by about 20%. Thus,sedimentation by gravitation in primarytreated sewage will hardly satisfy both thesecriteria simultaneously since a 20% decreasein COD is accompanied with only about 25-30% reduction in suspended solids(Christoulas et al. 1998). However, reductionin suspended solids by 90% in secondarytreatment is optional according to theUWWTD (while 70-90% reduction in BOD5and/or 75% reduction in CODCr is required),which would result in 15-20% and 35-40%reduction in nitrogen and phosphorus,respectively. Since nitrogen is the limiting

35

nutrient for algal utilisation as discussedabove, its reduction is of concern in therecipient. Finally, it may be mentioned thatpollutants in sewage (microbes as well asorganic and inorganic micropollutants) aremainly associated with particulate matter inthe sewage water.

Effects of suspended solids in the marineenvironmentAs shown in chapter 2, where total amount oforganic matter was discussed in relation topossible oxygen depletion, the sewage solidswill hardly have any detectable effects in therecipient if they are assumed to be solely insuspension (oxygen depletion was based ontotal organic matter by way of COD). Even inthe near field of the initial dilution, themaximum concentration of COD is wellbelow the amount of 1 mg/L, i.e. a level atwhich experiments show no effects of COD inseawater (Quetin and De Rouville 1986) and adecrease in oxygen will not be detectable. Ifthe solids settle down onto the sediments,however, they may affect both the benthicbiota and sediment chemistry. The followingconsequences may occur, see for exampleCSTT (1997), most likely interacting but theirrelative importance will vary according to thenature of discharge and the nature of thereceiving environment:

i) Physical: Particles in sufficientquantity may induce changes inspecies composition or relativeabundance as a result of physicalinterference with e.g. feeding or tubebuilding activities.

ii) Physico-chemical: At least in stableareas, a buildup of organic matterwithin sediments may, throughmicrobial actions, result in increasedoxygen demand. Chemical changesassociated with anaerobic metabolismmay increase the toxicity of sediments,resulting in favouring of more tolerantor adaptable taxa over others.

iii) Chemical: Accumulation withinsediments of certain particle-boundcontaminants, if present in sufficientlyhigh concentrations in sewagedischarge, may induce a toxicresponse in “sensitive” species.

iv) Organic: Inputs of organic matter mayprovide a direct food source for certainbenthic macrofaunal taxa, which maythen be favoured over others.

These effects are often seen by examining thebenthic community (CSTT 1997; Mearns andWord 1982) where the guidelines of Pearsonand Rosenberg (1978) are used whereincreased biomass, decreased biodiversity anda shift from the more sensitive suspension-feeding to the more tolerant deposit-feedingbenthic communities are among the responsesobserved with decreased distance from thesource of organic matter. These guidelines ofPearson and Rosenberg are best suited forareas of low turbulent energy, whereas areasof high energy like the disposal areas ofsewage from Reykjavík, are less amenable topredictions as they include gravelly seabedsand disperse the effluent constituents moreeffectively. In this regard CSTT (1997)defined adverse effect when primarytreatment induces change in communitystructure at a distance greater than 100 metresfrom the outlet, effects which would not haveoccurred had secondary treatment beeninstalled. As a working tool in theseinvestigations, the Infaunal Trophic Index(ITI) is frequently applied. This index isbased on four groups of feeders, i.e.suspension feeders (group I), surface detritusfeeders (group II), surface deposit feeders(group III), and subsurface deposit feedersincluding the characteristic Capitella capitataat high organic loads (group IV) (Mearns andWord 1982). The index takes on values fromzero (domination by subsurface deposit-feeding invertebrates) to 100 (domination bysuspension-feeders) at unaffected areas. Theindex has been shown to be inverselyproportional to the concentration of organic

36

matter in the sediments (Mearns and Word1982) and correlate well with the betterknown Shannon-Wiener index (Cromey et al.1998). Thus, ITI between 80 and 100indicates unaffected area (suspension-feeders), ITI between 60 and 80 occur nearoutfalls and non-outfall sites and is stillnumerically dominated by detrital-feederswhile ITIs between 30 and 60 designatemoderately affected or “changed” areas(surface deposit-feeding infauna), and below30 (dominated bu subsurface deposit-feeders)the ITI reflects “degraded” area. Empiricalrelationships were found between the totalemission of suspended matter from oceanoutfalls and both the elliptical area affected aswell as the total excess standing crop oforganisms resulting from the organic load insoft-bottom Californian sites (Mearns andWord 1982). Thus, like all models proposedhitherto, this model applies only to the moresensitive soft-bottom areas in relatively lowturbulent areas. Additionally, this very simpleempirical model does not take any currents intconsideration, and thus must be oversimplisticin spite of astonishingly good correlation foreight outfall sites in southern California.These relationships were advised by the initialCSTT in 1994 (CSTT 1994) to be used forUK-outfalls. However, in its revised form in1997 it was stated that these relationships had“overpredicted the benthic effects” but thatthese relationships could still be used to

predict the worst case scenario (CSTT 1997).It is worth while to apply these relationshipsto the Ánanaust site but emission rates of thepresent primary treatment is similar to thelowest rate found in southern California forwhich the model was founded. The total areaoccupied by total excess biomass, areas withITIs lower than 60 and lower than 30 togetherwith extensions of the areas on both sides ofthe diffuser, i.e. the long axis of the ellipses,are given in the following table.The table shows that even in this worst casescenario, the model does predict the affectedarea to extend to less than 100 metres on eachside of the diffuser. Therefore, according toUK guidelines, secondary treatment of sewagewould not impart any environmentalimprovement farther away than 100 m fromthe diffuser and, therefore, primary treatmentof sewage is deemed sufficient even at thisstage of evidence. Secondly, the area ofpossible degradation according to this model,i.e. ITI<30, would not be observed even in theclosest vicinity of the diffuser. Also evidentfrom this model prediction is a much smalleraffected area than could be foreseen fordissolved constituents of the sewage, chapter2.1.

As shown below, even these low values arenot at all likely to occur in the recipient ofÁnanaust and Klettagarðar.

Person Total emission Total excess Total Biomass Area Long axis Area Long axisequivalents of SS biomass area density ITI<60 ITI<60 ITI<30 ITI<30

p.e. tons/year tons wet weight km2 g/m2 km2 m km2 m100,000 2373 12.3 0.98 12.6 0.031 31 0.0005 0.5150,000 3559 25.0 1.44 17.4 0.074 74 0.0014 1.4

37

Physical behaviour of suspended particlesThe physical behaviour of the suspendedsolids is of considerable importance. Severalinvestigations all over the world have shownthat between 10 and 20% of the suspendedsolids of sewage may settle out of suspensionin the water column (Berelson et al. 2002;Baker et al. 1995; Mearns and Word 1982)while about 80% remain in suspension andabout 3% of SS of primary treated sewagemay rise to the surface (Baker et al. 1995).Therefore, at least about 80% of thesuspended solids are transported from therecipient in a similar fashion as dissolvedconstituents albeit at a somewhat slower rate.However, particulates decay in the watercolumn and CSTT (1997) assumes a decayrate of 0.3 d-1 while Farley (1990) assumes0.05 d-1 in cold water and 0.1 d-1 in surfacewater off southern California. The particlesthat settle are heterogenous in compositionand irregularly shaped. Additionally, andmore importantly, Baker et al. (1995) showedby experiments that 90% of the sinkingparticles have a diameter less than 80 µm andLavelle et al. (1988) showed experimentallythat 80-85% of the particles in sewage havediameters below 63 µm and in UK, more than90% of particles in primary treated sewage areless than 100µm (Cromey et al. 1998). Thisis of great importance since particles smallerthan about 100µm are not transported bybedload if deposited but taken directly intosuspension and, when in suspension aspertains to all sewage particles from the jetsof the diffuser, they are transported very longdistances with very small currents before theyare deposited. Current velocities of 1 cm/s orhigher are sufficient to keep mineral sedimentparticles less than 100 µm in suspension (seetextbooks on physical sedimentology, e.g.Anon. 1999) and to retain them in suspensionas long as the current is above 1 cm/s. Thisholds true for mineral particles which haveconsiderably higher density and therebygreater settling velocities than the suspendedsolids normally found in sewage. As anexample, a mineral particle of 100 µm has asettling velocity of about 10 mm/s (e.g. Anon.1999) while the highest velocity recorded for

sewage particle from untreated and primarytreated STP is 2.4-3 mm/s (Lavelle et al.1988; Baker et al. 1995). Furthermore,inorganic sediment particles less than 63µm(>80% of sewage particle sizes) only need acurrent of 0.7 cm/s to stay in suspension and acurrent of 0.1 cm/s keeps sediment particlesless than 20 µm in flowing suspension (Anon.1999). Therefore, the predictions belowassuming all the settlable sewage particles tosettle at velocities less than 1 cm/s based onbehaviour of mineral sediment particles, is avery conservative approach.Also of concern regarding sewage particles istheir possible flocculation into larger particleswith greater sinking rates. Flocculation willnot happen, however, if the concentration ofsuspended solids is below about 7 mg/L(Lavelle et al. 1988), a condition occurringinstantaneously when the sewage leaves thediffuser from Ánanaust resulting in levelsbelow 1 mg/L. Additionally, the experimentsdone by Baker et al. (1995) were carried outwith indiluted sewage and thus givingmaximum sinking rates and thereforemaximum amounts of solids that might settleto the bottom.

Possible settling of suspended solids at theÁnanaust siteContinous current measurements in the calmperiod of July and August 1994 showed thatspeeds lower than 1 cm/s at 1m above theseabed occurred for less than 22% of the time(Guðjón Atli Auðunsson 2005), see figure 10(current measurements were performed at 15m above the seabed but calculated for 1 mabove the seabed by using the common vanKarman function assuming the lowestmeasureable speed of 1.1 cm/s to be zero at 1m above the seafloor). This is alsomanifested by the fact that the seabed is ofsand and gravel in a relatively thin sheetabove a hard bottom (Kjartan Thors 1994 and2003) as a result of frequent currents fastenough to remove particles greater than 100µm from the bottom, i.e. currents greater than10-20 cm/s. This is also borne out by theslow burial rate of organic matter in the areaor less than 1 g C m-2 y-1 as elucidated in

38

Guðjón Atli Auðunsson (2001b) with aresulting flux of organic matter to thesediments of less than 10 g C m-2 y-1 in spiteof the large primary production in the area ormore than 300 g C m-2 y-1. Thus, less than4% of the primary production seems to reachthe sediments in the area, the rest isdecomposed in the water column and/oradvected away which further reveals the highenergy of the waters north of Reykjavik.

Settling rates of phytoplankton are in therange of 0.006-0.03mm/s (Heip et al. 1995;Wassmann 1991), which is similar to thesettling rates of the fine fraction (<63µm) ofsewage particles, which is 0.007-0.03 mm s-1

but this fraction makes up about 80% of thesewage solids (Lavelle et al. 1988). Thecoarse fraction (>63µm) had a settlingvelocity in the range 0.6-3.0 mm s-1 (Lavelle

et al 1988; Baker et al. 1995) or two orders ofmagnitude higher than detritus fromphytoplankton. The lowest rate found byBaker et al. (1995), 0.02 mm/s, is most likelydue to aggregation of particles in theundiluted sewage in their experiments.

The data shown in figure 10 can be used toestimate how much of the settling solids(maximum 22% of the total solids (Baker etal. 1995)) will have the chance of reaching theseafloor at the Ánanaust site. Only currentsare used for these predictions here and not theeffects of wave motion. As shown below inchapter 5.3, the latter are of great importancein keeping particles in suspension andespecially eroding them from the seabed ifthey had the chance to settle. Thus, the worstcase scenario is once again used for theprediction of the fate of sewage particles.

0

10

20

30

40

50

60

70

80

10 30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390 410

Time length of periods with currents < 1 cm/s, minutes

Num

ber

ofpe

riod

s

Current measurements at Ánanaust site 28/05/94-16/08/94: 49 days.Currents less than 1 cm/s at 1 m above the bed occur 21,7% of thetime.

Figure 10 Frequency of time periods with currents less than 1 cm/s at 1 m above bottom at the Ánanaust-disposal site.

