Fire Safety Resilience Assessment of Residential Self-Built Houses according to the TOPSIS Method

Abstract

Fire extinguishing in residential self-built houses is difficult, and casualties are likely to occur. The aim of this study was to improve the fire safety resilience of residential self-built houses and decrease fire safety hazards. Firstly, the connotation of fire resilience of residential self-built houses was scientifically defined according to resilience theory. Then, a fire safety resilience evaluation index system with resistance, response and learning as the core was developed based on the properties of residential self-built houses, legal norms and relevant literature. At the same time, to weaken the effects of subjective factors, a fire safety resilience evaluation model of residential self-built houses was established based on a combination weighting–similarity to ideal solution (TOPSIS) method to quantify the influence of each index and divide the resilience level standards. Finally, a practical case of a residential self-built house in Jiaozuo City was adopted for resilience assessment for the verification of the scientificity and rationality of the model. The results showed the following. (1) In the evaluation system, the proportion of the coping and learning stages was 75%, and improving this factor significantly promoted the fire safety resilience of residential self-built houses. (2) Through TOPSIS, the fire safety resilience of residential self-built houses was classified into four groups of poor, medium, good and excellent. (3) By performing an example test, the fire safety resilience level of this residential self-built house was found to be medium, and the evaluation results were basically consistent with the actual situation of the building, which verified the scientificity and rationality of the proposed model.

1. Introduction

With the rapid advancement of the urbanization process in China, the original urban–rural division pattern is being gradually disturbed and the urban–rural relationship is integrating. With the promotion of the balanced development of the economy and society in China, the problem of housing difficulty has emerged. Driven by the profit of a large population movement, several self-built houses around cities are rising every day. These places integrate accommodation, production, storage, operation and other functions, but they suffer from inherent deficiencies in fire safety conditions such as a lack of fire protection and evacuation facilities as well as other fire safety hazards. Based on Fire Rescue Bureau statistics, in 2021, 156,000 self-built house fires occurred in China, accounting for 60.5% of the total number of fires in residential places throughout the year. Moreover, 848 people were killed, which accounted for 58.1% of the total number of deaths in residential fires [1]. Therefore, residential self-built houses have become fire-prone areas, and their fire safety ability should be urgently improved.

Today, academic research on fire safety in residential self-built houses has primarily focused on microperspectives such as fire case analysis, scene simulation, safety hazards, prevention, and control measures. Dongsheng [2] studied the common fire points and causes of disasters by reviewing specific case scenarios. His aim was to provide the optimal emergency decision for subsequent disposal of similar cases. In addition, some researchers have used fire dynamics simulation software (FDS 6.7.7) to simulate the whole process of the fire scene. According to the key numerical changed, fire location [3], direction of smoke flow [4] and fire spreading law [5] were obtained, and continuous optimization measures were proposed based on simulation results. Other researchers have found that residential self-built houses often have inherent safety risks, such as complex personnel [6], mixed functions [7], inadequate supervision, and weak fire awareness [8]. To address these risks, fire control efforts should take into account both system implementation and daily management [9]. Research has revealed a tendency for fragmentation, which was significant for improving the fire safety of specific cases.

However, little research has been conducted on the macrolevel safety of residential self-built houses with fire incidents. Zhenzong [10] performed fire safety evaluations on 20 selected residential self-built houses using the Delphi method and provided targeted optimization suggestions. Huaqing [10] used the analytic hierarchy process (AHP) to screen and construct a comprehensive index system with 26 indicators across six aspects including the fire extinguishing system, fire prevention system, and safety evacuation system, among others, to evaluate the fire safety of residential self-built houses. Although the aforementioned research provided a basis for addressing fire hazards during the construction of self-built houses, it did not consider the entire process of fire prevention, response, and restoration to a safe state and neglected the structural features of the buildings themselves. In addition, since fire accidents are prevalent in residential self-built houses, a unified fire evaluation standard was urgent, and there was ample room for further development in research on an overall fire evaluation system for residential self-built houses.

