case study of uttarakhand earthquake 2017

Spoke to officials & took stock of the situation in the wake of the earthquake felt in various parts of North India. Narendra Modi (@narendramodi) February 6, 2017

• CM Rawat added that his administration is still collecting information.
• Uttarakhand Chief Minister Harish Rawat spoke to India Today, saying there is nothing to be worried about. Rawat added that all concerned departments have been alerted and that no loss of life has been reported yet.
• The NDRF has been put on high alert and Home Minister Rajnath Singh has asked for a detailed report on earthquake.
• DG NDRF said two teams have been rushed to Uttarakhand's Rudraprayag district, the epicentre of tonight's earthquake.
• Speaking to India Today Ajay Josh, SP Pithoragarh said, "No reports of damage to property or life have been received so far. Relief and rescue teams have been put on alert. Electricity and telephone e-service remain unaffected by the earthquake. We are keeping close watch on the situation."
• Delhi Chief Minister Arvind Kejriwal has also tweeted about the earthquake.

Earthquake in Delhi NCR. I pray for everyone's safety

Arvind Kejriwal (@ArvindKejriwal) February 6, 2017

Uttarakhand witnessed catastrophic floods and landslides, which killed over 5000 people in 2013. The four districts that were worst affected were during 2013 floods were Rudraprayag, Chamoli, Uttarkashi and Pithoragarh. The hilly state is scheduled to hold an Assembly election later this month.

Also read: Beware North India! Earthquake of 8.2 magnitude might hit the Himalayas soon

Watch video: No loss of lives, says CM Harish Rawat

Uttarakhand has a history of earthquakes but nobody cares!

RUDRAPRAYAG: In the last six years, Uttarkashi, Chamoli, Rudraprayag and several other areas in Uttarakhand have been repeatedly jolted by earthquakes ranging from magnitudes of 4 to 5.1 on the Richter scale. The most recent one on February 6, with a magnitude of 5.8, was enough to bring back memories of the previous times.

Though leading experts studying the fragile Himalayan eco-system are repeatedly giving a clear warning of a major earthquake anytime, politicians, who are busy fighting for votes and promising the moon to the electorate, have no plans in place to get the required infrastructure or to take steps for a better understanding of the eco-system.

As is expected, the day after the earthquake, netas from all parties did not forget to express their ''solidarity'' with the people by visiting their houses in their respective constituencies. For the people and experts, this meant nothing.

''We are expecting a major earthquake, something above six on the Richter scale, which can cause considerable damage. Himalayas are sensitive and unfortunately, our governments have not planned anything despite being hit by major earthquakes in 1991 and 1999. This is a cycle and every 10 years, Uttarakhand will be hit by a major earthquake. It can happen today, tomorrow or in the near future but it will happen,'' says Dr Anil P Joshi, founder of the Himayalan Environmental Studies and Conservation Organisation (HESCO), which has been working in Uttarakhand and other Himalayan states for the past three decades.

A recipient of the Padmashri award, Dr Joshi says that if the government does not regulate the construction of high-rise buildings and hydroelectric projects, it will surely pay a heavy price. ''Uttarakhand falls in the category of high seismic zone and a major earthquake will repeat. By regulating the construction in the hills, we can minimize the effects of earthquake. But unfortunately, our governments wake up only when disaster strikes,'' he says.

After the devastating 1999 earthquake, the Centre had constituted a committee of eminent environmentalists to come up with various methods through which seismic activity could be studied in great depth and appropriate measures could be taken to minimise damage in the wake of major earthquakes.

One such proposal was to set up an observatory centre in the Himalayas to continuosly monitor seismic activites, study the Himalayan eco-system and suggest ways and means to minimise damage. Importantly, the idea was to predict the occurrence of an earthquake.

“I was part of this committee and though all members had unanimously suggested setting up an observatory centre in the Himayalas, the suggestion remained only on paper. Successive governments forgot about it. Even if it is set up now, it would contribute significantly in understanding seismic activities and predicting earthquakes. But the governments have to be serious. They are only interested in getting votes and are least bothered about what happens to the people. Despite the Kedarnath tragedy, we have not learnt any lessons,'' says Dr Joshi.

This committee was formed under the ministry of Science and Technology and Dr Joshi says that the government should wake up now. ''Otherwise, we will pay a heavy price. In the name of development, building high-rises is an open invitation to disaster that is lurking around the corner,'' he warns.

''The whole construction management depends upon the status and nature of the region, demand, resources and, most importantly, the ecological dimensions. This is more important as the Himalayas are a fragile and ecologically sensitive zone. Besides, the constant threat of floods and earthquakes makes it more vulnerable compared to other geographical locations in the country," he pointed out.

He said the recent disaster in Uttarakhand along with the past ones necessitate a total review of the construction approach.

Dr Joshi points out that advanced construction and management approaches might be suitable for cities, but in rural India and especially in the Himalayas, these are not of much use.

Quoting studies, another environmentalist, Ajay Singh Rawat says

Uttarakhand is among the most seismically active parts of the country and many earthquakes of the magnitude 5.5 on the Richter scale have hit the hill state since the year 1900.

''The state straddles several active parallel thrust faults that form the ranges of the Himalayas. These faults have been formed in the highly folded strata of these mountains. The main active features, as far as the ecosystem is concerned, is the main boundary thrust as well as the main frontal thrust. Any slippage between these has generated major earthquakes in the past” he said.

He says the last time an earthquake of the magnitude 7.5 on the Richter scale hit the hill state was some 200 years back.

''One cannot stop nature but yes, we can minimise the damage by taking many steps, particularly, in regard to the construction activity which is going on unabated in the hills,'' he says.

According to one report, of the 13 districts in Uttarakhand—nine of which are in the hills—Chamoli and Bageshwar come under 100 per cent hyper sensitive seismic zones while five districts, including Dehradun, Champawat, Nainital, Uddhamsingh Nagar and Haridwar come under 100 per cent sensitive zone.

''Sensitivity percentages in Rudraprayag and Pithoragarh are 98.3 and 94.9 respectively. Similarly, it is 96.8 in Tehri and Pauri, 83.3 in Uttarkashi and 81.6 in Almora,'' the report says adding that the unplanned constructions in

Dehradun was also making it more vulnerable to earthquakes where the population has increased manifold in the last four decades.

''They have promised jobs, spoken about taking steps to stop migration from the

hills, no one speaks about what steps can be taken to minimise damage and loss of lives in the hills. I think had any party included this aspect in their election manifesto, they would have certainly got several more votes,'' said Amar Chauhan, a hotel manager in Rudraprayag.

WHAT EXPERTS SAY...

The construction approach should be decentralised with the use of local materials and centuries-old local wisdom which reflects in the old structures in the hills. The contractual time-bound approach and ignorance of local conditions and wisdom, especially in public infrastructure, have jeopardised the construction values.

The factors of fragility of the Himalayas and sensitivity of the hills were completely ignored while constructing roads in a hurry. Earlier, the roads were built manually which did not cause much damage to the hills, but the use of JCB machines, electric drillers and dynamites has shaken the base rocks and hill sides, triggering landslides and heavy rains. Looking at the sensitivity of the Himalayas, a green road construction approach must be followed. There should be green treatment of damaged road walls as well.

By treating the roadsides with necessary vegetation and by providing mechanical support, the problems of frequent road debris in every rain can be prevented. In many European countries and in China, road construction is inclusive of green treatment. In the fragile Himalayas, aggressive mechanisation cannot be allowed to preserve longevity of roads.

MAJOR EARTHQUAKES IN THE REGION

July 06, 1505 - Lo Mustang-Globo area, China

Magnitude: 8.2

Heavy damage reported in regions along the China-Nepal border. Tremors felt strongly in many parts of north India and damage was reported from Agra, Delhi, Dholpur, Gwalior and Mathura.

1751 - Daba area (Xizang), China

Magnitude: 7.0

This earthquake has been discovered from Tibetan writings and describes damages in and around the Guge area of southern Xizang (or Tibet) along the border with India.

September 01, 1803 - Kumaon- Gharwal area, Uttarakhand

This earthquake is believed to be one of the strongest in this region. Between 200 and 300 were killed and several villages were buried by landslides and rockfalls. The Badrinath temple located nearly 40 km north of Chamoli was severely damaged. Tremors were felt as far away as Kolkata. Due to liquefaction effects at Mathura in Uttar Pradesh, this earthquake is often wrongly placed in the Mathura area.

May 26, 1816 - Gangotri area, Uttarakhand

Magnitude:  6.5

This earthquake was located south of Gangotri, in the glaciers surrounding the Badrinath peak.

June 16, 1902 - Pokhra - Kainur area, Uttarakhand

Magnitude: 6.0

This earthquake was located south-east of Pauri in Uttarakhand.

June 13, 1906 - Gangotri area, Uttarakhand

Magnitude: 6.1

This earthquake was located near Gangotri, in the glaciers surrounding the Badrinath peak.

July 27, 1926 - Near Changabang Peak, Uttarakhand

Magnitude: 6.5

This earthquake was centred in the vicinity of the Changabang Peak, which lies in the vicinity of Nanda Devi National Park in Uttarakhand.

October 08, 1927- Indo-China border

Magnitude:  6.1

This earthquake was centred north of the town of Dakar, Uttarakhand.

June 04, 1945 - Near Nanda Devi Peak, Uttarakhand

This earthquake was centred in the vicinity of the peak Nanda Devi (Elevation: 7,817 metres)

December 28, 1958 - Rameshwar-Devi Dhura area, Uttarakhand

This earthquake is called the Kakpot earthquake. More than a dozen buildings collapsed. Fissures and landslides were generated in an area within 150 kilometres of Kapkote

June 27, 1966 - Athpali-Dhung area, Uttarakhand

Magnitude: 6.2

This earthquake was centred in Far-western Nepal, along the border with

Uttarakhand

October 19, 1991 - Pilang-Bhatwari area, Uttarakhand

Magnitude: 6.8

768 people were killed and nearly 5,000 injured in this earthquake in Uttarkashi district. Some 18,000 buildings were destroyed in the Uttarkashi-Chamoli region. Landslides and rockfalls were widespread in the Gharwal Hills. Tremors were felt over a wide area of northern India, western Nepal and Pakistan. Minor damage was reported from New Delhi and Chandigarh.

January 05, 1997 - Dharchula area, Uttarakhand

Magnitude: 5.6

Tremors felt strongly in many parts of Uttaranchal, including Nainital, Kumaon and Terai areas.

March 28, 1999 - Chamoli-Pipalkoti area, Uttarakhand

Magnitude: 6.4

115 people killed in the Gharwal region. The quake was felt strongly in Uttar Pradesh, Chandigarh, Delhi and Haryana.

March 30, 1999 - Chamoli-Pipalkoti area, Uttarakhand

Magnitude: 4.9 

50 people were injured in this tremor which was an aftershock of the event on March 28, 1999. Several buildings developed further cracks and many damaged houses at Maithana village collapsed. At Barai in Chamoli district, 20 houses

collapsed and 11 developed cracks, while at Kotiyal 4 houses collapsed and 85 developed cracks.

March 31, 1999 - Chamoli-Pipalkoti area, Uttarakhand

Magnitude: 3.0

One person was killed and several injured in a house collapsed at Hat Pipalkot in Chamoli district. Felt at Chamoli and Rudryaprayag.

May 27, 2003 - Bangina region, Uttarakhand

Magnitude: 5.0

A moderate earthquake struck the Gharwal Himalayas on May 27, 2003 at 09:53 AM local time.

December 14, 2005 - Pokhri-Gopeshwar region, Uttarakhand

A moderate earthquake struck the Gharwal region of Uttarakhand, on December 14, 2005 at 12:39 IST causing minor damage to property in some parts of Uttarakhand. The earthquake had a magnitude of 5.0 and was felt at many places in Uttarakhand as well as in Delhi.

August 5, 2006 - Thal area, eastern Uttarakhand

Magnitude: 4.4

A light earthquake struck the Nepal-India border, on August 5, 2006

causing damage to property in parts of eastern Uttarakhand, India.

July 22, 2007 - Surka Ridge, Uttarakhand

A moderate earthquake struck the Yamnotri region in Uttarkashi district, Uttarakhand, on July 22, 2007 causing a few injuries and minor damages to property in parts of Uttarakhand.

Follow The New Indian Express channel on WhatsApp  

Download the TNIE app to stay with us and follow the latest

Related Stories

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

The PMC website is updating on October 15, 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Sensors (Basel)
• PMC11174874

Uttarakhand State Earthquake Early Warning System: A Case Study of the Himalayan Environment

Pankaj kumar.

1 Centre of Excellence in Disaster Mitigation & Management, Indian Institute of Technology Roorkee, Roorkee 247667, India; ni.ca.rtii.md@ramukp

2 Department of Earth Sciences, Indian Institute of Technology Roorkee, Roorkee 247667, India

Mukat Lal Sharma

3 Department of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee 247667, India; [email protected] (M.L.S.); [email protected] (R.S.J.)

Ravi Sankar Jakka

4 Department of Mathematics, Indian Institute of Technology Roorkee, Roorkee 247667, India; ni.ca.rtii.am@ahbitarp

Associated Data

URLs of the BhuDEV mobile application are as follows: For Android users— https://play.google.com/store/apps/details?id=com.iitr.eews&pcampaignid=web_share (accessed on 28 April 2024); For iPhone users— https://apps.apple.com/in/app/bhudev/id1661902248 (accessed on 28 April 2024).

The increased seismic activity observed in the Himalayas, coupled with the expanding urbanization of the surrounding areas in northern India, poses significant risks to both human lives and property. Developing an earthquake early warning system in the region could help in alleviating these risks, especially benefiting cities and towns in mountainous and foothill regions close to potential earthquake epicenters. To address this concern, the government and the science and engineering community collaborated to establish the Uttarakhand State Earthquake Early Warning System (UEEWS). The government of Uttarakhand successfully launched this full-fledged operational system to the public on 4 August 2021. The UEEWS includes an array of 170 accelerometers installed in the seismogenic areas of the Uttarakhand. Ground motion data from these sensors are transmitted to the central server through the dedicated private telecommunication network 24 hours a day, seven days a week. This system is designed to issue warnings for moderate to high-magnitude earthquakes via a mobile app freely available for smartphone users and by blowing sirens units installed in the buildings earmarked by the government. The UEEWS has successfully issued alerts for light earthquakes that have occurred in the instrumented region and warnings for moderate earthquakes that have triggered in the vicinity of the instrumented area. This paper provides an overview of the design of the UEEWS, details of instrumentation, adaptation of attributes and their relation to earthquake parameters, operational flow of the system, and information about dissemination of warnings to the public.

1. Introduction

The economic repercussions of disasters have significantly increased manifold and are projected to grow continuously in the coming years [ 1 ]. In 2012, the climate-related financial deficit was 1% of the gross domestic product (GDP) of developing countries [ 2 ]. Moreover, losses attributed to seismic activities are also on an upward trajectory. The Himalayan region has experienced numerous devastating earthquakes in the past, and many experts predict their recurrence in the future [ 3 , 4 , 5 , 6 , 7 , 8 ]. Although the occurrence of earthquakes cannot be prevented, their impact on society can be reduced through the application of innovative technological solutions.

The inherent unpredictability of earthquakes renders them more perilous than other disasters. Seismologists have scrutinized many precursors to predict earthquakes, yet none achieved full confidence in all scenarios. Hence, a novel method emerged known as the Earthquake Early Warning (EEW) system. This system serves as a tool to mitigate the risks associated with earthquakes by providing early warnings to the users, potentially saving lives. Fully developed EEW systems worldwide have demonstrated their effectiveness in reducing casualties during earthquake events [ 9 , 10 , 11 ]. The EEW system is a live earthquake monitoring system capable of detecting the onset of an earthquake, estimating its probable magnitude, and issuing warnings before substantial ground shaking reaches the users’ locations [ 12 ]. Its primary focus is on providing timely alerts with a sufficient lead time required to take precautionary measures and shut down key facilities, rather than precisely determining earthquake parameters.

The fundamental principle of an EEW system relies on the differing propagation speeds of primary (also known as longitudinal, non-damaging, P-wave) and secondary (also known as shear, transverse, damaging, S-wave) waves. These waves are produced because of stress release during earthquakes. S-waves inflict notable damage as they travel roughly half the speed of P-waves and are considerably slower than electromagnetic signals. Telecommunication operates at the speed of electromagnetic signals, akin to the speed of light. Therefore, a system can be developed to take advantage of the speed of telecommunication to fetch data from sensors to a server and subsequently, perform analysis and issue warnings before damaging waves reach the users’ locations. The server comprises an assembly of high-performance computers for computational tasks and other telecommunication equipment such as switches and routers. The modern-era telecommunication and data transmission facilities between the sensors and server instill confidence in the efficacy of an EEW system. EEW algorithms operate on real-time data and decision-making modules issue warnings based on predefined threshold parameters [ 13 ]. It is preferable to establish precise threshold values of EEW parameters to obtain accurate, reliable, and fast results [ 14 ]. Lead time or warning time depends on factors such as the epicentral distance from the target locations and the geographical distribution of the stations around the epicenter.

Initially, an EEW system was developed for the Tohoku Shinkansen railway system in Japan in 1982 by employing the front detection technique [ 15 ]. Mexico’s seismic alert system in 1991 also adopted a similar approach [ 16 ]. The Urgent Earthquake Detection and Alarm System (UrEDAS), which employs an improved front detection technique, commenced operations for the Tokaido Shinkansen line in 1992, utilizing the initial three seconds of P-wave motion following the P-wave onset [ 17 ]. After observing the successes of the EEW systems in Mexico and Japan, many other countries embarked on endeavors to develop EEW system customized to their respective regions. Significant progress has been evident in this field over the past three decades. EEW systems are operational in specific countries like Japan [ 18 , 19 ], Taiwan [ 20 , 21 , 22 ], Mexico [ 9 , 23 , 24 ], and South Korea [ 25 ], issuing nationwide warnings that cover a significant portion of the public. However, some countries issue region-specific warnings due to localized seismic vulnerability in certain areas, such as Anatolia in Turkey [ 26 , 27 ], Southwest Iberia [ 28 , 29 , 30 ], Southern Italy [ 31 , 32 , 33 ], Vrancea in Romania [ 34 , 35 , 36 ], Beijing in China [ 37 , 38 ], Chile [ 39 ], Costa Rica [ 40 , 41 ], Switzerland [ 42 ], Nicaragua [ 43 ], Israel [ 44 , 45 ], and the West Coast of the United States of America [ 11 , 46 ]. The advancement of EEW systems is pivotal for reducing seismic risk however enhancing the capacity to provide more lead time is equally essential [ 12 ]. An EEW system’s effectiveness in risk reduction relies on how quickly information is provided to the users, enabling them to have more lead time to react. This can be achieved by developing improved algorithms, estimating EEW attributes, calculating earthquake metadata, and issuing warnings rapidly.

2. Seismic Activity in the Himalayas and the Region of Interest

The Himalayas stand as one of the world’s most seismically active regions. Ongoing convergence over the past roughly 50 million years has led to the formation of numerous complex tectonic clusters in the region. The collision of the continental plate boundaries, specifically the Indian and Eurasian plates, led to the subduction of the Indian plate beneath the Eurasian plate [ 47 , 48 ]. Following the collision period, the Indian tectonic plate has been steadily advancing northward at an average rate of about 50 mm per year [ 49 ]. As a result, deformation occurred in the southward direction leading to the formation of faults, folds, and notable structural characteristics within the Himalayan orogenic belt [ 50 ]. The significant tectonic features include the Main Frontal Thrust (MFT), Main Central Thrust (MCT), Main Boundary Thrust (MBT), and the Indus-Tsangpo Suture Zone (ITSZ) [ 51 , 52 , 53 ]. The ITSZ dips southward, whereas MCT and MBT dip northward [ 54 ]. The MCT was active during the earlier period and is considered the oldest thrust system. Conversely, the MBT currently displays an active thrust system. The MFT represents the southernmost and youngest thrust system. The movement of tectonic plates induces strain along fault plate boundaries, resulting in the accumulation of stress [ 55 ]. Earthquakes occur as a result of the catastrophic failure of rocks and the sudden release of accumulated stress [ 56 ].

The geodynamic activity in the Himalayas has resulted in numerous devastating earthquakes, with one of the most recent being the Nepal earthquake of 25 April 2015, of a magnitude M w 7.6, along with its subsequent aftershocks, which caused substantial economic loss to the Himalayan country. Following this earthquake, the Nepalese government conducted a post-disaster need assessment, estimating the total direct and indirect economic losses to be close to USD 7 billion, equivalent to approximately one-third of the country’s GDP in the year [ 57 ]. Among the notable past earthquakes in the Himalayas is the Kangra Valley earthquake of 4 April 1905, of a magnitude M s 7.8. This event marked the first of several devastating earthquakes of the 20th century. According to estimates by the then Punjab government, that earthquake resulted in approximately 20,000 casualties among the population of 375,000. The economic cost of recovery due to this earthquake was estimated to be around 2.9 million rupees in 1905 [ 58 ].

Similarly, the 12 June 1897, Great Assam earthquake of a magnitude M w 8.0 was very devastating, and its tremors were felt up to the Peshawar region (now in Pakistan). The then British government surveyed post-earthquake damage and stated in the reports that the infrastructure of the eastern region was majorly devastated [ 59 ]. The 15 August 1950 Tibet-Assam Great earthquake of magnitude M w 8.6 shocked the eastern Himalayan region [ 60 ] and caused 1526 casualties and extreme economic loss of around USD 25 million at that time [ 61 ]. The 15 January 1934 Bihar-Nepal earthquake of a magnitude M w 8.0 rattled the border region, and tremors were felt over an area of approximately 4,920,000 km 2 in India, Nepal, and Tibet [ 62 ]. That seismic event resulted in significant destruction, generating numerous fractures and triggering landslides [ 63 ], killing approximately 12,000 people [ 64 ]. In the 1990s, two strong earthquakes rattled the Garhwal region of Uttarakhand. The Uttarkashi earthquake of M w 6.8 on 19 October 1991, and the Chamoli earthquake of M w 6.6 on 28 March 1999, caused significant losses in the Uttarakhand region [ 65 , 66 , 67 ]. Sharma (2003) estimated the return periods of moderate to great earthquakes, with the conclusions aligning with other studies [ 68 , 69 , 70 , 71 ], expressing the increased frequency of occurrence and postulating urgent need and immediate attention to mitigation measures [ 72 , 73 , 74 , 75 ]. The epicenter of the 2015 Nepal earthquake was approximately 450 km from the Uttarakhand region. Considering that Uttarakhand shares similar seismological conditions with Nepal, the occurrence of an earthquake of similar magnitude in Uttarakhand could have devastating consequences. Hence, a lot of efforts are being made to curtail exposure and vulnerability in anticipation of future earthquakes in Uttarakhand.

