Construction Research Congress

Transcription

1 Construction Research Congress Sensitivity Analysis of Factors Affecting Decision-Making for a Housing Energy Retrofit: A Case Study Amirhosein Jafari, S.M.ASCE 1 ; Vanessa Valentin, Ph.D., M.ASCE 2 ; and Mark Russell, Ph.D., M.ASCE 3 1 Dept. of Civil Engineering, Univ. of New Mexico, MSC , Albuquerque, NM jafari@unm.edu 2 Dept. of Civil Engineering, Univ. of New Mexico, MSC , Albuquerque, NM vv@unm.edu 3 Dept. of Civil Engineering, Univ. of New Mexico, MSC , Albuquerque, NM russ1307@unm.edu Abstract In the United States, over 60% of the housing inventory is more than 30 years old. One way to improve energy efficiency of those aged buildings is through housing retrofits. One of the main challenges of housing retrofit projects is making the decision about the amount of investment that results in maximum long-time benefits. In terms of life-cycle cost for a housing retrofit, different factors may affect the type of retrofitting alternatives to be implemented in the project. This research first introduces an optimization model for decision-making in housing retrofit. The model incorporates the use of genetic algorithm for selecting the optimum retrofitting plan based on the minimum life-cycle cost of the building. Then using a case study through this research a sensitivity analysis is performed to evaluate the impact of different factors such as service life of the building, homeowner s available budget, and discount rate of the building location on the suggested optimum retrofitting alternatives. The initial results illustrate that the retrofitting efforts could be more feasible when the service life of the building is high and the amount of discount rate is low. The results can help homeowners to make a more accurate decision for a housing energy retrofit. INTRODUCTION Buildings, particularly residential buildings, are major consumers of energy (Syal et al. 2014) and therefore have a significant adverse impact on the environment. Based on the American Housing Survey (USCB 2013), over 60% of the housing inventory in the U.S. is more than 30 years old. Therefore a large number of these homes are energy inefficient (Syal et al. 2014). Since most of the energy consumption in a building occurs during the operation phase (Menassa 2011), there is significant potential for transforming the existing buildings to accomplish energy efficiency objectives (Bruce et al. 2015). It can result in decreasing building costs as well as reducing the adverse impact of existing buildings on the environment (Jafari and Valentin 2015).

2 Construction Research Congress Considering the built environment sustainability, energy retrofitting has received increasing attention as a method to reduce energy consumption and operation cost of a building. A retrofit is the physical and operational upgrade of a building s energy-consuming equipment to reduce the amount of energy needed to perform particular tasks (Syal et al. 2014). Any retrofitting project is always surrounded by uncertainties in many variables: the energy analysis requires an estimation of energy consumption through retrofitting efforts; the life-cycle cost of the project can change by designed service life of the building or the amount of discount rate; and the amount of retrofitting efforts can change by the amount of available homeowner budget. For these reasons it is essential to understand how the retrofitting strategy may change with certain parameters. While many studies have looked at decision-making in retrofitting (Abdallah et al. 2014; Asadi E 2012; Jafari and Valentin 2015; Juan et al. 2010; Kaklauskas et al. 2005; Kumbaroglu 2012; Syal et al. 2014), there is limited knowledge on evaluating the impact of different design factors (e.g. service life of the building, homeowner s available budget, and discount rate of the building location) on decision-making for retrofitting a building. The concept of understanding how retrofit decisions may be influenced by external factors comes before the selection of an optimum retrofitting strategy. In order to fill this knowledge gap, the research described in this paper aims to determine the relative impact of the design factors on the optimum retrofitting plan. There are a number of methods and models developed for evaluating existing building conditions and supporting decisions to building retrofit. Several studies have proposed methods to select the most suitable solution for a retrofitting project; including: multiple criteria analysis based on the assessment of economic, ecological, social, and political conditions (Kaklauskas et al. 2005); integrated decision support system based on the trade-offs between renovation cost, improved building quality, and environmental impacts (Juan et al. 2010); intelligent decision support system based on experts query (Syal et al. 2014); determining the value of the investment in retrofits based on the perceived benefits of the investment (Menassa 2011); technoeconomic evaluation method based on the economically optimal set of retrofit measures (Kumbaroglu 2012); multi-objective model using genetic algorithm and artificial-neural network based on the quantitative assessment of technology choices (Asadi E 2012). Recently, Abdallah et al. (Abdallah et al. 2014) presented an optimization model that identifies an optimal set of building sustainability measures that are capable of minimizing the annual operational costs of the building. Despite of the significant contributions of these studies, the available literature is limited in: (1) using total LCC for decision-making; and (2) presenting a quantitative model to select a retrofitting plan among available retrofitting options. The objective of this study is to (1) introduce an optimization model to select the optimum housing retrofit alternative to minimize total LCC of a building; and (2) perform a sensitivity analysis to evaluate the impact of different factors (e.g. service life of the building, homeowner s available budget, and discount rate of the building location) on the suggested optimum retrofitting alternatives, using a case study.

