1 Introduction

Egypt is very rich with archaeological sites which contribute significantly to the tourism industry. However, the management and conservation of these sites is a challenging task due to the ever-increasing threats such as population growth, urbanization, pollution, natural hazards, among others. These threats result in the gradual deterioration of these sites, and hence results in significant economic losses.

Egypt is classified as an area of moderate seismicity with limited number of damaging earthquake events through its history. Many buildings were designed considering gravity loads only till the inclusion of the first seismic provisions in design codes in 1989 in addition to the existence of non-engineered buildings especially in rural areas. Hence, countless buildings are recognized to be seismically vulnerable. Assessment of the seismic risk is a rising concern considering the deteriorated conditions of many historical structures in Old Cairo. Assessment studies are considered as major element in devising prevention strategies and forecasting damage expected during seismic events. Vulnerability reflects the probability of a structural system to take damage upon event occurrence. It is usually considered as the opposite of system resilience. Evaluation of seismic vulnerability could follow detailed approaches which is usually used for individual buildings. However, detailed evaluation cannot be practically carried out at a wide level. Simplified procedures are employed for larger scale analyses to provide knowledge of seismic vulnerability. Commonly, a vulnerability index is defined to reflect the capacity of a building to endure earthquake actions associated with an expected limit state.

The research project ICONIC [1] was initiated as a collaborative research effort between Cairo University, Egypt and Imperial College, London in addition to Athar Lina initiative aiming at proposing and applying an interdisciplinary approach for the management and conservation of the UNESCO world heritage site of Historic Cairo: Al Ashraf Street within Al Khalifa district. The project included several work packages and activities including management plans, structural health monitoring, structural assessment, experimental studies in addition to numerical analyses.

The current study is initiated by reviewing seismicity of Egypt and its relation to the study area. Observations related of consequences of previous events on historic structures are explored. The following section provides overview of seismic vulnerability assessment method and details of the method adopted in the current research. Section 4 and 5 list building stock inventory and characteristics of the investigated buildings. Information related to the structural system, construction material, usage, degree of deterioration…etc., was collected. Details regarding the inspected buildings are exhibited and analyzed. In Sect. 6, the vulnerability indices are calculated, and vulnerability map for the area is developed. This provides a clear presentation of the critical zones of a large-scale territory. In the following section, local mechanisms of failure are investigated for the inspected buildings.

2 Seismicity of Egypt and Cairo

Seismic hazard in historic Cairo is considered low to moderate with earthquake magnitudes not exceeding 7; however, it is posed as one of the main threats to existing structures due to their high vulnerability and deteriorating condition. In addition, the population in Egypt is condensed within a very narrow belt along Nile Valley and Delta where most of earthquakes occurred. Soil characteristics lead also to seismic amplification. Many studies investigated the seismicity and mechanisms of earthquakes generally in Egypt. The historical seismicity before availability of records was established by checking major events described in historical records [2,3,4,5]. Historical documents indicate that large destructive earthquakes occurred quite frequently with documentation of earthquake events dating back to the last 48 centuries or more [5]. Ambraseys et al. [3] assigned a magnitude of 5.8 to the largest known earthquake that occurred in 1847. Studies show that, in the time period 1906–1987, 956 earthquakes of magnitude ≥ 3 were reported. Twenty-two earthquakes with intensity ≥ VII and probable magnitude ≥ 4.5 are expected to have occurred during 320–1906. In addition, four major earthquakes with magnitude ≥ 5.0 were reported during the period 1778–1905. Figure 1 shows the estimated number of earthquakes in the time period 1906–1987 as reported by [6]. Recently, the most remarkable earthquake events included the 12 September 1955 offshore Alexandria Earthquake, the 31 March 1969 Red Sea Earthquake, the 9 December 1978 Gilf El-Kebeir Earthquake, the 14 November 1981 Kalabsha (Aswan) Earthquake, the 2 July 1984 Dabbab Earthquake, the 12 October 1992 Dahshour (Cairo) Earthquake, the 22 November 1995 Gulf of Aqaba Earthquake and the 28 May 1998 Ras El-Hikma Earthquake. Details of magnitudes, human losses and recorded damage are summarized in [7]. Researchers also investigated the seismic active zones in Egypt and the tectonic situation [6, 8,9,10]. The reported most active zone in Egypt is the Gulf of Suez with historically recorder earthquakes in the magnitude range 4.0–6.2 [9]. Egypt has a unique tectonic situation being located at the zone of convergence of the African and Eurasian Plates, the strike-slip movement along Gulf of Aqaba-Dead Sea Transform Fault to the east and Gulf of Suez Sea rift, Fig. 2.a. Hence, seismic activity of Egypt is mainly due to the interaction and relative motion between these plates. Considering the above, regions of high activity can be identified on the Gulfs of Suez and Aqaba, the junction between Red Sea, Gulf of Suez, and Gulf of Aqaba in addition to the area around Dahshour area, Fig. 2.b. Meanwhile, moderate seismic activity is reported in Aswan, Abu-Dabbab and Abu-Simbel [11]. Generally, earthquake activity in Cairo and its surroundings is lower than other regions in North Egypt. Most of the recorded events are sourced from the southwestern part of Cairo and around Dahshour area. The available studies show that seismic threats originate from shallow-crustal, intra-plate events concentrated around south-west of Cairo [12].

