Abstract
An economic feasibility optimization method for the life cycle cost (LCC) has been developed to apply energy saving techniques in the early design stages of a building. The method was developed using default data (e.g., operation schedules), energy consumption prediction equations and cost prediction equations utilizing design variables considered in the early design phase. With certain equations developed, an LCC model was constructed using the computational program MATLAB, to create an automated optimization process. To verify the results from the newly developed assessment tool, a case study on an office building was performed to outline the results of the designer’s proposed model and the cost optimal model.
1. Introduction
1.1. Introduction
Korea is the world’s fifth largest oil importer and it lacks fossil fuel reserves. Oil imports accounts for a large proportion of overall imports of Korea, and thus, fluctuating oil prices greatly affect the overall balance of payments [1]. The country’s continued reliance on depleting fossil fuel energy sources may also undermine its energy security, and as a result the government plans to decrease fossil fuel dependence in the long run. The government encourages buildings to use less fossil fuel by setting energy policies, regulating the minimum U-value and advocating the use of renewable energy sources. The government is also offering incentives and tax benefits. However, building owners and builders manage only to meet the regulations and have no interest to exceed these minimum targets. This is attributed to green premium, the initial cost that rises with use of the technologies necessary to build a nearly Zero Emission Building (nZEB), which occur in the construction stage. Decision makers simply link the green premium and the subsequent rise of the initial cost to a reduced economic benefit.
In fact, energy performance of a building which has been improved by green premium and the economic benefit do not have a conflicting but a causal relationship This is due to the fact that the time and money needed to operate a building is quite substantial and investing in green premium can considerably reduce the operational costs in the operating stage. This is particularly true when the oil price is expected to increase, and with the additional cost of CO2 emissions, the green premium effect is expected to further increase [2].
Decision makers have yet to fully recognize the life cycle concept incorporating the green premium and the operating costs comprehensively. Insufficient technologies and experts to evaluate their economic effect also limit the information that can be accessed. Though the outline of the initial investment is decided at the initial design stage, there is no system that supports decision making at that stage and this causes hesitation in people to invest in the green premium.
Therefore, providing information on the economic advantages coming from the green premium can be a solution in reducing the obstacles to realize a nZEB. As it is complicated to consider all of the energy consumed and the economic benefits generated by a building during its life cycle, an assessment tool must to be developed that would help general decision makers easily grasp the effects [3].
The current study attempted to develop an assessment model which can be used to calculate the LCC (initial investment cost and energy cost) and thereby assist in decision making during the early design phase. The assessment tool which was developed should operate together with a program that can draw an optimized design. The energy saving design for a building envelope typically takes into account different design variables, such as heat insulation, natural lighting and natural ventilation in the early design stage. The energy saving technologies trade off their heating and cooling load, and requires many cases to be designed and evaluated. This process takes time and repeated trials and is therefore omitted in the early design phase most of the time. A tool is required which can be connected to an auto–optimized design process that helps to generate the cost-optimal design without going through a large number of iterations. Therefore, a supportive decision making tool was developed in the present study.
1.2. Research Scope
A baseline building was selected for the present study to develop a cost optimization model and 153 office buildings in Seoul were surveyed to validate the model. The scope for applying each design variable was set, according to current legal requirements for minimum performance and the latest technology available in the market for the maximum performance.
The study developed a simple prediction equation that allowed calculation of energy consumption and LCC based on the baseline building. This was derived to be easily operated by any architect, as it comprised design variables considered in the early design stage. To develop the energy prediction equation, an orthogonal array was used to drastically reduce the number of experiments to 81 from 324 (=282,429,536,481), the number required for energy simulation. EnergyPlus was used for the computer simulation and the relative importance of each design variable was identified using Analysis of Variance (ANOVA). The ANOVA result was used as multiple regression analysis data to develop the prediction equation.
To calculate initial investment in the LCC model, a prediction equation was also developed. To develop the prediction equation, several cases were established where technological elements designed to improve energy performance, were designed for the baseline model before entering quotation data. A regression formula was then generated through statistical analysis of the calculated data. The equation parameters were made to correspond to the energy prediction equation as far as possible and important factors were included to the LCC which would not affect the energy performance.
The LCC model was constructed by using the energy performance prediction equation and the initial investment cost equation developed above. Based on this LCC model, an optimization model was developed by applying a multi-objective optimization method. Finally, a program was constructed and verified.
2. Literature Research
2.1. Definition and Research of Nearly Zero Emission Building
The concept of “Zero Emission” was first proposed by Pauli (United Nations University) in 1994, and it was meant to describe the process where a minimum amount of waste would be generated, eventually creating zero waste. “Zero Emission Building” in architecture means a building with the objective of creating zero CO2 emissions during the life cycle of the building. When it is explained that “Zero Emission Building” widened the scope of CO2 emitted by the building into three that are operation energy, embodied energy and transportation energy [4], scope 1 or the operation energy used by the building accounts for 70% of all CO2 emissions and has the greatest importance [5].
Buildings in South Korea emit more than 25% of all greenhouse gases. With the idea that there plenty of room to reduce the greenhouse gases by constructing green buildings, a great deal of effort has been made to reduce the greenhouse gas emissions from buildings. Besides, the Ministry of Land, Infrastructure and Transport and the Ministry of Trade, Industry and Energy have designated zero energy as a new industry and made joint efforts to realize the goal through cooperation as one of the key projects to create a future growth engine [6]. “Zero Energy Building” basically means a building where the sum of used energy and produced energy is 0 (Net Zero) [7]. Given the current technologies and economic feasibility, however, the policy defined it as a building that consumes the least energy (Nearly Zero, 90% reduction) [8], with maximized heat insulation performance, minimized building energy load (passive), use of renewables, such as photovoltaic (active) and minimum energy needs for proper functioning [9].