3 Construction Research Congress MODEL DEVELOPMENT This section introduces an optimization model to select the optimum housing retrofit alternative to minimize total LCC of a building based on the expected service life of the building and investor s available budget. The model is developed in three steps: (1) model indexes identification determining required metrics to optimize model objective; (2) formulation of the model including decision variables, objective function, and constraints; and (3) model implementation explaining how the formulated model can be implemented and optimized. Model indexes identification In this step, the required metrics for the optimization of model objective are determined. Five main cost elements are considered, including: initial investment cost (IC), energy consumption cost (EC), maintenance cost (MC), replacement cost (RC), and tax rebates (TR). Model formulation The model is developed to select the optimum combination of retrofitting activities. Therefore, a binary variable is used for indicating whether a retrofitting activity should be implemented. For instance, an activity with a value of 1 represents that this activity is part of the optimum retrofitting strategy, and a 0 value, represents that activity is not part of the optimum retrofitting strategy. The decision variable of i th retrofitting activity will be shown by x i. Therefore, the number of variables is equal to the number of available retrofitting activities for a building. The present model is developed to minimize the total LCC of a building based on the retrofitting efforts. Total LCC can be calculated by Eq. 1, as the objective function of the model: = (1) Where LCC is the net present value of the total life-cycle cost, IC is the initial investment cost, PV EC is the present value of energy consumption cost, PV MC is the present value of maintenance cost, PV RC is the present value of replacement cost, and PV TR is present value of the tax rebates of the building during its service life. The present model is developed to comply with two main constraints as below: 1. It is assumed that home owner has a limited budget (LB) to implement a retrofitting strategy, then initial investment cost of retrofitting strategy should be less than available budget (see Eq. 2): (2) 2. There may be activities grouped in a bundle. Under this condition, only one activity should be selected from a bundle. Eq. 3 shows the constrains that apply bundling to the method: 1 (3) Model implementation The optimization problem of the present model is a constrained nonlinear optimization with a non-differentiable objective function. The genetic algorithm (ga) is a method able to solve nonlinear constraint problems. Therefore, in order to optimize the objective function of the model and find the optimum combination of