Fig. 1
figure 1

Temporal annual variation of seismic activity, m ≥ 3.0, based on instrumental data: a Gulf of Suez; b other active zones in Egypt [9]

Full size image
Fig. 2
figure 2

Tectonic setting and seismic activity in Egypt

Full size image

Seismic records in Egypt are usually divided into two phases: historical or pre-instrumental phase (before 1900) and the instrumental phase (after 1900 till now). During the historical phase, information regarding earthquakes is inferred from books, papyri, Arabic chronicles in addition to evidence provided by temples and monuments. Most of the reported earthquake occurrences were during the Islamic period after 622 A.D. Details of such events could be found in different studies. During the instrumental phase, records of seismic activities led to recording seismic events even in uninhabited area and discovering new possible sources. Instrumentation started by component Milne-Shaw seismograph at Helwan in 1899 and developed till covering the whole territory of Egypt with strong motion network started after the 1992 Cairo Earthquake. Afterwards, the Egyptian National Seismic Network (ENSN) was established in 1997. Since then, full record of all seismic activities has been established.

The main parameters of seismicity of a region are geographic distribution of earthquakes’ foci, magnitude, occurrence over time in addition to mechanisms and damage produced by them. The focal mechanism is considered One of the most important parameters as it helps in defining nature of earthquake faulting and mapping the tectonic structure. Saweris et al. [7] compiled data from different studies into a single map showing focal sources across Egypt and around Cairo area. A total number of 688 focal mechanisms were reported covering different active seismogenic zones. Most of them had a magnitude greater than or equal to 3.0. Studies related to occurrence over time report a wide range of return period values for earthquakes with magnitude 6.0 in Suez Gulf ranging between 53 to 210 years [6, 9, 13]. Meanwhile, the maximum expected magnitude for the rest of active zones in Egypt is 6.1 ± 0.3 for a time span of 1668 years. Generally, major earthquakes are expected on average once in eight and in 21 years for Suez Gulf and rest of Egypt, respectively [9].

While Egypt is a country of moderate seismicity, limited number of damaging earthquakes were experienced among which Dahshour 1992 and Aqaba 1995. During these events, many historic structures suffered from damage and/or partial collapse. Damage and collapse of structures were especially documented upon the 12 October 1992 Dahshour (Cairo) Earthquake [14, 15]. Reports showed that at least 560 people lost their lives, and many thousands were injured. Recorded damage included collapse of top parts of minarets, deep through cracks, wall out-of-plan movements, bulging of walls, cracking of domes, and collapse of wooden roofs. Figure 3 shows sample of the observed damage upon the earthquake event.

Fig. 3
figure 3

Miscellaneous observed damage upon 1992 Cairo earthquake [14]

Full size image

3 Seismic vulnerability assessment methods

Methods used to derive vulnerability of a structure can be divided into four main groups: judgmental, analytical, empirical and hybrid methods [16]. In judgmental methods, vulnerability indices are determined based on information on damage level estimated by experts through in-situ surveys. The reliability of this method is questionable as it may result in different conclusions based on the opinion of assigned experts. However, it is suitable for scanning structures at urban scale [17]. Analytical methods employ statistical of results of numerical analyses on detailed or simplified structural models [18,19,20]. The reliability of results depends on the ability to capture the real performance of the concerned structure. Empirical methods depend on surveys and observations after an earthquake event. Hence, vulnerability indices are calculated based on statistical procedures. These are considered the most realistic approaches as they include real response of exposed building [21]. Hybrid methods combine several methods such as empirical and judgmental procedures [22, 23].