Regarding global resilient city construction [11], the fire safety resilience of residential self-built houses should not be limited to the response levels of fire safety hazards and fire prevention measures, but should also be reflected in predisaster risk investigation, postdisaster recovery, resistance reconstruction and learning level. Accordingly, the resilience theory was adopted to explore fire safety of residential self-built houses, the connotation of fire safety resilience was defined, and a fire safety resilience evaluation index system was constructed with resistance, response and learning as the framework [12]. Using the combination weighting–TOPSIS method, the fire safety level of residential self-built houses was investigated. From the perspective of resilience, a decision-making basis was provided to optimize the fire safety resilience of residential self-built houses and proposed new ideas for fire protection (Figure 1).

2. Analysis of Fire Safety Resilience of Residential Self-Built Houses

2.1. Concept and Classification

Generally, residential self-built houses refer to the buildings constructed by units or individuals who own the land and organize and employ others to carry out the construction. Self-built houses comprise the main architectural form in China, especially in rural areas where residences are almost self-built to meet their living needs [13]. Based on their application, they can be roughly classified into the following three categories.

(1)

Self-housing. An independent house for owner occupation only. It mainly refers to houses with a maximum height of 24 m, 7 layers and a single-story building area not exceeding 500 square meters.

(2)

Group rental housing. A type of residence in which rooms are divided into several small rooms by changing the structure and layout of the house and are rented separately in the form of a room or bed.

(3)

Three-in-one places. One or more functions in accommodation, production, warehousing and operation, illegally mixed in the same space. This space could be either a separate building or part of a building with no effective fire isolation between accommodation and other functions.

2.2. Analysis of Fire Safety Problems of Residential Self-Built Houses

(1) Fire safety conditions of residential self-built houses are congenitally deficient. The majority of residential self-built houses are designed and constructed according to their own homesteads and lack scientific and unified planning. To cut costs, most design and construction personnel are nonprofessionals. Building fire protection design is defective and has great fire hazards. To compensate for demolition and rental profits, residents illegally build and raise floors, and the phenomena of veneer and handshake floors are common. There are a wide variety of problems such as insufficient fire spacing of buildings, fire lanes, fire compartments and safe evacuation which do not meet technical standards for fire protection. Additionally, early residential self-built houses are not included in government planning, and a lack of approval conditions and procedures and the absence of a standard acceptance process of professional departments result in fire safety hazards which cannot be discovered and rectified in time.

(2) Three-in-one places have different safety hazards and high fire risks. The main fire safety hazards are as follows: first, low fire resistance grade of buildings can easily cause large-area combustions. Usually, such places have fire safety hazards such as imperfect fire separation, private interlayers, and extensive application of flammable materials, which can easily result in large-scale combustion. Second, the fire load inside the place is large, making personnel escape and rescue difficult. Such places are crowded and have piles of goods and often only one safe exit. Once goods are on fire, the burning speed is extremely accelerated, making escape and rescue even more difficult. Third, another hazard lies in the failure and aging of electrical wiring. In the statistics of self-built house fires, electrical fires caused by different triggering factors rank first among all fire causes. Especially in profitable three-in-one places, there are many pieces of electrical equipment in life and production, and the laying of electrical lines is not standardized. Leakage and ignition can easily take place due to line overload, poor contact, aging and damage.

(3) Firefighting and evacuation facilities of residential self-built houses are insufficient. Firefighting equipment is a front-line guard, which is responsible for the initial fire extinguishing and controlling. Whether fire extinguishing equipment configurations meet the requirements and are in good condition will determine potential fire development. Safe evacuation facilities can decrease casualties and property losses due to fire and provide favorable firefighting and rescue conditions for firefighters, both of which are indispensable. To cut costs, most self-built houses do not configure fire and evacuation facilities in accordance with relevant standard specifications and do not carry out daily maintenance. Therefore, when a fire occurs, rescue efficiency is greatly reduced.

