- Email: [email protected]
T driven CCHP microgrid
Accepted Manuscript Research Paper Multi-objective Optimal Component Sizing of a Hybrid ICE+PV/T Driven CCHP Microgrid Hossein Yousefi, Mohammad Hasan Ghodusinejad, Alibakhsh Kasaeian PII: DOI: Reference:
S1359-4311(16)32003-8 http://dx.doi.org/10.1016/j.applthermaleng.2017.05.017 ATE 10328
To appear in:
Applied Thermal Engineering
Received Date: Revised Date: Accepted Date:
28 September 2016 30 April 2017 4 May 2017
Please cite this article as: H. Yousefi, M.H. Ghodusinejad, A. Kasaeian, Multi-objective Optimal Component Sizing of a Hybrid ICE+PV/T Driven CCHP Microgrid, Applied Thermal Engineering (2017), doi: http://dx.doi.org/ 10.1016/j.applthermaleng.2017.05.017
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Multi-objective Optimal Component Sizing of a Hybrid ICE+PV/T Driven CCHP Microgrid Hossein Yousefi a*, Mohammad Hasan Ghodusinejad a, Alibakhsh Kasaeian a a
Department of Renewable Energies and Environment, Faculty of New Sciences and Technologies, University of Tehran Address: North Karghar St., Faculty of New Sciences and Technologies, Tehran, Iran Tel: +9821 8609 3293 ; Email: [email protected]
Abstract This paper presents a multi-objective optimization approach in order to find the optimal capacities of a CCHP microgrid, to meet the power, heating and cooling loads of a large office building located in Tehran. The CCHP microgrid is modelled in two states: a) an Internal Combustion Engine (ICE) driven system; b) a solar assisted CCHP system including ICE and Photovoltaic/Thermal (PV/T) panels. The model is implemented in MATLAB and optimized via NSGA-II. The optimization is carried out to simultaneously minimize the Net Present Cost (NPC) and maximize the Primary Energy Saving (PES) of the system. Carbon Emission Reduction Ratio (CER) is also calculated to assess the environmental performance of the microgrid. The results revealed that, despite a slight increase in the average value of NPC, the CCHP system integrated with solar energy have a considerably better performance in energy saving and emissions than the ICE driven system. Keywords: CCHP microgrid, Internal combustion engine, PV/T panels, combined heat and power, Net present cost, Primary energy saving
1. Introduction Energy, has been a key input to the industrial, commercial and residential structures of today world. The urgent need for meeting the increasing energy demand and also, global concerns about environmental issues, has led the energy policies to the way that energy systems should be technically more efficient and environmentally cleaner. Microgrid is a key concept that can ease achieving these energy systems designing goals. Microgrids combine various energy sources in the best way to meet local energy demands. They can operate in both grid-connected and off-grid modes. The concept of microgrid was first introduced by the Consortium for Electric Reliability Technology Solutions, CERTS [1]. This consortium defines microgrids as clusters of generators, including heat recovery, storage, and loads, which are operated as single controllable entities [2]. Developing combined cooling, heating and power (CCHP) microgrids has attracted lots of attention. In comparison to conventional CCHP systems, CCHP microgrid is more efficient. It not only can meet cooling, heating and power demands, but also can interact with the grid, providing different services such as demand response programs and peak-shaving services [3-5]. On the other hand, to produce useful energy according to sustainable development policies, some issues should be considered including less or zero carbon emission, higher reliability of the energy networks, increased energy security and reduced costs. There have been different ways to attain the afore-mentioned goals. The three main concept suggested for sustainable development in energy systems are distributed generation (DG), cogeneration systems and renewable energy utilization. These are separate approaches and each can provide several advantages. Developing CCHP microgrids, all these three concept can be combined into one solution. A CCHP microgrid, can act as a DG unit and shares power with the main grid and can also improve the use renewable energy
sources, locally. Additionally, capturing power and heat simultaneously, increases the energy efficiency. Buildings, consuming 40% of global energy consumption, play an important role in energy market [6] and therefore, large buildings are good potentials for CCHP microgrids. 1.1. Previous Researches Optimal design of microgrids has been the subject of many researches within the recent years. Generally, researches in this area can be classified in three categories; sizing optimization, siting optimization and scheduling and economic dispatch of microgrid [2]. Therefore, there is a rich literature in analyzing CHP and CCHP systems. Numerous papers have analyzed conventional CCHP systems comprising of conventional CHP technologies [7-17], while, some other papers, have studied the performance of CCHP systems with hybrid CHP technologies or unique CHP components. Moghadam et al. [18] conducted a sizing optimization of a micro-CHP system based on a solar dish Stirling. The system was analyzed based on 3E analysis. The 3E analysis evaluated primary energy saving analysis, carbon dioxide emission reduction and payback period for return of investment. Three different scenarios was considered and it was found that the system can provide a good potential of energy saving and carbon emission reduction. Sanaye and Sarrafi [19] presented an energy, exergy and economic optimization of a solar-based CCHP system. The solar system was equipped with photovoltaic (PV), concentrated photovoltaic/thermal (CPVT), and evacuated tube (ET) collectors. Wouters et al. [20] presented a MILP model for optimal design of a CCHP multi-microgrid which contains different power, heating and cooling generation units and energy storage components. The model was applied to a South Australian case study and several scenarios were considered. Mohamed et al. [21] investigated the feasibility of CHP and CCHP systems, along with conventional heating and cooling systems of building.
Askari et al. [22] conducted the optimization of a hybrid micro-CCHP system comprised of solar PV, flat plate collector and natural gas generators, for a 5 story residential building. Jabari et al. [23] proposed a novel energy and exergy methodology for designing and optimal short-term planning of a heat pumpbased CCHP system driven by a solar dish Stirling heat engine. This CCHP system provides a good potential for reduction in CO 2 and NOx emission and an increase in the economic saving in fuel consumption. Evins [24] carried out a multi-level optimization for building design and energy system sizing and operation. The proposed system was comprised of several components including PV, solar collector, fuel cell, heat pump and CCHP-related components. Yang et al. [25] proposed a paradigm for energy hubs that contain CCHP components, renewable energy and energy storage. An operation model is suggested for regional multi-energy prosumers (RMEP) whose energy demands are served by interconnected energy hubs. Yang et al. [26] presented a study focusing on the optimal design of distributed energy resource systems with consideration of the uncertainties. A two-stage stochastic programming model was developed and applied to a hospital building in China to meet its cooling, heating and power loads. Results show that the uncertainties in load demands have a significant effect on the optimal system design, whereas the uncertainties in energy prices and renewable energy intensity have almost no effect.
1.2. This Work In this paper, a hybrid CCHP system is modeled and optimal component sizes are determined via a multi-objective optimization approach. The system is comprised of two types of CHP technologies; fossil fuel fired Internal Combustion Engine (ICE) and solar photovoltaic/thermal (PV/T) panels. There are some papers in previous researches that consider the integration of solar energy components into CCHP systems. These all have analyzed the
integration of photovoltaic panels and/or solar thermal collectors into conventional CHP components. Clearly, this would not lead to a hybrid CHP mechanism. In contrast, in this paper, the system would be a hybrid CCHP microgrid based on renewable and non-renewable CHP components. To assess the performance of the system, optimization is conducted in two scenarios. First, the system is modeled based on ICE, and then in second scenario, the system is modeled with ICE and PV/T. The total description of the system is presented in section 2. Mathematical models, related to both scenarios are covered in section 3. Optimization results are presented in section 4. Finally, conclusion of the paper is presented in section 5.
2. System Description The CCHP microgrid is modeled to meet the electrical, heating and cooling loads of an office building. The case study is one of the headquarters of Tehran Municipality which is a 7 story office building. The electrical load profile of the building is derived from previous years’ data of the building consumption and is illustrated in Fig. 1. The power is consumed in order to indoor and outdoor lighting, office facilities, HVAC components use, etc. The building is simulated in Hourly Analysis Program (HAP) [27] to create cooling and heating load profiles. The cooling/heating demand of the building is depicted in 2. As all simulations are conducted in time steps of 1 hour, load profiles are in hourly format. Currently the electricity demand of the building is totally met by the grid. A natural gas fired boiler and a direct-fired absorption chiller provide heating and cooling for the building, respectively. To model PV/T, solar irradiance data is needed. The hourly solar data is derived from meteorological TM2 file, the TMY library of TRNSYS 16 software [28]. Fig. 3 illustrates the Plane of Array irradiance.
As stated previously, the CCHP microgrid is modeled in two scenarios. Fig. 4 illustrates the schematic diagram of the microgrid components in both scenarios. In scenario 1, the ICE produces heat and electricity to feed the building loads and the boiler and grid act as heat and electricity backups, respectively. The cooling load of the building is whether met by indirect fired absorption chiller (AC) or electric chiller (EC). The total structure in scenario 2 is kept the same as the first scenario, except PV/T panel is added to the system to combine a hybrid CHP system in cooperation with the ICE. Technical specifications of the microgrid components are presented in Table 1. The data used for PV/T panels is derived from FOTOTHERM Cs panels with a nominal power of 250W. There are two different approaches for efficiencies and COPs of components in the literature. In some papers, these are considered to be fixed, but in some others, it is changed as a function of ratio of operation capacity to nominal capacity. In this work, the ICE operates in full load mode, so that the efficiency is as its nominal value. But for the chillers and the boiler, it is supposed not to be changed during the operation. 3. Mathematical Models In this section, the mathematical models of the CCHP microgrid are presented. The model is totally implemented in MATLAB. Two blocks of codes were created for modeling and optimization tasks. Fig. 5 depicts the interaction of modeling and optimization processes. 3.1. Genetic Algorithm Genetic Algorithm (GA) is the most known meta-heuristic optimization algorithm which is based on the survival of the strongest and fittest creature. The general procedure is based on four basic operators; initial population creation, selection, crossover and mutation. Several modified forms of GA have
been proposed by researchers to promote the performance of the algorithm. Deb et al. [29] proposed Non-dominated Sorting Genetic Algorithm II (NSGA-II) in 2002. The special feature of NSGA-II is the approach of ranking the population members; the solutions are classified to different groups with different domination levels and a rank is specified to each level. NSGA-II is a fast algorithm which can suitably solve sizing and scheduling problems of microgrids. In this paper, NSGA-II is utilized to solve the optimization problem. Table 2 provides the characteristics of the algorithm used in this paper. 3.2. Microgrid Modeling The performance of all components as well as cooperation of them within the CCHP microgrid have been modeled in two scenarios. The decision variables in the first scenario are nominal capacities of ICE and AC (in kW) while nominal capacities of boiler and EC are dependent variables and can be found during the problem solution. In the second scenario, the number PV/T panels would be added to decision variables. The range of these decision variables are presented in Table 3. The model in both scenarios is presented in the following.
