Developing operation of combined cooling,heat,and power system based on energy hub in a micro-energy grid: The application of energy storages

Abstract

This research presents optimum operation strategies for multi-energy systems as combined cooling, heat, and power systems with the approach of electricity supply priority regime. A grid with energy hub structure including electrical, thermal, and cooling hubs is explored. Accordingly, an energy hub structure integrated with storages and renewable resources is designed. The mathematical model of the operation system for the presented triple generation micro-grid is considered for energy flow to the energy grid. Considering the limitations of the storage system and the performance of the equipment as well as electricity and gas line, a dynamic optimum operation model is prepared on the basis of mixed integer linear programming and is solved in general algebraic modeling system optimization software so as to minimize energy supply costs. For the model verification, different scenarios are developed in a residential building for a typical summer day so that the renewable resources and storages are fed into the system gradually. According to the findings, as each element is included in the micro-energy grid, its operational parameters, viz. the cost of electricity, gas, and pollutant emissions, are improved remarkably. Scenario I includes a combined cooling, heat, and power system, Scenario II is supplemented with renewable solar and wind energy, and Scenario III includes electrical, heat, and cold storages in addition to renewable sources and combined cooling, heat, and power system. Results reveal a decrease in total productivity cost by 12.2% in Scenario II versus Scenario I and 10.9% in Scenario III versus Scenario II.

Keywords

Combined cooling heat and power system micro-energy grid energy hub energy storages renewable resources

Introduction

Recently, an avid interest has been provoked about energy security and environmental conservation throughout the world, mainly in the face of depleting fossil fuel reserves and environmental pollution. At present, most conventional energy infrastructures, like natural gas and electrical networks, are programmed and exploited separately and in isolation. This impairs the efficiency markedly and incurs high operating costs. Di Somma et al. formulate a stochastic mixed integer linear programming (MILP) problem to find the optimal bidding strategies in the day-ahead market in order to maximize the expected profit. The numerical results show that the method is efficient in finding the bidding curves in the day-ahead market.¹ Similarly, Duić and Rosen² discuss the optimization of energy structures through integrating renewable resources in order to help to improve energy productivity and develop energy systems. As the micro-grids and smart technologies develop, more use is made of the renewable energies that make the grids cleaner, more efficient, and more reliable. However, consumers may feel the need for the heating, cooling, and gas fuel in addition to electrical requirements, too. Traditional energy systems supply these different energy forms separately, which is a source of poor efficiency and costly operation. Recently, relying upon gas consumption, combined heat and power (CHP) resources have been developed as a highly efficient technology with a rapid pace. Combined cooling, heat, and power (CCHP) systems (including heat exchangers, boilers, and cooling equipment), when integrated into the networks, reduce the emission of carbon and other pollutants and, at the same time, help to supply very highly efficient and cost-effective thermal, cooling, and electrical loads.³ The efficiency of these systems is generally reported to range from 60 to 80%. This excellent efficiency is by itself a motive to grasp the opportunity to employ them in order to enhance the productivity and enable environmental and economic benefits.⁴ CHPs are a link between electrical and gas systems, tying diverse energy carriers to one another. With the integration of electrical, natural gas, and other systems, power systems have been replaced with new integrated systems, and various new forms of energy systems like multi-energy systems have been introduced.⁵ Mancarella⁶ implemented this concept within smart multi-energy systems. To have a look at the development of energy system, an energy network has been lately introduced in that a highly efficient power network is integrated with renewable resources. The mechanism of energy distribution by this network is analogous to how data are distributed in a computer system. The development of integrated energy systems is an approach to developing power grids. These integrated energy systems are composed of at least two sub-systems, i.e. electricity, cooling, heating, natural gas, etc. Thus, different forms of energy are coupled and are in exchange with one another, so the whole energy system is analyzed for the coordination of different energy carriers, higher efficiency, and cost saving.⁷

Recently, the concept hub or ring has been suggested and received extensive feedbacks.⁸ An energy hub is a unit where diverse energy carriers can be converted and stored. This system typically uses electricity and natural gas as its inputs, and its outputs are various services like the delivery of electricity, heat, and/or cooling.⁹ Extensive research has focused on energy hub including modeling, optimization of energy structures, and optimization of multiple energy carrier systems. The concept of energy hub as the interface in multi-carrier energy systems has motivated researchers to concentrate on multi-carrier energy systems with the purpose of achieving more efficient performance.¹⁰ Similarly, Bahrami and Safe¹¹ present a structure for the optimization of systems and modeling composed of multiple carriers and formulate a nonlinear programming model for optimizing energy flow based on energy hub concept. Rastegar and Fotuhi-Firuzabad¹² design a residential energy hub model for a house and explore the operation mode of the system. Bozchalui et al.¹³ explore the optimization models of residential energy hubs. It should be noted that the load curves for each hub system are the consumers’ preferences and demanded welfare which influence the estimation of demand and load of hub system. Rastegar et al.¹⁴ present another residential hub system for a smart house including hybrid vehicles and CHP. El-Zonkoly et al. propose a two-layer modified firefly based on an optimization algorithm to specify the optimum CHP sizes in an energy hub. Their proposed algorithm was, then, applied to a hospital.¹⁵ Wang et al. discuss an interval optimization algorithm based on an energy hub in a micro-grid. They indicate that the proposed method is functional for the application of economic and renewable energy dispatch of the micro-grid.¹⁶ But, the cited literature has merely focused on residential and small hub systems. Consequently, these models lack the resilience required for the perfect simulation of a real system.