(4) Personnel composition is complex, and fire safety awareness is weak. Human security factors mainly include three aspects: first, a high proportion of floating population. Floating population is a major feature of residential self-built houses. Familiarity of personnel with the surrounding environment, evacuation channels and fire facilities has a direct influence on the efficiency of emergency evacuation. The second aspect is a high proportion of vulnerable groups. The proportion of elderly, children, disabled and other vulnerable groups also affects emergency evacuation efficiency and their ability to save themselves. The third main aspect is weak fire safety awareness among residents. In the analysis of typical fire cases in residential self-built houses (

Table 1. Typical residential self-built house fire statistics.

Date	Location	Reason	Loss
11 January 2022	Biyong Town, Huaihua City	Careless with fire	Five dead and one wounded
18 March 2022	Manjiang Street Office of Dali City	Wiring troubles	Two dead
11 February 2022	A self-built apartment in Changping, Beijing	Electric vehicle fire	Five dead
22 February 2022	A group rental house in Mengcheng, Anhui	Aging circuits	Five dead and four wounded
25 June 2021	Zhenxing hall in Shangqiu City	Careless with fire	Eighteen dead and sixteen wounded
8 March 2020	A self-built store in Qiandongnan Prefecture	Careless with fire	Nine dead
5 May 2019	A group of rental houses in Guilin	Electric vehicle fire	Five dead and thirty-eight wounded

), careless use of fire is among the most important ones. Some residents have poor self-protection awareness, treat fire safety as a fluke and lack basic knowledge on fire prevention, firefighting and escape. Illegal parking and charging of electric vehicles and careless use of fire have occurred frequently.

2.3. Definition of Fire Safety Resilience of Residential Self-Built Houses

The word resilience originates from its Latin root resilire, which is a term originally applied in physics to describe the properties of springs, stability of their materials and ability to withstand external shocks [14]. Therefore, the resilience theory was continuously introduced into the fields of ecology, city, disaster prevention and mitigation, forming a resilience science.

According to resilience theory and the objective situation of residential self-built houses in China, fire safety resilience of residential self-built houses is defined as follows: when exposed to fire hazard factors, the building itself can absorb and resist some external forces and take timely measures to maintain and restore the fire safety status of residential self-built houses. Specifically, it includes checking and monitoring disaster-causing factors before the fire occurs and providing timely warnings; quickly enabling firefighting facilities to self-adjust and repair in case of fire to minimize casualties; rapidly restoring infrastructure functions of residential self-built houses after the fire is over; and at the same time, making up for the shortcomings of the construction of the fire protection system through experiential learning.

The mechanism of resilient city is resilience, coping and learning, as explained below. First, the resistance stage is when the disturbance of fire hazard factors is within a controllable range and contains a large number of unpredictable crises which need to be monitored and warned in time. Secondly, the response phase follows; that is, the resilience construction phase. When a fire occurs, the firefighting and disaster relief system is quickly activated for self-adjustment and repair to avoid irreversible effects. Thirdly, there is the learning stage; that is, the resilience cultivation stage. This stage restores the infrastructure function of residential self-built houses through postdisaster experience learning, making up for deficiencies and improving their fire safety resilience.

3. Construction of Evaluation Index System

3.1. Index Selection

Based on the characteristics and current situation of residential self-built houses, combined with four regulations and standards including the “Fire Safety Law of the People’s Republic of China” (revised in 2021), “Technical points of structural safety inspection of self-built houses (provisional)”, etc., relevant literature were reviewed [15,16,17,18,19], and field research and in-depth interviews were conducted. The fire safety resilience indicators of residential self-built houses were identified according to the resilience mechanism of resistance, response and learning, and they were constantly revised based on expert opinions. Finally, 27 indicators were adopted to construct a fire safety resilience system for residential self-built houses (Figure 2).

