An illustration of the optimization of combined cooling heating and power systems using genetic algorithm

Abstract

The performances of combined cooling, heating and power (CCHP) system are greatly dependent on its design, operation strategy and thermal and electric demands. This paper illustrates how the use of a genetic algorithm can provide speedy optimization, by applying it to two styles of buildings operated in different operation strategies. The primary energy consumptions of CCHP system following electric demand management (EDM) and thermal demand management (TDM) are firstly analyzed respectively. Then, sixteen hypothetical buildings are constructed to represent various energy demands. Primary energy saving (PES), annual total cost saving (ATCS), and CO₂ emission reduction (CO₂ER), are weighted to evaluate the integrated performances of CCHP system in comparison to separation production system. Finally, the optimized CCHP system for sixteen scenarios using GA are compared.

Practical application: This paper provides an optimization design method for CCHP system. The performance analysis of CCHP systems running different operation modes for different buildings is believed by the authors to contribute to a significant guide for the fundamental design of CCHP systems. Although sensitivity to a number of other important design considerations such equipment performance, possible future changes in operating conditions, changes to the price or carbon intensity of grid-supplied energy etc are not addressed, the conclusions present a simple and effective direction and the proposed optimization algorithm and the evaluation method for CCHP system can be extended to other buildings.

Keywords

Combined cooling heating and power system building type electric demand management thermal demand management

Introduction

CCHP system is a booming technology because of the outstanding energy efficiency, reduced environmental emissions, and relative independence from centralized power grids. It has been widely recognized as a key alternative for the world to meet and solve energy-related problems, such as increasing energy demands, increasing energy cost, energy supply security, environmental pollution, and climate change.^1–6 CCHP system has been introduced in China to various kinds of buildings such as hotel, office, and hospital.⁷

A good CCHP system must achieve economical saving, but more importantly must achieve real energy saving as well as reducing the emission of pollutants. In order to improve the performance of CCHP system, a large number of researchers studied CCHP systems from various aspects including model,^8–18 optimization,^16,18–25 operation strategy,^12,16,23 feasibility analysis,^19,26–36 multi-criteria evaluation,^4,5,37 etc. The performance of CCHP system is greatly dependent on its operation strategy and the characteristic of energy demands of building.^12,38–40 Two of the simplest operation strategies are to run prime mover in accordance to either electric or thermal demand. Cardona and Piacentino³⁸ referred to these two styles as electric demand management (EDM) and thermal demand management (TDM). The choice between EDM and TDM is usually governed by the load level of prime mover as well as a few extraneous circumstances such as the ability to sell back electricity to grid, the purchased price of electricity from grid, and the price of fuel.

CCHP systems are inherently complex and not easy to be optimized. This paper illustrates how the use of a GA can provide speedy optimization, by applying it to two buildings operated in either of two modes and evaluated on four different criteria (16 cases in all). Section “CCHP system” analyzes the primary energy consumptions of CCHP system following EDM and TDM and presents the evaluation criteria. Section “Energy demands of buildings” constructs 16 scenario buildings to represent the different characteristic of energy demands. Section “Results and discussion” optimizes the CCHP systems and discusses the performances.

CCHP system

In order to measure the benefits achieved by CCHP system in comparison to SP system, a SP system is selected as a reference system. The general structures of SP system and CCHP system are shown in Figure 1. The energy demands of building include: (1) electric energy use for lights and equipments, E (kWh); (2) cooling demand for space cooling, Q_c (kWh); and (3) heating demand for space heating and domestic hot water, Q_h (kWh).

Figure 1.

General structures of reference SP system and CCHP system.

In the SP system, the cooling system adopts electric chiller, the heating system utilizes gas boiler, and the electricity for the building and the electric chiller comes from the local centralized power grid.

The CCHP system consists of a power generation unit (PGU), a waste recovery system, a back-up boiler, a cooling system, and a heating system. The PGU is driven by natural gas to produce the electricity needed by the building. The high-temperature exhaust gas from the PGU is recovered to accommodate the thermal load for cooling in summer and heating in winter. The cooling system adopts absorption chiller to utilize the recovered heat. If the recovered heat does not completely satisfy the application needs, the back-up boiler can be used to supplement the additional heat. Similarly, when the amount of electricity from the PGU is not enough to the building, the additional electricity comes from the grid. Contrarily, the excess products could be stored or sold to other users when there are excess heat or electricity produced by the CCHP system. However, it is important to mention here that the excess products from the CCHP system are dissipated directly, and their energy saving and economic saving are not considered into the independent CCHP system. Consequently, the CCHP system must reduce another excess product when it satisfies one kind of energy demands of building.

During the following analysis of CCHP system, some important assumptions are followed:

The minimum technical limit of CCHP system is neglected. The CCHP equipments can operate anywhere between 0% and 100% of its rated capacity, and ramping rate for load adjustment is not included.

The CCHP system is assumed to be 100% reliable.

The efficiency drops of CCHP equipments at part load operation are neglected to simplify the analysis and calculation.

The sales of surplus heat or electricity form CCHP system are prohibited.

Primary energy consumption of SP system

Aimed to satisfy the energy demands of building, the primary energy consumed by the SP system in Figure 1(a) can be calculated as:

F^{SP} = \frac{E}{η_{e}^{SP} η_{grid}} + \frac{E_{p}^{SP}}{η_{e}^{SP} η_{grid}} + \frac{Q_{c}}{{COP}_{e} η_{e}^{SP} η_{grid}} + \frac{Q_{h}}{η_{b}^{SP} η_{h}^{SP}}

(1)

where

F^{SP}

is the primary energy consumption of the SP system (kWh),

E_{p}^{SP}

is the parasitic electric energy consumption (such as pumps and fans) (kWh),

η_{e}^{SP}

is the generation efficiency of the SP system,

η_{grid}

is the grid distribution efficiency,

η_{b}^{SP}

is the gas boiler efficiency,

η_{h}^{SP}

is the heating coil efficiency, and

{COP}_{e}

is the coefficient of performance (COP) of the electric chiller.

Primary energy consumption of CCHP system

The CCHP system usually operates following either electric or thermal demand. The operation strategy determines the performance of CCHP system directly. Herein, the two operation modes, EDM and TDM, are employed to compare the performances of CCHP system.