3.2.1. Internal combustion engine To achieve the best operation efficiency of the ICE, it is considered that the engine would operate full time in rated operating power. Besides, the operation mode of the ICE is set to be full load in order to export power to the grid. This can make the system as a distributed generation unit. In cases the generated power by the ICE is more than the internal load, excess electricity will be sold to the grid. Therefore, the power from the ICE in each time step can be found as: PICE (t ) PICE ,nom
(1)
The natural gas consumed by the ICE is calculated by: Q f (t )
PICE (t )
(2)
ICE ,e
Finally, generated heat by the ICE can be found: Q ICE (t ) Qf (t ) ICE ,t
(3)
3.2.2. PV/T Panels Some papers, in the literature, have studied the integration of solar components into the CCHP system [30-34]. The area of each panel considered in this paper, as in the company catalog, is about 1.6m2. Therefore, the total useful area of the solar system is found: APVT 1.6N PVT
(4)
Generated power of the PV/T is a function of the total area as well as the panel efficiency and solar irradiance and can be written as: PPVT (t ) APVT PVT (t )G PoA (t )
(5)
In PV/T panels, circulating fluid, whether water or air, have a positive effect and act as a coolant to reduce the cell temperature. But in this paper, this effect is neglected and it is considered that the increase in cell temperature will cause a decrease in cell efficiency. Therefore, PV/T efficiency in each time step is calculated as below [35]: PVT (t ) ref 1 Tcell (t ) T ref
where, T ref is the reference temperature equal to 25°C. To calculate the cell temperature, many relations have been developed. In this paper, the following empirical relation, which can be used for poly-crystalline silicon cells, is utilized for the calculation of the cell temperature [35]:
(6)
Tc (t ) 30 0.0175 G PoA (t ) 300 1.14 T a (t ) 25
(7)
where, the ambient temperature is derived from the same source as solar irradiance. Calculation of generated heat by PV/T panels is slightly different from conventional flat plate collectors. The PV/T generated heat can be determined as [35, 36]: Q PVT (t ) APVT G PoA (t )( ) 1 PVT (t ) U L T cell (t ) T a (t )
(8)
where, ( ) is considered 0.9 as in [37]. 3.2.3. Cooling balance The cooling task of the building is done by both AC and EC: Q EC (t ) Q AC (t ) QC (t )
(9)
It is considered that, in cooling task, AC is prior to EC. If the cooling load is less than the nominal capacity of the AC, then it is totally met by the AC and the EC would not operate: Q AC (t ) QC (t )
(10)
Q EC (t ) 0
(11)
In cases that the AC capacity is not sufficient to meet the cooling load, it will operate in nominal power and the EC would cover the deficiency: Q AC (t ) Q AC ,nom
(12)
Q EC (t ) QC (t ) Q AC (t )
(13)
The nominal capacity of the EC can be found after the simulation is done for the whole year:
Q EC ,nom max Q EC
(14)
3.2.4. Heating balance The heat demand of the building can be written as equation (15). It is considered that the COP of the AC remains constant in different loads. Qdem (t ) Q H (t )
Q AC (t ) COPAC
(15)
In the first scenario, the heating demand is met by the ICE and the boiler: Q ICE (t ) Q B (t ) Qdem (t )
(16)
If the generated heat by the ICE is sufficient to meet the heating demand, then the boiler will not operate. Besides, if the generated heat is more than demand, the excess heat is exhausted. In cases the heating demand is more than the generated heat, boiler will cover the deficiency. For the second scenario, besides the ICE and the boiler, PV/T can also generate heat for the building: Q ICE (t ) Q PVT (t ) Q B (t ) Qdem (t )
(17)
As in the first scenario, if the generated heat by the ICE and the PV/T is as much as the heating demand, then the boiler will not operate. Here again, the excess generated heat, if exists, will be exhausted. In cases that the generated heat is not sufficient, boiler will operate as a backup to meet the heating demand. The nominal capacity of the boiler in both scenarios can be fined as follows: Q B,nom max Q B
(18)
For each time step, the consumed fuel by the boiler can be calculated using the boiler efficiency:
Q f ,B (t )
Q B (t )
B
(19)
The fuel consuming components of the microgrid are ICE and boiler. Therefore, the total fuel consumed by the microgrid is found: Qf ,Total Qf Qf ,B
(20)
3.2.5. Electricity balance In both scenarios, it is considered that the priority is to feed the building electrical load by the ICE and PV/T generated power. Therefore, local generated power is firstly fed to the internal loads and the surplus power, if exists, will be sold to the grid. This will ensure the self-sufficiency of the building in providing its demand. The power demand of the building consists of the electrical load as well as the power needed to run the EC and is written as below: Pdem (t ) PL (t )
Q EC (t ) COPEC
(21)
The electricity balance, in the first scenario, is as equation (22). In each time step, the generated power by the ICE is comprised of two fractions; the power sold to the grid and the power provided to the building load. This is stated in equation (23). If the generated power by the ICE is enough to meet the load, then no power is bought from the grid and excess generated power, if exists, is sold to the grid. If the ICE power is not as much as to meet the building load, then additional power is bought from the grid. PICE L (t ) Pgrid (t ) Pdem (t )
(22)
PICE (t ) PICE L (t ) PICE grid (t )
(23)
In the second scenario, the electricity generated by the PV/T is also considered to provide power for the building. Therefore, the electricity balance in this scenario is as equation (24). As well as ICE, the PV/T generated power is comprised of two parts; the power sold to the grid and the power provided to the building load. This is stated in equation (25). PICE L (t ) PPVT L (t ) Pgrid (t ) Pdem (t )
(24)
PPVT (t ) PPVT L (t ) PPVT grid (t )
(25)
The priority, in this scenario is to meet the load by the ICE power. This is considered since the Feed-in-Tariff (FiT) for solar power is too more than the one for ICE power and selling surplus solar power is economically more beneficial. Therefore, in each time step, if the generated power by the ICE is enough to meet the load, then no power is bought from the grid and the total PV/T power, as well as potential excess ICE power, is sold to the grid. In cases of insufficiency of the ICE power, PV/T power is fed to the building. If the power generated by both ICE and PV/T is not enough for the load, the deficiency is provided by the grid import. 3.2.6. Objective function As this research is a multi-objective optimization problem, two objective functions is considered to be optimized simultaneously by NSGA-II. The first objective function is Net Present Cost (NPC) of the microgrid. The goal is to achieve microgrid components capacity with minimized NPC. The NPC for this research is formulated as: 1 NPC C cap ,i C rep ,i K i C o &m ,i Pnom ,i CRF (r , L ) i C ng NG C grid Pbought C ICE ,Sell PICE ,Sell C PVT ,Sell
1 PPVT ,Sell CRF (r , L )
(26)
where it contains two cost parts; the first is capacity related costs and the second is the costs related to the energy transactions between the microgrid and utility, i.e. electricity and natural gas import and export. Obviously, the term related to PV/T power export is not considered in the first scenario. K i and CRF (r , L ) are single payment present worth factor and capital recovery factor, respectively, and can be calculated as follows [32]: yi
1 nLi n 1 (1 r )
K i (r , Li , y i )
r (1 r)L CRF (r , L ) (1 r)L 1
(27)
(28)
where, L i is the lifetime of component i , y i is the number of replacements of component i during the project lifetime, and r is the real interest rate and is given by: r
IN IF 1 IF
(29)
Table 4 provides details about the economic parameters considered in this paper. The unit costs and lifetime of each component are also presented in Table 5. Based on Table 5, and the lifetime of the project, it is clear that the ICE needs replacement twice and the EC and AC once, while there is no need for replacement of the boiler and PV/T. The second objective function is to maximize the Primary Energy Saving (PES) of the microgrid. PES is considered to assess the energy saving performance of the CCHP microgrid comparing current energy system performance of the building. Therefore, PES is calculated as: Q PES 1 CCHP Q ref
100
(30)
in which, QCCHP and Q ref are the total primary energy consumed by the CCHP microgrid and current energy system of the building, respectively, and are given by: 8760 8760
QCCHP Q f ,Total (t )
P t 1
t 1
8760
Q ref
PL (t ) t 1
grid
grid
t 1
B
(31)
grid
8760
Q H (t )
(t )
8760
Q t 1
C
(t )
(32)
COPAC ,ref
where, COPAC ,ref is coefficient of performance of the current direct-fired AC of the building and is considered as 1. 3.2.7. Environmental performance index In order to assess the environmental performance of the CCHP microgrid, Carbon Emission Reduction Ratio (CER) is defined. CER is used to compare Carbon emissions of the microgrid and the current system. Thus, it is calculated by: CE CER 1 CCHP CE ref
100
(33)
where, CE CCHP and CE ref are Carbon emissions of the CCHP microgrid and current energy system of the building, respectively, and are given by: 8760
8760
t 1
t 1
CE CCHP EFgrid Pgrid (t ) EFng Q f ,Total (t )
CE ref EFgrid
8760 8760 Q H (t ) QC (t ) 8760 PL (t ) EFng t 1 t 1 COPAC ,ref B t 1
(34)
(35)
EFgrid and EFng are the emission factors of grid electricity and natural gas
consumption, respectively, and can be found on Table 6.
4. Results and discussion The model, in both scenarios, is applied to the case study building mentioned in section 2. Two sets of solutions for components sizes were obtained. Fig. 6 illustrates the Pareto Fronts obtained for both scenarios. As can be seen from the Fig. 6.a, solutions experience slight changes in PES and NPC, in scenario 1; where the average values for PES and NPC are approximately 5.063% and 1716442$, respectively. Thus, applying the first scenario system configuration can provide an average 5% annual energy saving. From Fig. 6.b, the diversity of solutions is more than previous scenario; where from about 5 to 9% energy saving is achievable with changes of about 1.72 to 1.86 million dollars in NPC. In this scenario, by adding PV/T to the system configuration, an average PES of about 7.24% can be obtained with average value of 1.78 million dollar for NPC. Component sizes for three different solutions in each scenario, shown in Fig. 6, are provided in Table 7. It can be seen from Table 7 that by transition from scenario 1 to 2, the average value of solutions for nominal power of ICE and AC are increased while there is a decrease in average values of boiler and EC. This is mainly due to the addition of PV/T to the system. PV/T can provide more heat for the building which feed the AC and will result in a decrease in boiler capacity and increase in AC capacity. Excluding number of PV/T panels, the values for other components capacities are not distributed variously. Thus, the main effect on the value of NPC and PES is due to the variation in number of PV/T panels. This variation also effects on the generated heat and power by PV/T panels. Fig. 7 illustrates the generated
power by PV/T panels, for the three solutions stated in Table 7. It is clear that the generated power from the solution with the highest number of panels, i.e. solution 3, is too more than the one the least number of panels, i.e. solution 1. This would be the same as for the generated heat by panels, where is shown in Fig. 8. In scenario 1, power supplying of the building is done by ICE and grid, while in scenario 2, PV/T is also added to provide electricity for the building. Fig. 9 depicts the shares of each component for monthly power consumption of the building, for solution 3. It can be seen for both scenarios that for months June to September, the grid provides more electricity than local power generating units, i.e. ICE and PV/T. But in other months, a considerable share of building power is provided by the ICE and the PV/T. In the first scenario, about 53.22% of annual building load is met by the ICE and the rest by the grid. On the other hand, in scenario 2, about 3.2% and 53.63% of annual load is provided by PV/T and ICE respectively. It is necessary to mention that there can be slight differences in the electrical demand of Fig.9.a and Fig. 9.b. in some months due to the difference in the capacity of EC in each scenario. It was stated that the heating of the building in scenario 1 is met by both ICE and boiler. In scenario 2, generated heat by PV/T panels also is added to the thermal load supply. The share of each component in monthly heat supply for the building heating is illustrated in Fig. 10 (for solution 3). It can be seen that for scenario 1, ICE is dominant to provide heat for heating of the building in all months. Here, on the whole, about 73% of the annual heating is conducted by the ICE which is a considerable share. For scenario 2, PV/T panels also provide remarkable heat for the building. PV/T panels annually provide about 14.5% of the building heating while the share of the ICE is 67.93% and the rest is met by the boiler. Again, here, a considerable portion of heating is provided the recovered heat of CHP components.