3.2. Establish Weight

3.2.1. Calculation Method

To prevent the interference of subjective factors when evaluating the results, each index was weighted by the combination weighting method. The analytic hierarchy process (AHP) was applied in the subjective part. AHP is a combination of quantitative and qualitative evaluation methods, the core of which was to gradually decompose complex and difficult-to-quantify problems into target, criterion and index levels. System indexes were compared and analyzed in pairs to provide reference for adopting the optimal scheme [20]. The entropy weight method (EWM) was adopted in the objective part. The basic principle of EWM is to determine the weight of each index based on the size of target information contained in each index [21]. Finally, subjective and objective preference coefficients were introduced to obtain the comprehensive index weight of fire safety resilience of residential self-built houses [22].

The calculation equation of comprehensive weight was:

$$W=\alpha W*+(1-\alpha )W^{\prime }$$

(1)

where $W*$ is the subjective weight calculated by AHP, $W^{\prime }$ is the objective weight calculated by the entropy weight method, and the value of preference coefficient $\alpha$ is 0.5.

3.2.2. Weight Calculation

Through a questionnaire survey of 20 experts including university researchers, community staff, technical personnel and government-related staff in the field of fire safety of self-built houses, the weight values of each index of the influencing factors of fire safety resilience of residential self-built houses were calculated by the combination weighting method formula (

Table 2. Comprehensive weight of influencing factors of fire resilience.

Criterion Layer	Index Layer		$\mathit{W}*$	${\mathit{W}}^{\prime }$	$\mathit{W}$
resist 0.2527	hazard factors 0.1065	electrical installation, use and maintenance, C1	0.01095	0.0288	0.0199
		combustible and explosive hazard, C2	0.01655	0.0152	0.0159
		fire load density, C3	0.01655	0.0163	0.0164
		examination and approval of buildings, C4	0.01655	0.04	0.0283
		use of buildings, C5	0.0033	0.0487	0.0260
	personnel condition 0.1463	personnel quality, C6	0.035	0.0353	0.0352
		careless with fire, C7	0.035	0.0315	0.0333
		familiarity of personnel, C8	0.035	0.0397	0.0374
		personnel load, C9	0.0231	0.058	0.0406
response 0.3983	fire prevention for building 0.1429	fire resistance level of structures, C10	0.02265	0.0435	0.0331
		fire compartment, C11	0.0191	0.0897	0.0544
		fire distance, C12	0.02265	0.0393	0.0310
		firefighting conditions, C13	0.04155	0.0074	0.0245
	fire protection facility 0.2555	fire water system, C14	0.0162	0.0063	0.0113
		smoke control and extraction system, C15	0.0162	0.0516	0.0339
		sprinkler system, C16	0.04505	0.0404	0.0427
		automatic fire alarm system, C17	0.04505	0.0457	0.0454
		fire separation settings, C18	0.0094	0.0691	0.0393
		firefighting equipment configuration, C19	0.0162	0.0222	0.0192
		exit passageway, C20	0.0162	0.0487	0.0325
		fire brigade arrival time, C21	0.04945	0.0132	0.0313
Learning 0.3490	system implementation 0.1853	the implementation of fire safety responsibility system, C22	0.0971	0.0271	0.0621
		approval of residential self-built houses, C23	0.2114	0.035	0.1232
	daily management 0.1637	fire hazard investigation and rectification, C24	0.0477	0.0404	0.0441
		daily maintenance of firefighting facilities, C25	0.0477	0.0211	0.0344
		fire emergency plan optimization, C26	0.037	0.0424	0.0397
		fire training and drills, C27	0.0477	0.0435	0.0456

).

From the criterion level, the proportions of response and learning stages were 0.3983 and 0.3490, respectively, while the corresponding value for the resisting stage was only 0.2527. The results showed that the disturbance of fire factors due to disaster-causing factors and personnel conditions was inevitable, but for residential self-built houses, decreasing the sensitivity of the system itself and improving its adaptability were key to enhance fire resilience.