Therefore, studies carried out in the country mainly focus on realizing nearly zero emission by analyzing and combining individual elementary technologies. Song [10] analyzed the energy saving effect of air conditioning technology when applied using simulation while Gang [11] studied the effect of ventilation, eave and lighting control. Though the final goal of the above studies was to make a zero energy building, policy studies were also conducted to overcome the gap between the goal and reality. Policy wise, Shin [12] analyzed the energy demand and consumption depending on combination of technologies to bring the zero energy housing into reality. In the meantime, Jo studied the policies of South Korea and other countries to build zero energy buildings and examined the local policies [13].
2.2. Decision Making for Realization of Zero Emission Building
As the early design stage decides 80% of the initial investment and most of the design decided in this stage influences energy consumption during the operating stage [14], it can be said that the operating cost is decided here as well. Many decisions are made in the early design phase. A high-performance envelope designed with high performance insulation materials can help to reduce the heating costs as well as the operating costs. Building exterior shading can also decrease the cost of the cooling system and also the energy consumption for cooling. Exterior shading can however increase the heating cost as it is likely to block solar heat during winter, which raises the LCC. Accordingly, design variables should be analyzed comprehensively as they have varying impacts on the initial investment, cooling/heating costs and lighting costs.
Financial problem is one of the major obstacles in the realization of nearly Zero Emission Buildings (nZEBs). Some investors believe that increasing initial investment to enhance building performance results in an economic loss. However, it is reasonable to calculate the building cost with the life cycle cost (LCC) that includes both the initial investment cost and the operational cost, as investment to improve the energy performance should eventually lead to a lower energy cost during operation. Thus, the return on investment calculation should be related not only to the investment costs, but should also take into account the operational and maintenance costs of buildings [15]. Reaping the most benefit out of the least investment cost is important to consider. In this concept, the “lowest LCC” option, which pertains to the entire life cycle of a building is the most economical one [16]. Recently, the most economical condition is also called cost optimal, with investors wanting their investment to be cost optimal.
Calculating the optimal LCC at the early design phase requires calculation of the initial investment and operational costs and this also calls for a significant investment of time and money in a database. This suggests that the most early design phases end with only a small portion of the design space examined and a large number of design solutions left untouched. There is therefore no guarantee that the final design will be an optimal one. Besides, it is almost impossible in Korea to make the best decision based on such analysis in the early design phase as a quick decision is usually required. If the optimization includes an automatic process of design creation, simulation and evaluation, building design optimization will become beneficial as it will help to create a design quickly which can realize the performance that the building seeks for and provide an overall knowledge of the whole design space [17]. The optimal design created by an automatic process facilitates the designer’s decision making that considers the reality of nZEB [18]. It also acts to relieve the designer of a very time-consuming and labor-intensive trial and error approach, which designs and evaluates a number of buildings to reach the same goal [19].
In the study of optimal life cycle cost, Wang [20] developed an LCC model, which considered cost effective design rather than only the energy performance in a single-objective function. The LCC here, which reflected a trade-off between the initial investment and the operating cost, was more advanced than other previous models. Znouda [21] developed a standard GA which included a tool (CHEOPS) that allowed load calculation for a simple building. This model produces both energy consumption and a LCC optimal model. Typically, the economic impact (LCC) and the environmental impact (energy consumption) were used as design goals in the multi-objective optimization model. Hasan et al. [22] combined the IDA ICE 3.0, energy simulation tool, and the GenOpt2.0, genetic algorithm program, to develop an LCC design process. In the multi-objective optimization model developed by Shi (2011) [23], EnergyPlus, a sophisticated and widely used BEPS software tool, was integrated into mode FRONTIER, an optimization suite which contains a series of pre-programmed optimization algorithms, to minimize the energy demand for space conditioning and insulation usage for office buildings in Southeast China. With only one insulation material, the insulation usage ishas essentially an equivalent term to the initial cost.
2.3. Tools for Fast Decision Making in the Early Design Phase
Most of the recently developed optimal LCC tools are calculated based on a very detailed database which includes the finishing paint coating and type of stone finishing [24,25]. It is quite difficult to use an optimal LCC tool that uses a detailed database at the early design stage. Therefore, data which cannot be determined and obtained at early design stage to calculate the LCC must be included in the tool [26]. This helps to minimize both time and effort needed to calculate the initial investment and operating cost to give results quickly, thereby resulting in quick decision making. Moreover, the fast automatic process of the optimized LCC analysis presents the optimal LCC or the lowest LCC that can be drawn using the given variables, providing a guideline for the designer.
3. The Framework of Decision Supporting Tools in the Early Design Phase
The basic framework of the decision supporting tools was established with an input module, an assessment module and an output module and core elements which were derived from the assessment tool comparative analysis [27].
3.1. Input Module
The items to calculate for the LCC analysis should be decided considering the analysis target. The initial investment can largely be divided into items which influence the building’s operating stage, i.e., energy consumption and those not. Considering the goal of the study, the items that did not affect the energy performance had a relatively low influence. Hence, these were not included in the LCC analysis and the initial investment that was entered into the input module was set up. The initial investment cost usually comprises civil engineering, architecture, mechanical system, electric system, landscape and interior construction. Out of these, civil engineering, landscaping and interior construction were excluded because their impact on the energy performance is relatively insignificant. In the electric construction, only the lighting system needs to be considered and should be addressed in the early design phase.
