0160-4120/93$6.00 +.00 [email protected]
Environment International, Vol.19,pp. 291-299, 1993
Printed in the U.S.A. All fightsreserved.
TREATMENT AND DISPOSAL OF THE WASTEWATER OF THESSALONIKI, GREECE A.D. Andreadakis Water Resources Division, National Technical University of Athens, Zografos, Athens 15773, Greece
T.E. Agelakis Hydroelektriki Ltd., Athens, Greece
D.V. Adraktas Adraktas and Associates, Athens, Greece
E19208-178 M (Received 15 August 1992; accepted 24 February 1993)
The wastewater from the city of Thessaloniki is discharged without treatment to the nearby inner part of the Thessaloniki Gulf. The existing treatment plant is designed for primary treatment and does not operate since the expected effluent quality is not suitable for safe disposal to the available recipients. The paper describes the results of a recent two-year study, aiming to investigate suitable treatment and disposal alternatives and to propose an appropriate upgraded treatment plant. The results of the investigation, which involved application of mathematical modelling of the quality parameters of the gulf, indicate that an environmentally acceptable and technico-economically sound solution consists of a biological treatment system with seasonal nitrogen removal, followed by discharge of the treated effluent to the central part of the gulf, through a marine outfall.
Thessaloniki is the second largest city of Greece, l o c a t e d in the n o r t h e r n part of the c o u n t r y with a population exceeding 1 000 000. The wastewater from the city has been, and to a large extent still is, discharged without treatment to the adjacent bay, which forms the inner part of the G u l f of T h e s s a l o n i k i (Fig. I). As a result of this uncontrolled discharge, serious deterioration of the bay waters has occurred, e v i d e n c e d by i n c r e a s e d levels of p h y t o p l a n k t o n concentrations, oxygen depletion of the bottom sea layers, and occasional offensive odors.
The need for remedial action has been recognized for some time and the construction of a treatment plant at a site approximately 7 km southwest of the city was completed in 1989. The plant operation consists of preliminary treatment (screening and grit and grease removal) followed by treatment in primary sedimentation tanks. In order to increase the efficiency of the primary treatment to around 40-45% in terms of B e D 5 removal, provision for aeration of the wastewater and the recirculated primary sludge from the sedimentation tank is made, in this way promoting the flocculation of the solids in the wastewater.
A.D. Andreadakis et al.
Upon completion of the treatment plant, it was realized that the originally selected effluent recipient, the Axios River, could not serve this purpose anymore, due to a combination of stricter standards imposed on the river quality and reduced summer flows. In view of this situation, the Ministry of Environment and Public Works, in cooperation with the Sewage Corporation of Thessaloniki, commissioned a study in 1989 aiming to investigate the required degree of
treatment prior to discharge to the Axios River or to an alternative recipient, and to design an appropriate upgraded treatment plant for the future anticipated total flow. The study was undertaken by a joint venture of two consultancies, Hydroelectriki Ltd. and Gofas and Partners, and was conducted over the period of 19901992. This paper presents the main results and conclusions of this study.
Fig. 1. Study area and segmentation of the Gulf of Thessaloniki.
Treatment and disposal of the wastewater of Thessaloniki
ALTERNATIVE TREATMENT--DISPOSAL SYSTEMS
The pollutional loads originating from the city wastewater were estimated by sampling and analysis and were found to be as follows: 50 000 kg BODs/d, 60 000 kg SS/d, 10 000 kg N/d, and 2 500 kg P/d. By the year 2 000, the anticipated loads will be 80 000 kg B O D s / d , 76 000 Kg SS/d, 16 000 Kg N/d, and 3 500 kg P/d. At a projected mean daily flow of 300 000 m3/d, the expected concentrations for BODs, SS, N, and P are 267 rag/L, 250 mg/L, 53 mg/L, and 11.5 rag/L, respectively. These pollutional loads enter the inner part of the Gulf of Thessaloniki resulting in serious deterioration of the bay waters. In addition to these loads, other sources of pollution include agricultural runoff (which is discharged to the sea through two pumping stations) and three rivers (Loudias, Axles, and Aliakmen), flowing into the central Gulf of Thessaloniki (Fig. 1). The poUutional loads from all the sources for the year 2000 are presented in Table 1 and show that, on a yearly basis, the wastewater from the city contributes to the organic load by 55% and to the nutrients load by about 35%. However, the contributions are higher during the summer months due to the reduced river flows, and in the case of nutrients, approaches a figure of 45-50%.