Vulnerability studies play a key role in reducing risks of natural hazards on human communities. Quantifying vulnerability of buildings dates to 1956 through the Modified Mercalli scale by Richter [24] followed by MSK scale in 1964. In these methods, buildings are divided into classes based on type, construction material and used lateral load resisting system. These scales are considered the base for one of the most common seismic vulnerability methods: the European Macro-seismic Scale, EMS-98. It provides three analysis levels for seismic vulnerability. In the first level, the evaluation is based on a vulnerability index which depends on building class within the Building Typology Matrix (BTM). Several factors, including construction local characteristics, state of maintenance…etc., are accounted for considering a behavior modifier factor. Hence, the Damage Probability Matrices (DPM) are built. Accordingly, for a given seismic input, the damage representation can be identified. In the second analysis level, mechanical methods are employed through detailed investigations of building parameters as detailed in HAZUS 99 [25] estimating potential building and infrastructure losses due to natural hazards.

Determination of seismic vulnerability was investigated and applied by many researchers through the Probabilistic Seismic Hazard Assessment (PSHA) for seismic hazard analysis. Some studies were general [26] while others focused on certain zones in Egypt [27, 28]. The Italian GNDT vulnerability method 1993 [29] represents one of the most common methods. In this method, a vulnerability index that relates seismic action intensity to expected damage of structure is defined. The index depends on 11 parameters including type and organization of resisting system, Quality of resisting system, conventional strength, building position and foundation, horizontal diaphragms…etc. Equation (1) shows the basic formula for calculation of the vulnerability index [29].

$${I}_{v}^{*}= \sum_{i=1}^{11}{C}_{vi}*{P}_{i}$$
(1)

where Cvi is the vulnerability class and Pi is the weight associated to the corresponding parameter as indicated in Table 1 and 2 for masonry and RC buildings, respectively.

Table 1 GNDT Method Classes and Weights for Masonry Buildings [30]
Full size table
Table 2 GNDT Method Classes and Weights for RC Buildings [31]
Full size table

In the current study, the GNDT approach including 11 parameters was adopted for seismic vulnerability determination. Values of weights are adopted as per Benedetti and Petrinin [30]. Collected data include information regarding the inspected building, type of construction material, number of stories, footprint dimensions, type and characteristics of lateral load resisting systems, characteristics of non-structural systems…etc. Accordingly, a class (A, B, C or D) is assigned to each parameter. For reinforced concrete buildings, different classes (Cvi) and weights were adopted following Cara et al. [31]. Different parameters and weights are considered to account for the difference between expected vulnerability of masonry and RC buildings with respect to importance of each parameter. These parameters are meant to describe the deficiencies and the faults of the structure depending on expert judgement. It is worth mentioning that methodology of GNDT was initially established for masonry structures and then extended to cover RC buildings [31].

The vulnerability index is then normalized as follows for masonry buildings [29]:

$${I}_{v}= \frac{{I}_{v}^{*}}{382.5}*100$$
(2)

For RC buildings, the vulnerability index is normalized as follows to transform the vulnerability index of RC structures to masonry vulnerability indices [31]:

$$ {\text{If}}\,I_{v}^{*} > - {6}.{5}:{\text{ I}}_{{\text{v}}} = - {1}0.0{7}*I_{v}^{*} + { 2}.{5175} $$
(3)
$$ {\text{If}}\,I_{v}^{*} < - {6}.{5}:{\text{ I}}_{{\text{v}}} = - {1}.{731}*I_{v}^{*} + {56}.{72} $$
(4)

For the buildings including mixed structural systems including both masonry and RC systems, the minimum of both vulnerability indices is considered.

4 Building stock inventory

The study area, Al Ashraf Street, is located in Al Khalifa district in Fatimid Cairo. In the current study, focus will be given to buildings located et al. Ashraf Street province. Masonry and reinforced concrete buildings were observed during the preliminary site visits.