Ippolito et al.¹⁷ presented a new device for the control of and connection to a supply utility grid of combined renewable energy sources based generators and electric storage systems. Fu et al. demonstrate that the operating nature of micro-grids drastically differs from conventional distribution systems for several reasons, including energy storage (ES), renewable sources, high penetration of distributed generations, and power electronics-based components. The micro-grid architecture and components, various control techniques, and protection design for micro-grid systems are also discussed in detail.¹⁸ Emodi et al.¹⁹ argue that at the demand side, consumers may incline to consume electricity more when prices are lower and temperatures are higher in summer and this poses a huge potential to reduce peak loads in summer. Vatankhah Barenji et al.²⁰ discuss an off-grid hybrid energy generation system to minimize the cost of energy generation. Graditi et al. present an optimal power dispatch problem on a 24 h basis for distribution systems with distributed energy resources that also includes directly controlled shiftable loads. In this paper, a new formulation of shiftable loads is employed.²¹

ES is becoming a key element of achieving the goals of energy sustainability including energy and cost savings.²² Eshraghi et al. implemented a demand response program with both time-of-use (TOU) and real-time-price pricing policies on an energy micro-grid including renewable resources and ESs. They report about 28% lower operational costs and pollutant emission.²³ Poullikkas reviews different types of batteries used for large-scale electricity storages. In particular, he presents the current operational large-scale battery ES systems around the world with their applications.²⁴ Graditi et al. present a study to evaluate the economic viability of using a battery ES system for load shifting applications at a consumer level, where TOU electricity tariffs are applied.²⁵ Marzband et al.²⁶ study the concept of home micro-grids. They adopt the game theory and formulate different players with competing objectives. El-Zonkoly²⁷ discusses that the proposed algorithm is to minimize the overall operational cost including consumed electrical/gas energy cost, generated energy cost, energy loss cost, and cost of unserved energy. Warren²⁸ demonstrates that demand-side policy plays an increasingly influential role as a complement to low carbon emission on the supply side in the revolution to a more environmentally sustainable energy system. Chen et al. discuss an optimization method proposed for a CCHP in a smart grid. It can be used for electricity on an hourly demand.²⁹ On the other hand, micro-grid has an important role in energy generation in residential buildings; for example, the micro-energy grid (MEG) system is used in 30–40% of the buildings in Europe and the UK.³⁰ The main advantage of an MEG versus other energy grids lies in the fact that they are generally composed of various forms of energy, as well as cooling, heating, and electric power. In fact, an MEG is a development notion of energy grids.³¹

To summarize, the following shortcomings can be identified in the existing literature with respect to energy hubs in a MEG:

The lack of the application of ES devices for optimum operation of the CCHP systems as in literature.^{10,14,15,19,21,26,28,29}

The lack of the application of an MEG modeling approach with the help of the energy hub model as in literature.^{1,3,4,7,9,13,16,18–20,22,24–26,28–39,47}

The lack of the application of renewable resources for optimum operation of the CCHP micro-grid as in literature.^{3–5,8,9,11,22,28,29,32,33}

The lack of the modeling of energy system with the help of MILP method to accomplish an optimum answer with global optimum type as in literature.^{2,3,5,7–12,14–22,24–33,35,37}

A look at the shortcomings and gaps in past studies (as reviewed above) and the analysis of the results prove the significance of the present research. The study seeks to evaluate the impact of adding (applying) ESs to an MEG in the presence of renewable resources. To give more resilience to the operation of the existing MEG, the structure of an electrical energy, heating, and cooling hub is implemented with electrical, heating, and cooling storages and wind and solar renewable resources. This will result in a higher efficiency of energy production and lower pollutant emission and operational costs in spite of all constraints of an energy grid. The problem is solved by the MILP method. Even when the problem is solved iteratively, one single answer is yielded by this method—a globally optimal answer that is generated rapidly with high responsiveness in real time.

The major contributions of this paper can be summarized as follows:

The application of three types of storage devices, including electrical, heating, and cooling, which is a real energy conversion model based on time series that improves the efficiency of the MEG.

The application of a multiple energy hub model in designing the studied micro-grid that enhances the operational flexibility of the MEG.

The application of renewable resources along with the ESs that increases electricity sale to the grid and reduces electricity purchase from the grid.

The application of MILP method that yields a globally optimal definitive, single answer with high speed in the form of real time.

System description and assumptions

Assumptions

The present research mainly considers electricity supply priority regime. Also, the electricity selling prices from/to the grid are assumed to be equal. The existing resources are supposed to have certainty in delivering the power, and discussion on uncertainty and contingency issues are ignored. Also, the simulated model was evaluated by energy demand in a typical summer day for a residential building located in the north of Tehran as a case study.

System description

Architecture of an energy hub integrated with CCHP and a multi-energy hub

A review of the literature reveals that CCHP systems do not typically include renewable resources and storages simultaneously. However, to improve the flexibility and reliability, the present study designs an energy hub system that includes storages, renewable resources, and CCHP, and to illustrate the inherent structure of the energy hub, a model is presented that is composed of an electricity hub, a heat hub, and a cooling hub, reflecting the exchange and power balance. As is depicted in Figure 1, the architecture of the presented energy hub system is designed on the basis of CCHP and other carriers are composed of five sections.

Figure 1.

The architecture of a micro-grid with small hub systems and CCHP. PV: photovoltaic; WT: wind turbine.

Energy resources including natural gas, photovoltaic (PV), wind turbine (WT), and electricity supplied from the grid by a transformer.

Energy conversion including gas turbine, boiler, electric chiller, heat exchanger, etc.

Energy accumulator/distributor in which electricity, cooling, and heat are accumulated and then, it is aggregated and distributed by electric, cooling, and thermal hubs, and the power balance is preserved at all moments.

ESs including electric, cooling, and heat storages.

Energy delivery including power, gas, and heat distribution networks.

The proposed energy hub (Figure 1) clearly depicts the relationships between the components and displays energy flow processes separately.

Methodology

Energy hub system modeling

Energy hub is composed of multiple energy exchangers and ESs. The model of ESs is formulated in the time series in terms of power transfer. Electricity tariff expressed in kW h is shown for different hours in Figure 3. The tariff of natural gas is 0.35 USD/unit and the penalty for CO₂ emission is 0.031 USD/kg.

Conversion model of natural gas

A portion of the procured natural gas is flown into the gas turbine for the cogeneration of electricity and heat and the remaining is burned in the gas boiler for heat generation

P_{gas} = P_{ge} + P_{gh}

(1)

where P_ge and P_gh are the natural gas burned in gas turbine and gas boiler, respectively.

In the model converting natural gas to electricity, P_ge is burned by the gas turbine to generate electricity, P_gt. (see Appendix) It can be observed in Figure 1 that

η_{ge} \times P_{ge} = P_{gt}

(2)

where η_ge denotes the efficiency of the gas turbine and P_gt is the electricity produced by the gas turbine.

