A model in Optimization in Power-Shared Community

...

one of them declares the energy needed for family i, one of them is for the amount of energy produced at family i and one of them represents the capacity of energy storage facilty.

The variables and their descriptions are shown below.

According to these parameters and variables, the mathematical representaion of the model is constructed. Firstly, the objective function is determined. Below figure, the objective function is shown.

This objective function minimizes the overall benefit which is composed of the cost of buying energy from state grid, selling energy to state grid and the cost of using energy storage facilities respectively. Since selling energy to state grid has a negative effect on the cost side and all other objective function components represent the cost, there is a negative sign on the second segment (selling energy to state grid). According to this formulation, a home which produces more than its demand may prefer to pay more on the utilization of energy storage facilities or to sell more energy to the state grid, which actually results in less cost of buying energy with high prices form the state grid.

In Figure 5 below, the model constraints are shown.

The first constraint is energy conservation constraint and it makes all inbound household energy to outbound energy. Second equation is for the energy storage facilities. It allows for both the inbound and outbound energy are sourced or destined from energy storage facilities as well as homes. Third constraint limits the amount of energy for each storage facilty by capacity at each period. In fourth constraint evolution of the energy for each storage facilty over time is presented. Fifth constraint associates the variables Z (accumulated energy family i has stored at time period t) with S (accumulated amount of energy stored at energy storage facility j at time period t). Sixth constraint shows how energy to be stored over the time period association with the seventh constraint (upper bound). Other constraints are non-negativity constraints.

In order to simulate this modeling part, we used GAMS and used random numbers for parameters. In our first trial, we determined the number of families as 25, number of energy storage facilities as 5 and time periods as 10. We got the following results:

Then, we try the same model for different set size of families while other set sizes and set values are same. In this second trial we chose 10 households. And we got the following result:

According to our chosen random parameter values, we concluded that when the family size increases while other parameters are same, households generate more energy and sell to the grid and as a result we get higher negative value of cost.

4) RESULTS

With the help of this paper, we would like to propose a new model based on energy-shared community. This model aims to determine the storage quota given to each family, and optimize the usage and storage of energy produced by families through solar panels in an energy-shared community. The overall aim is to minimize the cost in this energy-shared community.

From this point of view, after analyzing the model we concluded that this paper has its own weak and strong points. It has a weak point since in the modeling part the authors assume that the rate lij of energy loss depends on the distance dij between family i and facility j but in the modeling part, there is no relationship is shown between lij and dij. So, this is the weak point of this paper. If there is a relationship between those and the solver solves according to this criterion too, the model would be much more consistent than the current model.

Furthermore, in the paper, we cannot see any output results of the ‘minimized overall benefits’ which is indicated as an objective funtion. After problem description and modeling part, it jumps right onto IBM CPLEX Log results which is just showing the time needed to solve problem if we change either set of families or time periods. There is no optimized solution data which reflects that the minimized cost (overall benefit) in the end. Therefore, when they said that the problem solved efficiently, we cannot see any proven objects in the conclusion part. Actually, this is a very complicated problem to solve at first glance. At first step, we wrote the code but since our code violates the GAMS’ general solving criterias (line limit extension) without licence, then we tried to find another GAMS with licence to solve and obtain the result. Thus, after all these steps, we can indicate that according to our result, model is an optimized and efficient one consequently.

Also this paper has a strong point too. The strong point is that this modeling approach and solving parts can be easily converted into real life problems about power-shared communities too since the assumptions are suitable for real life problems and this Energy Optimization Model (EOM) can be solved optimally and efficiently.

REFERENCES

[1] Renewableenergyworld.com, (2014). Why is renewable energy important?. [online] Available at: http://www.renewableenergyworld.com/rea/tech/home [Accessed 20 Dec. 2014].

[2] Li, J., Liu, F., Ren, C., Wang, Q., He, M. and Chen, F. (2012). A model for energy optimization in power-shared community. Proceedings of 2012 IEEE International Conference on Service Operations and Logistics, and Informatics.

[3]Anon, (2014). [online] Available at: http://delivery.acm.org/10.1145/1130000/1129664/9254439.pdf [Accessed 20 Dec. 2014].

[4] Anon, (2014). [online] Available at: http://www.nrel.gov/docs/fy12osti/54570.pdf [Accessed 20 Dec. 2014].

APPENDIX

Appendix 1: