Elsevier

Energy

Volume 192, 1 February 2020, 116723
Energy

A three-stage optimization methodology for envelope design of passive house considering energy demand, thermal comfort and cost

https://doi.org/10.1016/j.energy.2019.116723Get rights and content

Highlights

  • A novel three-stage optimization method is proposed for passive house performance.

  • Energy demand, cost and comfort levels are considered.

  • GBDT based meta-models are used to improve accuracy and efficiency.

  • SA and RDA separately used for performance analysis and dimensionality reduction.

  • Can provide guidance for the ontology design of the passive house.

Abstract

Due to reducing the reliance of buildings on fossil fuels, Passive House (PH) is receiving more and more attention. It is important that integrated optimization of passive performance by considering energy demand, cost and thermal comfort. This paper proposed a set three-stage multi-objective optimization method that combines redundancy analysis (RDA), Gradient Boosted Decision Trees (GBDT) and Non-dominated sorting genetic algorithm (NSGA-II) for PH design. The method has strong engineering applicability, by reducing the model complexity and improving efficiency. Among then, the GBDT algorithm was first applied to the passive performance optimization of buildings, which is used to build meta-models of building performance. Compared with the commonly used meta-model, the proposed models demonstrate superior robustness with the standard deviation at 0.048. The optimization results show that the energy-saving rate is about 88.2% and the improvement of thermal comfort is about 37.8% as compared to the base-case building. The economic analysis, the payback period were used to integrate initial investment and operating costs, the minimum payback period and uncomfortable level of Pareto frontier solution are 0.48 years and 13.1%, respectively. This study provides the architects rich and valuable information about the effects of the parameters on the different building performance.

Introduction

Population growth, increased demand for indoor environment and global warming have led to a sharp increase in energy consumption for buildings heating and cooling, which accounts for 20% of global energy consumption [1]. Especially, the energy consumption by the residential building is increasing at approximately 30% annually worldwide [2]. Therefore, developing sustainable buildings has increasingly become a very important task, and Passive House (PH) has emerged as the preferred concept for architects and subject for researchers in most countries. PH are buildings that need 80%–90% less heating energy than conventional buildings to provide comfortable indoor conditions, while the incremental cost of their construction is only 5%–10% [3]. Many countries have introduced PH standards. For example, China has issued “Passive ultra-low energy green building technology guidelines” [4]. However, the standards merely set deterministic results to evaluate the design without pre-directing the design. Therefore, it is almost impossible to use standards to fully exploit large design spaces and to guide decision-makers. In addition, engineers and technicians put more focus on active equipment and facilities but ignore the idea that PH is based on improving passive performance [5]. The passive performance is the base of sustainable development, especially for residential buildings [6].

Current research on PH buildings mainly focuses on performance assessment such as energy consumption and thermal comfort. Many studies have shown the superior performance of PH buildings. Based on the performance simulation, the heating energy demand of the PH building is less than the requirement of the PH standard, and the annual cooling energy demand is also very low [7]. The measured data shows that the total energy consumption of PH buildings is reduced by about 50% compared with traditional buildings [8] and reduce heating energy consumption by about 65% and energy consumption by 35% compared with low-energy buildings. However, there are also studies that expose the negative side of PH buildings in the indoor thermal environment, especially during the transitional season. For example, based on in-use monitored data gathered in 21 months, a study evaluated the thermal comfort of a UK PH dwelling with vulnerable occupants [9]. This research indicated that indoor overheating risk (IOR) can occur if the dwelling is not managed correctly, energy and carbon savings should not be at the expense of thermal comfort. In Europe, IOR in PH buildings has been recorded as more widespread compared to traditional buildings [10,11]. Moreover, in Australia, the high-performance buildings do not directly encourage passive survivability and can even increase IOR compared to traditional buildings [12]. Even some studies have directly pointed out that improving building fabric (increased insulation and airtightness) increases IOR [13]. It should be mentioned that natural ventilation is very effective in reducing energy consumption and improving the indoor thermal environment [14]. Opening the window for ventilation during the night or uncomfortable daytime can maintain indoor thermal comfort [15]. In short, thermal comfort issues should be valued in high-performance buildings such as PH buildings.

Many recent scientific studies performed building performance optimization whether it is single or multi-objective. There are numerous metrics involved in assessing building performance: 1) energy metrics [[16], [17], [18], [19], [20]], including annual heating, cooling, lighting, and total building energy demand; 2) life cycle metrics [16], including life cycle costs, life cycle carbon emissions and life cycle energy consumption; 3) indoor thermal comfort metrics [17,18]. It should be mentioned that it is easy to transfer energy metrics into life cycle metrics with some simple additional information [21]. According to the research objects, these studies mainly focus on new residential buildings and renovated buildings. For example, a study developed a new methodology to optimize building life cycle cost, environmental impacts, and occupant satisfaction in the early design phase [17,22]. Another research proposed a set of optimization methods for energy-renovating buildings that focus on energy consumption, retrofit cost, and thermal discomfort hours [23]. However, the design priorities of PH and conventional buildings are different, in the related research of PH buildings, few studies involve integrated optimization of energy efficiency, thermal comfort, and economic benefit.