In the natural gas to heat conversion model, the natural gas, P_gh, is converted to heat by the gas boiler, while P_ge can produce heat and electricity concurrently in the gas turbine. (see Appendix) Then

η_{gh, gb} \times P_{gh} = H_{gb}

(3)

η_{gh, gt} \times P_{ge} = H_{gt}

(4)

where η_gh_, _gb and η_gh_, _gt represent the heat generation efficiency for the gas boiler and CCHP, respectively, and H_gb and H_gt represent heat generated by the boiler and gas turbine (see Appendix).

The total output power of CCHP can be related to the input power by equation (5)

P_{ge} = (P_{gt} + H_{gt}) / η_{CHP, tot}

(5)

Conversion model of electrical energy

In electrical energy equation between the hub and power grid using a transformer, the energy hub uses a transformer to exchange the electricity with the power grid. When there is an electricity shortage, the hub will purchase it from the power grid, but when there is an electricity surplus in the hub, it can sell it to the power grid (see Appendix).

η_{t} \times P_{grid} = P_{in}

(6)

where P_grid represents the electricity exchanged between the power grid and the energy hub, η_t represents the efficiency of the transformer, and P_in represents the actual input power to the hub from the grid. A positive P_in will imply the purchase of electricity from the grid and a negative P_in will imply its sale to the grid (see Appendix).

In electricity to cooling conversion model, the electrical chiller consumes electricity to supply its cooling load. The model of this instrument is well illustrated below

P_{ec} \times CO P_{ec} = c_{ec}

(7)

where COP_ec and c_ec denote the coefficient of performance for the electrical chiller (compression) and its power output (see Appendix), respectively. In addition to this model, the present paper uses cooling storages to improve the performance of cooling hub and to transfer the load to the off-peak hours which will reduce the costs remarkably.³²

Energy conversion model of ESs

ESs are crucial parts of an MEG and are capable of recognizing energy transfer at different time intervals.³³ Storages store the surplus or cheap energy and deliver it to the grid when there is an energy shortage or it is expensive. Inasmuch hybrid vehicles play the role of load and power input source for the energy hub, then this device can be regarded as a battery. This description has been applied in the operational optimization of the residential areas.⁴⁰ The ESs include not only electrical storages like batteries but also heat storages and cooling storages like ice storage tanks. The discharging and charging process is the same for all storages and all of them have a set of capacity limitations and technical constraints.³⁴ Accordingly, the following overall dynamic model describes the electrical, thermal, and cooling storages as a real model of energy transfer in time series

E_{x}^{t + 1} = E_{x}^{t} (1 - δ_{x}) + (P_{x, c}^{t} . η_{x, c} - \frac{P_{x, d}^{t}}{η_{x, d}}) . Δ t

(8)

0 \leq P_{x, c}^{t} \leq u_{x} . P_{x, c}^{\max}

(9)

0 \leq P_{x, d}^{t} \leq (1 - u_{x}) . P_{x, d}^{\max}

(10)

E_{x}^{\min} \leq E_{x}^{t} \leq E_{x}^{\max}

(11)

E_{x}^{0} = E_{x}^{24}

(12)

The subscript x shows the energy type, which can be electrical, thermal, or cooling. Equation (8) describes energy variations in ES over the period Δt before and after charging and discharging. Also, $δ_{x}$ represents the rate of energy wastage in storage, $P_{x, c}^{t}$ represents the charged power, $P_{x, d}^{t}$ denotes the discharged power, and $η_{x, c}$ and $η_{x, d}$ show the efficiencies of charging and discharging, respectively.

Equations (9) and (10) imply that the discharge and charge rates will not exceed the maximum possible rate. The binary variable $u_{x}$ , which can take the values of either 0 or 1, has been introduced into the equation to prevent the simultaneous occurrence of discharging and charging. Equation (11) shows the lower and upper limits of energy stored in storages, and equation (12) expresses that the mode of ES at the end of distribution period should be the same as the original energy. This research assumes the distribution period to be a typical day based on a residential building located in the north of Tehran which can be generalized to all building types and all days of the year.

Optimum operation model

This section describes the optimum model of operation on the basis of the hub model described in the last section. This is a real model for optimum distribution or operation in the MEG.

Objective function

The optimum operation objective of an MEG is to minimize the daily operation costs that include the cost of gas purchase (M_pg), the cost of electricity purchase (M_pe), and the cost of carbon emission (M_ce)

M = min (M_{pe} + M_{pg} + M_{ce})

(13)

In electricity purchase cost, M_pe is calculated by

M_{pe} = \sum_{t = 1}^{24} (P_{e}^{t} . P_{in}^{t} . Δ t)

(14)

where

P_{e}^{t}

i is the cost of electricity at hour t. As Figure 3 shows

P_{in}^{t}

denotes the value of input electricity to the energy hub at hour t (see Appendix). It implies electricity sale when it is negative and electricity purchase when it is positive.

In gas purchase cost, M_pg, as expressed in equation (15), the purchased gas is divided into two parts: $P_{gh}^{t}$ and $P_{ge}^{t}$ which are consumed by the gas boiler and turbine, respectively. Assuming that $P_{g}$ represents the base price of purchased gas from the network (as is evident in Table 3), the cost of gas is described as below

M_{pg} = P_{g} \sum_{t = 1}^{24} (P_{ge}^{t} + P_{gh}^{t}) = P_{g} . \sum_{t = 1}^{24} (\frac{P_{gt}^{t}}{η_{ge}} + \frac{H_{gb}^{t}}{η_{gh, gb}}) . Δ t

(15)

M_ce is carbon emission cost that is composed of the equivalent of carbon emission costs of the gas procured from the gas network and the electricity procured from the power grid. Assuming that

β_{g}

and

β_{e}

are the coefficients of emission for gas and electricity, respectively, and

ε

is the cost paid for CO₂ per kg (as shown in Table 3), then the cost of carbon emission can be described by the below equation

M_{ce} = ε . \sum_{t = 1}^{24} (β_{e} . P_{in}^{t} + β_{g} . (\frac{P_{gt}^{t}}{η_{ge}} + \frac{H_{gb}^{t}}{η_{gh, gb}})) . Δ t

(16)

Constraints and limitations

The constraints pertaining to the objective function include the energy balance constraint in small hub systems, the electricity technical constraints, heat and cooling storages, the constraints on the performance of all instruments of the hub, and the constraints on the grid between hub and gas line network and electricity. These are described below.