With respect to final effluent disposal possibilities, the following alternatives were investigated. Reuse for irrigation Reuse of the effluent from the treatment plant for
irrigation is an attractive alternative. On the one hand, it assists in preserving the available freshwater resources; on the other hand, it leads to a reduction of the amount of fertilizers used, and consequently, to a reduction of the nutrients load entering the gulf. However, there are considerable difficultiesin adopting reuse as a sole method of effluent disposal for the following main reasons. Although experience has shown that biologically treated effluents may be safely used for restricted irrigation (State of California 1978), it is not always easy to overcome the prejudices and fears of the farmers. This is even more true as it is practically impossible to exclude the possibility of unrestricted irrigation practices, accidental or not. The microbial characteristics of an effluent that is safe for unrestricted irrigation are not clearly defined yet. The WHO limit of 1000 coliforms per 100 mL (WHO 1989), achievable through clorinadon or use of stabilization ponds, is considered inadequate, although it has not been disproved by epidemiological studies. It receives strong criticism by the interna-
Table 1. Anticipatedpollutional loads to the Gulf of Thessaloniki(year 2000).
Pumping Stations Loudias Axios Aliakmon
6 3 10 4
I 1 5 2
524 240 831 650
161 315 1 500 190
450 400 000 540
tional scientific community, at least of the so-called developed part of the world. It is unlikely that the WHO criteria would be acceptable to the population of the study area, if on nothing else than purely socio-economical grounds. To meet some of the more conservative standards proposed, such as the California standards of 2.2 coliforms per 100 mL (State of California 1978), an advanced tertiary treatment system, including prior filtration and possibly coagulation-sedimentation, is required (Asano et al. 1992). This system will raise the treatment costs considerably. Normally, wastewater reuse can be practiced for only part of the year, which may not include some months critical for the assimilative capacity of the recipients, e.g. September. In addition to severe problems due to land availability constraints, the construction of effluent storage reservoirs increases the cost substantially. Reuse of wastewaters for irrigation, though ecologically attractive, has to be cost effective or at least economically comparable to other alternatives, based on the supply of fresh water. Considering the above, it was concluded that, although reuse for irrigation is a promising alternative, one could not expect it to be practiced in this case systematically on a large scale, at least for the time being or in the very near future. Discharge into the Axles River
The ecosystem of the river is protected under the RAMSAR agreement, thus an advanced treatment is required prior to discharging effluents to the river. In addition to carbon removal, nitrification is necessary throughout the year in order to avoid ammonia toxicity to fish. Since treatment systems achieving nitrification often suffer operational difficulties due to uncontrolled denitrification in the final clarifiers (especially at high temperatures ), the system adopted should also provide controlled denitrification of the nitrified effluent. The maximum permitted phosphorus concentration in the river cannot be met during the summer months; therefore, in addition to nitrogen, phosphorus removal should also be included in the treatment system. With respect to microbial pollution, the standards to be met are also high since the outlet of the river is adjacent to a sensitive sea shell growing coastal area. In the case of chlorination, a subsequent dechlorination stage is needed in order to reduce the residual chlorine in the river to acceptable levels for fish protection.
A.D. A n d r e a d a k i s et al.