3.1.1. Architecture
Extracting the variables that affect the energy consumption and cost is important. Architectural design factors at the early design phase include building shape, size, window/wall ratio, envelope and structure. Though the number of deciding factors is not large, they interact with one another, which in turn further affects the energy performance. Therefore, each factor needs to be systematically classified into two categories [28]. One is the factors which affect both the energy consumption and the architectural design, whilst the other has no relevance to the design but affects energy consumption. The factors affecting the design have to be further sorted first through literature review. After the sorting process, a sensitivity analysis should be performed to select the critical input parameters. The parameters selected through statistical analysis have relatively greater influence or sensitivity on the energy consumption than others. After sensitivity analysis, the parameters which had larger ratios were selected. A factor influencing the energy, though it is not one of the design factors, was also checked, as it had to be entered to predict the energy consumption. It is particularly important to pre-set the factors that affect energy consumption as a default database for prediction used at the early design phase even if they are not any design factor. The factors that do not include design factors but must be input are the internal loads, ventilation rate, indoor temperature set points and operating hours. If these inputs are set up as standard information, time and efforts to find and input them can be greatly reduced.
3.1.2. Mechanical and Electrical System
For mechanical systems, a Heating, Ventilating, Air-Conditioning, and Refrigeration (HVAC&R) system should be of primary interest as it is a complicated system with numerous variables and restrictions. According to past research, the HVAC&R system matrix has a combination of 16 AC sub-systems, 15 plant sub-systems and 4 transportation systems. The number of possible combinations that can be made is therefore 960 [29]. However, it was narrowed down to the realistic and universal systems which are used in Korea. Thus, a total of 56 sub-system combinations (4 × 14) can be produced. The present study used these 56 basic HVAC&R database sets to calculate the energy consumption of the HVAC&R system. In the electric system, only lighting system that can be decided in the early design phase needs to be considered and others should be set to a default value [30].
3.2. Assessment Module
A LCC process usually includes the following steps: definition of LCC analysis/identification of the goals, establishment of a basic assumption for the analysis, selection methodology for the LCC analysis (e.g., cost breakdown structure, identifying data sources and contingencies) and LCC modeling [31].
3.2.1. Definition of the LCC Analysis/Identification of Goals
Life cycle cost (LCC) is an approach that assess the total cost of an asset over its life cycle including initial investment, operating cost, maintenance costs. Nowadays, most builders are only concerned about the initial investment, working toward their minimization [32]. This leads to an emphasis on the initial cost, in detriment of other life cycle cost, and, for the same cases, to the supporting of solutions that require smaller investment but have higher operational costs (such as the application of less insulation resulting in higher need for heating and cooling energy) and also lower Zero Emission levels (like higher carbon emission).
Life cycle cost is an economic methodology for selecting the most cost-effective alternative over a particular time frame, taking into consideration its initial cost (construction), operational cost and maintenance cost. However, the material cost and installation cost of the initial investment as well as the energy operating cost are the most influential in reflecting the energy saving strategies into the building design. Therefore, only the costs of these two stages of the life cycle were calculated to analyze the LCC in this study (Figure 1).
Figure 1.
Scope of the Life Cycle Cost Analysis.
3.2.2. Establishment of the Basic Assumptions for Analysis
A discount rate should be established to calculate operating cost, with interest rate (10 years on average), oil, electricity and gas prices considered in the calculation [33]. South Korea is also expected to face carbon price soon which is not yet decided in the market. Therefore, this study employed a price from the CO2 emissions trading market of the EUA (European Union Allowance) [34].
3.2.3. Selection of Methodology for the LCC analysis
The process of predicting the LCC requires calculation of the costs that occur throughout the period in question at a certain time. The global cost method [35] was used to calculate the LCC. This calculation method includes a system that affects the building’s energy performance. In this study, the construction and facility costs that affect the energy performance were included in the calculation. All costs were deemed to occur when they are calculated. The global cost calculation method is as shown below:
where Cp = Conversion Factor of Present Value; Ci = Initial Cost; Co = Operation Cost; Cm = Construction Cost; CE = Energy Cost = CC × (EC + EH + EL + EE + CEM); CEM = CO2 Emission Cost = CC × EEmission, where EC = Cooling Energy; EH = Heating Energy; EL = Lighting Energy; EE = Electric Energy; EEmission = CO2 Emission.
Life Cycle Cost (X) = Cp (Ci + Co) = Cp (Cm + CE + CEm)
The discount rate coefficient Rd was used to refer to the replacement cost and the final value to the starting year. It is expressed as:
where Rr is the real interest rate in the year of calculation. When the annual costs occur for many years, such as the running costs, the present value factor, which is expressed as a function of the number of years n and the interest rate Rr can be calculated using Equation (3):
This study used the present value method to calculate the value of all future costs at the current value:
PF: Present value of a lump sum; A: Annual Value; F: Future Cost; P: Present Worth Factor; ir: interest; n: Analysis Period.
3.2.4. LCC Modeling
Table 1 presents the LCC analysis procedure for a nZEB in which the CO2 emission cost was considered.
Table 1.
LCC Analysis procedure of nZEB.
3.3. Output Module
A linear scaling transformation was applied to normalize the performance assessment results. The method is suitable when different attributes are mixed. The attributes consist of decision making factors which are in different units and cases where a minimization standard (e.g., expense) and a maximization standard (e.g., profit) that should be applied are mixed. Under a complicated linear scaling transformation in particular, the value is transformed between the range of the maximum and minimum value of each attribute. This means that each attribute transforms precisely to a value between 0 and 1. Hence, a complicated linear scaling transformation was applied to this study. Under the assumption that E is economic feasibility. for example, the maximum value of E is obtained when the LCC is at its lowest. Ej is the performance value of E for the alternative j. Emax and Emin represent the LCC of the baseline model and the optimized model, respectively. This suggests that Emin is the lowest possible life cycle cost among the selected variables, which signifies the optimal cost. If Emax and Emin are normalized using complicated linear scaling transformation, Equation (5) is obtained. When each performance result is transformed to a dimensionless score through normalization, the output value lies between 0 and 1. In this case, 0 represents the standard model, whilst 1 represents the target building with maximum performance.