The electrical power balance in the electrical hub is obtained from

P_{in} + P_{pv} + P_{wt} + P_{gt} + P_{es, d} = P_{ec} + P_{es, c} + P_{L}

(17)

in which the right-hand side represents power allocation and the left-hand side expresses the input electricity to the electrical hub.

Thermal heat balance in the heating hub is calculated by

η_{he} . H_{gt} + H_{gb} + P_{hs, d} = P_{hs, c} + H_{L}

(18)

in which the right-hand side shows the heat allocation and the left-hand side indicates the input heat to the thermal heating hub in which

η_{he}

denotes the heat exchanger efficiency,

H_{gt}

denotes the output turbine heat,

H_{gb}

shows the boiler output heat,

P_{hs, d}

indicates the output heat power of the heat storage (discharging process),

P_{hs, c}

shows the input heat power of the heat storage (charging process), and

H_{L}

represents the thermal load demand (see Appendix).

Cooling energy balance in the cooling hub is obtained from

C_{ec} + P_{cs, d} = C_{L} + P_{cs, c}

(19)

in which the right-hand side denotes the cooling allocation and the left-hand side represents the input cooling of the cooling hub.

It is considered as another constraint that all instruments should operate within their upper and lower limits. The mathematical model of optimum operation described above is an MILP problem. To avoid the simultaneous occurrence of the discharging and charging procedure in ES devices, binary variables (0–1) are included. The optimum distribution model is resolved with a C + Simplex (Cplex) solver in the general algebraic modeling system (GAMS) software, which is popular in solving mixed integer linear problems.

Model validation

Since there is a lack of an experimental set, it was impossible to validate the model by comparing simulation output with the behavior of a real system. In addition, given the models in the reviewed literature, no precise system was found that was simulated and properly confirmed. Thus, a combination of appropriate techniques was used for validation and verification (V&V). Validation aims to confirm that the conceptual model can be a real system. On the other hand, verification aims to confirm that it can be ensured that the computer program can be run correctly and the simulation by the program can be considered as expected. The V&V process has widely been used for simulation models in literature.^41–45 The present study used some practical techniques such as model verification and assessment and its performance evaluation.^35,45 The simulation output was examined by considering the following points:

It should be ensured that the primary assumption of the conceptual model and the equation used in the model are scientifically accepted and valid.

It should be checked whether the model logic is correct. This is performed by following different specific behaviors during simulation. The present operational study uses the diagrams of electricity/heat/cool generation by each element for a 24 h interval in order for better and more tangible observation of the verification process.

The experiment should show that similar relationships of the model occur in the real world. Sensitivity analysis (parameter variable) reveals the impacts of the variation of input and interval value on the behavior of the model and the output values. In this study, a model has been designed to validate the optimization model by optimizing for three scenarios, each with input variables differing from the other scenarios.

This section first presents the results of V&V for the energy flow model (system level) and the component level (sub-model level). Then, the optimization algorithm is validated by observing the model behavior for the proposed scenarios. For systematic V&V, the followings should be considered:

Component model (sub-model)

○ if the model output for each component (sub-model) is according to the theory (conceptual model) as expected.

Energy flow model (system level)

○ if power, heating, and cooling are in equilibrium at each time step.

○ if the operation of each individual component has been defined as per the operational strategy.

○ if the system operation has been defined as per the operational strategy.

Optimization algorithm

○ if the optimization algorithm response is sensitive and valid to the systematic variations of the input variables.

Scenario description

For V&V purpose, a test case was defined and the simulation output behavior was examined for a typical day quantitatively and qualitatively. The case study is on the data for energy demand of a residential building in the north of Tehran. As such, to find out the impacts of renewable energy resources and the storages on the optimal operation of an MEG, three case studies were designed as presented in Table 1. The ✗ sign shows that the hub model depicted in Figure 1 does not include the equipment or demand response program, but the ✓ sign shows that the hub in the studied state uses the equipment of demand response program.

Table 1.

Scenarios of micro-grid operation.

Scenarios	CCHP	PV + WT	ES	Electricity tariff
Scenario I	✓	✗	✗	TOU
Scenario II	✓	✓	✗	TOU
Scenario III	✓	✓	✓	TOU

CCHP: combined cooling, heat, and power; ES: energy storage; PV: photovoltaic; TOU: time of use; WT: wind turbine.

A V&V technique for optimization algorithms is sensitivity analysis. The model should supply valid and reasonable solutions when it encounters systematic variation of input variables, and the variations should be in the direction that is expected from a real system. Three optimization designs have been defined to test the optimization algorithm. The specifications of the three optimization scenarios and the reference system are presented in Table 1. The results of operation cost saving in each scenario versus its previous scenario, which is reasonable and expected, are consistent with the results of optimization by the software simulation. This implies the validity of the modeling and simulation method and the model designed in the present study.

More precisely, the hub stated in Scenario I includes none of the storages and renewable resources and uses just a single CCHP system. The studied hub of Scenario II is equipped with CCHP and renewable resources (wind and solar energy as shown in Figure 2) but has no storages. Finally, Scenario III uses not only CCHP, PV, and WT simultaneously but also the electric, heat and cooling storages. As can be observed in Table 1, Tables 2 and 3 display the parameters of storage devices and instruments, respectively. Scenarios I–III apply TOU pricing plan whose values can be found in Figure 3.

Figure 2.

Output electric power of PV and WT. PV: photovoltaic; WT: wind turbine.

Table 2.

Micro-grid storage parameters.^36–39,47

$δ_{es}$	$P_{es, c}$	$P_{es, d}$	$E_{es}^{\max}$	$E_{es}^{\min}$	$η_{es, d}$	$η_{es, c}$
0.01	500	700	1800	400	0.96	0.96
$δ_{hs}$	$P_{hs, c}$	$P_{hs, d}$	$E_{hs}^{\max}$	$E_{hs}^{\min}$	$η_{hs, d}$	$η_{hs, c}$
0.02	800	800	1800	400	0.98	0.98
$δ_{cs}$	$P_{cs, c}$	$P_{cs, d}$	$E_{cs}^{\max}$	$E_{cs}^{\min}$	$η_{cs, d}$	$η_{cs, c}$
0.02	700	800	1800	400	0.95	0.97

Table 3.