Combined with the proportion of index layer, the analysis procedure was as follows: (1) The overall proportion in the resistance stage was balanced, indicating that equal attention needed to be paid to the unsafe behaviors of people and unsafe state of things in case of residential self-built house fires. In this stage, weight distributions of individual indicators showed that the approval status of residential self-built houses and application of buildings were the main factors contributing to disasters. The safety of a building is determined by whether a building has been approved and the building application determined the complexity of its functions. The confusion of functions is the fire safety hidden danger of residential self-built houses, which is consistent with the actual situation. Among personnel conditions, indicators such as personnel quality, careless use of fire, familiarity with fire safety measures and personnel load had significant impacts on fire risk. In the assessment of fire resilience and safety in self-built houses, close attention should be paid to various indicators of personnel conditions. This included strengthening fire safety training for residents, improving personnel quality and enhancing safe fire usage awareness. (2) In the response stage, the proportion of fire protection facilities and equipment was remarkably higher than that of the building’s inherent fire performance and had great effect on the overall situation. In practice, it is difficult to change inadequate inherent fire conditions of buildings; therefore, it is necessary to make up for shortcomings through the provision of fire protection facilities and equipment. At the fire protection facilities level, the highest proportion belonged to automatic water sprinklers and fire alarm systems, which highlighted the importance of automation technology in fire safety and underscored the significant role of smart firefighting in enhancing fire resilience. Using smart firefighting led to reducing losses due to fires. (3) In the learning stage, the implementation of regulations and daily management were equally balanced, indicating that both aspects were important for the cultivation of fire resilience in self-built houses. According to individual indicators, the approval status of residential self-built houses had the highest proportion of 0.1232. The approval of residential self-built houses directly affected their construction conditions and fire prevention conditions. It was not only the source of fire risk of self-built houses, but also the key condition influencing fire safety resilience.

3.3. Index Level

Based on the objective situation of residential self-built houses and their own properties as well as relevant literature and evaluation criteria, the selected 27 resilience evaluation indicators were classified into 4 categories and the resilience level increased from 1 to 4 levels. In order to facilitate calculations, the lower limit of the quantitative index range was taken as its relative quantitative value and 1~4 points were given to the qualitative index based on its resilience level from low to high as its grade value.

3.4. Level Evaluation

3.4.1. TOPSIS Principle

TOPSIS is a multiobjective decision analysis method, the essence of which is to explore the superiority or inferiority of a solution by calculating its distance to the optimal solution [23]. Compared with other methods [24], TOPSIS comprehensively considers the relative relationship between the evaluation object and the positive ideal solution and the negative ideal solution, and can evaluate and sort each evaluation object more comprehensively. The basic steps were as follows:

① Weighted normalized matrix $Z_{ij}$ of $Y_{ij}$ was constructed where $W$ is the comprehensive weight given in Table 2.

$$Z_{ij}=Y_{ij}W$$

(2)

② Positive and negative ideal solutions $Z^{+}$, $Z^{-}$ of each index were determined where $z_j^{+}$ and $z_j^{-}$ are the positive and negative ideal solutions of the $j$th term of the matrix, respectively.

$$Z^{+}=\left[\max (z_{ij})\right]=\left[z_1^{+},z_2^{+},\cdots ,z_j^{+}\right]$$

(3)

$$Z^{-}=\left[\min (z_{ij})\right]=\left[z_1^{-},z_2^{-},\cdots ,z_j^{-}\right]$$

(4)

③ The distances $D_i^{+}$ and $D_i^{-}$ between each index and positive and negative ideal solution, respectively, were calculated as:

$$D_i^{+}=\sqrt{{\displaystyle \sum _{j=1}^n{(z_{ij}-z_j^{+})}^2}}$$

(5)

$$D_i^{-}=\sqrt{{\displaystyle \sum _{j=1}^n{(z_{ij}-z_j^{-})}^2}}$$

(6)

④ The closeness $S_i^{+}$ between each index level and positive ideal solution was calculated as:

$$S_i^{+}=\frac{D_i^{-}}{D_i^{+}+D_i^{-}}$$

(7)