Since the 1950 Assam earthquake, no great earthquake ( M w 8 or above) has occurred in the Himalayas. Srivastava et al. (2015) analyzed Himalayan seismicity by examining seismic patterns, local tectonics, global positioning system (GPS) observations, microearthquakes, paleo seismicity, and other pertinent datasets. They identified two distinct types of seismic gaps with unique characteristics [ 71 ]. Category-1 seismic gaps are found in regions where significant earthquakes ( M w ≥ 8) have either occurred historically or are anticipated in the future. Specifically, category-1 gaps include the Kashmir seismic gap, West Himachal-Pradesh seismic gap, Uttarakhand-Dharchula seismic gap, Central Nepal-Bihar seismic gap, Arunachal seismic gap, and Shillong seismic gap. Category-2 gaps delineate regions where significant earthquakes ( M w < 8) have either occurred historically or are expected to occur in the future. These include the Jammu, East Himachal-Pradesh, Western Nepal, and Sikkim-Bhutan seismic gaps. The region situated between the 1905 Kangra earthquake and the 1934 Bihar-Nepal earthquake has been identified as the central seismic gap [ 76 ]. GPS measurements conducted in the Himalayan regions indicate accumulation of strain in this area, possibly resulting in one or more significant earthquakes [ 77 , 78 ]. After a thorough examination of the intricacies involved, the utilization of constant seismicity and constant moment rate methods, along with time-dependent occurrence models, has revealed return periods of earthquakes with different magnitude ranges in the Himalayas [ 79 , 80 , 81 , 82 ].

The Uttarakhand-Dharchula seismic gap lies within the central seismic gap and has not witnessed noteworthy major earthquakes in recent recorded history. This seismic gap comprises an approximately 800 km long central segment commonly known as the Garhwal-Kumaon Himalaya. It is often described as an uninterrupted section of the Himalayan arc [ 69 ]. However, the Himalayan region is experiencing swift urban expansion, rapid settlement, and significant infrastructure development [ 83 , 84 , 85 ]. According to the 2011 census, the Indian Himalayan region spanning Kashmir in the north to Arunachal Pradesh in the east is home to around 52.8 million people [ 86 ]. With an annual growth rate of 3.30%, the population is projected to surpass 260 million by the end of 2061 [ 87 ]. In the event of a major earthquake occurring in the central seismic gap region, the potential for thousands of casualties and substantial economic losses amounting to billions of dollars [ 88 ] underscores the urgent need for implementing mitigation measures and disaster risk reduction strategies in this area. Recognizing the heightened seismic risk and dispersed vulnerability within the central seismic gap, an earthquake early warning (EEW) system for this region was proposed. Initially, a two-pronged approach was adopted. First, a feasibility study was performed about the available emerging solutions. Subsequently, the decision was made to establish the first Indian regional EEW system, UEEWS. This paper outlines the development phases of the system, including instances of successful notifications, alerts, and warnings issued.

3. Architecture of the Developed UEEWS

The UEEWS was developed and proposed to address the specific needs of Uttarakhand in central Himalaya after a comprehensive study of the existing EEW systems present in the world [ 89 , 90 , 91 , 92 ]. Due to high seismicity and the pressing need for mitigation measures, the central seismic gap region was chosen for establishing India’s first EEW system. Subsequently, a cost-benefit analysis was carried out, leading to the finalization of sensor selection [ 89 ]. For this, the available strong ground motion recording sensors underwent thorough scrutiny before the procurement process. The cost was a constraining factor, compounded by the significant number of sensors needed for establishing a regional earthquake early warning system. Consequently, low-cost micro-electromechanical systems (MEMS)-based sensors were finalized. The developed system follows the regional early warning approach; hence, a control room was set up at the Earthquake Early Warning System (EEWS) Laboratory in the Centre of Excellence in Disaster Mitigation & Management (CoEDMM), Indian Institute of Technology (IIT) Roorkee.

3.1. Seismic Network

During the initial instrumentation phase, the Garhwal region in the central Himalayas, spanning from Joshimath in the East (30.5616° N, 79.5594° E) to Mori in the West (31.0183° N, 78.0409° E) was selected. This area encompasses approximately 150 by 50 km 2 across five districts of Uttarakhand: Uttarkashi, Tehri, Rudraprayag, Pauri and Chamoli. As a dense network is preferred for a regional earthquake early warning system; therefore, low-cost MEMS-based sensors were selected. These sensors offer a relatively lower dynamic range compared with conventional high-cost force balance accelerometers (FBA) with larger dynamic range. The MEMS-based accelerometer (i.e., pAlert) fulfills the requirement of regional EEW systems and is cost-effective compared with conventional FBAs. These sensors comprise tri-axial MEMS accelerometers paired with a 16-bit 80 MHz CPU, offering an output resolution of 16 bits and a dynamic range exceeding 86 dB. The required EEW parameter (low-pass filtered vertical displacement, P d ) is efficiently calculated from the data retrieved from these sensors. These sensors have undergone testing and have also been successfully deployed in Taiwan’s EEW system [ 93 , 94 ].

The pAlert can transmit data to two servers over the internet by utilizing the pAlert-to-Earthworm module (PALERT2EW) running on the servers and automatically synchronizing its time through the network time protocol (NTP). PC-Utility 1.14 software is utilized for remote access to the pAlert and configuration purposes. In the initial phase, instrumentation of pAlert for the UEEWS was completed as a pilot project from 2014 to mid-2017. Figure 1 shows the location of 84 sensors installed in the first phase. These sensors are mounted on the ground floor of government-owned offices of the base transceiver station (BTS) of Bharat Sanchar Nigam Limited (BSNL) and the point of presence (PoP) of the state-wide area network (SWAN) available in the Garhwal region of Uttarakhand.

Location of the sensors installed in seismogenic areas of Uttarakhand.

To cover the remaining portion of the current seismogenic source, the instrumentation was expanded in mid-2017 to include the Kumaon region of Uttarakhand. The region encompasses four districts: Bageshwar, Pithoragarh, Champawat and Nainital. For this expansion, an upgraded version of the sensor (pAlert+) was chosen. These sensors feature 24-bit tri-axial MEMS accelerometers with an internal memory of 8 GB and dynamic range exceeding 100 dB. The sensors are mounted on the ground floor of BSNL’s BTS and the SWAN’s PoP in the Kumaon region. As shown in Figure 1 , an additional 86 sensors are installed in the Kumaon region, bringing the total to 170 sensors in the entirety of Uttarakhand. The inter-station spacing ranges from about 10 to 20 km. Consequently, the network now satisfies both criteria of the regional EEWS: the necessary sensor density and the spatial coverage of the seismogenic source in Uttarakhand, enabling the detection of large earthquakes in the instrumented region.

The logical structure for established telecommunication from sensors to the control room is shown in Figure 2 . In this figure, the sensors installed in the field are accelerometers, connected by a communication device-an asymmetric digital subscriber line (ADSL) or an optical terminal line (ONT) as per the availability and are connected to the optical line terminal (OLT) at the stations. Networking of the sensors installed in the field passes through different communication media lines like the BSNL cloud, multiprotocol label switching (MPLS) networking, then the OFC network from Haridwar to Roorkee to the installed router, switch, server, and multiple PCs for numerous operations in the laboratory.

Logical structure diagram of the collection system network.

3.2. Data Streaming

The decision to install sensors in BSNL’s BTS and the SWAN’s PoP buildings was driven by the provision of power availability, a dedicated virtual private network over broadband (VPNoBB), and the necessary security required for the sensor as well as associated accessories. All the sensors are installed on the ground floors of these buildings, transmitting ground-motion data to the server located in the EEWS Laboratory. This network operates continuously in real time, 24 hours a day, 7 days a week. Figure 3 illustrates the streaming of data from sensors to servers in real time using Swarm 3.2.0 software. The installed sensors transmit data every second, packaged into 1200-byte packets. As all sensors within the network synchronize their clocks with the UEEWS server, the NTP server adjusts each sensor’s time accordingly.

Streaming of data from sensors to the server in real-time during an earthquake event.

3.3. Data Processing

As the UEEWS marks the first EEW system developed for a Himalayan region, efforts have therefore been made to employ open-source software to facilitate easy adoption in other Himalayan areas. Earthworm 7.10, an open-source platform used to monitor earthquakes and volcanoes and process seismic data [ 95 , 96 ], serves as a module-based seismic network processing platform. Its open-source nature allows the customization of modules to meet specific requirements. Leveraging the flexibility of Earthworm, the central server is set up with the customized existing modules and incorporating new modules designed to suit our earthquake parameter estimation needs.

A two-pole Butterworth high-pass filter with a cut-off frequency of 0.075 Hz is applied over the streamed data in real time to remove low-frequency noises. The PALERT2EW module is used to fetch the streamed data into the shared memory called WAVE_RING, while the WAVE_SERVERV module is responsible for storing and providing seismic waveforms over time, displaying streamed data. Figure 4 illustrates the data flow from the field to the central server. Swarm 3.2.0, another open-source software, plots the accelerograms of the data, offers real-time streaming of data from installed sensors, and supports historical data analysis.

Flowchart of the UEEWS.

3.4. Event Detection

Accurately identifying the onset of the P-phase within continuously streaming seismic signals is crucial for the effectiveness of an EEW system. Many algorithms have been developed in the past for detecting the P-phase onset in seismic signals [ 97 , 98 , 99 , 100 , 101 , 102 , 103 , 104 , 105 , 106 ]. The UEEWS employs an enhanced version of the standard short-term average (STA) and long-term average (LTA) algorithm [ 21 , 107 , 108 ].

The PICK_EEW module is deployed to detect P-phase arrivals in the continuously streaming vertical channel data. This module also estimates peak amplitudes of P-wave displacement ( P d ), velocity ( P v ), and acceleration ( P a ) from 3-s time-windowed data following P-onset, then forwards this information to another shared memory named PICK_RING. The PICK_EEW module [ 21 ] is an improved version of the Earthworm’s default PICK_EW module. The PICK_EEW module applies the standard STA-LTA algorithm [ 107 , 108 ].

Due to potential variations in noise levels caused by natural surroundings and human activities at different installation sites, the PICK_EEW module incorporates two additional parameters, P a and P v , to mitigate false detections triggered by surrounding disturbances [ 21 ]. This module reads a configuration file containing the threshold values of the parameters to obtain a true pick of the P-onset. The parameter values vary for each station, so the configuration file contains their respective values. Thus, site-specific adjustments to the configuration file have been made based on the analysis of earthquake data collected since 2014, the year of commencement of this project. This involved replaying recorded data on an offline server to facilitate improvements in the configuration file for each station.

3.5. Estimation of Earthquake Parameters

A module named TCPD was created in the C programming language to estimate earthquake parameters, designated to integrate seamlessly with the Earthworm platform [ 96 ]. This module underwent modifications to align with the requirements of the UEEWS. This module accesses the PICK_RING and performs analysis to estimate earthquake parameters including location, magnitude, depth, and origin time. Subsequently, it updates the information obtained from this ring and transfers it to another memory location referred to as the EEW_RING.

3.5.1. Location Estimation

The process of estimation of the earthquake hypocenter involves two main steps. First, the Geiger method, an inversion process, is employed to determine the epicenter [ 109 ]. This method requires a P-wave velocity profile specific to the instrumented region. For this purpose, the half-space velocity model recommended by a study conducted for the Uttarakhand region has been adopted [ 110 ]. In the second step, the grid search method is utilized to ascertain the actual depth of the earthquakes, ranging from 0 to 50 km. Although alternative methods like the back azimuth method employing principal component analysis are advantageous for locating long-distance earthquakes, they are deemed unsuitable for near-source recordings [ 111 ]. In the grid search method, the theoretical travel time to each triggered station is computed and compared with the observed depth at each iteration. The depth of the earthquake is determined by identifying the step at which the minimum residual is encountered. Initially, at least four triggered stations are used, and the search point that yields the minimum residual is selected as the focal point. For the UEEWS, the search region is confined within a radius of 200 km from the position of the initial trigger with a depth of 50 km, ensuring that the search area remains within the array. However, in the case of earthquakes originating from outside the instrumented area, the search region is extended beyond the instrumented area, but within 200 km from the location of the first trigger. The limit of 200 km was set up after considering the area of the instrumentation. The expansion of instrumentation spans 280 km east-west and 130 km north-south. If an earthquake occurs in the border area, its vibrations will spread in all directions and will be recorded by sensors installed in the border area. With the help of this recorded data UEEWS server will calculate the location and other parameters of the earthquake.

This approach provides an epicenter’s accuracy of approximately 5 km within the search region. Hence, precise locations can be determined for earthquakes originating outside but in close proximity to the instrumented region. Its capability to achieve an accurate location with approximately 5 km precision in focal depth, coupled with its reduced computational requirements compared with classical methods, makes it well-suited for the EEW system [ 21 ].

This entire procedure is encapsulated within a function in the TCPD module, which is invoked when there are P-phase picks from at least four stations. Upon meeting this condition, the TCPD module initiates the process to locate the hypocenter. If the estimated root mean square (RMS) of travel-time residuals obtained from the inversion process surpasses the specified threshold, the pick with the highest travel-time residual is discarded. Meanwhile, if a new pick is identified, the procedure to locate the hypocenter restarts again and updates the previously estimated hypocenter.

3.5.2. Magnitude Estimation

Once the hypocenter is determined within the TCPD module, another function is invoked to estimate the magnitude. This function utilizes a regression model to calculate the magnitude of earthquakes. The mathematical expression of the employed model is as follows:

In this equation, R represents the hypocentral distance, calculated as the square root of the sum of the squares of epicentral distance ( d ) and focal depth ( h ), and is estimated as R = √( d 2 + h 2 ). P d is the estimated peak displacement using three seconds of data after P-onset. A, B, and C are coefficients in Equation (1) and vary depending on the dataset used as well as the region.

Initially, Wu and Zhao (2006) utilized this model to estimate P d -based magnitude using Southern California Seismic Network (SCSN) data in the following manner [ 112 ]:

A new model was devised to calculate magnitude based on the peak displacement ( P d ) of the P-wave’s vertical component, utilizing 70 records of 13 earthquakes that transpired in the Uttarakhand region between 2005 to 2020. The attenuation relationship of P d with hypocentral distance ( R ) and magnitude ( M w ) is given below (Equation (3)).

The dataset used in the development of this new model is limited and includes records of 3.6 < M w < 5.5 with 15 < R < 100. Thus, the model is a good fit for small earthquakes (3 < M w < 5). After inverting this new relationship, M Pd can be estimated, and it has a 1:1 relationship with catalog magnitude ( M w ).

Due to the insufficient number of large earthquakes recorded in the Indian dataset for the Uttarakhand region, the model proposed by Hsiao et al. (2011) is currently being utilized in the UEEWS to estimate P d -based magnitude (Equation (4)) [ 113 ].

3.5.3. Report Generation

The TCPD module generates a report and stores it in a location specified by the server administrator for archival purposes. Once the report is generated, a warning is promptly issued to the public. A threshold of magnitude five has been set for issuing warnings. The earthquake report contains essential parameters such as latitude and longitude of the epicenter, depth, origin time, list of triggered stations, P-wave arrival time, etc. A decision-making flowchart for issuing warnings is depicted in Figure 5 . The parameters of small earthquakes detected by the UEEWS are outlined in Table A1 . For comparison purposes, the source parameters for the same earthquakes, as reported by the National Center for Seismology (NCS), Government of India, are also provided in Table A2 .

The decision-making flowchart of the UEEWS.

4. Warning Modes

Warnings to the public can be disseminated through multiple modes, including application-specific sirens, television broadcasts, AM/FM radio, mobile messages, and mobile app. In the present setup of the UEEWS, individuals receive alerts by two modes: sirens and mobile app. Sirens are strategically installed at government-owned buildings, schools, hospitals, and residential complexes, while the mobile app is publicly available to download in smartphones.

4.1. Sirens

The EEWS Laboratory indigenously developed the siren units. The design of the UEEWS siren is divided into four main modules [ 114 ]. First, the controller board manages the siren’s operation and stores various warning sounds. The second component consists of an amplifier circuit, which boosts the sound signal from the controller board. Following this is the speaker/hooter, which converts these electrical signals into audible sounds. Lastly, the power supply unit converts 230 V AC into DC, necessary for powering the controller and amplifier. Figure 6 depicts a flowchart outlining these modules.

Flowchart depicting the modules of the UEEWS siren.

The controller board functions as a microcomputer, akin to a Raspberry Pi, linked to the internet through a LAN port. It includes general-purpose input-output (GPIO) pins for interfacing with other devices. A computer program communicates with the warning server over the internet to receive real-time warning messages and manages hooter/speaker relays via GPIO pins. Figure 7 illustrates the logic diagram outlining the functionality of the computer program.

The schematic diagram for the sirens.

Figure S1 illustrates the installation of sirens across various locations, including the State Emergency Operation Center (SEOC) in Dehradun, the capital city of Uttarakhand, all District Emergency Operation Centers (DEOCs) within the state, as well as key government buildings in the two major cities of the state, Dehradun and Haldwani. Figure S1 is given in the Supplementary Material .

4.2. Mobile Application

The Uttarakhand government, in collaboration with IIT Roorkee, has installed public sirens in two major cities, Dehradun and Haldwani, as well as at various DEOCs. However, due to the extensive area and large population, a considerable number of sirens would be needed to cover Uttarakhand, requiring significant investment of both funds and time. This prompted us to explore alternative options to enhance the effectiveness of the developed system. In today’s digital world, Android and iOS-based smartphones are exceedingly prevalent and owned by a significant portion of the population. Therefore, it was deemed important to issue warning messages on mobile app also. The developed app should include some unique features like loud distinctive sound along with voice message when warning is received, so that users get cautious immediately. The app should also provide guidance on actions to be taken in the event of an earthquake and instructions on identifying safe areas within the home for taking cover. Additionally, the app should feature a graphical interface displaying a visual representation of the remaining time before the S-wave reaches the user’s location.

Therefore, IIT Roorkee has developed a smartphone app called BhuDEV (Bhukamp Disaster Early Vigilantè) to disseminate earthquake early warnings, with the aim of alerting users so that they could take precautionary measures before the arrival of damaging waves at their locations. To receive such warnings, users need to install the app and provide essential details during the registration process. The app also features instructional videos for users to take proper safety steps for themselves as well as their loved ones. Currently, the app provides early warnings for damaging earthquakes originating exclusively in Uttarakhand.

When the server anticipates an imminent earthquake with a magnitude of five or above, public warnings are issued indicating potential damage. Similarly, notifications are sent out for earthquakes of lower magnitude, below five. This app needs the current location of smartphone users to dispatch SMS alerts to the registered contacts on press of SOS button to share their present location with the Uttarakhand State Disaster Management Authority (USDMA) during earthquake emergencies. Consequently, permission for location access should be given during app installation and maintained continuously. Hence, granting access to location sharing is essential. This sharing of location assists search and rescue teams in promptly commencing operations during post-earthquake operations. The app receives warnings via the internet, requiring users to always maintain an internet connection. However, data usage is limited to earthquake notifications and warnings only. Additionally, the app helps in simulating earthquake scenarios during mock exercises and provides video links for preparedness and better comprehension of earthquakes.

On 4 August 2021, the honorable Chief Minister of Uttarakhand officially launched this app to the public. The app can be downloaded either by scanning the QR code or by accessing it from the Play Store or App Store. The app icon is provided in Figure S2 for easy identification, and a few pages of the mobile app after installation are showcased in the Supplementary Material .

5. System Application Effectiveness—The Case of the Tehri Garhwal Earthquake Warning

Lead time denotes the duration of warning time offered to the users during earthquake events. It indicates how long it takes for S-waves to reach a specified location, which varies depending on the distance of human settlements from the epicenter of the earthquakes. Settlements at far distances get more lead time compared with those nearest to the epicenter. The current system operates based on the regional EEWS concept, requiring a few seconds to make decisions. These seconds include the total time required to gather three seconds of data from at least four sensors, and time lapsed in the following processes: transmission of data to the server, their processing, estimating magnitude and hypocenter, and subsequently relaying the decision to the warning server for issuing warnings to the public.

The lead time ( Tw ) can be calculated using the formula: Tw = Ts − Tr . Here, Tr represents the reporting time and is estimated as Tr = Td + Tpr , where Td signifies the time required to trigger and record an adequate length of waveforms, and Tpr is the processing time needed to process the waveforms for determining the hypocenter and magnitude. Ts denotes the travel time of the damaging S-wave, and for the advanced warning, it is essential to have Tw greater than 0. This condition implies that Ts must exceed Td + Tpr . Settlements with Tw < 0 are categorized as blind zones, suggesting the necessity for an onsite warning system. The EEWS Laboratory is currently engaged in developing such a system. The lead time is further explained below, accompanied by a detailed example.

Figure 8 shows a diagram illustrating discrete times based on data from the Tehri Garhwal earthquake that occurred on 6 November 2022, at 03:03:03 UTC, of a magnitude M w 4.5 and a depth of 5 km. From the epicenter, the Maneri (MNRB) station was 10.27 km away, the Chamba (CMBS) station was 40.87 km away, and the Artola (ARTB) station was 166.19 km away. In this earthquake, the first P-onset was detected at the MNRB station, while the last P-onset was registered at the CMBS station during the initial report generated by the TCPD module. Consequently, the time required for triggering a sufficient number of stations and recording sufficient primary wave data ( Td ) was 7.81 s from the origin time. The processing time ( Tpr ) needed to calculate the hypocenter and magnitude was 4.38 s. Hence, there was a lead time ( Tw ) of approximately 38 s for the ARTB station before the arrival of destructive waves, which took approximately 50 s ( Ts ) from the origin time to reach this station.

The diagram illustrates the data-recording time ( Td ), data-processing time ( Tpr ), event-reporting time ( Tr ), target-area lead time ( Tw ), and shear-wave travel time ( Ts ) during the Tehri Garhwal earthquake on 6 November 2022.

The TCPD module promptly generates earthquake reports upon detecting P-onset signals from at least four stations. It initiates the earthquake parameter estimation process and generates a report accordingly. Upon report generation, the information is relayed to a warning server, which, based on the estimated magnitude, issues either alerts or notifications. Warnings are issued for earthquakes with a magnitude exceeding five, while notifications are dispatched for those below magnitude 5.0. If the TCPD module detects another new P-onset, it reinitiates the parameter estimation process, generates earthquake reports, and disseminates warnings/alerts to the public.

