David M. Bush, Duke University, Durham, North Carolina
Richard D. Marshall, National Institute of Standards and Technology, Gaithersburg, Maryland
A number of lifeline systems in the U.S. Virgin Islands and Puerto Rico were damaged or seriously disrupted by Hurricane Hugo. Because Hugo was a relatively “dry” hurricane, damage to lifelines by stream flooding or erosion was minimal. Some coastal roadways were undercut by wave action or covered by sand deposits, but in most cases the cause of lifeline disruption was wind or wind-blown debris. Particularly hard hit were electrical distribution lines. In several cases, damage to electrical distribution systems contributed to or directly caused the failure of other lifeline systems. Descriptions of the most significant failures observed by the CND investigative team are presented in the following paragraphs.
The field study team found that the real extent of the destruction had been exaggerated. For example, the Team heard reports of massive landsliding and of a failed seawall, among other catastrophes, all of which turned out to be completely unfounded. However, substantial damage was done to the infrastructure, which adversely affected general living conditions. As reported in a special section on Hurricane Hugo in the September 24, 1989, edition of the San Juan Star, “collapse of the island's infrastructure left a large percentage of the population without power, water, garbage collection and other basic services.”
ELECTRICAL DISTRIBUTION SYSTEMS
Electrical distribution systems were the lifelines that suffered the greatest damage in Hugo. It was not possible to study this problem in detail in the Virgin Islands, but it is believed the damage in northeastern Puerto Rico, as well as the emergency response efforts, were typical. Figure 5-1 shows downed power lines directly east of San Juan, where the sustained winds probably did not exceed 70 knots (81 mph). Under these conditions, it is unexpected that such widespread failure
should occur. Fallen trees and other wind-blown debris collecting on the conductors contributed to line failures. However, the great majority of downed lines in Puerto Rico were in open areas and failed because the poles were carrying too many conductors. Failed poles included conventional treated-timber poles as well as prestressed concrete poles. Channeling of wind through hilly terrain was observed to have intensified damage in several places. Winds were channeled through some small hills with enough velocity to snap these large electric-supply poles. No trees in those areas fell on the electric lines; they were toppled by wind force alone. Less than 100 m to the east, in the shadow of the hills, a satellite dish remained, undamaged, atop a building.
At 0505 on Monday, September 18, as Hugo was beginning to attack the main island of Puerto Rico, Governor Rafael Hernandez Colon had the power system shut down as a means of preventing severe damage. This was reported in the San Juan Star as well as on CNN television. There were reports that this actually compounded the problem, because with the system off, breaks in electrical supply could not be monitored as they occurred. Therefore, in essence, the entire system had to be rebuilt and/or rechecked before service could be restored. This apparently led to much longer delays before service could be restored. As of September 24, most residents of the northeast coast were still without power. Consequently, danger of food- and waterborne infectious diseases increased, and residents were encouraged to boil water before drinking. This was not possible for much of the population, since they were without electricity.
According to a report in the September 24, 1989, edition of the San Juan Star, 47,500 homes and businesses were still without electricity, although 80 percent of San Juan had power. There was also only partial service on Culebra. It should be noted that repair efforts may have been hampered by the theft of copper wire by looters. In Humacao and Ceiba, it was reported that medicines requiring refrigeration had
spoiled and had to be thrown out. A San Juan Star article as late as September 28, 1989, reported that electrical service was still only 80 percent restored.
On October 8, 1989, the San Juan Star reported that residents of the city of Rio Grande, east of Loiza, stormed their city hall because they still lacked water and electricity. They were also upset because no FEMA center had been set up in Rio Grande, and residents had to travel to Loiza to apply for disaster relief. On October 22, 1989, a San Juan Star article reported that 25 percent of the electricity customers in Fajardo (9,000) were still without service. On October 29, 1989, a San Juan Star article reported that the damage to AEE poles and wires would cost $50 million to repair.
Recovery in some areas of Puerto Rico was rapid, and much of San Juan was back on line within 48 hours. In many areas, the repair effort involved the complete replacement of poles and conductors. Repair work in progress near the San Juan International Airport is shown in Figure 5-2. Significantly, most of the repair and replacement involved no change from or improvement to the original installation. Recovery in the Virgin Islands was not as rapid, but most electrical service on St. Croix and St. Thomas was restored within 3 months.
