Optimal Design and Operation of a Residential Hybrid Microgrid System in Kasuga City

Abstract

This study aims to find the optimal configuration of an autonomous microgrid system which can be used to meet the electricity demand of a small community in the city of Kasuga, Fukuoka Prefecture, Japan. To this aim, an optimization model was developed, using the least cost perceptive approach and the load patterns of the residential end-uses. The proposed microgrid system in this study consists of a cluster of loads and micro sources such as the wind turbines, solar photovoltaic panels, battery storage, and a diesel generator. In this research, the effect of the demand response capabilities and patterns on optimal power generation from the proposed microgrid was investigated by introducing a standard demand pattern for a family, including three residents, living in a standard Japanese house in Kasuga city.

1 Introduction

After the Great East Japan Earthquake, energy security and vulnerability have become the significant challenges facing the energy system in Japan. Consequently, the rapid increase in power demand was met through higher electricity generation from fossil fuels (coal, oil, natural gas, etc.). It was associated with severe negative environmental impacts in terms of increased greenhouse gas emissions (Phurailatpam et al. 2018). Shifting to Renewable Energy Resources (RERs) can be considered as a vital solution to the problems arisen from the extensive use of fossil fuels in the electric power industry in Japan.

To promote the introduction of RERs on a large scale, following the collapse of public trust in nuclear power due to the Fukushima Daiichi nuclear disaster, the government of Japan has introduced several energy resilience programs, particularly in the residential and commercial sectors in urban areas. Among all new concepts in power generation, the residential community Microgrids serve as the self-sustaining and autonomous hybrid power systems which can use a variety of RERs such as solar and wind and operate disconnected from the utility grid when outages occur. However, integrating several energy sources in microgrids also increases the complexity of the system. It is because that the main drawback of RERs is their unpredictable natures, which raises concerns about the reliability of the power to the user (Singh et al. 2016). Besides, the increasing penetration of RERs with intermittent power generation brings more significant technical challenges to maintaining the balance of power supply and demand in the power system. To this aim, diesel generators and battery storages are widely used as backup power to cover the supply and demand mismatch and improve the system stability in microgrids. However, the main disadvantages of using diesel generators with microgrid systems are their costs and environmental impacts from burning fossil fuel such as gas oil, kerosene, or fuel oil (Singh et al. 2016).

The optimal design of a microgrid in autonomous mode needs a solid understanding of the dynamic nature of the distribution network, which challenges the stability and control effectiveness of the system. In addition to the reliability of the system, the economic viability of the system plays a key role in cost-effective design and usage of a microgrid (Yoshida and Farzaneh, 2020; Shaqour et al., 2020).

Based on the above discussion, this paper will address a simulation modeling approach which can be used to estimate the optimal configuration of a proposed microgrid consisting of solar photovoltaics, wind turbines, batteries, and diesel power to meet the power demand of a selected standard Japanese house in Kasuga city in Japan. Figure 33.1 shows the overall configuration of the proposed microgrid in this research.

Fig. 33.1

Microgrid system

The optimal design of the proposed microgrid will be carried out based on finding the minimum cost of electricity generation by the system.

2 Mathematical Modeling of the Proposed System

2.1 Wind Power Generation

The amount of power output of a wind turbine is directly proportional to the wind speed, which can be calculated using the following formula (Dhundhara et al. 2018):

$${\text{P}}_{\text{wt}} ({\text{V}}) = \left\{ {\begin{array}{*{20}l} {\frac{{\text{P}}_{\text{r}} ({\text{V}} - {\text{V}}_{\text{CIN}} )}{{\text{V}}_{\text{rat}} - {\text{V}}_{\text{CIN}} },} \hfill & {{\text{V}}_{\text{CIN}} \le V \le {\text{V}}_{\text{rat}} } \hfill \\ {{\text{P}}_{\text{r}} ,} \hfill & {{\text{V}}_{\text{rat}} \le V \le {\text{V}}_{\text{CO}} } \hfill \\ {0,} \hfill & {V \le {\text{V}}_{\text{CIN}} \,and\,V \ge {\text{V}}_{\text{CO}} } \hfill \\ \end{array} } \right.$$

(33.1)

where,

P_r:

Constant power [kW]

V_CIN:

Cut-in speed [m/s]

V_rat:

Rated wind speed [m/s]

V_CO:

Cutout speed [m/s].

$${\text{V}} = {\text{V}}_{{{\text{ref}}}} \left( {\frac{{\text{H}}}{{{\text{H}}_{{{\text{ref}}}} }}} \right)^{{\alpha }}$$

(33.2)

V referes to the wind speed at the height of H. V_ref is the wind speed measured at the reference height, H_ref, and $\upalpha$, which is the power-law exponent (Hiendro et al. 2013).

$$\varvec{P}_{{\varvec{wt}}} = \frac{\varvec{\rho}}{{\varvec{\rho}_{0} }}\varvec{P}_{{\varvec{wt}}} \left( {\varvec{STP}} \right)$$

(33.3)

where, $P_{wt} \left( {STP} \right)$ is the wind turbine power output at standard temperature and pressure [kW]. $\rho$ shows the actual air density [kg/m³]. $\rho_{0}$ is defined “the air density at standard temperature and pressure (1.225 kg/m³). Table 33.1 gives the technical specifications of the Air dolphin pro wind turbine which was used in this study.