During the Tehri Garhwal earthquake on 6 November 2022, the TCPD module generated six reports. Seismic data from this earthquake were recorded at 96 stations. Post-processing was performed to estimate the timings of P-onset and S-onset from the accelerograms of each recording station. Following this, the time taken for the S-waves to reach these stations after issuance of notification was estimated. The lead time was then calculated based on the time difference between the issuance of warnings and the arrival of S-waves. Figure 9 illustrates the lead time for all functioning stations at that time. The pink oval indicates a blind zone area where no notifications could have been provided because the processing time by EEWS to estimate parameters and issue warnings was more and, in the meantime, S-waves had already crossed those locations.

Estimated lead time for cities during the Tehri Garhwal earthquake on 6 November 2022. The plots ( i – vi ) represent the lead times users at different locations got during this earthquake. The pink ovals represent the blind zone.

Figure 10 displays a screenshot of the notifications received during this event on a user’s mobile app. Notifications issued to the public commence from the third generated report onwards, while for the initial two reports, UEEWS waits to confirm that they are indeed generated because of an earthquake. The user’s mobile time is shown in the hh:mm format without seconds. For sake of precise information, the timing of the notifications issued by the server is provided in the Indian Standard Time format.

The screenshot of the notification received on the mobile app by a user.

6. Performance of UEEWS

The dense instrumentation of the UEEWS ensures that any earthquake above magnitude 4.0 in the instrumented region would be recorded by at least four stations. On 8 February 2020, 01:01:50 UTC, a light earthquake with a magnitude M w 4.7 and a depth of 48.2 km struck Pithoragarh district of Uttarakhand. Figure 11 illustrates the epicenter and the triggered sensors’ locations, while Figure 12 displays the recorded accelerograms by the vertical channel of the sensors. An early notification was automatically issued solely to a close group of researchers since the system was not unveiled to the public at that time.

The positions of the triggered sensors along with the epicenter of the earthquake that occurred on 8 February 2020, in Pithoragarh. Each triggered sensor is labeled with three rows: the first row represents the station’s short code, the second row indicates the measured peak ground acceleration (PGA) in gal, and the third row denotes the epicentral distance in km.

Recorded vertical channel accelerograms at different stations during the Pithoragarh earthquake on 8 February 2020.

Subsequently, on 11 September 2021, a magnitude M w 4.7 earthquake struck in the Chamoli region. This event marked the first earthquake following the launch of the mobile app and the inauguration of the system to the public. Since it was a minor earthquake, alert messages were disseminated to the public. This event represents the first incident during which UEEWS successfully issued notifications to the public for the first time.

The UEEWS has produced reports for 19 earthquakes from 2015 to the present, as indicated in Table A1 . Among these, 14 earthquakes were triggered in the Uttarakhand region, while 5 occurred in the Nepal region. In line with the UEEWS objectives, our analysis focuses solely on earthquakes triggered in the Uttarakhand region. The comparison of earthquake information (magnitude, location) estimated by the UEEWS and NCS is shown in Figure 13 , revealing significant variation in the epicenter and depth. The location estimation section describes the procedure to estimate the epicenter and depth. The variations in hypocenter determination by the UEEWS and NCS were influenced in part by differing velocity models and phase-picking methods. Additionally, the UEEWS employs cost-effective MEMS-based accelerometers, whereas NCS utilizes high-end state-of-the-art broadband seismographs and strong motion accelerographs. Furthermore, the UEEWS relies on the real-time analysis of the initial segment of seismic records, while NCS utilizes the complete earthquake waveform and provides source parameters in ~5–10 min after the earthquake, with an average time of about 8.0 min [ 115 ].

Comparison between earthquake information, including ( a ) epicenters, ( b ) depths, and ( c ) response times estimated by the UEEWS server and published on the NCS website. In ( c ), the response time is estimated based on the first reports. The dashed line in ( c ) represents the average response time i.e., 13.1 s, excluding the maximum response time of one report, which was 22.38 s.

The average disparities in the estimated depth and epicenter between the UEEWS and NCS were 11.1 ± 13.6 km and 24.5 ± 22.6 km, respectively. The response time of the system is calculated as the duration between the origin time and the moment when the initial report is generated on the server. The average response time based on the reports generated during fourteen earthquakes ( Table A1 ) was estimated as 13.6 ± 3.8 s. In one event (23 May 2021), the average response time was 22.38 s and after excluding this event, the average response time came down to 12.8 s.

Magnitude estimates are derived from P-wave data, leading to significant uncertainties in magnitude determination. This discrepancy arose from the use of various regression equations (Equations (2)–(4)) at different times and their subsequent updates. The estimated M Pd from the UEEWS server and magnitude ( M w ) from the NCS catalog can be plotted in a 1:1 relationship ( Figure 14 ). In UEEWS, the magnitude estimation relies solely upon the initial portion of the P-wave following the P-onset; therefore, the estimated magnitude exhibits variation compared with the magnitude based on the moment estimated by NCS ( Figure 14 a). The standard deviations between the estimated magnitudes of the initial and final reports from the UEEWS and the magnitudes in the NCS catalog were 0.76 and 0.58, respectively. The discrepancy in the determined earthquake source parameters fluctuates with the earthquake reports generated by the server and gets improved as more triggered sensors are detected. The earthquake records given in Table A2 , underwent re-analysis on the offline server, utilizing the model outlined in Equation (4). It resulted in notable enhancements in the estimated magnitudes. Consequently, the standard deviation of the first and final reports compared with the NCS catalog magnitude improved to 0.47 and 0.30, respectively ( Figure 14 b).

Comparison of earthquakes’ magnitude estimated by ( a ) the UEEWS in real time and NCS, and ( b ) re-running the recorded data in offline mode and NCS. The filled circles indicate the estimated magnitude in the final report, and the open circles depict the estimated magnitude in the first report.

7. Discussion

The developed UEEWS has successfully issued notification for light earthquakes triggered in the instrumented region of Uttarakhand. Since the launch of the mobile app, BhuDEV, no strong earthquake has triggered in the Uttarakhand region; hence, notifications were issued for only light earthquakes of magnitude below five. However, warnings were issued for three moderate and two strong earthquakes triggered in the Nepal region. Despite their epicenters being 100 to 200 km away from the instrumented region, seismic waves were recorded by the installed sensors in Uttarakhand. UEEWS server analyzed the received seismic data while PICK_EEW module identified P-onset. Subsequently, the TCPD module utilized the seismic data associated with these P-onsets to estimate the earthquake parameters and generated earthquake reports accordingly. Upon detection of a new P-onset from a different station, the TCPD module repeated the same process to estimate the earthquake parameters and subsequently issued new notifications and alerts based on the estimated magnitude.

However, the estimated parameters provided by UEEWS did not align closely with those estimated by other agencies such as NCS. This disparity arose from the fact that UEEWS bases its parameter estimations on primary waves recorded exclusively by MEMS-based sensors. In contrast, NCS utilizes seismograms recorded by the state-of-the-art broadband seismographs, which offer a high dynamic range for more accurate estimation of earthquake parameters. The low-cost MEMS-based sensors offer cost-effectiveness but have limited capacity to accurately estimate complex earthquake parameters. Nevertheless, these sensors effectively serve the intended purpose of a regional EEWS.

The primary objective of the UEEWS project is to provide timely warnings to the public before the arrival of damaging waves. This objective is being achieved with the current instrumentation. However, due to the constraints of MEMS-based sensors, the expected results for several other EEW parameters cannot be produced consistently. Therefore, these parameters are omitted, and the EEWS relies primarily on peak displacement as the key EEW parameter.

8. Conclusions

In conclusion, the UEEWS stands as one of the pioneering initiatives in seismic hazard mitigation efforts in the Himalayan region. Leveraging insights from global EEW systems, the UEEWS has been meticulously crafted to address the unique seismic challenges of the region. The work on developing EEWS for Uttarakhand was started in 2014 under a pilot project entitled “Development of Earthquake Early Warning System for North India”. Under this project, instrumentation in the Garhwal region was started and completed by mid-2017. The sensors were mounted on concrete platforms built on the ground floors of the BSNL and SWAN buildings. These buildings were selected due to their affiliation with telecommunication services, provision of electricity, and assurance of security for the installed sensor assemblies. A server has been set up in the EEWS Laboratory at CoEDMM, IIT Roorkee, along with other supporting communication devices. The network is maintained as private for security purposes. Seismic data is continuously streamed to the central server via dedicated lease lines in real time, having latency in milliseconds only. Different modules are executed on the open-source platform, Earthworm. After the successful development of the pilot project, the complete system was taken over by the Government of Uttarakhand and its instrumentation was extended into the Kumaon region; thus, a total of 170 sensors have been deployed across Uttarakhand under UEEWS. Despite the rugged terrain of the Himalayas posing challenges to internet connectivity for streaming ground motion data, the system remains under constant surveillance to ensure its operational integrity.

This system is currently full-fledged operational, issuing warnings to the public through blowing installed sirens and sending warnings on the mobile app installed in the smartphones by the users. This mobile app provides an SOS button useful for users during earthquakes. Upon pressing this button, the user’s location is transmitted to the disaster management authority and two registered relatives. The mobile numbers of relatives are entered during the mobile app registration process. Upon receiving the location information of the earthquake victim, the disaster management authority initiates search and rescue operations swiftly. The developed system serves not only to issue warnings to the public but also to establish a robust ground motion database, proving invaluable for earthquake and civil engineering applications [ 116 ]. This database can be utilized for analyzing seismic hazards in the region and formulating new ground motion prediction equations [ 117 , 118 , 119 , 120 ]. The developed system offers lead times ranging from seconds to tens of seconds to urban areas, towns, and rural villages within the state. One significant outcome of this initiative is the increased public awareness regarding natural hazards, especially earthquakes, as evidenced by the significant number of downloads of the mobile app. As a testament to its success, the system continues to evolve, striving for greater effectiveness and resilience in mitigating seismic risks and safeguarding communities in Uttarakhand.

9. Future Outlook

The present UEEWS relies on the VPNoBB service provided by BSNL and the SWAN network. In addition to this, communication may face disruptions due to various factors such as extreme weather conditions, landslides, damage to optical fiber cables, power outages, interruptions in data streaming, and so forth. Therefore, exploring alternative options such as public networks, cloud-based services, and solar power backups could prove advantageous.

• Currently, the warning system does not provide information about the intensity of the earthquake at the user’s location. This feature could be incorporated once the prediction of strong ground motion and its conversion to intensity is integrated into the algorithm.
• At present, warnings are issued based on peak displacement ( P d ) of the first three seconds of P-wave data after P-onset from at least four sensors. However, there are various other attributes such as thepredominant period ( τ i p ), characteristic period ( τ c ), cumulative absolute velocity ( CAV ), squared velocity integral ( I V 2 ), log averaged period ( τ l o g ), root sum square cumulative velocity ( RSSCV ), etc., may be explored in the future.
• Due to the intricate nature of Himalayan tectonics, it is recommended to deploy a dense network featuring wider aperture arrays.
• The issuance of warnings should also be vetted for their societal and management implications.

Acknowledgments

The authors are thankful to USDMA, Dehradun, Uttarakhand for providing funds to run this project and the Ministry of Earth Sciences for funding the pilot project. The discussions held with NDMA, Uttarakhand administration, and NCS officials are gratefully acknowledged. The authors express gratitude to V P Dimri and H N Shrivastav for providing their expertise in identifying the region for instrumentation. The authors are also thankful to Ashok Kumar, Ajay Gairola, Himanshu Mittal, Bhanu Pratapa Chamoli, Bhavesh Pandey, and Govind Rathore for their support in establishing this system. The authors extend gratitude to the advisory committee, led by B K Gairola, for their invaluable suggestions and guidance. The help provided by the Centre of Excellence in Disaster Mitigation & Management and the Department of Earthquake Engineering, IIT Roorkee, is thankfully acknowledged.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/s24113272/s1 , Figure S1: Location of installed siren units in Uttarakhand; Figure S2: The icon and QR codes for the Bhukamp Disaster Early Vigilantè (BhuDEV) app. A few pages of the app are also shown.

List of earthquakes for which reports were generated by the UEEWS server.

* The same serial number multiple times indicates that the UEEWS generated more than one report for that earthquake.

Information about the earthquakes is from the National Center for Seismology (NCS), Ministry of Earth Sciences, Government of India, which determines and reports source parameters of the earthquakes in this region.

Funding Statement

This project “Earthquake Early Warning System for Uttarakhand” is being funded by USDMA, Dehradun, Government of Uttarakhand under the grant number USD-1077-DMC.

Author Contributions

Conceptualization, K. and P.K.; Methodology, K. and P.K.; Software, P.K.; Validation, P.K.; Formal analysis, M.L.S.; Investigation, P.K.; Data curation, P.K.; Writing—original draft, P.K.; Writing—review & editing, P.K., K., M.L.S., R.S.J. and P.; Supervision, K. and M.L.S.; Project administration, K.; Funding acquisition, K. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Conflicts of interest.

The authors acknowledge no conflicts of interest recorded and declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Detailed Report-Study of Causes & Impacts of the Uttarakhand Disaster on 7th Feb 2021

• Wadia Institute of Himalayan Geology

Abstract and Figures

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations
• Dmitry Ye. Chelyshev
• Jessica Munch

• Dan H. Shugar

• M. J. Westoby
• EARTH SURF PROC LAND

• C V Sangewar
• PHOTONIRVACHAK-J IND
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

• Open access
• Published: 24 February 2022

Environmental and economic impact of cloudburst-triggered debris flows and flash floods in Uttarakhand Himalaya: a case study

• Vishwambhar Prasad Sati   ORCID: orcid.org/0000-0001-6423-3119 1 &
• Saurav Kumar 1  

Geoenvironmental Disasters volume  9 , Article number:  5 ( 2022 ) Cite this article

13k Accesses

15 Citations

1 Altmetric

Metrics details

This paper examines the environmental and economic impact of cloudburst-triggered debris flow and flash flood in four villages of Uttarkashi district, Uttarakhand Himalaya. On 18th July 2021 at 8:30 p.m., a cloudburst took place on the top of the Hari Maharaj Parvat, which triggered a huge debris flows and flash floods, affecting 143 households of four villages of downstream areas. Immediately after the cloudburst occurred, the authors visited four affected villages—Nirakot, Mando, Kankrari, and Siror. A structured questionnaire was constructed and questions were framed and asked from 143 heads of affected households on the impact of debris flows and flash floods on people’s life, settlements, cowsheds, bridges, trees, forests, and arable land in and around the villages. The volume of debris, boulders, pebbles, gravels, and mud was assessed. It was noticed that all four villages got lots of destructions in terms of loss of life—people and animals, and property damage—land, crops, and infrastructural facilities. This study shows that the location of the settlements along with the proximity of the streams, which are very violent during the monsoon season, has led to the high impact of debris flow on the affected villages. We suggest that the old inhabited areas, which are located in the risk zones, can be relocated and the new settlements can be constructed in safe places using suitability analyses.

Introduction

Cloudburst, a geo-hydrological hazard, refers to a sudden and heavy rainfall that takes place within a short span of time and a particular space (Sati 2013 ). The intensity of rainfall is often more than 100 mm/h (Das et al. 2006 ). The disruptive events, cloudbursts occur during the monsoon season in the Himalaya and trigger debris flows, flash floods, landslides, and mass movements (Fig.  1 ). Fragile landscape, rough and rugged terrain, and precipitous slope accentuate the magnitude of geo-hydrological hazards. Cloudburst-triggered debris flows, flash floods, landslides, and mass movements have become more intensive and frequent worldwide, mainly in the mountainous regions, causing large-scale destruction of people, land, and property (Houghton et al. 1996 ; Wang et al. 2014 ; Mayowa et al. 2015 ; Malla et al. 2020 ; Sim et al. 2022 ). Similarly, the Himalayan region is prone to the occurrences of cloudburst-triggered hazards, causing huge loss of life and property and degradation of forest and arable lands (Bohra et al. 2006 ; Allen et al. 2013 ; Balakrishnan 2015 ; Ruiz-Villanueva et al. 2017 ).

Cloudburst-triggered hazards in the Uttarakhand Himalaya

The Uttarakhand Himalaya, one of the integrated parts of the Himalaya, is the most fragile landscape and prone to geo-hydrological hazards—cloudbursts, avalanches, and glacier bursts (Sati 2019 ). It receives many hazards mainly cloudburst-triggered debris flows, flash floods, landslides, and mass movements during the monsoon season every year. The intensity, frequency, and severity of these hazards have been observed to increase during the recent past. Devi ( 2015 ) stated that the changing monsoon patterns and increasing precipitation in the Himalaya are associated with catastrophic natural hazards. However, these hazards are the least understood because of the remoteness of the areas and lacking meteorological stations (Thayyen et al. 2013 ).

The Uttarakhand Himalaya has many eco-sensitive zones, vulnerable to natural hazards mainly for geo-hydrological hazards. Every year, many cloudburst events occur here, cause to roadblocks, land degradation, forest and cropland loss, and losses of life and infrastructural facilities. One of the most devastating cloudburst-triggered debris flow events of this century occurred on the night of 16th and 17th June 2013 in the famous Hindu pilgrimage ‘Kedarnath’, which killed more than 10,000 people and devastated the entire Mandakini and Alaknanda river valleys (Upadhyay 2014 ; Sati 2013 ). The entire region had received 16 major geo-hydrological and terrestrial hazards within the last 50 years (Bhambri et al. 2016 ). Some of the devastating cloudburst-triggered debris flows and flash floods that occurred in the Uttarakhand Himalaya are Rudraprayag on 14th September 2012, Munsiyari on 18th August 2010, Kapkot on 19th August 2010, Nachni on 7th August 2009, Malpa and Ukhimath on 17th August 1998, Badrinath on 24th July 2004, and the Alaknanda River valley on 1970. About 20,000 people died and a huge loss of property took place due to these calamities (Das 2015 ). It has been noticed that these catastrophic events occurred mainly during the three months of the monsoon season—July, August, and September.

Debris flows and flash floods caused by glacier-bursts incidences were although not much frequent and intensive yet, during the recent past, their number has increased owing to changes in the climatic conditions. The increasing number of infrastructural facilities on the valley bottom has accelerated damages owing to exposed elements in risk-prone areas (Sati 2014 ; ICIMOD 2007a , b ; Chalise and Khanal 2001 ; Bhandari 1994 ; Uttarakhand 2017 ). Many drivers exist, which affect the severity of cloudburst-triggered hazards in the Uttarakhand Himalaya. Growing population and the construction of settlements and infrastructural facilities on the fragile slopes and along the river valleys have also caused severe hazards. The Uttarakhand region is home to world-famous pilgrimages and natural tourism. Mass tourism during the rainy season enhances the intensity of disasters.

Several studies have been carried out on glacier-bursts and cloudburst-triggered debris flows and flash floods in the Himalaya (Shugar et al. 2021 ; Byers et al. 2018 ; Cook et al. 2018 ; Asthana and Sah 2007 ; Bhatt 1998 ; Joshi and Maikhuri 1997 ; NIDM 2015 ; IMD 2013 ; Khanduri et al. 2018 ; Sati 2006 , 2007 , 2009 , 2011 , 2018a , b , 2020 ; Naithani et al. 2011 ). These studies were conducted from broader perspectives, mostly covering the entire Himalaya. However, the present paper looks into the case study of four villages of the Uttarakhand Himalaya, which were severely affected and damaged by cloudburst-triggered debris flows and flash floods, which occurred on July 18th, 2021. It analyses the environmental impact of cloudbursts in terms of forest and fruit trees dislocation, land degradation, and soil erosion—arable, forests, and barren land of the four affected villages. It also evaluates the human and economic losses like the killing of people, loss of existing crops, and damage of houses and cowsheds, respectively. The study suggests policy measures to risk reduction and rehabilitation of settlements from danger zones to safer areas after suitability analysis.

The Uttarakhand Himalaya is located in the north of India and south of the Himalaya. It is also called the Indian Central Himalayan Region. Out of the total 93% mountainous area, 16% is snow-capped, called the Greater Himalaya. The terrain is undulating and precipitous and the landscape is fragile, vulnerable to natural hazards. This catastrophic event occurred in the four villages of Uttarkashi district. The Uttarkashi town lies about 10 km downstream of the affected villages. A National Highway number 108, connecting Haridwar and Gangotri, is passing through Uttarkashi town. The four affected villages—Nirakot, Mando, Kankrari, and Siror are located in the upper Bhagirathi catchment, which is prone to geo-hydrological hazards. The slope gradient of these villages varies from 15° to 70°. Indravati is a perennial stream, a tributary of the Bhagirathi River that meets Bhagirathi from its left bank. All three Gadheras (streams)—Mando, Diya, and Siror are seasonal but violent during the monsoon season. Nirakot (1530 m) village is located in the middle altitude of the Hari Maharaj Parvat (2350 m) in a steep slope, Mando village (1180 m) is located on the left bank of the Bhagirathi River along the Mando Gadhera with gentle to a steep slope, Kankrari (1620 m) village is located on the moderate to the gentle slope on the bank of the Diya Gadhera, and Siror village (1280 m) is situated on the left bank of both Bhagirathi and Siror Gadhera with gentle to the steep slope (Fig.  2 ). One of the prominent eco-sensitive zones of the Uttarakhand Himalaya, the ‘Bhagirathi Eco-Sensitive Zone’ is 120 km long, spanning from Uttarkashi to Gaumukh, along the Bhagirathi River valley (Sati 2018a , b ). The rural people depend on the output of the traditional farming systems, often face intensive natural hazards. The settlements are located either on the fragile and steep slopes or on the banks of streams, which are very violent during the monsoon season when a heavy downpour occurs. Therefore, heavy losses of life and property in these areas are common, taking place every year.

Location map of cloudburst source and hit areas and their surroundings

Methodology

This study was empirically tested and a qualitative approach was employed to describe data. A structured questionnaire was constructed. The main questions framed and asked from the heads of households were—human and animal death, damage to self property—houses and cowsheds, and existing crops—cereals, fruits, and vegetables. Loss to public properties such as bridges, public institutions, and forest land was assessed. Based on the questions framed, we surveyed 143 heads of households of four villages, which were partially or fully affected due to cloudburst-triggered debris flow. These villages are Nirakot, Mando, Kankrari, and Siror. To assess the debris and the damaging areas, the authors travelled from the source areas to the depositional zones and measured the volume of debris—boulders, pebbles, sands, and soils using a formula; circumference = 2πR and area = π * R 2 . The slope gradient, accessibility, economic conditions, and climate of the villages were assessed and based on which, the susceptibility analysis of the villages was carried out. The villages were divided into very high susceptibility, high susceptibility, and moderate susceptibility levels. Both environmental degradation and economic losses in four villages were assessed. We used Geographical Positioning System (GPS) to obtain the data of altitude, longitude, and latitude. Two maps—case study villages and the major cloudburst incidences—2020 and 2021 were prepared and data were also presented using graphs. Photographs of four villages were used to present the destruction of villages due to the cloudburst event.