EDM operation mode

The maximum input fuel energy of the PGU is assumed to $F_{max}$ (kWh). The PGU operates according to the electric demand in EDM operation mode so that the input fuel energy, $F_{pgu}$ (kWh), can be determined by the electric load. When $F_{pgu}$ is less than $F_{max}$ , the electricity generated by the PGU is just to satisfy the electric load and the parasitic electric energy. The recovered heat is utilized to produce cooling and/or heat. When the recovered heat is not enough, the back-up boiler runs to supplement the additional heat. When the electric demand is more than the maximum generation electricity, the additional electricity comes from the local grid. Therefore, the CCHP system would not export excess electricity in EDM operation mode while the excess heat may be produced and dissipated directly.

The operating condition and the primary energy consumption are expressed as follows:

Test condition A : E + E_{p} \geq F_{max} η_{e}

(2)

If Test condition A = True then

F = F_{max} + \frac{E + E_{p} - F_{max} η_{e}}{η_{e}^{SP} η_{gird}} + \frac{Q_{b}}{η_{b}} \cdot U

(3)

where

Q_{b} = \frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}} - F_{max} (1 - η_{e}) η_{rec}

(4)

U = {1, Q_{b} \geq 0 0, Q_{b} < 0

(5)

where U is the operation status of the back-up boiler,

E_{p}

is the parasitic electric energy of the CCHP system (kWh), F is the primary energy consumption of the CCHP system (kWh),

Q_{b}

is the supplemental heat from the gas boiler (kWh),

η_{e}

is the PGU generation efficiency,

η_{b}

is the back-up boiler efficiency,

η_{h}

is the heating coil efficiency,

η_{rec}

is the heat recovery system efficiency, and

{COP}_{ch}

is the absorption chiller COP.

If Test condition A = False then

F = \frac{E + E_{p}}{η_{e}} + \frac{Q_{b}}{η_{b}} \cdot U

(6)

where

Q_{b} = \frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}} - \frac{E + E_{p}}{η_{e}} (1 - η_{e}) η_{rec}

(7)

TDM operation mode

The PGU operates according to the thermal demand of building in TDM mode so that the input fuel energy is determined by the thermal load. Similarly, the recovered heat from the PGU is just to satisfy the thermal demand when the input fuel energy is less than the maximum capacity. The generated electricity is provided to the building and the distribution system. When the electricity is not enough, the additional electricity comes from the grid. When the thermal load is more than the maximum recovered heat, the additional heat is supplemented by the back-up boiler. Therefore, the CCHP system would not exhaust excess heat in TDM operation mode while the excess electricity may be exported.

Similarly, the operating condition and the primary energy consumption are expressed as follows:

Test condition B : \frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}} \geq F_{max} (1 - η_{e}) η_{rec}

(8)

If Test condition B = True then

F = F_{max} + \frac{\frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}} - F_{max} (1 - η_{e}) η_{rec}}{η_{b}} + \frac{E_{grid}}{η_{e}^{SP} η_{gird}} \cdot V

(9)

where V is the on-off switch status of the grid connecting CCHP system,

E_{grid}

is the electricity from the grid (kWh), and

V = {1, E_{grid} \geq 0 0, E_{grid} < 0

(10)

E_{grid} = E + E_{p} - F_{max} η_{e}

(11)

If Test condition B = False then

F = \frac{\frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}}}{(1 - η_{e}) η_{rec}} + \frac{E_{grid}}{η_{e}^{SP} η_{gird}} \cdot V

(12)

where

E_{grid} = E + E_{p} - \frac{\frac{Q_{c}}{{COP}_{ch}} + \frac{Q_{h}}{η_{h}}}{(1 - η_{e}) η_{rec}} η_{e}

(13)

Evaluation criteria of CCHP system

To measure the benefits achieved by the CCHP system in comparison to the SP system from energy, economy, and environment, the following criteria are used.

primary energy saving (PES) PES is defined as the percentage of the saving energy of the CCHP system in comparison to the SP system to the energy consumption of the SP system. It is written as:

PES = \frac{F^{SP} - F}{F^{SP}} \times 100 %

(14)

annual total cost saving (ATCS) The annual total cost including annual capital cost and operation cost, is calculated to:

ATC = R \times \sum_{k = 1}^{l} N_{k} C_{k} + \sum_{i = 1}^{365} \sum_{k = 1}^{24} \times (E_{ik, grid} C_{ik, e} + F_{ik} C_{ik, f})

(15)

where

ATC

is the annual total cost of the CCHP system (Yuan) and (1$ ≅ 7Yuan),

N_{k}

is the installation capacity of the kth equipment (kW),

C_{k}

is the capital cost per unit of the kth equipment (Yuan/kW), l is the number of equipments,

C_{ik, e}

and

C_{ik, f}

are the hourly prices of electricity and natural gas (Yuan/kWh), and

E_{ik, grid}

and

F_{ik}

are the hourly demands of electricity bought from the grid and natural gas (kWh). The capital recovery factor, R, is defined to:

R = \frac{i (1 + i)^{n}}{(1 + i)^{n} - 1}

(16)

where i is the interest rate and n is the service life (year). Herein it is assumed that the values of i and n are equal to all kinds of equipments. Similarly, ATCS is defined to the percentage of the saving annual total cost of the CCHP system in comparison to the SP system to the annual total cost of the SP system. It is expressed to:

ATCS = \frac{{ATC}^{SP} - ATC}{{ATC}^{SP}} \times 100 %

(17)

where

{ATC}^{SP}

is the annual total cost of the SP system.

CO₂ emission reduction (CO₂ER) The amount of CO₂ emission from the CCHP system can be estimated to^12,41

{CO}_{2} E = μ_{{CO}_{2}, f} F_{on - site} + μ_{{CO}_{2}, e} E_{grid}

(18)

where

{CO}_{2} E

is the CO₂ emission of the CCHP system (g),

μ_{{CO}_{2}, f}

and

μ_{{CO}_{2}, e}

are the emission conversion factors of fuel and electricity from grid, respectively, (g/kWh),

F_{on - site}

is the on-site fuel consumption of the CCHP system (kWh), and

F_{on - site} = F_{pgu} + F_{b}

. CO₂ER of the CCHP system in comparison to the SP system is calculated to:

{CO}_{2} ER = \frac{{CO}_{2} E^{SP} - {CO}_{2} E}{{CO}_{2} E^{SP}} \times 100 %

(19)

where

{CO}_{2} E^{SP}

is the amount of CO₂ emission from the SP system.