The study area was divided into five zones (A, B, C, D and E) as shown in Fig. 4. Zones were defined considering the main monuments across the street: Fatma Khatun Dome, Al-Ashraf Khalil Dome, Shagaret El Dor Dome‬‏ and Sayyeda Soukaina Mosque. Table 3 shows the number of buildings considered in each zone.

Fig. 4
figure 4

Distribution of zones at studied area

Full size image
Table 3 Number of buildings assigned at each zone
Full size table

4.1 Characteristics of study area

The main characteristics of the inspected buildings were registered using visual inspection checklists. Focus was given to evaluating the 11 parameters related to the GNDT method using prepared checklists. In addition, data were collected considering usage of building, structural material, type of vertical and horizontal structural systems, number of stories…etc. Table 4 lists the main characteristics of inspected buildings. The following observations can be outlined:

  • Most of the inspected buildings were used as apartment buildings (88.2%). Other usages of areas included mosques (3.9%), Qubba (2.9%), cementery (1.0%), Wekala (1.0%), Minaret (1.0%) in addition to closed buildings (2.0%). Most of the buildings included random commercial activities at the ground floor level. Some workshops were also observed.

  • The number of floors of inspected buildings ranged between one to eight floors. Most of the buildings had 1 to 4 floors (75.5%). Meanwhile, it is worth noting that even medium rise buildings did not include a clear and continuous lateral load resisting system. This is attributed to the fact that these buildings were randomly built without following the common engineering practice including design, issuing permits, supervision on construction…etc.

  • When inspecting the type of vertical structural system, it was found that 10.8% of the buildings had mixed structural systems including masonry at the bottom floors and reinforced concrete columns and beams at the upper floors or vice versa. This was attributed for the fact that several recent buildings were built as extension of existing old buildings. Meanwhile, 14.7% of the inspected buildings included reinforced concrete frame systems composing of columns and beams. However, the quality of design and construction is questionable for most of the buildings. The most observed structural system (68.6%) was wall-bearing.

  • Most of the inspected buildings (58.8%) were constructed using block masonry. Reinforced concrete buildings represented 14.7% of the inspected building stock. Plastered dressed stone and dressed stone materials composed 9.8% and 7.8% of building population, respectively. Rest of buildings were constructed using different materials -including wood, steel, adobe, and bricks- which indicated construction activities during different time periods.

  • Limited gaps were observed between adjacent buildings. In addition, they are characterized by different number of floors and plan dimensions. This will result in different periods during an earthquake excitation which might lead to severe pounding effects.

  • Some buildings have heavy overhangs (long and wide balconies) which can lead to increased seismic effects.

Table 4 Building stock statistics
Full size table

5 Vulnerability assessment study results

Table 5 illustrates the distribution of the different parameters for the 102 buildings in the current study. The following observations regarding each parameter can be summarized as follows:

  • Parameter 1 (P1) examines the degree of organization of the building in the vertical direction and whether it includes a system capable of resisting lateral load actions. The inspected building stock can be divided into masonry buildings in addition to reinforced concrete buildings. Most of the reinforced concrete buildings were built and constructed under poor design and construction processes. Some of the buildings did not even include a clear system for resisting lateral loads and were classified as Class D. None of the buildings was classified as Class A due to the lack compliance with the seismic regulations for the new buildings. Rest of the buildings were classified as either Class B or C. This was based on the quality of the connection between orthogonal walls in masonry buildings or the existence of beams connecting columns in reinforced concrete buildings. Accordingly, classification of this parameter was either Class B, Class C and Class D with percentages 40.2%, 50.0% and 9.8%, respectively.

  • Parameter 2 (P2) reflects the type and shape of materials of elements forming walls. Class A represent good-quality and well-woven brick walls, meanwhile, class D represent irregular and badly woven poor-quality brick walls or stone masonry. For the considered building stock, the four categories were observed with percentages 7.8%, 34.3%, 35.3% and 22.5% for classes A, B, C and D, respectively.

  • Parameter 3 (P3) evaluates the resistance of the lateral load resisting elements while considering several parameters including number of floors, average floor height in addition to the area of the resisting system in both directions. Most of the inspected buildings did not include a clear lateral load resisting system leading to D classification for most of the inspected buildings, 74.5%. Rest of buildings were considered as Class C.