39

Firstly, the height(s) at which solids will startto fall has to be estimated. The density ofsewage solids has been estimated to be in therange 1.04-1.5 g/cm3 (Lavelle et al. 1988) andthus both buoyancy and jet momentum fluxfrom the diffuser orifices takes them toheights well above the diffuser (at least toabout 8 m for the lighter particles at the worstconditions during stratification in mid-summer (Koh 1982; Neves et al. 1995 and2002)). To be very conservative, a height of1.5 m above the seabed or 0.3 m above the topsurface of the diffuser pipe is assumed.Secondly, the empirical relation between massfraction of settling sewage solids of primarytreated sewage as a function of settling rate inBaker et al. (1995) is assumed to be valid inthe area. This way one can calculate themaximum mass of solids reaching the seabed.The highest settling rates of sewage solids arebetween 2.2 and 2.4 mm/s (this applies toabout 0.3% of the total solids). Solids cantherefore settle out on the occasions whencurrents less than 1 cm/s prevail for more than10 minutes. Similarly, the particles of lowestsettling rates of sewage solids (0.024-0.028mm/s) constitute 0.77% of the total solids ofthe sewage. These would settle to the bottomif currents (much) less than 1 cm/s prevail formore than 890 minutes (which will neveroccur according to figure 10). Summing up, amaximum of 2.4% of the total solids willhave any chance of reaching the bottom evenat this worst case scenario at the disposal siteof Ánanaust STP. Assuming further anaverage current speed of 10 cm/s, theabsolutely highest carbon flux of 0.1-0.2 g Cm-2 d-1 will occur in a narrow band between50 and 100 m from the diffuser (once again aworst case scenario). Applying the empiricalrelationships for on one hand the oxygendemand resulting from this maximum carbonflux to the sediments and on the other handthe diffusion of oxygen into the sedimentboundary layer (dependent on currents at 1 mabove the bed) as described in Findlay andWatling (1997) for Maine, USA, and furtherverified in New Zealand (Morrisey et al.2000), it is evident that the current will supplythe sediments with much higher oxygen

fluxes than required by the settling solidsfrom the sewage at all times. The Findlay-Watling model relates carbon flux to thesediment to both sediment chemistry andeffects on benthic biota where the keycriterion is that the carbon flux should notexceed the oxygen supply for more than 2hours since the most sensitive organisms wereassumed to be harmed permanently if exposedto reduced oxygen for longer periods.Therefore, according to the Findlay-Watlingmodel elaborated to predict effects fromcarbon flux into sediments below salmon net-pens, the sewage will not give rise to anyorganic enrichment in the sediments as it willbe decomposed aerobically and further, noeffects will be observed in the benthic biota.In other words, the layer of seawater closest tothe sediment will not be depleted of oxygento a degree affecting the benthic fauna.

Another approach is to distribute themaximum amount of solids that could settleto the bottom, i.e. 2.4% of the total solids,onto the total area predicted to be affected bythe Mearns-Word model above giving about0.06 g C m-2 d-1 on average. This is wellbelow the amount of organic matter shown bymesocosm-experiments to give little or noeffects on the macrobenthic communities or0.1 g C m-2 d-1 while 1 g C m-2 d-1 enrichedthe sediment community, and loadings over1.5 g C m-2 d-1 produced degraded conditions(Kelly and Nixon 1981; Oviatt et al. 1987;Maughan and Oviatt 1993; Cromey et al.1998). These figures are in line with anempirical relationship found by Findlay andWatling (1997) between oxygen demand ofthe sediments caused by flux of organicmatter and carbon flux, where no effects onoxygen saturation above the seabed occurs forcarbon fluxes below 0.37 g C m-2 d-1.

Erosion or resuspension of settled sewageparticles at the Ánanaust siteFinally, as regards predictions of the fate ofsewage solids at the Ánanaust site, the periodsof currents below 1 cm/s are most of the timefollowed by maximum currents within twohours, i.e. periods of low currents occur at

40

slack water followed by maximum currents at1-2 hours after after slack water. Figure 11shows that even if solids might settle, theywill most of the time be resuspended oreroded from the sediments again within twohours and transported away. This occurssince the critical resuspension speed forsewage particles has been measured to be 5.0cm/s (Burt and Turner (1983) as cited inCromey et al. 1998) which will most oftenfollow the periods of minimum currentspeeds. This resuspension rate for sewageparticles is less than half that forcorresponding sizes of sediment particles, seediscussion on the Hjulström-curve in e.g.Anon. (1999). A closer look at the periods ofcurrents less than 1 cm/s shows that the timebetween them is in 47% of cases less than 50minutes, and in 42% of cases half a tidal cycleas expected. The period from the onset ofcurrents less than 1 cm/s until the currentbecomes 5 cm/s at 1 m above the bottom maybe considered as the residence time of sewageparticles on the seafloor. Figure 12 showsthat 97% of the periods are less than 10 hours,

82% are less than 5 hours, and that 50% of theperiods are less than two hours. In oneinstance was there a period of 52 hoursestimated residence time ocurring from02/08/94 (15:00) to 04/08/94 (17:50). Windspeed in this period was on average 2.4 m/s(1.0-3.6 m/s) to be compared with the averageof 4.2 m/s for the two months of currentmeasurements in July and August 1994.During this long period, an estimated flux tothe sediments is about 0.2 g C m-2 d-1 from150 thousand person equivalents assuming theconservative area of 1.4 km2 from Mearns andWord (1982) to be affected. The calculatedaverage current 1 m above the bottom was1.34 cm/s during this period. This carbonflux is well below the threshold that mightresult in any decrease in oxygen levels abovethe sediments according to the empiricalmodel of Findlay and Watling (1997).Actually, the oxygen supply to the sedimentsduring this longest period is about fifty timeslarger than the carbon flux requires foraerobic decomposition.

0

2

4

6

8

10

12

14

16

18

0 50 100 150 200 250 300 350 400Direction of current, clockwise from north

Fre

quen

cy,%

(cur

rent

spee

d,cm

/s)

Frequency of speeds <1cm/sFrequency of all speedscm/s (85% percentile of all speeds)

Figure 11 Frequency of currents less than 1 cm/s and all currents at the Ánanaust-disposal site at differentcurrent directions (intervals of 20°) at 1 m above the bottom. Also shown is the 85% percentile current velocity.

41

0

5

10

15

20

25

30

35

40

45

0 10 20 30 40 50 60

Residence time, hours

Fre

quen

cy,%

Figure 12 Frequency of periods from the onset of currents less than 1 cm/s until the current has reached 5 cm/sor more in the period 28/06/94-16/08/94.

In addition to the assumptions of equalling thebehaviour of sewage particles with mineralparticles as regards their stay or residence insuspension as well as not accounting for wavemotion in keeping the particles in solution,there is also considerable turbulent mixingduring slack water when currents changedirection rapidly. In spite of the predictionsabove representing absolutely worst casescenarios, no effects are foreseen on thesediments by emissions from150 thousandperson equivalents at the Ánanaust site by thepresent treatment of sewage, i.e. screeningfollowed by sand sedimentation and fatfloatation. Thus, the adaptation of UWWTD(CEC 1991) by Iceland, i.e. that screening isequivalent to primary treatment when sewageis discharged to less sensitive areas (article20.2 of regulation 798/1999), is thereforefully justified. Furthermore, it is not to beexpected to find any effects on the benthiccommunity nor sediment chemistry as a resultof the ocean outfall off Ánanaust at currentemission rates nor future discharge from

150,000 p.e.. Finally, further reduction insuspended matter than the present daytreatment offers will not result in anyenvironmental improvement in the recipientsince no effects are foreseen nor detectedexperimentally, see 5.3, 5.4 and 6.

Using the minimum exchange rate estimatedin 2.1, one may calculate the averageconcentration of suspended solids in the nearfield from future 150,000 p.e. as 120 µg/L.This level corresponds to about 50 µgparticulate carbon (POC) per litre. Here it isassumed that a negligible part of thesuspended solids of the sewage will settleonto the seafloor. This may be compared withthe natural content of offshore water(33.82‰S35.15‰) around Iceland or 29-420 µg POC/L (Unnsteinn Stefánsson 1981),where POC increases with primary productionor Chl a as expected. Regression of POCwith primary production or Chl a in May-June1976 gives a value of about 100 µg POC/L atzero primary production (Unnsteinn

42

Stefánsson 1981), which is an estimate ofwinter values. Higher naturally occurringPOC-values may be expected closer to shoredue to both resuspension from bottom andfrom fluvial and aeolian inputs from land.

In the far-field estimated in 2.1, theconcentration is only about 20 µg POC/L.Particulates decay in the water column andCSTT (1997) assumes a decay rate of 0.3 d-1

while Farley (1990) assumed a rate of 0.05 d-1

in cold water and 0.1 d-1 in surface water offsouthern California. Therefore, an averageexcess POC due to the sewage from 150,000p.e. in the far-field is 10-20 µg/L resulting in10-20% increase compared with off-shorewinter values but most probably the relativeincrease at the Ánanaust site will beconsiderably lower. The suspended matterderived from sewage, however, is of a differentnature to the naturally occurring suspendedmatter especially during summer months whenprimary production is effective. Finally, thePOC excess due to effluents from the future400,000 p.e. in the Kollafjörður, will onaverage be about 0.3-0.7 µg/L, which would beextremely difficult to detect experimentally.

5.2 Preexaminations and experimentallayoutIn February 2000, sampling of surface waterabove and around the present diffuser ofsewage from the Ánanaust STP took place forthe analysis of nutrients and microorganismswith additional examination of verticalprofiles for turbidity, temperature and salinity(Jón Ólafsson and Sólveig R. Ólafsdóttir2000; Vatnaskil 2000). Turbidity (measuredas loss of transmission of light at 670 nm) wasusually greatest at the bottom as expected butnever high in surface water. Sewage particlesmay have contributed to this bottom turbiditybut it stems also from incoming seawater intothe area. The most important source of theturbidity, however, is erosion of the seafloor.The so-called benthic nepheloid layer iscreated by the naturally increasing turbidity ofseawater with proximity to the seafloor.Transmittance, i.e. the proportion of light at

the wavelength of 670 nm that penetratesseawater, is a measure of suspended particlesin the water. Measurements of transmittancerequire calibration against direct analysis ofsuspended particles at each site and each timeif the data are to be converted to levels ofsuspended particles. Such a calibration wasnot performed on this occasion. At stationswhere the highest dissolved nutrientconcentrations were found, the turbidity wasvariable down the water column but not high.The study showed, not unexpectedly, thatsuspended particles behaved differently fromdissolved nutrients (Jón Ólafsson and SólveigR. Ólafsdóttir, 2000). During the February2000 survey the discharge from the ÁnanustSTP was about 95,000 person equivalents(Guðjón Atli Auðunsson 2000).

A preexamination of the Ánnaust-disposal sitewas carried out by videophotography from aROV on 20. May 1994 (Kjartan Thors 1994)and a similar study was undertaken at theKlettagarðar disposal site in February andMarch 1999 (Kjartan Thors 1999). Thesediment at the Ánanaust disposal site wasrich in shell fragments and at some places thecoarse sediment was overlain by finermaterials. As mentioned above, the sedimenthas ripples with wave heights of 10-20 cmand wavelengths of 20-50 cm. The rippledsurface is seemingly burrowed by organisms.

The examination in 1994 did not result in anestimate of the residence time of thesediments. Such an estimate is however ofconsiderable importance since a prerequisitefor possible oxygen depletion and effects onthe benthic biota caused by sewage disposal isthe temporary deposition of organic matter onthe seafloor as discussed above. Such asituation was deemed very improbable aselucidated above. Nevertheless, it wasconsidered important to account for thebehaviour of the sediment particles on theseafloor and to estimate the vertical flux andcomposition of the particles at the disposalsites as a baseline for later monitoring of theareas. In principle, the flux of the bedloadmay be estimated from the height of the

43

ripples and their rate of movement. Similarstudies have been carried out in for exampleBoston-Harbour-Masssachusetts Bay (Anon.1998).

In November 1995 an examination of fluxesby way of sediment traps was started at theplanned disposal site of the Ánanaust STP andthese studies continued until December 1996.For the estimation of sediment residencetimes and bedload transport, a programme oftime-lapse photography of the seafloor wasstarted in May 1996, and these studiescontinued until June 1997. In the period July1998 to April 1999 parallel studies werecarried out on the by then planned disposalsite from the Klettagarðar STP. These studiesprovided information on a period of leastactivity during the summer and a period ofgreatest activity during the winter. Evaluationof the data from the photograpy is found inKjartan Thors (2000) while the data for thesediment trap studies are found in Guðjón AtliAuðunsson (2001b) where also acomprehensive discussion is given on the fateof organic matter at the disposal sites ofsewage from Reykjavik and comparisonsmade with coastal areas in different parts inthe world. Corresponding studies wereconducted at the disposal site of the ÁnanaustSTP after the discharge of sewage started(June 2000 - April 2001) (Kjartan Thors2003; Guðjón Atli Auðunsson 2005). Belowis a short summary of the results of thesestudies.