Table 33.1 Wind turbine input data (Zephyr AIRDOLPHIN Introduction. http://www.shida-kk.com/images/Airdolphin.pdf)

2.2 PV Power Generation

The amount of power output form the PV array is expressed by the following equation (HOMER 2019):

$${\text{P}}_{{{\text{pv}}}} = {\text{G}}_{{{\text{pv}}}} {\text{f}}_{{{\text{pv}}}} \left( {\frac{{{\text{I}}_{{\text{T}}} }}{{{\text{I}}_{{{\text{T}},\,{\text{STC}}}} }}} \right)[1 + \upalpha _{{\text{p}}} ({\text{T}}_{{{\text{C}} - }} {\text{T}}_{{{\text{C}},\,{\text{STC}}}} )]$$

(33.4)

where,

G_pv:

Rated capacity of the PV array power, under standard test conditions [kW]

f_pv:

PV derating factor [%]

I_T:

Solar radiation incident on the PV array [kW/m²]

I_T,STC:

Incident radiation at standard test conditions [kW/m²]

α_p:

Temperature coefficient of power [%/°C]

T_C:

PV cell temperature [°C]

T_C,STC:

PV cell temperature under standard test conditions [25 °C].

PV cell temperature is the same as the ambient temperature at night, but it can increase during sunny days. Following equation represents the energy balance of the PV array. The input data used in this study for the simulation of the power out of a selected PV panel (Panasonic HIT 244α) is given in Table 33.2.

Table 33.2 PV Input data (Hiendro et al. 2013) (Panasonic Residential fuel cell. http://panasonic.co.jp/ap/FC/en_index.html.)

$$\uptau \, \upalpha \, {\text{G}}_{{\text{t}}} = \upeta _{{\text{c}}} {\text{G}}_{{\text{t}}} + {\text{U}}_{{\text{L}}} ({\text{T}}_{{\text{c}}} - {\text{T}}_{{\text{a}}} )$$

(33.5)

where,

τ:

Solar transmittance the PV array [%]

α:

Solar absorptance of the PV array [%]

GT:

Incident solar radiation incident on the PV array [kW/m²]

η_C:

Electrical conversion efficiency of the PV array [%]

U_L:

Overal heat transfer coefficient of the PV [kW/m²°C]

T_C:

PV cell temperature [°C]

T_a:

Ambient temperature [°C]

2.3 Diesel Generation

The fuel consumption of diesel generator is calculated as follows (Jamshidi and Askarzadeh 2019):

$${\text{Cons}}_{\text{G}} = {\text{B}}_{\text{G}} \cdot {\text{P}}_{{{\text{N}}\_{\text{G}}}} + {\text{A}}_{\text{G}} \cdot {\text{P}}_{\text{G}}$$

(33.6)

where,

Cons_G:

Fuel consumption of diesel generator [L/h]

P_{N_G}:

Nominal power of diesel generator [kW]

P_G:

Power output of diesel generator [kW]

A_G:

0.2461 [L/kWh] and BG:0.081451 [L/kWh].

The thermal efficiency of the diesel generator is estimated by using the following equation:

$$\upeta_{\text{G}} = \frac{{{\text{P}}_{\text{G}} }}{{{\text{Cons}}_{\text{G}} }}$$

(33.7)

2.4 Battery Modeling

In this investigation, the Kinetic Battery Model was used to determine the amount of energy that can be charged by or discharged from the storage bank in each time step. The Kinetic Battery model explains the behavior of a lead acid battery based on the operation of two storage tank. The first tank contains “available energy”, or the amount of chemical energy which is available and can be converted into DC electricity. The second tank represents “bound energy,” or energy that is chemically bound and therefore not immediately available for withdrawal. The following equations show the maximum charged by and discharged from the battery bank in each time step (Das et al. 2017):

$${\text{Q}}_{{1,{\text{end}}}} = {\text{Q}}_{1} {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} + \frac{{({\text{Qkc}} - {\text{P}})(1 - {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} )}}{\text{k}} + \frac{{{\text{Pc}}({\text{k}}\Delta {\text{t}} - 1 + {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} )}}{\text{k}}$$