  • Parameter 4 (P4) accounts for the influence of the terrain and foundation. The terrain of land at the area was estimated from previous studies to be in the range between 10 to 30%. Accordingly, class B was assigned to most of the buildings except for two cases in which building ground water level was around 1 m above ground water level. For these cases, a lower class (C) was assigned to account for the expected deteriorated condition.

  • Parameter 5 (P5) reflects the quality of connection between horizontal diaphragms and vertical resisting elements in addition to the rigidity of the horizontal diaphragms. Class A represent rigid horizontal diaphragms effectively connected to the vertical elements. Class B is the same as A with the consideration of staggered tops. Class C represent deformable horizontal diaphragms effectively connected to the vertical elements. Class D considers any horizontal diaphragm badly connected to the vertical elements. Based on these definitions, 30.4% of the inspected buildings were classified as class D. Some did not include any proper connection between the horizontal diaphragm and the vertical masonry wall. Many of them included a thin reinforced concrete slab poured directly on bearing masonry walls. Some of the masonry buildings were in good condition due to recent rehabilitation practices.

  • Parameter 6 (P6) is calculated based on building dimensions in plan. Factors β1 and β2 were calculated for the different buildings. The percentages of buildings for classes A, B, C and D were found equal to 20.6%, 37.3%, 31.4% and 10.8%, respectively.

  • Parameter 7 (P7) assesses the irregularity of buildings in plan. It depends on the existence of porches, arcades or changes in floor area. Most of the buildings were classified in class B with a percentage of 85.3%. Only 2.0% of building stock was categorized in class C or D. Most of the inspected buildings had regular rectangular shape in plan with minor changes in area when comparing the different floors.

  • Parameter 8 (P8) accounts for the relationship between master walls and transverse walls. It was calculated based on the ratio relating building inter-axis distance to the dimension of the master wall. The estimated percentage of classes A, B, C and D were 2.9%, 22.5%, 60.8% and 13.7%, respectively.

  • Parameter 9 (P9) classifies shape of roof of building. Classes were assigned based on description of each class category with respect to roof shape. Most of the buildings (96.1%) fall in class B. The remaining percentage was divided equally between classes A and D.

  • Parameter 10 (P10) accounts for non-structural elements or extensions that may cause damage to personal or property. Most of the buildings were considered in class B. This represents buildings without chimneys and with balconies that can be considered as integral part of the horizontal system of the building.

  • Parameter 11 (P11) measures the state of conservation of the building according to the extent of observed defects. 2.9% of the inspected buildings were classified as D due to severe deterioration of material in addition to lead-free walls. Buildings with cracks (2–3 mm) were considered as class C and were estimated as 30.4% of the inspected building stock.12.7% of the buildings were in good condition without visible defects especially for the newly conserved buildings.

Table 5 Distribution of different parameters considered in applied methodology
Full size table

The normalized vulnerability index as calculated for the different buildings is illustrated in Fig. 5. The reported percentages range between 15 and 75. It is also observed that most of the buildings have vulnerability index value ranging between 30 to 40. The minimum and the maximum vulnerability indices were 17 and 70.6, respectively. The mean value of the vulnerability index is 40.2, meanwhile, the standard deviation associated with the data is 11.36. Figure 6 shows the distribution of buildings considering the different vulnerability index values.

Fig. 5
figure 5

Vulnerability index as per GNDT method—Number of buildings

Full size image
Fig. 6
figure 6

Vulnerability Index as per GNDT method—Percentage of buildings

Full size image

Figure 7 shows the vulnerability index mapping for the study area. The map indicates clearly the areas where more vulnerable buildings are located. Zone B which had the poorest conservation condition. The condition of masonry buildings in this zone was deteriorated with visible cracks and collapsed parts. The highest vulnerability index was recorded for Building B-5 with a percentage of 70.6%. Zones C and D had the same average for the vulnerability index around 35. Meanwhile, Zone E had increased values for the vulnerability index with an average of 39.4.

Fig. 7
figure 7

Vulnerability index map (Iv)

Full size image

6 Identified local mechanisms of collapse

Possible local failure mechanisms were explored in the current study. Accordingly, the following main mechanisms were outlined:

  • LMC1: Diagonal cracking in Unreinforced masonry walls.