5.3 Flux of suspended load studied withsediment traps

Fluxes and erosionIn the first study at the Ánanust disposal site,5 sediment traps were deployed, one at about3700 m from shore (middle of the planneddisposal site) and four at 500m distance fromthe first one, i.e. to the west, north, east andsouth of the middle of the disposal site. In alater study at this site, only three traps weredeployed. These were located at the east,west and north positions. On the by thenplanned disposal site of Klettagarðar STP,four traps were deployed around the middle ofthe planned diffuser at 5500 m from shore, i.e.at a distance 500 m west, north, east and southof the middle station. The traps were visitedmonthly and the total amount of material,particle size distribution and nutrients (LOI,organic carbon, organic nitrogen, andphosphorus) were analysed while traceelements (Cu, Cd, Zn, Hg, Cr, Pb and Ni)were analysed in selected samples.

The sediment trap studies at the Ánanaustdisposal site show large amounts of particlesin suspension, see figure 13, and a higher fluxduring winter than during summer.

The fraction of fine sediment particles, i.e. siltwith sizes less than <63µm, depends on theweather conditions each time and ranged fromas little as 4% up to 95%. The materials inthe traps are mostly due to erosion from theseafloor as shown by the good correlationbetween flux and wave height (at Garðskagi)to a power of about three where more than80% of the variation is explained by waveheight (Guðjón Atli Auðunsson 2001b and2005). This is a well known semi-empiricalrelationship for bedload transport, see forexample Anon. (1999).

44

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

5/11-1

3/12/95

5/11-17/1

2/95

13/12

/95-1/ 2

/96

17/12/9

5-1/2

/96

1/2/96-25

/3/96

1/2/96

-26/3

/96

25/3/96

-24/4

/96

24/4/

96-2

5/5/96

24/4/

96-6/

6/96

25/5/96

-3/7/96

6/6/96-3/7

/96

6/6/96-6/ 7

/96

3/7/96-15

/8/96

6/7/96

-15/8

/96

15/8/96

-13/9

/96

13/9/96

-16/10

/96

16/1

0/96-1

7/11/9

6

16/10

/96-6/

12/96

26/05

/00-2

4/06/00

24/06/0

0-27/7/00

27/0

7/00-26

/08/00

26/08

/00-2

3/09/00

23/0

9/00-23/1

0/00

23/10

/00-0

1/12/00

23/10

/00-02

/12/00

01/12/0

0-09/01/2

001

02/12

/00-10/01

/2001

02/12/0

0-09/0

1/200

1

09/01

/2001-

01/03/

01

10/01

/2001-

01/03/

01

01/0

3/01-15

/04/01

Wav

ehe

igh

t 1/3

(med

ian

,m)

0

1

2

3

4

5

6

7

8

ln(f

lux,

gm

-2d-1

)

Wave height

Ln(flux)

Figure 13 Comparison of significant wave height at Garðskagi and total flux of suspended material in the traps at the Ánanaust-recipient before (’95-’96) and after (’00-’01) the discharge started.

0

1

2

3

4

5

6

7

5/11

-13/1

2/95

5/11

-17/12/9

5

13/12/9

5-1/2

/96

1 7/12

/95-

1/2/96

1/2/9

6-25

/3/96

1/2/9

6-26

/3/96

25/3

/96-2

4/4/96

24/4

-25/5

24/4

-6/6/9

6

25/5-3/

7/96

6/6-

3/7/96

6/6-

6/7/96

3/7-

15/8/96

6/7-

15/8/9 6

15/8

-13/9

/96

13/9

-16/1

0/96

16/10

-17/11

/96

16/1

0-6/1

2/96

2/7/9

8-2/8

/98

2/8/9

8-6/9

/98

6 /9/98

-6/10

/98

6/10

/98-8

/11/98

8/11

/98-1

1/12/9

8

8/11

/98-14/1

2/98

11/1

2/98-

19/1/99

14/1

2/98-

19/1/99

19/1/99

-07/0

3/99

8 /11/9

8-19

/1/99

26/05/0

0-24

/06/00

24/0

6/00-

27/7/00

27/0

7/00-26

/08/00

26/08/0

0-23

/09/00

2 3/09/0

0-23

/10/00

23/10/0

0-01

/12/00

23/1

0/00-

02/12

/00

01/12

/00-0

9/01/

2001

02/12

/00-10 /01

/2001

02/12

/00-0

9/01/

2001

09/01

/2001

-01/03

/01

1 0/01/2

001-01

/03/01

01/0

3/01-

15/04

/01

Win

dsp

eed

(med

ian)

,m/s

0

1

2

3

4

5

6

7

8

ln(fl

ux,g

m-2

d-1)

Wind speed, m/s, SA-NV

ln F(total flux)

Figure 14 Comparison of wind speed from SE and NW and flux of suspended material in the traps at the Ánanasut-recipient before(‘95-‘96) and after discharge was started (‘00-‘01) together with data for the Klettagarðar-recipient before the discharge started(‘98-‘99).

45

In addition, materials transported to the areasfrom other marine waters as well as from landmay contribute. The total flux and flux of siltparticles behave very similarily at both theÁnanaust disposal site and the Klettagarðardisposal site. The wave height at Garðskagialso correlates well with wind speed atReykjavik harbour. Figure 14 shows thatspeeds of winds from SE and/or NW, i.e.parallel to the north coast of Reykjavík, alsocorrelate well with the total flux. Thus,synergic effects are at work between windsand currents where particles are eroded fromthe bottom material by the wave action whiletidal and residual currents transport it away.

Even gravel and pebbles are transported intothe water column to heights of at least of 45cm above the bottom, i.e. height of the traps.When the behaviour of silt fraction iscompared to total flux it may be seen that at aflux of 200 g m-2 d-1, the fine fraction starts todecrease, i.e. sand and gravel start to moveupward from the bottom. This fluxcorresponds to a wave height of 2.2 m atGarðskagi but wave heights are 35% of thetime higher than 2.2 m in the three periodsinvestigated. Similarily, this flux correspondsto wind speed of 5 m/s in Reykjavík but windspeeds higher than 5 m/s occurred 40-50% ofthe time during the sediment trap studies atÁnanaust. Monthly averages of wind speedsin Reykjavík 1961-2005 are 70% of the timehigher than 5 m/s where lower wind speedsoccur predominatly in June and July, seefigure 7.

At SE-wind speeds >3m/s the flux increasesfast and so does wave height from 1.2-1.5 m.Wave heights <1.5 m increase the fluxessignificantly but not wind speeds <3 m/s. Themonthly average wind speed in the years1961-2005 is 3 m/s or lower only 0.3% of thetime. At the lowest recorded wave heights atGarðskagi, i.e. 0.4-0.5 m (0.4m 0.1% of thetime and 0.5 m 0.5% of the time) there isstill a flux of silts of about 0.5 g m-2 d-1 thatmay occur for 1-3 days (this happened onlyonce during these three study periods, i.e. atthe end of June 2000). As organic carbon is a

constant fraction of the silt or about 8.5%,there is a flux from the bottom of about 0.04 gC m-2 d-1 on these rare occasions. This flux islittle less than the maximum settling rate oforganic carbon from the sewage during thelongest estimated residence time of sewageparticles on the seafloor or 52 hours under thevery conservative approach for settling ratedescribed in 5.1.

More than 90% of the time, the wave height isgreater than 0.9 m and corresponds to a fluxof more than 4.5 g m-2 d-1. Thus, more than90% of the time, the erosion rate of carbonfrom the bottom sediments is greater thanabout 0.5 g C m-2 d-1. This rate is muchgreater than the maximum settling rate underthe worst case scenario described in 5.1 andtherefore it is highly improbable that possibleadditional organic matter could be observed inthe sediment traps.

These experiments verify that there isconsiderable movement of seawater above thesediments caused by wave motion in additionto the movement accounted for by currents.This in turn manifests that the settling ratesand residence times in 5.1 are greatlyoverestimated and represent truly a worst casescenario. These observations explain the verylow burial rate of organic carbon in thesediments in this area and therefore low fluxof carbon to the sediments (Guðjón AtliAuðunsson 2001b) in spite of high primaryproductivity. Organic matter is mostlyoxidised in the water column due to longresidence times caused by the great turbulentmixing which in turn supplies ample amountsof oxygen for their aerobic decomposition. Ifthe organic matter settles on the seafloor, it isshortly thereafter eroded from the bottom intothe water.

From these observations it may be concludedthat there is always considerable flux of siltparticles at the seafloor except, very rarely, forvery short periods, 1-3 days as an absolutemaximum when a minimum flux is observed.During these periods, the oxygen supply to thesediments is always well above the oxygen

46

demand and thus reduced oxygen above thesediments due to sewage particles alone willnot occur.

As described above, the area in SE-Faxaflói,where the disposal sites are located, is one ofthe most productive marine areas aroundIceland with primary production of more than300g-C m-2 y-1 to be compared with about 218g C m-2 y-1 on average within the the 200 mdepth contour (Þórunn Þórðardóttir 1994). Inspite of this high primary productivity,oxygen depletion has not been observed in thearea due to high currents and fast waterexchange. Episodically, some oxygenundersaturation may be observed in this areaas could be seen in 1966 when someundersaturation took place close to the bottomdue to fast decomposition of organic detritusfrom phytoplankton and related materials,starting at the end of June, peaking in middleof August, and ending in late October(Unnsteinn Stefánsson, Þórunn Þórðardóttir &Jón Ólafsson 1987). The minimum oxygensaturation was about 80%. It may also bementioned that the rate of decomposition isfaster the higher the primary productivity.

Nutrients in the trapped materialsOrganic carbon, organically bound nitrogen,and phosphorus of the organic matter indicatethat almost solely organic matter of marineorigin is present at the disposal sites prior tosewage release as well as after dischargesstarted from the Ánanaust STP (localRedfield-relationships). The natural ratios ofthese nutrients in living marine biota may,however, vary to a degree that encompassescomposition of sewage particles (Guðjón AtliAuðunsson 2005). Furthermore, seasonaldifferences may be observed as can variationfrom one year to another. Thus, evaluation ofthese data may be a complicated task. Whencomparing the composition of the organicmatter in the sediment traps before and aftersewage discharge started it turned out that nochanges could be observed in the total amountof organic matter. Furthermore, both nitrogenand phosphorus were significantly lower withrespect to carbon after the release than before.

The linear relationship between nitrogen andcarbon was lower by a constant but the rate ofchange of nitrogen with carbon was the same,i.e. the same rate of change prior to and afterrelease of sewage and the same as in typicalorganic matter of marine origin. Whencompared with the Ánanaust study 1995-1996, lower nitrogen by a constant relative tocarbon was also observed at the Klettagarðar-disposal site prior to any disposal, i.e. 1998-1999. Phosphorus also correlates withnitrogen in the same manner before and afterdischarges started and in the same manner asthese nutrients relate in the ocean south ofReykjanes (Selvogsgrunn). Theseobservations bring out the variations betweenyears in the age of organic matter in the areabut phosphorus and nitrogen are ususallyreleased faster from the particles during theirdecomposition, i.e. the ratio of carbon toeither nitrogen or phosphorus increases withage.

In summary, no changes in supply nor incompostion of organic matter caused bydischarge of sewage from 100,000 personequivalents could be detected in the area.This is in line with predictions on thebehaviour and fate of sewage particles in thearea discussed above.

Trace elements in the trapped materialsSome trace elements (Hg, Cr, Pb, Cd, Cu, Zn,Ni) were analysed in selected trap samplesbefore and after release of sewage and theresults show that their concentrations were notsignificantly different from their levels insurface sediments from uncontaminated areasaround Iceland (Guðjón Atli Auðunsson2005). The levels of nickel are, however,generally somewhat higher in the sedimentsnorth of Reykjavik than in most sedimentsaround Iceland. The reason for this is atpresent not known but this nickel does notstem from the current discharge of sewage(Guðjón Atli Auðunsson 2005). Lead andmercury increase with organic carbon as dozinc and copper although their rate of increasewas slower. This behaviour indicates thatthese metals originate in organisms in the

47

area. Nickel and chromium, however,decrease in concentration with organiccarbon, which indicates that their behaviourmay be explained by the geochemistry of thearea. Since organic matter on the seafloor ismade up of the marine biota of the area aswell as faecal pellets, these results show thatneither particulate nor dissolved traceelements of the sewage are taken up by thelocal biota. This is in line with the predictedrate of dilution of these metals but already inthe near-field the levels of all these metals arewell below their natural concentrations inoffshore ocean water (Guðjón Atli Auðunsson2006). These results also indicate that thesediments are well oxygenated as predictedabove because suboxic or anoxic sedimentsaccumulate metals.When the results for trace elements insediments from unaffected areas aroundIceland are compared with sediment qualityguidelines from all over the world it becomesclear that quality guidelines need to beestablished on a regional basis. Suchguidelines for some elements, e.g. copper andzinc, do not appear to be suitable for prevalentconditions around Iceland.