Results and analysis

Major cloudburst incidences in the uttarakhand himalaya.

Past incidences depict that the Uttarakhand Himalaya suffered tremendously due to cloudburst-triggered calamities. We gathered data on the major cloudburst incidences in Uttarakhand in the monsoon seasons of 2020 and 2021 from the state disaster relief force (SDRF), Dehradun. From May to September 2020, 13 major cloudburst incidences were noticed in Uttarakhand (Table 1 ). These incidences resulted in the death of 22 people and 77 animals, and 19 houses were fully damaged. Similarly, from May to September 2021, 17 major cloudburst incidences were occurred in the Uttarakhand Himalaya, resulting in the death of 34 people and 144 animals, and 106 houses were buried. Besides, it caused a huge loss to public property and landscape degradation.

The economic losses in 2021 were much higher than the losses in 2020 (Fig.  3 ). In 2021, the frequency and intensity of cloudburst-triggered calamities were also higher. The loss of animals was quite high both the years. Houses that collapsed due to calamity were six times higher in 2021 than in 2020. The loss of human life was substantial in both years. Several bridges were washed away.

Loss of human lives, livestock, houses and bridges due to cloudburst in Uttarakhand during the 2020 and 2021

District-wise major cloudburst events of 2020–2021 are shown in the map of the Uttarakhand Himalaya (Fig.  4 ). A total of 30 major cloudburst incidences were recorded, out of which, 17 occurred in 2021. The Uttarkashi district received the highest incidences (07), followed by the Chamoli district (05). Dehradun and Pithoragarh districts have recorded 04 incidences each. Rudraprayag 03 and Tehri, Almora, Bageshwar have recorded 01 each. It has been observed that cloudburst-triggered incidences mainly occurred in remote places along the fragile river valleys and middle slopes.

Location map of cloudbursts hit areas in 2020 and 2021

Case study of affected villages

On July 18, 2021, a cloudburst hits the Hari Maharaj Parvat (hilltop) at an altitude of 2350 m at 8:30 p.m., which triggered huge debris flows and flash floods. The four villages—Nirakot, Mando, Kankrari, and Siror of Uttarkashi district, located down slopes of the hilltop and close to the Uttarkashi town, were severely affected due to debris flow (Table 2 ). At the cloudburst hit area, it formed three gullies, which later on merged into three streams, along which these villages are located. Debris, from the source i.e. hilltop of Hari Maharaj Parvat, equally flew in three directions. Since the cloudburst event occurred at 8:30 p.m., the people did not have time to move with their movable property and therefore, the magnitude of damage was enormous.

The villages are located from the altitudes of 1180 m (lowest) to 1620 m (highest). Mando village is located at 1180 m, Kankrari village at 1620 m, Nirakot at 1530 m, and Siror has 1280 m altitude. The two villages—Nirakot and Mando have west-facing slopes, Kankrari has a south-facing slope, and Siror has a north-facing slope. These villages are located along the tributaries of the Bhagirathi River, with 2 to 5 km distance from the road. The intensity and volume of debris were different in different villages, therefore, the casualties and losses were also varied. The villages are surrounded by agricultural and forestlands. The farmers mainly grow subsistence cereal crops—paddy, wheat, pulses, oilseeds, fruits, and vegetables. Forest types comprise pine (sub-tropical) and oak and coniferous forests (temperate), used for fodder, firewood, and wild fruits.

Located at the high-risk zones, these villages face several disaster incidences every year. Out of the total 143 heads of households surveyed, more than 80% of heads were in favour of rehabilitating them in the safer areas. They wanted to relocate their houses and cowshed within the village territory with financial assistance from the state government. The streams, along which the settlements are constructed, are fragile and highly vulnerable to landslide hazards. Further, the cloudburst incidences are increasing due to climate change, the heads of households perceived.

Figure  5 shows four villages—Nirakot, Mando, Kankrari, and Siror, which were severely affected by cloudburst-triggered debris flow and flash flood. The volume of debris and boulders can be seen in all the villages. These villages are surrounded by dense sub-tropical and temperate forests that vary from pine to mixed-oak and deodar. Kharif crops were growing in the arable land whereas a large cropland has been washed away.

Cloudburst affected villages a Nirakot, b Mando, c Kankrari, d Siror; Photo: by authors

Impact of cloudburst-triggered debris flow and flash flood

Environmental impact.

The environmental impact of cloudburst-triggered debris flow and flash flood in four villages of Uttarkashi district was analyzed (Table 3 ). The major variables were the number of forest trees dislocated, total land degradation, land degradation under existing crops, number of fruit trees dislocated, land degradation under arable land, number of buildings were damaged, number of bridges damaged, and boulders’ volume. Forest trees, which dislocated were pine in the middle altitude and mixed-oak and deodar in the higher altitude. A total of 770 forest trees were dislocated from all four villages, out of which, 500 were from the Kankrari village (highest). The lowest trees dislocated were from Siror village (70). The total land degradation from the cloudburst hit areas to the affected areas was huge, however, we have measured the land which was within and surrounding each village. The total land degradation was 52.5 acres with the highest in Kankrari (45 acres) and the lowest in Siror (0.5 acres). The land degradation under existing crops was 22.6 acres in all four villages, varying from 0.1 acres in Siror to 20.6 acres in Kankrari. The total number of fruit trees dislocated was 486. Land degradation under arable land was 22.6 acres. It includes the area under existing crops both agriculture and horticulture. A total of 19 buildings were damaged whereas a total of 14 bridges, connecting the affected villages were washed away.

Economic impact

The economic impact due to cloudburst calamity was tremendous in the forms of a household affected, loss of human and animal life, building loss, forest loss, loss of existing crops including fruits, loss of arable land, and loss of bridges (Table 4 ). The value of all these assets was calculated in Indian Rupees (INR) at the current price. The total number of households affected was 143, of which, 100 households belonged to the Kankrari village (highest) and three households (lowest) were from Siror village. Four people died due to the calamity—three women from Mando village and 1 man from Kankrari village. Two cows from Mando village died. The total loss from the collapse of the building was 1.7 million INR, with the highest (1.1 million INR) from Kankrari village. A total of 0.77 million INR was lost due to forest loss, and the loss from existing crops was 3.35 million INR. Loss from dislocation of fruit trees was noted high, which was about 0.5 million INR. A large portion of arable land was flown which value was 11.3 million INR. About 14 million INR was lost due to the collapse of bridges. As a whole, about 31.62 million INR was lost due to cloudburst calamity. Per household loss by the cloudburst calamity was noted 0.22 million INR.

Average circumference, area, and volume of boulders

We calculated the average circumference, area, and volume of boulders in the case study villages using a formula: circumference = 2πR; Area = π * R 2 ; volume = length × width × depth (Table 5 ). We noticed that the highest average area of boulders was in Mando village, which is 28.3 m 2 followed by Kankrari 19.6 m 2 , Nirakot 12.57 m 2 , and Siror 7.1 m 2 . In terms of the total volume of debris, it was the highest in Kankrari village, followed by Mando, Nirakot, and Siror villages.

Figure  6 shows the average diameter of boulders in the cloudburst-affected villages. We drew the figure with a scale of 1 cm is equal to 1 m. The average biggest diameter of boulders was found in Mando village (6 m), followed by Kankrari (5 m) and Nirakot (4 m) villages. The average smallest diameter of boulders was found in Siror village (3 m).

Village-wise average diameter of boulders

Susceptibility analysis

Based on the above description, susceptibility analysis of the case study villages was carried out (Table 6 ). The main variables of susceptibility were slope gradient, accessibility of villages, economic conditions of households, and climatic conditions. We noticed that Nirakot village has very high susceptibility, Kankrari has high, and Siror and Mando have moderate susceptibility.

The Uttarakhand Himalaya is highly vulnerable to geo-hydrological disasters because of its geological formation (Vaidya 2019 ). It is an ecologically fragile, geologically sensitive, and tectonically and seismically very active mountain range (Sati 2019 ). The geo-hydrological events—cloudbursts and glacier bursts-triggered catastrophes are very common and devastating. The monsoon season poses severe threats to natural hazards because of heavy downpours. About 93% of the Uttarakhand Himalaya is mountainous mainland, of which 16% is snow-capped. The undulating and precipitous terrain and remoteness are the most vulnerable for disaster risks.

This study reveals that most of the cloudbursts incidences in 2020–21 occurred mainly in the remote mountainous districts of the Uttarakhand Himalaya. The villages in the Uttarakhand Himalaya are located on the sloppy land and along the river valleys, which are fragile and very vulnerable to disasters. The rivers flow above danger marks during the monsoon season cause threats to rural settlements. The roads of Uttarakhand are constructed along the river banks and on fragile lands. These roads lead to the highland and river valley pilgrimages where the number of tourists and pilgrims visit every year mainly during the monsoon season. There are many locations along the river valleys where the houses are constructed on the debris, deposited by rivers during debris flow events. Therefore, the environmental and economic losses due to debris flows and flash floods are high. The construction of hydropower projects along the river valleys without using sufficient technology further accentuates the vulnerability of debris flows and flash floods. One of the recent examples is the Rishi Ganga tragedy in Chamoli district where more than 200 people died with a huge loss to property (Sati 2021 ). We observed that the cloudburst triggered calamity in 2021 was higher than in 2020. The trend of occurring natural hazards has been increasing. Similarly, the intensity and frequency of natural hazards were observed high.

The present study shows that the environmental and economic loss in the four villages of the Bhagirathi River valley was huge due to cloudburst-triggered debris flows and flash floods. Almost every household of the villages were affected by cloudburst calamity. There were large forest and arable land degradation, forest and fruit trees were dislocated, loss of life—human and animal, and the houses and bridges were collapsed. The calamity also poses threat to the future, in terms of, the large deposition of debris including boulders, pebbles, and gravels in the villages along the streams and gullies. The rural people are poor and their livelihood is dependent on practicing subsistence agriculture. Many of them are living below the poverty line in these villages. Because the existing crops have been lost, they are facing food insecurity. Further, the psychological problems are immense. The fear of another calamity is always there in the mind of people as all villages are situated in very high to moderate susceptible areas. The national highway is passing through the right bank of the Bhagirathi River and the affected villages are situated on the left bank. The connectivity problem is immense all the time in these villages. The entire rural areas of the Uttarakhand Himalaya are facing similar problems.

Cloudburst-triggered debris flows and flash floods are natural calamities in the Himalayan regions. They occur naturally and cannot be stopped. The losses—environmental and economic are also huge. However, the severity of these natural calamities can be minimized. For example, the high impact of cloudburst-triggered debris flow on the four study villages was mainly due to their location along the streams and on the fragile slopes. This can be avoided by constructing the settlements in safer places generally away from the violent streams. In the disaster risk zones, scenario analysis can be carried out under which, identifying driving forces of disaster risks is the first step. Then, the critical uncertainties are to be identified, and finally, a possible scenario can be developed. Nature-based eco-disaster risk reduction can be adopted to prevent further disaster risks. A large-scale plantation drive in the degraded land will restore the fragile landscape. Both pre and post-disaster risk reduction measures can be adopted to reduce the economic and environmental impact of debris flows. There must be policies implementation programmes for providing immediate relief packages for the affected people in terms of food and shelters. In a long run, susceptibility analyses should be carried out to understand the risk to the settlements so that the settlements can be replaced on the safer side if needed. A special budget can be allocated to hazard-prone villages during adverse situations.

Allen SK, Rastner P, Arora M, Huggel C, Stoffel M (2013) Lake outburst and debris flow disaster at Kedarnath, June 2013: hydrometeorological triggering and topographic predisposition. Landslides 13(6):1479–1491

Article   Google Scholar  

Asthana AKL, Sah MP (2007) Landslides and cloudbursts in the Mandakini Basin of Uttarakhand Himalaya. Himal Geol 28:59–67

Google Scholar  

Balakrishnan S (2015) Chennai flood of 1–5 December 2015: an extreme climatic event? Curr Sci 110(1):9

Bhambri R, Mehta M, Dobhal DP, Gupta AK, Pratap B, Kesarwani K, Verma A (2016) Devastation in the Kedarnath (Mandakini) Valley, Uttarakhand Himalaya, during 16–17 June 2013: a remote sensing and ground-based assessment. Nat Hazards 80:1801–1822

Bhandari RK (1994) Landslide hazard mapping in Sri Lanka—a holistic approach. In: Proceeding of national symposium on landslides in Sri Lanka, pp 271–284

Bhatt O (1998) Living in the shadows of death: landslides in Uttrakhand. The calamity-prone Central Himalayas, Lok Soochana Evam Sahayata Kendra, Gopeshwar, Chamoli, Uttarakhand, India, pp 1–4

Bohra AK, Basu S, Rajagopal EN, Iyengar GR, Gupta MD, Ashrit R, Athiyaman B (2006) Heavy rainfall episode over Mumbai on 26 July 2005: assessment of NWP guidance. Curr Sci 90(9):1188

Byers AC, Rounce DR, Shugar DH, Lala JM, Byers EA, Regmi D (2018) A rock fall-induced glacial lake outburst flood, Upper Barun Valley, Nepal. Landslides 16:533–549. https://doi.org/10.1007/s10346-018-1079-9

Chalise SR, Khanal NR (2001) An introduction to climate, hydrology and landslide hazards in the Hindu Kush-Himalayan Region. In: Tianchi L, Chalise SR, Upreti BN (eds) Landslide hazard mitigation in the Hindu Kush-Himalayas. ICIMOD, Kathmandu, pp 51–62

Cook KL, Andermann C, Gimbert F, Adhikari BR, Hovius N (2018) Glacier lake outburst floods as drivers of fluvial erosion in the Himalaya. Science 362:53–57

Article   CAS   Google Scholar  

Das PK (2015) Global warming, glacial lakes and cloud burst events in Uttarakhand–Kumaon Himalaya: a hypothetical analysis. Int J Environ Sci 5(4):697–708

Das S, Ashrit R, Moncrieff MW (2006) Simulation of a Himalayan cloudburst event. J Earth Syst Sci 115(3):299–313. https://doi.org/10.1007/BF02702044

Devi R (2015) Spatio temporal occurrences of cloud burst In the Himachal Himalaya. Int J Res Soc Sci 5(1):886–894

Houghton JT et al (1996) The IPCC second assessment report. In: Houghton JT, Meira Filho LG, Callander BA, Harris N, Kattenberg A, Maskell K (eds) Climate change. Cambridge University Press, New York, p 572

ICIMOD (2007a) Inventory of glaciers, glacial lakes and identification of potential glacial lake outburst flood (GLOFs) affected by global warming in the mountains of the Himalayan Region (DVD ROM). ICIMOD, Kathmandu

ICIMOD (2007b) Flash flood hotspot mapping in the Hindu Kush-Himalayan region (draft DVD ROM). ICIMOD, Kathmandu

IMD (2013) A preliminary report on heavy rainfall over Uttarakhand during 16–18 June, 2013. India Meterological Department, Ministry of Earth Sciences, New Delhi

Joshi V, Maikhuri RK (1997) Cloudburst: a natural calamity—a case study from Uttarakhand Himalaya, UP. J Indian Build Congr 4:207–217

Khanduri S, Sajwan KS, Rawat A, Dhyani C, Kapoor S (2018) Disaster in Rudraprayag District of Uttarakhand Himalaya: a special emphasis on geomorphic changes and slope instability. J Geogr Nat Disasters 8:1–10

Malla SB, Dahal RK, Hasegawa S (2020) Analyzing the disaster response competency of the local government official and the elected representative in Nepal. Geoenviron Disasters 7:15. https://doi.org/10.1186/s40677-020-00153-z

Mayowa O, Pour SH, Mohsenipour S, Harun M, Heryansyah SB, Ismail HT (2015) Trends in rainfall and rainfall-related extremes in the east coast of peninsular Malaysia. J Earth Sys Sci 124(8):1609–1622

Naithani AK, Rawat GS, Nawani PC (2011) Investigation of landslide events on 12th July 2007 due to cloudburst in Chamoli District, Uttarakhand, India. Int J Earth Sci Eng 4:777–786

NIDM (2015) Uttarakhnd disaster 2013. National Institute of Disaster Management Ministry of Home Affairs, Government of India, New Delhi

Ruiz-Villanueva V, Allen S, Arora M, Goel NK, Toffel M (2017) Recent catastrophic landslide lake outburst floods in the Himalayan mountain range. Prog Phys Geogr 41(1):3–28

Sati VP (2006) Natural hazards in an ecologically mountain terrain: a case for the Pindar Basin of Uttarancal Himalaya. ENVIS Himal Ecol 14(1):22–30

Sati VP (2007) Environmental impacts of debris flows—a case study of the two debris-flow zones in the Uttarakhand Himalaya. In: Chen-lung C, Major JJ (eds) Debris-flow hazards mitigation: mechanics, prediction, and assessment. Science Publishers, Rotterdam, pp 715–723

Sati VP (2009) Atmospheric and terrestrial natural calamities in the Himalaya: several devastating incidences overview. In: Proceedings of international conference on ‘mitigation of natural hazards in mountain areas’ Bishkek, SALAM, Kyrgyzstan. http://www.geomin.cz/conference/menu/Sati_paper_eng.pdf . (in Russian)

Sati VP (2011) Climate disasters in the himalaya: risk and vulnerability. In: International conference on climate change and natural hazards in mountain areas. Dushanbe, Sept. 19–21, 2011. http://www.mountainhazards2011.com/souboryeditor/Paper%20Vishwambhar%20P.%20Sati.pdf

Sati VP (2013) Extreme weather-related disasters: a case study of two flashfloods hit areas of Badrinath and Kedarnath valleys, Uttarakhand Himalaya, India. J Earth Sci Eng 3:562–568

Sati VP (2014) Landscape vulnerability and rehabilitation issues: a study of hydropower projects in the Uttarakhand region, Himalaya. Nat Hazards 75(3):2265–2278. https://doi.org/10.1007/s11069-014-1430-y

Sati VP (2018a) Cloudburst triggered natural hazards in Uttarakhand Himalaya: mechanism, prevention and mitigation. Int J Geol Environ Eng 12(1):45–38

Sati VP (2018b) Carrying capacity analysis and destination development: a case study of gangotri tourists/pilgrims’ circuit in the Himalaya. Asia Pac J Tour Res APJTR 23(3):312–322. https://doi.org/10.1080/10941665.2018.1433220

Sati VP (2019) Himalaya on the threshold of change, advances in global change research series, scopus indexed. Springer International Publishers, Bern

Sati VP (2020) Increasing incidences of cloudburst triggered debris-flows/flash floods in Uttarakhand Himalaya, India. Int J Interdiscip Res Innov 8(2):8–16

Sati VP (2021) Are hydropower projects sustainable in Uttarakhand? Himachal Star, Feb 10, 2021, P 2, Dehradun; Delhi Post, March 01, 2021. https://delhipostnews.com/are-hydropower-projects-sustainable-in-uttarakhand/

Shugar DH, Jacquemart M, Shean D et al (2021) A massive rock and ice avalanche caused the 2021 disaster at Chamoli, Indian Himalaya. Science 373:300–306

Sim KB, Lee ML, Wong SY (2022) A review of landslide acceptable risk and tolerable risk. Geoenviron Disasters 9:3. https://doi.org/10.1186/s40677-022-00205-6

Thayyen RJ, Dimri AP, Kumar P, Agnihotri G (2013) Study of cloudburst and flash floods around Leh, India, during August 4–6, 2010. Nat Hazards 65(3):2175–2204

Upadhyay K (2014) A year later, no lessons learnt. The Hindu (available at 498. https://www.thehindu.com/opinion/op-ed/a-year-later-no-lessons499learnt/article6120397.ece ). 500

Uttarakhand (2017) Retreat from http://www.q8india.com/blog/2017/08/15/uttarakhandcloudburst

Vaidya RA (2019) Disaster risk reduction and building resilience in the Hindu Kush Himalaya. In: Wester P, Mishra A, Mukherji A, Shrestha A (eds) The Hindu Kush Himalaya assessment. Springer, Cham. https://doi.org/10.1007/978-3-319-92288-1_11

Chapter   Google Scholar  

Wang XJ, Zhang JY, Shahid S, Guan EH, Wu YX, Gao J (2014) Adaptation to climate change impacts on water demand. Mitig Adapt Strat Glob Change. https://doi.org/10.1007/s11027-014-9571-6

Download references

Author information

Authors and affiliations.

Department of Geography and Resource Management, Mizoram University (A Central University), Aizawl, 796004, India

Vishwambhar Prasad Sati & Saurav Kumar

You can also search for this author in PubMed   Google Scholar

Contributions

The first author analyzed data, prepared maps, and diagrams, and wrote the manuscript. The second author conducted a field survey and collected data. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Vishwambhar Prasad Sati .

Ethics declarations

Ethics approval.

This article does not contain any studies with human participants performed by any of the authors.