  • LMC2: Toe compression failure in Unreinforced masonry walls.

  • LMC3: Bed-joint sliding in Unreinforced masonry walls.

  • LMC4: Critical vertical crack between building parts.

  • LMC5: Cracks around Wall Openings.

  • LMC6: Outward Bulging of Walls.

  • LMC7: Outward Detachment of Walls.

  • LMC8: Pounding Cracks between Two Adjacent Buildings.

  • LMC9: Excessive deflection of overhangs.

  • LMC10: Deterioration of construction material and loss of parts.

  • LMC11: Soft story formation

These local mechanisms were identified against each of the inspected buildings in case of observed deterioration. Some buildings had visible cracks and signs of partial failure that maps to one of the identified local failure mechanisms. However, others do not have visible signs of failure. Some buildings were either in good shape or recently rehabilitated. Meanwhile, possible local failure mechanisms can be assigned based on structural system, status, spacing between adjacent buildings…etc. Accordingly, their susceptibility to formation of soft story or pounding was reported. Samples of the observed local mechanisms of collapse are identified in Fig. 8a–u.

Fig. 8
figure 8figure 8figure 8figure 8figure 8

Observed local mechanisms of collapse

Full size image

Figure 9 and Table 6 summarize the assigned buildings to each local failure mechanism. The most frequent case was the deterioration of construction material which may lead to sudden loss of strength. This is followed by cracks around openings which was observed in many buildings. A similar number of cases was assigned the local mechanism of vertical and diagonal cracks which were visible in many buildings. Excessive deflection of overhangs was also observed in 8 buildings. Signs of repair was observed in the form of additional support elements. Several inspected reinforced concrete buildings involved use of mixed systems including reinforced concrete columns in addition to masonry floors. This inhomogeneity in the structural system makes it susceptible to formation of soft story in the event of earthquake excitation. Moreover, many buildings were built adjacently without a proper separation joint. This led to crack due to pounding observed in several buildings. Only two occasions were found for outward bulging of walls and one occurrence of toe compression failure in unreinforced masonry wall.

Fig. 9
figure 9

No. of assigned buildings to each identified local mechanisms of collapse

Full size image
Table 6 Number of assigned buildings to each identified local mechanisms of collapse
Full size table

7 Conclusions

The aim of the current project is to tackle technically advanced yet locally applicable interdisciplinary approaches for threats faced by Historic Cairo including rise of underground water, lack of management system, absence of comprehensive urban conservation plan and inappropriate restoration works while using an interdisciplinary approach. The current study focused on evaluating the seismic vulnerability of a building stock within the area of study. The inspected buildings included masonry in addition to RC buildings. The well-known GNDT method was employed through extensive site surveys to estimate the vulnerability index of more than 100 buildings. The study started by reviewing seismicity of Egypt in general and Cairo specifically showing the attained damage during previous earthquake events. It was shown that lack of adequate seismic design resulted in severe damage in some cases even during earthquakes having moderate magnitude. Afterwards, a large-scale vulnerability study was performed. The high values of vulnerability index, exceeding 70%, pointed the level of seismic risk and low seismic resistance of buildings within the study area. Many buildings possess characteristics that would lead to poor seismic performance including limited spacing between buildings, heavy long overhangs, discontinuity of lateral load resisting system…etc. Moreover, several inspected buildings do not have a clear lateral load resisting system and/or random building extensions using different construction materials. Generally, the area involved buildings of different construction materials including block masonry, dressed stone and reinforced concrete shows the construction works performed in different historic eras. Observations during site visits were employed to study the possible local mechanisms of failure for the considered area with the most observed mechanism related to deterioration of construction material and loss of parts. Other mechanisms included diagonal cracking of unreinforced masonry walls, vertical cracks between building parts, cracks around openings, pounding cracks, excessive deflections, and outward bulging of walls. The results of the study will be incorporated to identify needed rehabilitation works as part of the second phase of the current research project. The current study represents continuous efforts to record condition of buildings in old Cairo and the rigorous planning for possible rehabilitation works to reduce seismic vulnerability of old masonry historic structures in order to preserve lives and heritage within the area.