5.4 Benthic photographyThe time-lapse photographs were examinedwith respect to signs of bedload transport.These signs are all types of changes on theseafloor, transport of sand and particulargrains, especially shells and fragments ofshells, and formation of ripples in the seafloormaterial and erosion of the seafloor.Additionally, fine materials are seen asresuspended solids or turbidity. Informationon both wind speeds and wind directions inReykjavík were used to assess their possibleeffects on the transport of bottom sedimentsin the area. It has not been possible to findany confident relationship between winds andsediment transport nor between tidal currentsor wave height (at Garðskagi) and movementof the sediment.

For the Ánanaust disposal site, both beforeand after discharges started, it could be seen

that transport of the sand and gravel bottom ofthe investigated area was frequent in theperiod of the research, especially duringwinter. The nature of the sediments,especially their coarse grain size, suggestshigh-energy conditions not likely to allowsedimentation of fines

From September to the end of March,frequent periods of sediment transportoccurred, often several times each month,scenarios that may be divided into 1)formation of ripples, 2) erosion of ripples, and3) sedimentation. The last event wascharacterised by sand that covered the wholearea of the photograph and masked olderformations, see figure 15. Quite frequently acloud of sediment particles in suspension wasformed above the bottom and often so densethat the bottom could not be seen (1 meterfrom the lens of the camera). These resultsdemonstrate considerable movement ofsediments during the winter time and verifythat fines (fine sand, silt, and clay) do notreside in the area but are transported away bycurrents (Kjartan Thors 2000 and 2003).

The situation was much calmer in the areasduring the summer 1996, spring 1997, spring2000 and summer 2001. For example in theperiod between end of May to the end ofAugust 1996, hardly any ripples were formedin the sediment except for one time, i.e. at theend of July and beginning of August. Theformation of ripples is an indication offorceful transport of bottom sediments causedby currents and this was seen only this time.The same is true for the period from themiddle of April to 24th of June 1997, whensediment transport of the aforementioned typewas only observed once, i.e. end of May andbeginning of June. This does not mean thatno sediment transport takes place during thesummer outside these periods. Anexamination of the photographs reveals thatchanges of the bottom sediments are frequentalthough they are not as decisive as they are inwinter time, see figure 16.

48

Figure 15 Photographs from December 2000.a) December 11 at 13:06. Bottom with sand and gravel with shells and shell- fragments. Some ripples may be seen inthe sediments.b) December 16 at 13:06. Some turbidity above the bottom. The rope has been moved. New ripples may be noticed onthe bottom.c) December 18 at 13:06. The turbidity has disappeared but the new sandripples are clearly seen.

a

b

c

49

Figure 16. Photographs from June 2000.a. 4 June. Ripples in the sand rupturing.b. 6 June. Newly formed ripples seen in the sand.c. 7 June. Ripples disappear again.

a

b

c

50

A series of photographs from 25th of June to3rd of July 1996 contains 77 photos. In 33 ofthe pictures, changes from the precedingpicture are observed. In some of these cases,the changes are relatively small but in othersthe changes are more marked. The changesindicate either erosion or sedimentation, butclearly indicate transport of sediments. Aseries of pictures from 6th of July to 15th ofAugust the same year contain 78 pictures ofwhich 51 show evidence of sedimenttransport.

It is therefore evident that sediment transportoccurs frequently in the spring and duringsummer months although not as vigorous asseen during winter months, often startingalready in September. Not all transport iscaused by waves and currents. In some of thecases where changes in the bottom sedimentsare observed, these are probably caused bybenthic organisms, bioturbation (Graf andRosenberg 1997). In many of the pictures,starfish, gastropods, crabs and fish includingflatfish, are seen which undoubtedly affect thebottom sediments and move sediment grains.Although the effects of these organisms arenot very important in transporting sediments,they dislodge sediment grains which aresubsequently more prone to transport by theflowing seawater. When the changes of thebottom sediments are considerable, these canonly be explained by motion of the seawater.

One event is worth mentioning regardingsedimentation and movement of sedimentsalthough no explanation is provided at thistime. In pictures from the springs of 1996 and1997 a brownish hue is seen on the bottom inevery second picture but less conspicuous orabsent in other pictures. The surface spread isseen during the day but has disappeared thefollowing night. The nature of this featureshas not been confirmed by direct experimentsbut there is strong evidence that epipelagicdiatoms are the reason for this effect sincethey attach themselves to the surface ofsediment grains and move to the top during

the day but to the underside of the grains atnight.

The sedimentation/erosion of bottomsediments at the planned disposal site fromthe Klettagarðar STP was studied by thesame method. Pictures were taken of thebottom twice every 24 hours in the periodJuly 2nd 1998 to February 1999.

The pictures show small changes of theseafloor in summer and autumn 1998.Occasionally, some turbidity was seen in theseawater but it disappeared after a short whilewithout visible settling on the floor. Thebottom at this site is sandy with shellfragments and shells and does not seem tochange much as mentioned above. There isconsiderable activity of organisms at this site,fish are frequently seen but the camera framemight have affected that. Starfish arefrequently seen but stay only for a short timein the path of the camera. The alterations onthe bottom are slight enough to be causedsolely by the passing organisms.During winter, i.e. from middle of November1998, the frequency and time length of turbidwaters increases and the waters become richerin suspended particles. As was seen duringthe summer months, the suspended matterdoes not seem to settle on the floor and all ofit seems to disappear from the site. Themotion of the seawater starts to relocate thecoarser materials on the bottom and in themiddle of December, ripples on the seafloorstart to form but these are signs of sedimenttransport. New ripples are also seen in theend of January and beginning of February1999.The main conclusion is that silty materials,i.e. materials finer than sand, do notaccumulate at this site, the planned disposalsite from the Klettagarðar STP. Turbidity thatforms occasionally as a result of theinteraction between winds and currents, isflushed away. During windy periods,especially during winter, the motion of theseawater is strong enough to relocate thebottom sediments.

51

From the discussion above, it may beconcluded that net sedimentation rate at theKlettagarðar disposal site is extremely slow ornone at all. This is supported by the fact thatmodern sediment at this site is very thin, onlyfew tens of centimeters. Modern timeencompasses approximately the last tenthousand years and therefore, in the timeframe of discharge from the Klettagarðar STP,the sedimentation rate is not detectable.

Summary of the sediment bedload studiesThe bottom photography, which is mainlydirected at the coarser sediments on theseafloor, shows that sand and gravel is movedseveral times each year, especially during thewinter time, with the sand more frequentlythan the gravel. Fine materials settletemporarily and disappear long before thecurrents start to move the sand. The net effectof this process is transport of coarse materialthrough the area.The photographs only gives indirectinformation on the fate of the sediments thatemerge from the sewage diffusers. The siltymaterial which is often seen in suspension atthe bottom, seems to settle only for a veryshort time on the seafloor. It is reasonable toconclude, therefore, that the sedimentsemerging from the sewage diffuser will notaccumulate at the disposal site.This conclusion is in agreement with thepredictions of 5.1 that sewage particles willnot settle on the seafloor at the disposal site ofthe Ánanaust site and on the rare occasionsthey might do so, their residence times wouldalways be less than 10 hours and in a fluxonto the sediments that would hardly beobservable by the camera.

6. BIOLOGICAL SURVEYS OF THEBENTHOSIn the summer of 1995, the benthos of theplanned disposal site of the Ánanaust STPwas investigated for the purpose ofdetermining its potential value forconservation. The study was also meant to

serve as a baseline for monitoring of the areaafter sewage disposal started (JörundurSvavarsson 1996). The post-examination ofthe disposal site of Ánanaust STP took placein the summer of 2000, two and a half yearsafter disposal started (Jörundur Svavarsson2002). Parallel investigations were performedat the planned disposal site from theKlettagarðar STP in 1999 (JörundurSvavarsson 2000).

The sandy bottom further out than 3700 m atthe disposal site of Ánanaust STP isunsuitable for monitoring since uncertaintyassociated with the sampling limits itsusefulness. The problems of sampling wasdescribed in 1.3. Therefore, only the areawithin 3700 m was investigated which meansthat only the first 100 m of the diffuser areawas investigated since the diffuser starts at3600 m. Sandy bottom beyond 3700 moffshore is generally limited in the number ofspecies and, as stated above, reflects a ratherharsh environment. Most benthic studiesaround Reykjavík have been carried out onmuddy sediments which have greater speciesrichness; and studies on benthic communitiesin relation to sewage outfalls discussed in 5.1are generally on sediments rich in mud. Forthis reason, investigations at the Ánanaust sitewere carried out by photographing the bottomand by studies on kelp-holdfast wherever theycould be found.

The outer limit of the brown kelp Laminariahyperborea is reached at 3540 m. At thisdistance the plants are small and scattered.The biota on the rocky bottom from 3540 to3700 m offshore is without kelp and thereforethe diverse biota associated with theirholdfasts is not available. The number ofspecies and species density was in 1995 ratherlow in this area. The kelp forest gets moredense in direction to land and plants becomelarger while it again turns more scattered stillcloser to shore with barren zones. In an areabetween 3200 to 3540 m from the shoreline,considerable biodiversity exists on the rockyand sandy bottom where horse mussels(Modiolus modiolus) dominate, inhabiting

52

pockets in the sand and are used by kelp assubstratum. In spite of considerable speciesdiversity in the kelp forest, the area is notconsidered unique and similar in biodiversityto other areas of similar sediments aroundIceland. Within 3200 m from the shoreline,the bottom becomes more sandy.In 2000 no changes in the benthic communitycould be detected as a result of the dischargeof sewage effluents. The number of speciesand number of individuals on kelp-holdfastsin 2000 after discharge started, was inagreement with what had been found prior torelease of sewage in the area, i.e. in 1995.Furthermore, there was not a significantdifference in biodiversity or biodensitybetween the disposal site and the referencesites west and east of the study area.Natural variability in the benthos is, however,considerable and environmental stress noteasily detected above this natural variabilityeven on soft bottoms (Livingston 1982). In2000 the number of horse mussels haddecreased significantly from 1995 but mostprobably due to a large number of the star fishAsterias rubens in 1995 feeding on horsemussels. This natural event is furtherdemonstrated by a similar decrease in horsemussels at the reference sites west and east ofthe disposal area.The absence of significant change in thebenthic communities is in good accordancewith predictions on the fate of sewageparticles, i.e. they do not settle in the area andare thereby not available to the organisms.

The benthos at the disposal site from theKlettagarðar STP is homogenous but withconsiderable diversity. The investigation atthe Klettagarðar site was carried out by grabsampling although difficult to carry out on thehard bottom with a thin layer of sand. Thebiodiversity is not among the highest foundon sandy bottoms in shallow waters in SW-Iceland. Probably the diversity of the areareflect young individuals of species that willpresumably not thrive there as adults. Mostof the species found in the area have beenfound elsewhere in shallow waters aroundIceland. No species were found unique

enough to warrant conservation of the area offKlettagarðar.

Many investigations have been carried outaround and in the vicinity of Reykjavik asregards the possible effects of the pastdisposal of sewage on organisms on the beachand benthos off shore and baseline datacollected (Sólmundur Einarsson 1973; EinarJónsson 1976; Arnþór Garðarsson and KristínAðalsteinsdóttir 1977; Agnar Ingólfsson1977; Arnþór Garðarsson, Jónbjörn Pálssonand Agnar Ingólfsson 1974; Guðmundur VíðirHelgason and Jörundur Svavarsson 1991;Karl Gunnarsson and Konráð Þórisson 1976;Guðmundur V. Helgason and ArnþórGarðarson 1992; Karl Gunnarsson 1993;Guðmundur V. Helgasaon and ArnþórGarðarson 1995). Of the investigations thathave taken place, only those related tomacroalgae on the beach closest to the outletof sewage in Skerjafjörður, showed effectsthat could be associated with sewage disposal(Karl Gunnarsson and Konráð Þórisson 1976)but the benthos in Viðeyjarsund and Eiðsvíkhad possibly been affected by pollution(Guðmundur V. Helgason and ArnþórGarðarson 1992). The present disposal ofsewage has relieved the pressure on the areasthat were affected by sewage disposal throughtemporary outlets close to or onto beaches andthereby of limited initial dilution. All sewageis now released through diffusers at thedisposal sites of the Ánanaust andKlettagarðar STPs, where mixing withseawater is fast and efficient andaccumulation of organic matter on thesediments is extremely limited, nil for morethan 97% of the time. As discussed under5.1, the accumulation of sewage particles inor onto the sediments is a necessaryprerequisite for them to exert any effects onthe benthos. Therefore, the net effect on thebiosphere has been scaled down extensivelyand the sites of present disposal are evidentlyunder no detectable pressure from theeffluents of the outfall. This was to beexpected from the predictions of initialdilution in chapter 2, verified by distributionof coliforms and nutrients, and fate of

53

particulates from the sewage in 5.1 verified bythe results from the sediment trap studies in5.3, and benthic photography in 5.4.