Energy demands of buildings

Two styles of buildings are defined to the application and analysis of the CCHP systems. One is a hotel building and another is an office building both in Beijing, China. Both of buildings have the same area, construction, and characteristic parameters. General information of the simulated buildings are presented in Table 1. The hotel building includes guest rooms, office rooms, dinning halls, ballrooms, and meeting rooms. The hotel always operates during the entire year while the office building operates only from 8:00 to 20:00 on weekdays and does not operate on weekend. The hourly energy consumptions of the two buildings are estimated in software DeST.⁴² The annual energy demands of the buildings are summarized in Table 2. It can be found that the energy demand of the hotel is largely more than the office, and the annual heating demand of the hotel reaches four times as much as the office.

Table 1.

General description of the simulated buildings using DesT.

Location		Beijing, China
Orientation		L-building, long side aligned with north
Total area		9433 m² (six floors)
Glass area		30% in each wall
Hotel building	People	Two people/room for everyday
	Occupancy schedule	Until^a (fraction)^b: 7 (1), 9 (0.5), 11 (0.3), 12 (0.1), 14 (0.5), 16 (0.3), 18 (0.1), 21 (0.3), 23 (1)
	Electric equipment	500 W
	Equipment schedule	Until^a (fraction)^b: 5 (0), 7 (0.2), 16 (0.05), 19 (0.3), 22 (1), 23 (0.5)
	Lights	6 W/m²
	Lights schedule	Until^a (fraction)^b: 5 (0), 6 (0.2), 15 (0.1), 17 (0.3), 21 (1), 22 (0.7), 23 (0.5)
	Thermostat set point	22–24℃
Office building	People	0.1 people/m² for weekdays, 0 for weekend.
	Occupancy schedule	Until^a (fraction)^b: 7 (0), 8 (1), 12 (0.3), 16 (1), 17 (0.3), 18 (0.1), 21 (0)
	Electric equipment	10 W/m²
	Equipment schedule	Same as for occupancy
	Lights	10 W/m²
	Lights schedule	Same as for occupancy
	Thermostat set point	Winter^c: 20–22℃; Summer^d: 24–26℃; Spring and autumn^e: 20–26℃

The hour of the day until the specified fraction is considered.

The fraction of the total value of the variable that is considered in the calculation for that specific period of time.

November 15–March 15.

June 1–August 30.

March 16–May 31 and August 31–November 14.

Table 2.

The annual energy demands of the simulated buildings.

Building	Annual electric demand, kWh	Annual cooling demand, kWh	Annual heating demand, kWh
Hotel	872,773	886,127	1,472,099
Office	497,619	404,968	339,733

The energy demands of the buildings are shown in Figures 2 and 3. To show the energy characteristic of the buildings, the energy curves in typical days are shown in Figure 2. The typical day is defined to the representative day when the characteristic of energy demands of building can be reflected and demonstrated. Here, 18 January 18 and 2 August when the heat load and the cooling load are the maximum, respectively, are selected to represent the corresponding typical days in winter and summer. In Figure 2, (a) represents in winter and (b) represents in summer. The daily energy demands and the cumulative curves are displayed in Figure 3, where (a) represents the daily energy demands and (b) represents the cumulative curves. It is notable that the energy demands in typical days are only used to show the characteristic of the buildings in Figure 2, while the calculation and analysis are based on the annual energy demands in Figure 3.

Figure 2.

The hourly energy demands of buildings on the typical day in winter and in summer.

Figure 3.

The daily energy demands in all-year and the energy demands cumulative curves of buildings.

From these profiles in Figures 2 and 3, the hotel has the following characteristics: (1) There needs heat to produce domestic hot water during the entire year; (2) The daily electric demand is stable relatively and is less than the heating and cooling loads; (3) The daily heating and cooling loads both fluctuate more greatly than the daily electricity demand; (4) The cooling peak is greater than the heating peak because of the hot climate in Beijing; and (5) In spring and autumn, the three energy demands for cooling, heating, and electricity are almost same and they fluctuate only between a little range.

The following characteristic of the office building can be derived: (1) The daily electric and cooling demands are stable relatively; (2) The daily heating load fluctuates more greatly than the electric and cooling demands; (3) The heating peak is greater than the cooling peak; and (4) In spring and autumn, the electric demand is more than the heating and the cooling.

The comparison between the two buildings can derive the following characteristics: (1) The energy consumption of the hotel is continuous while the office’s is intermittent. There are energy demands of the hotel during the entire year while the office does not need energy supply at non-working time and on weekend; and (2) the ratio of electricity to heating or cooling of the office building is greater than that of the hotel.

Aimed to compare more different buildings, some load scenarios are assumed. Each of load scenarios is represented by an “energy demand vector” composed of a triplet of the form

D = (d_{1}, d_{2}, d_{3})

(20)

where d₁ indicates the electric load level, d₂ represents the cooling load level, and d₃ represents the heating load level. The different load levels are referred to the base energy demands of buildings indicated by the vectors

D_{1, hotel} = (1, 1, 1)

and

D_{1, office} = (1, 1, 1)

in Figures 2 and 3.

In order to point out the impact of the energy demands on the CCHP system’s performance, the scenarios in Table 3 are analyzed and compared. Based on the basic energy demands in Figure 2, the following scenarios with respect to the hotel are discussed:

Scenario D ₁ _,hotel : energy demand vector $D_{1, hotel} = (1, 1, 1)$ .

Scenario D ₂ _,hotel : energy demand vector $D_{2, hotel} = (1, 0.5, 0.5)$ . The heating and cooling load patterns of the hotel building are uniformly halved with respect to the values specified in Scenario D₁ _,hotel , while the electric demand of the hotel remains unchanged.

Scenario D ₃ _,hotel : energy demand vector $D_{3, hotel} = (1, 2, 1)$ . The cooling load pattern of the hotel is uniformly doubled with respect to the values specified in Scenario D₁ _,hotel , while the electric and thermal demands of the hotel remain unchanged.

Scenario D ₄ _,hotel : energy demand vector $D_{4, hotel} = (1, 1, 2)$ . The thermal load pattern of the hotel is uniformly doubled with respect to the values specified in Scenario D₁ _,hotel , while the electric and cooling demands of the hotel remain unchanged.

Similarly, the following scenarios with respect to the office building are defined and listed in Table 3.

Scenario D ₁ _,office : energy demand vector $D_{1, office} = (1, 1, 1)$ .

Scenario D ₂ _, _office : energy demand vector $D_{2, office} = (1, 0.5, 0.5)$ .

Scenario D ₃ _, _office : energy demand vector $D_{3, office} = (1, 2, 1)$ .

Scenario D ₄ _, _office : energy demand vector $D_{4, office} = (1, 1, 2)$ .

Table 3.

