User Behavior Assessment of Household Electric Usage

. Energy resilience is one of the famous issues among researchers and practitioners in energy sector. With enabling new technologies in power engineering for smart grid such as distributed generation, distributed storage, and intelligent information and management, each household community can establish a resilience energy production, distribution, and consumption. A household in smart grid system behaves as a customer and producer at the same time. This condition enabled them to reduce the power shortage in the peak hours, reduce CO2 pollution using renewable electricity, and minimizing electricity USAge by changing life style. In developing countries, the amount of electricity supply is less than its demand. Most of the demand comes from the household that has peak load on nighttime. Keywords: User behavior, Game theory, Smart grid, Heating and cooling appliances, Energy resilientdoi:10.12695/ajtm.2013.6.2.1 How to cite this article:Mulyono, N. B. (2013). User Behavior Assessment of Household Electric Usage. The Asian Journal of Technology Management 6 (2): 65-71. Print ISSN: 1978-6956; Online ISSN: 2089-791X. doi:10.12695/ajtm.2013.6.2.1


logies.
to an es like ission, 13). rom a users. marter t grid) flows mation twork. lity to ficient range in the The introduction of electric generation based on renewable resources like sun, water, and wind, is one of the major parts of smart grid. However, the amount of electricity generated from renewable sources (called micro grid) is fluctuating and uncontrollable (Molderink et al, 2010). In this situation, additional production of electricity power is necessary to reduce demand fluctuation, supply uncertainty, and avoid power shortage for some users.
Each of the users in smart grid can become a producer of electricity by installing renewable electricity source. We can call this user as the active user. They have an ability to manage their appliance load, generate a small amount of electricity from renewable sources, and keep their own energy in term of electricity and heat. Active users in decentralized power system have two obvious advantages over passive user in centralized ones, such as improve energy usage and contribute to emission reduction (Gu et al, 2014). Active users can both buy and sell electricity to the other active users or to the electric company.
Besides increasing the production capacity of the renewable energy in house, the household tenants have to be able to manage their electricity usage trough changing their behavior. Understanding the behavior of the user in using electricity is important in this situation. By adjusting their behavior, they can maximize utilization of renewable energy sources, minimize usage of electricity, and help the other user in needs. Similar with lateral transshipment of inventory system in disaster recovery (Mulyono and Ishida, 2014), each active user can mutually help the other active users having electric shortage.
We proposed a methodology to understand the behavior of the active users in using electricity. We utilize game theory to model the behavior and interaction between active users on their way to maximize their benefit over electricity consumption and production. The active users are players in a game that defined by a common goal with different constraints and conflicting objectives. We use game against nature to predict the behavior of electricity consumption of active users. We are focusing on the cooling and heating appliances since those appliances are major contributor of electricity consumption.
The remaining of this paper is structured as follows. The following section provides an overview of several related work on smart grid. Section three introduces a model of user interaction to minimize the electricity consumption and support the other active users.
Section four describes an implementation of the model trough simulation process. We conclude this paper with a discussion of the result.