From the capacities obtained in Table 7, it is clear that the EC owns a bigger share than the AC to provide cooling for the building. Fig. 11 illustrates the daily cooling provided by the EC and AC for both scenarios (solution 3). From Table 7 and Fig. 11, the share of AC in the second scenario is clearly more than the one in the first scenario. 4.1. Environmental Assessment As stated previously, to assess the environmental performance of the microgrid, CER index is utilized. Therefore, CER is calculated for all solutions in both scenarios. Fig. 12 depicts the values of CER in comparison with PES, for both scenarios. Due to the repeated and close solutions in the first scenario, as well as PES, CER is also calculated as an approximate fixed value. In this scenario, a CER of about 3.5% is achievable; it means that 3.5% annual reduction in Carbon emission, i.e. about 48650 kg/year less CO 2, can be a result of using a CCHP microgrid based on ICE. In scenario 2, the values for CER are more various. From Fig. 12.b, a range of about 3.3% to 7.5% of CER can be achieved, with the application of CCHP microgrid based on ICE and PV/T. Therefore, from about 46000 to more than 103000 kg/year reduction in CO2 emission would be available using different capacities derived in scenario 2 for microgrid design. CER is tightly relative to the value of PES. Therefore, the more energy is saved by the CCHP system, the less Carbon would be released to the environment. Besides, from Table 7 and Fig. 12.b, it can be said that the number PV/T panels has the most effect on the emission of the microgrid. Since, by increasing the number of panels from solution 1 to 3, CER experiences an improvement of more than 120%. 4.2. Comparison of the Results 4.2.1. Comparison of two scenarios
In order to compare the CCHP microgrid in two scenarios, 5 parameters were chosen and a comparison was applied to the results based on the following equation: F
F2 F1 100 F1
(36)
where, F is the value of each parameter and the numbers represent the number of each scenario. Here, scenario 1 was considered as the base state and the results of scenario 2 were compared with scenario 1. Five parameters selected for comparison are NPC, PES, CER, total purchased electricity and total sold electricity. The average value of these parameters for all solutions in both scenarios was calculated and used for comparison. Fig. 13 shows the result of comparison. In average, 4% increase can be seen in NPC as a result of transition from scenario 1 to 2. This is a slight negative change, shown in red in Fig. 13, while other parameters experience considerable improvements. The value of PES and CER is increased by about 43% and 52%, respectively. These improvements can also be seen in the values of imported and exported electricity from/to the grid, as the building purchased 17.31% less electricity while selling 184.14% more electricity to the grid.
4.2.2. Comparison with other works As the proposed system of this paper is unique, two papers, that have the closest scope to this work, were selected in order to compare the results. The environmental and energy saving results are extracted form Li et al. [9] and Rodríguez et al. [31] and presented in Table 8. As can be seen from Table 8, the values determined for the assessment of environmental and energy consumption performance of the systems are in different ranges. This is naturally due to
numerous differences in the problem, including the definition of indices, input data such as loads and irradiation, predefined energy management rules, etc. Meanwhile, the consistent fact in the three works is that, by adding solar components to the CCHP system, the performance of the system in emission and energy consumption improves. 4.3. Sensitivity Analysis This research was conducted in a deterministic approach. But the nature of each energy system contains several uncertainties. In order to assess the impact of some economical parameters on the performance of the CCHP system, a sensitivity analysis is carried out. Six parameters were selected and each was set to 50% higher and lower than the reference value. For each state, the optimization process was applied and repeated three times and the average value of NPC was calculated to analyze the effect of the change on the value of NPC. The parameters of sensitivity analysis are presented in Table 8. Fig. 14 illustrates the results of sensitivity analysis. It can be seen that the grid electricity price has the highest effect on the value of NPC, where a 50% decrease will lead to about 34% decrease in NPC while 50% increase results in about 10% increase in NPC. The second level of influence belongs to the ICE capital investment. A considerable point to mention is the effect of selling price of PV/T power. While it has a little effect on the NPC, it influences inversely and less price of PV/T power leads to more NPC. 5. Conclusion In this paper, a CCHP microgrid is modeled and an optimal sizing problem is solved using NSGA-II. Two different configurations are considered for the CCHP system. The first is a system fully based on fossil fuel, i.e. a nonrenewable CHP component. The second is modeled as a hybrid CCHP
microgrid using a renewable CHP component in addition to the non-renewable one. As the optimization is applied, two sets of solutions are obtained for two scenarios. From the details derived from the solutions, the following points are concluded: Application of CCHP systems in buildings is extremely beneficial as far as the energy saving and environmental objectives are concerned. In this paper, a conventional CCHP system with ICE, offered an average 5% energy saving and 3.5% less emission. Using PV/T panels as the second CHP component in addition to the ICE increased the annual energy savings up to 9.24% and the Carbon emission reduction up to 7.46%. For a sample solution, about 54% of the building electrical load was supplied by ICE in scenario 1. This is approximately summed with about 3.2% solar power in scenario 2, increasing the local power share up to about 57%. This means more than half of the annual electrical demand of the building is met locally which is an acceptable rate of self-sufficiency in electrical power supply. For a sample solution in scenario 1, about 73% of annual heat needed for space heating is provided by the ICE recovered heat. In the second scenario, the share of ICE decreased to 68%, while 14.5% of the demand is met by PV/T and therefore, about 82.5% of the space heating is provided by recovered heat of the ICE and PV/T. Notable points are attained by comparing two scenarios. Adding PV/T panels to the structure of the microgrid caused a 4% increase in the average value of NPC. In contrast with this adverse change, several benefits could be achieved. In solar-assisted CCHP microgrid, energy saving is increased and emissions are decreased considerably. Power transaction with the grid was also improved; the imported power from the
grid decreased by 17.31% while the exported power to the grid increased by 184.14%. In other words, a little increase in NPC, not only offers better energy saving and environmental characteristics, but also improves the power self-sufficiency of the building. As a future extension to this work, some ideas can be applied to the research, including considering other different dispatch strategies for the ICE such as electrical load following or thermal load following, examining the effect of adding electrical and thermal storage systems, considering PV and solar thermal components as well as PV/T panels to conduct a size optimization for these three components, etc.
Nomenclature Acronyms CCHP Combined Cooling, Heating and Power CHP Combined Heat and Power NPC Net Present Cost PES Primary Energy Saving CER Carbon Emission Reduction NSGA-II Non-dominated Sorting Genetic Algorithm II Symbols Electrical power (kW) P Thermal power (kW) Q Efficiency (%) Time step (h) t Area of solar panels (m2) A Number of solar panels N Solar irradiance (W/m2) G Temperature coefficient (power) (%/°C) Temperature (°C) T Heat loss coefficient (W/m2.K) UL
COP C K CRF r L
NG y IN IF CE EF
Coefficient of performance Cost related to the component Single payment present worth factor Capital recovery factor Real interest rate (%) Lifetime (years) Annual natural gas consumption (m3) Number of replacements Interest rate (%) Inflation rate (%) Carbon emission (kg) Emission factor (kg/kWh)
Subscripts ICE PVT B AC EC nom
f e t PoA ref cell
a C H L dem grid cap rep
o &m i ng bought Sell CCHP
Internal combustion engine Photovoltaic/thermal Boiler Absorption Chiller Electric Chiller Nominal capacity Fuel Electrical Thermal Plane of Array Reference value Solar cell Ambient Cooling load Heating load Electrical load Demand Grid Capital cost Replacement cost Operation and maintenance cost Index of components Natural gas Bought power from grid Sold power to grid CCHP system
References [1] Fathima, A. Hina, and K. Palanisamy. "Optimization in microgrids with hybrid energy systems–A review." Renewable and Sustainable Energy Reviews 45 (2015): 431-446.
[2] Gamarra, Carlos, and Josep M. Guerrero. "Computational optimization techniques applied to microgrids planning: a review." Renewable and Sustainable Energy Reviews 48 (2015): 413-424. [3] Gu, Wei, et al. "Modeling, planning and optimal energy management of combined cooling, heating and power microgrid: A review." International Journal of Electrical Power & Energy Systems 54 (2014): 26-37. [4] Dincer, I., and C. Zamfirescu. "Renewable‐energy‐based multigeneration systems." International Journal of Energy Research 36, no. 15 (2012): 1403-1415. [5] Wu, D. W., and R. Z. Wang. "Combined cooling, heating and power: a review." progress in energy and combustion science 32, no. 5 (2006): 459-495. [6] Nejat, Payam, et al. "A global review of energy consumption, CO 2 emissions and policy in the residential sector (with an overview of the top ten CO 2 emitting countries)." Renewable and Sustainable Energy Reviews 43 (2015): 843-862. [7] Ünal, Ali Nadi, İbrahim Ersöz, and Gülgün Kayakutlu. "Operational Optimization in Simple Trigeneration Systems." Applied Thermal Engineering 107 (2016): 175-183. [8] Farahnak, Mehdi, et al. "Optimal sizing of power generation unit capacity in ICE-driven CCHP systems for various residential building sizes." Applied Energy 158 (2015): 203-219. [9] Li, Miao, et al. "Optimal design and operation strategy for integrated evaluation of CCHP (combined cooling heating and power) system." Energy99 (2016): 202-220. [10]
Ascione, Fabrizio, et al. "Multi-stage and multi-objective optimization for energy retrofitting
a developed hospital reference building: A new approach to assess cost-optimality." Applied Energy 174 (2016): 37-68. [11]
Hajabdollahi, Hassan, Abdolsaeid Ganjehkaviri, and Mohammad Nazri Mohd Jaafar.