- -
- Establishing the minimum performance: Minimum performance means the lowest performance realized by the suggested model. Therefore, the LCC is the largest model and the value of the universal standard model that complies with the minimum legal requirements in general has the minimum performance.
- -
- Establishing the optimal performance: The optimal performance implies the best performance of the suggested model conditions, meaning that the building is cost-optimal with the lowest LCC under the given condition. Therefore, an optimization design method was applied and the economic performance resulting from its best solution was set as the optimal performance. The optimization design is determined by the objective function for many decision variables:
3.4. Framework for the Decision Tool for Use at the Early Design Stage
The basic framework is established with an input module, an assessment module and an output module and core elements (Figure 2). An assessment software program for nZEB is developed based upon this framework derived from a logical calculation flowchart.
Figure 2.
Framework of the Decision Tool.
4. Development of the Decision Support Tool
4.1. Input Module
4.1.1. Description of the Baseline Building
In Korea, it is mandatory to submit an “energy conservation plan” when applying for a construction permit to build an office building larger than 5000 m2. In this study, the building data of 158 typical office buildings with their design drawings and energy conservation plans were thoroughly examined among these data, the contents of 30,000–50,000 m2 of floor area are mentioned in this text. At this time, the Baseline is the building which is the Mean Average within the database.
The baseline buildings served as the basis for calculations applying the methodology. The baseline building was defined to reflect the typical buildings of the building stock of Seoul, Korea [36]. The characteristics of the baseline building were determined by careful examination of a typical design. Survey results showed that most of the office buildings in Korea are of basic module type consisting of a central building core. Figure in the Table 2 is a simplified model for simulation and Table 2 shows a description of the baseline model and operating features.
Table 2.
Brief Description of the Baseline Model.
4.1.2. Design Variables
In this study, the building data of 153 typical office buildings with their design drawings and energy conservation plans were thoroughly examined, and all the variables which contribute to building energy consumption were identified. Among them, 24 factors are selected as energy related design parameters based on sensitivity analysis. These variables are related to architects’ early design decision and, at the same time, are able to be defined in quantifiable objective values. A number of precedent researches were also examined to verify the validity of the findings [37,38,39] (Table 3). Also, the practically applicable ranges of each design variable in the building were set based upon investigated data, which encompassed performance from the minimum level regulated by the building code to cutting edge commercialized technologies.
Table 3.
Applicable ranges of the design variables.
4.1.3. Sensitivity Analysis
The parameters which have significant contribution at the 5% significant level (p-value) in ANOVA were selected as the contributing energy factors, and the non-significant terms (p-value greater than 0.05) were eliminated. Only nine parameters were selected as significant factors for heating energy consumption. For cooling energy consumption, 13 parameters were found to be significant. They are; Floor to Floor Height, Window Wall Ratio (Façade), Orientation, Window Wall Ratio (South), Window Wall Ratio R (North), Ventilation, U-value of the Wall, U-Value of Window (North), Solar Heat Gain Coefficient-South for heating energy. For cooling energy, Gross Floor Area, Surface Floor Ratio, Floor to Floor Height, Width Depth Ratio, Window Wall Ratio (East), Window Wall Ratio (West), Daylighting, U-Value of Wall, Window U-factor (South), Solar Heat Gain Coefficient (West), Solar Heat Gain Coefficient (South), Visual Light Transmittance (West), Visual Light Transmittance (South) will be selected (Table 4 and Table 5).
Table 4.
Coefficient (a) for Heating Load.
Table 5.
Coefficient (a) for Cooling Load.
4.2. Assessment Module
4.2.1. Initial Cost
Costs related to building elements with no influence on the energy performance, such as e floor covering and wall painting, while the costs that remain the same for all measures/packages/variants, e.g., earthworks and foundations, staircases, and demolition costs can be excluded. The costs must be market-based and coherent regarding the location and time of the investment costs, running costs, energy costs, and CO2 cost. Data sources can be an evaluation of recent construction projects, an analysis of standard offers of construction companies, or existing cost databases which have been derived from market-based data gathering.
(1) Architecture Cost Prediction Equation
It is not easy to estimate the construction cost. In particular, in the early stage of design, the cost is calculated excluding the design cost. This is because many undetermined design factors are generally present which affect the construction cost. Therefore, it becomes difficult to generate reliable cost estimates. In the current study, a construction cost prediction equation was developed for the early design stage using statistical methods and actual construction cost estimated data. The reliability of the prediction equation was increased by producing data through interpolation. Variables used in the construction cost prediction equation are the factors from the early design stage which affect energy consumption during the operating stage. The predicting equation was developed through the following four steps:
- -
- Estimation of a first-order model based on actual data volatility
- -
- Generation of model development data and step by step data interpolation
- -
- Development of an initial investment cost estimation model and equation
- -
- Correction of the model based on estimation results
- Estimation of the first-order model based on actual data volatility
The first-order model based on actual data volatility was estimated in the first step of the analytical process, namely validating the relationship between the construction cost and the various independent variables for the data enrichment or interpolation. Interpolation is essential for the estimation of the initial investment as data extraction for each condition is limited [40]. The construction variability based on the baseline floor area (1500, 2000 or 2500 m2) was the most fundamental and significant part and was set as a reference point for data interpolation done later. This variability arose due to a lack of measured values when changes in the independent variables occurred along with the changes in the baseline floor area. First, an estimation equation between the construction cost and the changes in the independent variable was derived. No data were available to represent the changes in the baseline floor area and the floor height. Therefore, a regression equation for changes in the construction cost according to changes in the floor height, and the relation between the change in the floor height and the floor area was applied at 1500 m2, 2000 m2 and 2500 m2. Figure 3 shows a graphical representation of the derivation of the first-order model for changes in the construction cost with changes in certain independent variables (Figure 3).