Parameter of the instruments.^36,37,39,46

$H_{gb}^{\max}$	$P_{gt}^{ma x}$	$P_{wt}^{\max}$	$P_{pv}^{\max}$	$P_{ec}^{\max}$	$η_{ge}$
800	1000	200	180	500	0.3
$η_{gh, gt}$	$η_{gh, gb}$	$CO P_{ec}$	$η_{CHP, tot}$	$η_{t}$	$ε$
0.4	0.9	4	0.8	0.98	0.031
$β_{g}$	$β_{e}$	$P_{gas}^{\max}$	$P_{grid}^{\max}$	$P_{g}$
0.23	0.972	3400	1500	0.35

Figure 3.

Electricity tariff with TOU plan.^38,39

Furthermore, Figure 4 displays the demand for electric, heat and cooling loads for the MEG in a residential building on a typical summer day in the north of Tehran so as to express the consumers’ behavior for the simulation.²³ As well, it is supposed that the loads, outputs of instruments, and gas and electricity tariffs are fixed in each time interval.

Figure 4.

Electric, heat, and cooling loads profile of a residential building on a typical summer day in the north of Tehran for model verification.²³

Results of simulation

After performing the simulations for the three scenarios, the results can be drawn separately. In each individual scenario, the amounts of electric, heat and cooling powers were recorded.

Scenario I

The output electrical power of transformer whose values are presented in Table 4 is fed into the MEG. As was mentioned, the electricity in Scenario I is generated only by the CCHP unit and power grid and is supplied to the MEG.

Table 4.

The decision variables of Scenarios I, II, and III.

Time (h)	Pin (kW)			Pgt (kW)			Hgb (kW)			Pec (kW)
Time (h)	S1	S2	S3	S1	S2	S3	S1	S2	S3	S1	S2	S3
1	754.5	666.5	666.5	0	0	0	690	690	609.8	174.5	174.5	174.5
2	661.8	601.8	601.8	0	0	0	610	610	610	161.8	161.8	161.8
3	559.0	479.0	479.0	0	0	0	600	600	608.1	149.0	149.0	149.0
4	523.6	440.6	440.6	0	0	0	595	595	603.1	163.6	163.6	163.6
5	621.8	583.8	990.5	0	0	0	590	590	598.1	181.8	181.8	186.0
6	731.8	640.8	1268.0	0	0	0	605	605	613.1	201.8	201.8	329.0
7	945.4	800.4	1427.7	0	0	0	680	680	688.1	265.4	265.4	392.7
8	314.2	157.2	149.5	670.3	670.3	677.9	0	0	0	274.5	274.5	274.5
9	241.8	56.8	49.1	750	750	757.6	0	0	0	301.8	301.8	301.8
10	92.7	−99.2	−106.8	778.1	778.1	785.7	0	0	0	330.9	330.9	330.9
11	47.3	−145.6	−53.1	806.2	806.2	813.9	0	0	0	363.6	363.6	389.8
12	−3.8	−206.8	−1203.9	848.4	848.4	1000	0	0	0	374.5	374.5	229.0
13	−84.7	−271.7	−1093.4	876.5	876.5	1000	0	0	0	381.8	381.8	296.2
14	−187.5	−325.5	−388	937.5	937.5	1000	0	0	0	360	360	360
15	−185.8	−341.8	−62.5	942.1	942.1	668.5	0	0	0	336.3	336.3	337.8
16	−77.2	−229.2	271.5	853.1	853.1	853.1	0	0	0	290.9	290.9	306.3
17	−20.2	−174.2	445.3	834.3	834.3	842.0	0	0	0	269.0	269.0	396.3
18	66.1	−50.8	478.4	787.5	787.5	885.4	0	0	0	263.6	263.6	390.9
19	208.6	181.6	−899.7	764.0	764.0	1000	0	0	0	292.7	292.7	147.2
20	218.9	110.9	−738.2	843.7	843.7	1000	0	0	0	272.7	272.7	188.6
21	320.4	170.4	−4.5	825	825	1000	0	0	0	245.4	245.4	245.4
22	385.3	277.3	32	754.6	754.6	1000	0	0	0	220	220	220
23	1048.1	1018.1	1146.7	0	0	0	740	740	0	218.1	218.1	238.4
24	910	855	70.3	0	0	0	700	700	219.8	200	200	115.3

As is evident in Table 4, due to the higher price of electricity at hours 08:00 to 22:00, the operator orders the CCHP unit to assume the main responsibility for the supply of electricity. The fifth column of Table 4 presents the output electrical power of this unit. The load unsupplied by this unit is purchased from the power grid.

The output thermal power of the gas boiler is displayed in the eighth column of Table 4. As expected, given the high efficiency of the boiler in heat supply and its lower costs, this equipment is in priority for the supply of thermal load. Therefore, the gas boiler first supplies the thermal load and then if required, the load shortage is supplied from the waste heat energy of the CCHP unit. Since the CCHP unit operates at 08:00 to 22:00, it can deliver thermal power during these hours.

Now, it is dealt how cooling power of the MEG is supplied. This is performed by the compression chiller. The input electrical power of this chiller for the supply of cooling load is presented in the 11th column of Table 4. The compression chillers are known as one of the best equipment in the cold industry due to their relatively high efficiency. The use of Scenario I will entail an operational cost of $21,996.210. The cost of electricity purchase in this scenario is $4838.04, the cost of gas purchase is $16,575.632, and the cost of carbon emission penalty is $581.538.

Scenario II

This scenario utilizes renewable energy generation resources effectively and dynamically in addition to the use of the CCHP unit. Since our objective is to operate the MEG, the investment costs are not considered. So, these resources can be considered free resources on the horizon of the system operation. Here again, the representation trend of the powers is followed. The output power of the power transformer that flows into the MEG is presented in the third column of Table 4.

As the figures in the third column of Table 4 shows, the power sale is significantly higher in this scenario than Scenario I. This by itself suggests the great benefits of the use of renewable resources. The output electrical power of the CCHP is shown in the sixth column of Table 4. To avoid the high costs due to higher prices of electricity in certain hours, the CCHP unit is taken into use to do its main responsibility, i.e. electricity generation, in order to reduce the MEG operational costs. The output thermal power of the gas boiler is presented in the ninth column of Table 4. The excellent efficiency of this equipment in heat generation makes it the main unit of thermal power generation. The cooling load is supplied by a powerful, highly efficient compression chiller. To represent the trend of cooling load supply, the input electrical power to the compression chiller is presented in the 12th column of Table 4.

