OPTIMIZATION DESIGN OF GRANULAR ACTIVAED CARBON USING GENETIC ALGORITHM

Integrated advanced wastewater treatment provides important fundamental solutions to problems associated with water scarcity prevailing in arid and semi arid climatic regions. This was accomplished through treated water with specific specification and characteristics suitable to be used for agricultural, domestic, and industrial purposes. This study is concerned with the use of genetic algorithm procedure for the optimum design of integrated advance wastewater treatment units, with their various types and characteristics. The aim of optimum wastewater treatment units design is to attain optimum values of certain pre defined objective function


1.
In most Mediterranean countries, the scarcity of water together with the high cost associated with collecting and using the limited surface rainwater for irrigation have become real constraints for our irrigated agriculture. Because of this, particular emphasis is placed on the water use efficiency and the cultivation of crops with high return per square meter and volume of water (Morrison et al., 2009).
Drought attributable in significant part to climate change is already causing acute water shortages in large parts of Australia, Asia, Africa, and United States (Alderfasi, 2009).
Reuse of effluents from municipal or sanitary wastewater treatment plant is becoming a matter of great importance as available and economical resources for a variety of applications (Lin, 2007, Metcalf& Eddy, 2003). Water reuse figures are basically classified into; Unavoidable water reuse and intentional water reuse. The location of several cities on a single river, stream, or lake leads to unavoidable water reuse when each city uses the water body as a water supply and receiving body for wastewater (Culp et al., 1979). Intentional water reuse holds the key to the efficient and effective utilization of the limited freshwater resources by making available a new valuable source of water to augment existing supplies and a major source of supply for the future. It includes the reuse of water for industrial, agricultural, domestic, recreational, and ground water recharge purposes (Culp et al., 1979).

Activated Carbon Adsorption
Adsorption is the process of collecting soluble substances within a solution on a suitable interface. In AWWT, adsorption with activated carbon-a solid interface, is used principally for the removal of refractory organic compounds, as well as residual amounts of inorganic compounds such as nitrogen, sulfides, and heavy metals (Benefield, 1989). The removal of taste and odor compounds from wastewater is another important application, especially in reuse applications. Activated carbon is produced by heating char to a high temperature and then activating it by exposure to an oxidizing gas at high temperature. The two most common types of activated carbon are granular activated carbon (GAC), which has a diameter greater than 0.1 mm, and powdered activated carbon (PAC), which has a diameter of less than 200 mesh (Álvarez , et al., 2011). In this study GAC has been selected over powdered carbon primarily because GAC is normally regenerated and reused, while powdered carbon is discarded after use. Also, GAC has the additional advantage of providing a margin of safety in operation that powdered carbon does not provide. Sudden changes in influent composition are common in wastewater treatment. Granular carbon has the capacity to withstand substantial changes in the influent composition with a much reduced effect on the effluent quality. This aspect and the availability of the regeneration technology were the major factors in the selection of granular carbon.
A bed column is often used to bring the wastewater into contact with the activated carbon. There are three types of activated carbon columns; fixed-bed, expanded-bed, and movingbed. In fixed-bed columns, the water is applied to the top of the column and withdrawn from the bottom. In expanded bed columns, the influent is introduced at the bottom of the column and is allowed to expand. While, in the moving bed columns, spent carbon is continuously replaced with fresh carbon. Spent granular carbon can be regenerated by removal of the adsorbed organic matter from its surface through oxidation in a furnace. The capacity of the regenerated carbon is slightly less than that of the virgin carbon (Álvarez et al., 2011).
The design of GAC contactor is based on contact time, hydraulic loading rate, carbon depth and number of contractors. At least two parallel carbon contactors should be used, because then one or more units can remain in operation while one unit is taken out of service for maintenance or for removal and regeneration of spent carbon

ADVANCED WASTEWATER TREATMENT
Since the objective of this study is to develop a genetic algorithm for the optimum design of advanced wastewater treatment (AWWT) plant, it is necessary to define;  Water reuses criteria which are the aim of AWWT.  Required quality of reclaimed water.  AWWT processes; types and design criteria

Water Reuse Classifications
Reclamation and reuse of effluents from municipal wastewater treatment plant is becoming a matter of great importance in the available and economical resources for a variety of applications (Zhang, 2004). These applications are illustrated in Figure 1 below:-

Development of Genetic Algorithm for Granular Activated Carbon Treatment Plant Design
The basic steps of genetic algorithms development are namely (Chambers,  5. Selection of the (chromosomes) strings for reproduction, 6. Crossover of the selected strings, and 7. Mutation of the strings. The adoption of these steps is dependent on type of genetic algorithm (GA) application. This is particularly true for advanced treatment technologies capable of treating wastewater to a degree of quality appropriate for a specific reuse, making the selection of the most suitable sequence of processes for any potential reuse situation more complex.