Figure 3.
First-order model estimation process based on volatility.
- Generation of model development data and step-by-step data interpolation
This step refers to the process of filling the areas which lacked data for specific conditions through a what-if analysis, using the model created in the previous step. A what-if analysis is a process of finding an outcome when values are altered with respect to the actual values. This kind of analysis is often applied in time-series forecasting models. However, since this model does not have time-series data, a linear model was applied [41]. The result from the previous first-order model estimation process was applied to perform data interpolation based on a construction cost relations estimation equation to populate the area for which data did not exist. The results obtained from the regression estimation equation were applied and data interpolation was performed in the following order: the baseline floor area, floor height, width depth ratio, window wall ratio, U-value of wall and wall insulation. This series of steps was applied sequentially and repeatedly (Figure 4).
Figure 4.
Model development data and step-by-step data interpolation.
- Development of an estimation model for the initial investment cost
Based on the integrated data extracted from the analysis process in phases 1 and 2, a linear model was established. Since time series factors were absent, the ARIMA model was not applied. Instead, a linear model, which has a high potential for expansion, was estimated [42]. The regression model was used as a basis for derivation of the Best Linear Unbiased Estimator (BLUE) from several linear models [43].
- Correction of the model based on estimation results
In this final step, the derived initial investment cost estimation model was examined and validated for any model adjustment. By hypothesizing a particular condition, discrepancies between the estimates derived from the model and the real values provided at the early design phase were calculated and examined. The average error rate was 3%–5%, and to improve this, the model was adjusted through random simulations of the estimated values by changing the intercept term and the beta coefficient. Thus, the model created from step 3 was adjusted to a final form to derive the initial investment cost model estimation equation (Figure 5).
Figure 5.
Final correction of the model based on estimation results.
- Prediction Equation
A mathematical model was developed through a multiple regression analysis (Table 6). Multiple regression analyses were conducted using the data set, and the load prediction equations were derived. The coefficient of determination (R2) between the regression models was 0.845 for the heating load prediction and 0.998 for initial cost prediction. As a result, the regression equations were found to have a considerable predictive power.
Table 6.
Coefficient of Equations of the Initial Cost Prediction.
The equation is y = (−467,611,530.097) + 599,921.127 × Gross Floor Area + 142,207,711.837 × Floor Height + 25,200,179.482 × Width Depth Ratio + (−73,285.833) × Window Wall Ratio (East) + (−85,622.281) × Window Wall Ratio (West) + (−119,988.105) × Window Wall Ratio (South) + (−113,804.377) × Window Wall Ratio (North) + (−112,897,138.380) × U-value of Wall + (−30,919,301.401) × U-value of Window.
(2) Mechanical System Cost Prediction Equation
This study used 56 basic HVAC&R database sets, as mentioned above. For the facility system construction cost for each set, the capital cost is calculated for baseline and basic HVAC&R system, which consists of typical floor, mechanical room, and riser construction cost. The construction cost is then used to predict 56 combinations of HVAC&R system initial construction cost (Figure 6) [44].
Figure 6.
Analysis of construction cost of HVAC system boundary.
(3) Electrical system
Assuming that electrical construction cost is rarely an influencing factor to be decided in the early design stage, the electrical construction cost was taken as a constant, except for the lighting system and the cost of the lighting system. The constant was 178,251 won/m2. The value for the lighting system and the cost of the lighting system was automatically selected from 48,413 won/m2, 33,889 won/m2 and 17,659 won/m2, depending upon the selected lighting density [45].
4.2.2. Operating Cost
The energy cost of operating was calculated considering the expected energy consumption, the energy price inflation rate, inflation rate and the discount rate that reflected the interest rate and operational period. The building load prediction equation was developed with the variables mentioned above and the energy consumption was calculated through a connection with the disclosed equipment system database.
(1) Building Load Prediction Equation
The energy load prediction equations developed in the present study can be easily utilized to estimate heating and cooling load of office buildings in the central climatic zone of Korea during the early design stage. The number of tests needed to analyze the sensitivity was 324 (=282,429,536,481), which was reduced to 81 using an orthogonal array. The tests applied an EnergyPlus simulation. The prediction equation was developed using ANOVA and multiple regression analysis (Table 7) [46]. The criterion used for the improvement of the prediction was the coefficient of determination (R2). R2 represents the square of the correlation between the predicted value and actual value. It is expressed as a decimal number between 0.00 and 1.00. 1.00 means perfect prediction in the model. The p-values were R2 = 0.753 for heating energy consumption and R2 = 0.602 for cooling energy consumption.
Table 7.
Equations for Energy Consumption Prediction.
(2) System Load Prediction Equation
Energy simulation is a considerably accurate tool, but in order to use this, a professional knowledge and detailed input variables are needed. Therefore, an easily and conveniently available tool is highly desirable at the early design phase for practical purposes.
To compensate for the drawbacks of energy simulation, a database was developed based on the comparison analysis of different combinations of Heating Ventilation Air Conditioning & Refrigeration (HVAC&R) systems. This database was opened to public. As shown in Table 8, 56 possible combinations (4 Air conditioning × 14 Plant system) were chosen as the most widely used systems in Korea [44].
Table 8.