Energy demand vectors of load scenarios.

Building	Scenarios	Electric load	Cooling load	Heating load
Hotel	Scenario D ₁ _,hotel	1	1	1
	Scenario D ₂ _,hotel	1	0.5	0.5
	Scenario D ₃ _,hotel	1	2	1
	Scenario D ₄ _,hotel	1	1	2
Office	Scenario D ₁ _,office	1	1	1
	Scenario D ₂ _,office	1	0.5	0.5
	Scenario D ₃ _,office	1	2	1
	Scenario D ₄ _,office	1	1	2

Thus, eight scenario buildings that have different energy demands are constructed according to the baseline buildings. The annual ratios of heat to electricity of scenarios are summarized to Table 4. Additionally, the eight scenarios are doubled to compare the performances of CCHP system running in different operation modes, EDM and TDM.

Table 4.

The ratios of thermal demand to electric demand of buildings.

	Scenario D ₁			Scenario D ₂			Scenario D ₃			Scenario D ₄
	Q/E	Q_h/E	Q_c/E	Q/E	Q_h/E	Q_c/E	Q/E	Q_h/E	Q_c/E	Q/E	Q_h/E	Q_c/E
Hotel	3.48	2.34	2.19	1.74	1.17	1.10	4.62	2.34	4.38	5.83	4.68	2.19
Office	2.16	2.65	1.98	1.08	1.32	0.99	3.10	2.65	3.97	3.38	5.29	1.98

Results and discussion

Optimization results

The characteristic parameters of the CCHP system and the SP system, the capital cost, and the energy prices are listed in Tables 5 –7, respectively. Based on the energy demands of the building and the operation strategy, the equipment capacities are optimized and determined in GA to maximize the following integrated benefit achieved by the CCHP system (the detail GA optimization procedures of CCHP system can be found in Appendix 2 or the Ref. 43):

max J = ω_{1} \cdot PES + ω_{2} \cdot ATCS + ω_{3} \cdot {CO}_{2} ER

(21)

where J is the objective function,

ω_{1}

ω_{2}

, and

ω_{3}

are the corresponding weights of PES, ATCS, and CO₂ER, and all of weights are equal to 1/3 here.

Table 5.

Input values employed for the energy used calculations for SP system and CCHP system.

System	Variable	Symbol	Value
CCHP system	PGU efficiency	$η_{e}$	0.25^a
	Waste heat recovery system efficiency	$η_{rec}$	0.8
	Cooling system COP	${COP}_{ch}$	0.7
	Heating coil efficiency	$η_{h}$	0.8
	Boiler efficiency	$η_{b}$	0.8
	Heat/electricity ratio	2.40
SP system	PGU efficiency	$η_{e}^{SP}$	0.35^b
	Cooling system COP	${COP}_{e}$	3.0
	Heating coil efficiency	$η_{h}^{SP}$	0.8
	Boiler efficiency	$η_{b}^{SP}$	0.8
	Grid transmission efficiency	$η_{grid}$	0.92
CO₂ emission conversion factor	Natural gas, g/kWh	$μ_{{CO}_{2}, f}$	220
CO₂ emission conversion factor	Electricity from gird, g/kWh	$μ_{{CO}_{2}, e}$	968

It is assumed that the PGU efficiency is kept constant in this computation and analysis although it is not constant when the PGU operates either under TDM or EDM.

The coal power plant produces the electricity to send to the centralized power grid in the reference SP system in this analysis. Different values lead to the difference of the calculation results.

Table 6.

Unit price of the facilities.²⁴

Facility	Prime mover	Heating coil	Boiler	Absorption chiller	Electric chiller
Unit price^a (Yuan/kW)	6800	200	300	1200	970

The date of unit price is 2008.

Table 7.

Unit price of electricity and natural gas.²⁴

Natural gas	Electricity (6:00–21:00)	Electricity (22:00–5:00)
Unit price^a (Yuan/kWh)	0.194	0.964	0.435

The date of unit price is 2008.

The optimal capacities are listed in Table 8, which shows that the capacity of the PGU in EDM mode is greater than or equal in TDM mode. Herein, it is assumed that the optimal capacities are just the installation capacities. The investment costs are summarized into Figure 4. It can be found that the investment costs of CCHP system in EDM mode and TDM mode for each scenario are almost equivalent, and the EDM mode is a little higher than or equal the TDM mode. The investment cost changes of scenarios of the hotel and the offices with the build energy demands are similar. When the cooling demand of the building is the main consumption, the investment cost will increase dramatically and the investment cost of scenario D₃ is the highest. When the heating load of the building increase, the investment cost of CCHP system increase smoothly and the difference between D₁ and D₄ is not large.

Figure 4.

The investment costs of CCHP systems in EDM and TDM operation modes.

Table 8.

The capacities of equipments in CCHP systems.

Scenarios	Prime mover (PGU + heat recovery system)			Auxiliary boiler F_b (kW)	Absorption chiller Q_c (kW)	Heating coil Q_h (kW)
Scenarios	Heat (kW)	Electricity (kW)	F_max (kW)	Auxiliary boiler F_b (kW)	Absorption chiller Q_c (kW)	Heating coil Q_h (kW)
Scenario D ₁ _,hotel _, _EDM	458	191	764	2724	1640	872
Scenario D ₁ _,hotel _, _TDM	341	142	569	2870	1640	872
Scenario D ₂ _,hotel _, _EDM	372	155	621	1183	820	436
Scenario D ₂ _,hotel _, _TDM	372	155	621	1183	820	436
Scenario D ₃ _,hotel _, _EDM	501	209	836	5600	3280	872
Scenario D ₃ _,hotel _, _TDM	335	139	559	5807	3280	872
Scenario D ₄ _,hotel _, _EDM	501	209	836	3038	1640	1744
Scenario D ₄ _,hotel _, _TDM	321	133	535	3263	1640	1744
Scenario D ₁ _,office _, _EDM	563	234	939	1301	821	1279
Scenario D ₁ _,office _, _TDM	520	216	867	1348	821	1279
Scenario D ₂ _,office _, _EDM	322	134	538	595	411	639
Scenario D ₂ _,office _, _TDM	322	134	537	596	411	639
Scenario D ₃ _,office _, _EDM	567	236	946	2578	1643	1279
Scenario D ₃ _,office _, _TDM	541	225	903	2256	1643	1279
Scenario D ₄ _,office _, _EDM	562	234	938	3299	821	2557
Scenario D ₄ _,office _, _TDM	492	205	821	3380	821	2557

Performance analysis

The energetic, economic, and environmental performances and the integrated performances for the 16 scenarios are calculated and displayed in Figure 5, where (a) represents the performances in EDM operation mode and (b) represents the performances in TDM operation mode.