Discharge into the sea In this alternative, it was decided to discharge the wastewater through an o u t f a l l at a depth of some
20 m, located outside the bay in the proximity of the existing drainage pumping stations (Fig. 1). By placing the discharge point away from the bay, a significant improvement of the water quality of the bay can be expected due to the enhanced dispersion of pollutants. The existing primary treatment facility, in conjunction with disposal through the marine outfall, is sufficient to prevent carbon-related impacts, such as oxygen depletion of the receiving water body. However, it was decided to adopt at least biological treatment, mainly for the following two reasons. The first is the recent EEC Directive (EEC 1991) regarding treatment of urban wastewaters, according to which any treatment less than conventional biological treatment should be considered as an exception. The second, probably more important reason, is the need for effective pathogen reduction, in order to protect the sea shell cultures in the vicinity of the discharge point (at a distance of about 4 km to the south). The dilution provided by the combined effect of the initial and subsequent surface dispersion (of the order of 500) is not sufficient to bring about a faecal coliform reduction adequate to conform with the strict limits imposed to the sea shell areas (maximum median concentration of 10-14 FC per 100 mL of water and 300 FC per 100 mL of interior liquid in the shells). The required additional pathogen reduction can be achieved by disinfection. This creates the need for a good quality effluent with low BOD5 and SS concentrations for effective pathogen destruction and reduced trihalomethanes formation (especially in the case of chlorination). To address the eutrophication problem, it was necessary to evaluate the impact of various alternative treatment schemes including a conventional biological treatment for carbon removal and systems for advanced treatment with varying degrees of nutrients removal. The evaluation was done on the basis of the results of a eutrophication-dissolved oxygen mathematical model of the gulf. Mathematical model: The mathematical model used incorporates physical transport and physicochemical and biological reactions and is based on the concept of conservation or continuity of mass. It was necessary to use a finite difference description of the spatial derivatives to solve the set of equations obtained. In practice, this is equivalent to considering the continuous body of water as a network of finite completely mixed cells or segments connected by dis-
Treatment and disposal of the wastewater of Thessaloniki
persive and convective transport. In finite difference form the continuity equation can be written as: Vs dCs = 2 Q s j (asj Cs + ~sj Cj) + ~ s j
(Cj - Cs) + rs + Ws
rAl = glAl - (KdA+ RA + KsA) A1
gl = Ilmax0 (t-2o) f (I)
NI Pt KN+NI Kp+PI
= concentration of species in cell s (mg/L) = concentration of species in cells j in
V, Q,j a,j
contact with cell s (mg/L) = volume of cell s (m 3) = net advective flow from cell s to j (m3/d) = a finite difference weighting factor
1-a,j dispersion coefficient (mZ/d)
L, Lj L,j r, w,
= = = = =
cross-sectional area between s and j (m 2) characteristic length of cell s (m) characteristic length of c e l l j (m) (L,+Lj)/2 reaction term for cell s (mg/L) loading into cell s (mg/L).
Using the above finite difference approach, a system of equations was obtained describing the transport and reactions of each species (chlorophyll, BODs, dissolved oxygen, phosphorus, nitrogen, and transparency) within each cell. The segmentation of the Gulf is shown in Fig. I. Equation 1 was applied for chlorophyll (A), BOD5(C), ammoniacal nitrogen (NH), oxidized nitrogen (NO) organic phosphorus (OP), inorganic phosphorus (IP), dissolved oxygen (DO), and transparency in terms of the Secci Disk (SD), resulting in a set of nonlinear expressions. The reaction terms describe phenomena such as phytoplankton growth, death, endogenous respiration and settlement, transformation of nutrients (uptake, hydrolysis, nitrification, denitrification), settlement of nutrients, release of nutrients due to phytoplankton death, photosynthesis, organic carbon degradation, oxygen consumption, surface reaeration, etc. As an example, Equation 1 applied to segment 1 concerning chlorophyll is presented in the form: Vt dA1 d----~ - Q12 (a12 A1 + [312A2) + Q14 (a14 A1 + ~t4 A4)
+ El2 (A2- A1) + E14 (A4- A1) + rat + WA1
Fe -Io -Io f(I) = ~ l exp('~'s exp-KH1)-exp--~-
gffi,~ 0 KN Kp F K H, Io I, T
= = = = = = = = = =
Maximum phytoplankton growth rate (/d) Temperatute coefficient (nondimensional) Monod kinetic constant for N (mg/L) Monod kinetic constant for P (mg/L) Photoperiod (nondimensional) Light extinction coefficient (/m) Mean depth of segment 1 Surface light intensity (J/cruZ.d) Critical sunlight intensity (J/cm2.d) Homogenous segment temperature (°C).