7. MONITORING OFCONTAMINANTS WITH BLUEMUSSELS

Monitoring with musselsEnvironmental effects of sewage disposal aregenerally reflected in changes in the speciesliving at the disposal sites. In order tofacilitate the study of these effects and notleast to make the evaluation of the data lessdemanding, organisms may be deployed atdifferent distances from source as well as in areference site unaffected by sewage. Variousbiological, physiological and biochemicalchanges in the organisms may be monitoredfor this purpose but most commonly theaccumulation of certain pollutants is lookedfor since that is often the direct cause of mostof the detrimental changes in the biological,physiological and biochemical parameters.Information on accumulation also revealswhether the chemicals found in the organismsare bioavailable to them, often theprerequisite for their possible harm toorganisms. The nature of the effects mayoften be related to the nature of and/or levelof the substances found in the organismssince much is known through experimentsabout detrimental substances found in sewage.

The most commonly used organism for thispurpose is the blue mussel (Mytilus edulis)but other species have also been used. Bluemussels at SW-Iceland have been used fortrace metal studies and regional monitoringprocedures developed (Jón Ólafsson 1983 and1985). By placing the mussels at differentdistances from the sewage source, distributionof sewage is revealed. Additionally, thehealth of the organisms is evaluated bothgenerally and in relation to the type and levelsof chemicals found in their tissues. If thesechemicals correlate with sewage exposure,e.g. faecal coliforms, their source in the sewer

system may be located and their releaseterminated. Studies with blue mussels arecommonly used in Iceland for monitoringindustrial emissions and discharges fromvarious activities. Additionally, blue musselsin the wild are widely used for long termmonitoring. By this way a good Icelandicdata bank is available, which, together withinternational data banks, greatly facilitateevaluation of the results.Since the accumulation of pollutants inmussels is rapid and the analytical methodsvery sensitive, relatively short period (abouttwo months) will suffice to find out whetherand to what extent the organisms in the areaare affected by the sewage or otherdischarges. If further sensitivity than thatobtained in two months is required, the timeperiod of deployment may easily be increased.Furthermore, for increasing the certainty inthe assessment, the cages may be placed atstations where effects of sewage are mostprobably found irrespective of whetherappropriate species for monitoring arenaturally present at the stations or not. It willtherefore take a relatively short time to findout whether sewage treatment is effective ornot and to reveal possible effects of sewagedisposal on the organisms in the recipient and,if not satisfactory results are obtained, onemay react rapidly to improve the situation(e.g. by readjustment of the sewage treatmentprocess, readjustment of the route of thepipeline or by preventing disposal ofhazardous substances).

Comprehensive studies have been carried outas regards past disposal of sewage fromReykjavik for both classification of theplanned disposal sites from Ánanaust STPand Klettagarðar STP as sensitive or lesssensitive as well as creating baseline data forlater monitoring at these sites (Guðjón AtliAuðunsson 1994b and 2001; Guðjón AtliAuðunsson and Hannes Magnússon 1995aand 1995b). Similar studies have also beencarried out in relation to sewage inHafnarfjörður (Guðjón Atli Auðunsson 1996).In the summer of 2000, examination bymeans of deployed blue mussels along the

54

pipeline from the Ánanaust STP took placewhere organochlorines, PAHs, trace elementsand microbiological indicators were analysedin the whole soft tissue of the blue mussels(Guðjón Atli Auðunsson 2006). Togetherwith growth rates, death rates, andmorphometry, proximate chemicalcomposition was also determined.

A short review of the main conclusions drawnfrom the monitoring of mussels is as follows.

Bacterial indicator organismsThe situation prior to disposal of sewage bythe ocean outfalls may be summarised asfollows (Guðjón Atli Auðunsson 1994b;Guðjón Atli Auðunsson and HannesMagnússon 1995a and 1995b)

a) The distribution of bacterial indicatororganisms off Klettagarðar and Ánanaust wasgreater than expected from their distributionin surface water at the same season. This isprobably due to the fact that musselsaccumulate these bacterial indicatororganisms whereupon their concentrationsbecome about ten times (coliforms) and 100times (enterococci) that of the surface waterthey live in.b) The level of bacterial indicator organismsdecreased continuously with distance fromshore both off Klettagarðar and Ábnanaust.c) Total coliforms in blue mussels within4000 to 5000m from shore were in all caseshigher than Icelandic guidelines for bivalvesstipulate (1 MPN/g).Figure 17 shows the distribution of faecalcoliforms in blue mussels off Ánanaust in1995. The release of sewage to the coast atthat time was about 15 thousand personequivalents distributed over a long distancealong the shore (Guðjón Atli Auðunsson1994).

The lower diagram shows the distribution oftotal coliforms in blue mussels in the periodJuly to September 2000 after release ofsewage started at the beginning of 1998 andduring this sampling period it amounted to100,000 p.e. (Guðjón Atli Auðunsson 2000and 2002). The diffuser is at 3600 to 4100 mand the maximum concentration of totalcoliforms was found above the landward endof the diffuser at 3600 m (Guðjón AtliAuðunsson 2006). The concentration is lowto the east, i.e. in the direction of the residualcurrent. Only one sample could be obtainedto the west of the diffuser (a cage was lost)but its value was fairly high. A single sampleis only an indication since the variability inmeasurements of bacteria is very high. Forthis reason, this single measurement is notshown in the diagram. There is considerablevariation in these measurements but goodcorrelations were found between totalcoliforms, faecal coliforms, enterococci andE.coli indicating a common human source(Guðjón Atli Auðunsson 2006) where thefollowing was always found:

total coliforms faecal coliforms enterococci E.coli

As seen in figure 17, the maximum at 500 mfrom the coast from 15 kpe is about onefourth of the maximum from 100 kpe.However, a sample taken 50 m from apreliminary outlet of 50-60 kpe fromLaugarnes in 1994 gave total coliforms onaverage 60 MPN/g in the mussels (GuðjónAtli Auðunsson, 1995a), indicating further thegreater dilution efficiency of the presentdischarges. These results on bacteria inmussels together with the measured per capitavalues of the bacteria confirm the the averagedilution of about 1000 in the near field(Guðjón Atli Auðunsson 2006).

55

Total coliforms off Ánanaust 24/07-06/09 1995Averages of four sampling occasions

0

5

10

15

20

25

0 1000 2000 3000 4000 5000 6000

Distance from shore along planned pipeline, m

Tot

alco

lifor

ms,

MP

N/g

ww

EW

Red points show mussels 250 m eastand west of station 4000 m

Total coliforms off Ánanaust 25/06-10/09 2000Averages of four sampling occasions

0

10

20

30

40

50

60

70

80

90

100

0 1000 2000 3000 4000 5000 6000

Distance from shore along the lay of pipeline, m

Tot

alco

lifor

ms,

MP

N/g

Green points show mussels alongthe diffusor: 3600-4100 m

500 m east from middle of diffusor

Figure 17 Distribution of total coliforms along the pipeline off Ánanaust. In the upper diagram, a release of anestimated 15,000 p.e. occurred at various places along the shoreline. In the lower diagram, a discharge of about100,000 p.e. took place along the 500 m diffuser.

56

Trace elementsA comprehensive review on trace elements inall the investigations that have been carriedout using blue mussels in the recipient(Guðjón Atli Auðunsson 2006) shows thatexcept for the metal silver, the past disposalof sewage could not be found to have had anyeffect on their levels in blue mussels inrelation to unpolluted areas around Icelandand when compared with international databanks on trace elements in mussels. A slightincrease in lead was observed closest to shore(and sewage outlets) prior to the transfer ofdischarge to the present disposal sites. Thatslight increase in lead remained after thepreliminary outlets were removed indicatingeither lead in past time sewage or, moreprobably, various human activities close toshore (Guðjón Atli Auðunsson 2006). Thecalculations on dilution in chapter 2 togetherwith results of the levels of trace elements inthe sewage itself, predict that all the traceelements are diluted to levels much lowerthan their current levels in ocean water exceptfor silver. Thus, the data on currents in thearea predict the outcome. This holds trueeven though the trace elements are bound to

particles which distribute differently todissolved species. This is due to the facta thatparticles are always suspended in the recipientand that many trace elements are releasedfrom the particles when entering seawater(Sólveig R. Ólafsdóttir and Jón Ólafsson1999). As shown in figure 18, the presentdischarge of sewage has relieved andimproved the situation for silver quitemarkedly, where mussels closest to shore offÁnanaust are as low as the reference sitewhile the levels at the disposal site are thesame as prior to discharge of sewage in thatarea. For reasons not yet elucidated, therewas an elevation of copper, nickel, mercuryand cobalt in the middle of the disposal sitefrom the Klettagarðar STP prior to anysewage release at the site (Guðjón AtliAuðunsson 2001a).The levels of trace elements in blue musselsare in all cases lower or much lower than theirstipulated maximum limits for seafood forhuman consumption (maximum limits existfor lead, cadmium and mercury in mussels(Commission regulation (EC) No 466/2001 of8 March 2001)).

0

2

4

6

8

10

12

0 1000 2000 3000 4000 5000 6000

Distance from shore along the pipeline, m

Ag (

Sund

in)/A

g (re

fere

nce

site

)

Sundin '93, '94 og '98

Ánanaust 2000

Figure 18 Concentration of silver in blue mussels as a raio of silver in the mussels along the pipelines (Sundin: bothoff Ánanaust and Klettagarðar) to the concentration of silver in mussels at the reference site (Hvalfjörður).

57

Since the levels of trace elements off the northcoast of Reykjavik and in the near field of theÁnanaust site are the same as in non-affectedareas around Iceland, except for silver, theyare also low in comparison with, for example,Norwegian quality guidelines, rendering thearea fit for aquaculture and recreationalpurposes (Molvær et al. 1997), i.e. the criteriafor highest quality. This holds also true forsilver at the Ánanaust site.

In general, it may be stated that in spite of therelatively large release of some of theinorganic trace elements by sewage (apartfrom silver which is at extremely low levels inthe sewage of Reykjavik), they do notaccumulate in the blue mussels except forsilver. This is due to two reasons: their lowconcentrations in sewage caused by dilutionof sewage prior to release and relatively fastand efficient dilution of the sewage when itenters the recipient. Nevertheless, silveraccumulates to a very high degree in mussels.

Organochlorine micropollutantsSeveral classes of organochlorine compoundshave been analysed in the blue mussels.

A comprehensive review on organicmicropollutants in all the investigations thathave been carried out on blue mussels in therecipient (Guðjón Atli Auðunsson 2006) maybe summarized as follows.

As regards PCBs, DDTs, and HCB, theirlevels fall off at similar rates with distancefrom shore, both off Klettagarðar andÁnanaust, and the levels of the reference sitein Hvalfjörður is obtained withinapproximately 4.5-5 km distance from thecoast. Other compounds like chlordanes,toxaphenes, HCHs, and trans-nonachlor alsodecrease with distance from shore albeit atdifferent rates. This holds true both beforeand after the release of sewage started. Apartfrom the results for the study off Laugarnes in1993 (west of the planned STP at Klettagarðarwhere a preliminary outlet was situated priorto the installation of the Klettagarðar STP inspring 2002) neither increase nor decrease has

been discerned with time in these threeclasses of compounds in the period 1993-2000. These compounds were found toincrease towards shore. The present sewagewater is not the source of the compoundssince no change is observed in the gradient offÁnanaust in 2000, two years after the STPthere started its operation. The source may besewage from earlier times that may havecontained these substances with resultingaccumulation in the sediments. It may also bevarious activities on land or at shore that mayhave emitted these pollutants. Whether thesituation will be the same for the gradient offKlettagarðar awaits to bee seen when cagedmussels will be deployed along the pipelineand at the disposal site. These results werealso predictable from the levels found in thesewage itself and the dilution calculated in 1.6since that dilution takes these compoundsbelow levels found in ocean water aroundIceland.

The results show that the levels closest toshore range from being two times (HCB) tofive times (DDE) higher than in wild musselsat the reference sites but the PCBs are up tofour times the levels of the reference sites. Noincrease is detected above the diffuser at theÁnanaust site.

According to Norwegian guidelines fororganochlorines in blue mussels (Molvær etal. 1997), the areas north of Reykjavik, therecipient, fulfill the highest standards and maybe rated as suitable for aquaculture andrecreational fishing.The levels of the organochlorines in bluemussels off the north coast of Reykjavik, evenclosest to the shore, is below (PCBs) or farbelow (DDTs, HCB, and HCHs) thestipulated maximum limits for thesecompounds in seafood for humanconsumption in Iceland.

