Resilience of Ports Against Natural Hazards
By Deepak Pant, Ph.D., P.Eng., Greg Brunelle, MS, MA, Youngsuk Kim, Ph.D., Jaskanwal P. S. Chhabra, Ph.D., & Shabaz Patel, MS
Introduction
More than 90% of the world’s trade moves by sea, making the continuous operation of maritime ports a critical piece of the global economy. Ports also serve as a focal point for providing supplies and humanitarian support to communities after natural disasters. Unfortunately, the vast majority of the world’s ports are located in areas at high risk for natural hazards such as earthquakes, floods and windstorms. Moreover, climate change and the corresponding increase of sea-levels are leading to flooding at some of the major ports around the world. Economic losses resulting from port disruptions will be significant, and the functional recovery following a disaster will be a very time-consuming process. For example, the 1995 Kobe earthquake caused significant damage to the Port of Kobe and led to $5.5 billion in direct economic losses from damaged components (Burden et al. 2001) in addition to indirect economic losses due to the lack of functionality for an extended period of time. In this article, we will analyze the factors that contribute to more resilient ports following natural disasters.
The resilience of infrastructure can be defined as “the ability to suffer less damage and recover more quickly from adverse events” (Fisher et al. 2017). In order to evaluate the resilience of ports against natural hazards, their damage and recovery must be properly modeled. For this, it is important to understand the physical vulnerabilities of their components, typical recovery durations, and the factors controlling recovery. It is also crucial to understand that different perils impact ports in different ways and controlling factors governing recovery could be different for different perils.
This article provides an overview of key components of ports and their operations, a summary of the review of historical damage and recovery of ports from natural hazards, and the identification of key factors that control their recoveries in order to understand port resilience more deeply.
Overview of Major Components of Ports and Their Operations
Ports are waterfront structures which can spread over a wide area. There are six main components of a port: (1) wharves and other waterfront structures such as breakwaters and bulkheads (2) cranes, (3) cargo handling equipment such as container handlers, (4) fuel facilities, (5) warehouses, and (6) office/management buildings. A schematic of a port with different components is shown in Fig. 1. Ports are generally divided into terminals where different types of cargo can be handled. Wharves are the key components of a port where ships dock and cranes unload containers. Areas of wharves where ships dock are called berths, and the number of berths per wharf varies by the size of the ships docked there. Once cargo is removed from ships, it is taken to container yards where it is stored temporarily and picked up by trucks, trains or barges. Unlike container yards, warehouses are typically used for long-term storage. Fuel facilities can be located within the port or outside the port and are needed for fueling of ships.
Review of Historical Damage and Recovery
We carried out an in-depth review of historical damage and recovery durations of ports around the world due to earthquakes, windstorms, and floods. A summary of the review highlighting some of the key events is discussed in this section.
Earthquakes
Ports have suffered various degrees of damages in past earthquakes. While all the components of ports discussed in the previous section are vulnerable to earthquakes, wharves and cranes are the most vulnerable components in terms of port functionality. There are three main causes of damage to ports during earthquakes: (1) liquefaction, (2) ground deformation not affiliated with liquefaction, and (3) ground shaking (Pachakis and Kiremidjian 2004). While cranes are generally considered safe against seismic forces (Kosbab 2010), they can be damaged by shaking as well as due to damage of wharves, which are the most vulnerable components of a port.
Despite this vulnerability, we found in our research that U.S. ports performed well in past earthquakes because of their inherent redundancy and because they have not yet been subjected to damaging ground motions.The 1989 Loma Prieta earthquake affected the port of Oakland in that one container terminal suffered moderate damage. Nonetheless, the cargo from the damaged terminal was diverted to other terminals at the port, resulting in zero downtime (Pachakis and Kiremidjian 2004). The 1994 Northridge earthquake impacted the port of Los Angeles similarly; one container terminal suffered some minor damage, and overall no reports of downtime at the port could be found.