The convective (Qsj) and dispersive (Esj) terms connecting the segments were determined from hydrodynamic models (Gannoulis 1991). The external variables, which include temperature, solar intensity, photoperiod, and boundary concentrations were determined from analysis of field measurements (Gannoulis 1991). A detailed description of the structure of the model is presented elsewhere (Andreadakis 1991). The validation of the model was carried out using available experimental data for the period of 1984-91 (Gannoulis 1991). The results indicate a satisfactory proximity of the measured and calculated values. Measured concentrations of chlorophyll-a in three stations, one at the borders of the bay and the central gulf (cells 1-6), a second one in the central gulf (cells 7-17), and a third one in the outer gulf (cells 18-33) ranged between 0.5-7 gg/L, 0.4-3 ~tg/L, and 0.4-2 gg/L, respectively. These values are close to the model predictions of 0.23-8.04 gg/L, 0.22-3.14 mg/L, and 0.22-2.77 gg/L. Satisfactory agreement was also obtained for the other parameters (nitrogen, phosphorus, transparency, and dissolved oxygen). Measured and predicted N/P ratios were low, indicating that nitrogen is the limiting nutrient. Following the validation procedure, the model was used for the prediction of the impact of various treatment-disposal alternatives. The results led to the following conclusions:
A.D. Andreadakii et al.
(1) In all cases, the quality parameters of the outer gulf remain practically unchanged and insensitive to the treatment applied. (2) Biological treatment without nutrient removal and discharge through an outfall, located near the existing pumping stations, results in substantial decrease of the organic carbon concentrations along the west coastline and in the bay (reductions up to 70% when compared to existing concentrations). These reduced carbon concentrations lead to increases of the dissolved oxygen concentrations. With respect to nutrients and phytoplankton, the concentrations remain high, not much different from the observed current concentrations, indicating that the improvement of the disposal method (outfall in the central gulf instead of surface discharge in the bay) offsets the higher future nutrient loads. (3) Biological treatment with seasonal nitrogen removal (complete nitrification-denitrification for approximately six months of the year) has only a marginal additional beneficial effect with respect to
carbon and dissolved oxygen, but results in considerable reduction of the summer chlorophyll-a concentrations (in the order of 30% for the bay and 28% for the central gulf). (4) Addition of phosphorus removal to the previous treatment scheme gives similar results to the system without P removal, indicating that nitrogen is the limiting factor for eutrophication. (5) Biological treatment with nitrogen removal throughout the year reduces summer chlorophyl-a concentrations even more with overall reductions in the order of 36% and 34% for the bay and central gulf, respectively. (6) An advanced system consisting of biological treatment, nitrogen and phosphorus removal, and filtration, with the effluent discharged to Axios and through the river to the gulf, results in the highest reduction of chlorophyl-a concentrations (in the order of 46% for the bay and 44% for the central gulf).
Table 2. Comparison of the main alternative treatment and disposal systems.
the year. Discharge
to the central gulf
to the Axios river
Required sludge age (days)
Effluent BOD5 (mg/L)
Effluent SS (mg/L)
Nitrogen removal (%)
15 - 85
Phosphorus removal (%)
I n i t i a l cost(mil.Drachmas) Annual operational costs (mil.Drachmas)
Treatment and disposal of the wastewater of Thessaloniki
-t "~ 0.7. o
t? E o2E u) 0.lr,-
Cent ral Gulf
m Outer Gulf
.(~ 6 ¸
a sL e-
QJ U cO
~:~:e:~.:Q 3 IIIIIIIIIIIIIIIII llllllllllllllll Ill
Present Biological Biological Discharge Biological Discharge
situation. Treatment. Discharge through outfall. Treatment and seasonal N removal. through outfail. Treatment and N,P removal. through the Axios river.