Basic HVAC&R database set.
The energy consumed by operating the water heating and lighting systems, and the electrical distribution for the equipment use in the building was calculated by using data from an energy-consumption survey of 40 office buildings [47] (Table 9).
Table 9.
Lighting and Electrical Energy Consumed.
4.3. Output Module
Finding an optimal model with the given variables requires calculation with application of the LCC evaluation equation to the optimization methodology. To undertake this optimization methodology, objective function and constraints which are appropriate for the given question are required.
4.3.1. Optimization Methodology
A multi-objective optimization function should be applied that uses multiple local minima solutions to produce an optimal value of the building energy cost and initial investment. Solution Space incorporates the features for discrete variables [48,49,50]. In this context, a genetic algorithm (GA) was believed to be the most appropriate among the many optimization models for calculating the optimal value of the discrete variables [51]. The overall process of the GA has been shown below (Figure 7).
Figure 7.
Process of GA algorithm.
4.3.2. Objective Function of the LCC and Constraints
As indicated earlier, an objective function was considered in the model, namely minimization of cost and energy consumption. The objective function can be expressed as shown in Equation (6):
where, X1 = Life Cycle Cost = Cp(Ci + Co) = Cp(Ci + CE + CEm); Cp = Conversion Factor of Present Value; Ci = Initial Cost; Co = Operation Cost; CE = Energy Cost; CEm = CO2 Emission Cost.
Objective function is f(x) = min (X1)
4.3.3. Constraints
The constraints on the design variables essentially delineate the bounds of the feasible range for each variable, which is influenced by market availability of the relevant products and the applicable codes and regulations. Constraints on the selected design variables are shown in Table 10.
Table 10.
Constraints for the selected design variables.
5. Development of a Fast Decision Support Tool and Validation
5.1. Fast Decision Support Tool
The LCC Model was developed using an Excel-based program. Thus, the program can be executed in a very simple manner. Each module was divided into separate Excel sheets which consisted of an Input Module, an Assessment Module and an Output module. This tool was developed for use at the early design phase. The input module comprised only the variables considered at the early design stage to reduce the time of analysis. For the assessment module, a cost/energy prediction equation was developed that does not affect the energy but allows the users to calculate the output by simply adjusting the variables. A computational optimization model calculation program (MatLab) was linked with the output module (Excel sheet) to provide an automated optimization process.
In most cases, the proposed designs, which should meet actual land conditions, designer’s intentions and legal obligations, do not fit the optimization model condition. This optimization model will be used for designing and recognizing how far a proposed design lies from the optimized model. The baseline result which is shown in the output module will be used as starting point to measure how far the proposed design was improved over the baseline building. The LCC of the design proposed by the designer will probably be positioned between the LCC of the baseline building and that of the optimal building. The LCC that is the closest to the optimal building is the optimal design, that is, the lowest LCC. This can then serve as a yardstick to show what performance the proposed design can attain and also offer a guideline to the designer. As shown in Figure 8, the tool developed in the current study was output graphically when a design was compared with the optimized building and the baseline building (Figure 8).
Figure 8.
Fast Decision Support Tool developed using excel-based program (output module).
5.2. Case Study
The present study also verified the optimization model by running the model with a MatLab program. Case A was designed with the actual preference of the client (or designer), whereas case B was the model that allowed a significant LCC saving, which was obtained from the optimization model. Case B, despite its high initial investment cost, provided a higher economic efficiency in terms of life cycle with a lower energy cost than that of case A. When case B increased the efficiency by selecting a chilled beam to enhance the envelope performance (a better U-value of the wall and enhanced window and air tightness), the initial investment grew by 300,000 won/m2. The lighting facility cost, when LED lightings replaced the general ones, also rose by 300,000 won/m2. The initial investment was finally increased by 300,000 won/m2, and the energy consumption decreased by 300,000 won/m2 with the enhanced envelope performance and 300,000 won/m2 with the improved facility efficiency. This meant that with 40 years of the building’s operation there will be a drop in the total operating cost by 60%. For case B, the total LCC was reduced by 35% compared to case A (Table 11). A continuous comparison between case A (proposed model) and case B (optimized model) enabled the decision makers to find an alternative which is closer to a nZEB. A comparison with the basic model is not described here.
Table 11.
Evaluation Results.
6. Conclusions
Multi-objective optimizations have been applied with multiple optimization targets (objective functions) through use of genetic algorithm [6], the Taguchi-ANOVA method [9] and the PSO algorithm [10] as optimization techniques. Recent studies on life cycle cost buildings published in other countries [11,12,13,14,15] estimated all costs, including building energy and reduction of CO2 emissions. However, Korean studies have limited their scope to optimal construction cost, focusing on structure, materials and construction. Especially, past studies rarely used a generic algorithm (GA), which is known to be a good method to identify the cost optimization model based on discrete variable data.
Therefore, LCC evaluation and a cost optimal design tool with GA algorithm (MATLAB program) were developed in the present study which can be used in the early design phase. With this tool the client can recognize the financial benefit of nZEB rapidly in the early design phase and invest to achieve a higher performance from the nZEB. This newly developed tool can also provide the necessary guidance to a designer on the building design of a nZEB and enable them to take effective decisions without spending much time and effort.
It has been found in this study that more design variables must be considered in the early design phase to be able to easily use the tool for fast decision making. Programs for different purposes should be developed that can facilitate decision making for the design of other buildings as well as office buildings. It is essential to continue to update the database of the developed programs as the performance of the technology applied to the baseline building will improve and the unit cost will fall as the technology develops. Therefore, an up-to-date database needs to be maintained through market surveys.
Acknowledgments
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2013R1A1A3013119).