A look at the trend of supplying the electrical, thermal, and cooling loads and the comparison of the results reveal the favorable impact of the use of renewable energy resources (WT, solar cells, etc.) on the operation of the MEG. The cost of electricity purchase is $2222.440, the cost of gas purchase is $16,576.632, and the cost of carbon emission penalty is $494.246. The application of Scenario II will incur an operational cost of $19,293.317.

Scenario III

In this section, the electrical, thermal, and cooling storages are included in the MEG to improve the operational costs by optimum charging and discharging. The output electrical power of the power transformer is shown in the fourth column of Table 4. The improved chance of electricity sale is evident in the figures of this column.

The other electrical power supplier of the MEG is the CCHP unit whose electrical power delivery is presented in the seventh column of Table 4. Due to higher costs of electricity at hours 08:00–22:00, the gas turbine unit is ordered to generate as much as possible in order to minimize the operational costs. The unsupplied surplus load at these hours is purchased from the grid. The figures pertaining to the supply of thermal load by the invaluable gas boiler unit are shown in the 10th column of Table 4. This unit has the main responsibility for the heat supply due to its excellent efficiency. The unmet surplus thermal power is supplied by the CCHP unit when it is operating. The 13th column of Table 4 presents the input electrical power of the compression chiller. These figures show that the cooling load is optimally supplied by the compression chiller. This chiller is capable of supplying cooling power with a certain coefficient of performance because of its high efficiency.

It is better to present the modes of storages as well as their charging and discharging status (electrical, thermal, and cooling storages) and to examine them individually. This begins with the electrical storage and shows its energy at each hour as well as its charging and discharging status. The discharged and charged electrical power and the electrical energy existing in the electrical storage are depicted in Figure 5 so as to clarify how the operator of the MEG operates the storages.

Figure 5.

The energy stored in the electrical storage and the discharged and charged power at each hour including E_es (kW h), P_es,d (kW), and P_es,c (kW).

It is evident that the electrical storage is charged at hours when there is surplus electricity or the electricity is cheap, but it delivers its electricity to the MEG at hours when the electricity is expensive so as to both minimize the costs and improve the system efficiency remarkably.

The charged and discharged thermal power of the thermal storage, as well as the amount of stored energy and the storage mode, are illustrated in Figure 6 which demonstrates its usefulness to the grid operation.

Figure 6.

The energy stored in thermal storage and the discharged and charged power at each hour including E_hs (kW h), P_hs,d (kW), and P_hs,c (kW).

Figure 6 displays that at peak hours, the heat is discharged to the MEG. Here, this storage stores the heat energy at hours when the generated heat exceeds the load supply level according to the thermal load profile. Then, it discharges this heat into the grid when the MEG fails to meet the load power demand.

The abovementioned trend is repeated for the cooling storage. Figure 7 shows the discharged and charged cooling power and the cooling energy stored in the cooling storage.

Figure 7.

The energy stored in the cooling storage and the amount of discharged and charged power at each hour including E_cs (kW h), P_cs,d (kW), and P_cs,c (kW).

If Scenario III is exploited, the operational cost will be $17,184.632. In this scenario, not only no cost is paid to the purchase of electricity from the public grid but it also entails an income of $554.812 for the sale of electricity to the public grid. The cost of gas purchase is $17,268.124 and carbon emission penalty is $471.320 in this scenario. This scenario uses thermal, cooling, and electrical ESs, but maintenance and personnel costs are ignored. These costs are recommended to be addressed in future research.

Discussion

The integrated representation of input electrical power of the MEG under Scenarios I, II, and III in Figure 8 reveals that the amount of electricity purchased from the grid is decreased and the amount of electricity sold to the grid is increased remarkably when the renewable resources are added to the combined system of electricity, heat, and cooling generation in Scenario II as compared to Scenario I and when the ES system is added in Scenario III as compared to Scenario II. This enables reducing the total operational cost of the MEG.

Figure 8.

Input electrical power of the MEG during the day under three studied scenarios, P_in (kW).

According to Figure 9, gas turbine is capable of generating more electricity in Scenario III than in the other scenarios during day because power can be stored in and retrieved from electric storage system. The operational cost is improved remarkably by purchasing less electricity and selling more electricity to the power grid, specifically at hours when electricity cost is higher as per the tariffs shown in Figure 3.

Figure 9.

Output electrical power of the gas turbine, during the day under three studied scenarios, P_gt (kW).

It is observed in Figure 10 that at hours 08:00–22:00, the gas boiler is operated in none of the scenarios due to the supply of heat load by the gas turbine, which results in saving of fuel consumption. Also, the application of the thermal storage in Scenario III enabled lower use of the gas boiler as compared to Scenarios I and II.

Figure 10.

Output thermal power of the gas boiler, during the day under three studied scenarios, H_gh (kW).

According to Figure 11 that shows the input electrical power of the compression chiller, the input electrical power does not differ between Scenarios I and II, but it is much lower in Scenario III than Scenario II at 12:00 and 19:00 due to the use of the cooling storage system. It is known that the electricity price is maximum at these hours (Figure 3). It is evident in Figure 7 that the power discharging rate of the cooling storage maximizes at these hours, i.e. 12:00 and 19:00, implying that the operator uses the cooling storage at these hours to supply the cooling load due to the high price of the electricity and its high operational costs so as to operate the electrical chiller to a lesser extent and to reduce electricity consumption at peak hours. Therefore, the gas boiler is not operated when the heat load is supplied by the CCHP system, which results in saving of fuel consumption. Also, the application of the thermal storage enables less use of the gas boiler as compared to its nonuse.

Figure 11.

Input electrical power of the compression chiller in a typical day under three studied scenarios, P_ec (kW).

Therefore, the input electrical power of the compression chiller is reduced due to the exploitation of the cooling storage system.

A look at the combined results of the simulation of the MEG under three different scenarios demonstrates the improvement in gas and electricity costs and carbon emission cost and the reduction of operational cost of the whole system in the three scenarios (Table 5).

Table 5.

Costs of operation ($).