Fig. 2. Effect of the treatment scheme on the water quality of the gulf.
The results of the comparative evaluation of the alternative schemes are summarized in Table 2 and Fig. 2. On the basis of this evaluation, taking also into account technico-economic comparisons, it was decided to adopt the alternative solution which consists of biological treatment with seasonal nitrogen removal and effluent disposal to the central gulf. The
biological system selected was the activated sludge process with aerobic sludge retention time ranging between seven days (summer months) and ten to twelve days (winter months), assuring practically complete nitrification throughout the year. Nitrogen removal is achieved by prcdenitrification with anoxic reactors ahead of the aeration compartments. The total volume of the anoxic reactors varies between
60 000 m 3 (summer months) and 20 000-40 000 m 3 (winter months), so that the nitrogen removed can vary between 85% or more during summer months and 40-45% on average during winter months.
Description of the proposed treatment plant The main units of the proposed treatment plant are the following:
Preliminary treatment: The three existing screw pumps, with some minor motor and transmission modifications, are capable of lifting 7m3/sec at a height of 9 m. Preliminary treatment of the wastewater includes screening and grit and grease removal. The screening installation consists of four (three existing and one new) mechanically raked bar screens. Each screen has a total width of 2 m and the clearance between bars if 25 ram. Removal of grit and grease if effected in two (one existing and one new) aerated units, each consisting of two longitudinal chambers, 30 m long. The total volume of the four grit removal chambers is 1350 m 3, resulting in a residence time of 6 and 3 min for the mean DWF and peak flow, respectively. Grease and oil, contained in the wastewater, is removed by four 2-m wide quiescent zones, installed longitudinally along the grit removal chambers and subsequently pumped to the digesters. Flow measurement occurs in three (two existing and one new) Venturi channels. Wastewater flow in excess of 4.5 m3/sec, which is the peak design flow of the biological stage, is diverted into one of the three Venturi channels and subsequently guided into two stormwater retention tanks. Two electromagnetic flowmeters are installed at the two conduits, which transport the treated wastewater from the outlet pumping station to the outfall. Primary treatment : Two new 47-m diameter primary sedimentation tanks, 5080 m 3 each, will be added to the three existing ones. During normal operating conditions (Q<4.5 m3/see), the flow is divided through a distribution chamber to the three primary sedimentation tanks which operate at a peak hydraulic loading of 75 m/d and a minimum retention time of one hour. The expected removal efficiency of the primary sedimentation is 50% for suspended solids and 30% for BeDs. The two other tanks serve as storm tanks for retention-sedimentation of the excess (over 4.5 m3/sec) flow, which may occasionally reach the treatment plant and be lifted by the inlet pumps. The settled wastewater is returned to the inlet of the treatment
A.D. Andreadakis et al.
plant by gravity during low flow periods and the tanks remain empty until the next storm event. Part of the wastewater flow (up to one-third of the total) can bypass the primary sedimentation tank and reach an intermediate pumping station, which is used for feeding the biological reactors. This arrangement enables the optional introduction of raw sewage into the activated sludge units, thus assisting the control of the denitrification process as well as sludge bulking.