3.4.2. Standard Division of Resilience Level

① The initial matrix $Y_{\mathrm{ij}}$ was constructed. According to the index level described in Section 2.3, the initial matrix $Y_{\mathrm{ij}}$ was established as:

$$Y_{\mathrm{ij}}=\left[\begin{array}{cccccccccccccccccccccccccc}1& 1& 0& 1& 0& 1& 0& 0& 1& 0& 0& 1& 1& 1& 1& 1& 1& 1& 1& 0& 1& 1& 1& 1& 1& 1\\ 2& 2& 300& 2& 0.3& 2& 0.3& 4& 2& 1300& 6& 2& 2& 2& 2& 2& 2& 2& 2& 5& 2& 2& 2& 2& 2& 2\\ 3& 3& 600& 3& 0.6& 3& 0.5& 8& 3& 1900& 9& 3& 3& 3& 3& 3& 3& 3& 3& 10& 3& 3& 3& 3& 3& 3\\ 4& 4& 1200& 4& 0.8& 4& 0.8& 12& 4& 2500& 13& 4& 4& 4& 4& 4& 4& 4& 4& 15& 4& 4& 4& 4& 4& 4\end{array}\right]$$

② The initial matrix was standardized and the values of $D_i^{+}$, $D_i^{-}$ and $S_i^{+}$ of each resilience level were calculated using Equations (2)–(7). Resilience level was positively correlated with closeness degree. Therefore, the fire resilience level of residential self-built houses could be classified according to the size of $S_i^{+}$ (

Table 3. Resilience level standard.

Level	${\mathit{D}}_{\mathit{i}}^{+}$	${\mathit{D}}_{\mathit{i}}^{-}$	${\mathit{S}}_{\mathit{i}}^{+}$	Description
poor	5.1962	0.000	0.000	Fire probability in residential self-built houses is high, and the ability of the building to absorb and resist is poor. It is impossible to take timely and effective countermeasures, and it takes a long time to return to a safe state.
medium	3.4122	1.8086	0.346	Fire occurrence probability is high, and the building’s own absorption and resistance ability is medium. It is difficult to take timely and effective countermeasures, and it takes a period of time to return to a safe state.
good	2.598	2.8244	0.521	Fire occurrence probability is small, and the building itself has good absorption and resistance. Timely response measures can be taken, and time to return to a safe state is short.
excellent	0.000	5.1962	1.000	Building is almost a no fire risk, and its absorption and resistance are excellent. Timely and effective response measures can be taken to quickly return to a safe state.

).

4. Empirical Study

4.1. Overview of the Study Area

For the verification of the rationality of the model, a self-built house in Jiaozuo was selected as case for analysis. The self-built house was located in the commercial center near a university in this city. It was surrounded by a comprehensive commercial street integrating shopping, catering and entertainment facilities. The road was narrow and the flow of people was high, which was prone to congestion. It was an important personnel-intensive place and was among the key fire prevention units in local area. The building was a typical three-in-one place, which was established in 2013, and its usable area was approximately 720 m². There were four floors in total, the first floor was for stores on the outside, and the inside was for residents to live on their own. The 2nd–4th floors were for group rentals, which could accommodate 40 tenants. According to the above analysis, through further on-the-spot investigation and in-depth interviews combined with the fuzzy evaluation criteria described in Section 2.3, the classification was completed (

Table 4. Study area index level division.