(33.8)

$${\text{Q}}_{{2,{\text{end}}}} = {\text{Q}}_{2} {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} + {\text{Q}}(1 - {\text{c}})(1 - {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} ) + \frac{{{\text{P}}(1 - {\text{c}})({\text{k}}\Delta {\text{t}} - 1 + {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} )}}{\text{k}}$$

(33.9)

where,

Q₁:

Available energy at the beginning of the time step [kWh]

Q₂:

Bound energy at the beginning of the time step [kWh]

Q_{1_end}:

Available energy at the end of the time step [kWh]

Q_{2_end}:

the bound energy at the end of the time step [kWh]

P:

Power into (positive) or out from (negative) the storage bank [kW]

Δt:

Time step [h]

k:

Constant coefficient which shows how quickly the storage can convert bound energy to available energy or vice versa

c:

Battery capacity ratio.

The total amount of energy stored in the battery bank at each time step is the sum of the available energy and bound energy:

$${\text{Q}} = {\text{Q}}_{1} + {\text{Q}}_{2}$$

(33.10)

The battery state of charge (SOC) is given as follows, whereby Q_max is the maximum (rated) capacity of the battery bank:

$$SOC = \frac{Q}{{Q_{max} }} \times 100\left( \% \right)$$

(33.11)

The following equation gives the maximum amount of power charged by $({\text{P}}_{\text{Maxchar}} )$ and discharged $({\text{P}}_{\text{Maxdiscahr}} )$ from the battery bank which can be absorbed over a specific length of time (Δt) (Table 33.3):

Table 33.3 Battery input data (Dhundhara et al. 2018)

$${\text{P}}_{\text{Maxchar}} = \frac{{{\text{kQ}}_{1} {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} + {\text{Qkc}}(1 - {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} )}}{{ 1- {\text{e}}^{{ - {\text{k}}\Delta {\text{t}}}} + {\text{c}}({\text{k}}\Delta {\text{t}} - 1 {\text{ + e}}^{{ - {\text{k}}\Delta {\text{t}}}} )}}$$

(33.12)

$$P_{Maxdischar} = \frac{{ - {\text{kcQ}}_{max} + kQ_{1} e^{ - k\Delta t} + {\text{Qkc}}(1 - {\text{e}}^{ - k\Delta t} )}}{{1 - {\text{e}}^{ - k\Delta t} + {\text{c}}({\text{k}}\Delta {\text{t}} - 1 + {\text{e}}^{ - k\Delta t} )}}$$

(33.13)

3 Optimization

3.1 Cost Analysis

In this study, the Levelized Cost of Energy (LCOE) was considered as a measure for comparative analysis of various power generation technologies (i.e. solar, wind, and diesel) in the proposed microgrid. The LCOE represents the net present value of the total cost of the microgrid over its lifetime, divided by the total output power from the system (DOE Office of Indian Energy Upfront Capital Costs for Renewables.,https://www.energy.gov/sites/prod/files/ 2015):

$${\text{LCOE}} = \frac{{{\text{S}}_{\text{CL}} }}{{{\text{S}}_{\text{EL}} }} = \frac{{\mathop \sum \nolimits_{{{\text{i}} = 1}}^{\text{n}} \frac{{{\text{I}}_{\text{t}} + {\text{M}}_{\text{t}} + {\text{F}}_{\text{t}} }}{{(1 + {\text{r}})^{\text{t}} }}}}{{\mathop \sum \nolimits_{{{\text{i}} = 1}}^{\text{n}} \frac{{{\text{E}}_{\text{t}} }}{{(1 + {\text{r}})^{\text{t}} }}}}$$

(33.14)

where,

${\text{S}}_{\text{CL}}$:

Sum of costs over the lifetime ($)

${\text{S}}_{\text{EL}}$:

Sum of electrical energy produced over the lifetime (kWh)

${\text{I}}_{\text{t}}$:

Investment expenditures in the year t ($)

${\text{M}}_{\text{t}}$:

O&M expenditures in the year t ($)

${\text{F}}_{\text{t}}$:

Fuel expenditures in the year t ($)

${\text{E}}_{\text{t}}$:

Electrical energy generated in the year t (kWh)

r:

Discount rate (%)

n:

Expected lifetime of system or power station.

For the battery storage, the following formula may be used to compute the LCOE (What is the true cost to you behind energy storage?,https://www.solarpowerworldonline.com/ 2016):

$$LCOE_{battery} = \frac{P}{{C_{a} \times C_{y} \times E_{f} \times DoD}} + T_{AC}$$

(33.15)

where,

P:

Price of the battery storage ($)

C_y:

Number of full charge and discharge cycles expected over the guaranteed life of the battery

C_a:

Capacity of the battery storage (kWh)

E_f:

Storage efficiency (%)

DoD:

Depth of discharge (%)

T_AC:

Total ancillary costs ($).