Biological treatment: The design of the biological treatment stage was based on requirements of complete nitrification for temperatures higher than 12°C and 85% nitrogen removal (through denitrification) during six months of the year, with temperatures over 18°C.The total volume of the biological reactors is 140 000 m 3, divided into eight parallel 17 500 m 3 units. Each unit is separated into three zones, the denitrification zone (DN) which is permanently anoxic, the nitrification-donitrification zone (NI-DN) which can operate under either aerobic or anoxic conditions, and finally the nitrification zone (NI) which remains permanently aerobic. The volume of the DN zone of each biological reactor is 2 500 m 3, divided into two equal 1250 m 3 compartments, the NI-DN zone consists of two compartments, 2 500 m 3 each, and the NI zone has a volume of 10 000 m 3. This arrangement allows the volume of the aerobic (aerated) zone in each o f the 8 r e a c t o r s to vary b e t w e e n 10 000-15 000 m 3 and the volume of the anoxic zone between 7 500-2 500 m 3, respectively. Though not required, it was decided to retain an anoxic compartment (DN-zone) even during the colder (T
Treatment and disposal of the wastewater of Thessaloniki
termediate 300-m 3 upflow nonaerated compartment in order to eliminate the dissolved oxygen of the mixed liquor.
Final clariflers: Eight 54-m diameter and 3.5-m sidewater depth circular clarifiers will be constructed, with a total area of 18 300 m 2 and a volume of 76 000 m3. The clarifiers will be equipped with diametrical dual speed scrapers for efficient sludge removal. Recirculation of settled sludge up to 180% of the design flow is possible through five screw pumps.
Sludge treatment: Sludge treatment includes thickening, anaerobic digestion, and dewatering by beltfilterpresses. Thickening of primary sludge is performed in the four existing 15.5-m diameter gravity thickeners. The excess biological sludge (approx. 33 Mg/d) will be thickened up to a solids content of 7% through centrifuges or belt thickeners. The total capacity of the anaerobic digesters will be increased form 15 400 - 23 100 m3 by adding one new digester to the two existing ones. The biogas produced (25 000 m3/d) will be used for heating the digesters and running some of the compressors used for aeration. Dewatering of the post-thickened sludge will be carried out by five filterpresses (three existing and two new) up to a solids content of 30%.
Disinfection: The existing chlorination installation will be used for disinfection of the treated wastewater. The application of UV and ozone for disinfection were also considered; but due to their high cost and the ongoing intensive research in this area, the adoption of an alternative disinfection method was postponed. Sand-filtration and UV-disinfection may be a suitable future solution, especially if the treated wastewater is to be used for irrigation purposes.
The existing treatment plant for the wastewater of Thessaloniki is designed for primary treatment, producing an effluent which cannot be discharged into the originally selected recipient, the Axles River. Safe discharge to the river requires further treatment consisting of biological treatment with nitrogen and phosphorus removal, thus increasing substantially both the initial and operational costs. Disposal of the effluent to the central Gulf of Thessaloniki through a marine outfall is an alternative which, on the basis of the results obtained from a mathematical quality model of the gulf, can prevent adverse environmental impacts at a reduced operational cost. The results of the mathematical model indicate that nitrogen is the limiting nutrient and that a system including biological carbon and partial nitrogen removal is sufficient for a satisfactory control of eutrophication. The proposed treatment system was designed on the activated sludge principle with nitrogen removal through nitrification-denitrification, taking also into consideration the existing facilities. REFERENCES Andreadakis, A.D. Mathematical modelling of the Gulf of Thessaloniki. Report to the Ministry of Public Works of Greece. Athens, Greece; 1991. Asano, 1".; Leong, L.Y.C.; Rigby, M.G.; Sakaji, R.H. Evaluation of the California wastewater reclamation criteria using enteric virus monitoring data. Water Sci. Technol. 26: 1513-1524; 1992. EEC (Commission of the European Communities Council). Directive 91/271 concerning urban wastewater treatment. EEC, Brussels; 1991. Gannoulis, I. Oceanographic data and environmentalimpact from the sewerage system of Thessaloniki. Report to the Ministry of Public Works of Greece, Thessaloniki, Greece; 1991. State of California. Wastewater reclamation criteria. Department of Health Services, Berkeley, CA; 1978. WHO (World Health Organization). Health guidelines for the use of wastewater in agriculture and aquaculture. Technical Report Series 778. World Health Organization, Geneva; 1989.