Competing interests

There is no competing interests in the manuscript.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Sati, V.P., Kumar, S. Environmental and economic impact of cloudburst-triggered debris flows and flash floods in Uttarakhand Himalaya: a case study. Geoenviron Disasters 9 , 5 (2022). https://doi.org/10.1186/s40677-022-00208-3

Download citation

Received : 20 September 2021

Accepted : 15 February 2022

Published : 24 February 2022

DOI : https://doi.org/10.1186/s40677-022-00208-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Debris flow
• Flash flood

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

Earthquake-safe Koti Banal architecture of Uttarakhand, India

Related Papers

Paritosh Yadav

Smriti Saraswat , Gautam Mayuresh

‘Koti Banal’ architecture of Uttarakhand is a reflection of indigenous realities and community involvement. It demonstrates a profound knowledge of local materials and native sensibilities. Investigations suggest that this is an earthquake-safe construction style done in timber and stone, which evolved as early as 1000 years ago. This paper is an attempt to study the Koti Banal architecture of Uttarakhand and understand the craft nurtured by the indigenous communities using locally available materials in response to earthquakes. In fact, the Koti Banal architecture is much like the framed construction of modern times. The structural design suggests that these buildings responded well to the forces likely to act upon them during an earthquake. The paper further investigates what are the modifications that have happened in this style of architecture with respect to morphology and materials, through three case studies done in the Garhwal region. This is primarily a descriptive research based on a case study (field study) approach, which focuses on traditional knowledge systems; indigenous building materials; community involvement; and, craft skills of Uttarakhand. Keywords Koti Banal Architecture; Uttarakhand; Indigenous; Craft; Materials; Community; India

WIT Transactions on Ecology and the Environment

Bamidele I O Dahunsi

Subhankar Nag

Vinod S Shah , V. R. Shah

International Journal of Disaster Risk Reduction

Bashir Ahmad

1.1 ABSTRACT Buildings in the old walled city of Ahmedabad are placed in dense street formation and have a specific manner of construction using brick and wood. Structural systems could be studied for earthquake resistance. These structures have survived several natural calamities that have extensively damaged many buildings in the new city. If indigenous techniques of building are not studied vital information about how to build here would be lost. Idealized models of systemic behaviour (e.g. the Indian standards code) only partly explain the traditional construction of Ahmedabad. These models are based on scientific observations of contemporary materials and systems but fail in correctly assessing the vernacular. New models with alternative theoretical interpretations and 3D simulation techniques for analysis, may clarify the role of different materials and their construction systems within a structure. The quest of this study is to find out whether these traditional structural systems provide a greater degree of resistance to earthquakes. 1.2 INTRODUCTION: Vernacular Architecture in walled city of Ahmedabad is an example of a typical construction, which is non-engineered construction and has evolved over a period of time in response to cultural, physical and environmental conditions. Any non-engineered construction by definition is one, which occurs without any technical assistance. This construction is generally a result of traditions improved over a period of time. Often this kind of construction fails to respond to the natural calamities like earthquake, because of higher return periods of such calamities. In past earthquakes, it is a known experience of people in many parts of the world that vernacular structures have often shown better seismic resistance against the modern construction. Effort in this paper is to analyze and discuss the seismic resistance of Ahmedabad pole houses, which have survived in past earthquakes so far. Modern buildings are designed for seismic resistance using the building design codes. The codes are based on the seismic zone of region, the seismic force is considered as a mainly lateral load, and structure is assumed elastic. For non-engineered construction it is very difficult to model and analyze the structure as per the available codes, as these structures use materials and methods of construction that are difficult to analyze for behaviour, as properties of material used are not known. Furthermore, their behaviour may not be elastic because the nature of construction is composite. Earlier codes were silent about the effect of configuration on the behavior of structures in earthquake, but recent codes have recognized the importance of configuration of structures under lateral loads. The paper presents the study of the configuration of the Ahmedabad pole house and its comparison with codal provisions.

Sajjad Gulzar

Vinod S Shah , riyaz tayyibji , V. R. Shah

ISVS e-journal (Scopus Indexed)

Dr. Mohammad Arif Kamal , Anuska Mahalik

The old settlement of Srinagar, Jammu & Kashmir, India, is an open museum for vernacular buildings and has become an integral part of the local cultural heritage. The region is the heart of the entire urban area. It is also often the city's oldest part, constantly shaped with time. The old settlement of Srinagar has always been an important geographical site, despite its varied physical and historical characteristics. Srinagar symbolizes the city's historical heritage as it is located in the heart of the city. Moreover, they are a manifestation of architectural systems optimized over time for the climate and the threat of natural disasters like earthquakes. This research examines the vernacular architecture of the old settlement in terms of climatic and seismic responsive designs. Its aim is to understand the vernacular architecture of the old settlement, Srinagar. Its objectives are to unravel the vernacular ideas used by the people to make the structures seismic and climateresponsive according to the climate and topography. It employs a documentary survey and observations as research methods. The research findings show that traditional architecture proves to be climate responsive and energy conscious due to the inclusion of certain principles and resources. Therefore, the vernacular construction systems like Taq and Dhajji-Dewari; used to resist the ground movement must be utilized in today's construction. The research concludes with the need of preserving lost vernacular architecture.

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

RELATED PAPERS

Shulanki Pal

Vasudha Gokhale

IJSTE - International Journal of Science Technology & Engineering | Volume 6 | Issue 7 | January 2020 ISSN (online): 2349-784X

Ravi Bhushan Bhardwaj

Rishiraj Das

Paarija Saxena

Medha Gokhale

Frontiers in Built Environment

raffaele carabbio

Krishna G. Parajuli

Journal of Traditional Building, Architecture and Urbanism

Neelima Yadav

Journal of Civil Engineering and Environmental Technology ; pp. 320-324, p-ISSN: 2349-8404; e-ISSN: 2349-879X; Volume 5, Issue 6.

Janmejoy Gupta

SCITECH Nepal

Ram P R A S A D Suwal , Ramprasad Suwal

Indo Nordic Author's Collective

Dr. Uday Dokras

Ar. Tania Bera

Earthen Heritage, New Technology Management, Turkey

Adnan Anwar , Aurangzeb Khan

IRJET Journal

Rajeev Parashar

APS University Rewa

Aniruddh Bahuguna

Masters Thesis

Maanana Atrey

Smart Innovation, Systems and Technologies

Ravi Mokashi

IAEME PUBLICATION

IAEME Publication

Selcuk University, the department of Architecture

Dicle Aydin

Muskan Tiwari

sandra tonna , Claudio Chesi

RELATED TOPICS

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

• Corpus ID: 132023966

EARTHQUAKE DISASTERS IN HILLY AREAS (CASE STUDY UTTARAKHAND) –Part II

• S. Gupta , A. Jalali
• Published 9 June 2016
• Environmental Science, Geology, Engineering
• Paripex Indian Journal Of Research

2 Citations

A study on disaster detection and management using wsn in himalayan region of uttarakhand, multi-hazard disaster studies: monitoring, detection, recovery, and management, based on emerging technologies and optimal techniques, related papers.

Showing 1 through 3 of 0 Related Papers

Code-based brittle capacity models for seismic assessment of pre-code RC buildings: comparison and consequences on retrofit

• Original Article
• Open access
• Published: 16 September 2024

Cite this article

You have full access to this open access article

• Santa Anna Scala 1 ,
• Maria Teresa De Risi   ORCID: orcid.org/0000-0002-8882-6127 1 &
• Gerardo Mario Verderame 1  

The existing Reinforced Concrete (RC) buildings stock is often characterized by a significant seismic vulnerability, due to the absence of capacity design principles, even in regions with high seismic hazard, such as Italy. Approximately 67% of existing RC buildings in Italy have been designed without considering seismic actions (GLD), resulting in very low transverse reinforcement amount in beams and, particularly, in columns. Additionally, beam-column joints typically totally lack stirrups. Consequently, shear failures under seismic actions are very likely for this pre-code building typology, often limiting their seismic capacity. However, the assessment of shear failures in beams/columns or joints varies significantly from code to code worldwide. The main goal of this work is to quantify the impact of different code-based brittle capacity models on the seismic capacity assessment and retrofit, focusing on GLD Italian pre-1970 RC buildings. This comparative analysis is carried out by first considering three current codes, emphasizing their, even significant, differences: European (EN 1998-3-1. 2005), Italian (D.M. 2018), and American (ASCE SEI/41 2017) standards. Then, shear capacity models prescribed by the current drafts of the next generation of Eurocodes are implemented and compared to the current models. The assessment includes: ( i ) a parametric comparison among models; ( ii ) the evaluation of case-study buildings capacity in their as-built condition and after shear strengthening interventions. The latter is performed on 3D “bare” models, due to the lack of practical guidance in most codes on modelling masonry infills.

Avoid common mistakes on your manuscript.

1 Introduction

Post-earthquake surveys worldwide have brought to light the significant impact of shear failures in beams, columns, or beam-column joints on the seismic performance of existing reinforced concrete (RC) buildings. Recent catastrophic seismic events (e.g., Verderame et al. 2014 ; Masi et al. 2019 ) have highlighted the detrimental effects of inadequate transverse reinforcements or the absence of seismic detailing, especially in joint regions and columns, on the structural response. Consequently, ensuring the accurate estimation of shear capacity is imperative for a comprehensive evaluation of the structural performance (Karakas et al., 2022 ; Lupoi et al. 2004 ) and the design of effective retrofitting strategies.

In the literature, a multitude of shear capacity models for existing beam-column elements have been formulated by means of an empirical approach. Several of these (Priestley et al. 1994 ; Sezen and Moehle 2004 ; Biskinis et al. 2004 ; Kowalsky and Priestley 2000 ) forecast a deterioration in shear strength under seismic cyclic loading as ductility demand rises. Nonetheless, despite being conceptually rooted in the same theory, these models exhibit significant differences. Shear resistance is primarily attributed to two factors: one associated with the presence of transverse reinforcements, and the other reliant on the concrete resisting mechanisms. Nevertheless, some models relegate the strength cyclic degradation to the concrete resisting contributions only (e.g, Priestley et al. 1994 ), owing to a progressive reduction in load-carrying capacity with crack propagation. Other models extend the shear strength degradation to the contribution of transverse reinforcement, considering the potential loss of anchorage and bond capacity of the reinforcement due to concrete cracking (Sezen and Moehle 2004 ; Biskinis et al. 2004 ). In addition, among the various capacity models, there is not unanimous agreement on the definition of resistance contributions, especially about the concrete strength contribution. For example, Sezen and Moehle ( 2004 )’s model evaluates this contribution using the Mohr’s circle approach including the influence of axial load. In contrast, Biskinis et al. ( 2004 )’ s model considers the axial load with an additional resistance contribution, not subjected to any cyclic degradation effect.

In addition to empirical capacity models, the shear strength of RC elements can be assessed through the application of the Modified Compression Field Theory (Vecchio and Collins 1986 ) or its simplified variants (Bentz et al. 2006 ; CSA Standard, 2006 ; Model Code, 2010 ; Marcantonio et al. 2015 ). This theory represents the latest advancement of an approach that originated in the early 1900s, according to which the shear strength of a RC element is governed by a truss mechanism (Ritter 1899 ; Morsch 1909 ) with compressive stresses inclined at 45° to the longitudinal axis of the element. This very first model neglected any contribution by the cracked concrete, potentially resulting in overly conservative estimates of shear strength for elements with limited transverse reinforcement. Moreover, studies from the 1980s revealed that the inclination angle is seldom exactly 45°. The necessity for a rational determination of this angle gave rise to the Compression Field Theory (CFT, Collins 1978 ), subsequently modified to consider the tensile stresses in the cracked concrete (MCFT, Vecchio and Collins 1986 ). The method estimates the inclination angle based on the strain distribution in the cross-section of the element and it has been suggested by the Model Code ( 2010 ) - with different levels of approximations (Model Code, 2010 ; Biskinis and Fardis 2020 ).

In last years, many experimental works have been conducted focusing on the estimation of the shear strength of beam-column joints without transverse reinforcement, which typically characterize existing buildings. Most of these studies focus on exterior joints (e.g., Vollum and Newman 1999 ; Pantelides et al. 2002 ; Tsnos, 2007 ), generally more vulnerable, identifying the parameters that can most significantly affect joint strength. Among these parameters, in addition to joint configuration (interior or exterior) and concrete compressive strength (Priestley 1997 ; Kim and LaFave 2012 ), the joint aspect ratio, the beam longitudinal reinforcement ratio (Park and Mosalam 2012 ), and the column axial load (Priestley 1997 ) were identified as key factors, as also confirmed by Jeon et al. ( 2014 ) based on a wide experimental dataset. Nevertheless, there is not a full consensus within the research community about the influence of some factors on the joint shear strength. For example, several studies acknowledge the influence of axial force only on the deformability of the joint and not on its strength (e.g. Fujii and Morita 1991 ; Park and Mosalam 2012 ). An increase in column axial load has not or limited influence on interior and exterior joints strength respectively, according to Fujii and Morita ( 1991 ). A detrimental effect on the joint shear strength due to high axial loads has been observed by Li et al. ( 2015 ).

The discrepancies among the different brittle capacity models proposed in the literature, both for beam/column elements and beam-column joints, have been integrated into various national standards (e.g., EN 1998-3, 2005 ; D.M. 2018 ; ASCE/SEI-41, 2017 ; NZS 3101, 2006 ), and, thus, the assessment of the seismic capacity of RC buildings is based on technical codes that rely on (often very) different capacity models.

This study aims at investigating the potential impact of using different code-based brittle capacity models first in terms of parametric comparison and, then, by applying them on the seismic assessment and retrofit of RC case-study buildings. This assessment is conducted through nonlinear static pushover analysis within the N2 framework (Fajfar 2000 ) on Italian pre-1970 (“pre-code”) case-study buildings with different numbers of stories, in both the as-built condition and after the implementation of a retrofitting strategy that addresses brittle tensile-only failures (De Risi et al. 2023a ). Three current code prescriptions are considered herein: the current Eurocode 8 (labelled “EC8 2005” in what follows) (EN 1998-3, 2005 ), Italian technical code D.M. ( 2018 ) (labelled “NTC 2018” hereinafter), and American standards “ASCE/SEI” (ASCE/SEI-41, 2017 ). Based on European and Italian codes (EC8 2005 ; NTC 2018) approach, the building capacity at the Severe Damage (SD) Limit State (LS) is always assessed as that corresponding to the first failure attained at that LS. It is worth noting that this choice of “failure” criterion is certainly conservative with respect to the “real” (sidesway or gravity load) collapse of a building (Shoraka, 2013 ), as well know, but it is also more conservative with respect to other code-based approach (e.g. Turkish TBEC 2018 , according to which a certain percentage of RC members can reach a given LS). Due to the main aim of this study, the sole distinction among the code cases applied herein lies in the implemented brittle capacity models, while the framework for determining the seismic capacity remains consistent with the European codes: the seismic demand is uniform across all code cases (in contrast to Dhanvijay and Nair, 2015 ), and the ductile capacity of beam/column elements is always defined as prescribed by EC8 2005 (and NTC 2018).

Within the codes framework, brittle failures are typically identified through post-processing the data obtained from (linear or) nonlinear analyses. However, it is worth noting that American standards also explicitly provide tools to model the nonlinear behaviour of shear-sensitive elements, including beam-column joints (i.e., scissor model, Alath and Kunnath 1995 ), providing the backbone for their implementation (ASCE/SEI-41, 2017 ; Hassan 2011 ; Hassan and Elmorsy 2022a , b ).

Lastly, it is worth noting that, very recently, some works from the literature focused on the analysis of the capability of code prescriptions to catch real capacity and seismic damage extend and severity in existing buildings. Cook et al. ( 2023 ) and Sen et al. ( 2023 ) analysed the results of structures experimentally damaged via shake tables testing, to compare the experimental response with the simulated outcomes following ASCE/SEI 41 application, aiming at promoting its improvement. Similarly, in European context, a challenging work is currently ongoing by several European research groups to update the current European standards (e.g., Fardis ( 2021 ), Biskinis and Fardis ( 2020 ), Franchin and Noto ( 2023 ), Maranhão et al. ( 2024 ), among others). Therefore, a focus on the brittle capacity models of the (current draft of the) incoming second-generation Eurocode 8 - currently ongoing and in its final steps of development - is carried out in this work. Some studies from the literature have already started analysing the main differences compared to the current version. For instance, the design of moment resisting frame RC buildings according to the second-generation code has been compared to the previous EC8 version in Maranhão et al. ( 2024 ). The next-generation EC8 will modify the current shear strength model of beam-column joints (Fardis 2021 ) and change significantly the brittle capacity model to be used for beam/column members, moving from the empirical model by Biskinis et al. ( 2004 ) (EC8 2005 ) to a MCFT-based approach (Biskinis and Fardis 2020 ). Such novelties could potentially be very impactful and, thus, they are investigated in this work.

2 Overview of the current code-based shear capacity models worldwide

In this section, a description of some of the main shear capacity models currently adopted worldwide for beams/columns and joints is provided. The capacity models adopted by the current European (EC8 2005 ), Italian (NTC 2018; Circolare 2019 ), and American (ASCE/SEI) technical standards are analysed. A parametric comparison is also carried out to identify hierarchies and trends in resulting strengths.

2.1 Beams/columns shear strength models

Nowadays, worldwide, the shear strength of existing RC beam/column elements is generally evaluated using “degrading” models (De Luca and Verderame 2013 ), which predict a decreasing shear strength as the plastic displacement demand increases.

According to EC8 2005, the shear strength (V R ) of a beam/column element is calculated as proposed by Biskinis et al. 2004 ( \(\:{\text{V}}_{\text{R},\text{B}\text{I}\text{S}}\) ), namely, as the sum of three contributions:

In Eq. ( 1 ), the coefficient γ el , accounting for uncertainties in fitting experimental data, is equal to 1.15 for primary elements. \(\:{\text{V}}_{\text{N}}\) is the contribution due to the presence of compressive axial load N (Paulay and Priestley 1992 ). The latter is limited to 55% of the maximum axial load that the concrete section can sustain (i.e., \(\:{\text{A}}_{\text{c}}{\text{f}}_{\text{c}}\) , being \(\:{\text{A}}_{\text{c}}\) the area of the concrete cross-section and \(\:{\text{f}}_{\text{c}}\) the concrete compression strength). \(\:\text{h}\) is the section height, x the neutral axis depth, and \(\:{\text{L}}_{\text{v}}\) the element shear span. The contribution of the post-cracking concrete resistance mechanisms, \(\:{\text{V}}_{\text{c}}\) , can be expressed as α times √f c \(\:{\cdot\:\text{A}}_{\text{c}}\) . The term α depends on the total geometric percentage of the longitudinal reinforcement, \(\:{{\rho\:}}_{\text{t}\text{o}\text{t}}\) , and on the slenderness of the element, \(\:{\text{L}}_{\text{V}}/\text{h}\) . Lastly, the contribution of the transverse reinforcement, \(\:{\text{V}}_{\text{w}}\) , is the same proposed by Ritter-Morsch model (Ritter 1899 ; Morsch 1909 ). Thus, it depends on the stirrups area ( \(\:{\text{A}}_{\text{s}\text{w}}\) ), yielding strength ( \(\:{\text{f}}_{\text{y}\text{w}}\) ), and spacing (s), and on the internal lever arm (z) - assumed hereinafter as 0.9 times the cross-section effective depth, d.

According to EC8 2005, V R degrades by means of the coefficient, \(\:\text{k}.\) The latter decreases as displacement ductility demand (µ Δ ) increases (Fig.  1 ), moving linearly from 1 (no degradation) to 0.75 (maximum degradation).

Shear capacity model according EC8-2005 and NTC 2018 ( a ); degradation coefficient according to EC8 2005 and ASCE/SEI-41 ( b )

It is worth underlying that the EC8 2005 provides materials strengths reduction factors for safety check at SD LS. In particular, the mean strengths resulting from in-situ tests must be divided by the partial safety factor (γ c  = 1.50 for concrete and γ s  = 1.15 for steel) and by the Confidence Factor (i.e., CF) depending on the Knowledge Level (KL). In this work, a comprehensive KL has been always assumed, and, thus, CF = 1.00.

The same shear capacity model is (partially) adopted also by Italian technical code (NTC 2018), which introduces a modification for low µ Δ levels, by using the truss model of shear resistance with variable inclination diagonals (Biskinis and Fardis 2004 ). The latter is hereinafter referred to as Variable Inclination Truss (i.e., VIT) model. In particular, V R is the same provided by EC8 2005 model when µ Δ  ≥ 3 (i.e., V R, NTC18 =V R, EC8 ). When µ Δ  ≤ 2, V R is the maximum between the values provided by EC8 2005 model and VIT model. Lastly, for intermediate ductility demand, V R, NTC18 is obtained by linearly interpolating between the two models (Fig.  1 a). As well known, according to VIT model (prescribed for not-seismic loadings by NTC 2018 and EN 1998-1, 2004 ), the shear strength ( \(\:{\text{V}}_{\text{R},\text{V}\text{I}\text{T}}\) ) is the minimum between compressed strut strength, \(\:{\text{V}}_{\text{R}\text{c}}\) , and the tensile strength of the transverse reinforcement, \(\:{\text{V}}_{\text{R}\text{s}}\) , i.e. (in case of stirrups):

In Eq. ( 2 ), \(\:{\theta\:}\) is the inclination angle of the compressed struts with respect to the longitudinal axis of the element, b the web width, \(\:\stackrel{-}{{\nu\:}}\) is equal to 0.5, and \(\:{{\alpha\:}}_{\text{c}}\) is a function of N. According to Italian code, \(\:\text{c}\text{o}\text{t}{\theta\:}\) in Eq. ( 2 ) mut be limited between 1.00 and 2.50.

Similarly to European code, Italian guidelines also require that materials strengths must be divided by the partial materials factors and by the CF for safety checks at SD LS.

The model adopted by ASCE/SEI ( \(\:{\text{V}}_{\text{R},\text{A}\text{S}\text{C}\text{E}}\) ) is based on Sezen and Moehle ( 2004 )’s model, i.e. an additive degrading model relying on two contributions: \(\:{\text{V}}_{\text{c}}\) , due to concrete post-cracking mechanisms and axial load, and \(\:{\text{V}}_{\text{w}}\) .

where \(\:{\text{V}}_{\text{w}}\) has the same meaning of Eq. ( 1 ). According to this model, the degradation coefficient \(\:\text{k}\) is equal to 1.00 for µ Δ  ≤ 2, and 0.70 for µ Δ  ≥ 6, varying linearly between these two bounds (Fig.  1 b). ASCE/SEI model predicts a higher strength degradation compared to EC8 2005. Indeed, on one hand, the “residual” strength is derived from a lower degradation coefficient \(\:\text{k}\) ( \(\:\text{k}=0.70\) ); on the other hand, this coefficient multiplies all the strength contributions (even that related to the axial load).

The material strengths to be used for assessment are, also in this case, the reduced strengths. However, while γ c assumes the same value provided by European codes, γ s is higher (i.e., 1.25). Furthermore, this standard prescribes that, in case of “comprehensive” knowledge (maximum level), CF is equal to the 1.

In existing structures, especially if designed for gravity loads only, structural elements often have low transverse reinforcement ratios. In this hypothesis, the strength provided by the VIT model, coincides with \(\:{\text{V}}_{\text{R}\text{s}}\) evaluated with \(\:\text{c}\text{o}\text{t}{\theta\:}=2.5\) . Thus, in these cases, \(\:{\text{V}}_{\text{R},\text{V}\text{I}\text{T}}=\text{min}({\text{V}}_{\text{R}\text{c}};{\text{V}}_{\text{R}\text{s}})={\text{V}}_{\text{R}\text{s}}={\text{V}}_{\text{w}}\text{c}\text{o}\text{t}{\theta\:}=2.5{\text{V}}_{\text{w}}\) .