Conflicts of Interest
The author declares no conflict of interest.
References
- International Energy Agency (IEA). TASK 40/Annex 52 (2008). Towards Net Zero Energy Solar Buildings, IEASHC Task 40 and ECBCS Annex 52. 2011. Available online: http://www.iea-ebc.org/fileadmin/user_upload/docs/Facts/EBC_Annex_52_Factsheet.pdf (accessed on 16 December 2016).
- Tatari, O. Cost premium prediction of certified green buildings: A neural network approach. Build. Environ. 2011, 46, 1081–1086. [Google Scholar] [CrossRef]
- Black, R.; Brown, G.; Diaz, K.; Gibler, K.; Grissom, T. Behavioral Research in Real Estate: A Search for the Boundaries. J. Real Estate Pract. Educ. 2003, 6, 85–112. [Google Scholar]
- Kang, S. A Study on the Design Process of Zero Emission Buildings. Master’s Thesis, Chung-Ang University, Seoul, Korea, 10 December 2007. [Google Scholar]
- King, P.; Clare, M. Zero Carbon Task Group Report—The Definition of Zero Carbon; UK Green Building Council Press: London, UK, 2008. [Google Scholar]
- MOTL. Zero Energy Building Certification Launching (2016), Korean Government Announcement. Available online: https://www.molit.go.kr/USR/NEWS/m_71/dtl.jsp?lcmspage=1&id=95078225 (accessed on 16 December 2016).
- Torcellini, P.; Pless, S.; Deru, M. Zero energy buildings: A critical look at the definition. In Proceedings of the ACEEE Summer Study, Pacific Grove, CA, USA, 14–18 August 2006.
- Grözinger, J.; Boermans, T.; Ashok John, J.; Wehringer, F.; Seehusen, J. Towards Nearly Zero-Energy Buildings Definition of Common Principles Under the EPBD—Final Report; ECOFYS Germany GmbH: Köln, Germany, 2013. [Google Scholar]
- KICT. What Is Zero Energy Buildings? Korea Energy Agncy: Kyungki, Korea, 2016. Available online: http://nzeb.kict.re.kr/zeroEnergy/about.php (accessed on 16 December 2016).
- Song, D. High-Performance HVAC system for Net Zero Energy Building (NZEB). Architecture 2014, 58, 26–31. [Google Scholar]
- Kang, J.; Kim, D.; Thorsten, S.; Park, C. Assessment of energy saving potentials in terms of room’s thermal characteristics using building simulation. In Proceedings of the AIK Conference, Seoul, Korea, 23–26 October 2013; pp. 371–373.
- Shin, H.; Chang, K. Analysis of Energy Consumption and Cost based on Combination of Element Technologies for Implementing Zero-Energy House. J. KIAEBS 2014, 9, 163–170. [Google Scholar]
- Cho, D. Policies at Home and Abroad for Dissemination of nearly Zero Energy Buildings. Architecture 2014, 58, 10–15. [Google Scholar]
- Elkington, J. Towards the sustainable corporation: Win-win-win business strategies for sustainable development. Calif. Manag. Rev. 1994, 36, 90–100. [Google Scholar] [CrossRef]
- Cam, C.N. From global climate change to low carbon cities the triple bottom line revisit. In Proceedings of the First International Conference on Sustainable Urbanization, Hongkong, China, 15–17 December 2010.
- Dean, J.; VanGeet, O.; Simkus, S.; Eastment, M. Design and Evaluation of a Net Zero Energy Low-Income Residential Housing Development in Lafayette, Colorado; Technical Report NREL/TP-7A40-51450; National Renewable Energy Laboratory: Golden, CO, USA, 2012.
- Reijinders, L.; van Roekel, A. Comprehensiveness and adequacy of tools for the environmental improvement of buildings. J. Clean. Prod. 1999, 7, 221–225. [Google Scholar] [CrossRef]
- Deru, M.; Torcellini, P. Improving Sustainability of Buildings through a Performance-Based Design Approach. In Proceedings of the World Renewable Energy Congress VIII (WREC 2004), Denver, CO, USA, 29 August–3 September 2004.
- Chidiac, S.E.; Catania, E.J.C.; Morofsky, E.; Foo, S. A screening methodology for implementing cost effective energy retrofit measurein Canadian office buildings. Energy Build. 2011, 43, 614–620. [Google Scholar] [CrossRef]
- Wang, J.J.; Jing, Y.Y. Review on multi-criteria decision analysis aid in sustainable energy decision-making. Renew. Sustain. Energy Rev. 2009, 13, 2263–2278. [Google Scholar] [CrossRef]
- Znouda, E.; Ghrab-Morcos, N.; Hadj-Alouane, A. Optimization of Mediterranean building design using genetic algorithms. Energy Build. 2007, 39, 148–153. [Google Scholar] [CrossRef]
- Hamdy, M.; Hasan, A.; Siren, K. Applying a multi-objective optimization approach for Design of low-emission cost-effective dwellings. Build. Environ. 2011, 46, 109–123. [Google Scholar] [CrossRef]
- Shi, X. Design optimization of insulation usage and space conditioning load using energy simulation and genetic algorithm. Energy 2011, 36, 1659–1667. [Google Scholar] [CrossRef]
- European Committee for Standardization. Sustainability of Construction Works—Sustainability Assessment of Buildings—Part 4: Framework for the Assessment of Economic Performance; British Standards Institute: London, UK, 2012. [Google Scholar]
- IMMOVALUE: Improving the Market Impact of Energy Certification by Introducing Energy Efficiency and Life-Cycle Costs into Property Valuation Practice, IEE/07/553. 2010. Available online: http://www.immovalue.org (accessed on 16 December 2016).