Figure 5.

Annual performance of the CCHP systems in comparison to the SP systems for different buildings in EDM and TDM operation modes.

The average values of PES, ATCS, CO₂ER, and the integrated performance of the CCHP systems for all of scenario buildings in the two different operation modes are 5.8%, 3.1%, 23.1%, and 10.6%, respectively. It is known that the CO₂ER of the CCHP system in comparison to the SP system is the most outstanding. The average performances for all of scenario buildings in EDM mode are 5.4%, 4.3%, 25.3%, and 11.7% while the corresponding performances in TDM mode are 6.2%, 1.8%, 20.8%, and 9.6%, respectively. It can be found that the performance in EDM mode is better than in TDM mode except to the PES. Similarly, the average performances of CCHP systems for the hotel scenarios are 7.7%, 3.2%, 24.1%, and 11.7% while the performances of the office scenarios are 3.9%, 3.0%, 22.0%, and 9.6%, respectively. It can be concluded that the CCHP system is more suitable to the hotel building than the office building.

When Scenario D₁ is compared with Scenario D₃, it can be found that the performances of CCHP systems for hotel and office decrease with the increase of cooling load. For example, the PES, ATCS, CO₂ER, and the integrated performance for the hotel in EDM mode decrease 10.1%, 11.7%, 6.5%, and, 9.4%, respectively. There are even some negative values in Scenario D₃, which means that the CCHP system cannot achieve real benefit in comparison to the SP system. Especially, the increasing of cooling load of building would not save primary energy and cost in TDM operation mode. When the PGU runs in TDM mode and the annual ratio of heat to electricity of Scenario D₃ is greater than that of the PGU, 2.40, the excess electricity is produced by the PGU in order to satisfy the thermal demand. It is noted that the PES and the ATCS of Scenario D₃ hotel in TDM mode are less than the office’s because the difference between the heat/electricity ratio of hotel and the PGU is greater than the difference between the office’s and the PGU. From the whole comparison, the PES, ATCS, CO₂ER, and the integrated performance of the CCHP system for the two buildings in the two modes averagely decrease 9.2%, 8.2%, 5.1%, and 7.4%, respectively, when the cooling load increases by 100%. The primary energy consumption of the CCHP system will more markedly increase. The decreasing range of the CO₂ER is the least when the building needs more cool.

When Scenario D₁ is compared with Scenario D₄, it is seen that the performances of CCHP systems for hotel decrease with the increase of heating load while the performances of office increase. It is because that the heat/electricity ratio of hotel is larger than that of the PGU while the office’s ratio is less than that of PGU. For Scenario D₁, increasing appropriate heating demand will improve the utilization efficiency and there are not excess energy consumption to satisfy the energy demand. Additionally, the change trends differ when the building needs more heat or more cooling. When the heating load is doubled, the PES and the ATCS of the CCHP systems for the two buildings in the two operation modes averagely decrease 0.1% and 1.5%, respectively, while CO₂ER and the integrated performance increase 0.4% and 2.9%, respectively. Compared to the changes, 9.2%, 8.2%, 5.1%, and 7.4%, when the cooling increases, it can be seen that the impact of the cooling load on the saving performance is greater than the heating load. It is mainly determined by the COP difference between electric chiller and absorption chiller. When both the cooling load and the heating load decrease, the performance difference between Scenario D₁ and Scenario D₂ is similar to the difference between Scenario D₁ and Scenario D₄.

Additionally, the capacities of CCHP systems in EDM and TDM operation modes are compared. From Table 8, it can be found that the capacity of the PGU in EDM mode is greater than or equal in TDM mode. Taken Scenario D₁ as an example, the performance in EDM is better than in TDM mode except the PES in office (the explanation about the PES in office can be found in the latter section). The CO₂ emission factor of electricity from grid is greater than the supplemental gas. Consequently, the CO₂ER in EDM mode is greater than in TDM mode. The economic cost with the capacity of the CCHP system is vaulted relationship.⁴⁴ The ATCS begins to increase with the increase of capacity, reaches the maximum, and then decreases. The capacity in EDM mode is greater than in TDM mode and the ATCS in EDM mode is also greater than in TDM mode, which means that the optimal capacity locates the decreasing curve. Furthermore, the impacts of thermal demand of building on CCHP performance in EDM and TDM operation modes are compared. The impact in TDM mode for the hotel is greater than in EDM mode while the office is just contrary to the hotel. The reason is that the heat/electricity ratio of hotel is greater than that of PGU while the ratio of office is less.

When the components of the integrated performance in Figure 5 are compared, it can be noted that there are some negative values in these 16 scenarios while the CCHP system always reduces CO₂ emission in comparison to the SP system. The outstanding CO₂ER is the result that the CO₂ emission factor of gas is less than that of electricity from grid. From equations (18) and (19), it can be found that the CCHP system reduces CO₂ emission only when ${CO}_{2} ER$ is a positive value. The farther deduction can get the condition of reducing CO₂ emission as follows:

\frac{μ_{{CO}_{2}, f}}{μ_{{CO}_{2}, e}} \geq \frac{E_{grid}^{SP} - E_{grid}}{F_{on - site} - F_{b}^{SP}}

(22)

The left part is the ratio of CO₂ emission factors of fuel and electricity produced by the SP system. The right part is the ratio of electricity consumption’s difference of the CCHP system and the SP system and fuel difference. The CCHP system can really exhaust less CO₂ than the SP system only when the inequation in equation (22) is satisfied. After the input fuel and the electricity grid are determined, the left part is then fixed and unchanged. Therefore, the operation strategy of the CCHP system and the energy demands of building mainly influence the environmental benefit achieved by the CCHP system in comparison to SP system.

To show the performances of CCHP system more clearly, Scenario D₁ of hotel and office is selected to analyze the monthly performances. The monthly varieties of PES, ATCS, CO₂ER, and the integrated performance of CCHP system are displayed in Figure 6, where (a) represents EDM mode and (b) TDM mode.

Figure 6.

Monthly performance of the CCHP systems in comparison to the SP systems for hotel and office in EDM and TDM operation modes.

