Selçuk Soyupak

Case studies on the use of neural networks in eutrophication modeling

Abstract Artificial neural networks are becoming more and more common to be used in development of prediction models for complex systems as the theory behind them develops and the processing power of computers increase. A three layer Levenberg–Marquardt feedforward learning algorithm was used to model the eutrophication process in three water bodies of Turkey (Keban Dam Reservoir, Mogan and Eymir Lakes). Despite the very complex and peculiar nature of Keban Dam, a relatively good correlation (correlation coefficient between 0.60 and 0.75) was observed between the measured and calculated values. For Mogan and Eymir, which are much smaller and more homogenous lakes compared to Keban Dam Reservoir, correlation values as high as 0.95 were achieved between the measured and calculated values. Neural network models were able to model non-linear behavior in eutrophication process reasonably well and could successfully estimate some extreme values from validation and test data sets which were not used in training the neural network.

A three-dimensional water quality-macrophyte interaction model for shallow lakes

A dynamic three-dimensional water quality model for macrophyte-dominated shallow lakes was proposed and tested. The proposed model is capable of simulating macrophytes and its interactions with water quality constituents such as dissolved oxygen (DO), organic nitrogen, ammonia, nitrate, organic phosphorus, orthophosphate, biochemical oxygen demand, phytoplankton and the sediment layer in shallow lakes. An existing two-dimensional hydrodynamic model has been utilized in conjunction with the water quality model to simulate water levels, velocities and flow rates. The modelled macrophyte processes are photosynthesis, respiration, mortality and excretion. Hourly simulation of photosynthesis process has been realized. The hourly simulations need special attention to predict diurnal variations of DO in macrophyte dominated lakes. The proposed water quality simulation model was subjected to calibration, verification and prediction processes using the data collected from Mogan Lake, Turkey. Mogan Lake exhibits wide variations of macrophyte biomass seasonally. The lake also exhibits highly variable DO levels both seasonally and diurnally. Statistical error quantification methods have been utilized to test the goodness of fit between the water quality model predictions and field measurements. Good agreement has been achieved between model predictions and measurements. Moreover, uncertainty analysis has been carried out for macrophyte and DO constituents. The analysis showed that the magnitude of the saturated growth rate of macrophyte is the most sensitive model parameter both for macrophyte and DO. The proposed water quality simulation model gave some promising initial results as a management tool to predict the expected reductions for the undesired consequences of eutrophication problem.

Evaluation of eutrophication control strategies for the Keban Dam reservoir

Abstract The reservoir of the Keban Dam in eastern Anatolia is a very deep and large water body subject to significant seasonal water level fluctuations within a given hydrological year. At the present time, different sections of the reservoir have different trophic status and the upper portion of the reservoir, in particular, is eutrophic. In this regard, there is a need for the assessment of alternative strategies to be implemented for phosphorus and thus eutrophication control in this segment of the reservoir. In this study, mathematical modeling techniques are used to compare various alternatives in terms of their relative effectiveness for phosphorus control. With this intention, a two dimensional depth-averaged hydrodynamic model and a three-dimensional compartmental water quality model are coupled to represent the complex water body dynamics prevailing in the reservoir. After calibration and verification, the coupled model was used to simulate dissolved oxygen and euphotic chlorophyll-a concentrations under different scenario conditions corresponding to various strategies proposed for phosphorus control. The modelling effort has yielded valuable information that can be used by decision makers for the evaluation of different management alternatives including reduction of the phosphorus load coming from point as well as diffuse sources within the catchment area, use of baffles and construction of pre-dam structures.

Management strategies for the lower Seyhan catchment

Abstract The Seyhan river has a total catchment area of 20, 731 km2 with major pollution sources located downstream of the Seyhan Dam in Adana. The so-called Lower Seyhan, is under the threat of not only domestic and industrial wastewater discharges originating from the metropolitan area of Adana, but also agricultural wastes of the fertile Cukurova Plain. In this study, the present and future water quality characteristics of the Lower Seyhan have been investigated by means of a mathematical modelling study based on actual field data as well as in-situ water quality measurements. The model of choice was the QUAL2EU. The behaviour of the Lower Seyhan was studied under different pollution control scenarios in order to develop plausible water quality management strategies for the target year of 2010. As a result of these studies, two alternative systems were proposed for the management of the wastewater sources within the Lower Seyhan catchment. Alternative wastewater treatment plant designs and their cost implications were evaluated by means of the CAPDET-PC software package. A user-charging scheme and an optimisation model that can be used for demonstrating the economic viability of the proposed systems were also developed for obtaining a cost-optimum management policy for both municipalities and industrial establishments of the area.