For high plastic demands, Italian and European codes provide the same shear strength. On the contrary, a difference is observed when µ Δ  ≤ 3. This difference (Eq. ( 4 )) is maximized in absence of strength degradation ( \(\:\text{k}=1\) ) and can be expressed as the sum of three terms. They depend on five parameters: axial load ratio \(\:{\nu\:}=\text{N}/\left({\text{A}}_{\text{c}}{\text{f}}_{\text{c}}\right)\) , mechanical percentage of shear reinforcement \(\:{{\omega\:}}_{\text{s}\text{w}}={\text{A}}_{\text{s}\text{w}}{\text{f}}_{\text{y}}/\left(\text{b}\cdot\:\text{s}\cdot\:{\text{f}}_{\text{c}}\right)\) , the above-defined \(\:{\text{L}}_{\text{V}}/\text{h}\) , \(\:{{\rho\:}}_{\text{t}\text{o}\text{t}}\) , and the mean concrete compressive strength, \(\:{\text{f}}_{\text{c}\text{m}}\) .

In Eq. ( 4 ), \(\:{\Delta\:}{{\text{V}}_{\text{R}}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) is the difference between \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) and \(\:{\text{V}}_{\text{R},\text{V}\text{I}\text{T}}\) , and it is normalized with respect to the quantity \(\:{\text{A}}_{\text{c}}{\cdot\:\text{f}}_{\text{c}\text{m}}\) . So, when the combination of the five parameters above leads to positive values of ΔV R , \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) > \(\:{\text{V}}_{\text{R},\text{V}\text{I}\text{T}}\) . Note that \(\:{\text{f}}_{\text{c}}\) and \(\:{\text{f}}_{\text{c}\text{m}}\) represent both the concrete compressive strength, but the latter is a mean value (derived from in-situ tests), whereas the former is evaluated as \(\:{\text{f}}_{\text{c}}={\text{f}}_{\text{c}\text{m}}/\left(\text{C}\text{F}\cdot\:{{\gamma\:}}_{\text{c}}\right)\) .

The same normalized difference can be evaluated by comparing the non-degraded shear strengths resulting from the European and American codes, as shown in Eq. ( 5 ) (which results quite similar to Eq. ( 4 )):

Note that the expressions above assume that:

\(\:{{\rho\:}}_{\text{t}\text{o}\text{t}}\) is not lower than 0.50% (in tune with the definition of α in Eq. ( 1 ));

\(\:{{\omega\:}}_{\text{s}\text{w}}\) is compatible with the assumption of a weakly reinforced element (i.e., not exceeding 0.07, value which provides \(\:\text{c}\text{o}\text{t}{\theta\:}\) always limited to 2.5), as typical in existing buildings;

\(\:{\text{L}}_{\text{V}}/\text{h}\) ranges between 2 and 4 (considering the limitations of ASCE model);

the cross-section height, h, has been confused with its effective depth, \(\:\text{d}\) , for sake of simplicity.

Figure  2 shows the isocurves corresponding to \(\:{{\Delta\:}\text{V}}_{\text{R}}=0\) resulting from Eq. ( 4 ) (in grey scale) and Eq. ( 5 ) (in blue scale). They display when the code models provide the same strength. Three possible values of \(\:{\text{L}}_{\text{V}}/\text{h}\) and \(\:{\text{f}}_{\text{c}\text{m}}\) are assumed in Fig.  2 (i.e., \(\:{\text{L}}_{\text{V}}/\text{h}=2;3;4,\) and \(\:{\text{f}}_{\text{c}\text{m}}=10;20;30\:\) MPa). The axial load ratio \(\:{\nu\:}\) varies between 0 and 0.5. Each isocurve corresponds to a different value of \(\:{{\rho\:}}_{\text{t}\text{o}\text{t}}\) (ranging between 0.50% and 2.00%). It can be noted that:

Isocurves corresponding to \(\:{\varDelta\:V}_{R}=0\) , varying \(\:\frac{{L}_{V}}{h}\) , \(\:{f}_{c}\) , \(\:{\omega\:}_{sw}\) , \(\:\nu\:\) , and \(\:{\rho\:}_{tot}\) , resulting from the comparison between \(\:{V}_{R,EC8}\) with \(\:{V}_{R,VIT}\) (in gray scale) and with \(\:{V}_{R,ASCE}\) (in blue scale)

for low values of \(\:{{\omega\:}}_{\text{s}\text{w}}\) and high values of \(\:{\nu\:}\) , \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) results higher than the other two models. Actually, according to NTC 2018, for \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) > \(\:{\text{V}}_{\text{R},\text{V}\text{I}\text{T}}\) , \(\:{\text{V}}_{\text{R},\text{N}\text{T}\text{C}18}{=\text{V}}_{\text{R},\text{E}\text{C}8}\) , and, thus, both codes provide exactly the same shear strength;

the area where \(\:{{\Delta\:}\text{V}}_{\text{R}}>0\) covers wider ranges of \(\:{{\omega\:}}_{\text{s}\text{w}}\) and \(\:{\nu\:}\) values when \(\:{{\rho\:}}_{\text{t}\text{o}\text{t}}\) is high;

the latter effect is more pronounced when comparing \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) and \(\:{\text{V}}_{\text{R},\text{A}\text{S}\text{C}\text{E}}\) , and when considering upper bounds of \(\:{\text{f}}_{\text{c}\text{m}}\) and lower bounds of \(\:{\text{L}}_{\text{V}}/\text{h}\) .

Another useful representation of the differences among the considered models is shown in Fig.  3 . A “central” value of \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) (and \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{A}\text{S}\text{C}\text{E}}\) ) is calculated by using mean values (within the above-defined ranges of variation) of the 5 key parameters (central value of \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) and \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{A}\text{S}\text{C}\text{E}}\) results − 0.010 and 0.021, respectively). Then, the 5 parameters have been varied one-by-one to assume their upper or lower bound values (within the above-defined ranges of variation), and corresponding \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) (and \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{A}\text{S}\text{C}\text{E}}\) ) are evaluated. Lastly, the relative variation (Ω) of \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) (and \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{A}\text{S}\text{C}\text{E}}\) ) with respect to the central value is plotted in Fig.  3 a (and b). It is clear that \(\:{{\omega\:}}_{\text{s}\text{w}}\) has the greatest influence on \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{V}\text{I}\text{T}}\) , followed by ν, L V /h and ρ tot . The latter three parameters become more influent on \(\:{{\eta\:}}^{\text{E}\text{C}8-\text{A}\text{S}\text{C}\text{E}}\) , whereas f cm always has a quite small importance in these comparisons (especially when comparing EC8 and ASCE).

Tornado diagrams for sensitivity analysis

A similar comparison can also be carried out focusing on the residual shear strength. The latter comparison makes sense only if \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) and \(\:{\text{V}}_{\text{R},\text{A}\text{S}\text{C}\text{E}}\) are compared, since, when µ Δ  ≥ 3, \(\:{\text{V}}_{\text{R},\text{N}\text{T}\text{C}18}{=\text{V}}_{\text{R},\text{E}\text{C}8}\) . By using the maximum degradation factors (i.e., k equal to 0.70 and 0.75 respectively for \(\:{\text{V}}_{\text{R},\text{A}\text{S}\text{C}\text{E}}\) and \(\:{\text{V}}_{\text{R},\text{E}\text{C}8}\) ), a small modification is observed in the coefficients of Eq. ( 5 ) (the values 0.40, 11.40, and 0.12 are replaced with 0.28, 8.52, and 0.05, respectively). The isocurves in Fig.  2 tend to shift towards lower \(\:{\nu\:}\) and higher \(\:{{\omega\:}}_{\text{s}\text{w}}\) values, making the area with positive \(\:{{\Delta\:}\text{V}}_{\text{R}}\) much wider than that obtained for the non-degraded strength.

2.2 Beam-column joints strength models

The capacity models of beam-column joints prescribed by current global standards differ to each other significantly both for reinforced (Del Vecchio et al. 2023 ) and unreinforced joints, especially when comparing the European approach with the American one.

According to EC8 2005, the shear strength of a beam-column joint is evaluated as in EN 1998-1, 2004, for newly designed buildings, by means of two safety checks (related to a tensile and a compressive failure mode). These checks can be reformulated in terms of joint shear stress \(\:{{\tau\:}}_{\text{j}}={\text{V}}_{\text{j}}/{\text{A}}_{\text{j}}\) (where \(\:{\text{V}}_{\text{j}}\) is the joint shear load and \(\:{\text{A}}_{\text{j}}\) the joint horizontal area, according to EN 1998-1, 2004 ) and normal vertical stress \(\:{{\sigma\:}}_{\text{v}}=\text{N}/{\text{A}}_{\text{c}}\) (due to the axial force related to the column above the joint). By assuming a joint without stirrups (as typical in existing buildings), Eq. ( 6a ) represents the tensile failure check, whereas Eq. ( 6b ) the compressive failure check:

In Eq.s (6), η is equal to 0.60(1 - f ck /250) for interior joints and 0.48(1 - f ck /250) for exterior ones; f ct is the concrete tensile strength (according to EN 1992-1-1, 2004 ). This strength, as well as the concrete compressive strength \(\:{\text{f}}_{\text{c}}\) , is intended to be the mean strength reduced by CF and partial materials safety factors (EC8 2005 ). Thus, \(\:{\text{f}}_{\text{c}\text{t}}={\text{f}}_{\text{c}\text{t}\text{m}}/(\text{C}\text{F}\cdot\:{{\gamma\:}}_{\text{c}})\) with \(\:{\text{f}}_{\text{c}\text{t}\text{m}}=0.30\sqrt[3]{{\left({\text{f}}_{\text{c}\text{k}}\right)}^{2}}\) ). Conversely, \(\:{\text{f}}_{\text{c}\text{k}}\) is a characteristic concrete compression strength value, assumed equal to \(\:({\text{f}}_{\text{c}\text{m}}-8)\text{M}\text{P}\text{a}\) (EN 1992-1-1, 2004 ).

For existing buildings, the Italian standard (Circolare 2019 , C8.7.2.3.5) prescribes a double strength check for joints that are not fully confined according to Eq.s (6), as well. Nevertheless, it assumes \(\:{\eta\:}=0.50\) and \(\:{\text{f}}_{\text{c}\text{t}}=0.30\sqrt{{\text{f}}_{\text{c}}}\) (with \(\:{\text{f}}_{\text{c}}={\text{f}}_{\text{c}\text{m}}/(\text{C}\text{F}\cdot\:{{\gamma\:}}_{\text{c}})\) ). As a results, comparing Italian and European codes, a difference in safety checks results is obtained even if both use Eq.s (6). Figure  4 shows this difference, distinguishing between exterior (“EXT J”) and interior (“INT J”) joints, and assuming three f cm values to fix ideas.

Beam-column joint strengths according to EC8 2005 and NTC2018, given the joint configuration and the concrete compressive strength (tensile check with solid lines; compressive check with dotted lines)

For low values of f cm (f cm = 10 MPa), NTC 2018 model provides higher tensile joint strength (solid lines in Fig.  4 ) compared to EC8 2005; conversely, at higher values of f cm (f cm = 30 MPa), the hierarchy is reversed, with almost coincident strengths if f cm = 20 MPa.

Regarding the compressive safety check (dotted lines in Fig.  4 ), EC8 2005 model provides a lower resistance in the case of exterior joints and a higher strength for interior ones.

However, for both European standards, the joint strength results as the minimum between those produced by Eq.s (6), given the value of ν. In other words, for low values of axial load ratio, the joint strength is limited by that corresponding to diagonal cracking (i.e. tensile failure), while for high ν, the joint fails due to compression failure.

According to American standard (ASCE/SEI), a joint shear stress capacity equal to \(\:{\lambda\:}{\gamma\:}^{\prime\:}\sqrt{{\text{f}}_{\text{c}}\:}\) is assumed, being λ = 1 for normal-weight aggregate concrete. The γ′ coefficient (see Table  1 ) depends on various parameters: joint typology (i.e., interior, exterior, or knee joint), presence/absence of transverse beams, presence/absence of “conforming” transverse reinforcement. Note that according to the American Code, if the stirrup spacing in the joint is less than or equal to half the column cross-section height, then the joint is considered as conforming . Otherwise, the joint is nonconforming . Thus, unlike EC8 2005 and NTC 2018, American standard prescribes a single safety check (Eq. ( 7 )):

It should be noted that, while European standards lead to a joint strength variation with N, the American guideline always provides the same joint strength irrespective of the axial load level.

Another main difference of ASCE/SEI approach compared to European codes lies in the possibility of explicitly modelling the behaviour of the joint - possibility that could significantly impact the assessment outputs. This modelling is allowed by European standards as well, which, however, do not provide specific reference models.

In Fig.  5 , the trend of the joint strength (expressed as \(\:{{\tau\:}}_{\text{j}}/\sqrt{{\text{f}}_{\text{c}\text{m}}}\:\) ) is provided for fixed values of \(\:{\text{f}}_{\text{c}\text{m}}\) , depending on ν, according to all the strength models above.

Strength domains of beam-column joints according to NTC2018, EC8 2005 and ASCE/SEI, given the joint configuration and the concrete compressive strength

For European models, the joint strength is evaluated for each axial load value as the minimum between tensile and compressive strength. This type of representation can be considered as a strength domain. Indeed, considering a given joint typology (i.e., interior or exterior) and a given \(\:{\text{f}}_{\text{c}\text{m}}\) , the demand joint shear load (i.e., \(\:{\text{V}}_{\text{j}}\) ) and axial load (i.e., N) allow deriving the \(\:{{\tau\:}}_{\text{j}}-{\nu\:}\) coordinates of a “demand point”. If this point is inside or belongs to the boundary of the domain (related to the specific strength model), then the joint is on the safe side. The ascending branches of these domains represent the tensile check; the descending branches the compressive check.

Regarding American Code, joints are assumed as non-conforming herein, since, typically, stirrups in joints are totally missing in existing buildings. For interior joint (especially with transversal beams and low \(\:{\text{f}}_{\text{c}\text{m}}\) values), the joint strength is overestimated by ASCE/SEI compared with the European codes. Conversely, for exterior joints, the hierarchy among the models depends on ν and on \(\:{\text{f}}_{\text{c}\text{m}}\) values.

Moreover, ASCE/SEI provides a different strength for knee joints (i.e., γ′ = 4 √MPa). Being located on the top floor of the building, for these joints, zero axial load can be assumed. Thus, the joint strength according to European models can be obtained assuming \(\:{{\sigma\:}}_{\text{v}}=0\) in Eq.s (6) (i.e., ν = 0 in Fig.  5 ), generally resulting lower than joint strength by ASCE/SEI.

3 Shear capacity models according to the next-generation of Eurocodes

In the previous section, the capacity models prescribed by current standards have been analysed and compared. However, a paramount work is currently ongoing by European research groups to update the current European standards with a second-generation of Eurocodes in the next years. Significant changes will be carried out to the shear strength models of both beam/column elements and beam-column joints, as highlighted by the recently published works from the literature (Biskinis and Fardis 2020 ; Fardis 2021 ; Franchin and Noto 2023 ). Thus, in this section, the capacity models introduced by the incoming second-generation of Eurocodes will be first analysed, emphasizing their evolution with respect to the current version. The current available drafts of the second-generation of Eurocodes adopted herein are prEN 1998-3:2023, FprEN 1998-1-1: 2024 (along with its previous draft FprEN 1998-1-1: 2022 ), and FprEN 1992-1-1:2023, along with the relevant references from the literature, recalled in the next sub-paragraphs.

3.1 Beams/columns shear strength

In the second-generation of Eurocode 8-part 3 (prEN 1998-3: 2023 ), the shear capacity of existing RC beams and columns must be evaluated according to a model based on the variable inclination, θ, between the compression stress field in the member web and the member axis (Biskinis and Fardis 2020 ).

PrEN 1998-3:2023 prescribes to evaluate the shear strength, \(\:{\text{V}}_{\text{R},\text{E}\text{C}8-2\text{n}\text{d}}\) , according to FprEN 1998-1-1:2024, by using the mean values of the material properties and also following FprEN 1992-1-1:2023 suggestions, even if with some modifications explained below. \(\:{\text{V}}_{\text{R},\text{E}\text{C}8-2\text{n}\text{d}}\) can be expressed as in Eq. ( 8 ):

In Eq. ( 8 ), V N is evaluated similarly to Eq. ( 1 ), and the variable inclination θ has the same meaning of the VIT model, ranging between 1 and \(\:{\text{c}\text{o}\text{t}{\theta\:}}_{\text{m}\text{i}\text{n}}\) (see Eq. ( 9 )). The latter depends on the axial load, \(\:\text{N}\) .

However, \(\:\text{c}\text{o}\text{t}{\theta\:}\) may exceed the upper limit, \(\:{\text{c}\text{o}\text{t}{\theta\:}}_{\text{m}\text{i}\text{n}}\) , if the deformation state of the cross-section is analysed. In fact, the value of \(\:\stackrel{-}{{\nu\:}}\) is not necessarily a constant value (i.e., 0.5 as prescribed by the VIT model), and it can be obtained based on the state of strains of the member according to Eq. ( 10 ) (FprEN 1998-1-1: 2024 ):

where the reduction factor 1/1.6 is applied to account for cycling loading (Biskinis and Fardis 2020 ), and \(\:{{\epsilon\:}}_{\text{x}}\) is the average strain between the bottom and top chords, ranging between 0 and 0.02 (FprEN 1998-1-1: 2024 ). Note that, strictly speaking, according to FprEN 1998-1-1:2024 draft, \(\:\stackrel{-}{{\nu\:}}\) in seismic loading conditions should be always higher than 0.5/1.6 (= 0.31). However, the latter prescription was not present in the previous draft (FprEN 1998-1-1: 2022 ), nor in original works by Biskinis and Fardis ( 2020 ); additionally, it would result very close to the TIV model and in a not safe-sided prescription. Thus, it has not been applied in what follows.

\(\:{{\epsilon\:}}_{\text{x}}\) is calculated as in Eq. ( 11 ) (FprEN 1992-1-1: 2023 ):

where \(\:{\text{A}}_{\text{s}\text{t}}\) and \(\:{\text{A}}_{\text{s}\text{c}}\) are the areas of the longitudinal reinforcement in the flexural tension chord and flexural compression chord, respectively; \(\:{\text{A}}_{\text{c}\text{c}}\) is the area of the flexural compression chord. Lastly the “chord forces”, \(\:{\text{F}}_{\text{t}\text{d}}\) and \(\:{\text{F}}_{\text{c}\text{d}}\) , are defined as a function of the flexural ( \(\:{\text{M}}_{\text{E}\text{d}}\) ) and shear ( \(\:{\text{V}}_{\text{E}\text{d}}\) ) demand, and of axial force.

Moreover, in member end-zones expected to enter the inelastic range, the values of \(\:{\text{M}}_{\text{E}\text{d}}\) and \(\:{\text{V}}_{\text{E}\text{d}}\) from the analysis should be multiplied by the chord rotation ductility factor, \(\:{{\mu\:}}_{\varDelta\:}\) . It is worth noting that the approach proposed in Eq. ( 12 ) is a simplified approach, based on the assumption of the equal displacement rule. Nevertheless, the effective average strain \(\:{{\epsilon\:}}_{\text{x}}\) should be rigorously evaluated, considering the curvature and the neutral axis depth of the cross-section (Biskinis and Fardis 2020 ).

The additional tensile axial load, \(\:{\text{V}}_{\text{E}\text{d}}\text{c}\text{o}\text{t}{\theta\:},\) and factor \(\:\stackrel{-}{{\nu\:}}\) depend on \(\:\text{c}\text{o}\text{t}{\theta\:}\) , resulting in an iterative procedure to derive the inclination θ and, thus, the shear capacity \(\:{\text{V}}_{\text{R},\text{E}\text{C}8-2\text{n}\text{d}}\) .

Lastly, in the code-based safety check at SD LS, the shear resistance of existing members (prEN 1998-3: 2023 ), should be divided by the corresponding safety factor related to the resistance, γ Rd (prEN 1998-3: 2023 ). The latter accounts for uncertainty in the shear strength assessment and is evaluated as in Eq. ( 13 ) (Franchin and Noto 2023 ):

In Eq. ( 13 ), α R is the resistance sensitivity factor, equal to 0.85 according to FprEN 1998-1-1:2024 and Franchin and Noto ( 2023 ). The target reliability index in a 50-years reference period, β LS, CC, depends on both the considered limit state and the consequence class. According to the Annex F of FprEN 1998-1-1:2024, for SD LS and CC2 (second consequence class), β LS, CC is equal to 1.60. Lasty, the total logarithmic standard deviation \(\:{{\sigma\:}}_{\text{l}\text{n}\text{R}}\) for existing members with rectangular cross-section is equal to 0.40 (prEN 1998-3:2023- Table 8.5) when the KL3 is attained, as assumed herein. As a result, \(\:{{\gamma\:}}_{\text{R}\text{d}}=1.72\) is obtained herein.

3.2 Beam-column joints

According to PrEN 1998-3:2023, the shear resistance of existing beam-column joints should be evaluated as prescribed for new elements (prEN 1998-1-1: 2024 ). Based on prEN 1998-1-1:2024, for unreinforced joints a cracking strength (V Rj, cr ) only is provided. Vice-versa, in presence of transverse reinforcement, V Rj, cr can be overcome and joint strength estimated based on studies by Fardis ( 2021 ). Nevertheless, prEN 1998-1-1:2024 also specifies that, in a safe-side and simplified approach, joint strength can be calculated as the maximum between the shear resistance at the first cracking and a minimum value of joint strength (V Rj, min ), the latter related to the absence of transverse reinforcement and axial load:

In Eq. ( 14 ), α is equal to 0.5 for exterior joints and 1.2 for interior ones, h b and h c are the beam and column depth, respectively, and other parameters have been defined above. Eq. ( 14 ) is applied herein to calculate the shear strength of unreinforced joints according to the second-generation Eurocode. Thus, a first comparison with EC8 2005 can be easily carried out, as shown in Fig.  6 , in terms of shear stress τ j /√f cm , and, assuming four h b /h c ratios (i.e., 500 mm/[300 400 500 600]mm). It is worth noting that, ν in Fig.  6 is defined as a function of f cm , both for EC8 2005 (contrary to what Fig.  5 shows) and for the incoming-code, for sake of comparison. Additionally, the application of \(\:{{\gamma\:}}_{\text{R}\text{d}}\) factor for joints is not foreseen in the currently available drafts (i.e., \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1). However, it is reasonably very likely that in the final version of Eurocode 8, a \(\:{{\gamma\:}}_{\text{R}\text{d}}\) factor similar to those used to reduce the shear strength of beams/columns will be introduced. For this reason, herein, the joint strength has been assessed with a twofold assumption: \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1 and \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1.72.