When the performances of the hotel between EDM and TDM are compared, the trends of these performance curves are almost same, and they fluctuate with the ratio of heat to electricity. The trends of the PES and the CO₂ER of the two operation modes are similar except to the ATCS. In summer, the primary energy consumption and the CO₂ emission both are the most during the entire year and the energetic and environmental performances of CCHP system is the worst. Especially, there are some negative values of the PES in the summer, which means that the CCHP system cannot really save energy in comparison to the SP system. The ATCS during the transitional seasons is the lowest. From the comparison of annual performance, the corresponding difference between EDM and TDM are 2.5%, 1.1%, 4.9%, and 2.8%. It can be concluded that EDM mode of CCHP system for the scenario hotel is better than TDM mode.

When the performances of Scenario D₁ office between EDM and TDM are compared, the trends of these curves are also almost same. Only there is a little difference in the PES variations. But the performance curves of the office are very different to that of the hotel. The ATCS and the CO₂ER in the transitional seasons are the worst while the PES in the summer is the worst. From the comparison of annual performance, the corresponding difference between EDM and TDM are −3.4%, 4.9% 5.4%, and 2.3%. From the integrated performance of CCHP system, it can be also concluded that EDM operation mode is better than TDM. But the CCHP system in TDM mode for the office saves less energy than in EDM mode.

Similar to the fact that the CO₂ER of CCHP system is the most outstanding in Figure 5, it can be seen the curves of the monthly CO₂ER are located on top of other performance curves and the environmental benefit achieved by the CCHP system is the greatest. Because the emission factor of the CCHP system is much less than that of electricity produced by the SP system, the CCHP system always reduces CO₂ emission in comparison to the SP system for the 16 scenarios. When the monthly performance for the hotel and the office is compared, the benefits achieved by the CCHP system in comparison to the SP system from energy, economy, and environment are much different. The performance is closely related to the energy demands’ characteristic of building and the heat/electricity ratio. Generally, the adoption of the CCHP system in winter is more profitable to save primary energy and reduce CO₂ emission.

Discussion on the excess products from CCHP system

In the above analysis, the excess electricity and heat are not allowed to be sold back to grid or other users. To analyze the potential benefits of CCHP system, the excess products are allowed to be sold to other users. These CCHP systems for the 16 scenarios produce excess electricity in TDM mode or excess heat in EDM mode. The excess products of the CCHP systems are calculated and summarized in Table 9. The excess ratios (here the excess ratio is defined to the ratio of excess product to primary energy consumption) are listed in Table 9. It can be seen that the excess ratios in EDM mode of office scenarios are 19.95%, 20.60%, 12.31%, and 13.73%, respectively, and much larger than in TDM mode. Therefore, all annual PES of office scenarios in EDM mode are less than in TDM mode. However, it can be found that the ATCS in EDM mode are more than in TDM mode in Figure 5. It is because that the supplemental electricity from grid in TDM mode is much so that the operation cost is more than in EDM mode. The CCHP system for the office in EDM mode saves less primary energy and more total cost than in TDM mode. The CCHP system in TDM saves more primary energy while needing more cost. Consequently, the feasibility of the CCHP system must be determined by the weighted performances. The weights of criteria play vital roles to the integrated performance. In this paper, the weights of three criteria are same to reflect the same importance from energy, economy, and environment.

Table 9.

The excess products of CCHP systems for the 16 building scenarios.

Scenarios	Excess electricity (kWh/year)	Excess heat (kWh/year)	Excess ratio (%)	ATCS (%)
Scenarios	Excess electricity (kWh/year)	Excess heat (kWh/year)	Excess ratio (%)	0	20	50	80	100
Scenario D ₁ _,hotel _, _EDM	0	3.0045 × 10⁵	5.81	5.20	10.43	18.28	26.13	31.36
Scenario D ₁ _,hotel _, _TDM	1.8446 × 10⁵	0	3.47	4.11	6.00	8.84	11.68	13.57
Scenario D ₂ _,hotel _, _EDM	0	6.5408 × 10⁵	17.58	9.21	25.40	49.68	73.96	90.15
Scenario D ₂ _,hotel _, _TDM	0.50947 × 10⁵	0	1.44	8.23	9.03	10.24	11.44	12.25
Scenario D ₃ _,hotel _, _EDM	0	2.4443 × 10⁵	3.67	−6.46	−3.11	1.92	6.96	10.3
Scenario D ₃ _,hotel _, _TDM	2.0428 × 10⁵	0	2.97	−6.97	−5.32	−2.84	−0.36	1.29
Scenario D ₄ _,hotel _, _EDM	0	1.7267 × 10⁵	2.37	6.62	8.89	12.29	15.70	17.97
Scenario D ₄ _,hotel _, _TDM	2.3122 × 10⁵	0	3.05	5.71	7.48	10.13	12.78	14.55
Scenario D ₁ _,office _, _EDM	0	4.6783 × 10⁵	19.95	6.26	20.73	42.44	64.16	78.63
Scenario D ₁ _,office _, _TDM	0.79773 × 10⁵	0	3.53	1.35	3.07	5.66	8.25	9.97
Scenario D ₂ _,office _, _EDM	0	4.0136 × 10⁵	20.60	6.37	22.70	47.20	71.70	88.02
Scenario D ₂ _,office _, _TDM	0.21443 × 10⁵	0	1.18	1.14	1.75	2.66	3.57	4.18
Scenario D ₃ _,office _, _EDM	0	3.6113 × 10⁵	12.31	0.29	9.23	22.62	36.02	44.95
Scenario D ₃ _,office _, _TDM	1.6075 × 10⁵	0	3.66	−2.58	−0.73	2.04	4.81	6.65
Scenario D ₄ _,office _, _EDM	0	3.7983 × 10⁵	13.73	7.06	16.63	30.98	45.33	54.90
Scenario D ₄ _,office _, _TDM	1.0167 × 10⁵	0	3.70	3.69	5.48	8.17	10.85	12.64

The cost of heat is calculated to 1.379 Yuan/kW, and the selling price of the electricity is same as the buying price in Table 7. However, the selling price is usually lower than the buying price. Aimed to discuss the impact of the selling price on the ATCS, the selling prices of excess products are 20%, 50%, 80%, and 100% of buying price. The ATCS of these 16 scenarios are recalculated and shown in Table 9. It can be seen that the ATCS increases with the increase of the selling price, and especially the increasing extent in EDM operation mode is greater than in TDM mode. Some negative ATCS values change to be positive when the excess products can be sold back to grid or other heat users, such as Scenario D₃. Therefore, the energy policy allowing the electricity generated by the CCHP system to be sold back to the grid, and the technologies storing the excess electricity or heat are helpful to improve the performance of CCHP system.