Index	Description	Level	Index	Description	Level
C1	The service life is more than 8 years, with circuit protection.	2	C15	Inadequate facilities	2
C2	The distribution of dangerous goods is small and scattered.	3	C16	Automatic spray coverage is not complete.	2
C3	1100 (mj/m2)	2	C17	Alarm system, no video surveillance	2
C4	It has not passed the examination and approval, and it is self-built.	1	C18	Using ordinary fire shutters	2
C5	Three-in-one place	1	C19	It is configured according to mild hazard level standard, and the layout is reasonable.	2
C6	The proportion of vulnerable groups is 60%.	2	C20	The path is complex, and the debris is blocked and not easy to pass.	2
C7	Open fire is not often used, but safety awareness is weak	2	C21	12 min	2
C8	People who are familiar with the environment and escape routes account for 40%.	2	C22	Responsibility system is clearly implemented, and residents attach great importance to it.	4
C9	5 (person/100 m2)	3	C23	The approval system is partially implemented and lacks supervision.	3
C10	Most of them are B1 noncombustible materials and B2 combustible materials.	3	C24	Some fire dangers rectification opinions are carried out by residents.	3
C11	1350 m2	2	C25	There are regular maintenance arrangements.	3
C12	6 m	2	C26	There are reasonable and operable emergency plans, occasionally optimized.	3
C13	No fire fighting vehicle can be close to the rescue surface.	1	C27	There are regular personnel training and exercise programs, which are implemented.	4
C14	There is no fire pool, the pipe network is basically reasonable, and the water supply pressure is slightly insufficient.	2

).

4.2. Fire Resilience Level and Result Analysis

The initial matrix was constructed and standardized, and the distances between the fire resilience index of residential self-built houses and positive and negative ideal solutions were calculated using Equations (10) to (14) to be $D_i^{+}=3.161$, $D_i^{-}=2.459$ and $S_i^{+}=0.438$. The resilience level of residential self-built houses was found to be medium. The probability of fire occurrence was high, and the building’s own absorption and resistance ability was medium. It was difficult to take timely and effective countermeasures, and it would take a period of time to return to a safe state.

In-depth interviews with residents of the self-built building showed that a small fire had occurred since the building was built, which did not cause casualties, but economic losses were large. Consistent with the evaluation results of this research, the fire safety resilience level of residential self-built houses was medium and needed to be further improved. At the same time, it was also shown that the fire safety resilience evaluation model of residential self-built houses developed in this research had certain feasibility.

5. Conclusions

In the context of the current global construction of resilient cities, this research adopted the resilience concept to the field of fire safety in residential self-built houses. A fire safety resilience evaluation system was constructed to explore the fire safety level of residential self-built houses. In terms of resilience, the developed model provided a basis for the quantification of the fire safety resilience of residential self-built houses and put forward new ideas for their fire protection. The findings of this research were as follows:

The causes of fire were evaluated by combing the typical cases of self-built houses. Combined with relevant literature, four types of fire safety problems of residential self-built houses were identified, which included congenital deficiency of protection conditions, lack of facilities, complex personnel and chaotic functions.

To solve the fire safety problem of residential self-built houses, this research comprehensively considered resilience theory and the properties of building itself, and constructed a fire safety resilience index system for residential self-built houses in three stages of resistance, response and learning.

According to the combination weighting–TOPSIS method, the fire safety resilience evaluation model of residential self-built houses was developed, and four resilience levels were classified, which included poor, medium, good and excellent. The results showed that the fire safety resilience level of residential self-built houses was low, which was mainly manifested in the lack of coping and restoring learning ability.

The model was applied to a residential self-built house in Jiaozuo City as a research object. The results revealed that the fire resilience level of the building was medium, which was consistent with the actual situation and the objectivity and rationality of the model were verified.

According to the comprehensive consideration of the three dimensions of resistance, response and learning, this study provided some ideas and methods for quantifying the fire safety resilience of residential self-built houses, but there were also some limitations. (1) During index screening, certain indicators were obtained through field research and in-depth interviews in Jiaozuo City, and the research scope was narrow. Therefore, the proposed index evaluation system was only applicable to self-built houses with medium and backward economic levels, and there were big differences among residential self-built houses in first-tier cities. (2) The number of verified cases was small. In the future, empirical research can be conducted on different types of residential self-built houses in different regions to further optimize the fire safety resilience evaluation model of residential self-built houses.