"Assessment of new operational strategy in optimization of CCHP plant for different climates using evolutionary algorithms." Applied Thermal Engineering 75 (2015): 468-480. [12]
Wei, Dajun, et al. "Multi-objective optimal operation and energy coupling analysis of
combined cooling and heating system." Energy 98 (2016): 296-307. [13]
Abdollahi, Gholamhossein, and Hoseyn Sayyaadi. "Application of the multi-objective
optimization and risk analysis for the sizing of a residential small-scale CCHP system." Energy and Buildings 60 (2013): 330-344. [14]
Li, Longxi, et al. "Optimization and analysis of CCHP system based on energy loads coupling
of residential and office buildings." Applied Energy 136 (2014): 206-216. [15]
Moradi, Mohammad H., et al. "An energy management system (EMS) strategy for combined
heat and power (CHP) systems based on a hybrid optimization method employing fuzzy programming." Energy 49 (2013): 86-101. [16]
Alipour, Manijeh, Behnam Mohammadi-Ivatloo, and Kazem Zare. "Stochastic risk-
constrained short-term scheduling of industrial cogeneration systems in the presence of demand response programs." Applied Energy 136 (2014): 393-404.
[17]
Bianco, Vincenzo, Mattia De Rosa, Federico Scarpa, and Luca A. Tagliafico.
"Implementation of a cogeneration plant for a food processing facility. A case study." Applied Thermal Engineering 102 (2016): 500-512. [18]
Moghadam, Ramin Shabanpour, Hoseyn Sayyaadi, and Hadi Hosseinzade. "Sizing a solar
dish Stirling micro-CHP system for residential application in diverse climatic conditions based on 3E analysis." Energy Conversion and Management 75 (2013): 348-365. [19]
Sanaye, Sepehr, and Ahmadreza Sarrafi. "Optimization of combined cooling, heating and
power generation by a solar system." Renewable Energy 80 (2015): 699-712. [20]
Wouters, Carmen, Eric S. Fraga, and Adrian M. James. "An energy integrated, multi-
microgrid, MILP (mixed-integer linear programming) approach for residential distributed energy system planning–A South Australian case-study." Energy 85 (2015): 30-44. [21]
Mohamed, Ayman, et al. "The performance of small scale multi-generation technologies in
achieving cost-optimal and zero-energy office building solutions." Applied Energy 152 (2015): 94-108. [22]
Askari, I. Baniasad, M. Oukati Sadegh, and M. Ameri. "Energy management and economics
of a trigeneration system Considering the effect of solar PV, solar collector and fuel price." Energy for Sustainable Development 26 (2015): 43-55. [23]
Jabari, F., et al. "Optimal short-term scheduling of a novel tri-generation system in the
presence of demand response programs and battery storage system." Energy Conversion and Management 122 (2016): 95-108. [24]
Evins, Ralph. "Multi-level optimization of building design, energy system sizing and
operation." Energy 90 (2015): 1775-1789. [25]
Yang, Hongming, et al. "Optimal operation of DES/CCHP based regional multi-energy
prosumer with demand response." Applied Energy 167 (2016): 353-365. [26]
Yang, Yun, Shijie Zhang, and Yunhan Xiao. "Optimal design of distributed energy resource
systems based on two-stage stochastic programming." Applied Thermal Engineering 110 (2017): 1358-1370. [27]
http://www.carrier.com/commercial/en/us/software/hvac-system-design/hourly-analysis-
program/ [Accessed February 16, 2016] [28]
http://sel.me.wisc.edu/trnsys/features/ [Accessed February 16, 2016]
[29]
Deb, Kalyanmoy, et al. "A fast and elitist multiobjective genetic algorithm: NSGA-II." IEEE
transactions on evolutionary computation 6.2 (2002): 182-197. [30]
Di Somma, Marialaura, Bing Yan, Nicola Bianco, Peter B. Luh, Giorgio Graditi, Luigi
Mongibello, and Vincenzo Naso. "Multi-objective operation optimization of a Distributed Energy System for a large-scale utility customer." Applied Thermal Engineering 101 (2016): 752-761.
[31]
Rodríguez, Laura Romero, et al. "Analysis of the economic feasibility and reduction of a
building’s energy consumption and emissions when integrating hybrid solar thermal/PV/microCHP systems." Applied Energy 165 (2016): 828-838. [32]
Fux, Samuel F., Michael J. Benz, and Lino Guzzella. "Economic and environmental aspects
of the component sizing for a stand-alone building energy system: A case study." Renewable energy 55 (2013): 438-447. [33]
Barbieri, E. S., Y. J. Dai, M. Morini, M. Pinelli, P. R. Spina, P. Sun, and R. Z. Wang.
"Optimal sizing of a multi-source energy plant for power heat and cooling generation." Applied Thermal Engineering 71, no. 2 (2014): 736-750. [34]
Akbari, Kaveh, et al. "Optimal investment and unit sizing of distributed energy systems under
uncertainty: A robust optimization approach." Energy and Buildings 85 (2014): 275-286. [35]
Kalogirou, Soteris A. Solar energy engineering: processes and systems. Academic Press,
2013. [36]
Florschuetz, L. W. "Extension of the Hottel-Whillier model to the analysis of combined
photovoltaic/thermal flat plate collectors." Solar energy 22.4 (1979): 361-366. [37]
Duffie, John A., and William A. Beckman. Solar engineering of thermal processes. Vol. 3.
New York: Wiley, 2013. [38]
http://www.eia.gov/oiaf/1605/emission_factors.html [Accessed February 16, 2016]
[39]
"2006 IPCC guidelines for national greenhouse gas inventories.", Intergovernmental Panel On
Climate Change, ch. 2, 2006.
Figure Captions: Figure 1. Electrical load profile of the building. Figure 2. Cooling and heating load profile of the building. Figure 3. Plane of Array solar irradiance. Figure 4. schematic diagram of the microgrid: (a) scenario 1; (b) scenario 2. Figure 5. Interactions of modeling and optimization blocks. Figure 6. Pareto Front: (a) scenario 1; (b) scenario 2. Figure 7. Generated power by PV/T panels for three solutions. Figure 8. Generated heat by PV/T panels for three solutions. Figure 9. Building electrical load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 10. Building heating load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 11. Daily cooling provided by EC and AC: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 12. Determined values of CER and PES (a) scenario 1; (b) scenario 2. Figure 13. Comparison of five parameters in two scenarios. The parameters are: 1) NPC; 2) PES; 3) CER; 4) Purchased electricity; 5) Sold electricity. Figure 14. Results of sensitivity analysis.
Table Captions: Table 1. Technical specifications of the microgrid components Table 2. Genetic algorithm characteristics Table 3. Range of decision variables in the problem Table 4. Economic parameters of the project Table 5. Unit cost and lifetime of each component Table 6. Carbon emission factors [38-39] Table 7. Component sizes obtained from optimization for 3 solutions Table 8. Comparison of PES and CER (average) with similar works Table 9. Parameters of sensitivity analysis
Figure 1. Electrical load profile of the building.
Figure 2. Cooling and heating load profile of the building.
Figure 3. Plane of Array solar irradiance.
(a)
(b) Figure 4. Schematic diagram of the microgrid: (a) scenario 1; (b) scenario 2.
Figure 5. Interactions of modeling and optimization blocks.
Figure 6. Pareto Front: (a) scenario 1; (b) scenario 2.
Figure 7. Generated power by PV/T panels for three solutions.
Figure 8. Generated heat by PV/T panels for three solutions.
Figure 9. Building electrical load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 10. Building heating load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 11. Daily cooling provided by EC and AC: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 12. Determined values of CER and PES (a) scenario 1; (b) scenario 2.
Figure 13. Comparison of five parameters in two scenarios. The parameters are: 1) NPC; 2) PES; 3) CER; 4) Purchased electricity; 5) Sold electricity.
Figure 14. Results of sensitivity analysis.
Table 1. Technical specifications of the microgrid components PV/T reference electrical efficiency (%)
15.5
Temperature coefficient (% / °C)
0.43
Heat loss coefficient (W/m2.K)
9.12
ICE electrical efficiency (%)
30
ICE thermal efficiency (%)
50
Boiler efficiency (%)
80
Grid efficiency (%)
32
COP of absorption chiller
0.7
COP of electric chiller
3.2
Table 2. Genetic algorithm characteristics Population size
50
Number of generations
100
Crossover percentage (%)
70
Mutation percentage (%)
20
Mutation probability (%)
30
Selection method
Roulette Wheel
Selection pressure
8
Table 3. Range of decision variables in the problem
PICE ,nom (kW)
100-1000
Q AC ,nom (kW)
100-1200
N PVT (#)
0-300
Table 4. Economic parameters of the project Project lifetime (years) Inflation rate (%) Interest rate (%) Grid electricity purchase price ($/kWh) Natural gas price ($/m3) Solar electricity sell price ($/kWh) ICE electricity sell price ($/kWh)
20 12 20 0.055 0.043 0.26 0.03
Table 5. Unit cost and lifetime of each component Lifetime (years) ICE PV/T Boiler AC EC
8 20 20 15 10
Capital/Replacement Cost ($/kW) 1400 3000 10 200 180
O&M Cost ($/kWh) 0.02 0.01 0.008 0.01 0.008
Table 6. Carbon emission factors [38-39] Grid electricity EF (kg/kWh)
0.598
Natural gas EF (kg/kWh)
0.202
Scenario 2
Scenario 1
Table 7. Component sizes obtained from optimization for 3 solutions
PICE ,nom
N PVT
Q B ,nom
Q AC .nom
Q EC ,nom
NPC
PES
(kW)
(#)
(kW)
(kW)
(kW)
($)
(%)
Sol. 1
108
--
459
129
1123
1715800
4.9032
Sol. 2
117
--
444
137
1115
1716600
5.1008
Sol. 3
123
--
434
144
1108
1717600
5.1392
Avg.
115
--
447
135
1177
1716442
5.063
Sol. 1
133
35
413
195
1057
1740500
5.2916
Sol. 2
125
136
414
213
1039
1793700
7.6360
Sol. 3
124
217
406
217
1035
1840500
9.2430
Avg.
128
118
410
203
1049
1785136
7.24
Table 8. Comparison of PES and CER (average) with similar works CHP
CHP+PV
CHP+PV+ST
CHP+PV/T
Li et al. [9]
31.42
36.59
*
*
Rodríguez et al. [31]
-18.68
-1.94
9.84
*
5.06
*
*
7.24
Li et al. [9]
51.4
54.77
*
*
Rodríguez et al. [31]
7.57
18.86
27.56
*
This work
3.54
*
*
5.38
Energy Saving Index (%)
This work Environmental Index (%)
Table 9. Parameters of sensitivity analysis
1) Purchasing price of grid electricity
2) Purchasing price of natural gas
3) Selling price of PV/T power
4) Selling price of ICE power
5) PV/T capital investment
6) ICE capital investment
Highlights A novel hybrid CCHP system is modeled and optimized. Hybrid CCHP system is compared with a conventional CCHP system. Economic, energy saving and environmental indices are calculated and assessed.