Discussion on the different weights in GA optimization procedure

The above performance analysis is based on the optimization results in GA. Moreover, the weights in the objective function, equation (21), influence the optimization results directly. If the different importance is paid attention to the three characteristics including energy, economic, and environment, the performances of CCHP systems for different building will be different largely. To find the impact of the weights on the performance, Scenario D₁ of the hotel and the office in two operation modes is selected to analyze and discuss.

The optimization results, the investment costs, and the performances of CCHP systems when the criteria weights are different are summarized in Table 10. It can be found that the PGU capacities are different largely when the weights are different, especially the PGU capacity with others when the economic characteristic is the most important. Compared the data in Table 10, it can be seen that the optimal PGU capacities from the energetic and environmental aspects are both larger than that from the economic criteria. When the energy-saving and CO₂-emission-reduction are paid more attention than the economic performance, the higher investment is required and the ATCS of CCHP system in comparison to SP system is the minimum. Therefore, the PGU capacity should not too large from the point of the investment cost. The investment cost increases with the increase of the installation capacity of PGU and the relationship is shown in Figure 7. The investment cost changes almost linearly with the PGU capacity no matter which mode the PGU runs.

Figure 7.

The relationship between the PGU capacity and the investment cost of Scenario D_1.

Table 10.

The capacities, installation costs and performances of CCHP systems when the optimization objective functions are different.

Scenarios	Criteria weights ( $ω_{1}, ω_{2}, ω_{3}$ )	PGU F_max (kW)	Investment cos (×10⁶ Yuan)	PES (%)	ATCS (%)	CO₂ER (%)	Integrated performance (%)
Scenario D ₁ _,hotel _, _EDM	(1/3, 1/3, 1/3)	764	4.26	11.0	5.2	28.6	14.9
	(1,0,0)	1348	5.12	11.5	−0.9	29.3	13.3
	(0,1,0)	552	3.95	9.6	6.5	26.6	14.2
	(0,0,1)	1320	5.08	11.5	−0.6	29.3	13.4
Scenario D ₁ _,hotel _, _TDM	(1/3, 1/3, 1/3)	569	3.97	8.5	4.1	23.7	12.1
	(1,0,0)	565	3.97	8.5	4.1	23.7	12.1
	(0,1,0)	423	3.76	7.8	4.8	22.0	11.5
	(0,0,1)	625	4.05	8.4	3.5	23.9	11.9
Scenario D ₁ _,office _, _EDM	(1/3, 1/3, 1/3)	939	3.23	4.1	6.3	27.5	12.6
	(1,0,0)	1092	3.49	4.5	3.2	28.0	11.9
	(0,1,0)	755	2.95	3.0	7.8	25.3	12.0
	(0,0,1)	1081	3.47	4.5	3.3	28.0	12.0
Scenario D ₁ _,office _, _TDM	(1/3, 1/3, 1/3)	867	3.12	7.5	1.4	22.1	10.3
	(1,0,0)	977	3.28	7.9	−0.3	22.7	10.1
	(0,1,0)	426	2.47	2.3	3.9	15.0	7.1
	(0,0,1)	998	3.31	7.9	−0.6	22.7	10.0

Although the optimal capacities are largely different and the investment cost are also different, but there is a little difference between the integrated performances of CCHP system in comparison to SP system. The maximum difference of the integrated performance between the economic optimization and the integrated optimization reaches 3.2%, and others generally are less 1.0% except to the difference between the integrated optimization and the energetic optimization. As a whole, the environmental potential is the most remarkable and the economic performance is the lowest and even there is no economic benefit between these three characteristic benefits of CCHP system in comparison SP system.

Conclusion

This paper presented performance analysis and comparison of CCHP systems optimized by GA for different buildings. This analysis leads to the following conclusions that are more applicable to the PGU unit with a heat/power ratio of 2.4.

GA is a valid optimization method for CCHP system that needs no initial information and searches the global optimization solution. GA operates parallel from multi-points, overcomes the search blindness, and accelerates the search speed. The GA optimization method can be extended to cover some of the omissions mentioned above as long as the calculation of the fitness function is able to be finished.

The performance of CCHP system is greatly related to the ratio of cooling load to electric demand and the ratio of heating load to electric demand. The appropriate ratio of heat to electricity of building is helpful to improve the performance of CCHP system when the excess electricity is not allowed to be sold back to grid in EDM operation mode or the excess heat is directly exhausted in TDM operation mode.

The optimal operation mode of CCHP system is determined by many factors, such as the technical parameters of CCHP system, the energy demands of building, and the price of energy. The optimal capacity of CCHP system in EDM mode is greater than in TDM mode. Consequently, the capital cost in EDM mode is more. The excess ratios in EDM mode of scenarios having lower ratio of heat to electricity are much larger than in TDM mode, which would not really save the primary energy. When the excess products could be sold to other users, ATCS in EDM operation mode is greater than in TDM operation mode.

The optimal result and the performance change with the criteria weights in the optimization procedure of CCHP system. Generally, the environmental potential of CCHP system in comparison SP system is the most remarkable, and the economic performance is the lowest between these three characteristic benefits. The investment cost increases with the increase of the installation capacity of PGU. When the cooling demand of the building is much more than the heating demand, the investment cost will increase dramatically.

Footnotes

Acknowledgements

The authors wish to express their gratitude to the anonym referees for remarks and suggestions that improved this paper significantly.

Funding

This research has been supported by the Fundamental Research Funds for the Central Universities (13MS95), Beijing Natural Science Foundation (3122028), Special Funds for Excellent Doctoral Dissertation of Beijing (20121007901), and Hebei Natural Science Foundation (E2012502002).

Declaration of conflicting interests

None declared.

Appendix 1

References

Wang

. Combined cooling, heating and power: a review. Progr Energ Combust Sci 2006; 32: 459−495–459−495.

Chicco

Mancarella

. Distributed multi-generation: a comprehensive view. Renew Sustain Energ Rev 2009; 13: 535–551.

Joel

Augusto

. Trigeneration: an alternative for energy savings. Appl Energ 2003; 76: 219–227.

Wang

Jing

Zhang

. Integrated evaluation of distributed triple-generation systems using improved grey incidence approach. Energy 2008; 33: 1427–1437.

Wang

Jing

Zhang

. A fuzzy multi-criteria decision-making model for trigeneration system. Energ Policy 2008; 36: 3823–3832.

Medrano

Brouwer

McDonell

. Integration of distributed generation systems into generic types of commercial buildings in California. Energ Build 2008; 40: 537–548.