Damage to telephone systems was very similar to that experienced with electrical distribution systems since, in many instances, the poles were shared. Although several miles of trunk lines were down in Puerto Rico, many of the circuits
remained in service, and it was possible to call into areas that did not have electrical service for several days after Hugo. The level of damage to other communications facilities, such as microwave and radio towers, is not known. At least one microwave guyed tower north of Fajardo lost its upper section in sustained winds that probably were in the range of 85 to 95 knots (98 to 109 mph).
According to a San Juan Star article appearing on September 28, 1989, long-distance telephone service had been restored to Vieques on September 27. Service to Culebra was to be tested. Service to both islands had been cut off because of a destroyed communications antenna in El Yunque.
In the Virgin Islands, the telephone system was heavily damaged, and limited service to business establishments did not become available until December. On St. Croix, telephone service to many private residences had not been restored as late as March 1990.
At least two significant water supply failures occurred during the passage of Hugo. On St. Croix, the Virgin Islands Water and Power Authority (WAPA) power and water distillation facilities, located on the west side of Christiansted, were knocked out of operation. An aerial view of the facility looking to the east is shown in Figure 5-3. The fuel oil tank in the bottom center of the picture ruptured, and the containment, which was intended to function in such an emergency, also failed, probably before the tank ruptured. The result was a serious oil spill in the waters of Christiansted Harbor.
The problem was compounded by the loss of the Kings Hill water storage tank, approximately 3 km northeast of the airport. Because of the heavy demand for drinking water in preparation for Hugo, it was not possible to maintain a high water level at Kings Hill, thus making the steel tank highly vulnerable to wind damage. The result is shown in Figure 5-4. The almost complete disruption of water service forced residents of St. Croix to use home cisterns that had not been used for several years. As of March 1990, the Kings Hill tank had not been repaired, and a reinforced concrete tank was being considered as a replacement.
The Flooding of El Carraizo Pumping Plant
The other important failure of the disaster preparedness programs in Puerto Rico was the interruption of water services to the San Juan metropolitan area for 9 days as a result of the flooding of El Carraizo pumping plant, shown in Figure 5-5. The dam supplies San Juan's drinking water. During the emergency, some of the floodgates were inoperable, which caused water to spill over the dam, flooding the electric motors in the pumping plant and interrupting water services. The risk had been known for some time; a FEMA Interagency Hazard Mitigation Report had
identified the problem at El Carraizo Dam, as well as other potential flood sites throughout the island, as early as 1985. The CND investigative team confirmed FEMA's finding that the channels and flood-control structures near the Municipality of Catano were not adequately maintained. In the case of El Carraizo Dam, FEMA “uncovered many serious problems, including an earthen left bank which may be subject to erosion, an inadequate spillway, heavy siltation, rust deterioration of auxiliary floodgates, and a number of other serious maintenance/operations problems” (FEMA, 1985, p. 12).
At El Carraizo Dam, as with other dams in Puerto Rico, there is a significant potential for a truly catastrophic incident. Because the dam is only approximately 2 km upstream from the town of Trujillo Alto, tens of thousands of people could have perished if it had collapsed. Fortunately, the expected amount of rainfall did not materialize at that location. The dam's catchment area managed to withstand almost 2 inches of rain during Hugo. The situation at El Carraizo was similar to the tragedy on Las Americas Expressway near the Coamo River on the sourthern coast of Puerto Rico. On October 7, 1985, the Coamo River Dam overflowed, collapsing a span of the expressway and sending 29 persons to their death. In both cases, ample warnings about the unsafe conditions existed (FEMA, 1985, p. 35).
A San Juan Star special section on the hurricane on September 24, 1989, reported that the Autoridad de Aquaductos y Alcantarillados (AAA—Aquaduct and Sewer Authority) was the worst hit of the infrastructure and basic services on Puerto Rico. One in four people were still without water 1 week after the storm. Drinking water was dispensed from tank trucks, as shown in Figure 5-6. It is estimated that replacement of the five damaged motors alone at El Carraizo Dam will cost $200,000.
