Multi-purpose cyclone shelters in India

The Panchayati Raj and Drinking Water Department  looks into the repairing of cyclone and flood shelters across Odisha (Image: Odisha TV)
The Panchayati Raj and Drinking Water Department looks into the repairing of cyclone and flood shelters across Odisha (Image: Odisha TV)

An ‘Extremely Severe Cyclone Fani’ made landfall during morning hours on May 3, 2019, near Puri in Odisha. The landfall location was forecast accurately much in advance by India Meteorological Department. In anticipation of the strike, the state government carried out an extensive evacuation involving more than 1.5 million people.

Fani affected over 16.5 million people in more than 18,000 villages during which 64 people lost their lives, out of which 39 were from Puri district. The overall effort of the state government received much appreciation in minimizing loss of lives as compared to 1999 super cyclone.

Among various measures envisaged to deal with cyclone hazard, temporary evacuation to a safe shelter is the most widely practised one. It involves identification of a particular area as vulnerable or endangered based on the tropical cyclones' strike location in conjunction with its intensity forecast, resulting in storm surge etc.

The paper Role of multi-purpose cyclone shelters in India: Last mile or neighbourhood evacuation by Biswanath Dash and Ajinder Walia in Tropical Cyclone Research and Review, Volume 9, Issue 4, December 2020 deals with India’s strategy of having Multi-Purpose Cyclone Shelters (MPCS) along the coastline for tropical cyclone risk mitigation. These shelters are meant to provide refuge to vulnerable populations at the time of a cyclonic storm and otherwise to be used as a school, community centres etc.

This paper aims to examine the exact role which these MPCS seek to perform; as a safe shelter for people living in a tropical cyclone threatened region or meant for those who fail to evacuate due to various reasons.

As a part of the study, nine MPCS were visited during the fieldwork in ‘Puri’ district and except one, all of them were found to have been used to various extent during ‘Fani’. These MPCS were constructed after 1999 super cyclone with few being made operational as recent as 2015.

Barring one, all have a similar floor plan, comprising four big rooms and two smaller ones, and separate toilets on their first floor. The vertical elevation requirement of MPCS does not allow rooms on its ground floor. In contrast to the official estimate, accommodation capacities of each were estimated variedly by the local residents; from 200 to maximum 1000. Similarly, the total number of evacuees who stayed in each of them during ‘Fani’ was reported by locals to be between 25 to 800.

Delineation of an evacuation zone is complex

It brings to the fore uncertainties inherent in meteorological forecasts as well as in inundation models. In addition, there are various other considerations as well such as cost of evacuation, inconveniences to evacuees etc. In India, a vulnerable zone is evacuated using the strategy of moving endangered populations to a specially constructed public building, known as ‘cyclone shelters’.

They are generally located within or near one's neighbourhood though at times, they can be at a considerable distance too. This idea of a structurally engineered public cyclone shelter was originally formulated in Bangladesh during the 1960s and was replicated in India in 1977 in the aftermath of a deadly cyclone in Andhra Pradesh. It was soon realized that structural conditions of shelter deteriorate rapidly under saline environmental conditions. Lack of regular use renders them unsuitable at the time of need. This led to the conception of MPCS or their regular use as community centres, schools etc. to ensure periodic maintenance.

MPCS are widely seen to be far more effective as disaster risk reduction measures as compared to strategies such as cyclone track or surge prediction technologies and large programs are executed to ensure availability of MPCS for the most vulnerable population.

This paper seeks to examine an inherent ambiguity in this approach of having MPCS in the vulnerable coastal region of India from the perspective of their primary goal; a) are they a means to save lives of vulnerable people who have failed to evacuate for various reasons or b) do they provide an alternate means for coastal residents to take shelter as and when there is a cyclonic strike? Ambiguities over the purpose of MPCS have brought about a notion that the total number of cyclone shelters should match with the accommodation space requirement of a particular area.

This study scrutinizes feasibility of the second aspect of using MPCS for neighbourhood evacuation from the perspective of a) evacuee safety b) accommodating capacities, and c) sustainability and raises following questions; to what extent such MPCS often located close to the seashore, can provide safety in the face of a severe tidal surge? b) accommodation capacity of such shelters for a region will remain limited, and hence where do remaining populations take refuge and c) how realistic is it to expect that such public buildings will be adequately maintained to remain useful as a safe shelter under cyclonic weather?