Wang JJ, Jing YY, Zhang CF, et al. Distributed combined cooling heating and power system and its development situation in China. In: Proceeding of the ASME 2nd international conference on energy sustainability, Jacksonville, Florida, USA, 10–14 August 2008.

McCrorie

KAB

Underwood

Feuvre

RFL

. Small-scale combined heat and power simulations: development of a dynamic spark-ignition engine model. Build Serv Eng Res Technol 1996; 17: 153–159.

Moran

Mago

Chamra

. Thermoeconomic modeling of micro-CHP (micro-cooling, heating, and power) for small commercial applications. Int J Energ Res 2008; 32: 808–823.

10.

Fumo

Mago

Chamra

. Analysis of cooling, heating, and power systems based on site energy consumption. Appl Energ 2009; 86: 928–932.

11.

Fumo

Mago

Chamra

. Cooling, heating, and power energy performance for system feasibility. J Power Energ 2008; 222: 347–354.

12.

Mago

Fumo

Chamra

. Performance analysis of CCHP and CHP systems operating following the thermal and electric load. Int J Energ Res 2009; 33: 852–864.

13.

. Theoretical research of a silica gel-water adsorption chiller in a micro combined cooling, heating and power (CCHP) system. Appl Energ 2009; 86: 958–967.

14.

Fredrik

Jin

. Case study of energy systems with gas turbine cogeneration technology for an eco-industrial park. Int J Energ Res 2008; 32: 1128–1135.

15.

Gao

Jin

. System study of combined cooling, heating and power system for eco-industrial parks. Int J Energ Res 2008; 32: 1107–1118.

16.

Chicco

Mancarella

. Matrix modelling of small-scale trigeneration systems and application to operational optimization. Energy 2009; 34: 261–273.

17.

Calva

Núñez

Toral

MAR

. Thermal integration of trigeneration systems. Appl Therm Eng 2005; 25: 973–984.

18.

Sepehr

Masoud

Maziar

. Optimal design of gas turbine CHP plant with preheater and HRSG. Int J Energ Res 2009; 33: 766–777.

19.

Piacentino

Cardona

. On thermoeconomics of energy systems at variable load conditions: integrated optimization of plant design and operation. Energ Convers Manage 2007; 48: 2341–2355.

20.

Cardona

Piacentino

. Optimal design of CHCP plants in the civil sector by thermoeconomics. Appl Energ 2007; 84: 729–748.

21.

Rong

Lahdelma

. An efficient linear programming model and optimization algorithm for trigeneration. Appl Energ 2005; 82: 40–63.

22.

Piacentino

Cardona

. EABOT – energetic analysis as a basis for robust optimization of trigeneration systems by linear programming. Energ Convers Manage 2008; 49: 3006–3016.

23.

Kong

Wang

. Optimal operation of a micro-combined cooling, heating and power system driven by a gas engine. Energ Convers Manage 2009; 50: 530–538.

24.

Shi

Huang

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

25.

Arcuri

Florio

Fragiacomo

. A mixed integer programming model for optimal design of trigeneration in a hospital complex. Energy 2007; 32: 1430–1447.

26.

Tozer

James

. Absorption chillers applied to CHP systems. Build Serv Eng Res Technol 1995; 16: 179–188.

27.

Kong

Wang

Huang

. Energy efficiency and economic feasibility of CCHP driven by stirling engine. Energ Convers Manage 2004; 45: 1433–1442.

28.

Bruno

Ortega-López

Coronas

. Integration of absorption cooling systems into micro gas turbine trigeneration systems using biogas: case study of a sewage treatment plant. Appl Energ 2009; 86: 837–847.

29.

Costa

Paris

Towers

. Economics of trigeneration in a kraft pulp mill for enhanced energy efficiency and reduced GHG emissions. Energy 2007; 32: 474–481.

30.

Cardona

Piacentino

Cardona

. Matching economical, energetic and environmental benefits: an analysis for hybrid CHCP-heat pump systems. Energ Convers Manage 2006; 47: 3530–3542.

31.

Henderick

Williams

. Trigeneration in a northern Chinese village using crop residues. Energ Sustain Dev 2000; 4: 26–42.

32.

Rentizelas

Karellas

Kakaras

. Comparative techno-economic analysis of ORC and gasification for bioenergy applications. Energ Convers Manage 2009; 50: 674–681.

33.

Deng

Wang

. Exergy cost analysis of a micro-trigeneration system based on the structural theory of thermoeconomics. Energy 2008; 33: 1417–1426.

34.

Cardona

Piacentino

. A new approach to exergoeconomic analysis and design of variable demand energy systems. Energy 2006; 31: 490–515.

35.

Williams

Griffiths

Knight

. Combined heat and power: sizing plant for new hospitals. Build Serv Eng Res Technol 1996; 17: 147–152.

36.

Tozer

Serrano

MAL

Capilla

. Thermoeconomics applied to an air conditioning system with cogeneration. Build Serv Eng Res Technol 1996; 17: 37–42.

37.

Alanne

Salo

Saari

. Multi-criteria evaluation of residential energy supply systems. Energ Build 2007; 39: 1218–1226.

38.

Cardona

Piacentino

. A methodology for sizing a trigeneration plant in mediterranean areas. Appl Therm Eng 2003; 23: 1665–1680.

39.

Geng

. Energy utilization evaluation of CCHP systems. Energ Build 2006; 38: 253–257.

40.

Huang

. Influence of energy demands ratio on the optimal facility scheme and feasibility of BCHP system. Energ Build 2008; 40: 1876–1882.

41.

Chicco

Mancarella

. Assessment of the greenhouse gas emissions from cogeneration and trigeneration systems Part I: models and indicators. Energy 2008; 33: 410–417.

42.

University DdgiT. Building environmental system simulation and analysis – DeST, Beijing: China Architecture & Building Press, 2006.

43.

Wang

Jing

Zhang

. Optimization of capacity and operation for CCHP system by genetic algorithm. Appl Energ 2010; 87: 1325–1335.

44.

Ren

Gao

Ruan

. Optimal sizing for residential CHP system. Appl Therm Eng 2008; 28: 514–523.

45.

Gibbs MS, Maier HR, Dandy GC, et al. Minimum number of generations required for convergence of genetic algorithms. In: Proceedings of the 2006 IEEE congress on evolutionary computation, Vancouver, BC, Canada, 2006.

46.

Understanding TradingSolutions Concepts: overview, http://www.tradingsolutions.com/webhelp/concepts/concept_overview.html (2001, accessed 5 July 2013).