Costs	Scenario I	Scenario II	Scenario III
Cost of purchased electricity	4838.040	2222.440	−554.812
Cost of purchased gas	16,576.632	16,576.632	17,268.124
Cost of carbon emission	581.538	494.246	471.320
Total cost of operation	21,996.210	19,293.317	17,184.632

Conclusion

The present paper defined an energy hub-based MEG. All hub sub-systems were described and their individual effects were studied on the MEG. Furthermore, an optimum operation model was defined and implemented for the MEG with the objective of minimizing the daily costs, for which three scenarios were defined. Scenario I was related to the operation of a simple MEG with different energy carriers. It was implemented to minimize the operational cost and to maximize system efficiency. Scenario II addressed the introduction of renewable resources to the MEG presented in Scenario I. The results showed the decrease in costs and pollution emission at 2702.893 units and the increase in system efficiency as compared to Scenario I. In Scenario III, ESs were applied including electrical, thermal, and cooling storages. The charging and discharging of these storages at suitable times allowed reducing the operational cost by 2108.685 units and increasing system efficiency versus Scenario II. These were its major contributions. The application of a thermal, cooling, and electrical storage unit, PV, and wind modules during the operational period made the system more economical because of avoiding electricity and fuel purchase. For the system with the lowest operational cost (Scenario III), a cost of emission saving of 18.9% was obtained while total cost saving of 21.8% was accomplished in Scenario III as compared to Scenario I during the lifetime of the project. When compared to Scenario I as the reference scenario, in Scenario III although the cost of gas purchase was increased by 4.1%, the cost of electricity purchase from the grid was not only cut completely but also the system had an income of 554.812 units from the sale of electricity to the public grid.

The analysis of these three scenarios clearly shows the positive impact of each element, ESs and renewable energy resources added to the combined system of electricity, heat and cooling generation on the system operation. The results show that application of such systems in a residential building can be very attractive from an economic, energy, and environmental perspective. The performance of the polygeneration system is the highest in Scenario III and the lowest in Scenario I. This model can be generalized to all days of the year and all building types including larger and more complex systems. Future studies can focus on the planning for the collaboration of combined electricity and heat generation units with a demand response program in the presence of virtual power plant structure, the use of different combined micro-grids of electricity, heat, and cooling considering the transmission constraints, the execution of other control strategies of power in disperse production units, or on the dynamic investigation of the MEG. Also, future works can address the inclusion of further constraints to the operation of the system such as the inclusion of maintenance and personnel costs to operate the micro-grid equipment, the investigation of different ESs and various power plants such as solar plant, and the investigation of this subject matter for longer time intervals, e.g. one year, and more diverse climatic conditions.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Amirhossein Eshraghi received his Ph.D. degree in Electrical Energy System Engineering (Energy Modeling), Tehran, Iran, from Oct 2014 to Sep 2018. He has published several international papers. His research interests includes cogeneration cycles, smart grid and electricity market, energy storage, energy management, power system economics, and optimization theory and its applications.

Gholamreza Salehi received his Ph.D. in Mechanical Engineering - Energy System Engineering from the KNT University of Technology, Tehran, Iran in 2012. At present, he is working on the following research topics: the field of energy and energy conversion, renewable energy systems, and heat recovery and cogeneration cycles. He has contributed to 20 international papers and has written several books in Persian. He is also actively involved in energy-related industrial projects in oil and power industries and buildings.

Seyedmohammadreza Heibati had the research scholar postdoctoral position at the University of Ottawa, Canada, in 2017, and he received his second Ph.D. in Construction Engineering, Applied Research Program at Département de génie de la construction, École de Technologie supérieure, Université du Québec, Montreal, Quebec, Canada in 2020. His research interests includes building energy efficiency, building energy modeling and simulation, building energy optimization, zero solar energy house and HVAC systems designing.

Kamran Lari received his Ph.D. degree in 2004. He is presently the Head of the Department of Energy Systems Engineering, while he was the Dean of the Faculty of Marine Science and Technology, Islamic Azad University, North Tehran Branch, Tehran, Iran. Currently, he is an associate professor. His research interests are marine renewable energy, coastal engineering, and marine geodesy.

References

Di Somma

Graditi

Siano

Optimal bidding strategy for a DER aggregator in the day-ahead market in the presence of demand flexibility. IEEE Trans Ind Electron 2019; 66: 1509–1519.

Duić

Rosen

MA.

Sustainable development of energy systems. Energy Convers Manag 2014; 87: 1057–1062.

Chen

Gong

Wan

, et al. Performance study of a dual power source residential CCHP system based on PEMFC and PTSC. Energy Convers Manag 2016; 119: 163–176.

Shi

Huang

XH.

Sensitivity analysis of energy demands on performance of CCHP system. Energy Convers Manag 2008; 49: 3491–3497.

Krause

Andersson

Frohlich

, et al. Multiple-energy carriers: modeling of production, delivery, and consumption. Proc IEEE 2011; 99: 15–27.

Mancarella

MES (multi-energy systems): an overview of concepts and evaluation models. Energy 2014; 65: 1–17.

Quelhas

Gil

McCalley

, et al. A multiperiod generalized network flow model of the US integrated energy system: part I – model description. IEEE Trans Power Syst 2007; 22: 829.

Geidl

Koeppel

Favre-Perrod

, et al. Energy hubs for the future. IEEE Power Energy Mag 2007; 5: 24–30.

Geidl

Andersson

Optimal coupling of energy infrastructures. In: Power Tech, 2007 IEEE, Lausanne, pp.1398–1403. New York: IEEE.

10.

Daryani N and Tohidi S. Economic dispatch of multi-carrier energy systems considering intermittent resources. Energy & Environment 2019; 30(2): 341–362. https://doi.org/10.1177/0958305X18790959

11.

Bahrami

Safe

A financial approach to evaluate an optimized combined cooling, heat and power system. Energy Power Eng 2013; 5: 352.

12.

Rastegar

Fotuhi-Firuzabad

Load management in a residential energy hub with renewable distributed energy resources. Energy Build 2015; 107: 234–242.

13.

Bozchalui

Hashmi

Hassen

, et al. Optimal operation of residential energy hubs in smart grids. IEEE Trans Smart Grid 2012; 3: 1755–1766.

14.