Polyaromatic hydrocarbons (PAHs)PAHs were analysed in mussels along theocean outfall from Klettagarðar in the summerof 1998 (Guðjón Atli Auðunsson 2001) andoff Ánanaust in 2000 (Guðjón Atli

58

Auðunsson 2006). For both areas, theirbehaviour was similar to for example thePCBs and at a distance 4000-5000m from theshoreline, these substances had reached thelevels found at the reference site(Hvalfjörður). No change in the gradient wasobserved above the diffuser at the disposalsite off Ánanaust. Thus, the present sewagewater is not the source of these compoundsclosest to shore but various human activitiesclose to the shoreline. Samples closer toshore than about 4000m had more than fourtimes the levels found in the reference site.The outermost station off Klettagarðar,farthest away from sewage disposal, showedan elevation in PAH-level which mostprobably is due to release of oils or oil-relatedmaterials.

According to Norwegian guidelines for PAHsin mussels (Molvær et al. 1997), the area offÁnanaust, even closest to shore, fulfilled thehighest quality criteria and thus consideredsuitable for aquaculture and recreationalfishing. The levels off Klettagarðar werehigher, however. Only after discharge ofsewage has started in the planned disposal sitefrom the Klettagarðar STP will it beconclusive if and to what degree the sewageitself is the source of these substances closestto coast. For several decades comprehensivehandling of petroleum products took place inthe area around and off Klettagarðar. Thedisturbance caused by removal of the large oildepot in the area during the investigation forKlettagarðar may have affected the resultsobtained there. However, comparisons withmussels from other areas around N-Atlantic,including mussels for human consumption,show that even the highest levels found inmussels off the north coast of Reykjavík, i.e.off Klettagarðar, are very low (Guðjón AtliAuðunsson 2006).

Maximum limits have recently beenestablished for benzo(a)pyrene in mussels asfood commodity (Commission regulation(EC) No 208/2005 of 4 February 2005). Themussels off both Ánanaust and Klettagarðarcontain benzo(a)pyrene less than one-tenth of

these limits, even the samples containing thehighest levels off Klettagarðar.

59

8. CONCLUSIONS

Comprehensive studies have been carried outon sewage and the recipient of sewage fromReykjavik and neighbouring communities, thearea north of Reykjavík, Sundin. This reportfocuses on the recipient of sewage from thesewage treatment plant at Ánanaust, Skolpa,although results from the Klettagarðar site arealso presented.

The recipient is freely open to ocean waterand is characterised as a high energy area dueto strong tidal currents and wave motion. Theexchange rates are therefore fast and result ina dilution factor of about 1000 on averageabove the diffuser.

The bottom of the recipient area reflects ahigh energy where sand and gravel dominate.The bottom sediment is moved several timeseach year, especially during the winter time.

This study shows that the discharge ofnutrients (nitrogen and phosphorus) will notcause increased rate of algal growth in therecipient with resulting adverse effects. Thisis shown by empirical and semi-empiricalmodels of worst case scenarios. Thisindicates that the currently applied treatmentof sewage up to the planned maximum rate ofsewage discharge equivalent to 150,000person equivalents is more than sufficient tofulfill international criteria. Further treatmentof sewage than applied at present (screeningfollowed by sand sedimentation and fatfloatation) will not result in any detectableenvironmental improvement as regards algalgrowth.

Modelling dispersion of faecal bacteriaoutlines the largest area of sewage effects inseawater and thereby the dilution area. Theaccumulation of bacteria in deployed musselsduring summer time, during which shortesthalf-lives of faecal bacteria occur, shows thata much smaller area is affected than predictedon an annual basis. The present disposal ofsewage has resulted in a more confineddispersion area of faecal bacteria than past

near shore disposal. More stringent treatmentof sewage than applied at present, i.e.secondary treatment, will only marginallydecrease the dipersion zone.

Studies and modelling of currents and wavemotion indicate that the settling of sewageparticles onto the sediments is highlyimprobable. This settling of particles is aprerequisite for any detrimental effects on thebenthic community of organisms andsediment chemistry. Therefore, reducedoxygen above the sediments due to sewageparticles alone will not occur. A worst casescenario might result in the settling of sewageparticles to the bottom but only for a veryshort time (few hours during summer) and at arate experimentally known to cause no effectson organisms on or in the sediments nor resultin but very slight oxygen reduction above thesediments for such occasional short periods(hours). Sediment traps show unambigouslythat neither is there increase in organic matterat the disposal site nor a change in thecomposition (nutrients, trace elements). Asurvey of the benthic communities at thedisposal site did not indicate any effectscaused by the present discharge. Therefore,further treatment of sewage than applied atpresent will not result in any environmentalimprovement up to the planned maximum rateof sewage discharge equivalent to 150,000person equivalents.

Studies of the accumulation of trace elements,organohalogen compounds, and polyaromatichydrocarbons in blue mussels show that thepresent discharge of sewage renders therecipient to fit the highest environmentalquality criteria in Norway and well belowmaximum limits stipulated for seafood. Ofall the chemicals studied, only silver seems tobe affected by past discharge of sewage, asituation that has also been observedinternationally. The present disposal ofsewage has improved the situation markedlyas compared with past disposal throughprovisional outlets. More stringent treatmentof sewage or secondary treatment is not

60

expected to reduce these concentrations ofsilver.

The comprehensive studies on the watersnorth of Reykjavík receiving sewagedemonstrate that information on currents andpreferably also wave motion at planneddisposal sites is necessary for model-basedpredictions of the fate of sewage-derivedconstituents provided the composition andbehaviour of the sewage itself is known.These predictions are also sufficient ifevaluations of worst case scenarios giveunambigous results as regards acceptableeffects with respect to eutrophicationpotential, dispersion of faecal bacteria, changein macrobenthic communities, accumulationof sewage particles onto the sediments, andaccumulation of contaminants in sedimentsand biota. Only if one or more of theseeffects are not within acceptable limits, e.g.those set by experiments in internationallyaccepted research and/or set by internationalorganisations, should there be further studiesdirected specifically at these possible effects.

61

AcknowledgementsKjartan Thors, Geology office of KjartanThors, read the manuscript and madenumerous constructive and useful commentsfor which he is gratefully acknowledged.Snorri Páll Kjaran, Vatnaskil ConsultingEngineers, is gratefully thanked forcontributing and to give his permission to usethe material of chapter 3. Thanks are to JónÓlafsson, Icelandic Marine Research Instituteand University of Iceland, who made valuablecomments and suggestions. Jörundur

Svavarsson, Department of Biology,University of Iceland, read chapter 6 andmade useful comments. Gunnar SteinnJónsson, Environment and Food Agency ofIceland, and Hermann Þórðarson,Technological Institute of Iceland (IceTec) arethanked for their valuable suggestions andcomments on the manuscript. Finally, thanksto Sigurður Ingi Skarphéðinsson, OrkuveitaReykjavíkur (Reykjavik Energy), for fruitfuldiscussions and patience during this work.

62

9. REFERENCES

Agnar Ingólfsson 1977. Research in Skerjafjörður, II.Littoral zone organisms. Results of studies duringsummer 1975. University of Iceland BiologyDepartment, Report nr. 10. (in Icelandic)

Anon. 1998. Predicting the Long-Term Fate ofSediments and Contaminants in Massachusetts Bay.USGS fact Sheet FS 172-97. February 1998.

Anon. 1999. Waves, tides and shallow-waterprocesses. Bearman, G. editor. ButterworthHeinemann in association with The Open University,Walton Hall, Milton Keynes, England.

Arnþór Garðarsson and Kristín Aðalsteinsdóttir 1977.Research in Skerjafjörður, I. Bottom dwellingorganisms. Results of studies during June 1975.University of Iceland Biology Department, Report nr.9. (in Icelandic)

Arnþór Garðarsson, Jónbjörn Pálsson og AgnarIngólfsson 1974. Sampling and mapping of aquaticorganisms in the southern part of Leiruvogur nearReykjavík. University of Iceland Biology Department,Report nr. 2. (in Icelandic)

Baker, E.K., Harris, P.T., Kensett-Smith, B., Bagster,D.F., and Nobbs, D.M. 1995. Physical properties ofsewage particles in seawater. Mar.Poll.Bull., 30: 247-252.

Berelson, W.M., Johnson, K., Coale, K., and Li, H.-C.2002. Organic matter diagenesis in the sediments ofthe San Pedro Shelf along a transect affected bysewage effluent. Cont. Shelf Res., 22: 1101-1115.

Bickford, G.P. 1996. The effects of sewage organicmatter on biochemical processes within mid-shelfsediments offshore Sydney, Australia. Mar.Poll.Bull.,33: 168-181.

Boehm, A.B., and Weisberg, S.B. 2005. Tidal forcingof enterococci at marine recrational beaches atfortnightly and semidiurnal frequencies.Environ.Sci.Technol., 39: 5575-5583.

Brooks, N. H. 1960. Diffusion of sewage effluent inan ocean current. 1st Int. Conf. on Waste Disposal inthe Marine Environment, University of California,Berkley, Calif., 246–267.

Burt, T.N., and Turner, K.A. 1983. Deposition ofsewage flux on a rippled sand bed. HydraulicsResearch Report IT248.

Carvalho, J.L.B., Roberts, P.J.W, and Roldao, J. 2002.Field oservations of Ipanema outfall. J. HydraulicEngineering, 128: 151-160.

C.E.C. 1991. Council Directive of 21 May 1991concerning urban waste water treatment(91/271/EEC). Official Journal of the EuropeanCommunities, L135 of 30.5.91, 40-52.

Christoulas, D.G., Yannakopoulos, P.H., andAndreadakis, A.D. 1998. An empirical model forprimary sedimentation of sewage. Env.Int., 24: 925-934.

Chu, W.H., and Goldberg, H.B. 1974. Buoyant forcedplumes in a cross flow. J.Hydraul.Div.-Am.Civ.Eng.,100 (Hyg): 1204-1214.

City of Reykjavík’s Environmental and TechnicalSector 1997. Results of research on the recipient ofsewage from the STP at Ánanaust. Summaryprepared for the classification of the recipient.September 1997. (in Icelandic)

City of Reykjavík’s Environmental and TechnicalSector 2000. Results of research on the recipient ofsewage from the STPs at Ánanaust and Klettagarðar.Summary prepared for the classification of therecipient. October 2000.

Cromey, C.J., Black, K.D., Edwards, A., and Jack, I.A.1998. Modelling the deposition and biological effectsof organic carbon from marine sewage discharges.Est.Coast.Shelf.Sci., 47: 295-308.

Cromey, C.J., Nickel, T.D., and Black, K.D. 2002.DEPOMOD-modelling the deposition and biologicaleffects of waste solids from marine cage farms.Aquaculture, 214: 211-239.

CSTT (1994) Comprehensive Studies for the Purposesof Articles 6 of Directive 91/271 EEC, The UrbanWaste Water Treatment Directive. Marine PollutionMonitoring Management Group, ComprehensiveStudies Task Team. Dept of the Environment, NorthernIreland, Environment Agency, Scottish EnvironmentProtection Agency and Water Services Association, 1February 1994, pp. 41+figs.

CSTT (1997) Comprehensive Studies for the Purposesof Articles 6 & 8.5 of Directive 91/271 EEC, TheUrban Waste Water Treatment Directive, 2nd ed.Marine Pollution Monitoring Management Group,Comprehensive Studies Task Team. Dept of theEnvironment, Northern Ireland, Environment Agency,Scottish Environment Protection Agency and WaterServices Association, 13 January 1997, pp. 51+figs.

63

Degens, E.T. and Mopper, K. 1976: FactorsControlling the Distribution and Early Diagenesis ofOrganic Material in Marine Sediments. In ChemicalOceanography (J.P.Riley and R.Chester, eds.), Vol. 6,Ch. 31, pp. 59-113.

Einar Jónsson 1976. Pollution studies inSkerjafjörður, The effect of runoff on bottomdwelling organisms (summer 1976). Marine ResearchInstitute, Report nr. 4. (in Icelandic)

Farley, K.J. 1990. Predicting organic accumulationsin sediments near marine outfalls.J.Environ.Engineering., 116: 144-165.

Findlay, R.H., and Watling, L. 1997. Prediction ofbenthic impact for salmon net-pens based on thebalance of benthic oxygen supply and demand.Mar.Ecol.Prog.Ser., 155: 147-157.

Fischer, H.B., List, E.J., Koh, R.C.Y., Imberger, J.,Brooks, N.H. 1979. Mixing in inland and coastalwaters. Academic Press, New York, NY.

Graf, G., and Rosenberg, R. 1997. Bioresuspensionand biodeposition: a review. J.Mar.Syst., 11: 269-278.