4 Construction Research Congress retrofitting activities for a building, a genetic algorithm (Goldberg 1989) is used for the optimization algorithm. In this study, Matlab R2014a is used for implementing the genetic algorithm (GA) optimization for the model. For implementing the model, it is required to identify all possible retrofitting activities and for the building as well as all relevant costs. It is also required to estimate the impact of activities on energy consumption of the building. CASE STUDY In order to demonstrate the features of the developed method and perform a sensitivity analysis, an example case study is used in this study. The house being studied was originally constructed in 1964 as a ranch style home in Albuquerque, New Mexico. The home is 1600 square feet, has 3 bedrooms, 2 bathrooms, and is made of concrete blocks. The current heating is by gas furnace and cooling is provided by an evaporative cooling system (Russell et al. 2014). The annual electricity and gas consumption of the case home are 9,550 KWh and MBtu, respectively, which cost $1,900 per year in average (Jafari et al. 2014). In order to determine the potential sequence of activities to be implemented in the housing retrofit, this study starts by identifying the basic least expensive items from the house, works up through more complex items, and finishes with on-site renewable energy systems. The Build Green New Mexico criteria for a Green Building (BGNM 2012) document is used to evaluate the steps that could be taken to renovate the case home. Considering the home owner s preferences, 15 different major retrofitting activities - varying from low to high cost efforts - are selected as possible retrofitting activities for the case study. Figure 1 provides a summary of the planned activities for retrofitting the house Activity #03 includes different options to show how bundling will work with the model. Controllin g Lighting Appliance Insulation Windows & Doors Heating & Cooling Water Heating Renew. Options 02.Tune up HVAC Halogen 03.2 CFL LED 05.Replace clothes washer with an energy star one 06.Replace dishwasher Figure 1. Planned retrofitting activities SENSITIVITY ANALYSIS The purpose of this study is to understand the relative impact of the following contributing parameters on the optimum retrofitting strategy: the planned service life of the building; the economic situation, which translates into discount rate; and the amount of available budget for retrofitting. 12.Install ground source heat exchanger 13.Install evaporative cooler 14.Install solar thermal equipment 15.Install solar electricity equipmenth

5 Construction Research Congress After gathering data about possible retrofitting activities and related information, the model implementation process is used to select the optimum combination of retrofitting activities to minimize the LCC. In order to better understand the sensitivity of an optimum retrofitting strategy to changes in design parameters, parameters are changed one at a time, with all others remaining constant, and the changes in the optimal solution are noted. Impact of service life The service life of a building can vary according to client expectations and project characteristics. This time span typically vary from 25 to 50 years (Wang et al. 2012). In this study, in order to consider the impact of different service life periods on the optimum retrofitting strategy, five different scenarios are considered: The homeowner expects to operate the building for service life of (1) 5 years; (2) 10 years; (3) 15 years; (4) 20 years; and (5) 25 years. For the assumed discount rate of 2% and no limitation in the retrofitting budget for the case study, the results of the model are presented in Table 1. Table 1. Result of the optimum retrofitting strategy for different service lives Service (1) (2) (3) (4) (5) Life 5 Years 10 Years 15 Years 20 Years 25 Years Selected Activities 01.Install programmable 03.Replace lights with energy efficient ones (03.2. CFL) 01.Install programmable 03.Replace lights with energy efficient ones (03.2. CFL) 01.Install programmable 03.Replace lights with energy efficient ones (03.2. CFL) 04.Replace refrigerator with an energy star one 07.Insulate ceilings 10.Replace doors with 11.Replace windows with energy efficient 01.Install programmable 03.Replace lights with energy efficient ones 04.Replace refrigerator with an energy star one 07.Insulate ceilings 10.Replace doors with 11.Replace windows with energy efficient 14.Install solar thermal equipment 01.Install programmable 03.Replace lights with energy efficient ones (03.2. CFL) 04.Replace refrigerator with an energy star one 07.Insulate ceilings 10.Replace doors with 11.Replace windows with energy efficient 14.Install solar thermal equipment IC: $406 $3,642 $13,618 $18,565 $18,190 LCC: $9,153 $16,701 $20,271 $21,913 $22,732 LCC/Year $1,831 $1,670 $1,351 $1,096 $909 Energy Saving: 19% 43% 58% 70% 70% The optimum retrofitting plan for the scenario (1) with a service life of 5 years includes only 2 activities; therefore, the initial investment cost is small as well as the amount of energy saving. In the scenario (2), by increasing the expected service life of the building from 5 to 10 years, the optimum retrofitting plan includes two additional activities: insulation of walls and insulation of attic. As a result, the initial investment cost and energy saving increase. However, the amount of LCC per year decreases. As it is shown in Table 1, by increasing the expected service life of