Highlights A novel hybrid CCHP system is modelled and optimized. Hybrid CCHP system is compared with a conventional CCHP system. Economic, energy saving and environmental indices are calculated and assessed.
S1359-4311(16)32003-8 http://dx.doi.org/10.1016/j.applthermaleng.2017.05.017 ATE 10328
To appear in:
Applied Thermal Engineering
Received Date: Revised Date: Accepted Date:
28 September 2016 30 April 2017 4 May 2017
Please cite this article as: H. Yousefi, M.H. Ghodusinejad, A. Kasaeian, Multi-objective Optimal Component Sizing of a Hybrid ICE+PV/T Driven CCHP Microgrid, Applied Thermal Engineering (2017), doi: http://dx.doi.org/ 10.1016/j.applthermaleng.2017.05.017
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Multi-objective Optimal Component Sizing of a Hybrid ICE+PV/T Driven CCHP Microgrid Hossein Yousefi a*, Mohammad Hasan Ghodusinejad a, Alibakhsh Kasaeian a a
Department of Renewable Energies and Environment, Faculty of New Sciences and Technologies, University of Tehran Address: North Karghar St., Faculty of New Sciences and Technologies, Tehran, Iran Tel: +9821 8609 3293 ; Email: [email protected]
Abstract This paper presents a multi-objective optimization approach in order to find the optimal capacities of a CCHP microgrid, to meet the power, heating and cooling loads of a large office building located in Tehran. The CCHP microgrid is modelled in two states: a) an Internal Combustion Engine (ICE) driven system; b) a solar assisted CCHP system including ICE and Photovoltaic/Thermal (PV/T) panels. The model is implemented in MATLAB and optimized via NSGA-II. The optimization is carried out to simultaneously minimize the Net Present Cost (NPC) and maximize the Primary Energy Saving (PES) of the system. Carbon Emission Reduction Ratio (CER) is also calculated to assess the environmental performance of the microgrid. The results revealed that, despite a slight increase in the average value of NPC, the CCHP system integrated with solar energy have a considerably better performance in energy saving and emissions than the ICE driven system. Keywords: CCHP microgrid, Internal combustion engine, PV/T panels, combined heat and power, Net present cost, Primary energy saving
1. Introduction Energy, has been a key input to the industrial, commercial and residential structures of today world. The urgent need for meeting the increasing energy demand and also, global concerns about environmental issues, has led the energy policies to the way that energy systems should be technically more efficient and environmentally cleaner. Microgrid is a key concept that can ease achieving these energy systems designing goals. Microgrids combine various energy sources in the best way to meet local energy demands. They can operate in both grid-connected and off-grid modes. The concept of microgrid was first introduced by the Consortium for Electric Reliability Technology Solutions, CERTS [1]. This consortium defines microgrids as clusters of generators, including heat recovery, storage, and loads, which are operated as single controllable entities [2]. Developing combined cooling, heating and power (CCHP) microgrids has attracted lots of attention. In comparison to conventional CCHP systems, CCHP microgrid is more efficient. It not only can meet cooling, heating and power demands, but also can interact with the grid, providing different services such as demand response programs and peak-shaving services [3-5]. On the other hand, to produce useful energy according to sustainable development policies, some issues should be considered including less or zero carbon emission, higher reliability of the energy networks, increased energy security and reduced costs. There have been different ways to attain the afore-mentioned goals. The three main concept suggested for sustainable development in energy systems are distributed generation (DG), cogeneration systems and renewable energy utilization. These are separate approaches and each can provide several advantages. Developing CCHP microgrids, all these three concept can be combined into one solution. A CCHP microgrid, can act as a DG unit and shares power with the main grid and can also improve the use renewable energy
sources, locally. Additionally, capturing power and heat simultaneously, increases the energy efficiency. Buildings, consuming 40% of global energy consumption, play an important role in energy market [6] and therefore, large buildings are good potentials for CCHP microgrids. 1.1. Previous Researches Optimal design of microgrids has been the subject of many researches within the recent years. Generally, researches in this area can be classified in three categories; sizing optimization, siting optimization and scheduling and economic dispatch of microgrid [2]. Therefore, there is a rich literature in analyzing CHP and CCHP systems. Numerous papers have analyzed conventional CCHP systems comprising of conventional CHP technologies [7-17], while, some other papers, have studied the performance of CCHP systems with hybrid CHP technologies or unique CHP components. Moghadam et al. [18] conducted a sizing optimization of a micro-CHP system based on a solar dish Stirling. The system was analyzed based on 3E analysis. The 3E analysis evaluated primary energy saving analysis, carbon dioxide emission reduction and payback period for return of investment. Three different scenarios was considered and it was found that the system can provide a good potential of energy saving and carbon emission reduction. Sanaye and Sarrafi [19] presented an energy, exergy and economic optimization of a solar-based CCHP system. The solar system was equipped with photovoltaic (PV), concentrated photovoltaic/thermal (CPVT), and evacuated tube (ET) collectors. Wouters et al. [20] presented a MILP model for optimal design of a CCHP multi-microgrid which contains different power, heating and cooling generation units and energy storage components. The model was applied to a South Australian case study and several scenarios were considered. Mohamed et al. [21] investigated the feasibility of CHP and CCHP systems, along with conventional heating and cooling systems of building.
Askari et al. [22] conducted the optimization of a hybrid micro-CCHP system comprised of solar PV, flat plate collector and natural gas generators, for a 5 story residential building. Jabari et al. [23] proposed a novel energy and exergy methodology for designing and optimal short-term planning of a heat pumpbased CCHP system driven by a solar dish Stirling heat engine. This CCHP system provides a good potential for reduction in CO 2 and NOx emission and an increase in the economic saving in fuel consumption. Evins [24] carried out a multi-level optimization for building design and energy system sizing and operation. The proposed system was comprised of several components including PV, solar collector, fuel cell, heat pump and CCHP-related components. Yang et al. [25] proposed a paradigm for energy hubs that contain CCHP components, renewable energy and energy storage. An operation model is suggested for regional multi-energy prosumers (RMEP) whose energy demands are served by interconnected energy hubs. Yang et al. [26] presented a study focusing on the optimal design of distributed energy resource systems with consideration of the uncertainties. A two-stage stochastic programming model was developed and applied to a hospital building in China to meet its cooling, heating and power loads. Results show that the uncertainties in load demands have a significant effect on the optimal system design, whereas the uncertainties in energy prices and renewable energy intensity have almost no effect.
1.2. This Work In this paper, a hybrid CCHP system is modeled and optimal component sizes are determined via a multi-objective optimization approach. The system is comprised of two types of CHP technologies; fossil fuel fired Internal Combustion Engine (ICE) and solar photovoltaic/thermal (PV/T) panels. There are some papers in previous researches that consider the integration of solar energy components into CCHP systems. These all have analyzed the
integration of photovoltaic panels and/or solar thermal collectors into conventional CHP components. Clearly, this would not lead to a hybrid CHP mechanism. In contrast, in this paper, the system would be a hybrid CCHP microgrid based on renewable and non-renewable CHP components. To assess the performance of the system, optimization is conducted in two scenarios. First, the system is modeled based on ICE, and then in second scenario, the system is modeled with ICE and PV/T. The total description of the system is presented in section 2. Mathematical models, related to both scenarios are covered in section 3. Optimization results are presented in section 4. Finally, conclusion of the paper is presented in section 5.
2. System Description The CCHP microgrid is modeled to meet the electrical, heating and cooling loads of an office building. The case study is one of the headquarters of Tehran Municipality which is a 7 story office building. The electrical load profile of the building is derived from previous years’ data of the building consumption and is illustrated in Fig. 1. The power is consumed in order to indoor and outdoor lighting, office facilities, HVAC components use, etc. The building is simulated in Hourly Analysis Program (HAP) [27] to create cooling and heating load profiles. The cooling/heating demand of the building is depicted in 2. As all simulations are conducted in time steps of 1 hour, load profiles are in hourly format. Currently the electricity demand of the building is totally met by the grid. A natural gas fired boiler and a direct-fired absorption chiller provide heating and cooling for the building, respectively. To model PV/T, solar irradiance data is needed. The hourly solar data is derived from meteorological TM2 file, the TMY library of TRNSYS 16 software [28]. Fig. 3 illustrates the Plane of Array irradiance.
As stated previously, the CCHP microgrid is modeled in two scenarios. Fig. 4 illustrates the schematic diagram of the microgrid components in both scenarios. In scenario 1, the ICE produces heat and electricity to feed the building loads and the boiler and grid act as heat and electricity backups, respectively. The cooling load of the building is whether met by indirect fired absorption chiller (AC) or electric chiller (EC). The total structure in scenario 2 is kept the same as the first scenario, except PV/T panel is added to the system to combine a hybrid CHP system in cooperation with the ICE. Technical specifications of the microgrid components are presented in Table 1. The data used for PV/T panels is derived from FOTOTHERM Cs panels with a nominal power of 250W. There are two different approaches for efficiencies and COPs of components in the literature. In some papers, these are considered to be fixed, but in some others, it is changed as a function of ratio of operation capacity to nominal capacity. In this work, the ICE operates in full load mode, so that the efficiency is as its nominal value. But for the chillers and the boiler, it is supposed not to be changed during the operation. 3. Mathematical Models In this section, the mathematical models of the CCHP microgrid are presented. The model is totally implemented in MATLAB. Two blocks of codes were created for modeling and optimization tasks. Fig. 5 depicts the interaction of modeling and optimization processes. 3.1. Genetic Algorithm Genetic Algorithm (GA) is the most known meta-heuristic optimization algorithm which is based on the survival of the strongest and fittest creature. The general procedure is based on four basic operators; initial population creation, selection, crossover and mutation. Several modified forms of GA have
been proposed by researchers to promote the performance of the algorithm. Deb et al. [29] proposed Non-dominated Sorting Genetic Algorithm II (NSGA-II) in 2002. The special feature of NSGA-II is the approach of ranking the population members; the solutions are classified to different groups with different domination levels and a rank is specified to each level. NSGA-II is a fast algorithm which can suitably solve sizing and scheduling problems of microgrids. In this paper, NSGA-II is utilized to solve the optimization problem. Table 2 provides the characteristics of the algorithm used in this paper. 3.2. Microgrid Modeling The performance of all components as well as cooperation of them within the CCHP microgrid have been modeled in two scenarios. The decision variables in the first scenario are nominal capacities of ICE and AC (in kW) while nominal capacities of boiler and EC are dependent variables and can be found during the problem solution. In the second scenario, the number PV/T panels would be added to decision variables. The range of these decision variables are presented in Table 3. The model in both scenarios is presented in the following.