The cost items, including the capital investment, and operation costs of the main components of the proposed microgrid is given in Table 33.4. Based on these values, the LCOE for each component is derived and is shown in Table 33.5.

Table 33.4 Cost items of each component

Table 33.5 LCOE for each component

3.2 Optimization Model

The objective function of the optimization model is defined as the total cost of electricity supplied by the proposed system as follows (Farzaneh, 2019):

$$Total\,energy\,cost = \sum\limits_{t} {\sum\limits_{e} {Cost_{ e,t} } }$$

(33.16)

where,

$Cost_{e,t}$: the cost of using of each component ${\text{e}} \in \left[ {PV, Wind, Battery, Diesel } \right]$ at each hour t $\left( {0 \le t \le 23} \right)$.

In the above equation, the cost of electricity generated by each component is calculated as follows:

$$Cost_{ e,t} = LCOE_{ e } . E_{ e,t}$$

(33.17)

where,

$LCOE_{ e }$:

LCOE of component e,

$E_{e, t}$:

Amount of electricity generated by component e in time step t

The cost function is minimized subject to the following constraints:

$$E_{PV, t} + E_{Wind, t} + E_{Battery\_discharge, t} + E_{Diesel, t} \ge E_{Demand, t}$$

(33.18)

$$E_{Battey\_Charge, t} = E_{PV, t} + E_{Wind, t} - E_{Demand, t}$$

(33.19)

$$E_{Battery,charge,t} \le P_{Maxchar} .\Delta t$$

(33.20)

$$E_{Battery,discharge,t} \le P_{Maxdischar} .\Delta t$$

(33.21)

$$0.2 \le SOC \le 0.8$$

(33.22)

$$E_{{Diesel, {\text{t}}}} \le C_{load} P_{Diesel}$$

(33.23)

where,

$C_{load}$:

The load factor for emergency diesel usage, which is set to 0.7.

$E_{Demand, t}$:

Amount of electricty demand at hour t

4 Results and Discussions

In this research, a standard demand pattern, so-called “Normal schedule” was considered as the baseline load for a Japanese family, including three residents (Father, Mother, and a kid), living in an average size house (94.13 m², 4 rooms) located in Kasuga city, Fukuoka prefecture. The usage method of the main electrical appliances in the normal schedule is shown in Table 33.6. The rated electrical power of each appliance is given in Table 33.7. Accordingly, the load curve in the Normal schedule based on a daily average electricity demand in the selected case study is illustrated in Fig. 33.2. The total amount of electricity demand calculated here was roughly in agreement with the average Japanese household electricity demand based on collected data from the agency for Natural Resources and Energy in Japan [Annual Report on Energy (White Paper 2013)].

Table 33.6 Usage schedule

Table 33.7 Electric appliances rated power

Fig. 33.2

Hourly average demand load curve

The developed model was used to estimate the optimum size and capacity of the proposed microgrid system. The optimum configuration of the microgrid comprises a system with a 2.3 kW of the wind turbine, 4.3 kW of the PV panels, six units of 1kWh batteries and 0.5 kW of diesel generator. Based on the results of the optimization model, the hourly electricity supplied by the proposed microgrid in summer and winter is depicted in Fig. 33.3.

Fig. 33.3

Hourly electricity generation from the proposed microgrid

Figure 33.4 shows the monthly average electricity supplied by the proposed microgrid. The diesel generator is only used in summer due to lack of availability of wind energy.

Fig. 33.4

Monthly electricity generation

Figure 33.5 shows that the battery bank accounts for 50 and 64% of the total cost of electricity generation in summer and winter, respectively.

Fig. 33.5

Share of the different components in total electricity generation cost

The modeled LCOE is compared to the existing tariff structure in Japan in Fig. 33.6. The estimated LCOE’s of the proposed microgrid is on the order of 64.7 JPY/kWh in summer and 51.26 JPY/kWh in winter which is still well above the average electricity tariff in Japan, creating a cost gap that should be addressed through effective policy and incentive programs in the near future.

Fig. 33.6

The LCOE comparison

5 Conclusion

In this paper, modeling and simulation of a fully autonomous microgrid system were conducted in order to find the optimal configuration of the system based on the minimum total cost of the electricity generation. The representative micro-grid system considered includes a hybrid solar-wind-battery-diesel generation system, which currently is the most widely applicable and rapidly growing microgrid system type. The proposed microgrid was planned to be analyzed in a standard Japanese house in Kasuga city in Fukuoka prefecture. The results revealed the total power output from the system could sufficiently meet the demand load requirements of the selected household area. This simple analysis offered an estimate of the gap between current tariff regime in Japan and the LCOE of the system which can provide insight into the potential scale of incentives that may be required to enable micro-grid development under the current regulatory structure.