Guðjón Atli Auðunsson 1992. Analysis of chemicalconstituents in wastewater in Reykjavík. Performedfor the City of Reykjavík’s Environmental andTechnical Sector. December 1992. Report Rf 9. (inIcelandic)

Guðjón Atli Auðunsson 1994a. Analysis of chemicalconstituents in wastewater in the Gelgjutangipumping station. Work for the City of Reykjavík’sEnvironmental and Technical Sector. May 1994.Report Rf 50. (in Icelandic)

Guðjón Atli Auðunsson 1994b. Analysis of pollutantsin deployed mussels in the sea near Reykjavík . Workfor the City of Reykjavík’s Environmental andTechnical Sector. June 1994. Report Rf 51. (inIcelandic)

Guðjón Atli Auðunsson and Hannes Magnússon 1995a.Analysis of pollutants in deployed mussels in the seanear Reykjavík in 1994. Extension-I. Work for theCity of Reykjavík’s Environmental and TechnicalSector. September 1995. Report Rf 96. (in Icelandic)

Guðjón Atli Auðunsson and Hannes Magnússon 1995b.Analysis of pollutants in deployed mussels in the seanear Reykjavík in 1995. Extension -II. Work for theCity of Reykjavík’s Environmental and TechnicalSector. October 1995. Report Rf 98. (in Icelandic)

Guðjón Atli Auðunsson 1995. Analysis of pollutants indeployed mussels in the sea near Hafnarfjörður in1994. Work for the Environmental and Health Institute

of Hafnarfjörður. June 1996. Report Rf 121. (inIcelandic)

Guðjón Atli Auðunsson 1996. Chemical compositionof waste water in the Faxaskjól pumping station anddispersion of organic matter in the sea at Ánanaust.Work for the City of Reykjavík’s Environmental andTechnical Sector. October 1996. Report Rf 136. (inIcelandic)

Guðjón Atli Auðunsson 2000. Behaviour andcomposition of wastewater in the sewage treatmentplant at Ánanaust. Work for the City of Reykjavík’sEnvironmental and Technical Sector. March 2000.Report Rf 4-00. (in Icelandic)

Guðjón Atli Auðunsson 2001a. Research withdeployed mussels off Klettagarðar 1998. Work for theCity of Reykjavík’s Environmental and TechnicalSector. June 2001. Report Rf 10-01. (in Icelandic)

Guðjón Atli Auðunsson 2001b. Sediment trap studiesoff Ánanaust '95-'96 and Klettagarðar '98-'99:marine treatment of wastewater. Work for the City ofReykjavík’s Environmental and Technical Sector.October 2001. Report Rf 26 - 01. (in Icelandic)

Guðjón Atli Auðunsson 2002. Behaviour andcomposition of wastewater in the sewage treatmentplant at Ánanaust 2000-2001. Work for the City ofReykjavík’s Environmental and Technical Sector.March 2002. Report Rf 06-02. (in Icelandic)

Guðjón Atli Auðunsson 2005. Sediment trap studiesoff Ánanaust ’00-’01: marine treatment of sewage.Work for Orkuveita Reykjavíkur. ReportITÍ0605/EGK01. (in Icelandic)

Guðjón Atli Auðunsson 2006. Monitoringcontaminants with blue mussels: Ánanaust. Work forOrkuveita Reykjavíkur. June 2006. ReportITÍ0606/EGK02. (in Icelandic)

Guðmundur Víðir Helgason and Jörundur Svavarsson1991. Benthic organisms in Þerneyjarsund.University of Iceland Biology Department, Edition 30.(in Icelandic)

Guðmundur V. Helgasaon and Arnþór Garðarson 1992.Inspection studies in Viðeyjarsund and Eiðsvík 1985 -1992. Part I. Benthos. Performed for Reykjavik CityEnvironmental Health and Protection Office, January1992. (in Icelandic)

Guðmundur V. Helgasaon and Arnþór Garðarson 1995.Inspection studies in Viðeyjarsund and Eiðsvík 1985 -1992. Part II. Environmental factors and faecalbacteria. Performed for Reykjavik City EnvironmentalHealth and Protection Office, January 1995. (inIcelandic)

64

Heip, C.H.R., Goosen, N.K., Herman, P.M.J.,Kromkamp, J., Middelburg, J.J.and Soetaert, K. 1995.Production and Consumption of Biological Particlesin Temperate Tidal Estuaries.Oceanogr.Mar.Biol.Ann.Rev., 33: 1-149.

Hveragerðisbær and Háskólasetrið í Hveragerði 2004.The sewage treatment plant of Hveragerði. Annularreport 2004. (In Icelandic).

Icelandic Fisheries Laboratory 1981. Shellfishharvesting area. Sustainable development of the area128 c in Faxaflóí. Technical report nr. 127. (inIcelandic)

ISOTOPCENTRALEN 1971a. Recipient research inReykjavik in 1970. RAPPORT Köbenhavn16.08.1971. 545.07. (in Danish)

ISOTOPCENTRALEN 1971b. Recipient research inReykjavik in 1970. FORKORTETBILAGSSAMLING Köbenhavn 16.08.1971. 545.07.(in Danish)

ISOTOPCENTRALEN 1971c. Recipient research inReykjavik in 1970. KOMPLET BILAGSSAMLING.Köbenhavn 16.08.1971. 545.07. (in Danish)

A.B.Josefson and B.Rasmussen 2000. NutrientRetention by Benthic Macrofaunal Biomass ofDanish Estuaries: Importance of Nutrient Load andResidence Time. Est.Coast.Shelf Sci., 50: 205-216.

Jón Ólafsson, Guðjón Atli Auðunsson, StefánEinarsson and Magnús Daníelsson 1994.Organochlorines, heavy metals and nutrients inIcelandic waters in (Unnsteinn Stefánsson, editor)"Icelanders, the ocean and its resources", Societasscientiarum Islandica , Conference publication IV, 225-251. (in Icelandic).

Jón Ólafsson 1986. Trace metals in mussels (Mytilusedulis) from southwest Iceland. Mar.Bio., 90: 223-229.

Jón Ólafsson 1983. Trace metals in blue musselsfrom southwest Iceland. Marine Research Institute.Fjölrit nr. 10, 50 pp. (in Icelandic)

Jón Ólafsson 1987. Chemical Composition of ColdWater in Drillholes near Reykjavík , A Project forVatnsveita Reykjavíkur, January 1987.

Jón Ólafsson, Magnús Danielsen, Sólveig Ólafsdóttirand Þórarinn Arnarson 1996. Nutrients in the sea atÁnanaust in November 1995. Marine ResearchInstitute, Fjölrit nr. 50. (in Icelandic)

Jón Ólafsson and Sólveig R. Ólafsdóttir 2000. State ofthe seawater at the disposal site of sewage off

Ánanaust in February 2000. Marine ResearchInstitute, Fjölrit nr. 81. 49pp. (in Icelandic)

Jón Skúlason 1998. Sewage treatment plant atKlettagarðar. Study on unconsolidated layers.Performed for the City of Reykjavík’s Environmentaland Technical Sector. Almenna verkfræðistofan hf,1998. (in Icelandic)

Jörundur Svavarsson 1996. Benthic organisms at theplanned sewage outlet at Ánanaust. Performed for theCity of Reykjavík’s Environmental and TechnicalSector. University of Iceland Biology Department,1996. (in Icelandic)

Jörundur Svavarsson 2000. Benthic organisms at theplanned sewage outlet off Klettagarðar. Work for theCity of Reykjavík’s Environmental and TechnicalSector. University of Iceland Biology Department,2000. (in Icelandic)

Jörundur Svavarsson 2002. Benthic biota in therecipient of sewage from the outfall of Ánanaust-status after the operation started. Report to the Cityof Reykjavík’s Environmental and Technical Sector.Líffræðistofnun Háskólans. (in Icelandic)

Karl Gunnarsson og Konráð Þórisson 1976. The effectof sewage on the distribution and cover of littoralalgae near Reykjavik, Preliminary results. ActaBot.Isl., 4(1976)58-66.

Karl Gunnarsson, January 1993. Investigations inViðeyjarsund and Eiðsvík 1985-1992, Part II.Macroalgae. Environmental and Health Institute ofReykjavík, January 1993. (in Icelandic)

Karl Gunnarsson and Agnar Ingólfsson 1995.Seasonal changes in the abundance of intertidalalgae in southwestern Iceland. Bot.Marina, 38: 69-77.

Kelly, , J.R., and Nixon, S.W. 1984. Experimentalstudies of the effect of organic deposition on themetabolism of the coastal marine bottom community.Mar.Ecol.Prog.Ser., 17: 157-169.

Kjartan Thors, Árni Sigurðsson and Jón Jónsson 1994.Research on the seafloor at Ánanaust 1994. Reporton sampling for the City of Reykjavík’s Environmentaland Technical Sector. Marine Research Institute, June1994. (in Icelandic)

Kjartan Thors 1994. Videotaping of the seafloor atthe end of the proposed sewage outlet in Ánanaust.Report of photographs taken May 20, 1994 for the Cityof Reykjavík’s Environmental and Technical Sector.(Videotape included). Marine Research Institute, May1994. (in Icelandic)

65

Kjartan Thors 1999. The seafloor along the oceanoutfall from Klettagarðar. Work for the City ofReykjavík’s Environmental and Technical Sector.Geology office of Kjartan Thors. September 1999. (inIcelandic)

Kjartan Thors 2000. Ocean outfalls from Ánanaustand Klettagarðar: transport of sediments monitoredby photographing. Work for the City of Reykjavík’sEnvironmental and Technical Sector. Geology officeof Kjartan Thors. November 2000. (in Icelandic)

Kjartan Thors 2003. The Ánanaust outfall: Sedimenttransport examined by photography. Work for theCity of Reykjavík’s Environmental and TechnicalSector. Geology office of Kjartan Thors. June 2003.(in Icelandic)

Koh, C.Y., Brooks, N.H. 1975. Fluid mechanics ofwaste-water disposal in the ocean. Annual Reviews inFluid Mechanics 7, 187-211.

Koh, C.Y. 1982. Initial sesimentation of wasteparticulates discharged from ocean outfalls.Environ.Sci.Technol., 16: 757-763.

Kristinn Guðmundsson and Kristín J. Valsdóttir 2004.Measurements of primary production in the years1958-1999. Extent, methods and evaluation. MarineResearch Institute, Fjölrit nr. 107. (In Icelandic).

Kristinn Guðmundsson, Þórunn Þórðardóttir andGunnar Pétursson 2004. Computation of dailyprimary production in Icelandic waters; a comparisonof two different approaches. Marine ResearchInstitute, Fjölrit nr. 106.

Lavelle, J.W., Ozturgut, E., Baker, E.T., Tennant,D.A., and Walker, S.L. 1988. Settling speeds ofsewage sludge in seaweater. Environ. Sci.Technol.,22: 1201-1207.

Livingston, R.J. 1982. Long-term variability incoastal systems: background noise and environmentalchange. In: Mayer, G.F. (ed.) Ecological stress and theNew York Bight: science and management. Columbia,S. Carolina,: Estuarine Research Federation, 605-620.

Luketina, D. 1998. Simple tidal prism modelsrevisited. Est.Coast.Shelf Sci., 46(1998)77-84.

Maughan, J.T., and Oviatt, C.A. 1993. Sediment andbenthic response to waste water solids in a marinemesocosm. Water Env.Res., 65: 879-889.

Mearns, A.J., and Word, J.Q. 1982. Forecastingeffects of sewage solids on marine benthiccommunities. In: Mayer, G.F. (ed.) Ecological stressand the New York Bight: science and management.

Columbia, S. Carolina, Estuarine Research Federation,495-512.

Metcalf & Eddy 1991. Wastewater Engineering.Treatment, Disposal, and Reuse. Third Edition.Revised by G.Tchobanoglous and F.L.Burton.McGraw-Hill, Inc.

Molvær, J., Knutzen, J., Magnusson, J., Rygg, B., Skei,J., and Sørensen, J. 1997. Klassifisering avmiljøkvalitet i fjorder og kystfarvann. Veiledning.Statens forurensningstilsyn, SFT. 1997. (inNorwegian)

Morrisey, D.J., Gibbs, M.M., Pickermere, SE., andCole, R.G. 2000. Predicting impacts and recovery ofmarine-farm sites in Stewart Island, New Zealand,from the Findlay-Watling model. Aquaculture, 185:257-271.

Müller, G. 2003. Review. Sense or no-sense of thesum parameter for water soluble “adsorbable organichalogens” (AOX) and “adsorbed organic halogens”(AOX-S18) for the assessment of organohalogens insludges and sediments. Chemopshere, 52: 371-379.