The San Juan Star reported on September 27, 1989, that water finally started to flow on September 25. As of September 26, four of the five motors at El Carraizo Dam in Trujillo Alto were pumping 90 million gallons of water per day into the Sergio Cuevas filtration plant. Full capacity for the plant is 110 million gallons per day. By this time, water service had been restored to 90 percent of the city.
A report in the September 28, 1989, San Juan Star noted that the U.S. Army Corps of Engineers had distributed more than 2 million gallons of water. They made 230 runs with 33 tank trucks. Costs were paid by the federal government. On September 24, 1989, the San Juan Star reported that one in four people (800,000) were still without water. At that time, 60 tank trucks carrying 400 to 10,000 gallons each were dispensing water drawn from La Plata filtration plant in Toa Alta, west of San Juan. On October 8, 1989, the San Juan Star reported that in Rio Grande 75 percent of the rural residences and 10 percent of the urban residences were still without water. Some water was obtained from an AAA plant in El Yunque, but none came from El Carraizo Dam.
The incident at El Carraizo Dam was nearly tragic. Luis Ruiz Javier, Executive Director of AAA, was quoted in the September 24, 1989, San Juan Star as saying: “I doubt there is anyone who is dissatisfied with the service that the Aqueduct and Sewer Authority has provided.” Ruiz Javier has since been removed as executive director, and a current investigation is looking into the entire situation. It is clear that the condition of the floodgates at El Carraizo Dam was known well before Hugo. An October 15 San Juan Star editorial noted that the government of Puerto Rico had known since April 1989 that the emergency generator was not working. Everyone had
been advised—the Governor's office, the Senate, the Electric and Power Authority, the Secretary of the Department of Natural Resources, the Mayor of Trujillo Alto, and El Carraizo Dam operators—and yet no action was taken. In addition, other potential sources of water failed because of deteriorated storage tanks, wells, and conduits.
Airports and aircraft in Hugo's path suffered considerable damage. FAA facilities at Alexander Hamilton Airport on St. Croix were heavily damaged, and most of the aircraft guidance equipment had to be replaced. The control tower, which sits on top of the terminal building, lost all of its cab windows, and much of the instrumentation and communications gear suffered water damage. Although the airport was open to light aircraft almost immediately after the passage of Hugo, it was 6 days before a temporary air traffic control tower was operational. Military navigation and communications gear was used in the interim. A photograph of the airport terminal building taken 5 days after the passage of Hugo is shown in Figure 5-7. Not shown in the picture are numerous private and commercial aircraft that were destroyed on the ground. As of March 1990, the terminal building and control tower were back to normal service.
At Cyril E. King Airport on St. Thomas, the damage to FAA facilities was much lighter, and limited air traffic control service was restored within 24 hours. The control tower shown in Figure 5-8 lost some cab windows, probably as a result of windblown gravel from a parking lot on a nearby hill, and there was wind and water damage to antenna structures, signal and power cables, and control tower instruments. As noted earlier in this report, the actual wind speeds at Cyril E. King Airport were far lower than first reported. Figure 5-9 shows roof damage to the airport fire and rescue station. Damage to the terminal building was superficial. Shortly after Hugo passed over northeastern Puerto Rico, the news media reported that two new terminal buildings at the San Juan International Airport had been destroyed. This report grossly overstated the damage, because the brief disruption of airport functions was mainly due to loss of power, water, and sanitation services. Most FAA facilities were operational immediately after Hugo. The most significant building damage at the airport was at the cargo building shown in Figure 5-10. The top of the picture is to the south, and the most damaging winds were out of the north and northeast. As was the case on St. Croix, heavy damage was suffered by both private and commercial aircraft at San Juan.
A related incident was the condition at the International Airport. The San Juan Star reported on September 27, 1989, that restrooms at the airport had been closed because of lack of water. Cleanup began as soon as water service had been restored. The airport was serving as a de facto shelter for the many people evacuating the
Caribbean islands while waiting for flights back to the United States. Given its condition and lack of sanitation facilities, the airport was inadequate to serve this function. It should have been set up as an official shelter, but apparently was not. People were everywhere, sleeping in the hallways and on the floor. There was no water and other beverages were sold out or were only available at inflated prices.
OTHER TRANSPORTATION SYSTEMS
Damage to highways was relatively light. Traffic lights, signs, and route markers were blown down and, in some areas, large trees and power poles had to be removed from the roadways. However, most roadways were passable within 48 hours.