3.2.1. Internal combustion engine To achieve the best operation efficiency of the ICE, it is considered that the engine would operate full time in rated operating power. Besides, the operation mode of the ICE is set to be full load in order to export power to the grid. This can make the system as a distributed generation unit. In cases the generated power by the ICE is more than the internal load, excess electricity will be sold to the grid. Therefore, the power from the ICE in each time step can be found as: PICE (t ) PICE ,nom
(1)
The natural gas consumed by the ICE is calculated by: Q f (t )
PICE (t )
(2)
ICE ,e
Finally, generated heat by the ICE can be found: Q ICE (t ) Qf (t ) ICE ,t
(3)
3.2.2. PV/T Panels Some papers, in the literature, have studied the integration of solar components into the CCHP system [30-34]. The area of each panel considered in this paper, as in the company catalog, is about 1.6m2. Therefore, the total useful area of the solar system is found: APVT 1.6N PVT
(4)
Generated power of the PV/T is a function of the total area as well as the panel efficiency and solar irradiance and can be written as: PPVT (t ) APVT PVT (t )G PoA (t )
(5)
In PV/T panels, circulating fluid, whether water or air, have a positive effect and act as a coolant to reduce the cell temperature. But in this paper, this effect is neglected and it is considered that the increase in cell temperature will cause a decrease in cell efficiency. Therefore, PV/T efficiency in each time step is calculated as below [35]: PVT (t ) ref 1 Tcell (t ) T ref
where, T ref is the reference temperature equal to 25°C. To calculate the cell temperature, many relations have been developed. In this paper, the following empirical relation, which can be used for poly-crystalline silicon cells, is utilized for the calculation of the cell temperature [35]:
(6)
Tc (t ) 30 0.0175 G PoA (t ) 300 1.14 T a (t ) 25
(7)
where, the ambient temperature is derived from the same source as solar irradiance. Calculation of generated heat by PV/T panels is slightly different from conventional flat plate collectors. The PV/T generated heat can be determined as [35, 36]: Q PVT (t ) APVT G PoA (t )( ) 1 PVT (t ) U L T cell (t ) T a (t )
(8)
where, ( ) is considered 0.9 as in [37]. 3.2.3. Cooling balance The cooling task of the building is done by both AC and EC: Q EC (t ) Q AC (t ) QC (t )
(9)
It is considered that, in cooling task, AC is prior to EC. If the cooling load is less than the nominal capacity of the AC, then it is totally met by the AC and the EC would not operate: Q AC (t ) QC (t )
(10)
Q EC (t ) 0
(11)
In cases that the AC capacity is not sufficient to meet the cooling load, it will operate in nominal power and the EC would cover the deficiency: Q AC (t ) Q AC ,nom
(12)
Q EC (t ) QC (t ) Q AC (t )
(13)
The nominal capacity of the EC can be found after the simulation is done for the whole year:
Q EC ,nom max Q EC
(14)
3.2.4. Heating balance The heat demand of the building can be written as equation (15). It is considered that the COP of the AC remains constant in different loads. Qdem (t ) Q H (t )
Q AC (t ) COPAC
(15)
In the first scenario, the heating demand is met by the ICE and the boiler: Q ICE (t ) Q B (t ) Qdem (t )
(16)
If the generated heat by the ICE is sufficient to meet the heating demand, then the boiler will not operate. Besides, if the generated heat is more than demand, the excess heat is exhausted. In cases the heating demand is more than the generated heat, boiler will cover the deficiency. For the second scenario, besides the ICE and the boiler, PV/T can also generate heat for the building: Q ICE (t ) Q PVT (t ) Q B (t ) Qdem (t )
(17)
As in the first scenario, if the generated heat by the ICE and the PV/T is as much as the heating demand, then the boiler will not operate. Here again, the excess generated heat, if exists, will be exhausted. In cases that the generated heat is not sufficient, boiler will operate as a backup to meet the heating demand. The nominal capacity of the boiler in both scenarios can be fined as follows: Q B,nom max Q B
(18)
For each time step, the consumed fuel by the boiler can be calculated using the boiler efficiency:
Q f ,B (t )
Q B (t )
B
(19)
The fuel consuming components of the microgrid are ICE and boiler. Therefore, the total fuel consumed by the microgrid is found: Qf ,Total Qf Qf ,B
(20)
3.2.5. Electricity balance In both scenarios, it is considered that the priority is to feed the building electrical load by the ICE and PV/T generated power. Therefore, local generated power is firstly fed to the internal loads and the surplus power, if exists, will be sold to the grid. This will ensure the self-sufficiency of the building in providing its demand. The power demand of the building consists of the electrical load as well as the power needed to run the EC and is written as below: Pdem (t ) PL (t )
Q EC (t ) COPEC
(21)
The electricity balance, in the first scenario, is as equation (22). In each time step, the generated power by the ICE is comprised of two fractions; the power sold to the grid and the power provided to the building load. This is stated in equation (23). If the generated power by the ICE is enough to meet the load, then no power is bought from the grid and excess generated power, if exists, is sold to the grid. If the ICE power is not as much as to meet the building load, then additional power is bought from the grid. PICE L (t ) Pgrid (t ) Pdem (t )
(22)
PICE (t ) PICE L (t ) PICE grid (t )
(23)
In the second scenario, the electricity generated by the PV/T is also considered to provide power for the building. Therefore, the electricity balance in this scenario is as equation (24). As well as ICE, the PV/T generated power is comprised of two parts; the power sold to the grid and the power provided to the building load. This is stated in equation (25). PICE L (t ) PPVT L (t ) Pgrid (t ) Pdem (t )
(24)
PPVT (t ) PPVT L (t ) PPVT grid (t )
(25)
The priority, in this scenario is to meet the load by the ICE power. This is considered since the Feed-in-Tariff (FiT) for solar power is too more than the one for ICE power and selling surplus solar power is economically more beneficial. Therefore, in each time step, if the generated power by the ICE is enough to meet the load, then no power is bought from the grid and the total PV/T power, as well as potential excess ICE power, is sold to the grid. In cases of insufficiency of the ICE power, PV/T power is fed to the building. If the power generated by both ICE and PV/T is not enough for the load, the deficiency is provided by the grid import. 3.2.6. Objective function As this research is a multi-objective optimization problem, two objective functions is considered to be optimized simultaneously by NSGA-II. The first objective function is Net Present Cost (NPC) of the microgrid. The goal is to achieve microgrid components capacity with minimized NPC. The NPC for this research is formulated as: 1 NPC C cap ,i C rep ,i K i C o &m ,i Pnom ,i CRF (r , L ) i C ng NG C grid Pbought C ICE ,Sell PICE ,Sell C PVT ,Sell
1 PPVT ,Sell CRF (r , L )
(26)
where it contains two cost parts; the first is capacity related costs and the second is the costs related to the energy transactions between the microgrid and utility, i.e. electricity and natural gas import and export. Obviously, the term related to PV/T power export is not considered in the first scenario. K i and CRF (r , L ) are single payment present worth factor and capital recovery factor, respectively, and can be calculated as follows [32]: yi
1 nLi n 1 (1 r )
K i (r , Li , y i )
r (1 r)L CRF (r , L ) (1 r)L 1
(27)
(28)
where, L i is the lifetime of component i , y i is the number of replacements of component i during the project lifetime, and r is the real interest rate and is given by: r
IN IF 1 IF
(29)
Table 4 provides details about the economic parameters considered in this paper. The unit costs and lifetime of each component are also presented in Table 5. Based on Table 5, and the lifetime of the project, it is clear that the ICE needs replacement twice and the EC and AC once, while there is no need for replacement of the boiler and PV/T. The second objective function is to maximize the Primary Energy Saving (PES) of the microgrid. PES is considered to assess the energy saving performance of the CCHP microgrid comparing current energy system performance of the building. Therefore, PES is calculated as: Q PES 1 CCHP Q ref
100
(30)
in which, QCCHP and Q ref are the total primary energy consumed by the CCHP microgrid and current energy system of the building, respectively, and are given by: 8760 8760
QCCHP Q f ,Total (t )
P t 1
t 1
8760
Q ref
PL (t ) t 1
grid
grid
t 1
B
(31)
grid
8760
Q H (t )
(t )
8760
Q t 1
C
(t )
(32)
COPAC ,ref
where, COPAC ,ref is coefficient of performance of the current direct-fired AC of the building and is considered as 1. 3.2.7. Environmental performance index In order to assess the environmental performance of the CCHP microgrid, Carbon Emission Reduction Ratio (CER) is defined. CER is used to compare Carbon emissions of the microgrid and the current system. Thus, it is calculated by: CE CER 1 CCHP CE ref
100
(33)
where, CE CCHP and CE ref are Carbon emissions of the CCHP microgrid and current energy system of the building, respectively, and are given by: 8760
8760
t 1
t 1
CE CCHP EFgrid Pgrid (t ) EFng Q f ,Total (t )
CE ref EFgrid
8760 8760 Q H (t ) QC (t ) 8760 PL (t ) EFng t 1 t 1 COPAC ,ref B t 1
(34)
(35)
EFgrid and EFng are the emission factors of grid electricity and natural gas
consumption, respectively, and can be found on Table 6.
4. Results and discussion The model, in both scenarios, is applied to the case study building mentioned in section 2. Two sets of solutions for components sizes were obtained. Fig. 6 illustrates the Pareto Fronts obtained for both scenarios. As can be seen from the Fig. 6.a, solutions experience slight changes in PES and NPC, in scenario 1; where the average values for PES and NPC are approximately 5.063% and 1716442$, respectively. Thus, applying the first scenario system configuration can provide an average 5% annual energy saving. From Fig. 6.b, the diversity of solutions is more than previous scenario; where from about 5 to 9% energy saving is achievable with changes of about 1.72 to 1.86 million dollars in NPC. In this scenario, by adding PV/T to the system configuration, an average PES of about 7.24% can be obtained with average value of 1.78 million dollar for NPC. Component sizes for three different solutions in each scenario, shown in Fig. 6, are provided in Table 7. It can be seen from Table 7 that by transition from scenario 1 to 2, the average value of solutions for nominal power of ICE and AC are increased while there is a decrease in average values of boiler and EC. This is mainly due to the addition of PV/T to the system. PV/T can provide more heat for the building which feed the AC and will result in a decrease in boiler capacity and increase in AC capacity. Excluding number of PV/T panels, the values for other components capacities are not distributed variously. Thus, the main effect on the value of NPC and PES is due to the variation in number of PV/T panels. This variation also effects on the generated heat and power by PV/T panels. Fig. 7 illustrates the generated
power by PV/T panels, for the three solutions stated in Table 7. It is clear that the generated power from the solution with the highest number of panels, i.e. solution 3, is too more than the one the least number of panels, i.e. solution 1. This would be the same as for the generated heat by panels, where is shown in Fig. 8. In scenario 1, power supplying of the building is done by ICE and grid, while in scenario 2, PV/T is also added to provide electricity for the building. Fig. 9 depicts the shares of each component for monthly power consumption of the building, for solution 3. It can be seen for both scenarios that for months June to September, the grid provides more electricity than local power generating units, i.e. ICE and PV/T. But in other months, a considerable share of building power is provided by the ICE and the PV/T. In the first scenario, about 53.22% of annual building load is met by the ICE and the rest by the grid. On the other hand, in scenario 2, about 3.2% and 53.63% of annual load is provided by PV/T and ICE respectively. It is necessary to mention that there can be slight differences in the electrical demand of Fig.9.a and Fig. 9.b. in some months due to the difference in the capacity of EC in each scenario. It was stated that the heating of the building in scenario 1 is met by both ICE and boiler. In scenario 2, generated heat by PV/T panels also is added to the thermal load supply. The share of each component in monthly heat supply for the building heating is illustrated in Fig. 10 (for solution 3). It can be seen that for scenario 1, ICE is dominant to provide heat for heating of the building in all months. Here, on the whole, about 73% of the annual heating is conducted by the ICE which is a considerable share. For scenario 2, PV/T panels also provide remarkable heat for the building. PV/T panels annually provide about 14.5% of the building heating while the share of the ICE is 67.93% and the rest is met by the boiler. Again, here, a considerable portion of heating is provided the recovered heat of CHP components.