National Energy Authority, Hydraulic Laboratory,1967. City of Reykjavík’s Environmental andTechnical Sector current research in Fossvogur,Report nr. 1 research results. August 1967. (inIcelandic)

National Energy Authority, Hydraulic Laboratory,2006. Data bank of the Hydraulic Laboratory, handlingnr. 2006/41.

Neves, M.J., and Fernando, H.J.S. 1995.Sedimentation of particles from jets discharged byocean outfalls: a theoretical and laboratory study.Wat.Sci.Tech., 32: 133-139.

Neves, M.J., and Fernando, H.J.S. 1995. Prediction ofparticle deposition in effluent disposal systems.Presentation at 2nd International Conference on MarineWaste Water Discharges. MWWD 12002-IstanbulSept. 16-20.

Nixon, S.W. 1995. Coastal Marine Eutrophication: ADefinition, Social Causes, and Future Concerns.OPHELIA, 41: 199-219.

Nixon, S.W., J.W.Ammermann, L.P.Atkinson,V.Berounsky, G.Billen, W.C.Boicourt, W.R.Boynton,T.M.Church, D.M.Ditoro, R.Elmgren, J.H.Garber,A.E.Giblin, R.A.Jahnke, N.J.P.Owens, M.E.Q.Pilson,and S.P.Seitzinger 1996. The fate of nitrogen andphosphorus at the land-sea margin of the NorthAtlantic Ocean. Biogeochemistry, 35: 141-180.

66

Nordisk Ministerråd 1993. Miljøfremmende stoffer ikommunalt spildevand - stoffer, kilder ogbegrænsningsmuligheder. Nordisk Seminar- ogArbejdsrapport 1993: 515. Nordisk Ministerråd,København, 1993. (in Danish)

Okubo, A. 1971. Oceanic diffusion diagrams. Deep-Sea Res., 18(8), 789–802.

Okubo, A. 1974. Some speculations on oceanicdiffusion diagrams. Rapports et Procesverbaux desReunions Conseil International pour l’Exploration dela Mer, 167, 77–85.

OSPAR 2001. Annex 5: Draft Common AssessmentCriteria and their Application within theComprehensive Procedure of the CommonProcedure. Meeting Of The Eutrophication TaskGroup (Etg), London (Secretariat): 9-11 October 2001.OSPAR Convention For The Protection Of The MarineEnvironment Of The North-East Atlantic.

Oviatt, C.A., Quinn, J.G., Maughan, J.T., Ellis, J.T.,Sullivan, B.K., Gearing, J.N., Hunt, C.D., Sampou,P.A., and Latimer, J.S. 1987. Fate and effects ofsewage sludge in the coastal marine environment: amescosm experiment. Mar.Ecol.Prog.Ser., 41: 187-203.

Pearson, T.H., and Rosenberg, R. 1978. Macrobenthicsuccession in relation to organic enrichment andpollution of the marine environment.Oceanogr.Mar.Biol.Ann.Rev., 16: 229-311.

Petrenko, A.A., Jones, B.H., and Dickey, T.D. 1998.Shape and initial dilution of Sand Island, Hawaiisewage plume. J.Hydraulic Engineering, 124: 565-571.

Quetin, B., and De Rouville, M. 1986. SubmarineSewer Outfalls. Mar.Poll.Bull., 17, Nr. 4.

Roberts, P.J.W. 1991. Ocean Outfalls.Crit.Rev.Environ.Control., 20: 311-339.

Robets, P.J.W. 1999a. Modelling mamala Bay outfallplumes. II. Far field. J. Hydraulic Engineering, 125:574-583.

Robets, P.J.W. 1999b. Modelling mamala Bay outfallplumes. I. Near field. J. Hydraulic Engineering, 125:564-573.

Roberts, P.J.W., and Toms, G. 1987. Ocean OutfallSystem for Dense and Buoyant Effluents. J.Environ.Engineering, 114: 1175-1191.

Roberts, P.J.W., Snyder, W.H., and Baumgartner, D.J.1989a. Ocean outfalls. I. Submerged wastefieldformation. J. Hydraulic Engineering, 115, 1-25

Roberts, P.J.W., Snyder, W.H., and Baumgartner, D.J.1989b. Ocean outfalls. II. Spatial evolution ofsubmerged wastefield. J. Hydraulic Engineering, 115,26-48.

Roberts, P.J.W., Snyder, W.H., and Baumgartner, D.J.1989b. Ocean outfalls. III. Effect of diffuser designon submerged wastefield. J. Hydraulic Engineering,115, 49-70.

Sherwin, T.J. 2000. The significance of residualcurrents in the interpretation of the EU urbanwastewater treatment directive in coastal locations.Mar. Poll. Bull., 40: 17-21.

Sigurður Reynir Gíslason, Björn Þór Guðmundsson andEydís Salome Eiríksdóttir 1998. Chemicalcomposition of Elliðaár 1997-1998. The ScienceInstitute of the University of Iceland, RH-19-98. (InIcelandic)

Sigurjón Rist 1986. Chemical analysis of sufacewaters. Borgarfjörður and Elliðaár in Rekjavík.National Energy Authority OS-86070/VOD-03.Reykjavík, November 1986. (in Icelandic)

Sjötun, K., and Gunnarsson, K. 1995. Seasonalgrowth pattern of an Icelandic Laminaria population(section Simplices) containing solid and hollow-stipedplants. European Journal of Phycology, 30: 281-287.

Sólmundur Einarsson 1973. Results of research on theaquatic organisms in Fossvogur and Gufunes.Marine Research Institute (report for the Environmentaland Health Institute). (in Icelandic)

Sólveig R. Ólafsdóttir 2006. Concentrations ofnutrients in Icelandic waters. Marine ResearchInstitute, Fjölrit nr. 122. (In Icelandic)

Sólveig R. Ólafsdóttir and Jón Ólafsson 1999. Input ofdissolved constituents from River Þjórsá to S-Icelandcoastal waters. Rit Fiskideildar 16: 79-88.

Stockholm Vatten AB 1990. Wastewater fromhouseholds. April 1990. (in Swedish)

Strain, P.M., Wildish, D.J., and Yeats, P.A. 1995. Theapplication of simple models of nutrient loading andoxygen demand to the management of a marine tidalinlet. Mar.Poll.Bull., 30(1995)253-261.

Strain, P.M., and Yeats, P.A. 1999. The relationshipbetween chemical measures and potential predictorsof the eutrophication status of inlets. Mar.Poll.Bull.,38(1999)1163-1170.

Stacey, M.T., Cowen, E.A., Powell, T.M., Dobbins, E.,Monismith, S.G., and Koseff, J.R. 2000. Plumedispersion ina stratified, near-coastal flow:

67

measurements and modeling. Cont.Shelf Res., 20:637-663.

Svend-Aage Malmberg 1967. Current measurementsin the sea. Náttúrufræðingurinn, 37(1968)64-76). (inIcelandic)

Svend-Aage Malmberg 1968. Direct measurements ofocean currents. Marine Research Institute publication,9-19, 1968. (in Icelandic)

Tett, P. 2000. Marine eutrophication and the use ofmodels. In: Science and Environmental DecisionMaking, ed. Huxham, M. & Summer, D., AddisonWesley Longman/Pearson Education, London, 215-238.

Tett, P., Gilpin, L., Svendsen, H., Erlandsson, C. P.,Larsson, U., Kratzer, S., Fouilland, E., Janzen, C., Lee,J.-Y., Grenz, C., Newton, A., Ferreira, J. G., Fernandes,T. & Scory, S. 2003. Eutrophication and someEuropean waters of restricted exchange. ContinentalShelf Research, 23: 1635-1671.

Unnsteinn Stefánsson og Guðmundur Guðmundsson1978. The Freshwater Regime of Faxaflói,Southwest Iceland, and its Relationship toMeteorological Variables. Est.Coast.Mar.Sci., 6: 525-551.

Unnsteinn Stefánsson, Þórunn Þórðardóttir og JónÓlafsson 1987. Comparison of seasonal oxygencycles and primary production in the Faxaflói region,southwest Iceland. Deep-Sea Res., 34: 725-739.

Unnsteinn Stefánsson og Jón Ólafsson 1991. Nutrientsand fertility of Icelandic waters. Rit Fiskideildar,Bindi XII, no. 3.

Unnsteinn Stefánsson 1981. Particulate matter inIcelandic surface waters. Rit Fiskideildar, 6(1): 14-31

USEPA 1985. Coastal Marinas AssessmentHandbook. United States Protection Agency, RegionIV, Atlanta, Georgia.

Vatnaskil Consulting Engineers 1984. Numericalmodel for pollutant dispersion. Research for outlet atGrafarvogur. Work for the City of Reykjavík’sEnvironmental and Technical Sector, June 1984.84.06. (in Icelandic)

Vatnaskil Consulting Engineers 1989. Currentmeasurements in Eiðsvík and Gullinbrú. Letter toStefán Hermannsson, City of Reykjavík’sEnvironmental and Technical Sector, 27.10.1989. (inIcelandic)

Vatnaskil Consulting Engineers 1990. Pollution in thesea in Hafnarfjörður. Work for the city ofHafnarfjörður. September 1990. 90.15. (in Icelandic)

Vatnaskil Consulting Engineers 1991. Pollution in thesea from sewage outlets. Work for the City ofReykjavík’s Environmental and Technical Sector.December 1991. 91.11. (in Icelandic)

Vatnaskil Consulting Engineers 1992. Coastalcontamination from a wastewater outlet offshorefrom Eiðsgrandi. February 1992. Report no: 92.05.(in Icelandic)

Vatnaskil Consulting Engineers 1993. Sewagedisposal study U.S. Naval Station Keflavík, IcelandLocation of wastewater outlets. Prepared forDepartment of the Navy. Atlantic Division. NavalFacilities Engineering Command. September 1993.93.04. (in Icelandic)

Vatnaskil Consulting Engineers 1994. Ocean currentsin Faxaflóí and dispersion of pollutants from asewage outlet in Eiðsgrandi. Work for the City ofReykjavík’s Environmental and Technical Sector.December 1994. 94.17. (in Icelandic)

Vatnaskil Consulting Engineers 1995. Ocean currentsin Hafnarfjörður and dispersion of pollutants fromproposed sewage outlet in Hafnarfjörður. January1995. 95.01. (in Icelandic)

Vatnaskil Consulting Engineers 1996. Eiðsgrandi.Pollution from sewage outlet in Ánanaust. Work forthe City of Reykjavík’s Environmental and TechnicalSector. March 1996. 96.02. (in Icelandic)

Vatnaskil Consulting Engineers 1999. Ocean currentsin Faxaflóí and dispersion of pollutants from sewageoutlets in Ánanaust and Laugarnes. Work for the Cityof Reykjavík’s Environmental and Technical Sector.99.03. (in Icelandic)

Vatnaskil Consulting Engineers 2000. Dispersion ofpollutants from a sewage outlet in Ánanaust. Workfor the City of Reykjavík’s Environmental andTechnical Sector. April 2000 00.06. (in Icelandic)

Washburn, L., Stone, S., and MacIntyre, S. 1999.Dispersion of produced water in a coastalenvironment and its biological implications.Cont.Shelf Res., 19: 57-78.

Wassmann, P. 1991. Dynamics of primary productionand sedimentation in shallow fjords and polls ofWestern Norway. Oceanogr.Mar.Biol.Ann.Rev., 29:87-154.

Þórunn Þórðardóttir og Unnsteinn Stefánsson 1977.Productivity in Relation to Environmental Variables

68

in the Faxaflói Region 1966-1967. ICES C.M.1977/L:34.

Þórunn Þórðardóttir 1994. Phytoplankton andprimary production in the seas around Iceland in(Unnsteinn Stefánsson, editor) "Icelanders, the oceanand its resources", Societas scientiarum Islandica,Conference publication IV, 65-88. (in Icelandic)

Þórunn Þórðardóttir 1986. Timing and Duration ofSpring Blooming South and Southwest of Iceland. InThe Role of Freshwater Outflow in Coastal MarineEcosystems, S.Skreslet ed.. NATO ASI Series, vol. 7.Springer-Verlag Berlin Heidelberg 1986. pp. 345-360.

Zimmerman, J. T. F. 1986. The tidal whirlpool: Areview of horizontal dispersion by tidal and residualcurrents. Netherlands J. Sea Res., Amsterdam, 20,133–156.

Zeitoun, , M.A., Reid, R.O., McHilhenny, W.F., andMitchell, T.M. 1972. Model studies of outfall systemsfor desalination plants. Part III. Numerical anddesign considerations. Res. and Devel. Progress Rep.No. 804, Office of Saline Water, U.S.Dept. of Interior,Washington, D.C.

69