On the other hand, ports in Japan have experienced several damaging earthquakes. The 1995 Kobe earthquake impacted the Port of Kobe and caused widespread liquefaction and subsidence of wharves and caused moderate to severe damage to several cranes. While the repairs of damaged components were carried out for over two years, the full functional recovery was estimated to be achieved in less than two years. The 2011 Tohoku earthquake also caused widespread damage to ports. A combination of tsunami and ground shaking was the leading cause of the damage (Sugano et al. 2014). Wharves at the Port of Onahama were impacted by ground shaking and liquefaction which led to horizontal displacement of caissons, subsidence of backfill and some minor damage to cranes. The port of Onahama was able to resume full functionality in 6 months as measured by the volume of cargo handled by the port. During more recent events, such as the 2016 Kumamoto earthquake and the 2018 Hokkaido Eastern Iburi earthquake, the ports in Japan have performed well with some minor liquefaction and settlement of wharves, and the downtimes have been in the order of several days for the Kumamoto Port and the Tomakomai Port.
Port of Haiti, on the other hand, was devastated by the 2011 Haiti earthquake and both of the port’s wharves were completely destroyed and cranes were submerged in water and the recovery process was expected to take years (DesRoches et al. 2011).
In summary, earthquakes can cause significant damage to ports, and the recovery can take between several days to years depending on the degree of damage.
Windstorms
The main type of damage that windstorms can cause to ports are: (1) roof and wall damage of structures, particularly warehouses and (2) crane damage. Typically ports slow down their operations ahead of a large windstorm and are shut down during the duration of the storm. A comprehensive review of port slow-down and closure durations due to windstorms can be found in Verschuur et al. (2020). The focus herein is on the physical damage caused by windstorms and recovery durations following the storm. Damage to warehouses has occurred in almost all major hurricanes and typhoons around the world, but it typically does not cause functionality disruptions.
Cranes, on the other hand, are integral components of ports which are required for their functioning, and cranes have suffered damage in past windstorms. Although cranes are designed to resist wind loads, oftentimes there is a disconnect between crane design and crane-to-wharf connection design, which can cause failure of the crane-to-wharf connections referred to as tie-downs (McCarthy et al. 2007). Cranes are typically tied down to wharves ahead of strong windstorms to protect them from derailing and getting damaged.
In the US, there have been some instances of tie-down failures leading to crane damage. For example, the tie-down failure of one crane led to its severe damage at the port of New Orleans subjected to 2005 Hurricane Katrina (Cuffman et al. 2006).
One of the most devastating impacts of crane failures due to windstoms occurred at the Port of Busan in South Korea during the 2003 typhoon Maemi, which resulted in a sequential collapse of six cranes initiated by tie-down failures.
In Japan, some instances of crane collapses are reported, but they are mainly caused by not tying them down to the wharf ahead of the storm. For example, the 2018 Typhoon Jebi caused the collapse of 2 cranes at the Port of Amagasaki because the cranes were not tied down to the wharf ahead of the storm. Similarly, in 2006, one crane collapsed at the port of Niigata due to strong winds, and the recovery period for that crane was about 18 months.
Overall, the recovery data on cranes were less documented, but they are known to have long lead times. For example, the crane supplier Kalmar Global states that the lead times for purchasing a new crane are 12 to 24 months. In summary, windstorms can cause severe damage or collapse of cranes resulting in very long recovery times in the order of months to years.
Floods
The major impacts of floods on ports are: (1) shoaling of water channels, (2) flooding and debris accumulation on terminals, (3) floating of containers and damage to cargo handling vehicles, (4) damage to electrical equipment and (5) scouring and wash-out damage. Since most of the flooding instances at ports around the world have been caused by flooding due to tropical cyclones, the slow-down ahead of the storm and shutting down for the duration of the storm as discussed in the previous section applies here as well. The recovery durations discussed in this section refer to the durations after the flood waters have receded.
In the US, the ports have suffered various degrees of damages due to floods. The port of New Orleans was subjected to flooding due to the 2005 Hurricane Katrina, and almost one third of the port was destroyed due to floods. This was rather an extreme event from a view point of its regional impact, and the recovery was expected to take more than 6 months.