Formulation of Cost Function
The annual cost of water treatment includes the annualized capital cost, annual operation and maintenance cost, and land requirement cost."Capital costs" refers to the investment required to construct and begin the operation of the plant, principally materials, labor, and interest. Operation and maintenance costs include the costs associated with the labor, material, and energy required to operate and maintain the treatment plant (Sharma, 2010). The annual cost function for treatment unit-i can be written as:  Where CC i is the capital cost of treatment unit-i, CRF is capital recovery factor, m is the interest rate per year, and n is the number of years over which the cost will be spread. In this study, all the capital costs shall be spread over a period of 20 years at a 8 percent annual rate of interest. In this study, multiple hearth furnace was used for carbon regeneration. Hence, the capital cost of activated carbon treatment unit includes both the costs of granular activated carbon column and the multiple furnaces (CC GAC and CC MF ). Also, the annual operation and maintenance cost of activated carbon unit is the sum of annual operation and maintenance costs of granular activated carbon column and multiple furnace (OMC GAC

Results and Discussion
The design of AWWT units is illustrated in equations given above. Optimization using genetic algorithm (GA) was applied for the design of granular activated carbon. Two design aspects were considered. The first aspect includes the optimal design criteria for the desired quality of treated water while the second aspect includes the cost of the plant. The cost of plant is considered as the objective function which has to be minimized.
GA was implemented using MATLAB program Version 7.10.0.499 (R2010a) with the adoption of roulette wheel selection and single-point crossover (Goldberg, 1989). The crossover probability for single-point crossover was chosen to be 0.6. The population size was selected to be 60. This was done after performing several trains with population sizes of 20, 40, 60, 80 and 100 and it was found that when population size exceeds 60, the effect on results is insignificant.

Optimum Design of AWWT
AWWT incorporates granular activated carbon (GAC) process. Here, four design criteria govern the design of this treatments unit; EBCT, C dose , D ac , and L mf . According to GA terminology, the search is for a chromosome, consists of the four elements (genes), that minimizes the objective function along with satisfying all design criteria and effluent quality. During the application of GA in designing AWWT, the effect of varying the concentration of influent and effluent chemical oxygen demand (COD in and COD ef ) on optimum design criteria was studied. This was done by taking the minimum, averaged, and maximum values of COD in which are 65, 137, and 225 mg/l, respectively, and by considering COD ef of 5,8,11,14,17, and 20mg/l. The chosen COD in values where those of secondary effluent of Hamdan Sewage Treatment Plant in Basrah city. The adopted values of COD ef were selected to be within the accepted range of COD in water to be reused for agricultural , domestic, and industrial purposes (Alderfasi, 2009).
The effect of COD in and COD ef on activated carbon dose (C dose ) and loading rate of multiple hearth furnace (L mf ) are shown in Figures 2 through (7) for maximum, average, and minimum influent flowrates, respectively. The GA was applied for AWWT adopting the maximum, average and minimum values of the influent flow rates of Hamdan Sewage Treatment Plant in Basrah city (1.62, 0.81 and 0.54 m 3 /sec). From these figures, the optimum values of C dose and L mf were obtained and they are listed in Table 1. The optimum values of empty contact time (EMBT) and Diameter of contractor (D ac ) did not change with influent flow rate and they are 10 min and 6.1m, respectively.    respectively. From these figures it is shown that the treatment cost decreases with the decrease of influent flow rate which is a reasonable and logical result.

Conclusions
The development and application of genetic algorithm (GA) for the design of granular activated carbon of advanced wastewater treatment (AWWT) plant, reveal the following conclusions: 1-Genetic Algorithm was found to be powerful technique for operating and defining optimum values of the parameters used in design of the advanced wastewater treatment system. 2-A penalty function can be used to find the global optimum design of constrained problems such as of advanced wastewater treatment design and in conjugation with the genetic algorithm developed in this study.

3-
The optimum values of design criteria vary randomly with the increase of influent flow rate.  The optimum value of empty bed contact time was 10 min.  Dose of activated carbon was 25 mg/l  Diameter of contractor was 2.4m, and  Loading rate of multiple hearth furnaces was 30 m/min.

4-
The treatment cost decreases with the decrease of influent flow rate which is a reasonable and logical result. This outcome highlights the importance of applying GA for deciding the design criteria which minimize the treatment cost.