Computer-aided optimization is the earliest method used in building performance optimization (BPO). It is automatically optimized by coupling simulation software and optimization algorithms [24]. However, evolutionary algorithms usually still need a large number of cost function evaluations before a satisfying result can be obtained [25]. Moreover, higher time costs of dynamic building performance simulation reduce the effectiveness of BPO and especially its diffusion in professional practice [26]. A survey study based on architects, mechanical engineers, and green construction consultants demonstrated that slower optimization speed is the main reason for hindering the actual application of BPO [19]. More important, some optimization indicators cannot be directly obtained by running simulation software, generally, they need secondary processing. Therefore, traditional computer-aided algorithms are not suitable for multi-objective optimization problems. For this reason, an approximation of the optimization problem is required. Among existing approximation approaches, the functional approximation approach (a.k.a. meta-model or surrogate model approximation) is the most used in BPO [27]. The meta-model constructs a functional relationship between multiple inputs and multiple outputs, which improve efficiency by sacrificing precision [28]. It should be mentioned that the meta-model is not equal to the agent model. The former is generally used for building performance optimization problems, while the latter generally appears in research related to human behavior and perception [29]. Seeking appropriate algorithms to build a high-precision meta-model is getting more and more research attention [[30], [31], [32], [33]]. Several studies have employed a variety of algorithms to establish meta model, such as multiple linear regression [[34], [35], [36]], support vector machines (SVM) [33] and artificial neural networks (ANN) [17,23,37,38]. These meta-models are generally combined with optimization algorithms such as genetic algorithms (GA). For example, Asadi et al. present a multi-objective optimization model using GA and ANN to quantitatively assess technology choices in a building retrofit project [23]. Gou et al. established a model to optimize the thermal comfort and energy demand of new residential buildings, by using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) coupled with the ANN [17]. Although the meta-model has been generally accepted and applied, proposing new algorithms to build a more robust and highly accurate meta-model is still the research focus. Especially for multi-objective optimization problems, the robustness of the meta-model is more important [39].

Based on the above analysis, a summary of the existing problems and the corresponding innovations is as follows:

  • (1)

    From the optimization model:

    • The existing problems on PH buildings: Most research related to PH buildings focuses on performance assessments, some of which only raised the phenomenon of the indoor overheating risk in PH buildings, but did not propose solutions to solve the problem. In addition, PH standards have only constrained the range of energy consumption and certain design parameters, but it is difficult to guarantee an optimal solution. In summary, there is still a lack of systematic optimization methods to guide the passive design of PH buildings in engineering applications.

    • The corresponding innovation: This paper establishes an optimization model for PH buildings. Different from previous research, the optimization model considers the effect of windowing for natural ventilation on the indoor thermal environment. The relationship between 20 passive design parameters and two building properties including energy demand and thermal comfort was constructed and the optimization scheme was explored under the constraints of PH standards. Finally, an economic analysis of the Pareto frontier solution was carried out. The optimization framework produces more practical and detailed design guidance.

  • (2)

    From the optimization method:

    • The existing problems on passive optimization: Existing research generally uses meta-models coupled multi-objective optimization algorithm to improve optimization efficiency. Although the effectiveness of the meta-model has been proven in numerous architectural performance optimization studies, seek alternative algorithms suitable for all building performance to build robust and highly accurate meta-models is still the focus of current research.

    • The corresponding innovation: This paper proposes a three-stage optimization method (RDA-GBDT-NSGA) that simultaneously improves the optimization efficiency and accuracy. The GBDT machine learning algorithm (described in Section 2.3.2) is used to establish meta-models of building performance, which was first applied to the building optimization. From the two aspects of accuracy and robustness, its effectiveness is verified by comparison with the commonly used algorithms including ANN and SVM.

Section snippets

Method

The optimization framework for PH buildings mainly includes three parts: constructing optimization model, establishing optimization method and post-Pareto analysis (Fig. 1). This section details the involved methods.

Setting the building model

Since the optimization of passive parameters exists in the early stage of building design, as in most of the related existing studies, the simulation data is used to expand the research database [16]. The entire building performance is simulated using EnergyPlus (Ver.9.0.1), which is a highly validated simulation engine widely used in building energy analysis [70]. In order to represent the characteristics of most Chinese residential buildings, this paper takes the popular slab-type building as

Multivariate performance analysis

PRCC, the sensitivity index, derived from MLR is applied to analyze the main factors affecting building performance. The larger the sample size, the more stable the indicators will be. However, there is no a priori exact rule for determining the adequate sample size for LHS–PRCC index. A way to solve the problem is to systematically increase the sample size and check if the sensitivity index used can consistently capture and rank a similar set of most important effects [75]. Therefore, the

Conclusion

This paper proposes a multi-objective optimization method for PH design, and the cold climate zone is chosen as an example to construct the multi-objective optimization model. Annual energy demand and comfort level are used as optimization objectives, and the boundaries of design parameters, annual cooling, and annual heating energy demand are used as constraints. The proposed three-stage optimization design method can reduce the complexity of the model and improve optimization accuracy and

Declaration of competing interest

The authors declared that they have no conflicts of interest to this work.