From the capacities obtained in Table 7, it is clear that the EC owns a bigger share than the AC to provide cooling for the building. Fig. 11 illustrates the daily cooling provided by the EC and AC for both scenarios (solution 3). From Table 7 and Fig. 11, the share of AC in the second scenario is clearly more than the one in the first scenario. 4.1. Environmental Assessment As stated previously, to assess the environmental performance of the microgrid, CER index is utilized. Therefore, CER is calculated for all solutions in both scenarios. Fig. 12 depicts the values of CER in comparison with PES, for both scenarios. Due to the repeated and close solutions in the first scenario, as well as PES, CER is also calculated as an approximate fixed value. In this scenario, a CER of about 3.5% is achievable; it means that 3.5% annual reduction in Carbon emission, i.e. about 48650 kg/year less CO 2, can be a result of using a CCHP microgrid based on ICE. In scenario 2, the values for CER are more various. From Fig. 12.b, a range of about 3.3% to 7.5% of CER can be achieved, with the application of CCHP microgrid based on ICE and PV/T. Therefore, from about 46000 to more than 103000 kg/year reduction in CO2 emission would be available using different capacities derived in scenario 2 for microgrid design. CER is tightly relative to the value of PES. Therefore, the more energy is saved by the CCHP system, the less Carbon would be released to the environment. Besides, from Table 7 and Fig. 12.b, it can be said that the number PV/T panels has the most effect on the emission of the microgrid. Since, by increasing the number of panels from solution 1 to 3, CER experiences an improvement of more than 120%. 4.2. Comparison of the Results 4.2.1. Comparison of two scenarios
In order to compare the CCHP microgrid in two scenarios, 5 parameters were chosen and a comparison was applied to the results based on the following equation: F
F2 F1 100 F1
(36)
where, F is the value of each parameter and the numbers represent the number of each scenario. Here, scenario 1 was considered as the base state and the results of scenario 2 were compared with scenario 1. Five parameters selected for comparison are NPC, PES, CER, total purchased electricity and total sold electricity. The average value of these parameters for all solutions in both scenarios was calculated and used for comparison. Fig. 13 shows the result of comparison. In average, 4% increase can be seen in NPC as a result of transition from scenario 1 to 2. This is a slight negative change, shown in red in Fig. 13, while other parameters experience considerable improvements. The value of PES and CER is increased by about 43% and 52%, respectively. These improvements can also be seen in the values of imported and exported electricity from/to the grid, as the building purchased 17.31% less electricity while selling 184.14% more electricity to the grid.
4.2.2. Comparison with other works As the proposed system of this paper is unique, two papers, that have the closest scope to this work, were selected in order to compare the results. The environmental and energy saving results are extracted form Li et al. [9] and Rodríguez et al. [31] and presented in Table 8. As can be seen from Table 8, the values determined for the assessment of environmental and energy consumption performance of the systems are in different ranges. This is naturally due to
numerous differences in the problem, including the definition of indices, input data such as loads and irradiation, predefined energy management rules, etc. Meanwhile, the consistent fact in the three works is that, by adding solar components to the CCHP system, the performance of the system in emission and energy consumption improves. 4.3. Sensitivity Analysis This research was conducted in a deterministic approach. But the nature of each energy system contains several uncertainties. In order to assess the impact of some economical parameters on the performance of the CCHP system, a sensitivity analysis is carried out. Six parameters were selected and each was set to 50% higher and lower than the reference value. For each state, the optimization process was applied and repeated three times and the average value of NPC was calculated to analyze the effect of the change on the value of NPC. The parameters of sensitivity analysis are presented in Table 8. Fig. 14 illustrates the results of sensitivity analysis. It can be seen that the grid electricity price has the highest effect on the value of NPC, where a 50% decrease will lead to about 34% decrease in NPC while 50% increase results in about 10% increase in NPC. The second level of influence belongs to the ICE capital investment. A considerable point to mention is the effect of selling price of PV/T power. While it has a little effect on the NPC, it influences inversely and less price of PV/T power leads to more NPC. 5. Conclusion In this paper, a CCHP microgrid is modeled and an optimal sizing problem is solved using NSGA-II. Two different configurations are considered for the CCHP system. The first is a system fully based on fossil fuel, i.e. a nonrenewable CHP component. The second is modeled as a hybrid CCHP
microgrid using a renewable CHP component in addition to the non-renewable one. As the optimization is applied, two sets of solutions are obtained for two scenarios. From the details derived from the solutions, the following points are concluded: Application of CCHP systems in buildings is extremely beneficial as far as the energy saving and environmental objectives are concerned. In this paper, a conventional CCHP system with ICE, offered an average 5% energy saving and 3.5% less emission. Using PV/T panels as the second CHP component in addition to the ICE increased the annual energy savings up to 9.24% and the Carbon emission reduction up to 7.46%. For a sample solution, about 54% of the building electrical load was supplied by ICE in scenario 1. This is approximately summed with about 3.2% solar power in scenario 2, increasing the local power share up to about 57%. This means more than half of the annual electrical demand of the building is met locally which is an acceptable rate of self-sufficiency in electrical power supply. For a sample solution in scenario 1, about 73% of annual heat needed for space heating is provided by the ICE recovered heat. In the second scenario, the share of ICE decreased to 68%, while 14.5% of the demand is met by PV/T and therefore, about 82.5% of the space heating is provided by recovered heat of the ICE and PV/T. Notable points are attained by comparing two scenarios. Adding PV/T panels to the structure of the microgrid caused a 4% increase in the average value of NPC. In contrast with this adverse change, several benefits could be achieved. In solar-assisted CCHP microgrid, energy saving is increased and emissions are decreased considerably. Power transaction with the grid was also improved; the imported power from the
grid decreased by 17.31% while the exported power to the grid increased by 184.14%. In other words, a little increase in NPC, not only offers better energy saving and environmental characteristics, but also improves the power self-sufficiency of the building. As a future extension to this work, some ideas can be applied to the research, including considering other different dispatch strategies for the ICE such as electrical load following or thermal load following, examining the effect of adding electrical and thermal storage systems, considering PV and solar thermal components as well as PV/T panels to conduct a size optimization for these three components, etc.
Nomenclature Acronyms CCHP Combined Cooling, Heating and Power CHP Combined Heat and Power NPC Net Present Cost PES Primary Energy Saving CER Carbon Emission Reduction NSGA-II Non-dominated Sorting Genetic Algorithm II Symbols Electrical power (kW) P Thermal power (kW) Q Efficiency (%) Time step (h) t Area of solar panels (m2) A Number of solar panels N Solar irradiance (W/m2) G Temperature coefficient (power) (%/°C) Temperature (°C) T Heat loss coefficient (W/m2.K) UL
COP C K CRF r L
NG y IN IF CE EF
Coefficient of performance Cost related to the component Single payment present worth factor Capital recovery factor Real interest rate (%) Lifetime (years) Annual natural gas consumption (m3) Number of replacements Interest rate (%) Inflation rate (%) Carbon emission (kg) Emission factor (kg/kWh)
Subscripts ICE PVT B AC EC nom
f e t PoA ref cell
a C H L dem grid cap rep
o &m i ng bought Sell CCHP
Internal combustion engine Photovoltaic/thermal Boiler Absorption Chiller Electric Chiller Nominal capacity Fuel Electrical Thermal Plane of Array Reference value Solar cell Ambient Cooling load Heating load Electrical load Demand Grid Capital cost Replacement cost Operation and maintenance cost Index of components Natural gas Bought power from grid Sold power to grid CCHP system
References [1] Fathima, A. Hina, and K. Palanisamy. "Optimization in microgrids with hybrid energy systems–A review." Renewable and Sustainable Energy Reviews 45 (2015): 431-446.
[2] Gamarra, Carlos, and Josep M. Guerrero. "Computational optimization techniques applied to microgrids planning: a review." Renewable and Sustainable Energy Reviews 48 (2015): 413-424. [3] Gu, Wei, et al. "Modeling, planning and optimal energy management of combined cooling, heating and power microgrid: A review." International Journal of Electrical Power & Energy Systems 54 (2014): 26-37. [4] Dincer, I., and C. Zamfirescu. "Renewable‐energy‐based multigeneration systems." International Journal of Energy Research 36, no. 15 (2012): 1403-1415. [5] Wu, D. W., and R. Z. Wang. "Combined cooling, heating and power: a review." progress in energy and combustion science 32, no. 5 (2006): 459-495. [6] Nejat, Payam, et al. "A global review of energy consumption, CO 2 emissions and policy in the residential sector (with an overview of the top ten CO 2 emitting countries)." Renewable and Sustainable Energy Reviews 43 (2015): 843-862. [7] Ünal, Ali Nadi, İbrahim Ersöz, and Gülgün Kayakutlu. "Operational Optimization in Simple Trigeneration Systems." Applied Thermal Engineering 107 (2016): 175-183. [8] Farahnak, Mehdi, et al. "Optimal sizing of power generation unit capacity in ICE-driven CCHP systems for various residential building sizes." Applied Energy 158 (2015): 203-219. [9] Li, Miao, et al. "Optimal design and operation strategy for integrated evaluation of CCHP (combined cooling heating and power) system." Energy99 (2016): 202-220. [10]
Ascione, Fabrizio, et al. "Multi-stage and multi-objective optimization for energy retrofitting
a developed hospital reference building: A new approach to assess cost-optimality." Applied Energy 174 (2016): 37-68. [11]
Hajabdollahi, Hassan, Abdolsaeid Ganjehkaviri, and Mohammad Nazri Mohd Jaafar.