The port of New York and New Jersey was severely impacted by flooding from the 2012 Hurricane Sandy. The damage included debris accumulation on the terminal and the damage of electric equipment etc. The port was fully functional in 7 days.
In Japan, the 2018 Typhoon Jebi caused widespread flooding at a number of ports including the ports of Sakai Senboku, Hanan, Osaka, and Kobe. The damage at these ports included debris accumulation, flooding of cargo handling vehicles, and the damage to electrical equipment. At all of these ports, the limited functionality was restored within a day while full recovery took 8 days.
In summary, floods can cause significant damage to ports, but their recovery is rather quick and generally takes less than a week, but it could take longer depending on the severity of the flooding.
Key Controlling Factors
Based on the review of the historic damage and recovery, the following key factors controlling the resilience of ports to natural hazards were identified:
Repair of damaged components
Repair of damaged/affected components is one of the most important aspects governing the recovery of ports. Wharves, cranes, and electrical equipment were found to be the most vulnerable components of ports to natural hazards. There are other components of ports such as offices and warehouses which are also vulnerable to natural hazards, but they do not have much effect on the functionality of the ports because ports can function as long as ships can dock and cranes can unload the cargo.
Based on the review of historical events, for earthquakes, liquefaction of wharves and damage of cranes was the most dominant factor controlling recovery. For floods on the other hand, flooding of wharves and flooding of electrical equipment were detrimental in controlling the recovery. For windstorms, crane damage and, in particular, crane-to-wharf tie-down failure was the controlling factor.
Redundancy
Redundancy plays an important role in the resilience of ports against natural hazards. Ports typically have multiple wharves and cranes and thus damage/failure of some wharves and some cranes is not detrimental to the functionality of the entire port. An excellent example of the importance of redundancy is the Port of Oakland impacted by the 1989 Loma Prieta earthquake, where the port was fully functional following the event despite damage rendering one terminal inoperable because the demand from the damaged terminal was diverted to other non-damaged terminals.
Repair Vs functional recovery
Ports and businesses that depend on them are interested in the functional recovery durations rather than the actual repair durations. It is possible for ports to be functionally recovered while still being repaired, especially undergoing some cosmetic repairs following an event. It was revealed from the review of the historical damage and recovery data that while repairs and cleaning of debris can continue for a long period of time following the event, they should not impede the port functionality during that period of time. Cargo volume handled by a port could be a better indicator of the functional recovery of the port as opposed to the actual repair durations.
Scale of the disaster and interdependence on other infrastructure
It is also evident that not only the availability of parts, materials, and labours for repairs but also the availability of employees following disasters greatly depend on the size of the disaster and its impact on the larger community around the port. Since ports rely on other infrastructure components for their functionality, functioning of those infrastructure components is also important. For example, our review showed several instances where a port did not suffer any damage but was non-functional for several days after a disaster because of the flooded roads in the surrounding area or power outage (see for example ports impacted by 2017 Hurricane Harvey).
Pre-event slow-down and duration of hazard
Unlike earthquakes, which give little to no warning before they strike, windstorms and flooding can be predicted within a reasonable time before they strike. Typically ports slow down their operations ahead of a large windstorm and are shut down for the duration of the windstorm or flooding. This can result in several days of downtime before the actual physical recovery process can take place and thus must be considered in the resilience framework.
Conclusions
Evaluating the resilience of ports against natural hazards such as earthquakes, windstorms and floods is important. A systematic review of the damage and recovery durations from historical events as well as the identification of the key controlling factors discussed in this article serve as an important resource in developing predictive models. Our review highlights that the full recovery of ports can take from days to years depending on the type and the severity of the peril. Different components of ports are impacted by different perils in different manners. A number of external factors and dependency among infrastructure components were also found to be detrimental in the recovery process. Ports have redundancy built in them because they are typically composed of multiple wharves and cranes which must be considered in assessing their resilience. A holistic review considering all major hazards discussed in this article lays the foundation of developing tools and methods to assess port’s resilience against natural hazards.