Rastegar

Fotuhi-Firuzabad

Lehtonen

Home load management in a residential energy hub. Electr Power Syst Res 2015; 119: 322–328.

15.

El-Zonkoly

Abdelaziz

Eladl

AM.

Double-layer firefly algorithm for simultaneous optimal sizing and operation of energy hubs. Electr Power Compon Syst 2017; 45: 1846–1857.

16.

Wang

Fan

Han

, et al. Improved interval optimization method based on differential evolution for microgrid economic dispatch. Electr Power Compon Syst 2015; 43: 1882–1890.

17.

Ippolito

Telaretti

Zizzo

, et al. A new device for the control and the connection to the grid of combined RES-based generators and electric storage systems. In: 2013 International Conference on Clean Electrical Power (ICCEP), 2013, pp.262–267. New York: IEEE.

18.

Nasiri

Solanki

, et al. Microgrids: architectures, controls, protection, and demonstration. Electr Power Compon Syst 2015; 43: 1453–1465.

19.

Emodi NV, Chaiechi T and Alam Beg AR. The impact of climate change on electricity demand in Australia. Energy & Environment 2018; 29(7): 1263–1297.

20.

Vatankhah Barenji R, Ghadiri Nejad M, and Asghari I. Optimally sized design of a wind/photovoltaic/fuel cell off-grid hybrid energy system by modified-gray wolf optimization algorithm. Energy & Environment 2018; 29(6): 1053–1070.

21.

Graditi

Di Silvestre

Gallea

, et al. Heuristic-based shiftable loads optimal management in smart micro-grids. IEEE Trans Ind Inf 2015; 11: 271–280.

22.

Mahlia

Saktisahdan

Jannifar

, et al. A review of available methods and development on energy storage; technology update. Renew Sustain Energy Rev 2014; 33: 532–545.

23.

Eshraghi

Salehi

Heibati

, et al. An enhanced operation model for energy storage system of a typical combined cool, heat and power based on demand response program: the application of mixed integer linear programming. Build Serv Eng Res Technol 2019; 40: 47–74.

24.

Poullikkas

A comparative overview of large-scale battery systems for electricity storage. Renew Sustain Energy Rev 2013; 27: 778–788.

25.

Graditi

Ippolito

Telaretti

, et al. Technical and economical assessment of distributed electrochemical storages for load shifting applications: an Italian case study. Renew Sustain Energy Rev 2016; 57: 515–523.

26.

Marzband

Javadi

Pourmousavi

, et al. An advanced retail electricity market for active distribution systems and home microgrid interoperability based on game theory. Electr Power Syst Res 2018; 157: 187–199.

27.

El-Zonkoly

Optimal scheduling of observable controlled islands in presence of energy hubs. Electr Power Syst Res 2017; 142: 141–152.

28.

Warren P. Demand-side policy: Global evidence base and implementation patterns. Energy & Environment 2018; 29(5): 706–731.

29.

Chen

Luo

Yang

An optimization method for design and operation of combined cooling, heating, and power systems toward a smart grid. Sci Technol Built Environ 2016; 22: 766–782.

30.

Milcarek

Ahn

Zhang

Review and analysis of fuel cell-based, micro-cogeneration for residential applications: current state and future opportunities. Sci Technol Built Environ 2017; 23: 1224–1243.

31.

Elsied

Oukaour

Gualous

, et al. Optimal economic and environment operation of micro-grid power systems. Energy Convers Manag 2016; 122: 182–194.

32.

Chen

H-J

Wang

Chen

S-L.

Optimization of an ice-storage air conditioning system using dynamic programming method. Appl Therm Eng 2005; 25: 461–472.

33.

Lee

W-S

Chen

T-H.

Optimization for ice-storage air-conditioning system using particle swarm algorithm. Appl Energy 2009; 86: 1589–1595.

34.

Guo

Liu

Cai

, et al. A two-stage optimal planning and design method for combined cooling, heat and power microgrid system. Energy Convers Manag 2013; 74: 433–445.

35.

Ghaem Sigarchian S, Malmquist A, Martin V. Optimal planning and design method for complex polygeneration systems : A case study for a residential building in Italy. Energy Conversion and Management 2017; pp.1–35. Available at: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-228077

36.

Hall

Buckley

AR.

A review of energy systems models in the UK: prevalent usage and categorisation. Appl Energy 2016; 169: 607–628.

37.

Lopes

Moreira

Madureira

Defining control strategies for microgrids islanded operation. IEEE Trans Power Syst 2006; 21: 916–924.

38.

Sechilariu

Wang

Locment

Supervision control for optimal energy cost management in DC microgrid: design and simulation. Int J Electr Power Energy Syst 2014; 58: 140–149.

39.

Mesquita FGG. Design Optimization of Stand-Alone Hybrid Energy Systems: A Dissertation submitted under the scope of Mestrado Integrado em Engenharia Electrotécnica e de Computadores Major Energia. (Supervisor: Prof. Doutor Cláudio Domingos Martins Monteiro) Fevereiro de 2010.

40.

Brahman

Honarmand

Jadid

Optimal electrical and thermal energy management of a residential energy hub, integrating demand response and energy storage system. Energy Build 2015; 90: 65–75.

41.

Jagdev

Browne

Jordan

Verification and validation issues in manufacturing models. Comput Ind 1995; 25: 331–353.

42.

Sargent

RG.

Verification, validation, and accreditation: verification, validation, and accreditation of simulation models. In: Proceedings of the 32nd conference on winter simulation, 2000, pp.50–59. San Diego, CA: Society for Computer Simulation International.

43.

Kleijnen

JP.

An overview of the design and analysis of simulation experiments for sensitivity analysis. Eur J Oper Res 2005; 164: 287–300.

44.

Murray-Smith

DJ.

Testing and validation of computer simulation models. Berlin: Springer, 2015.

45.

Sargent

RG.

Verification and validation of simulation models. J Simulat 2013; 7: 12–24.

46.

Brander

Sood

Wylie

, et al. Technical paper| electricity-specific emission factors for grid electricity. Ecometrica, Emissionfactors.com, 2011.

47.

Eshraghi

Salehi

Heibati

Lari

An assessment of the effect of different energy storage technologies on solar power generators for different power sale scenarios: the case of iran, Sustainable. Energy, Technologies and Assessments 2019; 34: 62–67.