- Hofer, G.; Herzog, B.; Grim, M. Calculating life cycle cost in the early design phase to encourage energy efficient and sustainable buildings. In Proceedings of the IEECB Focus 2010, Frankfurt, Germany, 13–14 April 2010.
- Kang, H.; Rhee, E. A Comparative Analysis of Performance Assessment Tools for Establishing Evaluation Framework for Sustainable Buildings. Archit. Res. 2014, 14, 131–136. [Google Scholar] [CrossRef]
- Forsberg, A.; Malmborg, V.F. Tools for environmental assessment of the built environment. Build. Environ. 2004, 39, 223–228. [Google Scholar] [CrossRef]
- Kun, C.J. Development of an energy evaluation methodology to make multiple predictions of the HVAC&R system energy demand for office buildings. Energy Build. 2014, 80, 169–183. [Google Scholar]
- Jung, J.R. Evaluation of Alternative for Building Service System in High-Rise Building Based on Life Cycle Cost Analysis. Ph.D. Thesis, Yonsei University, Seoul, Korea, 5 August 2002. [Google Scholar]
- Yum, Y.S. Evaluation of Sustainable Building Technologies for Multi-Family Residential Building Based on Life Cycle Cost Analysis. Master’s Thesis, Chung-Ang University, Seoul, Korea, 4 August 2008. [Google Scholar]
- Sesana, M.M.; Salvalai, G. Overview on life cycle methodologies and economic feasibility for nZEBs. Build. Environ. 2013, 67, 211–216. [Google Scholar] [CrossRef]
- Lho, S.W. A Study of CO2 Emission Trading Market Focusing on EU ETS. J. Environ. Pol. Admin 2009, 17, 25–44. [Google Scholar]
- European Emission Allowance. Available online: https://www.eex.com (accessed on 16 December 2016).
- Lee, G.H.; Kwon, Y.C.; Rhee, E.G. A Study on the Economic Analysis Method of Building Energy Saving Measures Considering Environmental Costs. J. Archit. Inst. Korea 2001, 17, 217–224. [Google Scholar]
- The Bank of Korea Economic Statistics System. Available online: http://ecos.bok.or.kr/jsp/use/qnamgt/QNACtl.jsp (accessed on 15 April 2016).
- Korea Power Exchange. Survey on Electricity Consumption Characters of Home Appliances; Korea Power Exchange: Seoul, Korea, 2006; pp. 50–75. [Google Scholar]
- Lam, J.C.; Hui, S.C.; Chan, A.L. Regression Analysis of High-rise Fully Air-conditioned Office Buildings. Energy Build. 1997, 26, 189–197. [Google Scholar] [CrossRef]
- Min-Chul, K. A Study on the Character of Changes in Unit Plans Prepared Apartment Balcony Extension. Master’s Thesis, Kyungpook University, Daegu, Korea, 10 December 2007. [Google Scholar]
- Ho-Tae, S. A Study on the Development of Load Prediction Equation and Design Guidelines for the Energy Conservation of Office Buildings. Ph.D. Thesis, Department of Architecture Graduate School, Seoul National University, Seoul, Korea, 12 December 1994. [Google Scholar]
- Hofer, G. Integrated Planning for Building Refurbishment—Taking Life-Cycle Costs into Account (LCC-Reburb); European Commission: Brussels, Belgium, 2005. [Google Scholar]
- Hasan, A.; Vuolle, M.; Siren, K. Minimisation of life cycle cost of a detached house using combined simulation and optimization. Build. Environ. 2008, 43, 2022–2034. [Google Scholar] [CrossRef]
- Yoo, S.I. A Study on the Prediction of Required Total Cost and Avoidable Cost of a High-Rise Apartment Building. Master’s Thesis, Chung-Ang University, Seoul, Korea, 26 June 2000. [Google Scholar]
- Cho, J.K. Development of an HVAC & R Systems Energy Evaluation Methodology and Simulation Program for Office Buildings. Korean J. Air Cond. Refrig. Eng. 2013, 25, 363–370. [Google Scholar]
- Kim, Y.R. A Development of a Mobile Application for Building Energy Prediction Using Performance Prediction Model. Master’s Thesis, Chung Ang Univeristy, Seoul, Korea, 2014. [Google Scholar]
- Hwan, J.J. A Study on the Development of Expert System for Selecting and Modifying Orthogonal Array in Taguchi Method. Korea Intell. Inf. Syst. Soc. 1997, 1, 350–361. [Google Scholar]
- Hoseon, Y. Effects of Various Factors on the Energy Consumption of Korean-Style Apartment Houses. J. Air-Cond. Refrig. 2002, 40, 972–980. [Google Scholar]
- Leigh, S.B.; Won, J.S.; Bae, J.I. An Energy Management Process and Prediction of Energy Use in Office Building. J. Asian Archit. Build. Eng. 2005, 4, 501–508. [Google Scholar] [CrossRef]
- Wright, J.; Loosemore, H.; Farmani, R. Optimization of building thermal design and control by multi-criterion genetic algorithm. Energy Build. 2002, 34, 959–972. [Google Scholar] [CrossRef]
- Cai, Z.; Wang, Y. A multi objective optimization-based evolutionary algorithm for constrained optimization. IEEE Trans. Evolut. Comput. 2006, 10, 658–675. [Google Scholar] [CrossRef]
- Thomas, B.; Jan, G.; von Bernhard, M.; Nesen, S.; Ashok, J. Assessment of Cost Optimal Calculations in the Context of the EPBD (ENER/C3/2013-414); Final Report; ECOFYS Germany GmbH: Cologne, Germany, 2015. [Google Scholar]
© 2017 by the author; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).