6 Construction Research Congress building to 15 and 20 years (scenario (3) and (4)), the same results are observed. In the scenario (5) with an expected service life of 25 years, the suggested activities for the optimum retrofitting plan and the initial investment and energy savings are almost the same as in the scenario (4). Again, the amount of total LCC per year in decrease. Figure 2 shows the trend of optimum retrofitting investment cost and LCC per year in relation to expected service life of the building. By increasing the expected service life period (which means the homeowner will use the building for a longer time span) more retrofitting efforts will be needed to be implemented based on the developed model that can minimize total LCC per year of the building. Therefore by increasing expected service life of the building, the amount of initial investment cost will increase; however, the LCC per year of the building will decrease. In other words, if the homeowner wants to have a more beneficial housing retrofit, he has to plan to use the building for a longer time span. Otherwise, the investment cost will be less, but the long term benefits will be reduced. LCC per Year Service Life (Year) LCC/Year ($) Initial Investment Cost ($) $20,000 $15,000 $10,000 $5,000 $- Initial Investment Cost Figure 2. Initial Investment cost and LCC per year in relation to service life As Figure 2 shows, if the homeowner decides to invest less than $4,000 for retrofitting, the best expected service life of the building will be less than 10 years that would minimize the total LCC per year. If the homeowner decides to invest up to $14,000 for retrofitting, the best expected service life of the building will be between 10 to 15 years that minimize total LCC per year; and if the homeowner decides to invest around $18,000 for retrofitting, it would cost effective to plan to operate the building as much as 25 years. Impact of discount rate Analyzing long-term investments requires to compare costs and benefits that occur in different time periods. Therefore, an economic technique known as discounting is used to convert different costs and benefits occurred at different times (Ferreira and Santos 2013). The discount rate depends on the prevailing interest rate and the depreciation of the currency or inflation rate. This rate is not a constant term and may vary over the service life of the project. A discount rate of 2 or 3% above inflation is considered an appropriate value (Hojjat 2002). In this study, in order to consider the impact of different discount rates on the optimum retrofitting strategy, four different scenarios are considered: The economic situation implies a discount rate of (1) approximately zero; (2) 2 percent; (3) 4 percent; and (4) 6 percent. For the assumed service life of 20 years and no limitation in the retrofitting budget for the case study, the results of the model are presented in Table 2. The optimum retrofitting plan for the scenarios (1) and (2) (discount rate of approximately 0% and 2%) are same; however, the amount of calculated LCC has a

7 Construction Research Congress small difference. This small difference is because of differences in NPV calculation of future costs such as energy consumption costs. In the scenario (3), by increasing the discount rate from 2 to 4 percent, the optimum retrofitting plan has changed and the activity: Installation of solar thermal equipment is removed from the planned activities. As a result, the initial investment cost and energy saving decrease considerably. However, the amount of LCC is almost same (has only a very small decrease). In the scenario (4) with a discount rate of 6%, the suggested activities for the optimum retrofitting plan and the initial investment and energy savings are almost the same as in the 3rd scenario. Again, the amount of LCC has slightly decreased. Table 2. Result of the optimum retrofitting strategy for different discount rate Discount (1) (2) (3) (4) Rate 0% 2% 4% 6% Selected Activities 14.Install solar thermal equipment 14.Install solar thermal equipment (03.2. CFL) 14.Install solar thermal equipment IC: $18,565 $18,565 $13,618 $13,093 LCC: $23,402 $21,913 $20,707 $19,412 Energy Saving: 70% 70% 58% 54% Figure 3 shows the trend of optimum retrofitting investment cost and LCC in relation to different discount rate. By increasing the discount rate to 3%, no change in the optimum retrofitting plan is observed. However by increasing the discount rate to 4% and more, less retrofitting efforts will be needed to implement based on the developed model that can minimize total LCC per year of the building. In other words, in the locations with high amount of discount rate which means the future cash flow of savings from housing retrofit will be discounted higher (lower present value of the future savings from retrofitting), the homeowners may decrease the amount of retrofitting efforts to minimize LCC. LCC $24, $22, $20, $18, % 1.00% 2.00% 3.00% 4.00% 5.00% Discount Rate (%) LCC ($) Initial Investment Cost ($) $20,000 $15,000 $10,000 $5,000 $- Figure 3. Initial Investment cost and LCC in relation to discount rate Initial Investment Cost