The field team arrived in Puerto Rico just 3 days after Hugo. Many roads were still covered with debris, though main highways were open. It was reported in the San Juan Star on September 24, 1989, that damage to highways was estimated to be $40 million. Only one bridge—Route 187 at Boca de Cangrejos—was temporarily closed for repairs. Where Route 3, north of Humacao, runs right along the coast, the road was undermined slightly in places (Figure 5-11). Most of the road in this area is protected by a rock revetment; where the revetment does not exist, the road is in danger. It is the main road in the area, and could have hampered evacuation or emergency response had it been closed.
Route 187 from the International Airport east to Loiza is at very low elevations in places. The area near Pinones was subjected to the greatest amount of sand overwash observed after Hugo (see Figure 6-8 and Figure 6-9 in Chapter 6, Coastal Processes).
This was an obvious hindrance to evacuation during the storm and to accessibility afterward. Route 187 is the only road in this area and is backed by mangrove swamp for great distances. The commonwealth government favors developing this area more heavily, an action that could put many more people at risk.
The pier at Frederiksted, St. Croix, was also damaged. Wave action, probably sometime after 0800 GMT, removed several sections of the precast concrete deck. Recent repairs to this facility, which is shared by the U.S. Navy and cruise ships calling at St. Croix, had been completed in March 1990.
Ferry service between Puerto Rico and the islands of Culebra and Vieques was disrupted for several days. Typical of the damage caused by Hugo is the grounded ferry boat at Fajardo, shown in Figure 5-12. Two ferries were washed ashore, so there was only one trip per day from Fajardo to the islands. On September 27, 1989, the San Juan Star estimated over $50 million in losses to boats and another $25 million in damage to marinas.
OTHER LIFELINE SYSTEMS
Garbage collection was a significant problem that received little coverage. It was halted for almost a week, and residents were advised to take their own garbage to landfills. A San Juan Star article in the September 24, 1989, special section on Hurricane Hugo mentioned that the school system was especially hard hit for three reasons: (1) physical damage to the buildings, (2) no water or electricity, and (3) school rooms used as shelters. Of 166 shelters in Puerto Rico, 114 were in public
schools. Two-thirds of the public schools were still closed at the end of the week following Hugo, and 266,000 students would be staying home indefinitely.
The same section further reported that San Jorge and Teacher's hospitals had to stop admitting patients on September 20 because of lack of water. The hospital at Culebra suffered severe wind damage, as illustrated in Figure 5-13. However, government officials had taken preventive measures, storing critical medicines at three different secure locations. In the days following Hugo's impact there were reports of operational difficulties at a number of hospitals throughout the island, ranging from lack of blood and drinking water to electrical blackouts.
Figure 5-14 shows the uncovering of waste water pipes in the Isla Verde area (see also Figure 6-15 in Chapter 6, Coastal Processes). These pipes were buried in the sand in front of condominiums and are a potential hazard if the lines break and waste water contaminates fresh water.
River flooding is a common hazard associated with rainfall from hurricanes. Compounding the problem is the vigorous growth of water hyacinths in the freshwater rivers and streams of Puerto Rico. These water plants commonly get caught on bridge piers and pilings and increase upstream flooding, as well as causing structural damage to the bridges through blockage effects. After Hurricane David in 1979, the Dorado bridge over the Rio de la Plata was overwashed with water hyacinths and suffered a great deal of damage. Hugo did not cause severe flooding, so hyacinth buildup was not a severe problem this time.
In the Caribbean region, the electrical distribution system was the most seriously damaged lifeline during and after Hurricane Hugo. Others, such as communication systems and water pumping plants, suffered disruptions from the resulting lack of power. Most of these damages were caused by high winds or windblown debris. Many residents in the Virgin Islands and Puerto Rico were without electricity, water, or telephone service for periods ranging from a few days to a few weeks. Several hospitals suffered operational difficulties because of a lack of blood, water, or electricity.
Airports were considerably damaged by Hugo, although the extent of damage varied spatially. Highways and roads suffered little damage but were cluttered with windblown debris. Many traffic signals were damaged or destroyed in Puerto Rico; however, most roads were passable within 48 hours.