Strength domain for unreinforced beam-column joints: comparison between first- and second-generation of Eurocodes

Unlike EC8 2005, the current draft of the second-generation Eurocode does not explicitly provide a compression limitation for unreinforced joints, leading to a different shear strength-axial load trend for high axial load ratios (even for ν < 0.3 for exterior joints). Instead, the presence of a minimum strength leads to higher V j, EC8−2nd in the case of interior joint, especially for low axial loads and high values of the h b /h c . Therefore, moving from the first to the second generation, a lower number of joint failures can be expected for the top floors interior joints (if characterized by lower h c values, i.e., higher h b /h c ratios and lower axial loads). Moreover, the use of a \(\:{{\gamma\:}}_{\text{R}\text{d}}\) coefficient higher than 1 significantly reduces the joint strength, and, consequently, its hierarchy with respect to the current Eurocode (see Fig.  6 ).

4 Case-study buildings: description and modelling

According to the ISTAT ( 2011 ) census, roughly 1/3 of Italian RC buildings have been built before 1970, when most of the national territory (about 6700 municipalities) was classified as not-seismic prone area. About 2% of municipalities not seismically classified before 1970 are nowadays classified as first seismic zone, based on expected value of acceleration on stiff soil (a g ) with 10% probability of exceedance in 50 years exceeding 0.25 g. 23% is classified as second seismic zone (0.15 \(\:<\) a g \(\:\le\:\) 0.25 g), about the 60% as third (0.05 \(\:<\) a g \(\:\le\:\) 0.15 g) and the 17% as fourth (a g \(\:\le\:\) 0.05 g) seismic zone.

In this section, two case-study buildings have been selected to analyse the difference among code-based brittle capacity models described in Sects. 2 and 3. They are RC residential buildings designed according to the technical regulations in force in Italy until 1970 (Royal Decree, R.D. 2229, 1939 ), to withstand only gravity loads, located in the about 6700 municipalities mentioned above. Case-study buildings haves the same floor area, but different number of stories, Ns (2 and 4), being buildings with Ns ≤ 4 the most widespread in Italian building stock (ISTAT 2011 ).

4.1 Buildings description

The selected case study buildings are in line with the prevalent construction practices in force in Italy before ‘70s. Each structure has a Moment Resisting Frame (MRF) system, consisting of 2D parallel resisting frames in the longitudinal (X) direction (see Fig.  7 a), without interior beams in the transverse (Y) direction. Buildings are symmetric in both directions (X and Y). Floor slabs are 20 cm width, and the inter-story height is 3.00 m (see Fig.  7 b).

Plan view ( a ) of case-study buildings and related representative frames ( b ); cross-sections of typical beams ( c ) and columns ( d ) (dimensions in millimetres)

The cross-section dimensions and reinforcement details are based on a simulated design (Verderame et al. 2010 ; De Risi et al. 2023b ) according to the Italian code in force during the construction period (R.D.2229, 1939 ). A maximum allowable stress of 5.0 or 6.0 MPa was considered for concrete (depending on compressive loads or bending actions, respectively), and of 140 MPa for reinforcing plain bars (type AQ42). All the beams have a 30 × 50 cm² cross-section, with a geometrical percentage of longitudinal reinforcement (ρ l ) ranging from 0.40% to about 0.90%– see Fig.  7 c. Column sections (Fig.  7 d) vary from 30 × 30 cm² (for the upper storeys) to 30 × 40 cm² (for the central columns of the ground floor of the 4-storey building), with a decreasing reinforcement ratio from the ground floor of the 4-storey building (ρ l  ≈ 0.80%) to the last floor (ρ l  ≈ 0.60%). The minimum requirement specified by the R.D.2229 ( 1939 ) is adopted as transverse reinforcement (Fig.  7 c, d). Note that no transverse reinforcement was placed within beam-column joints, since the technical code in force at the construction time did not require any design nor reinforcement of joints. Additional information about the main buildings features is reported in De Risi et al. ( 2023a ).

Lastly, f cm and mean yielding strength of rebars (f ym ) used for buildings assessment are assumed equal to 20 MPa and 322 MPa, respectively, according to Verderame et al. ( 2010 ) and Masi et al. ( 2019 ), for the relevant time period.

Resulting first mode periods (T X and T Y ), mass participation ratios (m p, x and m p, y ) in both the main directions, and the seismic weight (W), ranging between 8.6 and 10 kN/m 2 , are also shown in Table  2 .

4.2 Modelling assumptions

Each building is modelled in the OpenSees platform (McKenna 2011 ) with 3D “bare” frames. Beams and columns are modelled as ductile elements using a lumped plasticity approach to simulate their flexural response (see Fig.  8 ). This approach is implemented by elastic BeamColumn Elements in series with Zero-Length Elements (featuring by the Pinching4 Uniaxial Material ) at both ends of each beam/column. The flexural response is a moment (M)-chord rotation (θ) relationship calibrated for RC elements reinforced with plain bars (Verderame and Ricci 2018 ), by means of a four-point envelope, integrated herein by an additional point corresponding to the first cracking (Fig.  8 ).

Adopted lumped plasticity approach ( a ); envelope of the flexural response of beams and columns ( b )

Masonry infills are only considered in terms of masses and loads. It is acknowledged that masonry infills play a crucial role on seismic performance of RC buildings. Nevertheless, despite decades of research about this topic, often codes worldwide do not provide comprehensive provisions for numerical modelling of masonry infills and relevant safety checks. This is the case of Italian code (D.M. 2018 ), and of the current European code (CEN, 2004 ) as well. As an example, no information about the in-plane nonlinear response, or, more simply, the elastic stiffness of the infill panels is provided within these codes. Additionally, even the evaluation of infills mechanical properties (e.g., Young modulus or compressive strength) in existing buildings is still a challenging issue in common practice. As a result, in common practice, both for the design of new buildings and the assessment/retrofit of existing ones, infills are neglected (except than as loads and masses). Since this work is intended to be a code-based study, masonry infills are not explicitly modelled. Nevertheless, it is worth noting that a comprehensive risk-based analysis should certainly consider explicitly the presence of infills, if the main aim is a more “realistic” assessment of seismic performance and its improvement, along with a realistic estimation of seismic losses (De Risi et al. 2020a ; Del Gaudio et al. 2021 ).

Similarly, following a typical practice-oriented approach, joints are assumed to be rigid elements, and the floors stiff in their own plane. Lastly, potential shear failures have been identified in post-processing, considering all models introduced in Sects. 2 and 3.

5 Seismic capacity of case-study buildings at SD LS

The code-based assessment at a given LS can be expressed through a capacity-to-demand ratio. NTC 2018 allows to synthetically express this ratio in terms of Peak Ground Acceleration (PGA). The demand mainly depends on the considered LS and the construction location and use. The capacity depends on the attainment of a certain failure condition, generally the first failure occurring at the considered LS (NTC 2018, EC8 2005). The adopted capacity model certainly affects the capacity. This is particularly true for the shear strength models, since brittle failures generally limit the seismic capacity of existing buildings (De Risi et al. 2023a ).

Therefore, in this section, the first achievement and the evolution of brittle failures at SD LS is illustrated, depending on the adopted shear capacity models. Then, the influence of the capacity models on the buildings seismic assessment is analysed, assuming as possible buildings locations all Italian municipalities classified as seismic-prone only after 1970.

5.1 SPO curves and failure mapping

The seismic capacity assessment is performed within the N2 framework (Fajfar 2000 ). Nonlinear static pushover analyses are carried out, considering a lateral load distribution proportional to the first vibration modal shape in each direction. The resulting capacity curves (obtained as suggested by European codes, NTC 2018 and EC8 2005), are shown in Fig.  9 , as spectral displacement (S d ) -versus- pseudo-acceleration (S a (T) up to the occurrence of the first ductile failure (DF) at the SD LS. As suggested by the Italian code, such failure occurs when the demand in terms of chord rotation θ reaches ¾ of the capacity calculated according to Biskinis and Fardis ( 2010 ). Since the focus of the present work is on brittle failures models, such definition of DF capacity point is always kept constant in what follows.

Capacity curves up to the first DF with the relevant collapse mechanisms; evolution of the brittle failures at SD LS according to all considered code-based capacity models

In addition to the capacity curves, also the relevant collapse mechanisms are shown in Fig.  9 . A global collapse mechanism is always observed in the transverse (Y) direction, whereas local mechanisms are observed in the longitudinal (X) direction. Each capacity curve also shows the achievement of all the failure typologies at SD LS, i.e. the first joint failure (JF), and the first shear failure (SF) in beams or columns, according to the considered capacity models. Moreover, the percentage of failing elements is provided in each step of the pushover analysis.

Regarding the beams/columns SFs (which occur only in the X direction on the lowest storeys of the case-study buildings), the most conservative capacity model for the analysed case-study buildings is that proposed by EC8 2005. Only according to this model, even the 2-story building exhibits SFs (especially in the longitudinal exterior beams at the first floor). Considering the other two codes, ASCE/SEI results the less conservative model.

About beam-column joints, current European technical regulations (NTC 2018, EC8 2005) prescribe a dual check. The tensile failure is hereinafter referred to as JF(T). The compression failure is labelled JF(C). It is worth noting that a joint does not necessarily reaches its maximum capacity when diagonal cracking first occurs (Hakuto et al. 2000 ). In these cases, the occurrence of JF(T) could severely limit the actual joint capacity.

Considering JF(T) according to NTC 2018, failures occur in both directions for all the case studies, with a maximum percentage of failing elements in X direction that exceeds 50% of all the joints. The number of failures is about the same moving from the NTC 2018 to EC8 2005 model. Indeed, the diagonal tensile check according to NTC 2018 and EC8 2005 provides very similar capacity when f cm = 20 MPa (Fig.  5 ).

On the contrary, JF(C), which represent a more appropriate failure criterion (Hakuto et al. 2000 ; Park and Mosalam 2012 ; NZS 3101, 2006 ), involves fewer joints (always below 10% of all the joints) of the tallest building, according to NTC 2018 and EC8 2005 models. The joints exhibiting JF(C) are typically interior joints with high axial loads. For this type of joints, EC8 2005 provides higher capacity than NTC 2018 (Fig.  5 ), thus delaying the first JF(C) (of the interior 8-11-14-17 joints at the first floor– see numeration in Fig.  7 a).

Only one safety check is performed according to the American code for joints. Its relevant failure evolution is plotted in Fig.  5 with JF(T) of European codes, since that generally limits the building capacity. ASCE/SEI results less conservative than European codes (see Fig.  5 ), especially for interior joints.

The seismic capacity assessment at the SD LS is, lastly, carried out according the next-generation Eurocodes, assuming a double option for \(\:{{\gamma\:}}_{\text{R}\text{d}}\) (1 or 1.72), as explained above.

With respect to the current EC8, the shear strength model by the second-generation Eurocodes leads to a lower number of SFs in beams/columns, even by using \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1.72. Such failures primarily involve the internal longitudinal beams of the central spans and the central columns (i.e., 8-10-14-16) on the ground floor of the 4-storey building. Note that, if \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1 was used, shear failures in beams/columns are not observed at all.

About JFs(T), the current European model provides intermediate results compared to those of the future Eurocode 8 considering the two \(\:{{\gamma\:}}_{\text{R}\text{d}}\) bounds (coherently with Fig.  6 ). However, for the 2-story building, the number of JFs(T) is approximately the same when applying the strength model of the current Eurocode or that of the second generation with \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1.72. As the number of stories increases, the maximum axial load increases at the bottom stories, and, thus, the unreinforced joint strength from the second-generation Eurocode tends to coincide with that related to cracking, leading to an increase in JF(T) failures even for interior joints (with \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1.72) (see Fig.  6 ). If \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1, a lower number of joint failures is always obtained with respect to the current Eurocode 8.

5.1.1 Influence of joints modelling on the buildings assessment

As clearly highlighted above, the shear failure of joints can significantly limit the building seismic capacity. It is worth noting that the safety check of beam-column joints is often very penalizing because of the use of a force-based approach (i.e., a comparison between shear load and shear strength), in conjunction with the definition of LS achievement when the first element fails. A possible alternative is offered, among the investigated codes, by ASCE/SEI guidelines. ASCE/SEI explicitly introduces the possibility to model the nonlinear response of beam-column joints, for example using the so-called scissors model, shown in Fig.  10 (a) (ASCE/SEI 41, 2017 ; Alath and Kunnath 1995 ). ASCE/SEI also explicitly provides the characterization of the joint nonlinear response and the joint shear strain (γ j ) thresholds to be used for each LS safety check, thus actually introducing the possibility of a displacement-based approach also for elements like joints.

Scissors model for beam-column joints ( a ); nonlinear response of the beam-column joint according to ASCE/SEI ( b )

Both the strain thresholds and the whole joint nonlinear response depend on the joint typology, the axial load ratio, the shear load level, and the presence of (conforming or not) stirrups. A typical nonlinear response of the beam-column joints according to ASCE/SEI suggestion is reproduced in Fig.  10 (b) (ASCE/SEI 41, 2017 ; Hassan 2011 ). These prescriptions by American guidelines have been applied to the buildings analysed herein for a brief comment, in this sub-section only, about the influence of joint modelling, by:

implementing the scissors model (Alath and Kunnath 1995 ) as in Fig.  10 (a);

converting the joint shear stress into joint moment (M j ) as suggested in the literature based on equilibrium equations (Celik and Ellingwood 2008 ; De Risi et al. 2017 ); and.

assuming that the joint rotational spring is equal to γ j (Celik and Ellingwood 2008 ; De Risi et al. 2017 ).

The joint spring is characterized based on f cm as concrete compressive strength, without any reduction coefficient. The achievement of SD LS is (conservatively) assumed herein as the attainment of the beginning of the joint softening response (i.e., when the second point of Fig.  10 (b) is reached for the first time in a joint spring). Pushover curves are updated following this modelling approach.

As a result, for the analysed 2- and 4-storey buildings, joint failures are not detected at all, thus highlighting the great difference in safety check depending on the check approach.

5.2 Seismic capacity

For each building and direction, the capacity curve is bi-linearized according to NTC 2018, obtaining an elastic-perfectly-plastic curve. Since the focus of the present work is on brittle failures models, the bi-linearization approach is kept always constant in what follows.

Starting from the inelastic capacity point, C IN (i.e., the attainment of the first failure at the SD LS on the elasto-plastic bilinear curve), the corresponding elastic capacity point, C EL , is derived by means of Vidic et al. ( 1994 ) relationships. Vidic et al. ( 1994 ) proposal depends on the ratio between the building effective period, T eff , and the corner period T C . The latter is a function of the building location according to NTC 2018, always used herein to characterize seismic hazard. Considering the demand elastic spectra at the SD LS (with return period 475 years) for all the considered sites– always assuming soil type A (NTC 2018, EC8 2005)–, the equal-displacement condition always applies herein (being T eff, X =0.55s and T eff, Y =0.99s for Ns = 2, and T eff, X =0.77s and T eff, Y =1.46s for Ns = 4). The elastic spectral pseudo-acceleration capacity, S a,C (T eff ), is shown in Fig.  11 for each building/direction/code.

As-built capacity in terms of S a (T eff ) at SD LS: X-( a ) and Y-( b ) direction

In almost all buildings/codes, the very first failure occurs on the linear branch of the bilinear capacity curve (resulting in C EL =C IN ). The exceptions are the first failures in the Y direction for all buildings, and in the X direction for the 2-story building, according to ASCE/SEI and next generation EC8 wth γ Rd  = 1. Note that in these cases, the values of S a, C (T eff ) were cut off from the plot, being very high (see Table  3 ). In general, the S a, C (T eff ) evaluated according to American standards is significantly higher compared to those obtained with (current or incoming) European models. Instead, the values provided by Italian and European standards are quite similar to each other, especially in Y direction (see Table  3 ).

The pseudo-acceleration spectrum passing through the elastic capacity point C EL allows associating a capacity PGA value (PGA C ) to each site (as described in De Risi et al. 2023a ). An example is shown in Fig.  12 a. Given C EL point and the spectral parameters of each site at SD LS (according to NTC 2018), as many spectra passing through C EL as the considered sites can be derived. Each of these spectra is characterised by a PGA C value. Thus, PGA C exhibits a certain variability, which clearly stems from the variability of the spectral shape associated with the building location (NTC 2018). Table  3 provides the median values and the 16th and 84th percentiles of PGA C values (PGA C,50 , PGA C,16 , and PGA C,84 , respectively) for both directions. The minimum value of PGA C is always in Y direction.

Example of PGA C derivation (4-storey building in X direction, according to EC8 2005 code) ( a ); PGA C depending on Ns and code, assuming A soil type ( b ) and varying the soil typology ( c )

In Fig.  12 b, the PGA C (hereinafter, the minimum value between the two directions) is provided, showing median values, 16th and 84th percentiles. The capacity values according to EC8 2005, NTC 2018 and next-generation Eurocode derive from the same kind of failure (i.e., JF(T) in Y direction). According to the ASCE/SEI code, too, the capacity at the SD LS is due to a JF, which nevertheless generally occurs for higher displacement demands (see Fig.  9 ), resulting in higher PGA C values (especially for Ns = 2). The coefficients of variation (CoV) of PGA C are quite small (about 18% for both case studies in accordance with ASCE/SEI and next-generation EC8 with γ Rd  = 1; about 10% for the other cases).

Lastly, Fig.  12 c shows the PGA C ratios between the value corresponding to flexible soil types - from B to D (NTC 2018) - and that related to soil A (Fig.  12 b), assumed as a reference. Moving from a rock soil (type A) to a more deformable soil (type D), PGA C progressively decreases for all codes, up to about 50%.

Lastly, the as-built assessment explained above has been repeated by changing f cm , assuming 10 MPa and 30 MPa, as in Sects. 2–3. Table  4 summarizes the results of this further analysis in terms of variation of PGA C,50 (soil A) with respect to the results presented above (for f cm =20 MPa), namely in terms of ΦPGA C,50 = PGA C,50,fcm /PGA C,50,fcm=20MPa . A lower value of f cm leads to lower PGA C (and more brittle failures); vice-versa if f cm increases. Nevertheless, such a variation has different weight depending on the considered building and, above all code. For the 2-storey building:

according to NTC 2018, the considered variation in f cm leads to percentage variation lower than 30% in PGA C ; the first failure always is a JF(T);

according to EC8 2005, higher variations are observed. When f cm decreases, in particular, JF(T) occurs even for gravity loads only, thus leading to a null PGA C (i.e., ΦPGA C,50 =0). Vice-versa, when f cm increases the first failure typology becomes a beam SF;

according to ASCE/SEI, a reduction in f cm leads to very premature JF; whereas, if f cm increases, joints failures disappear and the very first failure is a DF; similar outcomes are obtained according to the next generation EC8 if γ Rd  = 1.00 is assumed;

according to the next generation EC8 with γ Rd  = 1.72, JFs(T) are very sensitive to a reduction in f cm , leading to a significant PGA C reduction due to joint failures under gravity loads only (as for EC8 2005).

For the 4-storey building, similar outcomes are obtained, except for the NTC 2018 case. In this case a f cm reduction leads to considerable increments in axial load levels, and, thus, a very premature attainment of JF(C), even under vertical loads only (ΦPGA C,50 =0). Anyway, the “reference” case (f cm =20 MPa) only is analysed in what follows.

6 Retrofitting by solving shear failures

Building capacity can be improved in several ways, mainly grouped into four strategies: (i) increment of lateral strength and stiffness (e.g., by means of shear walls), (ii) increment of displacement capacity only (e.g., by fibre-reinforced polymer (FRP) wrapping or steel cages); (iii) mixed implementation of (i) and (ii); (iv) reduction of the demand (e.g., by using seismic isolators or dissipation devices). However, when shear failures significantly limit the building capacity, as in the above-analysed cases, a possible retrofitting strategy could just aim at solving the detected shear failures. This would make possible the achievement of (more favourable) ductile failures, even without any change in lateral stiffness nor in collapse mechanism. This latter retrofitting approach is one of the less invasive and less expensive strategies, and it is applied herein to analyse its effectiveness depending on the adopted code.

6.1 Retrofitting design

The main objective of the adopted strengthening strategy is the enhancement of the seismic capacity at the SD LS by solving all the (tensile-only) shear failures, without modifying the lateral stiffness of the structural elements. All details about the retrofitting design procedure can be found in De Risi et al. ( 2023a ).

FRP wrapping (e.g., Del Vecchio et al. 2015 ; Pohoryles et al. 2018 , 2023 ) is employed to mitigate shear failures in beams and columns. The number of uniaxial FRP fabrics has been designed according to CNR-DT 200/ 2004 guidelines. The plastic shear is used as shear demand for the design to convert shear-sensitive elements in ductile elements. The design results in a maximum number of uniaxial carbon-FRP plies (with high elastic modulus, 230 GPa, and an equivalent thickness equal to 0.166 mm) ranging from one to three. For columns, a continuous wrapping along the height is assumed. This FRP wrapping leads to an improved tensile strength ranging from 1.9 to 2.7 times the V w (as defined in Sect. 2) for columns, resulting in capacity-to-demand ratios ranging from 1.1 to 1.5. On the contrary, beams ending portions are wrapped (until a maximum extension of 50% of the beam length), namely only where the shear demand (assumed as plastic shear at the beams ends) overcomes the as-built shear strength. One FRP ply is always sufficient for beams, leading to capacities at least 1.8 times higher than the plastic shear load.

Pre-stressed steel strips, applied as “external stirrups”, are used to effectively solve tensile shear failures in beam-column joints. The number of pre-stressed strips is designed to prevent diagonal cracking of the joint or support the maximum tensile force coming from the converging beams (Verderame et al. 2022 ), in tune with CEN 2005. As a result, a maximum of about twenty pre-stressed 0.9 × 19 mm 2 strips of stainless steel (420 MPa yielding strength) is obtained. A maximum of three holes per beam is necessary for this intervention.

Nevertheless, the adopted techniques do not allow solving the compressive failures, as defined by CEN 2005 and NTC 2018. This failure can be an issue not for beams or columns (always characterised by a tensile shear failure in the investigated buildings), but for joints (especially when characterised by high axial load levels). In other words, if a JF(C) failure occurs before the first DF (according to CEN 2005 and NTC 2018), the building capacity is limited to the first JF(C), instead than the first DF (De Risi et al. 2023a ). Therefore, this intervention is intended to be applied to all shear-sensitive elements that fail during the pushover analysis up to the first DF (see Fig.  9 ) or the first JF(C), if any. It is also worth noting that, if JF(C) occurred for gravity loads only (as when f cm is very low), such strengthening strategy would have no sense, and thus, it should be replaced with a more comprehensive and likely “heavier” retrofitting approach.