Related Work
Development of smart power grid, that augments traditional power grid system, is one of the greatest inventions in the last decade. In contrast with traditional power grid system, that carries electricity power from a few central generators to a large number of users, smart grid uses two-way flows of electricity and information to create and distribute electricity. Smart grid includes the entire spectrum of the energy system from the points of generation to the points of consumption. Smart grid is decentralized electricity power system that uses two ways information and computational intelligence in an integrated fashion toward electricity generation, transmission, substations, distribution, and consumption to achieve electricity system that is safe, clean, secure, reliable, efficient, resilient, and sustainable (Gharavi and Ghafurian, 2011).
The International Energy Agency concludes that, although decentralized electricity system has a higher cost than centralized ones, it has potential to supply all demand with the same reliability but with lower capacity margin (International Energy Agency, 2002). We can simply define smart grid as a decentralized power grid that involves four operations like power generation, transmission, distribution, and control. Table 1 shows the difference between conventional power grid and smart grid (Farhangi, 2010). From the technical perspective, smart grid consists of three major systems such as smart infrastructure system, smart management system, and smart protection system (Fang et al, 2013).
Smart infrastructure system is the energy, information, and communication infrastructure underlying the smart grid. This infrastructure supports two-way flow of electricity and information.
Smart management system provides advanced management and control services and functionalities. Smart protection system provides advanced grid reliability analysis, failure protection, security, and privacy system. Table 1. Comparison between conventional grid and smart grid (Farhangi, 2010) In the beginning of smart grid development, many research focused on distributed generation management, energy storage and demand side management. Stability of the grid is studied intensively by (Azmy and Erlich, 2005) having a conclusion that electric generators are key to grid stability. In energy storage field, one of the hot topics is level out the electricity demand fluctuation by combining electricity storage with renewable resource such as windmill (Costa et al, 2008). Most literature in demand side load management used the agent-based solution having a hierarchical structure ensures the scalability of the solution (Molderink et al, 2010). Interaction of the electricity user also modeled using agent-based solution. Each of the smart grid users is an active user having the ability to consume, produce, and share electricity between them.
Game theory is the best tool to model the interaction of a user to maximize their payoff. Game theory related to the actions of decision makers who are conscious about their actions and its effect. A game consists of a principal and a finite set of players, each of which selects a strategy with the objective of maximizing his utility (Charilas and Panagopoulos, 2010). The utility function represents each player's sensitivity to others actions.

Behavior of Electricity Consumption
In every household, there are many electric appliances commonly used such as refrigerator, heater, air conditioner, microwave oven, television, water pump, and dehumidifier. Each of them consumes different range of electricity like refrigerator (725 watt), heater (750-1500 watt), air conditioner (900-1500 watt). Among all of them, the appliances for cooling and heating take the highest electric consumption level. It takes about 8-34% of the overall energy consumption and about 12-38% of the overall energy consumption in summer and winter, respectively (Takuma et al, 2006). Furthermore, the consumption level of those appliances is sensitive to the environmental temperature and lifestyle of the users. For that reason, we propose a game theory model against nature to understand the behavior of the users in using the cooling and heating appliances. Thermal comfort of the human depends on air temperature, radiant temperature, air velocity, and humidity (Health and Safety Executive, 2013). Comfortable air temperature is from 21 o C to 23 o C during winter, and from 23.5 o C to 25.5 o C during summer. In this research, we focus on one adjustable variable, which is air temperature.
The user of the household has three strategies such as not use the cooling/heating appliances, use the cooling/heating appliances economically and use the cooling/heating comfortably. If the user uses the cooling/heating comfortably, amount of electricity will increase at the maximum level. On the other hand, if the user uses the cooling/heating economically, he can reduce the cost while maintaining thermal comfort at a minimum level. Figure 1 illustrates tradeoff between comfort level and expenses. The more expenses spend to turn Those three tables imply that expected payoff is sensitive toward suffer cost. If average suffer cost is above the electricity cost, the user prefers to use cooling and heating appliances in a comfortable setting. On the other hand, the user does not prefer to use cooling and heating appliances if the average suffer cost is below the electricity cost. Figure 2 and 3 illustrate the trend of the suffer cost to the payoff and trend of the suffer cost to selected strategy, respectively. This figure clearly show that there is a trade off between suffer cost and payoff and the active user tend to use cooling/heating appliance comfortably as the suffer cost increases.

Conclusion
This paper successfully built deterministic model in power engineering based on game theory. The result shows unique behavior of the active users toward electricity usage especially in using heating and cooling appliances. The user strategy is highly affected by the suffer cost. This suffer cost is related with the physical condition of the user, their financial condition, and their life style.   This model is practically applicable to be used in smart grid with dual bus system. Future development of this research can be directed to the development of probabilistic model of mutual support system in electricity generation and distribution.