The study approach is qualitative and follows ‘multiple case study’ method. There were four ‘tropical cyclones’ of ‘Very Severe’ or above category which struck the east coast of India during 2013–2019, all of which have been considered. These are; a) Very Severe Cyclone ‘Phailin’ 2013; b) Very Severe Cyclone ‘Hudhud’ 2014; c) Very Severe Cyclone ‘Titli’ 2018 and d) Extremely Severe Cyclone ‘Fani’ 2019.

Evacuation for cyclone within a vulnerable zone

There are variations in strategies employed across countries and sometimes even within a country on how to carry out evacuation from a vulnerable area at the time of a cyclone. Prevalent approaches are chiefly of two types; persuading, facilitating and at times forcing residents living within the threatened area to a) move out of the danger zone or b) move into a designate specially designed public shelter located within the danger zone. The former wherein an area is completely evacuated is referred to as ‘horizontal or preventive evacuation’ whereas the latter is a form of vertical evacuation in which shelters are specially designed as elevated buildings that allow tidal water to pass through its base.

The two strategies are not mutually exclusive and can be employed together. In addition, people take shelter in one's own or neighbour's house-an approach known as ‘shelter-in-place’. There are times when warned population move into public buildings such as schools, community centres etc. which may or may not be certified as safe from a sheltering point of view. There are other forms of vertical evacuation as well such as climbing up trees, or to any other high ground, which are not recommended by authorities.

‘Horizontal evacuation’ is the preferred choice in developed countries such as the USA but it is rarely followed in countries such as India. This preference to move out of a danger zone is to avoid the risk of structural failure to designate cyclone shelters with evacuees inside.


A number of reasons are identified across cases for the non-use of available MPCS. They range from concerns over safety from a cyclone's impact point of view to inconvenience due to lack of space, lack of basic facilities; availability of alternatives in the form of other public buildings, or the use of one's house or that of others deemed more appropriate.


The general finding from this study suggests that MPCS within the neighbourhood are considered absolutely safe for cyclone evacuation if only they are in good conditions. Lack of clarity in policy and plans over the exact role of MPCS has contributed to a growing perception that closer the proximity to seashore, greater should be the emphasis on providing matching shelter accommodation keeping in consideration the size of the population.

The gap in this conception is that the requirement of additional shelter capacities are conflated with the goal to accommodate more such number of people who have failed to evacuate. The existing practice of promoting neighbourhood evacuation also raises another important question; MPCS which are structurally engineered to ensure adequate safety of evacuee, are they any different from other public buildings? If not, how do we account the safety of MPCS and the corresponding investment that is involved in their construction?

A related aspect concerns evacuee safety while being inside a MPCS during a cyclonic impact. Experiences of respondents across the four study areas who took shelter in MPCS particularly those located in the most impacted region point towards possible damage to these buildings and the risk of tidal inundation and/or building collapse.

So far as the maintenance challenges of MPCS are concerned, this study point to a basic flaw in conception. The idea of a MPCS as a school or other utility centre is fundamentally different from that of a house or hotel in terms of its public/private ownership and in its basic function. Evacuees in diverse settings of the four cases have similar complaint about their experience in designate shelters; leaking roofs and windows resulting accumulation of rainwater inside the shelter, poor sanitation etc.

MPCS were conceived at a time when a large part of coastal India was inaccessible, without much road infrastructure and concrete housing units. In such conditions, a structurally engineered building to withstand the impact of a storm was a good idea to save human lives. This policy however has not evolved, thus overlooking its usefulness under a given set of conditions.

From the time, ‘cyclone shelters’ were conceived, some of its basic limitations including its ability to withstand tidal impact, accommodation capacities etc. were identified, and yet other alternatives have not been pursued vigorously because of its perceived benefits in terms of saving human lives.


To conclude, this study makes three suggestions. First, the approach of using MPCS for general evacuation in areas which are vulnerable from tidal surge should be reviewed. They should be considered for those residents who fail to evacuate due to unanticipated reasons. Second, in spite of a large population in coastal India, horizontal or preventive evacuation needs to be considered in the most vulnerable region and this can be done in conjunction with an improvement in meteorological forecast services such as intensity, landfall location and tidal inundation, so those specific areas are more clearly identified, limiting unnecessary evacuation.

Finally, to ensure the maintenances of MPCS in saline conditions over a period of time requires new solution. MPCS’ use as public buildings such as school, community centres etc. are useful in its utility beyond cyclone period but it requires to be viable too in providing adequate safety during a cyclonic storm.