8 Construction Research Congress Impact of available budget For a retrofitting project, the available budget plays an important role because the amount of retrofitting investment cost cannot exceed the available budget. The developed decision-making model is considering a constraint to ensure that the assigned optimum retrofitting plan is always under the available budget. In this study, in order to consider the impact of available budget on the optimum retrofitting strategy, four different scenarios are considered: The homeowner s available budget is limited to (1) $5,000; (2) $10,000; (3) $15,000; and (4) there is no limitation. For the assumed service life of 20 years and discount rate of 2% for the case study, the results of the model are presented in Table 3. Table 3. Result of the optimum retrofitting strategy for different budgets Limited (1) (2) (3) (4) Budget $5,000 $10,000 $15,000 No Limit Selected Activities 14.Install solar thermal equipment 14.Install solar thermal equipment IC: $4, $9, $13, $18, LCC: $29, $28, $22, $21, Energy Saving: 46% 58% 59% 70% The optimum retrofitting plan for the scenario (1) with an available budget of $5,000 includes only five inexpensive activities; therefore, the initial investment cost is small (definitely under the available budget) as well as the amount of energy saving. In the scenario (2), by increasing the available budget to $10,000, the optimum retrofitting plan includes an additional activity: Insulation of solar thermal equipment. As a result, the initial investment cost and energy saving increase. However, it does not have a significant impact on total LCC. As it is shown in Table 3, by increasing the available budget to $15,000 (scenario (3)), the optimum retrofitting plan suggests different additional activities, which increases the initial investment cost. However there is no significant change in energy saving, the LCC drops significantly. In the scenario (4) with an unlimited budget for retrofitting, again it is suggested to install solar thermal equipment, which increase initial investment cost and energy saving but again has no significant impact on the total LCC. As a result, it can be mentioned that there is an activity in this example case ( Installation of solar thermal equipment ) that its implementation may not have a significant effect on LCC of the building, but can reduce the amount of energy consumption of the

9 Construction Research Congress house and its related impacts. Figure 4 shows the trend of optimum retrofitting investment cost and LCC in relation to available budget for housing retrofit. When there is no budget, there will be no retrofitting efforts. By increasing the available budget for retrofitting, more retrofitting efforts will be needed to be implemented based on the developed model that can minimize total LCC of the building. LCC $40, $30, $20, $10, $- Available Budget ($) LCC ($) Initial Investment Cost ($) $20,000 $15,000 $10,000 $5,000 $- Figure 4. Initial Investment cost and LCC in relation to available budget In addition, the results of Figure 4 can help homeowners to select the best retrofitting budget in order to minimize the LCC of the building. As it is shown, between budgets of $4,000 and $12,000, there is no significant change in LCC of the building. By assigning a budget more than $13,000 the LCC will drop significantly. Therefore, based on previous results for this case study, it can be said that the best decision for homeowner will be selecting one the following alternatives: Invest up to $4,000 to retrofit the building and use the building for up to 10 years; Invest around $14,000 to retrofit the building and use the building for up to 15 years; or Invest around $18,000 to retrofit the building and use the building for up to 25 years. CONCLUSION Since any retrofitting project is always surrounded by several uncertainties, this study focused on the importance of how retrofitting strategies may change with certain design parameters, including: the planned service life of the building; the economic situation, which translates into discount rate; and the amount of available budget for retrofitting. Therefore, an optimization model was developed to select the optimum housing retrofit measures based on the minimum LCC, a case study was considered, and then a sensitivity analysis was performed to evaluate the impact of these different factors on the suggested optimum retrofitting alternatives. The results showed that: (1) For a retrofitting project, the amount of available budget plays an important role in decision-making; (2) If the homeowners want to have a better payoff with a housing retrofit, they would have to plan to use the building for a longer time span. Therefore they would need to invest more but more long-term benefits would be achieved; and (3) In the locations with high amount of discount rate the homeowners may decrease the amount of retrofitting efforts to minimize LCC. As limitations of this research, the authors believe that the developed model needs to consider a database of retrofitting activities instead of just limited activities for a specific project. Initial Investment Cost

