CO 2 emission reduction and energy management for an integrated smart grid — Case of study: Rwandan electrical network

. Many scholars have been focusing on the energy management by Integrating a smart grid into a conventional electrical grid. They have showed that to meet a certain power demand of the consumers, using energy management, the electric utility can turn on some generators, which may have the least operation cost, while the generators with high operation cost are left to supply extra load demand in specific peak periods. Henceforth, the operation cost of its generation units is minimized. The issue remains at a level of relating the energy management to CO 2 emission. The present paper briefly discusses the Rwandan electrical network that still integrates the use of diesel generators. It estimates the amount of CO 2 emission that can be avoided once a PV system is integrated into the electrical network. The paper as well proposes an algorithm for energy management with consideration of CO 2 emission.


Introduction
The main goal of applying energy management has been to minimize the economic cost and losses [1]. It has been shown that this cannot be achieved without integrating the renewable sources [2]. All of utility companies focuses more on energy management to regulate the power flow in a way to minimize the energy losses on the network and increase the penetration level of renewable energy sources (such as PV and wind farms) in an efficient way. Whereas the end-users use energy management to minimize their electricity bill and schedule their load demand in an efficient way. Few utilities and end users focus on the amount of CO2 emission amount that they can avoid and achieve optimal operation cost while integrating the renewable sources. Some electrical networks intend to integrate diesel generator sets for meeting a high demand that can occur during a pick period. This the case of the Rwandan electrical network.

CO2 emission delivered by a diesel generator set
The previous section described clearly the role that diesel generators are playing in supporting the electricity production in Rwanda. The generators help to meet the demand during peak periods. However, the diesel generator is not environmental friendly. It has high CO2 emissions. The CO2 emission can differ from one generator to another due to their efficiency. The diesel generators are characterized by their efficiency and rate of specific fuel consumption. The efficiency of the diesel generator is influenced by thermal and mechanical efficacy of the generator [3]. The quality of the diesel generator might increase or decrease the thermal efficiency. Most of the time, the mechanical efficiency of diesel engine is around 80% to 85% whereas the generator efficiency is around 95% to 98%. This efficiency is calculated with reference to the rated power of the diesel generator. It will be a mistake to operate the diesel generator at lower percentage of its rated speed because it will result to lower efficiency of the generator. This must be taken into consideration while connected the generator to the grid. The generator must operate at least at its 70 and 90 % of its rated power. Even though, the fuel consumption of diesel generators differ from one type of engine to the other depending on technologies used by manufacturers, an equation of fuel consumption rate for diesel generator is common. Figure 1 found in [4][5] was used to formulate Eq.1 that explains the fuel consumption of the diesel.
Where FC is the fuel consumption, in [l/h], PBR is the rated brake power, in [kW], and a, b and c are constants to be fitted for each particular engine. Brake specific fuel consumption ( BSFC) is defined as a measure of the fuel efficiency of a diesel-fired generator whereas the brake power is developed by the engine at the output shaft. Figure 1 shows that the efficiency of diesel increases while its output power is closer to its brake output power [7]. In Rwanda, some diesel-fired generators are connected to slack busses and vary their corresponding generation capacities following the variation of power demand aiming at keeping the frequency within an acceptable limit; consequently, they are not always working closer to their nominal power ratings. This increases the fuel consumption as well as CO2 emitted in atmosphere.

Fig.2. Example of a real load profile from a remote petrol exploitation field [2]
Some scholars do not consider the quadratic term and the equation is pure linear. However, the author in [3], added the quadratic term in order to improve the fuel consumption for variable speed diesel generators. Depending on a load profile, the fuel consumption can be calculated using Eq.1. Figure 2 gives an example of a load profile of some generations in Rwanda. This can be based on estimating the CO2 emissions of the diesel generator in Rwanda. However, knowing the amount of CO2 emissions per litter supports Figure 2 in calculating the CO2 emissions using Eq.1. The fuel consumption of the diesel generator ranges between 2.4-2.8 kg of CO2/liter diesel [2]. Then considering the fuel consumption and the CO2 emissions per liter the green emission (GHG) can be estimated as follow: Where FC is the fuel consumption [l/h), CO2unit is the diesel CO2 emissions [kg/l], and Δt is time that the diesel is operating.
The same equation can be calculated using the energy consumed based on the load file in figure 2.
The Rwandan State-run Energy Development Corporation Limited (EDCL) has been working harder to increase the percentage of electricity access. Henceforth, the electrical generation capacity is being boosted using renewable and non-renewable sources. Among nonrenewable sources, Heavy Fuel Oil, peat and methane gas fired power plants are being used [8]. These generations are being used as baseload generators, working 24 hours per day. Table 1 presents the figures of the installed capacity of thermal power plants.  Figure 3 presents the projection that has been elaborated by the Rwandan State-run Energy Development Corporation Limited (EDCL) in order to meet the electrical demand. The present paper mainly focuses the diesel generators in order to investigate its contribution to CO2 emissions. Based on the author in [1], the emission factor considered for a diesel generator is 1.27 kg CO2/kWh and using the Eq.3 and the information in given by Table 1, the CO2 emission by the diesel generator is 1.761.744 kg per a day. This reveals that amount of CO2 emissions by the diesel generator by a year is enormous. The integration of renewable sources could help to reduce the CO2 emissions [10]. The following section discusses the amount of CO2 emissions once a PV system is integrated to the Rwandan electrical grid.