We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

Acknowledgment

This research has been supported by the “National Key R&D Program of China” (Grant No. 2016YFC0700100).

References (77)

  • A. Kaklauskas et al.

    Passive House model for quantitative and qualitative analyses and its intelligent system

    Energy Build

    (2012)
  • F. Harkouss et al.

    Passive design optimization of low energy buildings in different climates

    Energy

    (2018)
  • S. Gou et al.

    Passive design optimization of newly-built residential buildings in Shanghai for improving indoor thermal comfort while reducing building energy demand

    Energy Build

    (2018)
  • F. Bre et al.

    A computational multi-objective optimization method to improve energy efficiency and thermal comfort in dwellings

    Energy Build

    (2017)
  • X. Chen et al.

    Integrated energy performance optimization of a passively designed high-rise residential building in different climatic zones of China

    Appl Energy

    (2018)
  • D.A. Chi et al.

    Correlating daylight availability metric with lighting, heating and cooling energy consumptions

    Build Environ

    (2018)
  • Y. Huang et al.

    Energy and carbon emission payback analysis for energy-efficient retrofitting in buildings—overhang shading option

    Energy Build

    (2012)
  • E. Mostavi et al.

    Development of a new methodology to optimize building life cycle cost, environmental impacts, and occupant satisfaction

    Energy

    (2017)
  • E. Asadi et al.

    Multi-objective optimization for building retrofit: a model using genetic algorithm and artificial neural network and an application

    Energy Build

    (2014)
  • A. Prada et al.

    On the performance of meta-models in building design optimization

    Appl Energy

    (2018)
  • S. Attia et al.

    Assessing gaps and needs for integrating building performance optimization tools in net zero energy buildings design

    Energy Build

    (2013)
  • X. Chen et al.

    A multi-stage optimization of passively designed high-rise residential buildings in multiple building operation scenarios

    Appl Energy

    (2017)
  • J. Hester et al.

    Sequential early-design guidance for residential single-family buildings using a probabilistic metamodel of energy consumption

    Energy Build

    (2017)
  • E. Azar et al.

    Integrating and optimizing metrics of sustainable building performance using human-focused agent-based modeling

    Appl Energy

    (2016)
  • C.B. Zhou et al.

    Design optimization and field demonstration of natural ventilation for high-rise residential buildings

    Energy Build

    (2014)
  • S. Papadopoulos et al.

    Integrating building performance simulation in agent-based modeling using regression surrogate models: a novel human-in-the-loop energy modeling approach

    Energy Build

    (2016)
  • B. Eisenhower et al.

    A methodology for meta-model based optimization in building energy models

    Energy Build

    (2012)
  • X. Chen et al.

    A multi-stage optimization of passively designed high-rise residential buildings in multiple building operation scenarios

    Appl Energy

    (2017)
  • J.S. Hygh et al.

    Multivariate regression as an energy assessment tool in early building design

    Build Environ

    (2012)
  • S. Asadi et al.

    On the development of multi-linear regression analysis to assess energy consumption in the early stages of building design

    Energy Build

    (2014)
  • J.-S. Chou et al.

    Modeling heating and cooling loads by artificial intelligence for energy-efficient building design

    (2014)
  • C. Turhan et al.

    Comparative study of a building energy performance software (KEP-IYTE-ESS) and ANN-based building heat load estimation

    Energy Build

    (2014)
  • F. Ascione et al.

    CASA, cost-optimal analysis by multi-objective optimisation and artificial neural networks: a new framework for the robust assessment of cost-optimal energy retrofit, feasible for any building

    Energy Build

    (2017)
  • F. Kheiri

    A review on optimization methods applied in energy-efficient building geometry and envelope design

    Renew Sustain Energy Rev

    (2018)
  • M. Ferrara et al.

    Automated optimization for the integrated design process: the energy, thermal and visual comfort nexus

    Energy Build

    (2018)
  • W. Tian

    A review of sensitivity analysis methods in building energy analysis

    Renew Sustain Energy Rev

    (2013)
  • X. Chen et al.

    Approaching low-energy high-rise building by integrating passive architectural design with photovoltaic application

    J Clean Prod

    (2019)
  • M.-L. Pannier et al.

    Comprehensive assessment of sensitivity analysis methods for the identification of influential factors in building life cycle assessment

    J Clean Prod

    (2018)
  • Cited by (70)

    View all citing articles on Scopus
    View full text