"Assessment of new operational strategy in optimization of CCHP plant for different climates using evolutionary algorithms." Applied Thermal Engineering 75 (2015): 468-480. [12]
Wei, Dajun, et al. "Multi-objective optimal operation and energy coupling analysis of
combined cooling and heating system." Energy 98 (2016): 296-307. [13]
Abdollahi, Gholamhossein, and Hoseyn Sayyaadi. "Application of the multi-objective
optimization and risk analysis for the sizing of a residential small-scale CCHP system." Energy and Buildings 60 (2013): 330-344. [14]
Li, Longxi, et al. "Optimization and analysis of CCHP system based on energy loads coupling
of residential and office buildings." Applied Energy 136 (2014): 206-216. [15]
Moradi, Mohammad H., et al. "An energy management system (EMS) strategy for combined
heat and power (CHP) systems based on a hybrid optimization method employing fuzzy programming." Energy 49 (2013): 86-101. [16]
Alipour, Manijeh, Behnam Mohammadi-Ivatloo, and Kazem Zare. "Stochastic risk-
constrained short-term scheduling of industrial cogeneration systems in the presence of demand response programs." Applied Energy 136 (2014): 393-404.
[17]
Bianco, Vincenzo, Mattia De Rosa, Federico Scarpa, and Luca A. Tagliafico.
"Implementation of a cogeneration plant for a food processing facility. A case study." Applied Thermal Engineering 102 (2016): 500-512. [18]
Moghadam, Ramin Shabanpour, Hoseyn Sayyaadi, and Hadi Hosseinzade. "Sizing a solar
dish Stirling micro-CHP system for residential application in diverse climatic conditions based on 3E analysis." Energy Conversion and Management 75 (2013): 348-365. [19]
Sanaye, Sepehr, and Ahmadreza Sarrafi. "Optimization of combined cooling, heating and
power generation by a solar system." Renewable Energy 80 (2015): 699-712. [20]
Wouters, Carmen, Eric S. Fraga, and Adrian M. James. "An energy integrated, multi-
microgrid, MILP (mixed-integer linear programming) approach for residential distributed energy system planning–A South Australian case-study." Energy 85 (2015): 30-44. [21]
Mohamed, Ayman, et al. "The performance of small scale multi-generation technologies in
achieving cost-optimal and zero-energy office building solutions." Applied Energy 152 (2015): 94-108. [22]
Askari, I. Baniasad, M. Oukati Sadegh, and M. Ameri. "Energy management and economics
of a trigeneration system Considering the effect of solar PV, solar collector and fuel price." Energy for Sustainable Development 26 (2015): 43-55. [23]
Jabari, F., et al. "Optimal short-term scheduling of a novel tri-generation system in the
presence of demand response programs and battery storage system." Energy Conversion and Management 122 (2016): 95-108. [24]
Evins, Ralph. "Multi-level optimization of building design, energy system sizing and
operation." Energy 90 (2015): 1775-1789. [25]
Yang, Hongming, et al. "Optimal operation of DES/CCHP based regional multi-energy
prosumer with demand response." Applied Energy 167 (2016): 353-365. [26]
Yang, Yun, Shijie Zhang, and Yunhan Xiao. "Optimal design of distributed energy resource
systems based on two-stage stochastic programming." Applied Thermal Engineering 110 (2017): 1358-1370. [27]
http://www.carrier.com/commercial/en/us/software/hvac-system-design/hourly-analysis-
program/ [Accessed February 16, 2016] [28]
http://sel.me.wisc.edu/trnsys/features/ [Accessed February 16, 2016]
[29]
Deb, Kalyanmoy, et al. "A fast and elitist multiobjective genetic algorithm: NSGA-II." IEEE
transactions on evolutionary computation 6.2 (2002): 182-197. [30]
Di Somma, Marialaura, Bing Yan, Nicola Bianco, Peter B. Luh, Giorgio Graditi, Luigi
Mongibello, and Vincenzo Naso. "Multi-objective operation optimization of a Distributed Energy System for a large-scale utility customer." Applied Thermal Engineering 101 (2016): 752-761.
[31]
Rodríguez, Laura Romero, et al. "Analysis of the economic feasibility and reduction of a
building’s energy consumption and emissions when integrating hybrid solar thermal/PV/microCHP systems." Applied Energy 165 (2016): 828-838. [32]
Fux, Samuel F., Michael J. Benz, and Lino Guzzella. "Economic and environmental aspects
of the component sizing for a stand-alone building energy system: A case study." Renewable energy 55 (2013): 438-447. [33]
Barbieri, E. S., Y. J. Dai, M. Morini, M. Pinelli, P. R. Spina, P. Sun, and R. Z. Wang.
"Optimal sizing of a multi-source energy plant for power heat and cooling generation." Applied Thermal Engineering 71, no. 2 (2014): 736-750. [34]
Akbari, Kaveh, et al. "Optimal investment and unit sizing of distributed energy systems under
uncertainty: A robust optimization approach." Energy and Buildings 85 (2014): 275-286. [35]
Kalogirou, Soteris A. Solar energy engineering: processes and systems. Academic Press,
2013. [36]
Florschuetz, L. W. "Extension of the Hottel-Whillier model to the analysis of combined
photovoltaic/thermal flat plate collectors." Solar energy 22.4 (1979): 361-366. [37]
Duffie, John A., and William A. Beckman. Solar engineering of thermal processes. Vol. 3.
New York: Wiley, 2013. [38]
http://www.eia.gov/oiaf/1605/emission_factors.html [Accessed February 16, 2016]
[39]
"2006 IPCC guidelines for national greenhouse gas inventories.", Intergovernmental Panel On
Climate Change, ch. 2, 2006.
Figure Captions: Figure 1. Electrical load profile of the building. Figure 2. Cooling and heating load profile of the building. Figure 3. Plane of Array solar irradiance. Figure 4. schematic diagram of the microgrid: (a) scenario 1; (b) scenario 2. Figure 5. Interactions of modeling and optimization blocks. Figure 6. Pareto Front: (a) scenario 1; (b) scenario 2. Figure 7. Generated power by PV/T panels for three solutions. Figure 8. Generated heat by PV/T panels for three solutions. Figure 9. Building electrical load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 10. Building heating load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 11. Daily cooling provided by EC and AC: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3). Figure 12. Determined values of CER and PES (a) scenario 1; (b) scenario 2. Figure 13. Comparison of five parameters in two scenarios. The parameters are: 1) NPC; 2) PES; 3) CER; 4) Purchased electricity; 5) Sold electricity. Figure 14. Results of sensitivity analysis.
Table Captions: Table 1. Technical specifications of the microgrid components Table 2. Genetic algorithm characteristics Table 3. Range of decision variables in the problem Table 4. Economic parameters of the project Table 5. Unit cost and lifetime of each component Table 6. Carbon emission factors [38-39] Table 7. Component sizes obtained from optimization for 3 solutions Table 8. Comparison of PES and CER (average) with similar works Table 9. Parameters of sensitivity analysis
Figure 1. Electrical load profile of the building.
Figure 2. Cooling and heating load profile of the building.
Figure 3. Plane of Array solar irradiance.
(a)
(b) Figure 4. Schematic diagram of the microgrid: (a) scenario 1; (b) scenario 2.
Figure 5. Interactions of modeling and optimization blocks.
Figure 6. Pareto Front: (a) scenario 1; (b) scenario 2.
Figure 7. Generated power by PV/T panels for three solutions.
Figure 8. Generated heat by PV/T panels for three solutions.
Figure 9. Building electrical load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 10. Building heating load supply: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 11. Daily cooling provided by EC and AC: (a) scenario 1 (solution 3); (b) scenario 2 (solution 3).
Figure 12. Determined values of CER and PES (a) scenario 1; (b) scenario 2.
Figure 13. Comparison of five parameters in two scenarios. The parameters are: 1) NPC; 2) PES; 3) CER; 4) Purchased electricity; 5) Sold electricity.
Figure 14. Results of sensitivity analysis.
Table 1. Technical specifications of the microgrid components PV/T reference electrical efficiency (%)
15.5
Temperature coefficient (% / °C)
0.43
Heat loss coefficient (W/m2.K)
9.12
ICE electrical efficiency (%)
30
ICE thermal efficiency (%)
50
Boiler efficiency (%)
80
Grid efficiency (%)
32
COP of absorption chiller
0.7
COP of electric chiller
3.2
Table 2. Genetic algorithm characteristics Population size
50
Number of generations
100
Crossover percentage (%)
70
Mutation percentage (%)
20
Mutation probability (%)
30
Selection method
Roulette Wheel
Selection pressure
8
Table 3. Range of decision variables in the problem
PICE ,nom (kW)
100-1000
Q AC ,nom (kW)
100-1200
N PVT (#)
0-300
Table 4. Economic parameters of the project Project lifetime (years) Inflation rate (%) Interest rate (%) Grid electricity purchase price ($/kWh) Natural gas price ($/m3) Solar electricity sell price ($/kWh) ICE electricity sell price ($/kWh)
20 12 20 0.055 0.043 0.26 0.03
Table 5. Unit cost and lifetime of each component Lifetime (years) ICE PV/T Boiler AC EC
8 20 20 15 10
Capital/Replacement Cost ($/kW) 1400 3000 10 200 180
O&M Cost ($/kWh) 0.02 0.01 0.008 0.01 0.008
Table 6. Carbon emission factors [38-39] Grid electricity EF (kg/kWh)
0.598
Natural gas EF (kg/kWh)
0.202
Scenario 2
Scenario 1
Table 7. Component sizes obtained from optimization for 3 solutions
PICE ,nom
N PVT
Q B ,nom
Q AC .nom
Q EC ,nom
NPC
PES
(kW)
(#)
(kW)
(kW)
(kW)
($)
(%)
Sol. 1
108
--
459
129
1123
1715800
4.9032
Sol. 2
117
--
444
137
1115
1716600
5.1008
Sol. 3
123
--
434
144
1108
1717600
5.1392
Avg.
115
--
447
135
1177
1716442
5.063
Sol. 1
133
35
413
195
1057
1740500
5.2916
Sol. 2
125
136
414
213
1039
1793700
7.6360
Sol. 3
124
217
406
217
1035
1840500
9.2430
Avg.
128
118
410
203
1049
1785136
7.24
Table 8. Comparison of PES and CER (average) with similar works CHP
CHP+PV
CHP+PV+ST
CHP+PV/T
Li et al. [9]
31.42
36.59
*
*
Rodríguez et al. [31]
-18.68
-1.94
9.84
*
5.06
*
*
7.24
Li et al. [9]
51.4
54.77
*
*
Rodríguez et al. [31]
7.57
18.86
27.56
*
This work
3.54
*
*
5.38
Energy Saving Index (%)
This work Environmental Index (%)
Table 9. Parameters of sensitivity analysis
1) Purchasing price of grid electricity
2) Purchasing price of natural gas
3) Selling price of PV/T power
4) Selling price of ICE power
5) PV/T capital investment
6) ICE capital investment
Highlights A novel hybrid CCHP system is modeled and optimized. Hybrid CCHP system is compared with a conventional CCHP system. Economic, energy saving and environmental indices are calculated and assessed.
Highlights A novel hybrid CCHP system is modelled and optimized. Hybrid CCHP system is compared with a conventional CCHP system. Economic, energy saving and environmental indices are calculated and assessed.