Moving towards the next-generation Eurocode, shear strength of joints strengthened with pre-stressed steel strips has been assessed based on prEN 1998-1-1:2024 and Fardis ( 2021 ), assuming that steel strips act as exterior stirrups (as for CEN 2005 ). Joint shear strength of (thus reinforced) joints, evaluated according to prEN 1998-1-1:2024, increases with respect to unreinforced joints, and overcomes the joint maximum shear demand for the analysed buildings. Therefore, joint shear failures result completely solved after retrofitting according to the incoming code.

Contrary to European approaches, according to ASCE/SEI, the joint transverse reinforcement is conforming if, in the joint region, the spacing of the hoops does not exceed half of the height of the column’s cross-section. Therefore, it is assumed that joints strengthening - designed as described above - is able to transform non-conforming joints into conforming joints (Cosgun et al. 2019 ). A conforming joint has higher capacity (i.e., higher γ′ coefficients in Table  1 ) than a relevant non-conforming joint. This leads to a higher displacement capacity (Fig.  13 ), and, in tune, higher Sa c (T eff ) (Fig.  14 ), if compared with the ante-operam condition. Therefore, the post-operam capacity according to ASCE/SEI model is the minimum between the Sa c (T eff ) corresponding to the first DF and that corresponding to the occurrence of a conforming joint failure.

Capacity curves up to the first DF with relevant brittle failures at SD LS in the post-operam condition (second row) compared to the ante-operam condition (first row)

Post-operam capacity in terms of S a, C (T eff ) ( a ) and comparison with ante-operam capacity ( b )

Lastly, note also that any possible increment in column displacement capacity due to FRP wrapping is herein neglected, since it does not significantly affect the analysis of the effectiveness of the selected strengthening techniques.

6.2 Post-operam capacity assessment

Figure  13 shows the S d capacity increments, moving from ante - to post-operam condition.

Among current codes, the highest displacement capacity increments are observed for European/Italian codes (ranging from + 83% to + 94%), especially in Y direction (where no JF(C) occurs). About ASCE/SEI, displacement capacity increment reaches + 60% (for Ns = 4 in X direction); whereas, for Ns = 2, the displacement capacity increment is null or very low since, already in as-built condition, shear failures, if any, only occur very close to the first DF.

Figure  14 a shows that, for the 2-story building, after retrofit, S a, C (T eff ) is limited by the first DF failure for all considered codes and in both directions, thus reaching the same value for all codes (Fig.  14 b). The high ante-operam S a, C (T eff ) according to ASCE/SEI model results in a small capacity increment in Y direction (about + 20%). Vice-versa, in X direction, the ASCE/SEI-based S a, C (T eff ) remains unchanged between the ante- and post-operam conditions, since in both conditions, the first DF defines the capacity. On the contrary, the two current Italian and European codes provide about the same S a, C (T eff ) increment for the 2-storey building (about + 85% and + 95% in X and Y direction, respectively), moving from the first JF(T) to the first DF.

For the 4-storey building, the post-operam capacity is limited by the occurrence of a JF(C) in X direction according to NTC 2018 and EC8 2005, and of conforming JFs according to ASCE/SEI. The S a, C (T eff ) in X direction according to ASCE/SEI is due to the failure of an exterior joint. On the contrary, according to European standards, the post-operam capacity is associated with the JF(C) of an interior joint. In this case, the highest capacity increment (among current codes) is reached with EC8 2005 capacity model (about + 90%), which allows moving from the first SF to the first JF(C). In Y direction, instead, both the current Italian and European codes provide the same S a, C (T eff ) (corresponding to the first DF), higher than the ASCE/SEI-based S a, C (T eff ). In this latter case, indeed, exterior (conforming) joints fail before any element reaches its ductile capacity, limiting the corresponding capacity increment (about + 65%).

Lastly, Fig.  14 also shows how the current Eurocode leads to an as-built seismic capacity quite similar to that of the future Eurocode with \(\:{{\gamma\:}}_{\text{R}\text{d}}\) =1.72, in tune with observed failure evolution.

Even with the same retrofitting strategy, the resulting the post-operam capacity can be very different if second-generation Eurocodes is used (Fig.  14 ). For the 2-story building the capacity always corresponds to the first DF, thus leading to the same S a, c (T eff ) of the previously considered current codes. Vice-versa, for the 4-storey building, the capacity provided by the current Eurocode 8 is limited by the JF(C), contrary to what happens by using the second-generation Eurocodes in its current draft. This outcome leads to higher post-operam S a, c (T eff ) values if the incoming code is used.

7 Conclusions

Pre-code RC buildings are particularly vulnerable to shear failures during seismic events, thus emphasizing the paramount role of a reliable estimation of shear capacity of RC elements. The scientific literature proposes different capacity models, and technical codes worldwide have significant differences as well. This study provides an overview and a comparison of shear strength models adopted by Italian, European and American codes.

About the current shear strength models for low-standard beam/column elements, a first parametric comparison found that:

the current European standard is generally penalizing, when compared to American code;

the model adopted by Italian code provides intermediate resistances between the European and American standards; it is derived from European one but modifies the latter for low ductility demand.

Additionally, significant differences exist in the current capacity models used to assess unreinforced beam-column joints in the European and American contexts:

the joint resistance significantly varies with the axial load according to European and Italian models; on the contrary, in the American model, the axial load has not any role on unreinforced joint strength, which only depends on the joint geometrical configuration and the number of converging beams;

the current European and Italian codes generally provide a lower joint resistance compared to American standard, for interior joints; for exterior joints this comparison strongly depends on axial load ratio;

the current European and Italian models have same theoretical approach, but their hierarchy in terms of strength is strongly influenced by the joint configuration, concrete compressive strength, and axial load ratio.

All the models have been applied and compared to each other in terms of seismic capacity assessment of case-study pre-code RC buildings. They were designed for gravity loads only, with 2 or 4 stories. The assessment, at Severe Damage Limit State, based on pushover analyses, revealed that:

the seismic capacity is severely limited by joint failures in a force-based approach for almost all buildings/codes;

the seismic capacity in terms of elastic spectral acceleration based on the American standard overcomes that of European models (at least + 65%), while the Italian code generally falls between the other two current codes (but closely to the EC8 2005 outcome);

an explicit modelling of the beam-column joint behaviour and a displacement-capacity safety check approach is explicitly allowed by American standard only. It leads to very less conservative results for the investigated buildings than force-based safety checks.

The incoming second generation of Eurocodes has been also investigated and applied herein based on their current available drafts and background literature. With respect to the current European standards:

a generally less conservative safety check for beams/columns shear strength is obtained;

a similar outcome is confirmed for beam-column intersections, mainly due to the absence of an explicit diagonal compressive safety check for unreinforced joints.

The seismic capacity of the case-study buildings was also reassessed after implementing a retrofitting strategy that addresses all tensile-only brittle failures. It was found that:

post-operam capacity is due to the occurrence of the first ductile failure for the shortest building, whichever the code;

for the tallest building, post-operam capacity is due to the first joint compressive failure as for the current European and Italian standards (which has always to be checked also for reinforced joints), or to a conforming joint failure as for American standard;

the incoming shear strength model (next-generation Eurocode) for beam/column elements was found to be significantly less penalizing than the current one, thus also requiring fewer retrofitting efforts;

the current draft of the next-generation Eurocode, unlike the current European code, leads to higher post-operam capacities. This is particularly due to the new model adopted for beam-column joints, according to which joint strength can increase with transverse reinforcement.

Buildings seismic assessment herein has been performed on 3D models neglecting infills, due to the lack of practical guidance in most codes on modelling them and the code-based framework of this study.

Nevertheless, further works, if aimed at more comprehensive fragility analysis of as-built and retrofitted buildings, should consider the paramount presence of infill panels. Lastly, it is worth noting that some further details (e.g., safety factors) have to be still defined and could somehow modify the current available versions of the second-generation Eurocodes and, thus, in this eventuality, some discussed results could be affected.

Alath S, Kunnath SK (1995) Modeling inelastic shear deformation in RC beam-column joints engineering mechanics: Proceedings of 10th Conference: University of Colorado at Boulder, Boulder, CO, May 21–24, 1995. Vol. 2. New York, NY: ASCE: 822–825

ASCE SEI/41 (2017) Seismic evaluation and retrofit of existing buildings. American Society of Civil Engineers

Baradaran Shoraka M (2013) Collapse assessment of concrete buildings: an application to non-ductile reinforced concrete moment frames (Doctoral dissertation, University of British Columbia)

Bentz EC, Vecchio FJ, Collins MP (2006) Simplified modified compression field theory for calculating shear strength of reinforced concrete elements. ACI Struct J 103(4):614

Google Scholar  

Biskinis D, Fardis MN (2010) Flexure-controlled ultimate deformations of members with continuous or lap-spliced bars. Struct Concr 11(2):93–108

Article   Google Scholar  

Biskinis D, Fardis MN (2020) Cyclic shear resistance model for Eurocode 8 consistent with the second-generation Eurocode 2. Bull Earthq Eng 18(6):2891–2915

Biskinis DE, Roupakias GK, Fardis MN (2004) Degradation of shear strength of reinforced concrete members with inelastic cyclic displacements. Struct J 101(6):773–783

Canadian Standard Association (2006) CSA standard A23. 3–04–concrete design handbook

Celik OC, Ellingwood BR (2008) Modeling beam-column joints in fragility assessment of gravity load designed reinforced concrete frames. J Earthq Eng 12:357–381

Cirak Karakas C, Palanci M, Senel SM (2022) Fragility based evaluation of different code-based assessment approaches for the performance estimation of existing buildings. Bull Earthq Eng 20(3):1685–1716

Circolare 21/01/2019, n. 7 C.S.LL.PP. Istruzioni per l'applicazione dell'«Aggiornamento delle "Norme tecniche per le costruzioni"» di cui al decreto ministeriale 17 gennaio 2018. GU n.35 del 11-2-2019 - Suppl. Ordinario n. 5. (in Italian)

CNR-DT 200/2004 Istruzioni per la Progettazione, l’Esecuzione ed il Controllo di Interventi di Consolidamento Statico mediante l’utilizzo di Compositi Fibrorinforzati. (in Italian)

Collins MP (1978) Towards a Rational Theory for RC Members in Shear, Journal of the Structural Division, ASCE, V. 104, No. 4, Apr. pp. 649–666

Cook D, Sen A, Liel A, Basnet T, Creagh A, Koodiani HK, Smith R (2023) ASCE/SEI 41 assessment of reinforced concrete buildings: Benchmarking nonlinear dynamic procedures with empirical damage observations. Earthq Spectra 39(3):1721–1754

Cosgun C, Cömert M, Demir C, İlki A (2019) Seismic retrofit of joints of a full-scale 3D reinforced concrete frame with FRP composites. J Compos Constr 23(2):04019004

Article   CAS   Google Scholar  

De Luca F, Verderame GM (2013) A practice-oriented approach for the assessment of brittle failures in existing reinforced concrete elements. Eng Struct 48:373–388

D.M. 2018 Aggiornamento delle Norme Tecniche per le Costruzioni - D.M. 17/1/18. (in Italian)

De Risi MT, Ricci P, Verderame GM (2017) Modelling exterior unreinforced beam-column joints in seismic analysis of non‐ductile RC frames. Earthq Eng Struct Dynamics 46(6):899–923

De Risi MT, Gaudio D, C., Verderame GM (2020a) A component-level methodology to evaluate the seismic repair costs of infills and services for Italian RC buildings. Bull Earthq Eng 18(14):6533–6570

De Risi MT, Scala SA, Del Gaudio C, Verderame GM (2023a) Seismic retrofit of Italian pre-‘70 case-study RC buildings by solving shear failures: code-compliant assessment and economic effort. Int J Disaster Risk Reduct 97:104007

De Risi MT, Di Domenico M, Manfredi V, Terrenzi M, Camata G, Mollaioli F, Verderame GM (2023b) Modelling and seismic response analysis of Italian pre-code and low-code reinforced concrete buildings. Part I: bare frames. J Earthquake Eng 27(6):1482–1513

Del Gaudio C, De Risi MT, Verderame GM (2021) Seismic loss prediction for Infilled RC buildings via simplified analytical method. J Earthquake Eng, 1–34

Del Vecchio C, Di Ludovico M, Prota A, Manfredi G (2015) Analytical model and design approach for FRP strengthening of non-conforming RC corner beam–column joints. Eng Struct 87:8–20

Del Vecchio C, De Risi MT, Gaudio D, Ricci C, Di Ludovico P, M., Kang JD (2023) E-Defense 2015 ten-story building: beam–column joint assessment according to different code-based design. Bull Earthq Eng 21(15):6667–6698

Dhanvijay V, Telang D, Nair V (2015) Comparative study of different codes in seismic assessment. Comp Study Different Codes Seismic Assess 2(4):1371–1381

EN 1992-1-1:2004-12 (2004) Eurocode 2: design of concrete structures– part 1–1: General rules and rules for buildings. CEN, Brussels

EN 1998-1. CEN (2004) Eurocode 8 - design of structures for Earthquake Resistance, Part 1, General rules, seismic actions and rules for buildings. European Committee for Standardization

EN 1998-3. CEN (2005) Eurocode 8 - design of structures for Earthquake Resistance, Part 3, Assessment and Retrofitting of buildings. European Committee for Standardization

Fajfar P (2000) A nonlinear analysis method for performance-based seismic design. Earthq Spectra 16(3):573–592

Fardis MN (2021) Shear strength model for RC joints, consistent with the shear design rules for prismatic members in the second-generation Eurocodes. Bull Earthq Eng 19(2):889–917

fib Model Code (2010) Bulletin 65/66. Federation Internationale Du Beton, Lausanne: 2012, doi: 978-2-88394-105-2

FprEN 1998-1-1:2024 Eurocode 8: Design of structures for earthquake resistance—part 1–1: General rules and seismic action. Brussels: CEN; 2024

FprEN 1998-1-1:2022 Eurocode 8: Design of structures for earthquake resistance—part 1–1: General rules and seismic action. Brussels: CEN; 2022

FprEN 1992-1-1:2023 Eurocode 2: design of concrete structures—part 1–1: general rules, rules for buildings, bridges and civil engineering structures. Brussels: CEN; 2023

Franchin P, Noto F (2023) Reliability-based partial factors for seismic design and assessment consistent with second‐generation Eurocode 8. Earthq Eng Struct Dynamics 52(13):4026–4047

Fujii S, Morita S (1991) CComparison between interior and exterior RC beam–column joint behavior. Design of beam–column joints for seismic resistance (SP123), American Concrete Institute: Detroit, MI, pp 145–165

Hakuto S, Park R, Tanaka H (2000) Seismic load tests on interior and exterior beam-column joints with substandard reinforcing details. Struct J 97(1):11–25

Hassan WM (2011) Analytical and experimental assessment of seismic vulnerability of beam-column joints without transverse reinforcement in concrete buildings. PhD dissertation. California, USA: University of California, Berkeley

Hassan WM, Elmorsy M (2022a) Cyclic nonlinear modeling parameters for unconfined beam-column joints. ACI Struct J 119(1):89–104

Hassan WM, Elmorsy M (2022b) Probabilistic Beam–Column Joint Model for seismic analysis of concrete frames. J Struct Eng 148(4):04022011

ISTAT (2011) Istituto Nazionale di Statistica (National Institute of Statistics). 15° Censimento generale della popolazione e delle abitazioni. Dati sulle caratteristiche strutturale della popolazione, delle abitazioni e variabili (2011). (in Italian)

Jeon JS, Shafieezadeh A, DesRoches R (2014) Statistical models for shear strength of RC Beam-column joints using machine‐learning techniques. Earthq Eng Struct Dynamics 43(14):2075–2095

Kim J, LaFave JM (2012) A simplified approach to joint shear behavior prediction of RC Beam-column connections. Earthq Spectra 28(3):1071–1096

Kowalsky MJ, Priestley MN (2000) Improved analytical model for shear strength of circular reinforced concrete columns in seismic regions. Struct J 97(3):388–396

Li B, Siu-Shu Lam E, Wu B, Wang YY (2015) Effect of high Axial load on seismic behavior of Reinforced concrete Beam-Column joints with and without strengthening. ACI Struct J 112(6)

Lupoi G, Calvi GM, Lupoi A, Pinto PE (2004) Comparison of different approaches for seismic assessment of existing buildings. J Earthquake Eng 8(spec01):121–160

Maranhão H, Varum H, Melo J, Correia AA (2024) Comparative analysis of the impact of design and detailing provisions for RC moment resisting frames under the first-and second-generation of Eurocode 8. Eng Struct 306:117809

Marcantonio PR, Ožbolt J, Petrangeli M (2015) Rational approach to prediction of shear capacity of RC Beam-column elements. J Struct Eng 141(2):04014104

Masi A, Chiauzzi L, Santarsiero G, Manfredi V, Biondi S, Spacone E, Verderame GM (2019) Seismic response of RC buildings during the M w 6.0 August 24, 2016 Central Italy earthquake: the Amatrice case study. Bull Earthq Eng 17:5631–5654

McKenna F (2011) OpenSees: a framework for earthquake engineering simulation. Comput Sci Eng 13(4):58–66

Mörsch E (1909) Concrete-steel construction:(Der Eisenbetonbau). Engineering news publishing Company

NZS 3101 (2006) Concrete structures standards, vols. 1 and 2. Standards Association of New Zealand

Pantelides CP, Clyde C, Reaveley LD (2002) Performance-based evaluation of reinforced concrete building exterior jointsfor. Seismic Excitation Earthq Spectra 18(3):449–480

Park S, Mosalam KM (2012) Parameters for shear strength prediction of exterior beam–column joints without transverse reinforcement. Eng Struct 36:198–209

Paulay T, Priestley MJN (1992) Seismic design of reinforced concrete and masonry buildings. Wiley Interscience

Book   Google Scholar  

Pohoryles DA, Melo J, Rossetto T, D’Ayala D, Varum H (2018) Experimental comparison of novel CFRP retrofit schemes for realistic full-scale RC beam–column joints. J Compos Constr 22(5):04018027

Pohoryles DA, Minas S, Melo J, Bournas DA, Varum H, Rossetto T (2023) A novel FRP retrofit solution for improved local and global seismic performance of RC buildings: development of fragility curves and comparative cost-benefit analyses. J Earthquake Eng 1–19

prEN 1998-3:2023 Eurocode 8: Design of structures for earthquake resistance - Part 3: Assessment and retrofitting of buildings and bridges. Brussels: CEN; 2023

Priestley MJN (1997) Displacement-based seismic assessment of reinforced concrete buildings. J Earthquake Eng 1(01):157–192. https://doi.org/10.1142/S1363246997000088

Priestley MJN, Verma R, Xiao Y (1994) Seismic shear strength of reinforced concrete columns. J Struct Eng ASCE 120(8):2311–2329

R.D. Regio Decreto Legge n. 2229 Del 16/11/1939. Norme per la esecuzione delle opere in conglomerate cementizio semplice Od Armato. G.U. n. 92 Del 18/04/1940 (in Italian).

Ritter W (1899) Die bauweise hennebique (hennebiques construction method)

Sen A, Cook D, Liel A, Basnet T, Creagh A, Khodadadi Koodiani H, Smith R (2023) ASCE/SEI 41 assessment of reinforced concrete buildings: Benchmarking linear procedures and FEMA P-2018 with empirical damage observations. Earthq Spectra 39(3):1658–1682

Sezen H, Moehle JP (2004) Shear strength model for lightly reinforced concrete columns. J Struct Eng 130(11):1692–1703

TBEC-2018 (2018) Türkiye building earthquake code, Republic of Türkiye Ministry of Interior Disaster and Emergency Management Authority, Ankara

Tsonos AG (2007) Cyclic load behavior of reinforced concrete beam–column subassemblages of modern structures. ACI Struct J 104(4):468–478

Vecchio FJ, Collins MP (1986) The modified compression-field theory for reinforced concrete elements subjected to shear. ACI J 83(2):219–231

Verderame GM, Ricci P (2018) An empirical approach for nonlinear modelling and deformation capacity assessment of RC columns with plain bars. Eng Struct 176:539–554

Verderame GM, Polese M, Mariniello C, Manfredi G (2010) A simulated design procedure for the assessment of seismic capacity of existing reinforced concrete buildings. Adv Eng Softw 41(2):323–335

Verderame GM, Ricci P, De Luca F, Del Gaudio C, De Risi MT (2014) Damage scenarios for RC buildings during the 2012 Emilia (Italy) earthquake. Soil Dyn Earthq Eng 66:385–400

Verderame GM, Ricci P, De Risi MT, Del Gaudio C (2022) Experimental response of unreinforced exterior RC joints strengthened with prestressed steel strips. Eng Struct 251:113358

Vidic T, Fajfar P, Fischinger M (1994) Consistent inelastic design spectra: strength and displacement. Earthq Eng Struct Dynamics 23(5):507–521

Vollum RL, Newman JB (1999) Strut and tie models for analysis/design of external beam–column joints. Mag Concr Res 51(6):415–425

Download references

Acknowledgements

This study was developed within the activities of the ReLUIS-DPC 2024–2026 research programs, funded by the Presidenza del Consiglio dei Ministri—Dipartimento della Protezione Civile (DPC).

This work was supported by Reluis-DPC project, funded by Presidenza del Consiglio dei Ministri—Dipartimento della Protezione Civile (DPC).

Open access funding provided by Università degli Studi di Napoli Federico II within the CRUI-CARE Agreement.

Author information

Authors and affiliations.

Department of Structures for Engineering and Architecture, University of Naples Federico II, via Claudio, 21, Napoli, 80125, Italy

Santa Anna Scala, Maria Teresa De Risi & Gerardo Mario Verderame

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by all authors. The first draft of the manuscript was written by all authors and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Maria Teresa De Risi .

Ethics declarations

Conflict of interest.

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval

All principles of ethical and professional conduct have been followed.

Informed consent

Not applicable.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Scala, S.A., De Risi, M.T. & Verderame, G.M. Code-based brittle capacity models for seismic assessment of pre-code RC buildings: comparison and consequences on retrofit. Bull Earthquake Eng (2024). https://doi.org/10.1007/s10518-024-02016-6

Download citation

Received : 13 May 2024

Accepted : 04 September 2024

Published : 16 September 2024

DOI : https://doi.org/10.1007/s10518-024-02016-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Brittle capacity models
• Code-based assessment
• RC buildings
• Shear failures
• Shear strengthening
• Retrofitting design
• Second-generation eurocodes
• Find a journal
• Publish with us
• Track your research