Biogas Production from Agricultural Wastes: Semicontinuous Anaerobic Digestion of Sunflower Heads

Abstract Data for semicontinuous anaerobic digestion of sunflower heads were obtained in fermentors of 1 l of effective volume, maintained at 55°C. Experiments were carried out with a 2%percnt; sunflower head-water mixture for hydraulic retention times of 15, 10, and 8 days. Biogas quantities and compositions were measured each day until steady state was achieved. The steady state was achieved after three times of each hydraulic retention time (HRT). Gas productions of 0.180 m3 /m3 dig. day with 62%percnt; CH4, 0.280 m3/m3 dig. day with 58%percnt; CH4, and 0.375 m3/m3 dig. day with 48%percnt; CH4 were obtained, for 15, 10, and 8 days of HRT, respectively.

A comparative evaluation of different approaches for assessing soil acidification impacts

Abstract Long‐term impacts of acidic depositions on soils can be predicted with coupled applications of acid deposition models and approaches to assess soil acidification. Soil acidification assessment can be achieved by using the following approaches: Mechanistic modelling Experimental acid buffering capacity (ABC) method Qualitative evaluations The mechanistic modelling approach adopted in this study utilises predicted acid deposition rates and soil properties as well as dominant soil mechanisms including utilisation, immobilization, nitrification, dissociation, association, carbonate weathering, silicate weathering, aluminium hydroxide disintegration, and cation exchange. The experimental ABC method, on the other hand, was based on titration curves obtained by adding different dilutions of H2SO4 to known quantities of soil samples. The sensitivities of the soil samples to acidification were also evaluated using a qualitative approach based on the pH and cation exchange capacity (CEC). A comparative eva...

Chlorine modeling in water distribution networks using ARX and ARMAX model structures

AbstractChlorination is still the main method of disinfection worldwide. Utilization of chlorine modeling enhances the management of water supply systems toward the reduction of the risks of waterborne diseases and cancer from disinfection byproducts formation. In this paper, the results of chlorine modeling efforts were presented as an extension of a project for Konyaalti Water Distribution Network (KWDN) of Antalya City, in the south of Turkey, using autoregressive with exogenous input (ARX) and autoregressive moving average with exogenous input (ARMAX) model structures. The required data-sets were obtained from the existing online monitoring stations. The ARX and ARMAX model structures modified for time series applications were utilized to predict chlorine concentrations at the critical point of KWDN. Non-representative data-sets were initially identified and excluded from the database. Best fit and Akaike’s Final Prediction Error techniques were used as model selection criteria. ARX4-5-3 and ARMAX2-3-...

Effect of Pretreatments on the Semicontinuous Anaerobic Digestion of Sunflower Heads

Abstract The effects of hydraulic retention time and alkali treatment on methane production rate from the semicontinuous anaerobic digestion of 2% sunflower-head/ water mixtures were investigated. The experiments were carried out in laboratory-scale fermenters, fed with I liter of untreated, 2 g NaOH/l00 g total solids (TS), and 5 g NaOH/100 g TS alkali-treated sunflower-head/water mixtures, respectively, and maintained at 55 °C. Digestion experiments were performed for hydraulic retention times of 8, 10, and 15 days. The amount and composition of produced gas were measured until steady state was attained in each run. The steady-state methane production rates were found to decrease with hydraulic retention time and increase with alkali dosage used for pretreatment.

Pollutant transport study for Tuz Gölü

The results of a predictive pollutant transport study for Tuz Golu (Salt Lake), Turkey, is described. The present and future 5 Day Biochemical Oxygen Demand (BOD5) and Dissolved Oxygen (D.O.) profiles of the drainage canal of Konya Plain, which carries pollutants to Tuz Golu, are presented. The one‐dimensional pollutant transport model, which includes advection, dispersion, dilution, biochemical reaction and aeration mechanisms, was utilized to determine the profiles. Present and future BOD5 loads, under different water pollution control alternatives for Tuz Golu; and a comparison of the estimated values with the predictions under present conditions are also presented.

Assessment of soil acidification due to a natural gas‐fired power plant by using two different approaches

Potential soil acidification impacts of a proposed natural gas‐fired combined cycle power plant were assessed using an integrated approach coupling an atmospheric deposition model with soil acidification quantification. The deposition model was used to estimate the rates of nitrogen oxide (NOx) deposition on the air‐soil boundary. The expected changes in the soil column were then predicted by utilizing mechanistic and experimental methods, and the number of years required to reach critical pH values were predicted using the two methods mentioned above under different rates of acidic deposition. The number of years predicted by the mechanistic modeling approach was lower for all soils exhibiting calcareous character.