10 Construction Research Congress REFERENCES Abdallah, M., El-Rayes, K., and Liu, L. (2014). "Optimal Selection of Sustainability Measures to Minimize Building Operational Costs." Construction Research Congress 2014, Asadi E, D. S. M. G. A. C. H. D. L. (2012). "Multi-objective optimization for building retrofit strategies: A model and an application." Energy Build. Energy and Buildings, 44(1), BGNM, B. G. N. M. (2012). < Bruce, T., Zuo, J., Rameezdeen, R., and Pullen, S. (2015). "Factors influencing the retrofitting of existing office buildings using Adelaide, South Australia as a case study." Structural Survey, 33(2), Ferreira, A., and Santos, J. (2013). "Life-cycle cost analysis system for pavement management at project level: sensitivity analysis to the discount rate." INTERNATIONAL JOURNAL OF PAVEMENT ENGINEERING, 14(7), Goldberg, D. E. (1989). Genetic algorithms in search, optimization, and machine learning. Hojjat, A. (2002). "Life-cycle cost optimization of steel structures." International Journal for Numerical Methods in Engineering, 55(12), Jafari, A., and Valentin, V. (2015). "Decision-making life-cycle cost analysis model for energy-efficient housing retrofits." International Journal of Sustainable Building Technology and Urban Development. Jafari, A., Valentin, V., and Russell, M. (2014). "Probabilistic Life cycle Cost Model for Sustainable Housing Retrofit Decision-Making." International Conference on Computing in Civil and Building Engineering;Orlando, Florida. Juan, Y. K., Gao, P., and Wang, J. (2010). "A hybrid decision support system for sustainable office building renovation and energy performance improvement." ENERGY AND BUILDINGS, 42(3), Kaklauskas, A., Zavadskas, E. K., and Raslanas, S. (2005). "Multivariant design and multiple criteria analysis of building refurbishments." ENERGY AND BUILDINGS, 37(4), Kumbaroglu, G. M. R. (2012). "Evaluation of economically optimal retrofit investment options for energy savings in buildings." Energy and Buildings Energy and Buildings, 49, Menassa, C. C. (2011). "Evaluating sustainable retrofits in existing buildings under uncertainty." ENERGY AND BUILDINGS, 43(12), Russell, M., Valentin, V., Yu, K., Jafari, A., Folkman, J., and Maher, M. "Housing Refurbishment to Net Zero Energy Case Study." Proc., Proceeding of iisbe Net Zero Built Environment - 17th Rinker International Conference, March 6-7, Syal, M., Duah, D., Samuel, S., Mazor, M., Mo, Y., and Cyr, T. (2014). "Information Framework for Intelligent Decision Support System for Home Energy Retrofits." Journal of Construction Engineering and Management, 140(1), USCB (2013). "American Housing Survey for United States: 2011." < (January, 2015). Wang, N., Chang, Y.-C., and El-Sheikh, A. A. (2012). "Monte Carlo simulation approach to life cycle cost management." Structure & Infrastructure Engineering: Maintenance, Management, Life-Cycle Design & Performance, 8(8),