CO2 emission reduction after integrating renewable sources in smart grid
There are still many diesel generators in Rwanda to locally support the electrical network. Apart from the high operation costs of those generators, the PV systems, on grid or off grid, reduce substantially the greenhouse gas emission, other pollutants generated by the combustion engines. A simple methodology for estimating the total amount of greenhouse gas emissions (GHG) once a PV system is integrated is as follows: old = eng CO 2 eng (4) = (( eng − enPV ) CO 2 eng )+ enPV CO 2 PV (5) avoided = old − new (6) Where: Eeng and EenPV are energy provide by a diesel generator and a PV system respectively. CO2eng represents an amount of CO2 emissions from a diesel generator per kWh and CO2PV is amount CO2 emissions from a PV system per kWh. GHGold is amount of CO2 emissions of a diesel generator set installed to cover a specific electricity demand without a PV system. GHGnew is amount of CO2 emissions of a generator set installed to cover a specific percentage of electricity demand once a PV system is integrated into a grid. The difference between Eq.4 and Eq.5, expresses amount of CO2 emissions that can be avoided while a PV system is integrated in order to support a diesel generator. The case study was carried out for Sub-Saharan countries. The estimation was done based on the GHG emissions of a stand-alone diesel-generator set without a PV system. Integrating the PV system, which can cover the same electricity demand as the diesel generator reduces the amount of GHG [11] given by Eq. 6. This shows the impact of PV system installation on the environment.

Energy management in a smart grid (SG) with a consideration of reduction of COemission
Many of the mathematical techniques have been used for managing the energy on both production and load sides. Moreover, optimization techniques in electrical engineering have been used for decades with a main objective to optimize investment and keeping the generated energy cost at minimum level [12].
The Figure 4 illustrates the existing optimization techniques among which engineers can choose techniques answering to specific research questions in field of smart grid management.
The flowchart of the proposed algorithm to solve the energy management in a smart Grid is given in Figure 5. The general steps intended to solve the objective constrained problem that aims to minimize the electricity cost and reduces the CO2 emissions in a smart grids and find out operation modes of different loads with organizing between the considered productions systems are described in Figure 5.  Fig. 5.Flow chart of objective function optimization for a PV-diesel hybrid system without energy storage.
The objective function is formulated based on the grid power balance. In other words, the production of electricity must meet the electrical demand as shown in equation 7. The Figure 6 shows the interconnection between diesel generators with a PV system. Most objective functions are formulated based on the cost constraints functions. This paper tries to incorporate the objective function with CO2 emissions constraints. Where Pv, Pgen,i and Pload are production from a PV system, diesel generator production and power consumption at a load level, respectively. Eq.7 shows the balance of power between production and consumption. Based on the same power the objective function for the cost can be formulated.
Where X is the optimization problem design variable. Cpv, Ct and Mt are respectively the present cost of Photovoltaic system, the manufacturing cost of the components of all system and maintenance cost. The same variables are applicable to the diesel generator plus the fuel consumption price. Once, Eq.9 is considered as objective function. ≥ 0, This is the same at the diesel generator. However, the following constraints is related to the operating time for a PV diesel-hybrid system.

Conclusions
The Eq.8, 9 and 10 are true for the cost optimization but the fact is that Eq.11 can be used for letting generators in a network operate in average more than 14 hours a day. Moreover, if Eq.11 is applied the cost in Eq.8 is optimized. Furthermore, the Eq.6 is useful to evaluate the amount CO2 to be reduced. The future work is about the implement the flowchart in Fig.5.